Principles of Sheep Dairying in North America (A3767)

Principles of sheep dairying in North America

Yves Berger Pierre BillonFrançois BocquierGerardo CajaAntonello CannasBrett McKusickPierre-Guy MarnetDavid Thomas

A3767

ContentsIntroduction

Metric conversion chart 2

Chapter 1Sheep’s milk and its uses 3

Chapter 2Choosing a breed 13

Chapter 3Improving lactation through genetics 21

Chapter 4Nutrition of the dairy ewe 35

Chapter 5Effect of nutrition on milk quality 51

Chapter 6Udder morphology and machine milking 63

Chapter 7Management for better milk yield and quality 81

Chapter 8Mastitis: detection and control 87

Chapter 9Milking parlors and equipment 93

Chapter 10General management of dairy ewes 109

Chapter 11Artificial rearing of lambs 125

Chapter 12Economics of raising dairy sheep 133

Appendix AResources 143

Appendix BFloor plans and barn designs 145

Principles of sheep dairying in North America

P R I N C I P L E S O F S H E E P D A I R Y I N G I N N O R T H A M E R I C A

AuthorsYves M. Berger, Researcher and

Superintendent, Spooner AgriculturalResearch Station, University ofWisconsin–Madison, Spooner,Wisconsin, [email protected]

Pierre Billon, Project manager,Milking system specialist, Institut del’Elevage, BP 67, 35652 Le Rheu, [email protected]

Dr. François Bocquier, Professor ofAnimal Sciences, ENSA-INRAMontpellier, Agro. M-INRA2 place Viala, 34060 Montpellier, [email protected]

Dr. Gerardo Caja, Professor of AnimalProduction, Department of Animaland Food Sciences, UniversitatAutònoma de Barcelona, Bellatera,[email protected]

Dr. Antonello Cannas, Associate profes-sor of Anima Sciences, DipartimentoScienze Zootechniche, Via De Nicola,07100 Sassari, [email protected]

Dr. Brett McKusick, DVM, MPVM, Ph.D.Development manager, Orion PharmaAnimal Health, P. O. Box 42520101 Turku, [email protected]

Dr. Pierre-Guy Marnet, ScientificDirector/Professor,Mixed unit ENSAR/INRA of researchon milk production65 rue de St. Brieuc35042 Rennes, [email protected]

Dr. David L. Thomas, Professor of sheepgenetics and management,Department of Animal Sciences,University of Wisconsin–Madison,Wisconsin, [email protected]

Sheep dairying and cheese produc-tion from sheep’s milk has a longtradition in many of the countries

of southern and eastern Europe, theMiddle East and North Africa. The topfive countries for sheep milk produc-tion in 1996 were Turkey (922 millionkg), Italy (700 million kg), Greece (630million kg), Sudan (510 million kg) andSyria (499 million kg).

The top five countries for sheep milkcheese production in 1996 wereGreece (113 million kg), Italy (81million kg), France (41 million kg),Spain (41 million kg), and Syria (40million kg) (FAO, 1998). Famous sheepmilk cheeses produced in some ofthese countries are: Manchego (Spain),Roquefort and Ossau-Iraty (France),Pecorino Romano and Sardo (Italy),and Feta (Greece).

Once very much the domain ofMediterranean and Middle Easterncountries, sheep dairying now extendsto many other parts of the world. TheUnited Kingdom, Australia, NewZealand, Argentina, Uruguay, Iceland,Holland, Sweden, Norway and theUnited States are all countries withouta sheep dairying heritage (or long for-gotten tradition), but that support abudding sheep dairy industry with thepotential to flourish.

What has sparked the renewedinterest in sheep dairying? There is nosingle answer to this question.Common problems faced by sheepproducers, a slow and subtle changein consumer preferences, the develop-ment of new technologies and com-munication are certainly importantcontributors.

Domestic market forsheep’s milk cheesesSheep’s milk cheeses are becomingvery popular with U.S. consumers.Imports into the U.S. increased from16.7 million kg for a value of 64 milliondollars in 1987 to 30.1 million kg for avalue of 139.7 million dollars in 1997(FAO, 1998)—an 80% increase in justten years. Of the 18 countries thatreport imports of sheep milk cheeses,the U.S. imports the most—approxi-mately 2.5 times the amountimported by the second rankedcountry (FAO, 1998).

In 1989, under the guidance of WilliamBoylan of the University of Minnesota,two or three Wisconsin sheep produc-ers and a small cheese plant createdthe embryo of the sheep dairyindustry in the U.S. In 1998 there wereroughly 100 dairy sheep producers inthe U.S., producing approximately500,000 kg of sheep milk annually(representing 100,000 kg of cheese).

1

Sheep dairying:Building on tradition

Introduction

Sheep’s milk

cheeses are

becoming very

popular with

U.S. consumers.

To all indications, there is a largedomestic market for sheep’s milkcheese that could support a domesticdairy sheep industry several timesthat of the current industry. Butalthough growth in producernumbers has been constant, it has notbeen dramatic because of limitationson production and marketing.Production limitations have includedlow milk production of domesticbreeds, lack of financing and lack of aknowledge base or pool of sheepdairy experience to serve as a sourceof information. Marketing limitationshave included lack of an organizedmarketing system for sheep’s milk andfew processors.

Many of these problems are beingaddressed. Dairy sheep breeds withhigh levels of milk production are nowavailable in the U.S. The University ofWisconsin–Madison has establishedthe only dairy sheep productionresearch unit in the country with thephysical facilities located at theSpooner Agricultural Research Station.Considerable work on breed evalua-tion, milking techniques, manage-ment, physiology of lactation and pro-cessing properties of sheep’s milk isbeing conducted by scientists in theUW–Madison College of Agriculturaland Life Sciences, Department of Meatand Animal Science, Food Science andthe Center for Dairy Research.

While large strides have been madetoward developing an economicallyviable dairy sheep industry in the U.S.through generation of researchresults, applying those research resultsto production, and developing a milkmarketing organization still needssubstantial work. This publication ispart of this continuous effort. By com-piling the most recent knowledge ondairy sheep production, it providesexisting and future dairy sheep pro-ducers (small or large) with a tool tohelp them increase their knowledgeand make good management deci-sions.

ReferencesFAO. 1998. FAOSTAT Database

Collectionsapps.fao.org/lim500/agri_db.pl

2


This book deals only with problems very specific to dairy sheep and isnot intended to be a sheep production or “how to” handbook. Readersare advised to check another source for information on general sheepmanagement, health (except for mastitis), pasture management,lambing management, reproduction and related topics.

Metric conversions Weight

Gram (g) x 0.04 = ounce Kilogram (kg) x 2.2 = pound Metric ton (t) x 1.1 = ton (Imp.)

VolumeLiter (L) x 0.035 = cubic foot Hectoliter (hl) x 22 = gallon(Imp.) Hectoliter (hl) x 2.5 = bushel

Length Millimeter (mm) x 0.04 = inch Centimeter (cm) x 0.39 = inch Meter (m) x 3.28 = foot Kilometer (km) x 0.62 = mile

Area Hectare (ha) x 2.5 = acre

For thousand of years, sheep andgoat’s milk have been the stapleof life in many areas of the world.

Sheep were certainly the first animalsto be domesticated by humans intheir effort towards an agriculturalway of life. Eventually, the conserva-tion of milk by its transformation intocheese was discovered and the bestcheeses of the world were developed.

Although cows have replaced sheepas dairy animals because of theirhigher production potential, sheepdairying remains a strong and viableenterprise. Nowadays, sheep dairyproducts (cheese, yogurt, ice cream)are particularly in demand because oftheir rich flavor and exceptional nutri-tive value. Anyone who has tasted thefamous Roquefort, the hard Pecorino

Romano, the softer Pecorino Sardo,the melting Manchego, the tenderOssau-Iraty or the salty Feta, wants tohave more than just a taste. Inaddition, for the master cheesemaker,sheep’s milk is a dream come truebecause of its composition and itscheese making properties.

Composition of sheep’s milkSheep’s milk contains almost twice thesolids of cow’s milk, as well as highercasein and fat content. Sheep’s milkyields 18–25% cheese; that is, it takesonly 4–5 kg of milk to produce 1 kg ofcheese (it takes 10 kg of cow’s milk toproduce the same amount). Moreover,the higher casein content makes therennet coagulation time for sheep’smilk shorter and the curd firmer(Jandal, 1996). The gross compositionof sheep’s milk and milk of otherspecies is shown in table 1.

3

Sheep’s milk and its uses

Chapter 1

For the master

cheesemaker,

sheep’s milk is a

dream come true.

Table 1. Composition of sheep’s milk compared to other species.

Human Cow Sheep Goat Yak

Dry matter (%) 11.5-13.9 10.5-14.3 17.4-18.9 11.9-14.0 16.8-19.6

Fat (%) 3.7-4.6 2.8-4.8 6.0-7.5 4.1-4.5 6.5-7.8

Alb. Glob. (%) 0.8-1.7 0.3-0.8 0.9-1.1 0.4-1.0 0.6-1.9

Casein (%) 0.4 2.5-3.6 4.3-4.6 2.5-3.3 5.0-5.8

Lactose (%) 6.4-7.0 4.2-5.0 4.3-4.8 4.1-4.4 4.6-5.3

Ash (%) 0.2 0.7-0.9 0.9 0.8 0.9

Calcium (mg/l) 1360 2030

Sodium (mg/l) 460 360

Vit. A (mg/l) 0.3 0.5

Vit.E (mg/l) 7 15.8

Vit. C (mg/l) 22 40.0

Kcal/100g 73 73 113 77 114

Alfa-Laval (1981)

FatJandal (1996) shows that the fat ofsheep’s milk forms globules with a sizeranging from .5 to 25 microns with anaverage diameter of 3.3 microns(Assenat, 1985). This is smaller thanfat globules of cow’smilk (4.55microns). Thecolor of fat insheep’s milk isvery whitebecause of atotal absence ofb-caroten (Assenat,1985). Twenty percent ofthe fatty acids of sheep and goat’smilk are short-chain saturated fattyacids (C4:0 to C12:0) compared to 12%in cow’s milk (table 2). Lipases attackthe ester linkages of the short-chainfatty acids more rapidly, so these dif-ferences may contribute to more rapiddigestion of goat and sheep’s milk.

Moreover, Havel (1997) reports thatshort chain fatty acids have little effecton the atherogenic lipoprotein con-centration in blood plasma (choles-terol) of humans. The amount of cho-lesterol in sheep’s milk increases with

the amount of fat and hasbeen found to be

between 150 and300 mg/liter(Assenat, 1985).Sheep’s milk has

a higher propor-tion of short-chain

fatty acids such ascaproic, caprylic and capric

than cow’s milk (but less than in goatmilk) and gives its special taste andaroma. Some other compounds suchas phospholipids and phenols alsohave an important role (Kim Ha andLindsay, 1991).

ProteinsQuantitatively and qualitatively,proteins constitute the most impor-tant percentage of milk. Two types ofproteins can be found in milk: thecasein and the serum, or wheyproteins. In sheep’s milk, casein entersfor 80% of the total proteins. Becauseof a higher casein content, sheep’smilk has better coagulation propertiesand better cheese making potentialthan cow’s milk.

Caseins are made of 5 major compo-nents: ασ1, ασ2, β, κ and γ, and thepercentage of the 5 componentsvaries from one species to the other asshown in table 4 (Assenat, 1985). Thepercentage of ασ1 and ασ2 is higherin sheep than in goat’s milk but signif-icantly lower than in cow’s milk. Caseinb, however, represents 50% of the totalcasein in sheep’s milk compared to 2⁄3

4


Smaller fatglobule diameter and

greater percentage of shortchain fatty acids contribute toeasier and more rapid diges-

tion of sheep’s milk.

Table 4. Casein fractions according to the totalcasein.

Caseinfractions Cow Goat Sheep

ασ1 36 12.6 15.5

ασ2 9.5 14.7

β1 33 35.9 18.9

β2 39.4 28.2

κ 9.4 8.1 7.3

γ 6.8 3.9 15.4

Assenat (1985)

Table 3. Physical properties of milk

Properties cow sheep

Specific gravity 1.0231-1.0398 1.0347-1.0384

Viscosity, Cp 2.0 2.86-3.93

Surface tension (Dynes/cm) 42.3-52.10 44.94-48.70

Refractive index (nD20) 1.3344-1.3485 1.3492-1.3497

Conductivity (ohm-1cm-1) .0040-.0055 .0038

Freezing point -0.530 to -0.570˚C -0.570˚C

PH 6.65-6.71 6.51-6.85

Acidity (Lactic acid %) 0.15-0.18 0.22-0.25

Anifantakis (1985)

Table 2. Fatty acid composition of goat, cow andsheep’s milk (% per weight)

Fatty acid goat cow sheep

C4:0 (Butyric) 2.6 3.3 4.0

C6:0 (Caproic) 2.9 1.6 2.6

C8:0 (Caprylic) 2.7 1.3 2.5

C10:0 (Capric) 8.4 3.0 7.5

C12:0 (Lauric) 3.3 3.1 3.7

C14:0 (Myristic) 10.3 9.5 11.9

C16:0 (Palmitic) 24.6 26.5 25.2

C16:1 (palmitoleic) 2.2 2.3 2.2

C18:0 (Stearic) 12.5 14.6 12.6

C18:1 (Oleic) 28.5 29.8 20.0

C18:2 (Linoleic) 2.2 2.5 2.1

Jandal (1996)

in goat and 1⁄3 in cow’s milk. Thosevariations in percentages of caseinexplain the difference in micelle struc-ture and the absence of bitter taste insheep’s milk cheeses.

Most of the sheep’s milk produced inthe world (with the exception of theUnited Kingdom) is transformed intocheese, and is rarely consumeddirectly. For this reason, Bencini andPulina (1997) refer to the quality ofsheep’s milk as its capacity to be trans-formed into high quality products andto produce high yields of theseproducts (referred to as the process-ing performance of the milk). The pro-cessing performance of the milk (yield,composition and taste) mainlydepends on the milk’s clotting proper-ties, a combination of renneting time,rate of curd formation and consis-tency of the curd. The clotting proper-ties of the milk are widely affected byits composition.

Factors affecting the composition ofsheep’s milkBencini and Pulina (1997) cite severalfactors that affect the composition ofthe milk.

Somatic cell countIn sheep’s milk, only 10% of thesomatic cells are mammary gland cells(eosinophils, epithelial cells), normallysecreted together with the milk as aresult of cellular turnover in themammary gland. The remaining 90%of the somatic cells areblood cells(macrophages,leucocytes,lympho-cytes). Thesenormallycontribute tothe immunedefense of themammary gland, buttheir number increases consid-erably in the case of inflammatory or

pathological processes within themammary gland. Therefore, a highsomatic cell count is a sign of generalinfection in the animal. The mostcommon pathology of the mammarygland in sheep is mastitis (seechapter 8).

A high somatic cell count results inchanges in the composition of milk(table 5) with a higher pH, a reductionin fat, casein, total solids, solublecalcium, and an increase in totalnitrogen, non-protein nitrogen andwhey proteins (Pirisi et al., 2000). These

changes in composition lead to aconsiderable slow-

down of coagula-tion and

serious diffi-culties inthe struc-turing of

the curd andconsequently

a lengthening ofthe cheese making

process as well as a decreasein cheese yield (Pirisi, 2000). High

somatic cell count in cow’s milkhas been associated withproblems in the quality ofcheese. However, Pirisi et al.(1993, 2000) report that, in astudy conducted in sheep’s milk,although cheese yield wasreduced, ripened cheeses didnot show significant differencesfor chemical parameters andsensorial characteristics.Wendorff (2000), however, founda greater tendency for cheese todevelop a rancid flavor.

Some cheese makers might setan artificial limit of somatic cellsover which milk would be dis-counted. This is a very seriousincentive for producers to makeall possible efforts to decreasetheir overall (bulk tank) somaticcell count.

5

S H E E P ’ S M I L K A N D I T S U S E S C H A P T E R 1

In the U. S. states where

sheep’s milk is recognized as milk,

it cannot legally have a somatic cell

count higher than 1,000,000 cells/ml. A

higher level at two consecutive tests

could result in the revocation of the

milk producer’s license.

Table 5. Composition of sheep milk with different SCC

SCC<500 500<SCC>1000 1000<SCC>2000x 1000/ml

SCC x 1000/ml 229 ± 55 653 ± 250 1200 ± 214

Ph 6.52 6.62 6.68

Dry Matter g/100g 17.03 17.15 16.89

Lactose g/100g 4.74 4.54 4.38

Fat g/100g 6.61 6.34 6.36

True Protein g/100g 5.25 5.45 5.51

Casein g/100g 4.18 4.26 4.20

Soluble Casein % 6.51 6.98 7.77

Whey Protein g/100g 1.07 1.19 1.30

Non-Protein N gN/100g .06 .05 .05

Casein:Protein % 79.71 78.16 76.27

Urea mg/100ml 54.21 54.18 52.86

Total Ca g/l 2.21 2.14 2.26

Soluble Ca g/l .46 .38 .36

Pirisi et al. (2000)

Microbial countMany microorganisms present in themilk are advantageous for its transfor-mation into cheese (Lactobacillus spp.,Lactococcus ssp., Streptococcus spp.).However, others can cause serioushuman diseases (Salmonella, Listeria,Brucella), or create problems in thematuration of cheese(Enteriobactericeae, coliforms, psy-chrotrophs). Some psychrotrophicbacteria thrive at temperature below7˚C and produce enzymes that desta-bilize the casein and modify theclotting properties of the milk.

Breed of ewesThe breed of sheep can affect thecomposition of milk, mostly becausethere is a negative correlationbetween milk yield and milk compo-nents. Therefore, breeds highlyselected for dairy production tend tohave a lower concentration of fat,protein and total solids. As a conse-quence, with high milk production,the total amount of cheese producedfrom the milk will be higher but therelative yield of cheese from each literof milk will be lower.

Age and parityAlthough reports are somewhat con-tradictory, it seems that the milk ofyoung ewes contains a lower concen-tration of fat, proteins and total solids.The concentration of total solidsincreases with the parity number.

Stage of lactationThe amount of fat, protein, total solids,and somatic cells is high at the begin-ning and at the end of lactation andlow at the peak of lactation. The pro-cessing performance of the milk tendsto decrease as the lactation proceeds,with an increase in renneting time andrate of curd formation and a decreasein the consistency of the curd(Ubertalle, 1989, 1990 cited by Bencini

and Pulina, 1997).

Season of milkingMany researchers cited by Benciniand Pulina (1997) have shown

that sheep milk produced insummer has poor cheese making

performance due to long rennetingtime, poor consistency of thecurd, and high proteolytic andlipolytic activities. Mendia et al.,(2000) also found that Idiazabalcheeses made in Februaryearned higher sensory analysisscores for characteristic odor andtaste and higher sensory scores thancheese made in June. It seems thathot temperatures do not affect thecomposition of the milk as much asthe length of days. Long days result ina lower protein concentration andreduced secretion of fat and protein.

Nutrition Nutrition affects the total milk produc-tion as well as the quality of the milk.The concentration of fat in the milk iscorrelated with the concentration offiber in the diet (see chapter 5).

Conservation of milkFresh milkTo avoid the multiplication of bacteria,the milk must be cooled to 1̊ C–4˚Cand maintained at this temperaturefrom milking to delivery to the pro-cessing plant. The low temperaturewill prevent bacterial multiplicationfor 24–48 hours. However, thepresence of psychrotrophic bacteria,that is, those able to multiply, albeitslowly, at 5˚C, can result in markedincreases in bacterial count duringlonger storage periods. Cousins andMcKinnon (1977) have shown a signifi-cant increase in bacteria numbersafter 3 days at 5˚C and that the effectsof relatively small increases in storagetemperature were dramatic.

Ineffectively cleaned and disinfectedmilking equipment, particularly thebulk tank, is the major source of psy-chrotrophic bacteria in milk.

6


In the U.S. (at least in

Wisconsin) the legal limit for

bacteria count in the milk is less than

300,000 cells/ml for grade B milk and less

than 100,000 cells/ml for grade A milk. A

higher level at 2 consecutive tests could

result in the revocation of the milk

producer’s license. It is recommended that

fresh milk be chilled to 1˚C–4˚C

in the 2 hours following milking

and that it not be stored longer than

72 hours. Extreme care should be

taken in cleaning of the

equipment.

210 gallon bulk tank

Frozen milkIn North America, sheep dairy produc-ers are few, separated by great dis-tances and producingmodest amounts ofmilk daily.Therefore, thecollection offresh milk ona daily basisby a process-ing plant iseconomically dif-ficult for the timebeing. American produc-ers generally freeze the milk until asufficient quantity is stored fordelivery to a cheese maker. Moreover,the production of sheep’s milk is veryseasonal and fresh milk becomesunavailable from October to February.

The freezing of milk helps remedy theshortcoming of production. Bastian(1994) at the University of Minnesota,studied the effect of freezing on thequality of the milk for cheese making.He found that freezing and thawing ofsheep’s milk did not change rennetcoagulation properties, compared tofresh, unfrozen sheep milk. Wendorffand Rauschenberger, (2001) at theUniversity of Wisconsin–Madisondetermined the storage stability ofmilk frozen at 2 different temperatures(table 6). Samples of raw sheep milkwere frozen at -12˚C and at -27˚C.Samples were thawed at 1, 2, 3, 6, 9,and 12 months and analyzed for totalbacteria, coliform bacteria, acid degreevalue (ADV), and intact protein. Intactprotein was defined as the totalprotein content of milk minus theprotein present in sediment at thebottom of the recipient in which themilk was frozen. Results indicate thatmilk frozen in a standard home freezerat -12˚C was not as stable as milkfrozen in a commercial hardeningroom at -27˚C. After 6 months ofstorage at -12˚C, about one third of

the casein was destabilized and pre-cipitated out upon thawing. The rawmilk stored at the lower temperature

was stable up to 12 months.No evident protein

precipitation wasnoticed

throughoutthe study.

The bestfreezing is

obtainedwhen the milk is

placed in plasticbags approved for food

use with a cap (75 cm x 40 cm) andholding approximately 18–20 liters ofmilk. The bags are placed flat onshelves inside the freezer so that thethickness of the bags is no more than6–7 cm, allowing for a quick anduniform freezing. As soon as the milkis solid, the bags can be stacked easilyin the freezer.

7


It is recommended that

milk be frozen as quickly as

possible, and at a temperature of at

least -25˚C. In general, a home freezer

cannot provide this type of freezing. A

commercial grade freezer with

inside ventilation should

be used.

Sheep milk in FDA approved bags andplaced on shelves for quick freezing.

Sheep milk stored in a commercialfreezer at-25˚C.

Table 6. Properties of frozen raw milk stored at -12˚C and -27˚C forvarious time periods

Time of Coliforms SPC TCA-ppt.storage (Mo) (CFU/ml) (CFU/ml) ADV Protein

Stored at –12˚C

0 44 8200 .220 5.09

1 26 4100 .250 5.06

2 21 2500 .420 5.01

3 12 3400 .350 5.06

6 <1 2200 .410 5.02

9 <1 340 .420 3.37

12 <1 610 .490 3.90

Stored at –27˚C

0 44 8200 .220 5.09

1 10 4100 .260 5.05

2 9 3200 .320 4.99

3 12 3700 .290 4.96

6 8 2800 .310 4.89

9 8 2700 .280 4.92

12 5 1800 .350 5.07

Wendorff and Rauschenberger, (2000)

The type and size of freezing contain-ers, however, depends on the cheesemakers and on the thawing capabili-ties. Before deciding on a freezingmethod the producer or group of pro-ducers should discuss with the cheesemaker the best suitable method.

Uses of sheep’s milk Fluid productsSheep’s milk has some unique nutri-tional qualities that could be used inspecific markets. It is richer in vitaminsA, B and E, calcium, phosphorous,potassium and magnesium than cow’smilk. Sheep’s milk contains 1.08-1.44%whey proteins while cow’s milkcontains only 0.54-0.88%. It is alsoricher in C4-C12 fatty acids. Sheep’smilk provides some relief for allergysufferers who cannot tolerate cow’smilk proteins.

In spite of the added nutritional quali-ties, only small quantities of sheep’smilk are consumed as fluid milk. InSpain, for example, only 7.2% isconsumed this way. This is certainlydue to the fact that most of thesheep’s milk is produced in countrieswhere consumption of fluid milk hasalways been traditionally low.Moreover, with the high solids contentof sheep’s milk, it is more readilyaccepted for manufacturing of semi-solid or hard dairy products such asyogurt or cheese. In Great Britain,however, a large portion of the sheep’smilk produced is consumed as fluidmilk. The milk is pasteurized andstored in 1⁄2-liter carton containers(pint). The milk is then generally frozenand sold to health food stores.

Dried productsIn spite of the high solids content ofsheep’s milk, there is report of its usein the area of dried or non-fat milkproducts. There appears to be a signif-icant demand, at the current time, fordried sheep’s milk to blend with cow’smilk for specialty cheese production.Some concerns about these productsrevolve around the stability of themilk fat and shelf life.

YogurtsThe solids content of sheep’s milkmake it a natural for production ofpremium yogurt products similar tothe Greek-style yogurts. With solidscontent of 16–18% in the milk, yogurtscan be produced without the need ofadded solids or stabilizers. With thehigher fat in the sheep yogurt, thepotential harshness of the lactic acidmay be lessened. Sheep yogurt alsoshows a greater cold storage stabilityas shown in table 7. Sheep yogurtsneeds to be marketed as a premiumspecialty product to avoid competingwith commodity yogurts producedfrom cow’s milk. Low fat yogurts canalso be made after separation of thecream; the cream can be used for themanufacturing of butter.

8


Yogurt produced from

milk frozen and stored at -

27˚C was comparable to that

produced from fresh milk

(table 8).

Freshly made yogurts

Table 8. Characteristics of yogurts produced from ovine milk frozenand stored at -12˚C and -27˚C for 12 months.

Initial Stored StoredCharacterisics milk at -12˚C at -2˚C

Titratable acididty, % 1.25 0.90 1.18

Syneresis, % 75.1 79.7 77.5

Water holding capacity, % 28.5 25.7 30.4

Firmness, g 125 72 109

Wendorff and Rauschenberger (2001)

Table 7. Cold storage stability ofyogurts

Separated serum (ml)Type 1 day 10 days

Cow 41 40

Sheep 6 7

Goat 23 15

Kehagias et al (1986)

CheeseTraditionally, cheese production is thegreatest market for sheep’s milkthroughout the world. With its highsolids and smaller fat globules, sheep’smilk is an outstanding substrate formanufacturing high quality cheese.Normally, 15% solids in milk are aboutthe most efficient for obtainingmaximum output per vat per day,while allowing for sufficient syneresis(exudation of liquid) of the curd forproper moisture control in the finalcheese. Typical cheese yields for cowand goat milk are 9–10% while sheep’smilk yields approximately 18–25%according to the type of cheese.

Cheese yields can be estimatedaccording to the composition of themilk in fat and protein.

For the prediction of blue-veinedcheese yield (Roquefort) Pelligrini(1995) uses the following equation:

Blue-veined cheese yield = 0.05 Fat(g/Kg) + 0.32 Protein (g/kg)+ 1.81

For the prediction of hard or softcheese yields Jaeggi et al. (2004) haveshown that the Van Slyke formula canbe reliably used:

Where:RF = .84 for hard cheese and .82 forsoft cheese

RS = 1.08 for hard cheese and 1.01 forsoft cheese

RC = .96 for hard cheese and .96 forsoft cheese

A method of milk payment couldtherefore be based on either formulaaccording to the type of cheese madeby the manufacturer. A method ofpayment based on cheese yieldencourage production of milk withhigh fat and protein content.

9


Famous sheep’s milkcheeses produced in the worldWhite fresh cheeses

Burgos (Spain)

Villalon (Spain)

Cachat (France)

Perail (France)

Brined cheeses

Feta (Greece, Italy, France)

Teleme (Romania)

Sirene (Bulgaria)

Halloumi (Cyprus)

Hard and semi-hardcheeses

Pecorino Romano, Sardo,Siciliano, Toscano (Italy)

Kefalotyri (Greece)

Idiezabal (Spain)

Manchego (Spain)

Roncal (Spain)

Ossau-Iraty (France)

Blue-veined cheeses

Roquefort (France)

Cabrales (Spain)

Stretched curd cheeses

Kashkaval(Bulgaria/Romania/Macedonia)

Kaseri (Greece)

Whey cheeses

Ricotta (Italy)

Manouri (Greece)

Hard or soft cheese yield = [(RF x % Fat in milk) + (RC x % Casein in milk)] x RS]

(100 - % Moisture of cheese)

Composition of wheyOvine whey has a higher fat, proteinand lactose content than caprine orbovine whey (tables 9 and 10). This isexpected because sheep’s milkcontains higher fat, protein and ashthan the other two species. Ovinewhey contains more b-lactoglobulin(b-LG) and less serum albumin (SA)and immunoglobulins (IgG) as a per-centage of total whey protein thanbovine whey (table 9). There is asmuch a-Lactalbumin (a-LA) in ovinewhey as in bovine whey but signifi-cantly less than in caprine whey. Wheyprotein composition varies during lac-tation. The proportion of a-LAdecreases throughout lactation whileSA proportion increases. Also, b-LGgrows significantly during mid lacta-tion and then falls back to concentra-tions that are similar to those in earlylactation. Ovine whey protein concen-trate (WPC) has a better foam overrun,foam stability and gel strength thanbovine or caprine WPCs (Casper et al.,1999).

ConclusionSheep’s milk is a unique product withhigh nutritional qualities containingmore short chain fatty acids, moreprotein, more calcium and morevitamins than cow’s milk. It is recog-nized by many as non-allergenic,which could open the door to a largerfluid milk market. It is an outstandingproduct for the manufacture ofyogurts and cheeses, giving a cheeseyield twice as high as cow’s milk. Thecomposition of the whey could allowfor the manufacture of a vast array ofproducts, improving the financialreturn of the cheese maker.

10


Table 9. Gross composition of caprine and ovine wheys from themanufacture of specialty cheeses and bovine whey from the manu-facture of Cheddar cheese.

________ Goat ________ OVINE BovineComponents Cheddar Chèvre MANCHEGO CheddarPH 6.2 4.6 6.3 ND1

__________________ (% wt/wt) __________________

Total solids 6.61 6.40 7.46 6.7

Water 93.39 93.60 92.54 93.30

Fat 0.51 0.03 0.82 0.36

Ash 0.61 0.76 0.43 0.52

Lactose 4.71 5.07 5.16 4.50

Total protein 0.77 0.53 1.05 0.601Not determined Casper et al., 1998

Table 10. Distribution of whey proteins in specialty cheese wheyfrom caprine and ovine milk compared with that in Cheddar wheyfrom bovine milk.

Whey protein ———Goat——— OVINE Bovinefraction Cheddar Chèvre MANCHEGO Cheddar

__________________ (% of total protein __________________

Serum albumin 4.0 3.8 4.1 6.5

IgG 9.7 6.4 7.3 13.0

β-LG 58.6 59.2 74.0 64.9

α-LA 27.0 31.7 14.8 15.6

Casper et al., 1999

ReferencesAlfa-Laval, 1981. System Solutions for

Dairy Sheep. Alfa-Laval AgriInternational AB, Tumba, Sweden.

Alais C., 1984. Science du lait. Principesdes techniques laitières. 4thedition. SEPAIC, Paris.

Anifantakis E.M., 1985. Comparison ofthe physico-chemical properties ofewe’s and cow’s milk. Bulletin ofthe International Dairy FederationNo 202/1986.

Assenat L., 1985. Le lait de brebis: com-position et propriétés. In: Laits etProduits Laitiers: Vache, Brebis,Chèvre”. F.M. Luquet (Edt).Technique et DocumentationLavoisier, Paris, France.

Bastian E.D., 1994. Sheep milk coagula-tion: Influence of freezing andthawing. Cultured Dairy Products J.29 (4): 18-21

Bencini R. and G. Pulina, (1997). TheQuality of Sheep Milk: a review.Australian Journal of ExperimentalAgriculture, (1997), 37, 485-504

Casper J.L., W.L. Wendorff and D.L.Thomas. 1998. Seasonal changes inprotein composition of whey fromcommercial manufacture ofcaprine and ovine specialtycheeses. J. Dairy Sci. 81: 3117-3122

Casper J.L., W.L. Wendorff and D.L.Thomas. 1999. Functional proper-ties of whey protein concentratesfrom caprine and ovine specialtycheese wheys. J. Dairy Sci. 82: 265-271.

Cousins C.M. and C.H. McKinnon. 1977.Cleaning and disinfections in milkproduction. In: Machine Milking.Thiel C.C. and F.H. Dodd (Edts).National Institute for Research inDairying, Reading, England.

Havel R.J.. 1997. Milk fat consumptionand human health: NIH and otherAmerican governmental regula-tions. In: Milk composition,Production and Biotechnology.R.A.S. Welch, D.J.W. Burns, S.R. Davis,A.I. Popay and C.G. Prosser (Edts).CAB International, NY, USA

Jaeggi J.J., W.L. Wendorff, M.E. Johnson,J. Romero and Y.M. Berger.2004.Milk composition and cheese yieldfrom hard and soft cheese manu-factured from sheep milk. 10thGreat Lakes Dairy SheepSymposium. Proceedings. Nov 4-6,2004. Hudson, Wisconsin.JandalJ.M., 1996. Comparative aspects ofgoat and sheep milk. Small Rum.Research (1996), 22, 177-185.

Kim Ha J. and R.C. Lindsay, 1991.Contributions of cow, sheep andgoat milks to characterizingbranched-chain fatty acid and phe-nolitic flavors in varietal cheeses.J. Dairy Sci, 74:3267-3274

Mendia C., F.C. Ibanez, P.Torre and Y.Barcina. 2000. Influence of theseason on proteolysis and sensorycharacteristics of Idiazabal cheese.J. Dairy Sci. 83: 1899-1904

Noble R.C., W. Steele and J.H. Moore,1970. The composition of ewe’smilk fat during early and late lacta-tion. J. Dairy Res. (1970), 37, 297

Pellegrini O., 1995. Etude des relationsentre les caractéristiques physico-chimiques du lait de brebis et sesaptitudes fromageres. These deDocteur d’Etat. Institut NationalAgronomique, Grignon, France.

Pirisi A., G. Pirelda, F. Podda and S.Pintus, 1993. Effect of somatic cellcount on sheep milk compositionand cheesemaking properties.Somatic cells and milk of smallruminants. Proceedings. Bella, Italy,25-27 September 1993. EAAPPublications No.77: 245-251

Pirisi A., G. Pirreda, M. Corona, M. Pes,S. Pintus, A. Ledda. 2000Influence of somatic cell count onewe’s milk composition, cheeseyield and cheese quality. 6th GreatLakes Dairy Sheep Symposium.Proceedings. Nov. 2-4, 2000.Guelph. Ontario, Canada

Wendorff W.L., 1998. Updates onsheep milk research. 4th GreatLakes Dairy Sheep Symposium.Proceedings. June 26–27, 1998.Spooner, Wisconsin.

Wendorff W.L. and S.L.Rauschenberger. 2001. Effect offreezing on milk quality. 7th GreatLakes Dairy Sheep Symposium.Proceedings, Nov. 1-3, 2001. EauClaire, Wisconsin.

11


12


. Two questions a potential sheepdairy producer should ask are: 1)What breeds can be milked? and

2) Can domestic American breeds bemilked? The answer is that any breedof ewe can be milked—with varyinglevels of success.

The mechanism of milk ejection in aewe suckled by her lamb and a ewemilked by machine is quite different(see chapter 7).Therefore, a ewe thatapparently has enough milk to success-fully raise 2 or 3 lambs might dry upvery quickly as soon as her lambs areweaned and machine milking starts.The growth of her lamb during the first30 days of lactation is not necessarily asign of good milking ability when itcomes to commercial milk production.A ewe’s commercial milking abilityshould be determined during themachine-milking period only. Daily pro-

duction, total number of days beingmilked by machine, percentage of fatand percentage of protein are the mainfactors to consider.

Domestic breedsWhat breeds are available in the U.S.or might become available in the nearfuture?

Beginning in 1984, several U.S. sheepbreeds were evaluated at theUniversity of Minnesota for their milkproduction potential. Ewes werechosen from available breeds andmachine-milked twice a day followingweaning of their lambs at 30 days ofage. Ewes were subsequently milkedfor 120 days. Table 1 shows the per-formance of these breeds for milk pro-duction and milk composition over atwo-year period. With the exception ofthe Finn and Romanov breeds, allother domestic breeds studied seemto have an identical potential for com-mercial milk production.

The average daily milk production ofall ewes is .47 kg per day. With suchlow production one might wonder if itis economical to raise these breeds formilking. Early pioneers of sheepdairying in the U.S. were using popularbreeds such as Polypay and Dorset intheir first attempts at milking. Thesepioneers were soon looking for waysto improve average production. Buteven given low production, thesebreeds should not be disregarded alltogether. They offer certain advan-tages that the dairy-type breedsmight not possess, such as good adap-

13

Choosing a breed

Chapter 2

Table 1. Least-square means for several milk traits by breed (1989–1990).Milking period only.

Milk Fat Protein Lactose Solids Breed (liters) (%) (%) (%) (%)

Overall mean 57 6.6 5.8 4.7 17.9

Suffolk 69 6.7 5.9 4.7 18.1

Finnsheep 44 6.1 5.5 4.5 16.7

Targhee 62 6.9 5.9 4.8 18.4

Dorset 61 6.3 5.7 4.5 17.2

Lincoln 53 6.8 5.8 4.7 18.0

Rambouillet 65 6.6 6.1 4.9 18.3

Romanov 44 7.1 5.9 4.8 18.6

Outaouais 54 7.3 6.1 4.6 18.7

Rideau 77 6.6 5.8 4.8 18.0

W.J.Boylan (1995)

Any breed of ewe

can be milked

with varying levels

of success.

tation to a wide array of environ-ments, good lamb and/or wool pro-duction, ability to breed out of season(Polypay, Dorset), predictable behaviorand wide availability at a reasonableprice. If they are selected carefully,these breeds’ overall production couldincrease dramatically.

Jordan and Boylan (1988) suggest thatby selection and screening of the bestmilking ewes, overall milk productioncould increase by 30–40% in just a fewyears. Selection is a powerful toolsince there is always a large variation(CV = 30%) between individualanimals. In France, in 1969, the 800,000Lacaune ewes on which 65% of theFrench sheep dairy industry is basedwere producing only 80 liters per lac-tation. After progeny testing morethan 8500 rams since 1970, theLacaune ranked as one of the bestmilking breeds with more than 250liters per lacation by 1998.

Lower milk production may actuallybe quite acceptable in some segmentsof the industry. It is well known that astrong negative correlation existsbetween the total amount of milk andthe percentage of fat. Generally thehigher the production the less fat inthe milk. To produce a very highquality sheep’s milk cheese, the milkneeds to be high in butterfat (6–8%).Domestic breeds producing amoderate amount of milk do have ahigher butterfat content.

If the dollar value of the milk dependson its quality (the sum of fat andprotein being the total useful drymatter) the discrepancy between apure dairy breed such as the EastFriesian and a domestic breed is muchreduced in terms of overall return perewe. To remedy this problem, Frenchscientists included fat and protein intheir selection index of the Lacaunebreed as early as 1985, making thebreed a high milk producer with highfat and protein content.

The East Friesian(Ostfriesisches Milchschaf )This breed is now readily available inthe United States. Many entrepreneursin Canada or the U.S. sell live animals,embryos or semen of different origins(mainly from England, Holland andSweden through New Zealand).

It might be of interest to report hereon the breed’s controversial begin-nings as recorded by Flamant andRicordeau (1969) in their excellent lit-erature review on the East Friesian.

The breed has its origin in Germanyalong the North Sea coast. Someauthors think that the East Friesian isthe result of a cross between severalDutch breeds and a breed importedfrom the Gulf of Guinea in the early17th century. The progenies of thiscrossbreeding formed the nucleus ofthis new breed, which was fixed fairlyquickly since, as early as 1750, it wasexported toward Lithuania. The traitsthen were already the same as thosethat make the breed famous today:prolificacy, milk production and woolproduction. Other authors, however,observed the very distinct similaritiesbetween the East Friesian and otherbreeds traditionally milked for familyuse before the expansion of dairycows such as the Dutch Friesian, theBelgian Flammish and the Flander in

14


France. According to the same authorsthe East Friesian would be the lastrepresentative of this type of animal.

Whatever its origin, the East Friesian isconsidered by many to be one of thebest milking sheep in the world.Average production of 450–500 kg perlactation of 220–240 days and morehas been recorded. However, asFlamant and Ricordeau (1969) state,the total production includes thequantity of milk produced beforethe weaning of the lambs. Thisquantity is often estimated by mul-tiplying the quantity of milkobtained at the first testing by thenumber of days between the firsttesting and parturition (see chapter3). The production capability istherefore overestimated since thefirst testing often corresponds tothe peak of lactation.

Table 2 shows the milk production ofEast Friesian ewes in Germany. All datapresented in the table are old(between 1920 and 1950) but areapparently the only reliable informa-tion available for that period. The lackof reliable and new data on the milkproduction of the East Friesian breedis understandable given the absenceof a selection scheme and meaningfulmilk recording system since this time.Therefore, a certain degreeof heterogeneity has to beexpected in the breed.

15

C H O O S I N G A B R E E D C H A P T E R 2

Table 2. Milk production of East Friesian ewes in Germany

Number Total milk Lactation Fat Total fat Authors of ewes production (kg) Length (days) (g/kg) production (kg)

Spottel (1954)Milk recording 1929 59 639 63 40.6

Muhlberg (1934)Milk recording 1933 97 708 263 62 43.6

Leonhard (1954)Milk recording 1936 470 529

Ulrich (1953)Milk recording 1938 829 489 249 61

Spottel (1954)Milk recording 1942 544 471.5 64

Ebbinghaus (1949)450–550 200 60–90 27–40

Buitekamp (1952)Oriental Friesland 1951 287 556 245 62 34.6

Brauns (1953)Thuringe 1951 70 374 255 60 22.4

Schirwitz (1953)Saxonie 1953 507 393 64 25-27

Ver. Rhein. Schaf.Rhenanie 1958 582 64 37

Flamant and Ricordeau (1969)

Table 3 shows production data of afew East Friesian ewes in England. Asone can see there might not havebeen much progress in milk produc-tion after the 19th century.

Size and conformationThe East Friesian has a large frame.Adult females weigh between 70 and90 kg and adult rams weigh up to 120kg. The breed’s legs are long and thinand it has narrow hips. Its head, legsand tail are devoid of wool and shouldbe white with a pink, thin skin. Thelack of wool on its tail gave the EastFriesian the nickname “rat tail.” Allanimals are generally polled; somescurs can be found.

The udder is generally large, present-ing an important cistern capacity.There is a wide variation in udder mor-phology among individuals of thebreed. Often teats are implanted highand on the side making the completeemptying of the udder by themachine difficult. Milkers often haveto lift the bottom of the udder abovethe level of the teats either by hand orwith a “Sagi hook.”This results inslower milking that might be detri-mental to the farm operation.

WoolThe fleece is white with a long staple(10–15 cm) and medium quality(52/54 on the Bradford’s scale). Fleeceweighs between 4–6 kg. Handspinners generally appreciate thistype of wool. Some colored (black orbrown) East Friesian sheep can also befound.

ReproductionA prolificacy rate of 230% has beenreported, making this breed one of themost prolific. At the SpoonerAgricultural Station (University ofWisconsin), Berger (1998) reports pro-lificacy of 200% on 12-month-old and230% on adult crossbred ewes.Withtheir high milk production and highprolificacy, the East Friesian breed is anefficient lamb producer. Although it hasa rather poor carcass conformation,lambs produced from crossbreedingwith a terminal breed such as Suffolk,Hampshire or Texel have a remarkablegrowth with all desirable carcass traits.

16


Table 3. Milk production of 15 East Friesian ewes in England

Suckling and Age of ewes milking period Total kg of milk Average fat Protein (in years) (in days) for each period percentage percentage

9 51+144=195 224+228=452 6.71 5.36

8 59+167=226 201+443=664 6.93 5.48

3 53+175=228 313+438=751 5.85 5.19

1 52+175=227 68+385=453 5.72 5.28

3 61+172=233 140+683=823 5.61 4.84

3 58+172=230 290+480=770 6.22 5.23

4 41+154=195 86+319=405 5.08 5.01

3 51+168=219 275+294=569 5.22 5.04

3 53+168=221 127+277=404 5.88 5.39

1 48+154=202 86+168=254 6.22 5.41

4 63+190=253 334+581=915 6.27 5.46

2 55+190=245 220+450=670 5.92 5.57

1 48+190=238 77+281=381 6.61 5.77

5 37+190=227 100+281=N381 6.32 5.21

2 52+187=239 192+292=484 6.41 5.36

Olivia Mills (1989)

Although the East Friesian is

one of the highest milk pro-

ducing breeds, it has one of

the poorest fat and protein

contents (5.5–6.5% and 5%

respectively). Moreover, the

increment of the fat content

during the lactation is very

small (1–2%). The poor fat and

protein content is very detri-

mental to the production of

high quality sheep’s milk

cheese, which depends entirely

on fat and protein for yield,

flavor and texture.

According to Flamant and Ricordeau(1969) the East Friesian has a rathershort breeding season starting 12 to 18weeks after the longest day of the year.The best breeding period is betweenSeptember and November. Ewe lambsare precocious enough to be bred suc-cessfully at 7–8 months of age.

HealthThe East Friesian has a reputation forbeing fairly susceptible to pneumoniaand having trouble adapting to newenvironments. Flamant and Ricordeau(1969) note that the importation ofEast Friesians in many countries hasseldom resulted in a durable implan-tation of the breed in the country. Thisis particularly true in countries wherethe climate is radically different fromthe climate of the country of origin(Germany). Gootwine and Goot (1996),in their study of East Friesians in Israel,report a disappointing performancewith a prolificacy of only 160% and amilk production of 160 liters. Milk pro-duction decreased with increasingage rather than increasing as in otherbreeds. Katsaounis and Zygoyiannis(1986) reported especially poor viabil-ity of East Friesian sheep in Greece.

Boyazoglu et al. (1979) arrived at thesame conclusion in Sardinia. In coun-tries with climate rather similar toGermany (England for instance) itappears that the East Friesian is doingrather well. Olivia Mills (1989) reportsthat the harder it is kept, provided agood level of nutrition, the healthier itappears to be. However, this is onlythe author’s opinion and scientific andaccurate data are lacking.

Early results from the SpoonerAgricultural Research Station(University of Wisconsin) seem toindicate that lambs with more than50% East Friesian breeding may havereduced survival rates. Table 4

presents the survival rates of all lambsborn alive in the station flock in thewinter/spring of 1999, grouped bybreed of sire and expected proportionof East Friesian breeding in the dam.The survival rates varied from 100% to70% among the groups with thelowest survival rates for lambs withEast Friesian sires and East Friesian-cross dams. The various groups alsodiffer on the age of dams and lambingdates which may also affect lambsurvival. However, given these limita-tions of the data, the data have beenregrouped by expected proportion ofEast Friesian breeding in the lambsand presented in table 5.

17


Table 4. Arithmetic means for survival of lambs born alive by breed of sire and dam’s percentage of East Friesianbreeding at the Spooner Agricultural Research Station

————Survival rate, % ————Breed Dam’s % EF Dam Lambing No. lambs Birth to Weaning Birth toof sire breeding age, yr dates born alive weaning to 7/1/99 7/1/99

EF 0 1-9 2/4-5/28 60 95.5 93.0 88.4

EF >0 to <50 2-4 3/13-4/13 19 84.2 93.7 78.4

EF = or >50 1-2 2/4-5/18 132 82.1 85.5 70.2

Lacaune 0 3-5 4/7-5/1 45 95.5 100.0 95.5

Suffolk 0 2 2/26-3/20 10 100.0 100.0 100.0

Suffolk >0 to <50 2-4 2/2-3/27 135 97.0 99.2 96.2

Suffolk = or >50 2-4 2/2-3/26 70 97.1 99.0 96.1

Texel 0 2 1/10 1 100.0 100.0 100.0

Texel >0 to <50 2-4 3/25-4/5 11 90.9 100.0 90.9

Overall lamb survival of the flock 483 91.7 94.6 86.7

Percentage of dead lambs that died from pneumonia 45.9 91.3 63.3

Thomas et al. (1999)Table 5. Least squares means for lamb survival by percentage of East Friesianbreeding.

—————Survival rate, % —————Lamb’s % No lambs Birth to Weaning to Birth toEF breeding born alive weaning 7/1/99 7/1/99

0 56 96.4 ± 3.5 a 100.0 ± 2.9 a 96.4 ± 4.2 a

>0 to <25 146 96.6 ± 2.2 a 99.3 ± 1.8 a 95.9 ± 2.6 a

=>25 to <50 70 97.1 ± 3.1 a 98.5 ± 2.6 a 95.7 ± 3.8 a

50 60 95.0 ± 3.4 a 93.0 ± 2.8 a, b 88.3 ± 4.1 a

>50 151 83.4 ± 2.1 B 86.5 ± 1.9 b 72.2 ± 2.6 ba, b Within a column, means with a different superscript are different (P>.05)

Thomas et al. (1999)

Use in crossbreedingThe improvement in milk productionobtained by crossbreeding with EastFriesians is generally spectacular.Flamant and Ricordeau (1969) reportthat increases of 30–80% wereobserved on F1 issues of crossing localbreeds with East Friesian rams. In mostcases, however, it was not possible todetermine the effect of heterosis inthe superiority of F1 ewes, since pureEast Friesian ewes were not present.

Results of the crossbreeding experi-ment between East Friesian andDorset type ewes carried out at theSpooner Agricultural Research Station(University of Wisconsin–Madison) arereported in chapter 3.

The LacauneThe Lacaune was introduced in theUnited States in 1998 by the SpoonerAgricultural Research Station(University of Wisconsin–Madison)with the importation of two Lacaunerams from Canada and frozen semenfrom Great Britain.

The Lacaune is the most importantsheep dairy breed in France with800,000 ewes being milked mostly forthe production of Roquefort cheese. Itis important to note that before 1965the Lacaune breed, although used tra-ditionally as a milking animal, couldnot be considered a “dairy” animal.With the advance of milking tech-niques (the milking machine) and itsexpansion in the 1960s, as well as thehigh demand for sheep milk products,an intense selection program wasstarted. With an improvement of 6.3%per year (annual phenotypic gain3.9%, annual genetic gain 2.4%,Barillet, 1995), the milk production ofthe Lacaune breed increased from 80liters to 270 liters in about 30 years(figure 1 and table 6). Contrary to EastFriesian records, the numbers givenfor the Lacaune breed, always refer tothe milking period only (165 days)excluding the suckling phase (Barilletet al., 2000)a. In 1985, fat and protein

18


In all lamb growth intervals,

lambs with over 50% East

Friesian breeding had lower

survival rates than lambs with

less East Friesian breeding.

There was a tendency during

the postweaning period for

lambs of 50% East Friesian

breeding to have lower

survival rates than lambs of

less than 50% East Friesian

breeding.

Table 6. Summary of the evolution of the selection program of the Lacaune Lait.

1970 1980 1990 1999

number of ewes in milkrecording program 45,129 296,400 558,500 714,000(% of population) (9%) (49%) (74%) (90%)

numb er of ewesin AI program 31,100 93,700 294,000 322,300

number of ramsprogeny tested/year 40 332 453 450

Milk yieldin nucleus* (liters) 115 155 245 270

Milk yield in basepopulation* (liters) 110 125 200 220

* milking period only Upra Lacaune (1999)

300

200

100

0

milk

yie

ld (l

iter

s)

years65 70 75 80 85 90 95 99

base populationnucleus

Figure 1. Phenotypic evolution of the milk yield(milking period only) of the Lacaune

content was added to the selectionprogram to enhance the cheesemaking properties of the milk. Since2001 the selection scheme includesthe resistance to sub-clinical mastitisand udder morphology.

Size and conformationThe Lacaune is a large-frame breed.Adult females weigh 70–75 kg andmales around 95–100 kg. The carcassconformation of the dairy typeLacaune is average.

WoolThe Lacaune breed has very littlewool. Its head, legs and a goodportion of its belly are bare (see photoon p. 6). The fleece has a very shortstaple of medium quality and weighsno more than 1.5–2.5 kg. This lack ofwool has an enormous advantagewhen it comes to milking. However,management of the Lacaune breed inareas of the U.S with cold wintersneeds to accommodate the scantwool cover.

ReproductionThe Lacaune adult has an average pro-lificacy of 170–180% (induced estrus)with a rather long breeding seasonstarting early (June–July) making it anideal breed for late fall or early winterlambing. Ewe lambs can be bred verysuccessfully at the age of 7–8 monthsand have a prolificacy of 140% (Perret,1986).

HealthSince 1992, the dairy Lacaune breedhas been imported by 17differentcountries (Barillet et al, 2000)b for useeither as purebred or in crossbreedingsystems. Lack of adaptation or poorlivability of adult animals or lambshave not been reported. As shown intable 4, 50% of Lacaune lambs born in1999 at the Spooner Research Stationhave a good survival rate (95%). Nomajor health problem was recordedby the original importer of theLacaune breed in North America, adairy sheep producer in Ontario,Canada (Regli, 1999).

Use in crossbreedingVery little information is available onthe milking ability of Lacaune cross-bred ewes. Barillet (personal commu-nication, 2000) indicates that earlyresults seem to show a higher milkproduction in Sarda x Lacaune cross-bred ewes than in Sarda purebredewes while maintaining a high level ofbutterfat. Apparently the Lacaunecould be successfully used to increasemilk production of domestic breeds.The Spooner Research Station recentlybegan a comparison between EastFriesian x Dorset (or Polypay) andLacaune x Dorset (or Polypay).

19


It is important to note that

there are three types of

Lacaune in France. One has

been selected solely on its

milking performance (Lacaune

lait), one type on growth and

conformation (Lacaune

viande) and one because it

possesses a major gene for

prolificacy. Producers wishing

to introduce the Lacaune on

their farms should be aware of

the difference among the three

types. The meat type Lacaune

(Lacaune viande or prolific

Lacaune) does not possess any

dairy characteristics.

The Assaf The Assaf is a synthetic breed formedin Israel in the ‘60s and ’70s, composedof 5⁄8 Awassi and 3⁄8 East Friesian. TheAssaf became very popular in Israelmostly because of an increase in lambproduction due to the higher prolifi-cacy of the East Friesian. It was alreadyan excellent milker. The Assaf breedhas been exported to several coun-tries such as Portugal and Spain whereit can be found in large numbers.

The British MilksheepThe British Milksheep is a medium-large polled sheep with a predomi-nantly white face and legs. It wasdeveloped in the UK using the EastFriesian as one of the breed compo-nents to fill the demand for a highperformance crossing sire and for ahigh yield dairy ewe. The BritishMilksheep is highly prolific (200% inewe lambs to 300% in adult ewes). Itcan achieve an average milk yield of450 liters in a 7-month lactation with aparticularly high solid content(National Sheep Association, 1992). Noreal data concerning the milk produc-tion of the breed could be found. Thetraits do not appear to be definitivelyfixed and the percentage of EastFriesian blood could be variable.

Some British Milksheep animals arealready present in Canada.

ReferencesBarillet F.1995. Genetic improvement

of dairy sheep in Europe. FirstGreat Lakes Dairy SheepSymposium. March 30, 1995.Madison. Proceedings.

Barillet F. 2000a. Personal communication.

Barillet F., J.M. Astruc and C. Marie.2000. Amélioration génétiques desbrebis laitières. Cours CSAGAD,Nantes, October 2–6, 2000.

Barillet F., C. Marie, M. Jacquin, G.Lagriffoul and J.M. Astruc. 2000b.The French Lacaune dairy sheepbreed: use in France and abroad inthe last 40 years. 51st Annualmeeting of the EAAP, The Hague,The Netherlands, August 21–242000

Berger Y.M. 1998. Sheep Research. 10thAnnual Report. SpoonerAgricultural Research Station.University of Wisconsin–Madison.

Boyazoglu J.G., S. Casu and J.C.Flamant. 1979. Crossbreeding theSardinian and East Friesian breedsin Sardinia. Ann. Génét. Sél. anim.,1979, 11 (1), 23-51

Boylan W.J.. 1995. Sheep dairying inthe US. 1st Great Lakes DairySheep Symposium, Madison,March 30, 1995. Proceedings.

Flamant J.C. and G. Ricordeau. 1969.Croisements entre les races ovinesPrealpes du Sud et Frisone(Ostfriesisches Milchschaf ). I. Labrebis laitiere de Frise Orientale.Elevage en race pure. Utilisation encroisements, Ann. Zootech., 1969,18 (2), 107-130.

Gootwine E. and H. Goot. 1996. Lamband milk production of Awassi andEast Friesian sheep and theircrosses under Mediterranean envi-ronment.Small Rum. Res. 20:255-260

Jordan R.M. and W.J. Boylan. 1988. ThePotential for a dairy sheep industryin the Midwest. MinnesotaExtension Service. University ofMinnesota. AG-FO-3430.

Katsaounis N. and D. Zygoyiannis.1986. The East Friesian sheep inGrece. Res. and Develop. in Agric.3:19-30.

Mills, O.. 1989. Practical SheepDairying. Thorsons Publisherslimited, Wellingborough,Northamptonshire NN8 2RQ,England.

National Sheep Association. 1992.British Sheep.

Perret G..1986. Les races ovines.Publication ITOVIC 1986.

Regli J.G..1999. Farm adapted breeds: apanel presentation of flock per-formance records: Lacaune dairysheep. Proceedings of the 5thGreat Lakes Dairy SheepSymposium. November 4-6, 1999,Brattleboro, Vermont, USA, 51-54.

Thomas, D.L., Y.M. Berger and B.C.McKusick, 1999. Preliminary results:Survival of high percentage EastFriesian lambs. 5th Great LakesDairy Sheep Symposium.Proceedings. Nov. 4-6, 1999.Brattleboro, Vermont.

UPRA-Lacaune. 1999. La race Lacauneet ses terres d’élevage. Upra-Lacaune, Route de Moyares, Rodez,France.

20


.Producers have four options toimprove the milking ability of ewes:

(1) selecting a domestic breed to bemaintained in a purebred system;

(2) using an F1 crossbreeding systemwith an improving breed;

(3) creating a synthetic line afterdetermining which breed combi-nation results in the most efficientproduction system; and

(4) upgrading a local breed to a dairybreed by systematic mating of allfemale progenies with a purebredram of the improved breed.

Each option has its advantages andpitfalls, and the choice is not easy.Barillet (1995), one of the leading dairysheep geneticists, explains that if nomore than 50% germplasm of theimproved breed is desirable:“…improvement of native sheep pop-ulation through reliance on improvedbreeds appears in most cases to be

too difficult to manage (option 2: F1crossbreeding system). Breedingstrategies involve either the creationof synthetic lines through crossbreed-ing of local and imported productivebreeds to avoid having more than50% of the genes coming form theimported breed (option 3), or throughdairy selection of local breeds in theirnative area of production (option 1).The balance between cost/time andspecific situations must therefore betaken into account for any finaldecision. In practice the genetic andeconomic comparisons carried out inthe 1970s in Western Europe con-cluded most often that it was wiser torely on implementing the dairy selec-tion of the local breeds in their specificarea and condition of production.”

Without denying its accuracy, thisstatement should be examined in thecontext of the North American sheepproduction system. Selection in apurebred system or creation of a syn-thetic breed requires a long time anda complex organization that mightnot yet exist in North America’s bur-geoning dairy sheep industry.

21

Improving lactation through genetics

Chapter 3

Selection in a

purebred system

or creation of a

synthetic breed

requires a long

time and a complex

organization.

Genetic traits to considerThe main trait to consider for improve-ment of dairy ewes is milk yield. Witha moderately high heritability (0.30) itis quite possible to significantlyincrease the milk production of abreed. With an optimum selectionprogram, a genetic gain of 2.4%annually can be expected (Barillet,1995).

The fat and protein composition ofmilk determine its manufacturingqualities. Since the vast majority ofsheep’s milk is made into cheese and alesser amount into yogurt and icecream, milk composition is economi-

cally important to sheep milk proces-sors. Fat and protein composition havea heritability as high (or higher) asmilk yield (table 1). Therefore, progressfrom selection can be expected inthese traits.

Selection for high fat and proteincontent may become an issue in thevery near future because both traitscorrelate negatively with milk yield.Barillet (1995) reports a negativegenetic correlation of -0.40 for proteincontent and of -0.30 for fat content(figure 1). As milk yield improvesthrough genetic selection (or cross-breeding with a high producing breedsuch as the East Friesian), the fat andprotein content of the milk is

expected todecrease,making it lessdesirable forcheese manu-facturing.Therefore, itappears essen-tial that fat andprotein beincluded in aselectionprogram as

soon as possible but not beforetangible gains are obtained on milkyield because the genetic gains aresmall during the starting period(figure 2).

Although milk yield, fat and proteinpercentages are the main traits toconsider at the start of a breedingprogram, other economic traits arenow becoming more and more impor-tant and should not be ignored. Thesetraits range from feed efficiency andmilkability to udder health andgenetic resistance to diseases.

Feed efficiencyThe effect of selection for high milkproduction on feed efficiency hasbeen studied by Marie et al. (1996) inthe Lacaune breed. In a managementsystem in which there was no indi-vidual feeding according to the levelof production, the authors foundthat animals selected for high milkproduction tended to have betterfeed efficiency because more oftheir body reserves went towardmilk production. The authorsconclude that it is therefore not neces-sary to include feed efficiency in theselection program but that feed effi-

22


protein content

protein yieldfat yield

fat content

milk yield0.26

0.89

0.83

-0.31 -0.40

0.91

-0.03

0.63

3.0

2.5

2.0

1.5

1.0

.5

0

ann

ual

gen

etic

gai

n

(%o

fpo

pu

lati

on

mea

n)

years0 5 6 8 10 15 20

base population (d = 50%)

d = diffusion rate

base population (d = 100%)nucleus

Table 1. Heritabilities of milk traits in Lacaune first lactation

h2 estimated h2 estimatedTraits from 3-4 test days from all lactation

Milk yield 0.30 0.32

Fat percentage 0.35 0.62

Protein percentage 0.46 0.53

Fat yield 0.28 0.29

Protein yield 0.29 0.27

Barillet et al. (2000)

Figure 1. Genetic correlations of milk traits in Lacaunefirst lactation

Figure 2. Annual genetic gain for nucleus and base flock

Barillet et al. (2000)

Barillet (1995)

ciency should be monitored to avoidany possible genetic drift. For the lastfew years, because of better feed effi-ciency resulting from selection basedon milk production, Lacaune dairyfarmers have been able to reduce dras-tically the amount of expensive con-centrates in the ration of their eweswithout affecting the milk yield,although forages (hay) are given freely.

MilkabilityAn identical favorable indirectresponse on milkability has beenfound in the Lacaune when selectionfor high milk production is combinedwith simplified machine milking tech-niques (suppression of machine andhand stripping). Ewes were indirectlyselected for better milk flow, andlower stripping yield (Marie et al.,1998). However, de la Fuente et al.(1999) observed a degradation of theoverall udder conformation with anincrease of cistern height. The teatsalso tended to be in a more horizontalposition, leading, in the long term, tomore difficult milking routines (fallingoff of clusters). It seems that a selec-tion on udder morphology based onudder scores developed by de LaFuente (chapter 6) should accompanya selection on milk traits, but notbefore obtaining tangible improve-ment on milk yield.

Genetic resistance to diseaseMastitis. Somatic Cell Count in milkhas been established as a good indica-tor of udder health. A high somatic cellcount indicates some sort of mastitis(clinical or sub-clinical). A high somaticcell count not only decreases overallmilk production; it also influences thequality of the milk and can have somenegative effects on human health. TheUnited States and Canada have put amaximum limit on the number ofsomatic cells sheep’s milk can containand many cheese processors haveincluded a penalty in the payment ofmilk with high somatic cell count.Heritability of somatic cell count of0.10 to 0.18 has been estimated(Barillet et al., 1999) and indicates thatselecting against sub-clinical mastitis ispossible and desirable.

Scrapie. Scrapie is an infectious diseaseof sheep that attacks the centralnervous system and is always fatal. It isone of a number of transmissibleencephalopathies found in variousanimal species including humans.Upon necropsy, infected sheep showholes or vacuoles in the brain tissue.Scrapie has a very long incubationtime—several months to a few years—so the disease is seldom seen inanimals under 11⁄2 years of age.

While the disease affects relatively fewsheep, scrapie is of major concern tofederal animal health officials becausefeed ingredients made from scrapie-infected sheep in the United Kingdommay have been the initial cause of BSE(mad cow disease) in cattle in thatcountry. Consumption of meat fromBSE-infected cattle has been impli-cated as a cause of a newencephalopathy in humans in westernEurope. The presence of scrapie in theNorth American sheep populationalso limits the opportunities forbreeding sheep exports to manycountries.

Studies in Europe and North Americaindicate that certain alleles at theprion protein locus have an effect onan animal’s susceptibility to scrapie.Differences in amino acids of the prionprotein in at least two positions, orcodons, appear to affect susceptibilityto scrapie strain A and strain C. At the136 codon, two amino acids havebeen identified in sheep populations:alanine decreases susceptibility andvaline increases susceptibility toscrapie strain A. At the 171 codon, theamino acid arginine is associated withdecreased susceptibility, and theamino acid glutamine is associatedwith increased susceptibility to scrapiestrain C. Table 2 summarizes theseresults.

23

I M P R O V I N G L A C T A T I O N T H R O U G H G E N E T I C S C H A P T E R 3

Table 2. Genetics of scrapie susceptibility

Prion protein —Scrapie susceptibility—composition Genotype Strain A Strain C

Codon 136:

alanine (A)/alanine (A) AA Low susceptibility No effect

alanine (A)/valine (V) AV High susceptibility No effect

valine (V)/valine (V) VV High susceptibility No effect

Codon 171:

arginine (R)/arginine (R) RR No effect Low susceptibility

arginine (R)/glutamine (Q) QR No effect Low susceptibility

glutamine (Q)/glutamine (Q) QQ No effect High susceptibility

If strain A is the prevalent form ofscrapie, sheep of genotype AA atcodon 136 of the prion protein shouldbe selected, but if strain C is the preva-lent form, sheep of genotype RR atcodon 171 should be selected. Sheepwith genotypes AA,RR or AA,QRshould have a low susceptibility toboth strain A and C. Strain C scrapie isthe major strain in the U.S. because allrecent scrapie-positive sheep testedfor codon 136 and 171 genotype havebeen of genotype QQ at codon 171and either AA or AV at codon 136.

National or regional programs toincrease the genetic resistance toscrapie through prion protein geno-typing and selection of resistancegenotypes are underway in severalsheep breeds in several Europeancountries including the Manech dairybreed of the Pyrenees Region ofFrance (Smits et al. 2000). Manybreeders of blackfaced meat breeds(e.g. Suffolk, Hampshire) in the U.S. areDNA testing and selecting sheep forresistant genotypes, but there is cur-rently no active effort to genotypedairy sheep flocks.

Internal parasites. Ruvuna and Taylor(1994) reviewed the genetics ofparasite resistance. Internal parasiteshave a major effect on the efficiencyof sheep production. The majoreconomic costs associated withinternal parasitism are loss of produc-tion, veterinary and drug(anthelmintic) costs and ultimately,the death of infected animals. In theU.S. alone, losses in sheep and goatproduction due to internal parasiteswere estimated at $45 millionannually. Control of internal parasitescan be an important problem for dairysheep producers because there are noanthelmintics approved for ewesduring the commercial milking period.

Parasite eggs in sheep feces (EPG =eggs per gram of feces) is accepted asan accurate indicator of the number ofinternal parasites. Heritability of EPG ismoderate to high (25–40%) so selec-tive breeding for low EPG is expectedto increase resistance. Selectionstudies have verified this expectationwith a reported decrease in EPG of3.5–5.0% per year.

While selection for parasite resistancethrough selection for low EPG ispossible, such a selection program hasnot been undertaken on a commercialbasis. To obtain maximum responsefrom within-flock selection, all poten-tial breeding animals must have indi-vidual feces samples collected andassessed for EPG. The labor involved infeces collections and laboratoryanalyses are expensive.

A more attractive selection strategy isto select a group of ram lambs foryour own use or for sale to othersbased on high genetic value for lacta-tion traits. Determine EPG on thisselect group, and use or sell only ramlambs with the lowest EPG.

Selecting a NorthAmerican breed fordairy abilitySelection of a local breed for dairyability clearly offers many advantages:

■ The breed chosen is generally welladapted to its environment.

■ Animals are widely available andaffordable.

■ The improvement obtained is gen-erally permanent.

■ The improvement benefits manyproducers.

■ The traits to be improved havemoderate to high heritabilities (asshown in table 1).

The development of the Lacaunebreed into a dairy breed is a truesuccess story. In the late 60s theLacaune, although traditionally usedfor milking, was producing only 70 to80 liters of milk. By the late 90s, it wasone of the best dairy sheep in theworld, producing more than 270 litersin 165 days (milking period only).The introduction of fat and proteincontent in the index of selectionelevated the useful milk component.

The selection scheme used on thebreed is being adapted for otherbreeds (Manech in France, Latxa inSpain, Sarda in Italy) with varying levelsof success. Several factors, listed below,have contributed to the success of theLacaune program.They include:

■ The existence of a large populationof Lacaune. In 1970 there were5,661 producers milking more than500,000 ewes.

■ The willingness of most producersto increase the average milk pro-duction of each ewe.

■ The support of all sheep milkprocessors organized into acoherent industry.

■ A coherent and rigorous selectionprogram developed and sup-ported by INRA (National Institutefor Agronomic Research).

■ A vast array of support for therecording of performance, dataanalysis and progeny testing.

■ The development of artificialinsemination centers, estrus syn-chronization, collection of freshsemen on more than 450 rams,and systematic artificial insemina-tion of about 400,000 ewes (UPRA-Lacaune, 1999) for the rapid diffu-sion of genetic gain.

■ The development of technicalsupport for the producers onforage production, nutrition, man-agement and equipment.

24


The combination of factors allowedfor an improvement of 6.3% per yearconsisting of a phenotypic gain of3.9% (management, nutrition) and agenetic gain of 2.4%. Since 1995 thephenotypic gain has been negligible(Barillet, 1997).

Obviously none of these favorablefactors are present in the small dairysheep industry of North America.Producers interested in developing alocal breed into a dairy breed wouldhave to rely on systems requiringfewer resources such as a SireReferencing Scheme (with AI) or Ramcircle (without AI). In countries withoutrequirements for ram registration,cooperative nucleus schemes and/orsire referencing schemes seem likely tobecome the basis of wide-scale breedimprovement programs (Banks, 1997).

Sire referencing schemeAn individual flock owner workingindependently faces severe limita-tions. Individual flocks are often small,making it difficult to get good geneticcomparison between animals—andthe choice of animals is oftenrestricted. Typically, single flockowners practice selection at a verylow level with limited accuracy. This isespecially true with rams, which areselected from the “best” ewes withoutadjusting for non-genetics effects. Theaccuracy of estimated genetic merit ofa ram from the production of hismother is only √1⁄4h2 = 0.28 (if h2 =0.32). The producer might be forced toselect an animal that is not quite whatwas expected. The other alternativewould be to turn to another producerwith the uncertainty of the geneticmerit of the supplying flock. Thesolution to these problems lies in pro-ducers with similar goals workingtogether.

Sire referencing uses a team ofcommon sires over a group of flocks inorder to create genetic links betweenmember flocks. With genetic linksbetween flocks it becomes possibleto compare animals betweenflocks regardless of the differ-ent environment, manage-ment, nutrition or other non-genetic effects. EPDs(Expected ProgenyDifferences) can be calcu-lated. An EPD calculation has aprospective ewe or ram replace-ment would use the milk yields (orany other trait) of the individual’s dam,maternal grand-dam, paternal grand-dam, full-sisters, half-sisters and anyother female relatives with milk pro-duction records. Rams with thehighest EPD become reference ramsand are used on a certain percentageof ewes (generally the ones with thehighest EPDs) of each member flockeither with natural mating or, betteryet, with artificial insemination. Theconditions of use of the rams andreward to producers (payment, time ofuse, cost etc…) need to be sorted outby the group of producers.

EPD calculations require relativelysophisticated statistical techniquesand fairly large computing resources.EPDs are currently calculated by theNational Sheep Improvement Program(US) on a few meat breeds. It is up to agroup of sheep dairy producers toform an association and to work withNSIP (or another entity) for the calcu-lation of milk production EPD

The Sire Reference Scheme is not anew concept since most selectionschemes use similar principles ofnucleus and diffusion throughout thebase population. New Zealand,however, pioneered the method ofram circles working with smaller pop-ulation and achieved good results onmeat and wool breeds. Some similarbreeding programs are being devel-oped in other countries (Spain forexample) on commercial dairy sheepoperations.

Organization of a sire referencing schemeThe organization of a Sire ReferencingScheme should include include thefollowing steps:

1. Formation of a group of producerswith similar goals. Ten producerswith 200 ewes each would be agood starting number. Takingadvantage of groups alreadyexisting such as marketing cooper-atives would be desirable.

2. Definition of the selection criteria.Establishment of standardizedmethods of performance record-ing (milk testing and pedigreerecording).

3. Calculation of intra flock EPDs (orcalculation of milk productionadjusted for age) for the determi-nation of the best 30 ewes of eachflock (10%–20 % of the total flock).The best 30 ewes of each flock

25


In a selection

program, only producers

working together toward the

same objectives will achieve

tangible genetic gain.

Any type of selection

program is a long term opera-

tion and spectacular improvement

cannot be expected in just a few years.

It might take as long as 10 years to

observe genetic gains although phe-

notypic gains will occur as

soon as 5 years.

form a nucleus of selection, whichis kept fragmented rather thanbeing kept together on one farmonly. Generally the best ram lambswill come from this nucleus. Ewesof the nucleus are replaced withthe best EPDs ewes.

4. Determination of 3 to 4 initial ramsto use the first year to creategenetic links between flocks.

5. Collection of semen and insemina-tion with fresh or frozen semen ofthe best 30 ewes of each flock (seefigure 1). The rest of the ewes ineach flock are bred with rams ofthe producer’s choice.

6. Calculation of EPDs on the first lac-tation of progenies.

7. Selection of 3–4 ram lambs withbest EPDs.

8. Collection of semen of those ramsand insemination of the best 30ewes of each flock. The rest of theewes are bred with rams of theproducer’s choice, generally ramswith second best EPDs from hisown flock or purchased from anymember of the group.

Improvement is rather slow during thefirst 3–4 years because of the differentphases essential for the set up, but thespeed of improvement increasesrapidly.

Improvementthrough crossbreedingThere is no doubt that by using highmilk producing breeds such as theEast Friesian or perhaps the Lacaune,spectacular improvement can beachieved very quickly. An average milkyield of 160–180 liters per ewe can beobtained in just 3–4 years afterstarting with an original flock ofDorset type ewes as observed at theSpooner Agricultural Research Station.

Milk production of 1-, 2- and 3-year-old ewes (corresponding to theirnumber of lactation) is shown in table3. East Friesian cross ewes have a lac-tation length 30–40 days longer thanDorset type ewes and produced morethan twice as much milk. Fat andprotein percentage of milk from

Dorset type ewes is approximately 0.5percentage units higher compared tomilk from EF-cross ewes. No differencecan be found between milk produc-tion, fat and protein percentagesbetween ewes of different EF percent-age. Fifty percent EF ewes do notproduce more milk than 25% EF ewes.Higher milk production of crossbredewes with up to 50% EF breedingcompared to local ewes has beenreported by Ricordeau and Flamant(1969b), Kalaissakis et al. (1977),Katsaounis and Zygoyiannis (1986),Newman and Stieffel (1999). However,ewes with greater than 50% EFbreeding have been reported toproduce both less (Kalaissakis et al.,1977) and more (Ricordeau andFlamant, 1969b) milk than local breeds.Gootwine and Goot (1996) found thatpure EF and EF-cross ewes were either

26


Flock A

Reference sires

Flock B

Flock CFlock D

Progeny ofreference sires

Progeny of other rams

Flock E

Figure 3. Sire referencing scheme

inferior or similar to the improvedAwassi ewes (one of the best milkingbreeds in the Middle East) for milkyield. The poor lactation performanceof ewes of high percentage EFbreeding in these Mediterranean envi-ronments is thought to be due to pooradaptability of the breed to high tem-peratures (Boyazoglu, 1991).

In North America, the level of EFbreeding, however, still needs to bedetermined but it seems rather certainthat a high level of EF is not necessaryto achieve correct level of production.A high level of EF might result in lowerproductivity due to a lower degree ofadaptability to a new environment,and to a higher incidence of healthproblems (See Chapter 2, Choosing abreed).

Management of an F1crossbreeding systemThe management of a crossbreedingsystem in which the desired animalshould not have more than 50%breeding of the improved breed israther straight forward and much sim-plified if the producer can purchase F1replacement ewes. Starting with agroup of unimproved ewes (Dorset forexample) the producer would breedthese ewes with a purebred dairybreed ram. All females born are keptand put at milking as soon as possible.The following year the same system isused, mating only a certain number ofunimproved ewes to provide enoughF1 replacement ewes for the dairyflock. The producer who cannotpurchase F1 replacement ewes mustkeep a sufficient number of unim-proved ewes to provide replacementF1 ewes and replacement unimprovedewes. F1 ewes should be mated to aterminal sire breed (Suffolk,Hampshire, Texel) for the productionof slaughter lambs.

The management of F1 ewes has thegreat advantage of providing hybridvigor to the ewes and to their lambsborn from a terminal sire breed.However, although the milk produc-tion of these ewes would be at acorrect level, there is very little possi-bility of improvement since the dairytraits will come only from the sirebreed on which no selection pressureis applied for the time being in NorthAmerica. Dairy type rams used for theproduction of F1 ewes would need tocome from a proven selectionprogram. Moreover, not all ewe lambsput at milking will prove suitable forhigh milk production. A producer canexpect to eliminate about 20% of theewes put at milking the first year,forcing him to produce or purchasemore ewe lambs than necessary.

27


Table 3. Milk production of Dorset-cross and EF-cross ewes.

Number Milking only TotalBreed age of ewes length (days) milk (kg) % Fat % Protein

Dorset-cross 1 73 79 ± 5 62 ± 9 5.9 ± .6 5.3 ± .05

2 43 94 ± 7 91 ± 12 5.5 ± .7 5.8 ± .10

1⁄4 EF-cross 1 124 112 ± 4 139 ± 7 5.5. ± .4 5.1 ± .04

2 92 152 ± 5 206 ± 8 5.1 ± .5 5.4 ± .04

3 35 173 ± 7 246 ± 13 5.3 ± .7 5.1 ± .07

3⁄8 EF-cross 1 69 101 ± 5 122 ± 9 5.3 ± .5 5.1 ± .05

2 40 146 ± 7 190 ± 11 5.0 ± .7 5.3 ± .07

3 13 160 ± 12 250 ± 21 5.1 ± .5 5.2 ± .10

1⁄2 EF-cross 1 71 99 ± 5 128 ± 9 5.1 ± .5 4.9 ± .04

2 16 145 ± 11 187 ± 18 5.0 ± 1.1 5.4 ± .10

3 12 178 ± 12 250 ± 22 5.0 ± 1.3 5.1 ± .10

Berger, 1996, 1997, 1998

The management

of a first generation (F1)

crossbreeding system is theo-

retically simple but difficult to

sustain in a long term. Very

few examples of successful

long term systems

exist.useful animal purchased or developed by producer

X terminal breed

improved breed

X

local breed X

FI

Creating a synthetic breedThe lack of possible selection (that is,improvement of one or several traits)as well as the constraint for producersto keep breeds of several genotypeson the farm, or to purchase replace-ment ewe lambs leads to the creationof synthetic lines.

A synthetic breed is the combinationof several breeds (at least 2) obtainedby successive cross matings and forwhich traits have been fixed overseveral generations. An example of asuccessful composite breed in theUnited States is the Polypay. In Canadathe Arcott Rideau became a popularall-purpose breed. Israel created theAssaf from the Awassi and the EastFriesian (3⁄8 EF) and France developedthe FSL (Friesian x Sarda x Lacaune)which, although promising, was notdiffused. Producers chose to select theLacaune for milking performancerather than adapt to a breed withwhich they were unfamiliar.

Composite or synthetic breeds aregenerally created to respond to thedemands of producers facing newmarket strategies. Many breeds arecreated but very few are widelyaccepted. The creation of a syntheticbreed is generally carried out (but notalways) on research stations. Theprocess requires the breeder to:

1. Choose the initial breeds anddetermine which traits are interest-ing in each.

2. Evaluate the initial crossbredanimals on their performance andgeneral overall adaptation.

3. Determine the optimum breedingpercentage of each breed that thefinal breed should possess.

4. Multiply this optimum cross andstart an intense selection on thechosen traits. A minimum of 4 gen-erations are generally necessaryfor the traits to become fixed sothat the cross animal can be calleda “breed.”

5. Multiply the final animal to have asustainable population.

Considering 5 years of initial study, 2.5years per generation and 10 moreyears for the reproduction of animalsto a sustainable population, the totalnumber of years to develop a newbreed is roughly 20. This correspondsto the number of years it took todevelop the Lacaune into a dairybreed with an optimum selectionprogram.

It takes as long to create a new breed asto select a breed for a considered trait.The advantage of a synthetic breedresides in the combination of the mostfavorable traits of all the breedsinvolved in its creation.

The example of the FSL (Friesian,Sarda, Lacaune) helps show how anew breed is created. The Friesian waschosen for its milk yield, which at thetime (1965) was far superior to theLacaune. The Sarda was chosen for itssuperior milkability (rapid let down ofmilk at milking time). Finally theLacaune was included because ofbeing the traditional breed used bythe local dairy producers and for itsexcellent adaptation to its environ-ment. The following diagram showsthe successive crosses necessarybefore being ready for diffusion.

28


Sarde x Lacaune Friesian x Lacaune

F1 x Sarde F1 x Friesian

or 1⁄4 Lacaune, 3⁄4 Sarde X or 1⁄4 Lacaune, 3⁄4 Friesian

FSL1

FSL2 Selection

FSL3

FSL4 Multiplication

Upgrading a local breed withan improved dairybreedThe principle and management of anupgrading system are simple and veryattractive to many producers. Theprocess involves crossing domesticewes with rams (or semen) ofimproved dairy breeds. Crossing withthe dairy breed continues until ewesmake up a very high percentage ofthe introduced dairy breed, virtuallyindistinguishable from pure individu-als of the introduced dairy breeds.

The advantages of such a system arenumerous:

■ Any base population can be used.

■ Improvement in dairy traits areobtained as soon as the first gen-eration (F1) is born.

■ Improvement continues over theyears.

■ The new breed acquires all thetraits of the improved dairy breedin a few years.

■ Management is relatively simple.

However, there are also some veryserious considerations:

■ New rams from different lines arenecessary on a regular basis toavoid inbreeding. Rams should bechosen from a reliable source, andselected for milk production or anyother desirable traits. The numberof new rams can be limited if theupgrading program is accompa-nied by a Sire Referencing Schemewhich promotes the use of thesame rams in different productionunits.

■ Determine with absolute certaintythat the improved breed can adaptto a new environment. Chapter 2explained that the East Friesianbreed has serious adaptationproblems to some environmentsand that many countries haveabandoned or limited its use to nomore than 50% breeding in theirimprovement program as a result.The Lacaune is being evaluated atthe Spooner Agricultural ResearchStation, which is characterized byhot summers and extremely lowwinter temperatures. The littlewool cover of the breed could be a

concern in the typical sheep man-agement system of the region butmight be an advantage in otherparts of the country. Anotherbreed such as the Awassi (MiddleEast) is a fat-tail animal, which, asF1 (50%) would be accepted byproducers and by the industry, butwould have an adaptabilityproblem when managed aspurebred.

29


Local breed X Improved breed X

F1 X Improved breed

3⁄4 Imp. X Improved breed

7⁄8 Imp. X Improved breed

Considered purebred 15⁄16 Imp. X Improved breed or

7⁄8 Imp.or

15⁄16 Imp.

Whatever system a producer or

a group of producers uses, no

genetic improvement or deter-

mination of the best milking

animals can be conducted

without recording pedigree

and the performance of indi-

vidual animals.

Milk recordingAll parties involved must use the samelanguage to have a reliable compari-son of animals. These common termsare defined in the following section.

Definition of milk traits ■ The suckling length corresponds

to the lambs’ suckling period orthe simultaneous suckling andmilking period. If the lambs suckleonly during the colostral phase, thesuckling length is considered to bezero. If there is an initial sucklingphase, milk yield during this periodis equal either to the milk suckledif suckling only, or to the milksuckled plus that milked (shouldthere be partial milking during thesuckling period).

■ The milking-only length corre-sponds to the period during whichthe ewe is milked after the lamb(s)has (have) been weaned untildrying off.

■ The lactation length is equal tothe sum of the suckling lengthplus milking-only length: it is alsothe difference in days between thedate of lambing and the date ofdrying off.

■ The total milk yield per lactationis the sum of the milk yield of thesuckling period (milk suckled, ormilk suckled plus that milked) plusthe milk yield during the milking-only period. Only the milk yieldduring exclusive milking can bemeasured simply and accurately aspart of milk recording on farms.

Milk recordingmethodsIn 1992, the International Committeefor Animal Recording publishedguidelines for milk recording of sheep.The first test day for the flock takesplace 4–15 days after the start ofmilking for that year or season.Subsequent test days should takeplace at 28- to 34-day intervals until allewes are dried off. Three basic choicesare given for recording milk.

1. Method A4. This is the method ofreference. On each test day, milkyield is recorded at both milkings(a.m. and p.m.) and combined todetermine daily yield. This methodis the most time consuming andthe most expensive.

2. Method AC. Individual milk yield isrecorded at one milking only(either a.m. or p.m.), and total flockyield is determined by the quantityof milk in the tank after the 2milkings. The total milk yield of theflock is divided by the sum of theindividual yields. The resultingfactor is used to determine theindividual daily milk yield for theday of the control. This procedureeliminates the need to individuallyrecord ewes twice at each test day.Expenses and time are thereforereduced. However, the milk in thetank must come only from ewesthat were tested.

3. Method AT. This method is alsocalled the alternative method. Testsare performed at one milking only.On any given test day, ewes arerecorded at the a.m. milking. Onthe next test day ewes arerecorded on the p.m. milking. Ateach test the recorded figure ismultiplied by 2 in order to obtainthe total daily milk yield. Themethod is simple and accurate atthe condition that the alternationis respected. This method avoidsdifficult calculations.

Milk samples should be taken fromeach ewe, and analyzed for fat andprotein content, and for SCC, at leastthree times during the milking period.The precision of milk yield estimatesbetween the A4 and AC as well as theprecision of milk quality betweensamples taken at each test day andonly 3 times during the lactation isshown in table 4.

Milk yield can be recorded by weightor volume, although volume is pre-ferred. Since the rest of the sheepdairy world uses metric measure-ments, it is good to use the weightmeasures of grams or kilograms or thevolume measures of milliliters or liters.The volume to weight conversion fornormal sheep’s milk is: 1 liter = 1.036kilograms, or 1 liter = 2.28 pounds, or 1gallon (U.S.) = 8.64 pounds.

30


Any method is

acceptable but cannot be

changed in the middle of lac-

tation. The producer has to

decide which method to use

before the milking

season starts.

If the suckling period is not of

zero length, the milk yield in

dairy sheep takes into account

only the exclusive milking and

the length of the milking-only

period (which starts when the

lambs are fully weaned and

finishes when the ewe dries

off).

Individual milk production per milkingonly period can be estimated usingthe following formula (Fleischmannmethod):

Milk recording is best performed byprofessional organizations such asDHIA. For a minimal fee DHIA in theUnited States will perform milkrecording and milk sampling of dairysheep using its staff, equipment andlaboratories to analyze the milk. DHIAwill also calculate the total milk pro-duction of each individual ewe.

31


Table 4. Precision of estimating milk yield with AC testing and milk quality withpart-lactation samplings.

—————Comparison with A4 testing————

Loss of GeneticTrait precision Heritability correlation

Milk yield 1-2% 0.25 similar 0.99

Fat yield 3-5% 0.25 similar 0.99

Protein yield 2-3% 0.25 similar 0.99

Fat content 15-20% 0.35-0.40 decreasing 0.96

Protein content 10-15% 0.40-0.45 decreasing 0.98

Barillet (1991)

It is strongly rec-

ommended that dairy

sheep producers use a spe-

cialized organization such as

DHIA for the milk record-

ing of their ewes.

Estimated milk yield =

[production 1st test day x no.days between start of milkingand 1st test day]

+ [(production 1st test day + prod.2nd test day)/2 x no. daysbetween 1st and 2nd test day]

+ [(production 2nd test day + production 3rd test day)/2 x no.days between 2nd and 3rd testday]

+ ….

+ [( production next to last testday + production last test day)/2x no. days between next to lastand last test day]

+ [production last test day x no.days between last test day andend of milking]

Milk recording and milk sampling at La Fage Research Station (INRA),France

Electronic test jar

Adjustment factorsIf BLUP analysis (EPD calculation)cannot be done (for example, if pro-ducers do not belong to a geneticimprovement program) an animal’stotal production should be adjustedwith a correction factor for age. It isalso well known that the total com-mercial milk production is affected bythe type of lamb management (seechapter 10). The correction factors forlamb management are shown in table6. Therefore, estimated yields shouldbe adjusted for this non-genetic effectso ewes of different ages can be fairlycompared inside the same flock whereenvironmental factors are the same.Estimated lactation yields should bemultiplied by the appropriate adjust-ment factor shown in table 5 to adjustestimated milk yield to that expectedfrom a 4- to 7-year-old ewe.

The age of ewe adjustment factors intable 5 are based on a limited amountof European data and may be differentfor U.S. breeds of sheep and under U.S.production conditions. More refinedadjustment factors will be developed asU.S. milk production data become avail-able. In the interim, use of these adjust-ment factors is preferable over not usingany age of ewe adjustment factors.

Other information to recordMilk recording is a valuable improve-ment tool only if calculations of milkyield are performed accurately andattributed to an individual animal.Therefore the following information isneeded:

■ Identification of the animal. Eachanimal should be identified per-manently either by tattoo or by adouble ear tag. Electronic ear tagsshould be strongly consideredespecially if the flock is enrolled ina scrapie eradication program.

■ Date of lambing.

■ System of weaning (at milkingafter lambing, milking after 30 daysof suckling…).

■ Date of weaning or date of firstmilking.

■ Date of last milking (dry off ).

■ Sire and dam of animal.

Considering that the income derivedfrom the sale of milk in a dairy sheepoperation represents only 50–55% ofthe total income, lamb productioncannot be neglected and informationabout reproduction performance andgrowth of lambs should be recorded,such as:

■ number of lambs born; and

■ number of lambs weaned.

Individual producers should recordwhatever they believe to be valuableinformation for their operation. Sincecomputers are now widely present onevery farm, it is strongly suggestedthat a database system be used forrecord keeping, queries, forms andreports (see appendix A).

32


In any selection

program, recording the per-

formance of individual animals

and each animal’s pedigree (which

necessitates controlled matings)

is essential for the success of

the program.

Table 5. Multiplicative factors toadjust milk yield to a mature (4–7years of age) equivalent.

Ewe age, Adjustmentyears factor

1 1.44

2 1.24

3 1.13

4 to 7 1.00

8 and older 1.04

Thomas, 1996

Example:A 2-year-old ewe has a milk yield of206 liters. Her age adjusted milk yieldis 255 liters (206 x 1.24 = 255 liters).

Table 6. Multiplicative factors toadjust milk yield to the same lambmanagement systems

Management Adjustmentsystems factor

DY1 1.00

DY1 1.10

DY30 1.51

Berger and Thomas, 2004

Example:A 2 year old has an age adjusted milkyield of 255 liters. Her managementsystems adjusted and age adjustedmilk yield is 385 liters (256 x 1.51).

ReferencesBanks R.G. 1997. Genetics of lambs and

meat production. In: The Genetics ofSheep. L. Piper and A. Ruvinski(Edts). 1997. CAB International.

Barillet F., 1991. Rapport du group duCICPLB sur le contrôle laitier desbrebis:”La simplification ducontrôle laitier officiel de type A”.In: Performance recording ofanimals: State of the art, 1990.Proceedings of the 27th biennalsession of the InternationalCommittee for Animal Recording(ICAR), Paris, France 2-6 July 1990.EAAP Publication No. 50, 1991.

Barillet F.,1995. Genetic Improvementof Dairy Sheep in Europe. 1st GreatLake Dairy Sheep Symposium.March 30, 1995, Madison.Proceedings.

Barillet F., 1997. Genetics of milk pro-duction. In “The Genetics of Sheep”,Edts.: L. Piper and A. Ruvinski. 1997.CAB International.

Barillet F., R.Rupp, S. Mignon-Grasteau,J.M. Astruc, M. Jacquin and G.Lagriffoul. 1999. Genetic analysisfor mastitis resistance and somaticcell score in French Lacaune DairySheep. Proc. 6th InternationalSymposium on the milking ofsmall ruminants, Athens, Greece,September 26-October 1, 1998.EEAP Publication, No 95, 1999, 393-399, Wageningen Press.

Barillet F., J.M. Astruc and C. Marie.2000. Amélioration génétiques desbrebis laitières. Cours CSAGAD,Nantes, October 2-6, 2000.

Barillet F. and D. Boichard. 1987.Studies on dairy production ofmilking ewes. I. Estimates ofgenetic parameters for total milkcomposition and yield. Génét. Sél.Evol. 19 (3): 459-474.

Berger Y.M. 1996, 1997, 1998. SpoonerAg Research Station. University ofWisconsin. Annual Research Reports.

Berger Y.M. and D.L. Thomas. 2004. Milktesting, calculation of milk productionand adjustment factors. 10th GreatLakes Dairy Sheep Symposium.Proceedings. Nov 4-6, 2004,Hudson, Wisconsin.

Boyazoglu J., 1991. Milk breeds ofsheep. In: Genetic Resource of pig,sheep and goats. K. Maijal (Ed.),.Vol. 12. Elsevier, Amsterdam.

de la Fuente L.F., D. Perez-Guzman,M.H. Othmane and J. Arranz. 1999.Amélioration généique de la mor-pholgie de la mamelle dans lesraces Churra, Laxta et Manchega.Proc. 6th International Symposiumon the milking of small ruminants,Athens, Greece, September 26-October 1, 1998. EEAP Publication,No 95, 1999, 369-374, WageningenPress.

Gootwine E. and H. Goot. 1996. Lamband milk production of Awassi andEast Friesian sheep and theircrosses under Mediterraneanenvironment. Small Rum. Res.20:255-260

International Committee for AnimalRecording. 1992. InternationalRegulations for Milk Recording inSheep. (English translation). ICAR ,Via Alessandro Torlonia 15A, I-00161 Roma, Italy.

Kalaissakis P, T. Papadimitiou, J.C.Flamant, J.G. Boyazoglu, and N.Zervas, 1977. Comparaison desraces ovines Chios et Frisonne avecleur croisements en Grece conti-nentale. II. Production laitière.Ann. Génét. Sél. Anim. 9:181-201

Katsaounis N. and D. Zygoyiannis,1986. The East Friesland sheep inGreece. Res. And Develop. In Agric.3:19-30

Marie C., F. Bocquier and F. Barillet.1996. Influence du potentiel laitiersur les composantes de l’efficacitéalimentaire de brebis Lacaune.Proc. Renc. Rech. Ruminants,INRA, Institut de l’Elevage, 1996,3, 297-300

Marie C., M. Jacquin, M.R. Aurel, F.Pailler, D. Porte, P. Autan and F.Barillet. 1998. Déterminisme géné-tique de la cinétique du lait selonle potentiel laitier en race ovineLacaune et relations phénotyp-iques avec la morphologie de lamamelle. Proc. 6th InternatinalSymposium on the milking ofsmall ruminants, Athens, Greece,September 26-October 1, 1998.EEAP Publication, No 95, 1999, 391-388, Wageningen Press.

Newman S-A. N. and W. Stieffel. 1999.Milking performance of EastFriesian Poll Dorset cross ewehoggets. Proc. New ZealandSociety of Animal production 59:125-128

Ricordeau G. and J.C. Flamant. 1969a.Croisement entre les races ovinesPréalpes du Sud et Frisonne(Ostfriesisches). II. Reproduction,viabilité, croissance, conformation.Ann. Zootech. 18:131-149

Ricordeau G. and J.C. Flamant. 1969b.Croisement entre les races ovinesPréalpes du Sud et Frisonne(Ostfriesisches). III. Performanceslaitières. Ann. Zootech. 18: 151-168

Ruvuna F., and J. F. Taylor. 1994The genetics of parasite resistance.Sheep and Goat Res. J. 10(3):178-187

33


Smits, M.A., F. Barillet, F. Harders, M.Y.Boscher, P. Vellema, X. Aguerre, M.Hellinga, A. R. McLean, M. Baylis,and J. M. Elsen. 2000. Genetics ofscrapie susceptibility and selectionfor resistance. Book of Abstracts ofthe 51st Annual Meeting of theEuropean Association for AnimalProduction. No. 6:295.

Thomas D.L. 1996. Opportunities forgenetic improvement of DairySheep in North America.Proceedings of the 2nd GreatLakes Dairy Sheep Symposium.March 28, 1996, Madison,Wisconsin.

UPRA-Lacaune. 1999. La race Lacauneet ses terres d’élevage. Upra-Lacaune, Route de Moyares, Rodez,France.

34


Milk production with dairy ewesrequires more intensivesystems and nutrients per

animal than meat or wool productionsystems. During lactation, nutrientrequirements may be very high.Inadequate feeding may reduce boththe daily milk production and thelength of the lactation.

Adequate feeding requires properration balancing. This, in turn, requiresestimating animal nutrient require-ments, feed intake and the nutritivevalue of feed. Proper feeding strate-gies of the lactating ewe cannot bebased simply on those of dairy cows.Even though much of the informationavailable for dairy cattle is valid fordairy sheep, being aware of the differ-ences between the two species is vitalto avoid using improper feedingstrategies for the lactating ewe.

Dairy sheep are notsmall dairy cows Recommendations for feeding dairysheep are often derived from dairycows, whose nutrition and feedingmanagement have been studied moreextensively. Even though both sheepand cattle are ruminants with manysimilarities, they tend to have differentfeeding strategies and physiologicalfunctions.

Some of the most important differ-ences between the two species relateto their body sizes. Dairy sheep are, ingeneral, 10–12 times smaller thandairy cows. Many studies have shownthat in both species the total volumeof the gastrointestinal (GI) tract variesbetween 13–18% of the body volume(Parra, 1978, cited by Van Soest, 1994).As adult ruminants increase in size, GItract volume increases in direct pro-portion to body weight. This meansthat the GI tract of a 60 kg sheep is, onaverage, 10 times smaller than that ofa 600 kg cow. However, as the bodyweight increases, there is a less thanproportional increase in energyrequirement for maintenance.Maintenance energy requirements areusually proportional to the 0.75 powerof body weight (BW0.75, often calledmetabolic weight, MW). This meansthat maintenance requirements of a600 kg (MW= 121.2 kg) cow are only5.6 times higher than those of a 60 kg(MW= 21.6 kg) sheep. Dividing theweight of the GI tract by the mainte-nance energy requirements, it ispossible to estimate the digestive

35

Nutrition of the dairy ewe

Chapter 4

14

12

10

8

6

.4

2

0

120

100

80

60

40

20

0

body weight (kg)EWE COW

GI t

ract

wei

gh

t (k

g)

dig

esti

ve c

apac

ity

(kg

GI t

ract

/1M

calN

Em

ain

ten

ance

m

ain

ten

ance

NE

req

uir

emen

t(M

cal)

0 100 200 300 400 500 600 700 800

digestive capacity ca ity dige

GI tract weight GI t weigh

maintenance energy m a nrequirementsr t

Adequate feeding

requires proper

ration balancing.

Figure 1. Effect of body weight on gastrointestinal tract (GI) size, mainte-nance energy requirements (net energy, NE) and digestive capacity.

capacity (kg of GI tract available perunit of energy requirements).

The digestive capacity curve in figure1 shows that cattle tend to have morekg of GI tract available per unit ofenergy required for maintenance thansheep. This implies that cattle can“store” more feedstuff in the GI tractfor each unit (e.g. Mcal) of energyrequired for maintenance than sheep.This holds true even considering thatthe maintenance energy requirementsfor sheep per kg of MW are lower thanthose of cattle.

Fiber can be fermented only if it staysin the rumen for several hours. Thelonger it stays, the more it isdigested (up to a limit). In practice,if sheep and cattle are fed the samefibrous feedstuff, cows tend todigest better because they havelarger digestive tracts and they cankeep the feedstuff in the rumen fora longer time (table 1). The intake perkg of BW was higher in goats andsheep than in cattle. As a result, rumenretention time and dietary total tractdigestibility were highest in cattle. Thisdifference in digestibility is main-tained even when in both species theintake is much higher than that typicalof dry animals (Blaxter et al., 1966).

There are important practical implica-tions related to these facts. To com-pensate for their low digestivecapacity, sheep have to speed upthe passage of feedstuff in therumen (high passage rate, i.e. lowerretention time). Therefore, theyneed to eat more feed per day (as %of BW) than cattle to satisfy theirrequirements. Since the feed stays inthe rumen for a shorter period, eachkg of feed is digested less thoroughly.Despite this, due to the higher intakeof dry matter, the total amount ofnutrients digested per day is usuallyincreased. This explains why high pro-ducing dairy sheep may have a levelof intake between 5% and 7% of their

body weight, while high producingcows usually do not exceed 4%.

Another way sheep face this problemis to be more selective about what theyeat (Van Soest, 1994). Since sheephave less room for the feed per unit ofrequirement than cattle and they haveto speed up the passage rate, theynaturally tend to choose feeds orparts of feeds that are good qualityand highly digestible. Even if thefeed stays in the rumen for a shortertime, its digestibility is sufficient toallow the animal to meet its energyrequirements.

Sheep differ from cattle in chewingactivity as well. Sheep requirebetween 9 and 16 more times thancows to eat and ruminate 1 kg ofdry matter (De Boever et al., 1990).Sheep have to chew more than cattlebecause they are smaller animals andtheir chewing activity is less powerful.Sheep also have to grind the particlesmore finely than cattle to allow themto pass through the rumen and othercompartments of the foregut (VanSoest, 1994). This behavior was clearlyshown when lactating dairy cows(Holstein) and dairy sheep (Sardabreed) were fed a pelleted total mixedration as the only feed (table 2).

36


Table 2. Intake and chewing activity of cows and sheep fed the samepelleted total mixed ration as only feed.

Dairy cows Dairy sheep

Intake (kg of DM/day) 8.4 1.2

Eating time (min/day) 110.7 56.0

Rumination time (min/day) 19.4 78.5

Total chewing time (min/day) 130.1 134.5

Eating efficiency (min/kg of DM) 13.1 46.3

Rumination efficiency (min/kg of DM) 2.3 64.9

Total chewing efficiency (min/kg of DM) 15.4 111.2

Rossi, 1994, cited by Van Soest et al., 1994

Table 1. Apparent digestibility and retention times for ruminants fedthe same medium quality timothy hay ad libitum

Item Goats Sheep Heifer

Body weight 29 30 555

Intake of dry matterg/d 700 650 7830g/kg BW 24.3 21.7 14g/kg BW0.75 56 51 68

Digestibility (%)Dry matter 47 47 54NDF 44 44 52

Retention time of forage particlesRumen (hr) 28 35 47Whole GI tract (hr) 52 70 79Ratio: rumen/whole tract 54 50 59

BW = body weight NDF = neutral detergent fiber

Uden et al., 1982; Uden and Van Soest, 1982

While sheep spent more than an hourruminating 1 kg of dry matter, cowsruminated very little. Indeed, whilesheep were doing well with this dietand producing a good amount ofmilk, cows dropped milk yield, hadmilk fat depression, and showed clearsigns of acidosis.

Since there is a limit to the amount oftime an animal can spend ruminating(10–11 hours per day), intake tends tobe limited by the particle size ofcoarse diets containing long hay morein sheep than in cattle. This fact, aswell as the lower digestive capacity ofsheep, explains why grinding oftenincreases intake of forages and whythe response is stronger in sheep thanin cattle.

Greenhalgh and Reid (1973)compared the intake of sheep andcows fed 3 types of diets: (A) highquality; (B) medium quality; and (C)dehydrated ryegrass and a mix ofmedium quality ryegrass with barleypresented in either long or groundand pelleted form. Their results (table3) showed that grinding and pelleting:1) increased intake more in sheepthan in cows; 2) increased intake more

in young animals than in adultanimals; 3) increased intake in inverseproportion to dietary quality (B> A >C). Even in ground diets, however, thetotal daily intake of digested drymatter was higher in high quality thanin low quality diets.

Intense rumination activity in sheepcan also have important implicationswhen the diet includes grains.Rumination reduces the particle sizeand increases rumen digestibility ofgrains and therefore of starch. Sheeptend to chew grains more finely thancattle. This may explain why diets withhigh digestibility (> 66%) tend to bedigested better by sheep than bycattle, while cattle are more efficientwith diets that are harder to digest.(Mertens and Ely, 1982).

37

N U T R I T I O N O F T H E D A I R Y E W E C H A P T E R 4

Table 3. Effects of grinding and pelleting various diets on intake in sheep and cattle.___Sheep___ _____________Steers___

Age (months) 6 18 36 6 18 36Body weight (kg) 49 72 83 272 464 614

Diet Form Intake

A Long * g/kg of BW 21.9 18.1 23.8 20.5 19.9 15.7

Ground & pelleted** difference in % +59 +46 +29 +18 -17.1 +5

B Long* g/kg of BW 17.8 15.2 18.0 19.6 15.9 13.7

Ground & pelleted** difference in % +76 +74 +61 +31 +21 +30

C Long* g/kg of BW 22.0 17.5 24.6 20.5 19.7 17.3

Ground & pelleted** difference in % +49 +25 +11 +20 0 0

A= perennial ryegrass, 2nd cut, harvested 7 weeks after the 1st cut (NDF 59%, CP 19%, ADL 3.3%)

B= perennial ryegrass, 2nd cut, harvested 12 weeks after the 1st cut (NDF 64%, CP 16.6%, ADL 4.1%)

C= 60% hay B and 40% milled and pelleted barley

* = long (baled) for cows, coarsely chopped (5 cm screen) for sheep

** = ground (1.44 cm screen) and pelleted through a 16 mm die

Greenhalgh and Reid, 1973, modified

Differences betweencows and sheepSheep:

1. Must eat more than cows to

satisfy their maintenance

requirements. This results in

a higher passage rate of

feed and lower fiber

(forage) digestibility.

2. Tend to have more selective

feeding behavior than cows.

3. Are more affected in their

intake by particle size and

fiber content of the forages

than cows.

4. Have to spend more time

eating and ruminating each

kg of feed than cows.

5. Tend to have higher

digestibility of grains and

high energy diets than

cows.

Requirements ofthe lactating ewe

Energy requirementsEnergy requirements of lactating dairysheep are calculated in the same wayas those of lactating ewes of non-dairy breeds. Different organizationshave published equations to estimateenergy requirements of sheep. A com-parison of the requirements calcu-lated with the French (INRA, 1989),Australian (CSIRO, 1990) and British(AFRC, 1995) systems is reported intable 4.

The sheep NRC (1985) system doesnot specify energy requirements formilk production, probably because itwas written primarily for meat andwool sheep. For dry ewes, NRC (1985)tends to have higher requirements formaintenance than the other systems(see footnote in table 4). The CSIRO(1990) system is peculiar because itsenergy requirements for maintenancegrow in proportion to milk yield. Therationale behind this is that when thesheep produces milk, she requiressome extra energy. Indeed, whenanimals (not only ruminants) producemilk they have higher feed intakesand require extra energy to process

the extra feed (Ortigues and Doreau,1995). This leads to higher mainte-nance requirements during lactationthan during the dry period.

The requirements reported in table 4include some activity allowance forhoused sheep. If the ewes are grazing,an additional allowance should bemade for their extra movement.

On average, grazing activityincreases maintenance require-ments by 20% if the ewes are ongood quality flat pastures and by35–40% in more extensive, hillypastures (CSIRO, 1990).

If the ewes must walk long distancesto the pasture, a more precise calcula-tion can be done considering the fol-lowing values (CSIRO, 1990).

A component of maintenance require-ments often overlooked is that associ-ated with conditions of cold stress.Cold stress affects sheep muchmore than cattle. Indeed, since smallanimals have higher body surface perkg of BW than large animals, theydisperse more heat. Even though thewool of sheep is a much better insula-tor than the hair of cattle, its addi-tional insulation is less than the effectsof body size.

38


Activity Unity _________ Live weight _________

50 kg 60 kg 70 kg 80 kg

Walking (Mcal/mile) 0.04 0.05 0.06 0.07(horizontalcomponent)

Walking (Mcal/mile) 0.60 0.73 0.85 0.96(vertical component)

Table 4. Energy requirements for housed mature sheep (Mcal of ME/d).

FCM ** 50 kg of live weight * 60 kg of live weight * 70 kg of live weight *(6.5%) AFRC INRA ___CSIRO___ AFRC INRA ___CSIRO___ AFRC INRA ___CSIRO___

(kg/d) total total total maint. total total total maint. total total total maint.

0 1.53 1.79 1.57 1.57 1.76 2.05 1.79 1.79 1.99 2.30 2.01 2.01

1 3.22 3.54 3.45 1.74 3.45 3.80 3.65 1.96 3.67 4.05 3.90 2.19

2 4.90 5.29 5.33 1.91 5.14 5.55 5.51 2.13 5.36 5.80 5.78 2.36

3 6.59 7.04 7.22 2.08 7.13 7.30 7.36 2.30 7.05 7.55 7.66 2.53

* NRC (1985) = maintenance requirements (Mcal of ME/d): 50 kg = 2.00; 60 kg = 2.20; 70 kg = 2.40

** 6.5% FCM (6.5% fat-corrected milk) = actual milk yield x (0.3688 +0.0971 x % butterfat) (Pulina et al., 1989).

Cannas, 2000

The effect of cold stress in sheep wassimulated by Cannas (2000) by usingthe CSIRO (1990) model. The results(table 5) showed that lactatinganimals are less affected by cold stressthan are dry animals. This is becausethe high energy intake necessary tosustain milk production increases theheat produced by the body and bythe rumen and, therefore, alleviatesthe effects of cold stress. Wool depth isalso very important in reducing theeffects of cold stress, because of itsthermo-insulation properties.However, wind or rain can markedlyreduce its protective action. In thesimulation, the combined effects of allthese factors increased maintenancerequirements up to 3 times.

Protein requirementsCalculating the protein allowances forlactating ewes can prove to be adaunting task. Proteins supplied bythe diet are in part fermented in therumen and in part digested in theintestine. In most feeds, a totally indi-gestible fraction is also present.

The protein fermented in the rumen(degradable intake protein, DIP) isused by bacteria (if proper amounts offermented carbohydrates are present).Ruminal bacteria then pass to theintestine, where they represent amajor source of high quality proteinfor the ewe. The protein requirementsof the ewe are then satisfied in part byfeed protein that is not fermented inthe rumen and is digested in theintestine (undegradable intakeprotein, UIP) and, in part, by bacterialprotein digested in the intestine.

The problem is that the amount ofprotein fermented in the rumen (andconsequently, the amount of UIP) andthe ability of bacteria to use thatprotein is affected by many variableslike type and amount of feed eaten(usually related to milk production),feeding frequency and amount ofenergy fermented in the rumen. Inpractice, this means that it is difficultto estimate the amount of proteinneeded to meet the requirements oflactating ewes. Similar problems existfor cattle, but more information forcattle is available. Thus, most of theinformation used for lactating ewes isbased on information derived fromdry ewes or from dairy cows and maynot reflect actual requirements andfeed utilization.

39


Table 5. Effect of coat depth, wind, rainfall and current mean daily (24 h) temperature on cold stress requirements ofadult, non-lactating ewes of 50 kg of BW, with an MEI sufficient for maintenance in a thermo-neutral environment, andof lactating ewes weighing 50 kg, producing 1.5 kg/d of milk with 6.5% fat, and with MEI sufficient to satisfy mainte-nance and milk production requirements in thermo-neutral conditions (15–20 °C). Total maintenance requirements areexpressed as index, with maintenance requirements in thermo-neutral condition equal to 100.

_______________25 mm coat_______________ _______________50 mm coat_______________

Wind (km/h) calm 30 calm 30Rainfall (mm/d) 0 30 0 30 0 30 0 30

Adult, dry

Temp. +5 °C 115 134 234 247 100 103 183 195

Temp. +0 °C 129 149 267 280 100 114 208 220

Temp. -5 °C 144 164 300 313 109 124 233 245

Adult, lactating

Temp. +5 °C 100 100 125 133 100 100 107 114

Temp. +0 °C 100 100 137 145 100 100 116 123

Temp. -5 °C 100 104 149 157 100 100 125 132

Cannas, 2000

The French (INRA 1989), Australian(CSIRO, 1990) and British (AFRC, 1995)systems express protein requirementsin terms of metabolizable protein (thetotal amount of protein, of bacterialand feed origin, absorbed by theintestine). Table 6 reports the require-ments of metabolizable protein (MP)calculated by these systems. As in thecase of energy, the CSIRO (1990)system considers variable proteinrequirements for maintenance.Despite the different approach used,the INRA (1989) and the CSIRO (1990)systems predict very similar MPrequirements, while the AFRC givesthe lowest estimates for lactatingewes.

Using metabolizable protein may givemore precise estimates but requiresinformation sometimes not available.In this case, crude protein can be usedas a base for balancing the diet ofdairy ewes. The NRC (1985) uses crudeprotein (CP) instead of MP and givespractical estimates of protein require-ments for maintenance of dry females:

Feed protein values should beexpressed in terms of MP, the same asfor the requirements. Each of thesystems mentioned in table 6 uses adifferent approach to estimate MPvalue of the feeds. The description ofthe methods used by each system isbeyond the scope of this chapter.Briefly, all of them require the knowl-edge of the degradability in therumen of the feeds, which can beobtained experimentally by in vitro orin situ degradability measurements.When experimental measurementsare not available (most of the times for

field application of these systems), thevalues reported by the feedingsystems for feeds similar to thoseunder evaluation can be used. Due tothese difficulties, quite often rationbalancing is based on CP feed valuesonly. Being very clear that CP givesonly a rough idea of the protein valueof the feeds, some guidelines for itsuse in the field are reported here.

Optimal crude protein intake and con-centration in the diet can vary substan-tially depending on the intake of theanimals, the source of protein andenergy used in the diets, and thefeeding method.The NRC (1985)suggests CP concentration between13% (90 kg of BW) and 14.5 % (50 kg ofBW) for ewes producing 1.74 kg/d ofmilk and between 14% (90 kg of BW)and 16.2% (50 kg of BW) for ewes pro-ducing 2.6 kg/d of milk.These valuesmay be adequate in many situations.

This is supported by some experimen-tal results, both in early lactation(Gonzalez et al., 1984) and in late lac-tation (Pulina et al., 1990; Cannas et al,1998). Animals with high levels of pro-duction need diets with more UIPproteins. In these ewes, in fact, micro-bial protein may not be able to com-pletely satisfy the high proteindemand of the ewe.

40


Crude protein requirements

for milk production are around

120–125 g of CP per kg of milk

with 5% CP. If the milk has a

different protein content, CP

requirements for its produc-

tion should be proportionally

corrected.

However, in many other cases,

diets with up to 18.0–18.5% of

crude protein can give an extra

boost to milk production,

especially when protein

sources with low rumen

degradability are supplied.Table 6. Metabolizable protein requirements for adult sheep, expressed asg/d and relative to the AFRC requirement for dry sheep that was set to 100.MP requirements for wool production were not included.

5% true ___50 kg of body weight___ ___70 kg of body weight___

protein AFRC CSIRO1 INRA AFRC CSIRO2 INRAmilk total total maint. total total total maint. total

(kg/d) g/d of MP3

0 41 41 41 38 53 52 52 49

1 115 126 54 123 126 137 65 134

2 188 210 67 207 200 221 78 218

3 262 295 80 292 274 305 91 3031 based on the hypothesis that DMI is equal to 1.2, 1.8, 2.4, and 3.0 kg/d for dry ewes or lac-tating ewes producing 1, 2 and 3 kg/d of milk, respectively.2 based on the hypothesis of DMI is equal to 1.5, 2.1, 2.7, and 3.3 kg/d for dry ewes or lactat-ing ewes producing 1, 2 and 3 kg/d of milk , respectively.3 MP requirements to produce 1 kg of milk with 5% true protein: 74 g for AFRC; 71 g forCSIRO, and 85 g for INRA.

Body weight (kg) 50 60 70 80

Crude protein requirements (g/d) 95 104 113 122

Serra et al. (1998) defined practicaldietary CP concentrations for lactatingsheep of different body size and milkyield (table 7). Their values were basedon a extensive review of feedingexperiments on dairy sheep publishedin the literature and tend to be higherthan those reported by the NRC(1985) for sheep, implying a low effi-ciency of protein utilization by sheep.This may be justified by the fact thatsheep have high requirements ofsulphur-containing amino acids, suchas methionine, due to their wool pro-duction (Bocquier et al., 1987).Methionine is often the amino acidfirst limiting milk production even incows. Lynch et al. (1991) showed thatthe supplementation of rumen pro-tected methionine and lysine caused amarked increase in milk production inlactating sheep. In summary, proteinquality and amino acid compositionmay have large effects on milk pro-duction. An example of the combinedeffects of protein intake and qualityon milk yield is given in figure 2.

Fiber and non-structuralcarbohydrate requirementsInformation is scarce that defines theminimum fiber requirements of lactat-ing sheep. Pelleted complete dietswith NDF as low as 32% and smallparticle size were fed ad libitum as theonly feed to dairy ewes (Pulina et al.,1995). The level of intake was about4.75% of body weight and the eweshad similar milk production to theewes fed more fibrous diets. It is

important to notice that if concen-trates are fed separately from fibersources, they may induce acidosis orsub-acidosis even when as averagethe dietary fiber content is high. Theoptimal fiber intake to maximizemilk production is not known. Whendairy goats were fed diets containingfrom 14% to 26% ADF, there were nodifferences in intake and milk yieldcompared to those fed higher fiberdiets, while milk fat content wasincreased in the more fibrous diets(Santini et al., 1991).

Non structural carbohydrates (NSC).Supplied mostly by grains, NSC arecomposed mainly of starch andsugars. They tend to decrease whenthe fiber content of the diet increases.NSC are very important energysources for ewes and their rumenbacteria. However, excess NSC mayinduce acidosis and other digestiveand metabolic problems (Ørskov,1986). High roughage diets (60: 40forage to concentrate ratio) gavemuch lower milk yield than lowroughage diets (20: 80 forage to con-centrate ratio, lower NSC content) inFinn-sheep ewes in the first weeks oflactation (Brown and Hogue, 1985).

41


Table 7. Dietary CP concentration (as % of DM) suggested for different BW and milk yields.

5% true protein milk Body weight (kg)(kg/d) 30 35 40 45 50 55 60 65 70

0.5 16.6 15.8 15.1 14.8 14.5 14.0 13.7 13.3 12.9

1.0 17.7 16.9 16.5 15.9 15.6 15.0 14.5 14.3 13.9

1.5 18.5 17.7 17.4 16.7 16.4 15.9 15.7 15.2 14.8

2.0 19.1 18.7 18.1 17.7 17.2 16.6 16.4 15.9 15.7

2.5 18.9 18.3 17.8 17.5 17.0 16.6 16.4

3.0 18.6 18.0 17.6 17.3 16.9

3.5 18.3 17.8 17.6

4.0 18.0

Serra et al., 1998

3,1

2,9

2,7

2,5

2,3

2,1

1,9

1,7

daily intake of crude protein (g/d)

milk

yie

ld (k

g/d

)

165 (9.4%)

235(12.8%)

295 (16.1%)

350(19.1%)

groundnut mealgroundnut ean mealsoybean m

eed meallinseed mmealfish m

eat/bone mealmeat/blood mealblood m

Figure 2. Effect of different amount and source of protein in ewes (67 kgBW) in the first month of lactation. The lower protein level corresponds to abasal diet of hay, barley and molasses. Dry matter intake was restricted to 1.8kg per day for all diets. Numbers in brackets represents CP concentration inthe diets (adapted from Gonzalez et al., 1982, and Robinson, 1987b).

However, in dairy goats in the fourthmonth of lactation milk yield was onlyslightly higher when 45:55 forage toconcentrate ratio diets were comparedwith diets having a 75:25 ratio (Kawaset al., 1991). In contrast, Cavani et al.(1990) found higher intake and milkyield in East Friesian ewes fed, from thefifth to the seventh month of lactation,diets with 20% starch + sugarscompared to ewes fed diets with 35%starch + sugars, which in turn hadhigher positive BW variations.

In lactating dairy sheep, diets rangingfrom 14–21% CP were compared fromthe fifth to the eighth months of lacta-tion at two levels of NSC (as average,29% vs. 40% ) (Cannas et al., 1998). Theewes fed the diets with the lowestNSC concentration had higher intake(2411 vs. 2195 g/d) and producedmore milk (1428 vs. 1252 g/d). Thismay have been the result of too muchstarch in the rumen of high NSC diets,causing sub-clinical acidosis. Indeed, inthese diets milk fat, milk lactose andmilk pH were slightly lower than in thediets with the lower NSC concentra-tion.

It is also possible, however, that withthe high NSC diets the energy wasused more for body fat depositionthan for milk production, due to likelyhigh propionate production, followinga mechanism proposed by Ørskov(1986). In a recent experiment, Cannaset al. (2000) fed three diets, differingfor their forage to concentrate ratio(90:10, 70:30, 50:50) and chemicalcomposition (NSC ranging from 32%to 43% and NDF from 43% to 31%), todairy sheep in the last month of lacta-tion. The results showed that as theforage to concentrate ratio decreased,milk production decreased and bodyweight increased.

Therefore, in the second half of thelactation dietary NSC concentrationshould not be higher than 25–30%(maximum 20% of starch + sugars),

with the lower values for grass-baseddiets and the highest for legume-based diets. This is because legumeforages contain fairly high amounts ofpectins, that are included in the NSCfraction but induce different fermen-tation products than sugars andstarch. Comparing NSC utilization bycattle and sheep, it should be consid-ered that rumen digestibility of theNSC tends to be higher in sheep (thatchew grains intensely ) than in cows,while fiber digestibility tends to belower. This means that at similar NSCdietary concentrations sheep shouldhave a lower acetate to propionateratio than cows of comparable levelsof production.

Practical feeding ofthe lactating ewe

First part of the lactation(first 8–10 weeks)The first part of the lactation has beenstudied intensely because of itsinterest in wool and meat breeds. Milkproduction in this period, in fact, dra-matically affects lamb growth andbody weight at weaning. In the case ofdairy sheep, milk yield in early lacta-tion strongly affects the persistencyand the length of the lactation.

In the first weeks of lactation, intake isusually low but the requirements ofthe ewes are very high. Peak intakeusually occurs some weeks after thepeak of lactation. This brings the eweinto a negative energy balance. Milkproduction, then, relies in part on themobilization of body reserves. For thisreason, a very important aspect ofdairy ewe feeding is to allow theanimals to begin the lactation withappropriate amounts of body fatreserves (Louca et al., 1974; Robinson,1987a). Robinson (1987a) clearlydemonstrated the critical importanceof body fat reserves and energy intake

42


3.5

3

2.5

2

1.5

fat in body (kg)

max

imu

mm

ilk y

ield

(kg

/d)

5 10 15 20

ME intake7.2 Mcal/d6.0 Mcal/d4.8 Mcal/d

4% 20 g/d

14% 62 g/d

29% 104 g/d

4% 20 g/dd

33% 190 g/d

59% 360 g/d

The practical implication of

these trials is that during early

lactation large amounts of

grains (NSC up to 35–40%)

may help the ewe in negative

energy balance to produce

more milk, while later on large

amounts of grains (and then of

NSC) may be detrimental.

Figure 3. Effect of body fat reserves and daily intake of metabolizable energy(ME) on maximum milk yield in ewes in the first weeks of lactation. Thenumbers represent the percent of milk obtained from body fat mobilizationand body weight losses in grams per day (from Robinson, 1987a, modified).

in this period. In his experiments, milkproduction in ewes consuming highlevels of energy was independent ofbody condition, while milk productionof animals consuming low or mediumenergy rations was strongly affectedby it (figure 3). The lower the bodyreserves, the lower the amount of milkthat could be produced from fatmobilization. Excess body fat in thisperiod, however, may reduce thespace available for the rumen andnegatively affect intake (Stern et al.,1978).

Regardless of the condition score ofthe animals, at this stage it is critical toprovide diets that maximize intake, toavoid excessive negative energybalance and fat mobilization thatoccurs too fast. Even in the first weeksof lactation, very high intake (almost

7% of body weight) was obtained byewes nursing triplets when pelletedconcentrate was fed at will with afixed amount of hay (Hogue, 1994).This strategy promoted high growthrate of the lambs and even some bodyweight gain of the mothers (table 8)but was probably costly for practicalapplication. In general, high intake inthe first weeks of lactation can beachieved using high quality forages,finely chopped silages, chopped hay,and concentrates. To limit the risk ofgrain overload, cereal grains should bemixed with high energy feeds that areless prone to induce acidosis, such asbeet pulps, soybean hulls, or citruspulps.

Second part of the lactationDairy sheep nutrition in the secondpart of the lactation (from the thirdmonth until drying) has not beeninvestigated as completely as the earlylactation period. It is clear that dairysheep breeds have not been sub-jected to the same intense geneticselection that has occurred in dairycows. This means that the persistencyof lactation is often not as good as incows. In many dairy sheep breeds,after the first months of lactationthe ewes tend to use the nutrientsmore for body fat deposition thanfor milk production. This mechanismis even more evident when ewes ofmeat/wool breeds are used toproduce milk. In later lactation, dairysheep (Manchega breed) remarkablyincreased their milk yield whentreated with bovine somatotropin(bST), as shown in table 9 (Fernandezet al., 1995).

43


An easy and practical way to

make sure ewes have enough

body fat reserves at the onset

of lactation is to monitor their

body condition score (BCS)

throughout pregnancy and to

adjust the diet accordingly. At

lambing, dairy ewes should

have a body condition score

around 3.5 (figure 4) (INRA,

1989). At lower values of BCS,

milk yield may decrease

because of insufficient fat

reserves; at higher values it

may decrease because of low

feed intake.

Table 9. Milk yield and composition from ewes administered sustained-release bST.

bST bST wks 3 to 8 of lactation wks 11 to 23 of lactation*

80 mg/ 160 mg/ 80 mg/ 160 mg/0 mg 14 d 14 d 0 mg 14 d 14 d

Milk yield, ml/d 997 1198 1337 618 873 947

6% FCM, ml/d 1072 1301 1467 770 1071 1169

Milk fat, % 6.7 6.8 6.9 8.6 8.4 8.4

Milk fat, g/d 66 80 92 50 71 77

Milk protein, % 5.2 5.1 4.9 6.0 5.5 5.2

Milk protein, g/d 51 59 65 36 47 48

* For experimental purposes, the ewes were milked only once per day beginning at 18 weeks.This markedly reduced the difference in milk production between the control and the treatedanimals. Fernandez et al., 1995

Table 8. Observed feed intake and body weight gains of triplet-rearingewes and their lambs.

Feed intake of the ewesa Daily gain (41 days) Number of(kg DM/day) (grams/day) animals

Hayb 1.3 Ewes 250 14

Pelletsc 3.1 Lambsd 322 42

Total 4.4 3 lambs 966amean body weight at the beginning of the trial (1-2 weeks postpartum): 64.35 kgblimit fed chigh energy lamb pellets, fed ad libitum dlambs had access to pellets in acreep Hogue, 1994

Dairy cows treated with bST behave asgenetically superior cows (Peel andBauman, 1987) and tend to use thenutrients more for milk productionthan for body fat deposition. It ispossible that the high response ofbST-treated ewes is a sign that there isroom for genetic improvement of milkproduction. Basically this hormone isproducing a hormonal status that inthe future may be achieved by geneticselection.

This is probably due to the stimulatingeffect of the volatile fatty acidsproduced by grain fermentation(mainly propionate) on body fat depo-sition, as previously discussed. A betterfeeding strategy should be based onthe maximization of forage intake, theuse of by-products with fast-fer-mented fiber (such as beet pulps orsoy hulls) and the use of protein sup-plements whose protein content and

quality should be chosen consideringthe type of pasture or stored foragesavailable. Large amounts of proteinsupplements should be used if thepasture is made of mature grasses.When the pasture consists of legumesor grasses in early stages of growth,protein supplements are usuallyunnecessary. Protein supplementsused during the mating season canimprove reproductive parameters oflactating ewes (Molle et al., 1995), butonly if the protein content of thepasture is scarce.

Feeding housed lactating ewesThe lactation of dairy ewes can lastbetween 7 and 10 months. This meansthat, in most cases, ewes are fed storedforages (hay or silage) for part of thelactation.

The quality of the forages fed to lac-tating ewes is extremely important,especially if hay or silage make up alarge part of the diet. It is clear thatmilk yield and forage quality areclosely related. High quality foragesand small amounts of concentratesupplements allow milk productionlevels that cannot be obtained withlow quality forages, no matter howmuch concentrate is given.

A major problem of stored forages is

how to supply them. Many differentfeeding strategies are used on com-mercial dairies. The simplest strategy isto feed loose hay during the entireday and some concentrate supple-ments at milking. The most complexinvolves feeding total mixed rations.

Loose hay is often used as the onlysource of fiber in the diet. Because ofthe low digestive capacity of sheep,high intake of hay can be achievedonly if its quality is high (low fibercontent). If hay quality is low (highfiber content), intake will be low andthis can lead to low milk yield andhigher probability of acidosis evenwhen moderate levels of concentrateare used. One way to increase theintake of hay is to allow the animals toselect. The lower the quality of the hay,the higher the amount of hay the ewesdiscard. Van Soest et al. (1994) reportedpractical refusals for optimal lactationperformance in goats (table 10).

Quality is even more important forsilages because when chopped, theyare less readily selected. As for hay,silage feed intake can be increased byfurther reducing its particle size(Apolant and Chestnutt, 1985), unlessother factors controlling intake areinvolved (bad flavors and taste,molds). However, finely choppingshould not be considered a tool to

44


Experience and experimental

results, previously reported in

the NSC section, suggest that

feeding large amounts of grain

in the second part of the lacta-

tion stimulates fattening but

has negative effects on milk

synthesis.

To satisfy the requirements of

lactating small ruminants, it is

necessary either to provide

high quality hay or to accept

extensive selection and large

refusals. It usually ends up

costing more to feed “cheap”

low quality hay with high

refusal levels than to invest in

the production of excellent

forages.

Table 10. Estimates of practical refusals for optimal lactational performance ingoats.

Forage Predicteda Refusals Digestibility of Utilizationb

digestibility (%) (%) ingested forage (%) (%)

Alfalfa hay 65 15 69 59

58 25 66 50

50 35 60 39

Grass hay 70 20 75 60

60 35 69 45

50 50 60 30a From composition of the offered forage b Digested matter actually ingested as percent of amount offered

Van Soest et al., 1994

force sheep to eat poor quality feedsor an excuse to overlook the quality ofthe forages: What is eaten is notalways digested.

When the forages are given in a totalmixed ration (TMR), do not follow thesame procedure used for dairy cows.Sheep are more selective than dairycows and their intake is more affectedby particle size. If the particle size ofthe forages of the TMR is too large, it islikely that the ewes will first eat all theconcentrates. This may lead to acidosiseven when the average diet does nothave too much starch, as sometimes isobserved in Italian dairy sheep enter-prises.

When “cow-like”TMR diets are used fordairy sheep, another problem observedis low intake and low milk yield.Thisusually occurs because the particle sizethat maximizes intake and milk yield indairy cows is too coarse for lactatingewes.The strategy to use is to eitherproduce very finely chopped silages orto allow more grinding of the forages inthe mixer wagon.The result in most ofthe cases is an increase in both intakeand milk yield.This practical observa-

tion is supported by some experimentalevidence. Brown and Hogue (1985)compared TMR diets with two forage toconcentrate ratios (60:40 and 20:80) inwhich the forage (alfalfa hay) wasground either through a 32 mm screenor a 8 mm screen. Milk yield increased25% in the 8 mm diets, without anychange in intake. More extremegrinding may be beneficial ,too.

In a trial at Cornell, Dorset and Finnewes were fed grass hay that wasground through 12-mm (coarse), 2.4-mm (medium) and 1mm (fine) screens(Cannas, 1995). The reduction of theparticle size increased intake, milkyield and milk protein yield andmarkedly decreased ruminationactivity, while milk fat yield was notaffected (table 11).

It seems that sheep can produce welleven when fed diets that are veryfinely ground. On farms, it is almostimpossible to have diets ground asfine as in that trial. Dairy sheep pro-ducers should not be worried aboutgrinding feeds too finely for lactatingewes. Particles that are too coarse area much more likely problem.

Feeding grazinglactating ewesPasture management and grazingtechniques in producing ruminantshave been extensively analyzed inmany books and reviews and arebeyond the scope of this chapter. Amajor problem in feeding grazing lac-tating ewes is the choice of the amountand of the quality of the supplements.This section will try to give somecriteria on supplement management.

Nutritional indicators for choosing proper supplements The main problem to be faced withgrazing animals is that it is very diffi-cult to estimate both their intake andthe composition of their diet. For thisreason, any decision on quality andquantity of supplements is often theresult of a pure guess. A practical andoften more effective approach may bebased on the utilization of the infor-mation available to the dairy sheepbreeder to estimate the nutritionalstatus of the ewes and to define thecharacteristics of the supplements.Indicators commonly used are milkyield and quality, body condition andhealth status of the animals, andquality and quantity of the pasture.The relationships between milk pro-duction and nutrition are investigatedin details in another chapter of thisbook. Some other indicators are dis-cussed here.

The nutritional status of the ewescan be monitored by measuringtheir body condition score (BCS).This technique is very useful to checkif the ewes are losing too much bodyfat in the first part of the lactation or ifthey are over-eating and becomingtoo fat in the second part of the lacta-tion. The proper BCS of the ewes indifferent physiological stages is given

45


Table 11. Effect of dietary particle size on feeding behavior and milk pro-duction in lactating ewes in the 6th week of lactation. The results reportedare the means of the last experimental week covariated with those of thepreliminary period.

DietFine Medium Coarse sem

Dry matter intake (g/d) 4005 4132 3767 147

Rumination (min/d) 45a 165a 431b 38

Milk yield (g/d) 2400 2492 1991 192

Milk fat (%) 7.86ab 7.03a 9.08b 0.56

Milk fat (g/d) 187.7 185.1 178.6 18.9

Milk protein (%) 4.37 4.13 4.26 0.11

Milk protein (g/d) 105.3a 109.7a 83.8b 8.1abc Means with different subscript differ (P<0.05)

Diets: 54.9% grass hay, 30.1% cracked barley; 13.0% soybean meal, 2% mineral supplements (CP 16.4%, NDF 41.6%, ADL 3.05%); Hay ground through: 1 mm screen(FINE diet); 2.4 mm screen (MEDIUM diet); 12 mm screen (COARSE diet).

Cannas, 1995

in figure 4 (INRA, 1989). If the flock islarge, body condition can be moni-tored on only some of the ewes(about 20 % of the ewes in mediumsize flocks and 10–15% in larger ones).Supplements can then be dosedaccording to the body condition scoreof the animals and to the target BCS.

The feces are another good indicatorof the nutritional status of ewes.Liquid or loose feces often result fromexcessive protein in the diet. Ewesgrazing on young pastures rich insoluble proteins (first spring growth orregrowth after the harvest) often havethis type of feces. The use of readilyfermented carbohydrates (molasses,barley, oat) can reduce these problemsbecause bacteria need energy to use alarge amount of the fermentedproteins (Stephenson et al., 1992).These types of pasture are usually lowin fiber. The addition of some hay tothe diet may overcome this problem.Excess starch and subsequent acidosiscan also produce liquid or loose feces.In this case, however, it is oftenpossible to notice small particles ofundigested grains in the feces. Excessdietary NSC and rumen acidosis arealso indicated by the classic typicalbehavioral signs. The addition of fiberin the diet limits this problem. Pellet-like dry feces containing visible forage

particles indicate that fiber is notproperly fermented, either for its poorquality or for the lack of rumen-fer-mented dietary protein.

In dairy cows milk urea content isconsidered a good indicator of theprotein status of the animals (Roseleret al., 1993). Urea is produced in theliver to excrete N unused by theanimals. Most of the urea is excretedin the urine, but it also found in theblood and in the milk. Milk and bloodurea concentrations are closely associ-ated, although milk urea tends to be amore stable parameter and is moreeasily sampled. Blood and milk ureaare often reported as blood and milkurea nitrogen (N). Blood or milk ureanitrogen = blood or milk urea *0.4665. Some research on lactatingewes showed that blood urea was agood predictor of the protein status ofthe ewes (Egan et Kellaway, 1971).Cannas et al. (1998) reported that indairy ewes in mid-lactation milk ureaN was highly associated with thedietary CP concentration but lessassociated with the daily intake of CP(figure 5). Our unpublished calcula-tions, based on data from the scientificliterature, showed that whencompared to lactating cows fed dietswith equal dietary CP concentration,lactating sheep had 2–7 mg/dl more

urea N, with highest differences athighest CP concentrations. Cannas etal. (1997), Cannas et al. (1998) andUbertalle et al. (1998) found that milkurea was little influenced by proteinsource. First lactation ewes had highermilk urea N concentrations (4 mg/dl)than mature ewes (Cannas et al.,1997).

Other important information obtainedfrom the measurement of urearegards the reproductive efficiency ofthe ewes. In dry ewes, high blood ureaconcentrations have been associatedwith detrimental effects on earlydevelopment and survival of sheepembryos (blood urea N between 16and 23 mg/dl; Bishonga et al., 1994). Inlactating ewes, blood urea N and con-ception rates were inversely corre-lated, with significant effects for con-centrations higher than 23 mg/dl(Molle et al., 1998). Moreover, excessdietary proteins and high urea valuesare often associated with severaldiseases (e.g. mastitis, lameness) andcause an increase in energy require-ments (the energetic cost of excretionas urea of 100 g/d of dietary CP inexcess to protein requirements isequivalent to the cost of productionof 200 g/d of sheep milk). In a prelimi-nary way, a safe range of milk urea Nfor lactating sheep may be between 9and 18 mg/dl. Optimal values may becalculated by considering the optimaldietary CP concentrations reported intable 7 as the X value of the regressionequation reported within figure 5. Itseems that there are many reasons tojustify the application in the field ofthis index. It is important to noticethat individual values of milk ureahave little significance. At least 8–10sheep should be sampled to use milkurea as a nutritional indicator. Toreduce the analytical costs, individualmilk samples of ewes of the same age,stage of lactation and fed the same

46


5

4.5

4

3.5

3

2.5

2

1.5

1

bo

dy

con

dit

ion

sco

re

mating matinglambing 42 DIM 56 DIM mid- lactating

maximum mumminimum

Figure 4. Target body condition score during the production cycleof dairy ewes (based on INRA, 1989). DIM = days in milk.

diet can be pooled. Based on ourexperience, the application to sheepmilk of the commercial stripscommonly used to measure milk ureain cow milk has been unsuccessful.Milk urea is currently used in Sardinia(Italy) as a predictor of the nutritionalstatus of dairy ewes by the extensionservice of the local BreedersAssociation (Associazione RegionaleAllevatori della Sardegna). The valuesobserved throughout the year inSardinian flocks mimic almost per-fectly the availability of pasture and itsprotein content, with highest values ofmilk urea N (>28 mg/dl) in the earlygrowing season (March-April), whenboth grasses and legumes are veryrich in soluble protein, and lowest(<14 mg/dl) in the dry season (earlysummer) (personal communication,Associazione Regionale AllevatoriSardegna).

Concentrate feedingThe most common method to supplyconcentrates to dairy ewes is to givethem during milking (twice per day). Ifthe amount of grain or starchy feedsgiven each time is large (400-700 g),the ewes will likely have a surge ofpropionate production in the rumen.This may lead to reduced fermenta-tion of fiber and stimulation of bodyfat deposition. In the worst cases,acidosis (grain overload) may occur.Sometimes, even if the averageamount of grain supplied per animalis not too high, some animals, usuallythe most productive or competitive,may eat too much grain. To reduce thesurges of propionate production

(Ørskov, 1986) and the risks of lowrumen pH, excessive fattening andacidosis, the daily amount of grainshould be divided in more than twomeals. An extra meal could be given atnight in the barn or just beforebringing the ewes to the pasture, afterthe morning milking. The latter optionis particularly useful when the ewesare brought to the pasture severalhours after the morning meal. Thesupply of energy-rich concentratesjust before grazing is strongly sug-gested in the case of ewes kept onprotein-rich pastures. The goal in thiscase is not only to increase thenumber of meals, but also to optimizethe ruminal fermentation of theproteins of the pastures and to reducethe risks associated with excessiverumen ammonia concentrations.

In the case of highly fermentablegrains like barley, wheat and oat,using whole grains instead ofprocessed ones (cracked, steam-flaked, ground) is definitely benefi-cial for sheep (Vipond et al., 1985;Ørskov, 1986). Whole grains stimulaterumination and slow down ruminalfermentation. Sheep chew grains morefinely than cows and large losses ofwhole grains in the feces are unlikely.Only when slowly fermenting starchsources (corn and sorghum grains) aregiven, cracking or flaking may be ben-eficial, especially in animals with highintake and passage rate.

47


40

35

30

25

20

15

10

5

0

y =

dietary CP intake (g/d)

100 200 300 400 500 600 700

milk blood

Figure 5. Relationships between CP content of the ration, left, or daily CPintake right, and concentration of milk or blood urea N in housed lactatingewes (Cannas et al., 1998). Each square represents a treatment mean. If milkor blood urea nitrogen are used as predictors of CP % in the diet, the follow-ing equation should be used: Y = 0.48 X + 8.68, where Y = CP in the diet (%)and X = milk or blood urea nitrogen (ml/dl).

Milk urea gives important

information regarding the

type of supplements to use. In

the Sardinian case, for

example, the use of protein

supplements would be, for

most of the winter and the

spring, not only unnecessary

but harmful.

The high passage rate of feeds in lac-tating ewes poses limits to the utiliza-tion of some by-products with slowlydigested fiber and small particle size.Some of them (brewer grains, dis-tillers) may be eaten in large amountsbut may be poorly digested (VanSoest et al., 1994).

The utilization of bicarbonate can bebeneficial when high grain diets arefed. Rumen pH was increased andmaintained above the level inhibitoryto fiber digestion when a mix ofsodium bicarbonate (64%) and potas-sium bicarbonate (34%) was added atthe rate of 3.5% of the dry matter tothe diet of lambs fed large amounts ofbarley grains (Mould et al., 1983).

Acidosis and other digestive disordersare frequent in Mediterranean areasduring the winter, when the ewes arein the first months of lactation but thepasture is scarce due to the low tem-peratures. In this period, the ewes areusually fed with large quantities of hayand concentrates. However, the intakeof hay is frequently low due to its poorquality. In these cases, acidosis is fre-quently observed even when only400-600 g/d of concentrates aresupplied. With the aim of solving thisproblem, Rossi et al. (1991), developeda “safe” pelleted feed. This feed is madeof a mixture of energy and proteinsources plus about 30% choppeddehydrated alfalfa. The amount offiber and its particle size have beencalibrated to stimulate intake andrumination and to slow down therates of intake. The energy concentra-tion of this feed ranged between 1.55and 1.70 Mcal of NEL when no fatswere added. This product was used asthe only feed as long as 20 weeks(Rossi et al., 1991) or as pasture sup-plement, without giving any type ofdigestive or metabolic problem(Cannas et al., 1992; Calamari et al.,1991). When used as a complete feed,its intake ranged between 5.5% to 7%

of the body weight and alwaysinduced much higher milk yield thantraditional diets. In Italy, this type offeed is currently produced by severalfeed companies in different formula-tions. It has been used mostly inperiods when the pasture is scarce oris very young (high in soluble proteinsand low in fiber). In this case, it hasbeen beneficial in increasing milk pro-duction and in reducing many of thediseases related to nutritional stresses.

ConclusionsFeeding programs for lactating ewesshould consider the peculiar character-istics of sheep.There are substantial dif-ferences between sheep and cattle infeeding behavior and digestivecapacity.These differences are particu-larly important when ewes with highlevels of production are considered.Since very few studies have been con-ducted with high-producing ewes,much more research is needed. In par-ticular, more knowledge is needed todefine energy and protein require-ments of lactating ewes in the secondpart of lactation and fiber and NSCoptimal level throughout the lactation.In order to determine the quality andthe quantity of supplements requiredby grazing lactating ewes, it is impor-tant to use the available nutritionalindicators and to develop new ones.

ReferencesAFRC (1995)—Energy and protein

requirements of ruminants. CABInternational,

Wallingford U.K., S.M. Apolant andD.M.B. Chestnutt. 1985. The effectof mechanical treatment of silageon intake and production of sheep.Anim. Prod. 40:287-296.

Bishonga C., J.J. Robinson, T.G. McEvoy,R.P. Aitken, P.A. Findlay and I.Robertson. 1994. The effects ofexcess rumen degradable proteinin ewes on ovulation rate, fertiliza-tion and embryo survival in vivoand during in vitro culture. Anim.Prod. 58:447.(Abstr.)

Blaxter K.L., F.W. Wainman and J.L.Davidson. 1966. The voluntaryintake of food by sheep and cattlein relation to their energy require-ments for maintenance. Anim. Prod.8: 75-83.

Bocquier, F., M. Theriez and A. Brelurut.1987. Recommandations alimen-taires pour le brebis en lactation.In: Alimentation des ruminants:revision du systèmes et des tablesde l’INRA. Bull. Tech. Centre deRecherches Zootechniques etVeterinaires de Theix, INRA, (70):199-211.

Brown, D.L.and D.E. Hogue. 1985.Effects of roughage level andphysical form of diet on Finnsheeplactation. SID Research Digest, Fall1985: 11-14.

Calamari L., G.C.Rossi, A. Cannas and V.Cappa. 1991. Effect of the additionof fat and of the protein level of apelleted feed (Unipellet) on meta-bolic profile of Sarda lactatingewes (in Italian). Proceedings IXCongesso Nazionale ASPA, Rome,3-7 June 1991: 335-349.

48


Cannas A. 1995. Effects of the particlesize of the diet on feedingbehavior and milk production insheep. M.S. thesis, Cornell Univ.,Ithaca, NY.

Cannas A. 2000. Sheep and cattlenutrient requirement systems,ruminal turnover, and adaptationof the Cornell Net Carbohydrateand Protein System to sheep. Ph.D.dissertation, Cornell University,Ithaca, NY.

Cannas A., G. Annicchiarico, L. Taibi andS. Dell’Aquila. 2000. Effects of thedietary forage to concentrate ratioon milk production and live weightvariations in dairy ewes in the finalpart of lactation (In Italian).Proceedings the XIV CongressoNazionale SIPAOC, Vietri sul Mare(Italy), 18-21 October 2000.

Cannas A., A. Pes, S. Fresi, F. Serra andG. Pulina. 1997. Effects of feedinglevel and type of dietary proteinsupplement on milk urea contentin ewes. Volume of Abstracts of the48th Annual Meeting of theEuropean Association of AnimalProduction, Vienna 25-28 August1997:304.

Cannas A., A. Pes, R. Mancuso, B. Vodretans A. Nudda. 1998. Effect ofdietary energy and protein con-centration on the concentration ofmilk urea nitrogen in dairy ewes.J. Dairy Sci. 81:499-508.

Cannas A., G.C. Rossi, P. Brandano andG. Pulina. 1992. Effect of thefeeding method of a completepelleted feed (Unipellet) as supple-ment of grazing in dairy ewes.Studi Sassaresi, Ann. Fac. Agr. Univ.Sassari, 34:145-149.

Cavani C., L. Bianconi and D. Mongardi.1990. Soybean hulls and cereal dis-tillers in dairy sheep feeding (inItalian). Proceedings of the IXCongresso Nazionale S.I.P.A.O.C,Grado (VE), Italy, 20-22 June1990:6.9-6.10.

CSIRO (1990)—Feeding standards forAustralian livestock ruminants.CSIRO Publications, Melbourne,Australia.

De Boever J.L., J.I. Andries, D.L. DeBrabander, B.G. Cottyn and F.X.Buysse. 1990. Chewing activity ofruminants as a measure of physicalstructure—A review of factorsaffecting it. Anim. Feed Sci. Technol.27: 281-291.

Egan A.R. and R.C. Kellaway. 1971.Evaluation of nitrogen metabolitesas indices of nitrogen utilization insheep given frozen and dry matureherbages. Br. J. Nutr. 26: 335-351.

Fernandez N., M. Rodriguez, C. Peris, M.Barcelo, M.P. Molina, A. Torres and F.Adriaens. 1995. Bovine soma-totropin dose titration in lactatingdairy ewes. 1. Milk yield and com-position. J. Dairy Sci. 78: 1073-1082.

Gonzalez J.S., J.J. Robinson and I.McHattie. 1984. The effect of levelof feeding on the response of lac-tating ewes to dietary supple-ments of fish meal. Anim. Prod.40: 39-45.

Gonzalez J.S., J.J. Robinson, I. McHattieand C. Fraser. 1982. The effect inewes of source and level of dietaryprotein on milk yield, and the rela-tionship between the intestinalsupply of non-ammonia nitrogenand the production of milkprotein. Anim. Prod. 34: 31-40.

Greenhalgh J.F.D. and G.W. Reid. 1973.The effects of pelleting variousdiets on intake and digestibility insheep and cattle. Anim. Prod.16: 223-233.

Hogue D.E. 1994. Further observationson feeding highly productivesheep. Proc. Cornell NutritionConference, Ithaca, NY :140-142.

INRA. 1989. Ruminant nutrition: recom-mended allowances and feed tables.J. Libbey Eurotext, Paris.

Kawas J.R., J. Lopes, D.L. Danelon andC.D. Lu. 1991. Influence of forage-to-concentrate ratios on intake,digestibility, chewing and milk pro-duction of dairy goats. SmallRumin. Res. 4: 11-18.

Louca A., A. Mavrogenis and M.J.Lawlor. 1974. Effects of plane ofnutrition in late pregnancy onlamb birth weight and milk yield inearly lactation of Chios and Awassisheep. Anim. Prod. 17: 341-349.

Lynch G.P., T.H. Elsasser, C.J. Jackson,T.S. Rumsey and M.J. Camp. 1991.Nitrogen metabolism of lactatingewes fed rumen-protected methio-nine and lysine. J. Dairy Sci.66:2268-2276.

Mertens D.R. and L.O. Ely. 1982.Relationship of rate and extent ofdigestion to forage utilization: adynamic model evaluation.J. Anim. Sci. 54:895-905.

Molle G., A.Branca, S. Casu and S.Landau. 1998. Feeding and repro-duction in sheep: mating seasonand first three months of preg-nancy (in Italian). L’allevatore diovini e caprini, 15 (12):6-10.

Molle G., A. Branca, S. Ligios, M. Sitzia, S.Casu, S. Landau and Z. Zoref. 1995.Effect of grazing background andflushing supplementation onreproductive performance in Sardaewes. Small Rumin. Res.,17:245-254.

49


Mould F.L., E.R. ∆rskov and S.A. Gauld.1983. Associative effects of mixedfeeds. II. The effect of dietaryaddition of bicarbonate salts onthe voluntary intake and digestibil-ity of diets containing various pro-portions of hay and barley. Anim.Feed Sci. Technol. 10: 31-47.

NRC. 1985. Nutrient requirements ofsheep. 6th Ed., National AcademyPress, Washington, D.C.

Ørskov E.R..1986. Starch digestion andutilization in ruminants. J. Anim. Sci.63:1624-1633.

Ortigues I. and M. Doreau. 1995.Responses of the splanchnictissues of ruminants to changes inintake:absorption of digestion endproducts, tissue mass, metabolicactivity and implications to wholeanimal energy metabolism.Ann. Zootech. 44: 321-346.

Peel C.J., D.E. Bauman. 1987.Somatotropin and lactation.J. Dairy Sci. 70: 474-486.

Pulina G., A. Cannas, S.P.G. Rassu and G.Rossi. 1995. Effect of fibre andprotein content of a completepelleted feed on lactating dairyewes. Agr. Med. 125: 115-120.

Pulina G., G. Rossi, A. Cannas, G. Papoffand R. Campus. 1990. Effect of thecrude protein content of the dieton milk yield and quality of Sardaewes, (in Italian). Agricoltura Ricerca105: 65-70.

Pulina G., A. Serra, A. Cannas and G.Rossi. 1989. Determination andestimation of the energy value ofSarda ewe milk (in Italian). AttiSISVET 43:1867-1870.

Robinson J.J.. 1987a. Energy andprotein requirements of the ewe.In: Recent advances in animalnutrition - 1987 : 187-204. Editedby Heresign, W. and Cole, D.J.A.,Butterworths, London.

Robinson J.J.. 1987b. Nutrition ofhoused sheep. In: New techniquesin sheep production. 175-188.Edited by Marai, I.F. and Owen, J.B.,Butterworths, London.

Roseler D.K., J.D. Ferguson, C.J. Sniffenand J. Herrema. 1993. Dietaryprotein degradability effects onplasma and milk urea nitrogen andmilk nonprotein nitrogen inHolstein cows. J. Dairy Sci.76: 525-534.

Rossi G., A. Serra, G. Pulina, A. Cannasand P. Brandano. 1991. The utiliza-tion of a complete pelleted feed(Unipellet) in feeding dairy ewes. I.Effect of the addition of fat and ofthe protein level on milk yield andquality of Sarda ewes (in Italian).Zootecnica e Nutrizione Animale17: 23-34.

Santini F.J., C.D. Lu, M.J. Potchoiba andS.W. Coleman. 1991. Effects of aciddetergent fiber intake on earlypostpartum milk production andchewing activities in dairy goatsfed alfalfa hay. Small Rumin.Res. 6: 63-71.

Serra F., A.Cannas and G. Pulina. 1998.Feeding of dairy sheep: ration bal-ancing (in Italian). L’allevatore diovini e caprini, 15 (4):1-5.

Stephenson R.G.A., J.L. Huff, G. Krebsand C.J. Howitt. 1992. Effect ofmolasses, sodium bentonite andzeolite on urea toxicity. Aust.J. Agric. Res. 43: 301-314.

Stern D., J.H. Adler, H. Tagari and E. Eyal.1978. Responses of dairy ewesbefore and after parturition to dif-ferent nutritional regimes duringpregnancy. II. Energy intake, body-weight changes during lactationand milk production. Ann. Zootech.27: 335-346.

Ubertalle A., R. Fortina, L.M. Battaglini,A. Mimosi and M. Profiti. 1998.Effect of protein degradability onurea nitrogen in sheep milk.Scienza e Tecnica Lattiero-Casearia,49(2):67-81.

Uden P., T.R. Rounsaville, G.R. Wiggansand P.J. Van Soest. 1982. The meas-urement of the liquid and soliddigesta ritention in ruminants,equines and rabbits given timothy(Phleum pratense) hay. Br. J. Nutr.48: 329-339.

Uden P. and P.J. Van Soest. 1982.Comparative digestion of timothy(Phleum pratense) fibre by rumi-nants, equines and rabbits.Br. J. Nutr. 47: 267

Van Soest P.J.. 1994. Nutritional ecologyof the ruminant, 2nd Ed. CornellUniv. Press, Ithaca, NY.

Van Soest, P.J., B. McCammon-Feldmanand A. Cannas. 1994. The feedingand nutrition of small ruminants:application of the Cornell discountsystem to the feeding of dairygoats and sheep. Proc. 56th CornellNutrition Conference, Ithaca, NY:95-104.

Vipond J.E., E.A. Hunter and M.E. King.1985. The utilization of whole androlled cereals by ewes. Anim. Prod.40: 297-301.

50


Introduction

Like other dairy ruminants, dairyewe lactation curves, both interms of milk yield and milk com-

position, are conditioned by factorsincluding breed, stage of lactation,milking system and feeding (Flamantand Morand-Fehr, 1982; Treacher 1983,1989; Bocquier and Caja, 1993; Cajaand Bocquier, 1998). In addition, milkyield and milk composition (fat,protein, casein and serum proteins,but not lactose) correlate negatively(Barillet and Boichard, 1987; Molinaand Gallego, 1994; Fuertes et al., 1998).

This phenomenon generally appearsdue to improved management prac-tices. As a result, producers shouldstrive to find a balance between prac-tices that will increase milk yield andthose that increase the milk content.The producer’s income will reflect thecombination of prices related to bothmilk volume and its quality.

Because ewe’s milk is mainly used forcheese making, it is important to payattention to fat and protein contents.When routinely measured, these pre-cisely predict cheese yield (Pellegriniet al., 1997). In fact, the dairy sheepbreeder’s main objectives are to: 1)increase total milk dry matter output(cheese quantity); 2) stabilize yearround production of the milk content;and 3) maintain a high fat:proteinratio to ensure an adequate fatcontent for manufacturing processesand good ripening properties. Theprimary and long-term objective ofthe breeder is to improve the ewes’

dairy merit both in milk yield and milkcomposition.

As in the Lacaune breed, the objectiveof maintaining milk composition cameonly after a successful improvement ofmilk yield (Barillet et al., 1993). Otherobjectives can also include criteriasuch as milkability and mammarymorphology (Marie et al., 1999). Themerits of ewes’ milk quality differwidely between breeds, with majorvariations in milk yield, compositionand kinetics.

These discrepancies are compoundedby the large variety of productionsystems (Casu et al., 1983; Fernándezet al., 1983; Gallego et al., 1983, 1994;Labussière et al., 1983; Caja, 1994;Bocquier and Caja, 1993; Fuertes et al.,1998). Most dairy sheep productionsystems include a short lamb-sucklingperiod (3–5 weeks), and, afterweaning, a long milking period (4–8months), but “suckling-and-milking”can occur simultaneously during thefirst two months of lactation in somebreeds (Caja and Such, 1991; Sheath etal., 1995).

In regard to ewe milk composition, thelowest values in fat, protein and caseinare observed during this “suckling-and-milking” period (Gargouri et al.,1993 ; Bocquier et al., 1999 ; McKusicket al., 1999) or immediately afterweaning, and increasing later in thelactation stage. Slopes of the increas-ing curves of milk content are mainlyconditioned by the breed and level ofproduction (Bocquier and Caja, 1993).Whatever the influence of the above

51

Effect of nutrition on milk quality

Chapter 5

The merits of

ewes’ milk quality

differ widely

between breeds.

factors, feeding of the ewe modulatesboth the volume and the compositionof milk.

The aim of this chapter is to focus onthe known effects of nutrition ondairy ewe milk composition (seereview of Bocquier and Caja 1993,Bencini and Pulina 1997; Caja andBocquier, 1998). Results obtained indairy cattle and goats may not be suc-cessfully extrapolated to the dairyewe. In addition, because dairy ewesare fed mainly in large flocks, it is nec-essary to briefly analyze the effect ofthe flock structure (including days inmilk and parity) on the bulk milk com-position (Fraysse et al., 1996) and itsconsequences for feeding strategies ofdairy ewes. We artificially separatedthe global effects of nutrition from theeffects of specific nutrients that maybe effective for the manipulation ofewes’ milk composition.

Level of nutrition

Energy supply and milk compositionThe level of nutrition, mainly referredto as level of energy or of feed intake,is a major positive factor affecting milkyield and composition in dairy rumi-nants. Hence, a steep curve with anearly and high milk peak is observedwith a high nutrient supply during theearly lactation period. Conversely,nutrient shortage during pregnancyand early lactation lead to a low andlate peak milk yield.

Effects of nutrition on milk composi-tion are less clear because of interac-tions with the natural evolution ofmilk composition and throughindirect effects of nutrition on milkvolume (called dilution effect).Furthermore, during the middle andend of lactation, changes in nutrition

mainly affect the persistency and/orthe body reserves reconstitution. Thisis why limited effects are generallyobserved on milk yield or composition(Bocquier and Caja, 1993). Due to therespective variability of milk fat andprotein content, the possibilities ofaltering milk composition by feedingare higher for fat than for proteinand/or casein contents (Sutton andMorant 1989).

The specific effects of the level ofnutrition on milk composition in dairyewes are only partially known asrecently reviewed by Caja andBocquier (1998). In this sense, fewexperiments are based on individualfeeding of dairy ewes during themilking period and results obtained insuckling ewes are also taken intoaccount to obtain reliable conclusions.

Available references on the effects ofdifferent levels of nutrition in lactatingewes are summarized in table 1.

52


Table 1. Ranges of variation on milk yield and composition induced by the level of nutrition in lactating ewes.

Diet MilkLactation period Breed Energy Protein Yield Fat Proteinand reference (UFL/d)1 (gPDI/d)2 (l/d) (g/l) (g/l)

Suckling:

Robinson et al. (1974) Cheviot 2.14–2.27 188–265 2.4–3.1 76–74 54–50

Cowan et al. (1981) FxD 1.78–2.77 214–317 2.2–3.3 83–74 55–52

Cowan et al. (1981) FxD 2.28–2.33 241–277 3.3–3.5 84–92 53–56

Gonzalez et atl. (1984) FxD 1.66–2.36 183–260 2.3–2.6 90 50–52

“ “ “ 212–302 2.3–2.7 90 52–54

“ “ “ 239–339 2.5–3.1 90 53–54

Geenty & Sykes (1986) Dorset 1.99–2.00 146 2.4–2.5 76 40–39

“ “ 1.51–2.42 138–170 2.0–2.7 79–69 40–39

Perez–Oguez et al. (1994) Manch. 1.36–1.49 143–162 1.4–1.5 88–84 49

Milking:

Treacher (1971) Dorset 1.06–2.28 107–221 1.2–1.5 83–68 46–52

Bocquier et al. (1985) FxSxL 0.87–0.95 113–122 1.0 35–52 32

Geenty& Sykes (1986) Dorset 1.83 124 1.7 71 47

“ “ 1.69–2.10 132–158 1.5–2.0 71–65 53

Perez-Oguez et al. (1994) Manch. 1.41–1.50 147–164 0.6 92–99 57–58

FxD= Finnish landrace x Dorset horn; FxSxL= Finnish x Sardinian x Lacaune; Manch.= Manchega.1UFL : 1.7 NEL ; total requirements : 0.033 UFL/BW0.75 + 0.7 UFL/l of milk: 2PDI: Protein Digestible at the level of Intestine;Total requirements: 2.5 g /BW0.75 + 80 g/l (Bocquier et al, 1987b).

Existence of significant correlationbetween same milk components insuccessive controls (fat: r= +0.5;protein: r= +0.7; Barillet and Boichard,1987) suggest that effects of nutritionat an early stage of lactation may havecarry-over effects on milk compositionduring the milking period. Directevidence of such effects are lacking;however (Fraysse et al., 1996), even if itis obvious that it is of interest tooptimize nutrition during early lacta-tion because milk yield regularlydeclines.

In most dairy sheep breeds fed goodquality forages ad libitum, the balanceof energy reaches equilibrium within afew weeks after weaning (Caja, 1994;Bocquier et al., 1995) as a conse-quence of the evolution of voluntaryintake (Bocquier et al., 1987a, 1997;Pérez-Oguez et al., 1994, 1995; Caja etal., 1997) and milk yield. This may notbe the case when using large amountsof concentrate that induce a decline inforage consumption (Bocquier et al.,1983) or with poor quality forages.Milk fat content is negatively corre-lated (r=-0.87; P<0.05) to energybalance (-1 UFL/d=+12.2 g/l milk fat),this relationship being established(Bocquier and Caja, 1993) from avail-able references of suckling andmilking ewes in a wide range of netenergy balance (-1.5 to +1.5 UFL/d)and milk yield (0.6 to 3.5 l/d).Consequently, in most cases, a highlevel of dairy ewe nutrition willreduce the percentage of milk fat.

In comparison with fat content, and inagreement with cow and goat studies,the response of ewe milk proteincontent follows a positive relationship(r=+0.64; P<0.05 ; Bocquier and Caja,1993) with a lower and flatter slope. Asa consequence a high level of nutri-tion of dairy ewes generally producesa moderate increase in milk proteinand casein percentages. This was alsodemonstrated in both dairy goats(Flamant and Morand-Fehr, 1982) andcows (DePeters and Cant, 1992).

Effects of nutritionDairy ewes that graze in typical exten-sive or semi-intensive systems of theMediterranean area periodically expe-rience under-nutrition due to seasonalchanges in forages or by-productavailability (Caballero et al., 1992;Sheath et al., 1995). Moreover, in inten-sive large flocks of dairy ewes, evenwhen food supply is theoretically suf-ficient, the stage of lactation and com-petition for food between ewes oftenleads to some individual underfeedingsituations, especially in the cases ofthe most productive ewes in early lac-tation (or when rearing twins ortriplets) which have higher nutrientrequirements (Bocquier et al., 1995).

A negative energy balance producedby under-nutrition will result in adecrease in milk yield and proteincontent and in an increase in milkfat, in agreement with values shown intable 1. Slope of regression betweenmilk yield and fat percentage (-6.3 g/l)estimated by Bocquier and Caja (1993)from available data is higher thanobserved in the Lacaune population(-4.9 g/l; Barillet and Boichard, 1987)indicating that not only dilution-con-centration effects are involved in thisincrease of fat percentage. Increase ofblood free fatty acids, as a conse-quence of body fat mobilization, is animportant reason for observed highmilk fat percentage.

Effects of over-feeding Over-nutrition is also a consequenceof group feeding and is considered anormal way to restore body reservesin middle or late lactation. High levelsof intake during lactation are achievedwhen ewes have high quality dietsduring early lactation ( beforeweaning) (Pérez-Oguez et al., 1994,1995). As a general trend, when theenergy supply is increased, milkprotein content tends to increaseslightly and fat content tends todecrease, as described previously. Theexpected increment in milk proteincontent by increasing the level ofnutrition during the milking period isvery low as indicated in table 1 andBocquier and Caja (1993). Variations ofmilk content are lower than duringthe suckling period as a consequenceof differences in amplitudes of energybalance.

53

E F F E C T O F N U T R I T I O N O N M I L K Q U A L I T Y C H A P T E R 5

While under-nutrition is

mostly physiological at the

onset of lactation, its effects

during mid- or late lactation

are not well documented in

dairy ewes (Bocquier and Caja,

1993) or in cattle (Coulon and

Rémond, 1991). During this

period, a severe and chronic

under-nutrition of dairy ewes

strongly reduced the milk yield

(-31%) and increased milk fat

content in +9.6g/l (+16%),

while protein content of milk

was unchanged (Agus and

Bocquier, 1995).

It should be stressed that, in practicalconditions and as a consequence ofgroup feeding practices, the observedglobal effects of level of nutrition(over- or under-nutrition) are normallyhidden inside the feeding treatmentsand are mainly due to high yieldingewes. Individual intake of forage andconcentrate can differ according tofeed intake capacity. In these condi-tions a careful interpretation of data isrecommended.

Effects of the level ofdietary protein supply Analysis of ewes’ references (Bocquieret al., 1987b) indicates a quadratic rela-tionship (r2=0.97) between proteinsupply (in g PDI) over maintenancerequirements and milk protein yield.Mean estimation of PDI efficiency was0.56, which is close to the value (0.59)observed by protein balances (Bocquieret al., 1987). Marginal increase of proteinyield as a result of protein increment isalmost nil above 300 gPDI/d.There is,however, no significant effect of protein(PDI) balance on milk content either onfat or on protein in the compiled databy Bocquier and Caja (1993). Effects ofdietary protein level on milk productionof early lactating ewes are mainly attrib-uted to energy savings as a conse-

quence of an increase in body fat mobi-lization (Robinson et al., 1974, 1979;Cowan et al., 1981) and utilization(Geenty and Sykes, 1986).

Effects of the interaction betweendietary protein and energy werestudied by Cannas et al. (1998) in Sardaewes during the mid-milking periodand related to milk urea nitrogen. Eweswere fed in pens with whole pelleteddiets varying in two energy and fourprotein levels. Results are summarizedin table 2.

Milk yield tended to increase and milktrue protein to decrease with dietaryprotein level, in agreement withprevious conclusions. Milk yield seemsto reach a plateau above 19% of crudeprotein in the diet. Furthermore,energy level significantly reducedboth milk yield and milk fat. Milk fatvalues were low and close to thoseobserved in low fat syndrome,probably as a consequence of pelleteddiets and of high content in non-struc-tural carbohydrates. True milk proteindecreased with dietary protein levelbut was higher with the high-compared to the low-energy diet. Milkurea nitrogen, which is positively cor-related with protein in the diet, isbetter related to protein concentra-tion of the diet (r2=0.82) than withprotein intake (r2=0.56) giving an

effective indicator of N utilization. Milkurea of these ewes varied between12–27 mg/dl according to proteinlevel, which was lower than measuredin cows, and in general agreementwith measures on Lacaune ewes.

Effects of concentrate in the dietThe effect of concentrate is positivelyassociated with the energy level thediet produces as a result of its energydensity. Another effect is that milk fatmay be depressed and milk proteinincreased. Furthermore, using a highproportion of concentrates (>60% ofdry matter) in diets may depress boththe milk fat and protein contentsduring the first months of lactation(Eyal and Folman, 1978). These effectsmight be different according to theewe’s breed: higher for Awassi (fat: -28g/l; protein: -2 g/l) than for Assaf ewes(fat: -6 g/l; protein: +1 g/l).

Negative effects of concentrates onmilk production are attributed to aquick and phasic degradation of non-structural carbohydrates in the rumen,reducing dramatically the rumen pHand altering the amount and compo-sition of microbial protein synthesisand limiting the degradation of struc-tural carbohydrates. These adverse

54


Table 2. Effects of energy and protein content in the diet on milk yield and milk composition in dairy ewes.

Energy Crude protein (% DM) Meanlevel1 14 16 19 21

Milk yield (l/d) L 1.26 1.43 1.50 1.48 1.42

H 1.16 1.20 1.34 1.34 1.26

Milk fat (g/l) L 60 57 57 59 58

H 57 57 54 56 56

Milk true protein (g/l) L 55 54 54 52 54

H 57 54 53 54 55

Milk urea N (mf/dl) L 12.9 17.7 23.4 26.7 19.9

H 12.2 17.0 22.3 25.8 19.31L= 1.55 Mcal ENL/kgDM (i.e. 0.91 UFL/kgDM), H=1.65 Mcal ENL/kgDM (i.e. 0.97 UFL/kgDM).

Cannas et al., 1998

effects of excess concentrate may bepartially reversed by use of pH buffers(Hadjipanayiotou, 1988). During fulllactation, it is also observed ingroup-fed ewes that the level ofconcentrate, if moderatelyincreased, mainly affects the weightand body condition of lactatingewes, whereas bulk milk yield andcomposition are slightly or not sig-nificantly affected.

Consequences of group-feeding on nutritionalstrategies The dairy sheep allowances wereestablished for an individual ewe orgroup of ewes with similar perform-ance. They do not take into accountdifferences within the group of ewesto be fed (Bocquier and Caja, 1993;Bocquier et al., 1995). If possible,ewes should be allocated intohomogeneous groups according totheir characteristics (physiologicalstatus, prolificacy, stage of lacta-tion, milk yield or suckling litter sizeand body condition score).

When this allocation is not possibleand ewes’ performance varies widely,it is common to supply more feedthan the group’s average recom-mended allowances. For example, inLacaune dairy ewes, the main goal offeeding strategies is to give a diet thatis adequate for ewes that contributemost to milk production (those thatproduce about 10% more milk thanthe group average). Therefore, theenergy supply for such a group ofewes is calculated for individual milkproduction that is 10% above theactual mean milk yield.

The protein supply is generally calcu-lated for milk production that is 30%over the mean milk yield. This isbecause of marginal responses both inmilk yield and in protein content,although the excess dietary proteininduces protein waste, especially forthe low producing ewes of the group.Few comparative trials ofgroup-feeding strategies have beenconducted in dairy ewes.

Bocquier et al. (1995) conducted anexperiment to compare the effect oftwo strategies of group-feeding. Twosimilar groups of Lacaune dairy sheep(96 ewes each) were either fed alltogether (all levels of milk yield com-pounded) or after being separatedinto two subgroups according to milkyield (high and low). Total milk yieldand milk composition were identicalin both groups, but the “low-milkyield” subgroup showed a higherincrease in body weight and bodycondition score at the end of theexperiment.

Most of the beneficial effects ofgroup feeding are obtained whenconcentrates are saved, with dairyperformance generally maintainedor slightly improved.

On the other hand, at any given time,milk yield variability in a flock comesfrom lambing dispersion. Direct effectsof feeding on milk composition arehidden by the heterogeneity of per-formance. Studies conducted inFrance (Roquefort and Pyrenees) tomeasure the impact of within-flocklambing kinetic on annual milk pro-duction and its composition (Fraysseet al., 1996) factor this into indirectcomparisons of flock performances.

Effects of specific nutrients on the composition ofewes’ milk

Fat supplements Interest in adding fat supplements todairy sheep diets has increased in thepast years as a result of the availabilityof new preparations of fat as food forruminants and of favorable resultsobtained in dairy cows. Available infor-mation on dairy ewes is however,limited, and we especially focus oncalcium soaps of long chain fatty acids(CSFA). The effect of protected fat onewes’ milk production and composi-tion has been reviewed by Casals andCaja (1993) and Chilliard and Bocquier(1993). The main results regarding themilking period of dairy sheep aresummarized in table 3.

First references (Pérez Hernández etal., 1986) in suckling ewes tried toimprove lamb growth with contradic-tory results, but the most clearresponse was obtained in theimprovement of milk fat content indairy ewes. Lactational (suckling andmilking periods) effects of CSFAincluded in the concentrate fed toManchega dairy ewes grazing in semi-intensive conditions have beenreported mainly by Casals et al. (1989,1991, 1992ab, 1999), Cuartero et al.(1992), Gargouri et al. (1995), PérezAlba et al. (1997) and Osuna et al.(1998). The last authors compared theuse of oilseeds versus CSFA andLacaune versus Manchega dairy ewesin indoor conditions. McKusick et al.(1999) studied the effect of CSFA inconcentrate of East Friesian crossbredewes.

55


Although total milk yield was unaf-fected in all experiments, dietary CSFAsignificantly increased the milkcontents of fat and solids, in mostcases, and decreased milk proteincontent slightly in overall lactation.

Responses varied according to CSFAdose and lactation stage. Apparentefficiency of CSFA transfer to milk washigher in suckling than in milkingewes, and optimum intakes tomaximize milk fat production wereclose to 120 and 70 g CSFA/ewe/day,in suckling and milking respectively.The depressive effect of CSFA on milkprotein increased with time afterlambing, and optimum intake of CSFAthat maximized milk protein produc-tion were the same as for milk fat.

Milk casein also decreased with CSFAbut casein content as percentage ofmilk protein was unchanged in allcases. Fatty acids profile in milk andcheese changed with a strong increasein palmitic (C16:0) and oleic (C18:1)

acids and a decrease in the C6 to C14acids (Gargouri et al., 1995; Pérez Albaet al., 1997), but differences in fattyacids profile were not significant afterthe ripening of cheeses. Change in thefatty acids profile of milk depended onCSFA profile (Gargouri et al., 1995; PérezAlba et al., 1997). Special care must betaken in relation to changes in lipolysisrate or organoleptic characteristicsafter modification of fatty acid compo-sition in cheese.

More recently Osuna et al. (1998)studied the effects of feeding wholeoilseeds, to partially replace calciumsoaps of fatty acids, on dairy ewes’intake, milk production and composi-tion. In this aim, Manchega andLacaune dairy ewes were used duringthe mid-milking period to determinethe lactational effects of supplement-ing diets with fat coming from palmoil CSFA (5.5%) or from a mixture ofCSFA (2.5%) and whole cottonseed(11%) or CSFA (2.5%) and whole sun-flower seeds (4%). Diets were isoni-

trogenous (16%CP) and were offeredas a total mixed ration (71% forage:29% concentrate) where fat supple-ments were included. Ether extractincreased from 2.5% in control to 7%in fat supplemented. Results are sum-marized in table 4.

Due to the dietary fat, intake tendedto decrease, milk fat percentage andyield were increased, and caseincontent was reduced. Milk yield wasnot affected by treatments and nointeractions were found betweenbreed and fat supplementation, inspite of the respective differences(P<0.01) between Manchega andLacaune dairy ewes in milk yield (0.9and 1.6 l/d), and fat (8.8 and 7.2%) andprotein (6.2 and 5.6%) percentages,respectively in the control diet. A sig-nificant effect was detected on milkcasein as percentage of total proteinthat decreased as response to lipidsupplementation.

56


Table 3. Effects of calcium soap of long chain fatty acids on milk production of Manchega dairy sheep during milking.

Lactation period and reference Basal diet Lipid (g/d) Yield (%) Fat (g/l) Protein (g/l)

Casals et al. (1989, 1992a, 1999) Grazing 0 0.75 79 62“ 1601 0.78 97 56

Gr. + Protein suppl. 0 0.73 85 64“ 1601 0.69 100 59

Casals et al. (1991, 1992b) Grazing 0 0.74 74 60“ 401 0.83 82 59“ 801 0.70 94 60“ 1201 0.74 89 55“ 1601 0.71 94 56

Font et al. (unpublished) Grazing 0 0.51 99 65“ 721 0.53 105 61

Cuartero et al. (1992) Grazing 0 0.45 92 -“ 751 0.46 104 -

Gargouri et al. (1995) Grazing 02 0.94 82 67“ 721, 2 1.00 84 63

Gargouri (1997) Grazing 0 0.92 74 63“ 961 0.83 83 61

Perez Alba et al. (1997) Oat-vetch-hay 0 1.40 65 511663 1.56 68 49

1Calcium soaps of palm oil. 2Including 2% of animal fat and 3% of whole soybean seed in both concentrates. 3Calcium soaps of olive oil.

Effects of protein supplements Studies on the use of low degradableprotein supplements, protectedproteins or protected amino acids inmilk production of sheep are verylimited. Most of the references wereobtained from suckling ewes, alteringthe practical significance of data ofmilk composition. In addition, someresults are not significant or are con-tradictory.

In regard to low degradability proteinsupplements Robinson et al. (1979),Cowan et al. (1981), Penning andTreacher (1981), González et al. (1982),Hadjipanayiotou (1988, 1992) andPenning et al. (1988), and mostrecently Purroy and Jaime (1995),showed increases in milk yieldduring early lactation when theyincluded or substituted a degrad-able protein by fish meal (60-140g/d) as in lactating ewes. Milk com-position was, however unchangedin most cases and only significantlyimproved in the trials of Penning et al.(1988) and Purroy and Jaime (1995),when compared soybean and fishmeal in suckling ewes. These lastauthors reported significant increasesin milk protein (+2.9 g/l, +6.2%) butnot in milk yield, probably as a conse-quence of the reduction of under-nutrition (70-80% of energy require-

ments) applied in the experiment.Robinson et al. (1979) also found aslight increase (P<0.10) in milk proteinin ewes fed fish meal, when comparedwith those fed soybean or peanutprotein supplements. Effects of fishmeal are attributed to an increase inthe amount and profile of amino acidsabsorbed in the small intestine andthat are available for milk synthesis.

Protected proteins also gave interest-ing results, but in some cases are notsignificant or contradictory. Treatmentof protein supplements withformaldehyde must be done atoptimum doses (Caja et al., 1977).Ewes fed untreated soybeanscompared to those fed fish meal andformaldehyde-protected soybeans inChios dairy ewes were without signifi-cant effects on milk yield and milkcomposition (Hadjipanayiotou, 1992),even if milk fat and milk proteincontents were slightly higher in ewesfed formaldehyde-treated soybean.The use of formaldehyde-protectedsoybean in Chios dairy ewes innegative energy balance also did notaffect milk yield and composition(Hadjipanayiotou and Photiou, 1995).

Industrially protected soybean withlignosulphonate treatment is nowavailable for ruminants. Evaluation oftreated versus untreated soybeanswas done in Manchega dairy ewes fedwith poor quality forage at two levels

of supplementation with concentrate(Pérez et al., 1994, 1995). Values ofeffective degradability measured insacco for treated and untreatedsoybeans used in the experiment were0.30 and 0.56, respectively. Differencesbetween treatments were not signifi-cant, but a significant interaction(P<0.05) was observed in the milkyield comparisons between the levelof concentrate and degradability ofprotein. The highest values in milkyield were obtained with the highlevel of low degradability soybeansupplements. Milk composition wasunaffected by treatments.

More recently, protected amino acidshave been used in lactating ewes toincrease milk production duringsuckling (Lynch et al., 1991; Baldwin etal., 1993) or milking periods. (Bocquieret al., 1994). Lynch et al. (1991) studiedthe supplementation with Methionine(0.11%) and Lysine (0.28%) of two con-centrates for suckling ewes varying inits level of protein (10 and 16% crudeprotein). Results indicated a highermilk yield (+11%) in the ewes fed withthe high protein supplemented con-centrate, but the difference was notsignificant.

Milk protein was also unaffected byboth experimental treatments. Theinclusion of protected Methionine(0.2%) in the concentrate producedsmall (+2%) and non-significantincreases in milk yield and milkprotein as observed by Baldwin et al.(1993) in suckling Dorset ewes.

It has also been shown that theprotein content of milk can beincreased by adding 3 or 6 g/d of pro-tected Methionine at an early stage oflactation in Lacaune ewes (Bocquier etal., 1994) with ewes in positivenutrient balance (117-120% and 120-140% of energy and protein require-ments, respectively). The response toMethionine was higher when basaldiet was based on silage than on hay,

57


Table 4. Effects of feeding whole oilseeds and Calcium soaps of fatty acids onmilk production and composition of Manchega and Lacaune dairy ewes duringmid-milking.

Item Breed1 Control CSFA2 CSFA+WCS3 CSFA+SFS4

Milk yield (l/d) M 0.8 0.8 1.0 0.8

L 1.7 1.7 1.5 1.7

Milk fat (g/l) M 74 95 95 90

L 61 77 82 70

Milk protein (g/l) M 63 60 64 62

L 55 55 58 551M= Manchega, L= Lacaune; 2CSFA= Calcium soaps of fatty acids; 3WCS= Whole cotton seed;4SFS= Sunflower seed.

indicating that Methionine contentcould be the limiting amino acid inthis last diet. Milk yield and milk fatcontent were unaffected by the sup-plementation.

Conclusions and prospects The quality of milk can be defined inmany different ways according to itsfinal destination and/or to consumers’demands. In the future, however, on avery limited scale, some dairy ewesmay be bred for their milk properties,because of the feasibility of producingpharmaceuticals in the milk of trans-genic animals.

For the majority of breeders, theproblem is to produce milk on a largescale. For them, the changes in theway to produce milk evolved gradu-ally, based on scientific knowledgeand technical progress. A major stepwas increasing the productivity ofdairy ewes and controlling certainhealth aspects.

The second step was imposed bycheese manufacturers: Milk is nowgenerally paid for on its ability to betransformed into cheese; that is, its fatand protein content. Nowadays, thereare a wide variety of new objectivesemerging from social groups. Amongthem,“natural” production, animalwelfare, perennial land use and wastecontrol are often cited.

These objectives appear somehowconfusing because they may be con-tradictory or they may not be eco-nomically adapted to the presentcontext. This is the reason whybreeders defend their products andtheir income by well-defined new pro-duction rules determined collectivelyby breeders.

Hence, in France, it is illegal to treatdairy ewes with genetically engi-neered substances such as bST. Inaddition, in the Roquefort region of

France, the use of some constituentsof concentrates (ruminal-protected fator amino-acids) or animal by-productsis forbidden. The use of some sub-stances found in feed is also beingevaluated since these plants containparts of transgenic plants.

Sheep milk producers are locatedmainly in the Mediterranean area.Their breeding system relies on localsheep breeds that are well-adapted tothe environment and supported bylocal feed resources and cheesemaking and consuming traditions.

References Agus A. and F. Bocquier. 1995.

Contribution of body reserves tomilk production in underfed dairyewes. IV Symposium Internationalsur la Nutrition des Herbivores,Clermont-Ferrand, Septembre1995. Annales de Zootechnie, 44(Suppl.), 320.

Baldwin J.A., G.M. Horton, J.E. Wholt,D.D. Palatini and S.M. Emanuele.1993. Rumen protected methion-ine for lactation, wool and growthin sheep. Small Rumin. Res.,12: 125-132.

Barillet F and D. Boichard. 1987.Studies on dairy production ofmilked ewes. I. Estimates of geneticparameters for total milk composi-tion and yield. Gen. Sel. Evol.,19: 459-474.

Barillet F., S. Sanna, D. Boichard, J.M.Astruc, M. Carta and S. Casu. 1993.Genetic evaluation of the Lacaune,Manech and Sarda dairy sheepwith animal model. Proceed. 5thInt. Symp. on Machine Milking ofSmall Ruminants, Budapest, May14-20. Hungarian J. Anim. Prod., 1(Suppl.): 580-607.

Bencini R. and G. Pulina. 1997. Thequality of sheep milk: a review.Australian J. Exp. Agric., 37: 485-504.

Bocquier F. and G. Caja. 1993. Recentadvances on nutrition and feedingof dairy sheep, Proceed. 5th Int.Symp. on Machine Milking of SmallRuminants, Budapest, May 14-20.Hungarian J. Anim. Prod., 1 (Suppl.):580-607.

Bocquier F., D. Delmas and M. Thériez.1983. Alimentation de la brebislaitière: capacité d’ingestion etphénomènes de substitution chezla brebis Lacaune. Bull. Tech.CRZV-Theix, INRA, 52: 19-24.

Bocquier F., M. Thériez and A. Brelurut.1987 a. The voluntary hay intake ofewes during the first weeks of lac-tation. Anim. Prod., 44: 387-394.

Bocquier F., M. Thériez and A. Brelurut.1987 b. Recommandations alimen-taires pour les brebis en lactation.Bull. Tech. CRZV-Theix, INRA,70: 199-211.

Bocquier F., G. Delmas, B. Sloan, C.Vacaresse and E. Van Quackebeke.1994. Effet de la supplémentationen méthionine protégée sur la pro-duction et la composition du laitde brebis Lacaune. Renc. Rech.Ruminants, 1, 101 - 104

Bocquier F., P. Guillouet and F. Barillet.1995. Alimentation hivernale desbrebis laitières : intérêt de la miseen lots. INRA Prod. Anim., 8: 19-28.

Bocquier F., J.P. Guitard, C. Vacaresse, E.Van Quackebeke, G. Delmas, P.Guillouet, G. Lagriffoul, E. Morinand J.M. Arranz. 1997. Estimationde la capacité d’ingestion et desphénomènes de substitutionfourrage/concentré chez les brebisLacaune conduites en lots : compi-lation des données obtenues surdes rations à base d’ensilage. Renc.Rech. Ruminants, 4: 75-78.

58


Bocquier F., M.R. Aurel, F. Barillet, M.Jacquin, G. Lagriffoul and C. Marie.1999. Effects of partial milkingduring the suckling period on milkproduction of Lacaune dairy ewes.In: Milking and milk production ofdairy sheep and goats. F. Barilletand N.P. Zervas (Eds), EAAP Publ.No. 95 p. 357-262. WageningenPers, Wageningen, TheNetherlands.

Caballero R., J. Rióperez, E. Fernández,M. Arauzo and P.J. Hernaíz. 1992.Performance of Manchega ewesgrazing cereal stubbles and culti-vated pastures. Small Rumin. Res.,7 : 315-329.

Caja G. 1994. Valoración de las necesi-dades nutritivas y manejo de la ali-mentación de ovejas lecheras deraza Manchega. In: Ganado ovino:Raza Manchega. L. Gallego, A.Torres & G. Caja (Eds.). Mundi-Prensa, Madrid. p. 137-159.

Caja G. and X. Such. 1991. Situación dela producción de leche de oveja enel Mundo y clasificación de losprincipales sistemas de producciónde ovino lechero. Ovis, 14: 11-27.

Caja G., J.F. Gálvez, A. Argamenteríaand J. Ciria. 1977. Inhibition ofruminal deamination in vitro byformaldehyde treatment of sun-flower-seed, soya bean and fishmeals: response curves to protec-tive treatment. Anim. Feed. Sci. Tech.,2: 267-275.

Caja G., F. Bocquier, L. Pérez-Oguez andL. Oregui. 1997. Mesure de lacapacité d’ingestion durant lapériode de traite des brebislaitières de races méditer-ranéennes. Renc. Rech. Ruminants,4 : 84.

Caja G. and F. Bocquier. 1998. Effects ofnutrition on ewe’s milk quality.Cooperative FAO-CIHEAM Networkon sheep and goats, NutritionSubnetwork, Grignon, France, 3-5September. 1-16.

Cannas A., A. Pes, R. Mancuso, B. Vodretand A. Nudda. 1998. Effect ofdietary energy and protein con-centration on the concentration ofmilk urea nitrogen in dairy ewes.J. Dairy Sci., 81: 499-508.

Casals R. and G. Caja. 1993. Interés delempleo de los suplementos lipídi-cos en la alimentación de ovino ycaprino en zonas áridas. In:Nutrición de rumiantes en zonasáridas y de montaña y su relacióncon la conservación del medionatural. Congresos y Jornadas29/93. Junta de Andalucía,Consejería de Agricultura y Pesca.p. 173-193.

Casals R., G. Caja, X. Such, C. Torre. 1989.Efectos de la incorporación degrasa y proteína no degradablesen el concentrado de lactación deovejas de ordeño. ITEA Prod. Anim.,9 (Suppl.): 107-109.

Casals R., G. Caja, D. Guillou, C. Torreand X. Such. 1991. Variación de lacomposición de la leche de ovejasManchegas según la dosis delípidos protegidos. ITEA Prod.Anim.,11 (Suppl.): 331-333.

Casals R., G. Caja, X. Such, C. Torre andX. Fàbregas. 1992a. Lactationalevaluation of effects of calciumsoap and undegraded intakeprotein on dairy ewes. J. Dairy Sci.,75: 174 (Abstr. P 87).

Casals R., G. Caja, D. Guillou, C. Torreand X. Such. 1992b. Influence ofdietary levels of calcium soaps oflong chain fatty acids on lacta-tional performance of dairy ewes.J. Dairy Sci., 75: 174 (Abstr. P 88).

Casals R., G. Caja, X. Such, C. Torre andS. Calsamiglia. 1999. Lactationaleffects of calcium soap and unde-graded intake protein on dairyewes. J. Dairy Res., 66 : 177-191.

Casu S., R. Carta, G. Ruda. 1983.Morphologie de la mamelle etaptitude à la traite méchanique dela brebis Sarde. 3rd Int. Symp. onMachine Milking of SmallRuminants. Ed. Sever Cuesta,Valladolid. p. 592-603.

Chilliard Y. and F. Bocquier. 1993.Effects of fat supplementation onmilk yield and composition in dairygoats and ewes. Proceed. 5th Int.Symp.“La qualita nelle produzinidei piccoli ruminanti”. Camera diCommercio Industria ArtigianoAgricultura di Varese, 3 dicembre1993, Varese. p. 61-78.

Coulon J.B. and B. Rémond. 1991.Variations in milk output and milkprotein content in response to thelevel of energy supply to the dairycow: a review. Livest. Prod. Sci.,29: 31-47.

Cowan R.T., J.J. Robinson, I. Mc Hattieand K. Pennie. 1981. Effects ofprotein concentration in the dieton milk yield change in body com-position and the efficiency of uti-lization of body tissue for milk pro-duction in ewes. Anim. Prod.,33: 111-120.

DePeters E.J. and J.P. Cant. 1992.Nutritional factors influencing thenitrogen composition of bovinemilk: a review. J. Dairy Sci., 75:2043-2070.

Eyal E. and Y. Folman. 1978. The nutri-tion of dairy sheep in Israel. In: Milkproduction in the ewe. J.G.Boyazoglu & T.T. Treacher (Ed.).EAAP Publication, 23: 84-93.

59


Fernández N., J. Arranz, G. Caja,A.Torres and L. Gallego. 1983.Aptitud al ordeño mecánico deovejas de raza Manchega: II.Producción de leche, reparto defracciones y cinética de emisión deleche. 3rd Int. Symp. on MachineMilking of Small Ruminants. Ed.Sever Cuesta, Valladolid. p. 667-686.

Flamant J.C. and P. Morand-Fehr. 1982.Milk production in sheep andgoats. In: Sheep and goat produc-tion, I.E. Coop (Ed.), World AnimalScience, C 1. Elsevier SciencePublishing Company, Amsterdam,p. 275-295.

Fraysse J., G. Lagriffoul, F. Bocquier andF. Barillet. 1996. Brebis laitières:impact de la structure du troupeauet autres facteurs d’élevage sur lacomposition chimique du lait livré.INRA Prod. Anim., 9: 201-210.

Fuertes J.A., C. Gonzalo, J.A. Carriedoand F. San Primitivo. 1998.Parameters of test day milk yieldand milk components for dairyewes. J. Dairy Sci., 81: 1300-1307.

Gallego L., M.P. Molina, A.Torres and G.Caja. 1983. Evolución de lacantidad y composición de la lechede ovejas de raza Manchega desdeel parto. 3rd Int. Symp. on MachineMilking of Small Ruminants. Ed.Sever Cuesta, Valladolid. p. 285-297.

Gargouri A., G. Caja, X. Such, A. Ferret,R. Casals and S. Peris. 1993.Evaluation d’un systéme de traiteet allaitement simultanés chez lesbrebis laitières de race Manchega.5th Int. Symp. on Machine Milkingof Small Ruminants, Budapest.Hungarian J. Anim. Prod., 1 (Suppl.):484-499.

Gargouri A., G. Caja, X. Such, R. Casalsand A. Ferret. 1995. Efectos de lautilización de lípidos protegidos enla alimentación de ovejas deordeño en el sistema a medialeche. ITEA Prod. Anim., 16 (Suppl.):720-722.

Gargouri A., G. Caja, C. Gafo, X. Suchand A. Ferret. 1997. Modificacióndel perfil de ácidos grasos de laleche de oveja mediante el empleode jabones cálcicos de ácidosgrasos de cadena larga. ITEA Prod.Anim., 18 (Suppl.): 694-696.

González J.S., J.J. Robinson, I. McHattieand C. Fraser. 1982. The effect inewe of source and level of dietaryprotein on milk yield and the rela-tionship between the intestinalsupply of non ammonia nitrogenand the production of milkprotein. Anim. Prod., 34: 31-40.

González J.S., J.J. Robinson and I.McHattie. 1984. The effect of levelof feeding on the response of lac-tating ewes to dietary supple-ments of fish meal. Anim. Prod.,40: 39-45.

Hadjipanayiotou M. 1988. Feedingsystem largely based on concen-trates. I. Sheep. World Rev. Anim.Prod., 24: 75-85.

Hadjipanayiotou M.1992. Effect ofprotein source and formaldehydetreatment on lactation perform-ance of Chios ewes and Damascusgoats. Small. Rumin. Res., 8: 185-197.

Hadjipanayiotou M. and A. Photiou.1995. Effects of level of inclusionand formaldehyde treatment ofsoybean meal on the performanceof lactating Chios ewes in negativeenergy balance. Livest. Prod. Sci.,41: 207-215.

Labussière J., B. Bennemederbel, J.F.Combaud and F. Chevalière. 1983.Description des principauxparamètres caractérisant la pro-duction laitière, la morphologiemammaire et la cinétique d’émi-sion du lait de la brebis Lacaunetraite une ou deux fois par jouravec ou sans égouttages. 3rd Int.Symp. on Machine Milking of SmallRuminants. Sever Cuesta (Ed.),Valladolid. p. 625-652.

Lynch G.P., T.H. Elsasser, C.Jr. Jackson,T.S. Rumsey and M.J. Camp. 1991.Nitrogen metabolism of lactatingewes fed rumen-protected methio-nine and lysine. J. Dairy Sci., 74:2268-2276.

Marie C., M. Jacquin, M.R. Aurel, F. Paille,D. Porte, P. Autran and F. Barillet.1999. Déterminisme génétique dela cinétique d’émission du laitselon le potentiel laitier en raceovine de Lacaune et relationsphénotypiques avec la morpholo-gie de la mamelle. In : Milking andmilk production of dairy sheep andgoats. EAAP Publ. F. Barillet and N.P.Zervas (Eds.) No. 95 p. 381-387.Wageningen Pers, Wageningen,The Netherlands.

McKusick B.C., Y.M. Berger and D.L.Thomas. 1999. Effects of threeweaning and rearing systems oncommercial milk production andlamb growth. Proc. 47 th AnnualSponner Sheep Day, August 28,1999 : 33-48.

McKusick B.C., Y.M. Berger and D.L.Thomas. 1999. Rumen-Protectedbypass fat for dairy ewe commer-cial milk production. Proc. 5thGreat Lakes Dairy SheepSymposium. Nov 4-6 1999.Brattleboro, Vermont, USA.

60


Molina M.P. and L. Gallego. 1994.Composición de la leche: Factoresde variación. In: Ganado ovino:Raza Manchega. L. Gallego, A.Torres & G. Caja (Ed.). Mundi-Prensa, Madrid. p. 191-208.

Osuna D.R., R. Casals, G. Caja and S.Peris. 1998. Effects of feedingwhole oilseeds to partially replacecalcium soaps of fatty acids ondairy ewes intake and milk produc-tion and composition. J. Dairy Sci.,81 (Suppl. 1): 302 (Abstr. 1179).

Pérez Alba L.M., S. De Souza, M. Pérez,A. Martínez and G. Fernández.1997. Calcium soaps of olive fattyacids in the diets of Manchegadairy ewes: Effects on digestibilityand production. J. Dairy Sci., 80:3316-3324.

Pérez-Oguez L., X. Such, G. Caja, A.Ferret and R. Casals. 1994. Variaciónde la respuesta a la suple-mentación con proteína nodegradable en ovejas lecherassegún el nivel de concentrado. XIXJornadas de la SEOC, Burgos. Juntade Casilla y León. Consejería deAgricultura y Ganadería.p. 249-254.

Pérez-Oguez L., G. Caja, A. Ferret and C.Gafo. 1995. Efecto de la suple-mentación con proteína nodegradable en ovejas de razaManchega: 2. Ingestión de forraje.ITEA Prod. Anim., 16 (Suppl.): 12-14.

Pellegrini O., F. Remeuf, M. Rivemale, F.Barillet. 1997. Renneting propertiesof milk from individual ewes: influ-ence of genetic and non-geneticvariables, and the relationship withphysicochemical characteristics.J. Dairy Res., 64: 355-366.

Penning P.D and T.T. Treacher. 1981.Effect of protein supplements onperformance of ewes offered cutfresh ryegrass. Anim. Prod., 23: 374-775 (Abstr.).

Penning P.D., R.J. Orr and T.T. Treacher.1988. Responses of lactating ewesoffered fresh herbage indoors andwhen grazing, to supplements con-taining differing protein concen-trations. Anim. Prod., 46: 403-415.

Robinson J.J., C. Fraser, J.C. Gill and I.McHattie. 1974. The effect ofdietary crude protein concentra-tion and time of weaning on milkproduction and body weightchange in the ewe. Anim. Prod.,19: 331-339.

Robinson J.J., I. Mc Hattie, J.F. Calderon-Cortes and J.L. Thompson. 1979.Further studies on the response oflactating ewes to dietary protein.Anim. Prod., 29: 257-269.

Sheath G.W., M. Thériez and G. Caja.1995. Grassland farm systems forsheep production. In: Recent devel-opments in the Nutrition ofHerbivores. M. Journet, E. Grenet,M-H. Farce, M. Thériez & C.Demarquilly (Eds.). Proceed. 4th Int.Symp. on Nutrition of Herbivores.Clermont-Ferrand, INRA Editions,Paris. p. 527-550.

Sutton J.D. and S.V. Morant. 1989. Areview of the potential of nutritionto modify milk fat and protein.Livest. Prod. Sci., 23: 219-237.

Treacher T.T. 1971. Effects of nutritionin pregnancy and in lactation onsubsequent milk yield in ewes.Anim. Prod., 12: 23-36.

Treacher T.T. 1983. Nutrient require-ments for lactation in the ewe. In:Sheep production. W. Haresign(Ed.). Butterworths. London. p.133-153.

Treacher T.T. 1989. Nutrition of thedairy ewe. In: North AmericanDairy Sheep Symposium. W.J.Boyland (Ed.). University ofMinnesota, St. Paul. p. 45-55.

61


62


The mammary gland is an exocrineepithelial gland, exclusive tomammal species (animals able to

produce milk). It is quantitatively andqualitatively adapted to the growthrequirements and behavior of eachspecies. It shows histological similaritiesto other epithelial glands such as thesalivary and sweat glands. Milk secre-tion is described as the activity of acellular factory (the lactocyte), whichtransforms itself into the product (themilk).The entire process is controlledby integrated neuro-endocrine andautocrine systems. It mainly developsduring pregnancy and early lactation,and regresses very quickly after dry-off.

The anatomy and morphology of thesheep udder has been well known formany years (Turner, 1952; Barone,1978). Some examples of curiousselection on udder morphology havebeen studied (such as increasing pro-lificacy and number of teats). Earlyworks on the relationship betweenudder characteristics and milking per-formance in dairy ewes were carriedout in the ‘70s and early ‘80s (Sagi andMorag, 1974; Jatsh and Sagi, 1978;Gootwine et al., 1980; Labussière et al.,1981) resulting from efforts to adaptthe ewe to the milking machine.

With this aim, an internationalprotocol (M4 FAO project) was initi-ated to evaluate the dairy sheepudder in Mediterranean breeds(Labussière, 1983, 1988). Based on thisstandardized protocol, the udders ofmany dairy sheep breeds were sys-tematically studied in relation tomachine milking in the 3rd

International Symposium on MachineMilking of Small Ruminants held inSpain (Casu et al., 1983; Fernández etal., 1983a; Gallego et al., 1983a;Hatziminaoglou et al., 1983; Labussièreet al., 1983; Pérez et al., 1983; Purroyand Martín, 1983) and following sym-posiums (Arranz et al., 1989; Kukovicsand Nagy, 1989; Rovai et al., 1999) inEurope, and also in America(Fernández et al., 1999; McKusick et al.,1999).

Interest in the dairy sheep udder hasincreased in the last few years inwhich anatomy has been explored indepth (Ruberte et al., 1994b; Caja etal., 1999; Carretero et al., 1999), linearevaluation of udder traits has beenproposed (de la Fuente et al., 1996;1999; Carta et al., 1999) and thegenetic parameters evaluated(Gootwine et al., 1980; Mavrogenis etal., 1988; Fernández et al., 1995; 1997;Carta et al., 1999). Moreover, given thenegative effects observed in uddermorphology as a result of the increasein milk yield, main udder traits of dif-ferent breeds (Rovai et al., 1999) or ofgenetically isolated lines of the samebreed (Marie et al., 1999) are undercomparison. This chapter describesthe unique characteristics of the dairysheep udder and summarizes thecurrent implications of udder mor-phology on machine-milking.

63

Udder morphologyand machine milking

Chapter 6

Interest in the

dairy sheep udder

has increased in

the last few years.

Structure anddevelopment of themammary gland

Origin and growthThe mammary gland is formed by twomain structures: the parenchyma andthe stroma. The partitioning betweenboth structures defines the anatomi-cal and functional characteristics ofeach mammary gland. Theparenchyma is the secretory part ofthe gland. It is made up of tubulo-alveolar epithelial tissue, coming fromthe ectoderm layer of the embryo, andit consists of the tubular (ductal) andalveolar systems. The stroma is formedby other complementary tissues ofmesodermic origin such as blood andlymph vessels, and adipose, connec-tive and nervous tissues. Both struc-tures develop very early from theventral skin of the embryo andhalfway through the pregnancy a totalof eight pairs of isolated mammarybuds are present in all mammalembryos (Delouis and Richard, 1991).

Well-developed mammary buds areclearly observed in 2 cm long sheepembryos (near 30 days old) asreported by Turner (1952). An impor-tant differentiation in mammogenesisoccurs at this stage in the ruminants:The mammary parenchyma developscisterns. An involution process accord-ing to the species then begins, andonly the seventh mammary pairlocated in an inguinal position ismaintained in sheep (Delouis andRichard, 1991). Occasionally the sixthpair can also be maintained, produc-ing supernumerary teats. Anotherunique characteristic of sheep is thepresence of skin inguinal bags in thegroin behind each teat. These containsebaceous glands which produce ayellow and fatty secretion useful forthe care of the udder skin (Ruberte etal., 1994b).

At birth, the sheep udder showsclearly differentiated cisterns (Sinuslactiferus)1 and teats (Papilla mammae)and very incipient development of theductal system, with few primary ductssurrounded by numerous stroma-forming cells. After birth the uddergrows at the same rate as the body(isometric growth) until puberty, withproliferation and branching of the sec-ondary ductal system.

Puberty in most species is thequickest period of growth for ductsand stroma of the mammary gland(positive allometric growth), as aresult of the action of sexualhormones. Nevertheless, the futuremilk capacity of the udder can beimpaired at this stage by an exces-sive growth of the stroma (mainlyadipose and connective tissues) incomparison to the parenchyma(tubulo-alveolar epithelium).

This critical phase occurs earlier insheep than in cattle, with differencesbetween breeds. Thus, theparenchyma growth ends in sheepbefore puberty and, as a consequence,mammogenesis in sheep will beaffected by nutrition during and afterthe positive allometric growth phase(Bocquier and Guillouet, 1990).

The critical period for mammogene-sis is from 2 to 4 months old. Earlyonset of puberty will bring forwardthe decrease in mammary develop-ment. According to Johnson and Hart(1985) and McCann et al. (1989), arelative low growth rate (50% of highrate) from weaning (week 4) to theend of rearing period (week 20) willincrease the parenchyma growth andthe milk production in the first lacta-tion in non dairy ewe-lambs (figure 1).No negative effects were observed atthe beginning of puberty.Nevertheless a low growth rate beforeweaning also negatively affects mam-mogenesis (McCann et al., 1989).

Unfortunately there is no detailedinformation available on dairy sheep,but Bocquier and Guillouet (1990)reported that the restriction of con-centrate in Lacaune ewe-lambs,after they reach approximately 28to 30 kg, increases milk yield by10% in the first lactation.

64


2500

2000

1500

1000

growth rate (g/d)

milk

yiel

d (g

/d)

100 150 200 250 300

Figure 1. Effect of growth rate beforepuberty in ewe-lambs on milk yield atthe first lactation (McCann et al., 1989).

During the first and subsequent preg-nancies, the parenchyma shows anallometric growth where the placentaplays an important role. A specificovine chorionic somatotropinhormone (oCS), dependent on prolifi-cacy, can be obtained from the sheepplacenta after day 60 of pregnancy(Martal and Chene, 1993).Mammogenesis starts clearly in sheepbetween day 95 and 100 of preg-nancy, with detection of lactose (startof lactogenesis) after day 100 (Martaland Chene, 1993).

The presence of secretory lobes withalveolus in the extremes of the ductscan be observed in the second half ofpregnancy. Delouis and Richard (1991)estimate a change from 10–90% in therelative weight of the parenchymaduring pregnancy, where the lobulo-alveolar development of epithelialcells takes the place of the adiposetissue. The inverse process occursduring the dry period, with a completedisappearance of the alveoli in theewe after 3 to 4 weeks, and its replace-ment by adipocytes (Hurley, 1989).Moreover during the involutionprocess the mammary gland isinvaded by macrophages and lympho-cytes, the latter being essential for theproduction of immunoglobulins in thesynthesis of colostrum in the nextpregnancy.

Internal structure of themammary glandThe study of the internal structure ofthe ewe udder was first carried out invitro by anatomical dissection in deadanimals (Turner et al., 1952; Barone,1978; Tenev and Rusev, 1989; Ruberteet al., 1994b). This methodologyreveals the presence of two independ-ent mammary glands under a uniqueskin bag, each of them wrapped by abag of fibroelastic connective tissue(Apparatus suspensorius mammarum)and separated by a clearly definedand intermediate wall of connectivetissue (Ligamentum suspensoris uber).The strength of this ligamentnormally produces the presence ofan intermmamary groove (Sulcusintermammarius) between eachgland. This ligament plays an impor-tant role in the support of theudder, maintaining the uddertightly attached to the ventralabdominal wall. Each half uddershows internally a typical tubulo-alveolar structure with a big cistern(Sinus lactiferus) divided in two parts:glandular cistern (S. l. pars glandularis)and teat cistern (S. l. pars papillaris).Both cisterns are separated by amuscular sphincter of smoothmuscular fibers, traditionally known asthe cricoid fold, which plays an impor-tant role in milk drainage. It also helps

to keep the teat and gland morphol-ogy divided during machine milkingto avoid the appearance of clusterclimbing. The cricoid sphincter isnormally missing in goats and it is notvery effective in the conic teat udders,which are not favorable for machinemilking. Size and form of the glandcistern vary according to the breedand milking ability of the sheep, beinggreater and plurilocular in highyielding ewes (figure 2). Anothersphincter with smooth muscular fibersis present around the streak canal(Ductus papilaris) in the distal part ofthe teat, connected to the exterior bya unique orifice (Ostium papilare).

The last mammary gland structures inthe parenchyma are the secretorylobes, consisting of very branchedintralobular ducts and alveoli. Thealveolus is the secretory unit of themammary gland and consists of a bagof a unique layer of specialized cubicepithelial cells (the lactocytes) with aninside cavity (the lumen) in which themilk is stored after secretion.

65

U D D E R M O R P H O L O G Y & M A C H I N E M I L K I N G C H A P T E R 6

Figure 2. Comparison of cistern sizein dairy (right) and non-dairy (left)sheep udders in early lactation.

The mammary gland stores the milkextracellulary and this storage can beexplained using a model of twoanatomical compartments: ‘Alveolarmilk’ (secreted milk stored withinthe lumen of alveolar tissue) and‘Cisternal milk’ (milk drained fromthe alveoli and stored within thelarge ducts and the gland and teatcisterns). Short-term autocrine inhibi-tion of milk secretion in the mammarygland has been related to cisternalsize, the large-cisterned animalsbeing in general more efficient pro-ducers of milk and more tolerant tolong milking intervals and simpli-fied milking routines (Wilde et al.,1996).

Partitioning between cisternal andalveolar milk is usually determined bydrainage of cisternal milk, by using ateat cannula, and by milking alveolarmilk after an oxytocin injection(Ruberte et al., 1994a; Wilde et al.,1996). Nevertheless cisternal milkvolume can be increased in somebreeds by spontaneous liberation ofendogenous oxytocin as a conse-quence of milking conditionedbehavior or as a result of teat manipu-lation. This effect has been shown inLacaune but not in Manchega ewes(table 1) by Rovai et al. (2000), inaccordance with the milking ability ofeach breed, and the use of an oxytocinreceptors blocking agent for cisternaland alveolar milk determination is rec-ommended (Knight et al., 1994; Wildeet al., 1996). Values of cisternal milk insheep vary from 25 to 70% accordingto the breed (Caja et al., 1999; Rovai etal., 2000) but they are greater than50% in most dairy sheep breeds.Cisternal : alveolar ratio increases withlactation stage and parity in dairycows (Wilde et al., 1996), but no refer-ences are available on sheep.

The results in table 1 also indicate thatcistern size plays an important role inthe milk yield of the ewe. Thus, despitethe differences in milk production(100%) at the same stage of lactation(90 d), alveolar milk was very similar inthe two breeds, the difference beingonly 10% greater in Lacaune. On thecontrary, the difference in true cister-nal milk was 102% according to thedifference observed in yield. Thisseems to indicate that cisternal sizeis a direct limiting factor for milksecretion in dairy sheep and itsimportance is greater than theamount of secretory tissue (Rovai etal., 2000). A ratio of approximately 7.5between daily milk yield and cisternalmilk was obtained in both breeds.

In vitro anatomical studies are in somecases limited because the organ losestone and becomes flaccid, which isimportant in the case of the udder. Anin vivo image of the mammary glandstructures can be obtained by thenon-invasive technique of ultrasoundscans.

A method for sheep mammographywas proposed by Ruberte et al.(1994a) and its validity tested by Cajaet al. (1999). The method has beenused to show the milk ejection reflexin sheep (Caja and Such, 1999), tomeasure the cistern size and tocompare the internal morphology ofthe udder in different breeds of dairysheep (Rovai et al., 2000). This methoddemonstrates that the gland cistern isflat when empty after milking as aconsequence of the pressure of themammary suspensor system (figure 3).The method can also be used toestimate the distribution and move-ments of milk between the uddercompartments and for non-invasivedynamic studies on cisternal milk.

66


Table 1. Cisternal and alveolar distribution in dairy sheep at mid-lactationaccording to the breed and the method used.

Item Manchega Lacaune SEMControl Atosiban1 Control Atosiban

Number of ewes 10 10 -

Milk yield (l/d) 0.935b 1.871a 0.313

Alveolar milk (ml) 86.2b 104.0ª 88.8b 114.9ª 0.5

Cisternal Milk (ml) 121.8c 118.3c 299.2a 239.2b 1.2Area (cm2) 12.38b 13.06b 24.02a 23.25a 0.98

Cisternal: Alveolar (%) 59 :41 53 : 47 77 : 23 68 : 32 -1Oxytocin receptors blocking agent injected in jugulara, b, cDifferent letters in the same line indicate significant differences at P<0.05

Rovai et al., 2000

A different approach in the study ofthe cisterns and the lobulo-alveolarsystem in the mammary gland can beobtained by using the corrosionplastic cast method, normally used forthe anatomical study of soft tissues(Ditrich and Splechtna, 1989; Ruberteet al., 1994b; Carretero et al., 1999).This method consists of obtaining acast of the canalicular system of themammary gland in euthanizedanimals, after draining the milk fromthe udder. An epoxy resin is immedi-ately injected through the teat sphinc-ter to obtain the complete repletion ofall the tubulo-alveolar system(cisterns, ducts and alveoli). Udders areremoved after hardening of the epoxyresin and the organic tissue corrodedusing a KOH solution. The resulting

casts (figure 4) are used for macro-scopic and microscopic studies, whereanatomical details can be studied indepth.

The study of the ultrastructure of themammary gland is normally done byusing the scanning electronmicroscopy method used by Williamsand Daniel (1983), Caruolo (1980) andCarretero et al. (1999), which showedclear images of mammary alveoli insheep (figure 4). The method uses thecorrosion casts previously described,after conditioning for scanningelectron microscopy. Different degreesof development of the canalicularsystem are identified in theparenchyma of sheep mammaryglands during lactation.

Tubulogenic structures found byCarretero et al. (1999) in dairy sheepvaried in frequency and type accord-ing to stage of lactation but in allcases the sheep casts had the typicalappearance of a bunch of grapes asdescribed in the bibliography (figure5.1). All the alveoli seen in this workshowed a unique and independentlobular duct without fusion betweenadjacent alveoli. The development ofthe mammary gland ducts showed asimilar morphology to that previouslyreported in the development of thevascular system (mesodermic origin)in embryos of different species(Ditrich and Splechtna, 1989; Carreteroet al. 1995). Structures indicating anextensive proliferation of the canalicu-lar system were found by Carretero et

67


Figure 3. Scans of dairy sheep udders showing the gland and teat cisterns full of milk before milking (left) and empty after milking (right).

Ultrastructure of the mammary gland and changes during lactation in dairy ewes.

Figure 4. Cast of a dairy sheep udderobtained by the epoxy injection and corro-sion method (left) and detail of the ductalsystem with ducts and alveoli (right).

al. (1999) in Manchega and Lacaunedairy ewes between weeks 1(suckling) and 5 (start of milking) oflactation at the same time that a largenumber of alveolar sprouts wereobserved (figure 5.2).

Both dairy breeds showed the samemammary structures and followed thesame pattern of development duringlactation despite the differences inreported milk yield. The developmentof the mammary canalicular systemafter parturition has already beendescribed in primiparous sheep(Brooker, 1984) and goat (Knight andWilde, 1993), but Carretero et al. (1999)

used ewes that were in the third lacta-tion. The finding of cellular prolifera-tion at the time of maximum milk pro-duction, is also in accordance with theobservations of Knight and Wilde(1993). Franke and Keenan (1979)demonstrated that both situationscould be found even in a lactocyte.

The identification of concave valve-like structures as previously describedby Caruolo (1980) was also commonto all stages studied by Carretero et al.(1999). Nevertheless, these valve-likestructures do not appear in ourstudies at the level of the opening ofalveolus into the lobular duct, but at

the point where a lobular duct drainsinto a larger duct (figure 5.3).

This may indicate the existence of akind of cellular reinforcement toprevent milk leakage when the alveoliand ducts are full of milk.

At week 1 of lactation Carretero et al.(1999) reported the occurrence of‘intussusceptive growth’ at the level ofthe lobular ducts in sheep during thesuckling period. This new type ofgrowth leads to an increase in thenumber of tubules from preexistingones (figure 5.4). Intussusceptivegrowth has only been reported insome areas of the vascular system (i.e.,

68


Figure 5. Scanning electron microscopy images from epoxy casts obtained in ewes mammary glands at different stagesof lactation: 1) Lobular duct (L) and alveoli on wk 13 (bar = 40 mm); 2) Alveoli on wk 1 (bar = 0.2 mm); 3) Valve-like struc-ture in a duct (bar = 30 µm); 4) Intussusceptive growth in a lobular duct (L) on wk 1 (bar = 28 µm); 5) Alveolar sprouts ionwk 5 (bar = 60 µm); 6) Alveolus grooves on wk 13 (bar = 30 µm); 7) Collapsed alveolus on week 13 (bar = 20 µm).

lung vessels and widely duringembryo development) and it is char-acterized by the formation of pillars ofendothelial tissue in the lumen of theduct (Burri and Tarek, 1990; Patan etal., 1992). The pillars appear as trans-versal holes in the plastic casts. Thisconvergence in the model of develop-ment of two different cellular lines,mammary gland ductal and vascularsystem cells, is produced despite theirdifferent embryonic origin (ectodermand mesoderm origin, respectively).

At week 5 after parturition, correspon-ding to the first week of the milkingperiod after weaning, duct develop-ment by intussusceptive growthseemed to be complete and onlychanges in mammary alveoli wereobserved. In this way, fully developedalveoli together with others in the firstphases of development wereobserved at the same time and in thesame lobular duct. Nevertheless, struc-tures like angiogenic buds were fre-quently identified in the tubules atthis time. These buds appeared assemispherical enlargements that growfrom the ducts and become almostspherical by the narrowing of theirconnection with the duct. Then, thebud surface loses its smoothness anddevelops small sockets giving a golf-ball-like image identified as alveoli.Moreover frequently developingalveoli with sprouting shapes wereobserved on the surface of lobularducts at this lactation stage (figure5.5) as previously described in themammary gland of ewes by Alvarez-Morujo and Alvarez-Morujo (1982).Alveolar sprouts described byCarretero et al. (1999) are not compa-rable to the sprouts found in theextremes of lobular ducts producingthe longitudinal growth of lobularducts during puberty by Williams andDaniel (1983).

In mid-lactation (week 13), the mam-mogenic structures were notobserved by Carretero et al. (1999) inthe canalicular system of themammary gland, and the mostrelevant observation was the alveolimorphology. They were unilocular,spherical and with their externalsurface smoothed or grooved (figure5.6). The images are in accordancewith those obtained by Caruolo (1980)and suggest that grooves may be aconsequence of capillary vessels sur-rounding the alveolus. We alsoobserved, in a few cases, some flat-tened alveoli (figure 5.7) that may beconsidered collapsed (empty), but nofused alveoli were found.

External morphology of themammary gland

Udder typologyThe first practical utilization of uddermorphology on dairy sheep was madeby using tables of udder typology inAwassi and Assaf (Sagi and Morag,1974; Jatsch and Sagi, 1978), Sarda(Casu et al., 1983) and Manchega ewes(Gallego et al., 1983a, 1985), all ofthem based on four main udder types.A comparative table of these typolo-gies can be observed in Gallego et al.(1985). These typologies were lateradapted to the Latxa breed (Arranz etal., 1989) and Hungarian Merino andPleven (Kukovics and Nagy, 1989). Thetypology used in Sarda was evaluatedin field conditions (Casu et al., 1989)and extended to seven udder typesmainly based on teat position andcistern size (Carta et al., 1999) with theaim of improving the small discrimi-nating capacity of the previoustypologies. Nevertheless, the evalua-tion of sheep udders by morpholog-ical types is easy, quick and repeat-able with trained operators (Carta et

al., 1999; De la Fuente et al., 1999).Typology is recommended as a usefultool for the screening of animals, inthe standardization of machinemilking groups or in the choice ofewes at the constitution or acquisitionof a flock, and for culling of breedinganimals (Gallego et al., 1985; Carta etal., 1999).

69


A well shaped and healthy

dairy sheep udder for machine

milking should have:

■ Great volume, with globose

shape and clearly defined

teats.

■ Soft and elastic tissues, with

palpable gland cisterns

inside.

■ Moderate height, no sur-

passing the hock.

■ Marked intermammary

ligament.

■ Teats of medium size

(length and width),

implanted near to vertical.

Udder measurementsThe use of objective measurements tocharacterize dairy sheep udders andstudy their relationship with milk yieldor other productive traits has beenundertaken by different authors sincethe development of machine milking.The continuous nature of the meas-urements increases the discriminatingcapacity of each variable and the sig-nificance of correlation with the pro-ductive traits. The methodology gen-erally used corresponds to the stan-dardized protocol of Labussière (1983)with small variations incorporated insome cases (Gallego et al., 1983a;Fernández et al., 1983, 1995). Therepeatability of udder measurementsmade according to this methodologyis low for udder dimensions (r=0.17 to0.18), medium for teat dimensions andteat position (r=0.45 to 0.52), and highfor teat angle (r=0.65) and cisternheight (r=0.77), as calculated byFernández et al. (1995) in the Churradairy breed.

Table 2 summarizes the comparison ofmain objective udder measurementscarried out by Rovai et al. (1999) inManchega and Lacaune dairy sheepthroughout lactation, with the aim ofidentifying the most significant uddertraits in extreme yield conditions. Thestage of lactation produced significanteffects on all udder traits in accor-dance with Gallego et al. (1983a) andFernández et al. (1983, 1995).Nevertheless, despite the differencesin milk yield, breed effects on udderlength and distance between teatswere non significant, and only showeda tendency in teat angle. Similarresults were observed in regard toparity, where differences in teat angleand udder length were not significant.In contrast, differences in teat dimen-sions (width and length) and udderheight (depth and cistern height)were significant for breed and parity.These results agree with thoseobtained previously in differentbreeds (Labussière, 1988; Fernández etal., 1983, 1995) although teat anglewas affected by stage of lactation inother references (Casu et al., 1983;Gallego et al., 1983a; Labussière et al.,

1983; Fernández et al., 1989a, 1995).Other authors indicate that udderlength was not affected by the varia-tion factors analyzed.

In regard to the correlation coeffi-cients between udder traits, threenatural groups can be distinguishedas indicated by Fernández et al. (1995):1) udder size (height and width),which are high and positive; 2) teatsize (width and length), which aremedium and positive; and 3) cisternmorphology (height) and teat place-ment (position and angle) which aremedium and positive but show lowand negative correlation with teat andudder sizes. As udder width increases,cistern height and teat angle andposition decrease; and, as udderheight increases, cistern height andteat angle and position also increase.

When morphological traits are relatedto milk yield the greatest effects areobserved for udder width and height(Gallego et al., 1983a; Labussière et al.,1983; Fernández et al., 1989a, 1995;McKusick et al., 1999). Big volume andbig cistern udders produce more milk.Main effects of teat traits are relatedto milk fat (McKusick et al., 1999) andmilk emission during milking(Fernández et al., 1989a; Marie et al.,1999).

70


The most significant and

repeatable udder traits agreed

upon for a wide sample of

sheep dairy breeds are:

■ Teat dimensions (length)

and position (angle).

■ Udder height (also called

depth) and width.

■ Cisterns height.

Table 2. Mean values of udder traits and effects of breed, parity and stage oflactation in Manchega and Lacaune dairy sheep.

Item Breed Effect (P <)Manchega Lacaune Breed Parity Stage

Number of ewes 63 24 – – –

Milk yield (wk 4 to 20):Total, l/ewe 84.6a 153.2b 0.001 0.073 –Daily, l/d 0.82a 1.36b 0.001 0.017 0.001

Teat:Length, mm 33.6a 29.1b 0.003 0.025 0.001Width, mm 15.1a 13.2b 0.002 0.010 0.001Angle,º 42.5 44.1 0.065 0.487 0.052

UdderDepth, cm 17.2a 17.8b 0.001 0.001 0.001Length,cm 11.4 11.3 0.510 0.639 0.001Teat distance, cm 12.6 12.0 0.619 0.001 0.001Cistern height, mm 15.5a 20.0b 0.001 0.002 0.001

a, b Values with different letters in the same line differ (P < 0.05)

Rovai et al., 1999

Linear scoresThe main drawback of the uddertypologies is their use for the estima-tion of the genetic value of breedinganimals and when genetic and envi-ronmental effects need to be brokendown for selection. This problem hasbeen solved in dairy sheep, as in dairycows and goats, by using a breakdownsystem in which independent uddertraits are evaluated according to alinear scale of 9 points (De la Fuente etal., 1996).

The four udder traits considered by Dela Fuente et al. (1996; 1999) to be sig-nificant for machine milking are:udder depth or height (from theperineal insertion to the bottom ofthe udder cistern), udder attachment(insertion perimeter to the abdominalwall), teat angle (teat insertion anglewith the vertical), and teat length(from the gland insertion to thetip).The system also includes anexpanded typology to evaluate thewhole udder shape, in accordance

with the previously described optimalcriteria and udder types, but uses thesame linear scale of 9 points. Eachudder trait is evaluated independentlyby using extreme biological standards(table 3).

The desirable value is in some casesthe highest score (i.e., teat angle:vertical teats that scored 9 will reducecluster drops and make milk drainageeasier) or the average score in others(i.e. teat length: medium size teatsscored 5 and agree with a uniformcluster length). In udder height, givenits positive relationship with milk pro-duction an average score will also bepreferable.

This linear methodology has beenused in Spain for the evaluation of dif-ferent flocks (27 flocks and 10,040ewes) of Churra, Manchega and Latxadairy ewes (De la Fuente et al., 1999)and it is also being partially used inthe Lacaune breed for the evaluationof morphological traits in relation tomachine milking ability (Marie et al.,1999). Results for Spanish breeds are

shown in figure 7 according to lacta-tion stage and parity effects.

In regard to lactation stage, all linearscores decreased as lactation pro-gressed, udder height and udderattachment being the traits thatshowed the greatest decrease duringlactation. Teat size was only slightlymodified. This evolution agrees withthe loss of udder volume and milkyield but indicates a deterioration ofudder morphology for machinemilking as indicated by udder shape.Only udder height was improved.Regarding lactation number, udderheight increased dramatically in thefirst lactations, while other traitsdecreased and teat size was steadilyconstant. As a consequence, uddershape deteriorated and its scoredecreased rapidly from first to thirdlactation and stabilized thereafter.

71


Table 3. Linear scores for the evaluation of main udder morphological traitsin dairy sheep.

Morphological Score (1 to 9)trait 1 (Low) 5 (Average) 9 (High)

Udder height

Teat angle

Teat length

Udder shape

De la Fuente et al., 1996

The values of linear scores calculatedby Fernández et al. (1997) in theChurra dairy breed (table 4) were suffi-ciently repeatable (r= 0.48 to 0.64) andshowed intermediate heritabilityvalues (h2=0.16 to 0.24) as reported incattle. Coefficients of variation rangedbetween 18 and 37%. The authorsindicate that a single scoring per lac-tation would be sufficient in practice.

Udder shape, equivalent to a typologyof nine expanded categories, washighly repeatable and heritable, indi-cating its utility as a single trait fordairy sheep selection. Neverthelessudder shape showed high andpositive genetic correlation withudder attachment (r=0.55) and teatplacement (r=0.96), as a result of themain role of these traits in the defini-tion of udder shape. Consequently, theuse of the first four linear udder traitswill be sufficient to improve programsof udder morphology. Phenotypic andgenetic correlations showed thatselection for milk yield produces apoorer udder morphology, mainly inudder high and teat placement, givingas a result baggy udders that are inad-equate for machine milking.

Repeatabilities of udder linear scoresobtained in the Lacaune dairy breed(Marie et al., 1999) were also high(r=0.59 to 0.71) and show moderatephenotypic correlation with milk yieldin primiparous and multiparous ewes.Heritabilities of udder traits reportedin Assaf (h2= 0.23 to 0.42; Gootwine etal., 1980), Chios (h2= 0.50 to 0.83;Mavrogenis et al., 1988), and Sardawith the seven expanded typologies(h2= 0.55; Carta et al., 1999), gavehigher values but, as indicated by thelast authors, probably they were over-estimated.

72


Figure 7. Evolution of linear scores of main udder traits in Spanish dairy sheep.

6

5

4

3

lactation stage (month)

sco

re

0 2 3 4 61 5 7

udder height heighudder attachmentchment attaudder shape shateat lengtht lengthteat anglet angle

6

5

4

3

lactation number

sco

re

0 2 3 4 61 5 7

udder heighthudder attachmentmenter attaudder shapesteat lengthteat lenteat angletea g

The genetic variability and

heritability of the studied

udder traits indicate that the

efficiency of the breeding

programs could be improved

and some selection on udder

traits in long-term breeding

programs needs to considered.

Table 4. Genetic parameters of linear udder traits in dairy sheep.

Heritability Repeatability Correlation with milk yieldTrait (h2) (r) Phenotypic (rp)Genetic (rg)

Udder height 0.16 0.51 0.40 0.82

Udder attachment 0.17 0.48 -0.01 0.02]

Teat placement 0.24 0.64 -0.04 -0.34

Teat size 0.18 0.54 0.03 -0.16

Udder shape 0.24 0.62 0.03 -0.26

Fernández et al., 1997

De la Fuente et al., 1999

Machine milking Machine milkability is normally esti-mated by fractional milking (machinemilking, machine stripping, andextraction of residual milk after anoxytocin injection) or by analysis ofmilk emission curves obtained duringmachine milking without massage orextra stimulation of the mammarygland. The methodology proposed inthe M4 FAO Project (Labussière, 1983)is normally used as the standardizedmethod for both criteria.

Milk fractioningMilk fractions were mainly used as animportant indicator of milkability indairy sheep when the routinesincluded hand stripping as in the M4FAO project (Labussière, 1983).Reported values of milk fractioningvaried according to breed (Labussière,1988; Such et al., 1999a), milkingroutine (Molina et al., 1989) andmachine milking parameters(Fernández et al., 1999). Values of frac-tioning ranged normally from 60 to 75: 10 to 20: 10 to 15, formachine milking : machine stripping :residual milk, respectively.

The comparison of milking ability oftwo groups of ewes characterized bydifferent milk yield (Manchega, 0.6 l/d;Lacaune, 1.3 l/d), was carried out bySuch et al. (1999a) in late lactation(week 16) and under the same milkingconditions. Values of fractional milking(machine milk: machine stripping milk:residual milk) were 65:19:16 and68:21:11, for Manchega and Lacauneewes, respectively. No significant dif-ferences were observed according tobreed in percentages of milk fractions,except in the case of residual milk(figure 8). Both breeds gave onaverage 86% milk during machinemilking, but the Manchega breedretained more milk in the ductalsystem of the udder. This result wasobtained despite the differencesreported in milk yield and in absolutevalues of each fraction, as well as incistern size (table 1) and udder mor-phology (table 2), of each breed as dis-cussed previously. Differences inudder size and morphology explainthe increase in machine strippingmilk according to milk yield, andwere also reported by effect of lac-tation stage (Gallego et al., 1983b;Labussière, 1988).

As one conclusion, the obtainedresults show the unsuitability of themilk fractions as a main indicatorfor the evaluation of milkability inewes, fractioning probably being abetter indicator for the study ofmachine or milking routine effects,which were the same in this case.Moreover, Caja et al. (1999a) in goatand Fernández et al. (1999a) in sheep,reported significant differences in themachine stripping fraction accordingto milking routine or machine milkingparameters, respectively.

Milk emissionMilk emission is one of the mostinteresting criteria for studyingmilkability in the machine milkingof dairy sheep. Its main traits are con-sidered to be relevant for the designof milking machines and to adopt theoptimal milking routine in each breed.As milk yield strongly influences intra-mammary pressure, a strong effect ofmilk production on all milk flowparameters is also expected, as indi-cated by Marnet et al. (1999) andobserved clearly in dairy goats(Bruckmaier et al., 1994; Caja et al.,1999a). Moreover, milk emission willbe different for a.m. and p.m. milkings,and its curves should be analyzed sep-arately. Morning milking will increasemilk flow and milking time, butemission of alveolar milk will beobserved easily and separately in theafternoon.

Milk emission curves are obtained bymanual (Labussière, 1983; Fernándezet al., 1989b; Peris et al., 1996) or auto-matic methods (Labussière andMartinet, 1964; Mayer et al., 1989b;Bruckmaier et al., 1992; Marie et al.,1999). The flow from the right and leftmammary glands can be recordedseparately (Labussière and Martinet,1964; Labussière, 1983) or as a whole(Fernández et al., 1989b; Peris et al.,1996; Bruckmaier et al., 1992; 1996;

73


milking fraction

Manchega

Lacaune

machinemilk

machinemilk

in gland

machinestripping

milk

machinestripping

milk in gland

residential milk

in gland

volu

me

(%to

tal)

100

80

60

40

20

0

NS

NS

NS

NSP<.05

Figure 8. Milk fractioning obtained during machine milking of dairy sheepaccording to the breed at the same stage of lactation (Such et al., 1999a)

Marie et al., 1999; Marnet et al., 1999),but results and conclusions of flowmay be different in consequence(Rovai, 2000).

A good milk emission curve shouldmean a quick and complete milking,with a high milk flow rate and aneffective ejection of alveolar milkunder the action of the oxytocin.The milk emission pattern is related tothe structure of the udder (cistern size),to the teat traits (size and position) andto the neuro-hormonal behavior of theewe (Labussière et al., 1969; Bruckmaieret al., 1994, 1997; Marnet et al., 1998,1999). Globose and big cisternedudders with medium size, vertical andsensitive teats, that are able to openthe sphincter rapidly and widely at lowvacuums, are preferable.

An early typology of milk emissioncurves was proposed by Labussièreand Martinet (1964), and widelyadopted for the study of sheep dairybreeds (Labussière, 1983, 1988). Themilk emission typology considerscurves of different shape: 1 peak(single), 2 peaks (bimodal) and others,the last corresponding to animals withirregular or multiple milk emissioncurves (≥ 3 peaks). In some cases, aewe changes the milk emissiontypology on consecutive days, andmore than two recordings are recom-mended in practice. The first peakoccurs very early after cluster attach-ment and it is identified as cisternal

milk, which is drained after theopening of the teat sphincter. Thesecond peak corresponds to alveolarmilk and occurs as a consequence ofliberation of alveolar milk during theappearance of the milk ejection reflexby effect of released oxytocin(Labussière and Martinet, 1964;Labussière et al., 1969; Fernández etal., 1989b; Marnet et al., 1998).

Milking-related release of oxytocin hasbeen measured in dairy sheep byMayer et al. (1989a) and Marnet et al.(1998). The machine milk fraction isnormally greater and high milk flowmaintained during a longer time inthe bimodal ewes, which are consid-ered favorable for machine milking indairy ewes. Milking of ewes showing asingle milk emission curve can becompleted by using a milking routinewith machine or manual stripping(‘repasse’) after cessation of themachine milk flow, which is unfavor-able and increases dramatically thetotal milking time per ewe. Moreover,simplified milking routines (withouthand or machine stripping) are wellaccepted by bimodal ewes as indi-cated by Molina et al. (1989) inManchega dairy sheep.

Distribution of animals in a flockaccording to number of peaks has alsobeen used as an index of machinemilkability in dairy breeds as indicatedby Labussière (1988). Sheep breedswith a greater percentage of ewes

showing 2 peaks are the most appro-priate for machine milking.Nevertheless peak distribution in aflock changes according to the stageof lactation as observed by Rovai(2000) in a flock with breeds of differ-ent yield and milkability (table 4).Number of ewes in the 1 peaktypology increased at the end of lacta-tion and on the contrary the ≥ 3 peaksdecreased compensating the losses inthe 2 peaks group.

Machine milking parameters can alsomodify the milk flow characteristics indairy sheep, mainly the volume of thesecond peak and the milking time, asreported by Fernández et al. (1999) inManchega dairy ewes milked at differ-ent vacuum levels (36 and 42 kPa) andpulsation rates (120 and 180 P/min) .

Clear differences in milk emissioncurves during the p.m. milking wereobserved by Such et al. (1999b)according to breed, when Manchegaand Lacaune dairy ewes at the samestage of lactation were compared(figure 9) indicating the importance ofthis criterion on the evaluation ofmilkability. Daily milk yield at compari-son and percentage of bimodal ewesduring the comparison period were0.6 l/d and 38%, and 1.3 l/d and 83%,for Manchega and Lacaune ewesrespectively.

74


Table 4. Distribution (%) of milk emission curves obtained in dairy ewes during machine milking according to breed andstage of lactation.

Stage of (d) ______________Manchega______________ _______________Lacaune_______________

lactation 1 peak 2 peaks ≥ 3 peaks 1 peak 2 peaks ≥ 3 peaks

421 28.6 56.7 14.7 8.0 57.2 34.8

(62)2 (123) (32) (16) (115) (70)

70 29.6 64.2 6.19 9.8 49.4 40.8

(67) (145) (14) (25) (126) (104)

98 39.4 54.8 5.9 18.0 55.6 26.5

(74) (103) (11) (34) (105) (50)1 First week after weaning at day 35 2 Number of emission curves analyzed Rovai, 2000

Significant differences in the values ofmaximum milk flow (76 vs 129 ml/5s)and milk peak volume (207 vs 586 ml)were observed for the 1-peakManchega versus Lacaune ewes,respectively. The significant values forthe 2-peaks ewes were: first peak (72vs 94 ml/s; and, 171 vs 344 ml) andsecond peak (41 vs 83 ml/s; and, 78 vs239 ml), for Manchega vs Lacaune,respectively. Total emission time untila milk flow <10ml/s were: 1 peak (25vs 39 s) and 2 peaks (48 vs 56 s) forManchega vs Lacaune respectively.Observed differences in milk flowparameters between breeds were inaccordance with their milk yield.Nevertheless, despite the differencesof milk emission curves, the totalvolume of milk obtained in 1-peak vs2-peaks ewes were similar in eachbreed: Manchega (207 vs 249 ml) andLacaune (586 vs 583 ml) respectivelyfor 1-vs 2-peaks. Moreover maximummilk flow was the same in both breedsfor the 2 peaks ewes, despite the dif-ferences in yield. As a consequence, itcan be suggested that other factorsdifferent from milk ejection reflex aremainly conditioning the milk flowduring machine milking in dairy ewes.

Teat and cistern characteristics seemto be the most important factors inrelation to milk flow curves in dairysheep. Results of Marie et al. (1999) andMarnet et al. (1999) in Lacaune dairysheep, and Bruckmaier et al. (1994,1997) studying the effects of milkingwith or without prestimulation inSaanen dairy goats, and Friesian andLacaune dairy sheep, are in accordancewith these conclusions.

Marnet et al. (1999) indicate that lagtime between teat cup attachment andarrival of the first milk jets to therecording jar can be used as an indica-tor of milkability. Moreover significantcorrelation of lag time with vacuumneeded to open the teat sphincter(r=0.61), total milking time (r=0.86), andmean (r=–0.84) and maximum(r=–0.80) milk flow rates, wereobserved. A low but significant correla-tion between Somatic Cell Count andmaximum milk flow was also obtained(r= 0.39). Moreover, the vacuum valueneeded to open the teat sphincterseems to remain constant in eachanimal during lactation and is also pos-itively related with the teat congestionobserved after milking.The highestvacuum value needed to open the teatsphincter in this experiment was 36kPa, suggesting that the use of a lowmilking vacuum is possible in Lacaunedairy ewes.

Accordingly with these results, Marieet al. (1999) studied the main uddertraits and milk flow characteristics byusing an automatic milk recorder intwo lines of Lacaune dairy ewes differ-ing 60 l in genetic merit. Milk yield andmilking time averaged 0.94 l/d and 2min 44 s respectively. Average lag timewas 25 s for a minimum volume ofmilk of 160 ml. Maximum milk flow(0.87 l/min) was observed 27 s later(52 s from cluster attachment) inaverage. Lag time was negatively cor-related with milk yield (r= –0.26) andmaximum milk flow (r= –0.49).Measured repeatabilites for milk yield,lag time and maximum milk flow werehigh in the same lactation (r= 0.46 to0.59) and between lactations (r= 0.40to 0.75). Flow parameters variedaccording to milk yield as previouslyreported by Bruckmaier et al. (1994) ingoats, but the increase in milking timewas lower than in milk.

75


time (s)

1 peak 2 peak

flo

wra

te (m

l/5

s)

10 20 30 40 700 50 60

100

80

60

40

20

0

time (s)

flo

wra

te (m

l/5

s)

10 20 30 40 700 50 60

100

80

60

40

20

0

ManchegaM nch

LacauneL u

ManchegagM nch

LacauneL aune

Figure 9. Milk emission curves resulting from p.m. machine milking of dairy sheep according tobreed and number of peaks (Such et al., 1999b).

Correlation of udder traits with flowparameters obtained by Marie et al.(1999) were low (–0.3 to 0.3) andtended to increase in multiparousewes. An increase in teat angle wasassociated to a greater lag time(r=0.28) and a lower maximum milkflow (r=–0.26), both unfavorable traits.On the contrary, a very marked inter-mammary groove was correlated togreater milk yield (r=0.28) and milkflows (r=0.33 to 0.34), and lower lagtime (r=–0.23). As a final conclusionthe authors indicate that a goodudder shape tends to improve milka-bility and recommend the inclusion ofthis trait in genetic programs.

Bruckmaier et al. (1997) comparedmilk flow and udder anatomy, includ-ing ultrasound images, in Lacaune andFriesian dairy ewes. Both breedsshowed similar milk yield and cisternalareas. Nevertheless, milk flow waslower and stripping milk yield higherin the Friesian ewes as a consequenceof udder morphology that showed cis-ternal bags below the level of the teatchannel. The use of a prestimulationroutine failed to reduce stripping milkand total milking time but increasedmilk flow in both breeds. Oxytocinrelease was different in both breedsand a dramatic increase in blood con-centration was observed in Lacauneewes during teat stimulation and earlymilking, while only slight release wasfound in Friesian ewes. Duringmachine milking significant increasein oxytocin was observed in 88% ofLacaune but only in 58% of Friesianewes. The authors also indicate theoccurrence of single peak typologiesin milk emission with or withoutincreasing concentrations of oxytocinin both breeds.

ConclusionRelationships between morphologicand productive traits are evident indairy sheep as a consequence ofanatomical and physiological charac-teristics. Breed differences are alsodetected despite the differences inmilk yield. Phenotypic and geneticcorrelations indicate that selection formilk yield will produce a worse uddermorphology, mainly in udder heightand teat placement, causing baggyudders that are inadequate formachine milking. Teat and cisterncharacteristics appear to be the mostlimiting factors in machine milkabilityand especially in milk flow. Geneticvariability, repeatability and heritabil-ity of udder traits indicate that someselection pressure on udder traitsneeds to be considered. In practice theuse of four linear udder traits will besufficient to improve udder morphol-ogy in long-term breeding programs.

ReferencesAlvarez-Morujo A.J. and A. Alvarez-

Morujo. 1982. Variedades mor-fológicas del alveolo de la glándulamamaria. Zbl. Vet. Med. C. Anat.Histol. Embryol., 11: 65-75.

Arranz J., J.M. López de Munain and J.Lara. 1989. Evolución de las carac-terísticas morfológicas de la ubrede ovejas de raza Latxa a lo largodel periodo de ordeño. In: 4thInternational Symposium onMachine Milking of SmallRuminants. Kibbutz Shefayim, Tel-Aviv, Israel. pp. 80-93.

Barone R. 1978. Mamelles de la brebiset de la chèvre. In: Anatomiecomparée des mammifèresdomestiques. Tome III. J. Tixier etFils S.A., Lyon, France. pp. 491-495.

Bocquier F. and Ph. Guillouet. 1990.Nutrition et alimentation de labrebis laitière. In: SeminarioAvanzado sobre mejora de lacalidad en la producción del ovinode leche. CIHEAM-IAMZ, Zaragoza.38 pp.

Brooker B.E. 1984. An ultrastructuralstudy of the sinus epithelium inthe mammary gland of the lactat-ing ewe. J. Anatomy, 138:287-296.

Bruckmaier R.M., C. Ritter, D. Schamsand J.W. Blum. 1994. Machinemilking of dairy goats during lacta-tion: udder anatomy, milking char-acteristics, and blood concentra-tions of oxytocin and prolactin.J. Dairy Res., 61:457-466.

Bruckmaier R.M., G. Paul, H. Mayer andD. Schams. 1997. Machine milkingof Ostfriesian and Lacaune dairysheep: udder anatomy, milkejection and milking characteris-tics. J. Dairy Res., 64:163-172.

Burri P.H. and M.R. Tarek. 1990. A novelmechanism of capillary growth inthe rat pulmonary microcircula-tion. Anat. Rec., 228: 35-45.

Caja G. and X. Such. 1999. Curso deactualización sobre ordeñomecánico de ovino y caprino.SEOC-UAB, Bellaterra, Barcelona.CD Rom.

Caja G., J. Capote, L. López, S. Peris, X.Such and A. Argüello. 1999a. Milkpartitioning and milk flow rate ofCanarian dairy goats under oncedaily, twice daily milking frequen-cies. In: Milking and milk produc-tion of dairy sheep and goats. F.Barillet and N.P. Zervas (Eds.), EAAPPublication No. 95, WageningenPers., Wageningen. pp. 274-280.

76


Caja G., X. Such, J. Ruberte, A. Carreteroand M. Navarro. 1999b. The use ofultrasonography in the study ofmammary gland cisterns duringlactation in sheep. In: Milking andmilk production of dairy sheep andgoats. F. Barillet and N.P. Zervas(Eds.), EAAP Publication No. 95,Wageningen, Pers., Wageningen.pp. 91-93.

Caruolo E.V., 1980. Scanning electronmicroscope visualization of themammary gland secretory unitand myoephithelial cells. J. DairySci., 63:1987-1998.

Carretero A., H. Ditrich, F.J. Pérez-Aparicio, H. Splechtna and J.Ruberte. 1995. Development anddegeneration of the arterial systemin the mesonephros andmetanephros of chicken embryos.Anat. Rec., 243:120-128.

Carretero A., J. Ruberte, G. Caja, F.J.Pérez-Aparicio, X. Such, S. Peris, M.Manesse and M. Navarro. 1999.Study on the structure and thedevelopment of the canalicularsystem of the mammary glandduring lactation in Manchega andLacaune dairy sheep. In: Milkingand milk production of dairy sheepand goats. F. Barillet and N.P. Zervas(Eds.), EAAP Publication No. 95,Wageningen Pers., Wageningen.pp. 35-40.

Carta A., S.R. Sanna, G. Ruda and S.Casu. 1999. Genetic aspects ofudder morphology in Sarda primi-parous ewes. In: Milking and milkproduction of dairy sheep andgoats. F. Barillet and N.P. Zervas(Eds.), EAAP Publication No. 95,Wageningen Pers., Wageningen.pp. 363-368.

Casu S., R. Carta and G. Ruda.1983.Morphologie de la mamelle etaptitude à la traite mécanique dela brebis Sarde. In: 3rd InternationalSymposium on Machine Milking ofSmall Ruminants. Sever-Cuesta,Valladolid, Spain. pp. 592-603.

Casu S., F. Barillet, R. Carta and S.Sanna. 1989. Amélioration géné-tique de la forme de la mamelle dela brebis Sarde en vue de la traitemécanique: Résultats préliminaires.In: 4th International Symposium onMachine Milking of SmallRuminants. Kibbutz Shefayim, Tel-Aviv, Israel. pp. 104-133.

De la Fuente L.F., G. Fernández and F.San Primitivo. 1996. A linear evalua-tion system for udder traits of dairyewes. Livest. Prod. Sci., 45:171-178.

De la Fuente L.F., D. Pérez-Guzman,M.H.Othmane and J.Arranz. 1999.Amélioration génétique de la mor-phologie de la mamelle dans lesraces Churra, Latxa et Manchega.In: Milking and milk production ofdairy sheep and goats. F. Barilletand N.P. Zervas (Eds.), EAAPPublication No. 95, WageningenPers., Wageningen. pp. 369-374.

Delouis C. and Ph. Richard. 1991. LaLactation. In: La reproduction chezles mammifères et l’homme. Ch.Thibault and M.C. Levasseur (Eds.).INRA. Paris. p. 487-514.

Ditrich, H. And H. Splechtna. 1989. Theopistonephric blood vascularsystem of the chicken embryo asstudied by scanning electronmicroscopy of microcorrosioncasts and critical point driedpreparations. Scanning Microsc.,3: 559-561.

Fernández G., P. Alvarez, F. SanPrimitivo and L.F. De la Fuente.1995. Factors affecting variation ofudder traits of dairy ewes. J. DairySci., 78:842-849.

Fernández G., J.A. Baro, L.F. de laFuente and F. San Primitivo. 1997.Genetic parameters for linearudder traits in dairy ewes. J.Dairy Sci., 80:601-605.

Fernández G., D. Farias, M. López and C.Silveira. 1999b. Genetic improve-ment in milkability of theCorriedale breed: morphologicalstudy of the udder traits. In:Milking and milk production ofdairy sheep and goats. F. Barilletand N.P. Zervas (Eds.), EAAPPublication No. 95, WageningenPers., Wageningen. pp. 434-437.

Fernández N., J. Arranz, G. Caja, A.Torres and L. Gallego. 1983a.Aptitud al ordeño mecánico deovejas de raza Manchega: I.Biometría corporal, característicasmorfológicas de la ubre y cría decorderos. In: 3rd InternationalSymposium on Machine Milking ofSmall Ruminants. Sever-Cuesta,Valladolid, Spain. pp. 653-666.

Fernández N., J. Arranz, G. Caja, A.Torres and L. Gallego. 1983b.Aptitud al ordeño mecánico deovejas de raza Manchega: II.Producción de leche, reparto defracciones y cinética de emisión deleche. In: 3rd InternationalSymposium on Machine Milking ofSmall Ruminants. Sever-Cuesta,Valladolid, Spain. pp. 667-686.

Fernández N., G. Caja, A. Torres, M.P.Molina and L. Gallego. 1989a.Cinética de emisión de leche deovejas de raza Manchega: II.Relación con otros criterios deaptitud al ordeño mecánico. Invest.Agr. Prod. Sanid. Anim., 4:23-33.

Fernández N., G. Caja, A. Torres, M.P.Molina and C. Peris. 1989b. Cinéticade emisión de leche de ovejas deraza Manchega: II. Parámetros delas curvas de emisión de lechedurante el ordeño a máquina.Invest. Agr. Prod. Sanid. Anim.,4:9-21.

77


Fernández N., J.R. Díaz, C. Peris, M.Rodríguez, M.P. Molina and A.Torres.1999a. Machine milkingparameters for the Manchegabreed. In: Milking and milk produc-tion of dairy sheep and goats. F.Barillet and N.P. Zervas (Eds.), EAAPPublication No. 95, WageningenPers., Wageningen, pp. 65-68.

Franke, W.W. and T.W. Keenan. 1979.Mitosis in milk secreting epithelialcells of mammary gland. An ultra-structural study. Differentiation,13:81-88.

Gallego L., G. Caja and A. Torres. 1983a.Estudio de la tipología y caracterís-ticas morfológicas de las ubres deovejas de raza Manchega desde elparto. In: 3rd InternationalSymposium on Machine Milking ofSmall Ruminants. Sever-Cuesta,Valladolid, Spain, pp. 100-116.

Gallego L., A. Torres, G. Caja and M.P.Molina. 1983b. Fracciones de lecheen el ordeño de ovejas de razaManchega: cuantificación y evolu-ción durante la lactación. In: 3rdInternational Symposium onMachine Milking of SmallRuminants. Sever-Cuesta,Valladolid, Spain. pp. 149-163.

Gallego L., A. Torres, G. Caja, M.P. Molinaand N. Fernández. 1985. Reparto delas fracciones de leche y su com-posición según el tipo de ubre.ITEA Prod. Anim., 58:21-28.

Gootwine E., B. Alef and S. Gadeesh.1980. Udder conformation and itsheritability in the Assaf (Awassi xEast Friesian) cross dairy sheep inIsrael. Ann. Génét. Sél. Anim.,12:9-13.

Hatziminaoglou J., N. Zervas, E. Sinapisand P. Hatziminaoglou. 1983.Aptitude à la traite mécanique desbrebis de race Karagouniko(Grèce). Données préliminairesconcernantes la production et lacomposition du lait, la morpholo-gie des mamelles et la cinétique

d’émission du lait. In: 3rdInternational Symposium onMachine Milking of SmallRuminants. Sever-Cuesta,Valladolid, Spain, pp. 607-624.

Hurley W.L. 1989. Mammary glandfunction during involution. J. DairySci., 72:1637-1646.

Jatsh O. and R. Sagi.1978. Effects ofsome anatomical and physiologi-cal traits on dairy yield and milkfractionation in dairy ewes. In: Proc.2nd International Symposium onMachine Milking of SmallRuminants. INRA-ITOVIC, Alghero,Italy, pp. 60-79.

Johnsson I.D. and I.C. Hart. 1985. Pre-puberal mammogenesis in thesheep. I. The effect of level of nutri-tion and growth and mammarydevelopment in female lambs.Anim. Prod., 41:323-332.

Knight C.H. and C.J. Wilde. 1993.Mammary cell changes duringpregnancy and lactation. Livest.Prod. Sci., 35:3-19.

Knight C.H., K. Stelwagen, V.C. Farr andS.R. Davis. 1994. Use of an oxytocinanalogue to determine cisternaland alveolar milk pool sizes ingoat. J. Dairy Sci., 77 (Suppl. 1): 84(Abstr.).

Kukovics S. and A. Nagy. 1989.Relationships between sheepgenotype and udder type as wellas relative measurements of udder.In: 4th International Symposium onMachine Milking of SmallRuminants. Kibbutz Shefayim, Tel-Aviv, Israel, pp. 66-79.

Labussière J. 1983. Étude des aptitudeslaitières et de la facilité de traite dequelques races de brebis du BassinMéditerranéen. Projet M4 FAO. In:3rd International Symposium onMachine Milking of SmallRuminants. Sever-Cuesta,Valladolid, Spain, pp. 780-803.

Labussière J. 1988. Review of physio-logical and anatomical factorsinfluencing the milking ability ofewes and the organization ofmilking. Livest. Prod. Sci.,18:253-273.

Labussière J. and J. Martinet. 1964.Description de deux appareils per-mettant le contrôle automatiquedes débits de lait au cours de latraite à la machine. Premiers résul-tats obtenus chez la brebis. Ann.Zootech., 14 :63-126.

Labussière J., B. Bennemederbel, J.F.Combaud and F. De la Chevalerie.1983. Description des pricipauxparamètres caractérisant la pro-duction laitière, la morphologiemammaire et la cinétique d’émis-sion du lait de kla brebis Lacaunetraite une ou deux fois par jouravec ou sans égouttage. In: 3rdInternational Symposium onMachine Milking of SmallRuminants. Sever-Cuesta,Valladolid, Spain, pp. 656-652.

Labussière J., D. Dotchewski and J.F.Combaud. 1981. Caractéristiquesmorphologiques de la mamelledes brebis Lacaune. Méthodologiepour l’obtention des données.Relation avec l’aptitude à la traite.Ann. Zootech., 30:115-136.

Labussière J., J. Martinet and R.Denamur. 1969. The influence ofthe milk ejection reflex on the flowrate during the milking of ewes.J. Dairy Res., 36 :191-210.

Marie C., M. Jacquin, M.R. Aurel, F. Pailler,D. Porte, P. Autran and F. Barillet.1999. Déterminisme génétique de lacinétique d’émission du lait selon lepotentiel laitier en race ovine deLacaune et relations phénotypiquesavec la morphologie de la mamelle.In: Milking and milk production ofdairy sheep and goats. F. Barillet andN.P. Zervas (Eds.), EAAP PublicationNo. 95,Wageningen Pers.,Wageningen, pp. 381-388.

78


Marnet P.G., J.F. Combaud and Y. Dano.1999. Relationships between char-acteristics of the teat and milkabil-ity in Lacaune ewes. In: Milking andmilk production of dairy sheep andgoats. F. Barillet and N.P. Zervas(Eds.), EAAP Publication No. 95,Wageningen Pers., Wageningen,pp. 41-44.

Marnet P.G., J.A. Negrao and J.Labussière. 1998. Oxytocin releaseand milk ejection parametersduring milking of dairy ewes inand out natural season of lacta-tion. Small Ruminant Res.,28:183-191.

Martal J. and N. Chene. 1993. Placentaand lactation. In: Biologie de la lac-tation. J. Martinet and L.M.Houdebine (Eds.). INRA Editions.Paris, pp 31-58.

Mayer H., F. Weber and V.Segessemann. 1989a. Oxytocinrelease and milking characteristicsof Ostfriesian and Lacaune dairysheep. In: 4th InternationalSymposium on Machine Milking ofSmall Ruminants. KibbutzShefayim, Tel-Aviv, Israel,pp. 548-563.

Mayer H., F. Weber and V.Segessemann. 1989b. A method torecord and define milk flow curvesof sheep during routine machinemilking. In: 4th InternationalSymposium on Machine Milking ofSmall Ruminants. KibbutzShefayim, Tel-Aviv, Israel,pp. 564-575.

McCann M.A., L. Good, R.W. Harvey, E.V.Caruolo and D.I. Mann. 1989.Effects of rapid weight gain topuberty on reproduction,mammary development and lacta-tion in ewe lambs. Theriogenology,32:55-68.

McKusick B.C., Y.M. Berger and D.L.Thomas. 1999. Preliminary results:Effects of udder morphology oncommercial milk production ofEast Friesian crossbreed ewes. Proc.5th Great lakes Dairy SheepSymposium, Brattleboro, Vermont,pp. 81-90.

Molina M.P., X. Such, N. Fernández, G.Caja and A. Torres. 1989. Supresióndel repaso manual en el ordeñomecánico de ovejas de razaManchega. 1. Efecto sobre la pro-ducción de leche y el frac-cionamiento de la leche en elordeño. In: 4th InternationalSymposium on Machine Milking ofSmall Ruminants. KibbutzShefayim, Tel-Aviv, Israel,pp. 655-679.

Patan S., M.J. Alvarez, J.C. Schittny andP.H. Burri. 1992. Intussusceptivemicrovascular growth: A commonalternative to capillary sprouting.Arch. Histol. Cytol., 55:65-75.

Pérez J., J.L. Gómez and J. García. 1983.Contribución al estudio de la mor-fología de la ubre de la ovejaManchega. In: 3rd InternationalSymposium on Machine Milking ofSmall Ruminants. Sever-Cuesta,Valladolid, Spain. pp. 583-591.

Peris S., X. Such and G. Caja. 1996.Milkability of Murciano-Granadinadairy goats. Milk partitioning andflow rate during machine milkingaccording to parity, prolificacy andmode of suckling. J. Dairy Res.,63:1-9.

Purroy A. and J.L. Martín. 1983.Características de ordeño de laoveja Churra: Producción y com-posición de la leche, morfología dela ubre y cinética de emisión deleche. In: 3rd InternationalSymposium on Machine Milking ofSmall Ruminants. Sever-Cuesta,Valladolid, Spain, pp. 568-581.

Purroy A., J.L. Martín and J.J. Jurado.1982. Aptitud al ordeño mecánicoen ovejas de raza Churra. I.Características morfológicas de laubre. An. INIA/Ser. Ganadera,16:41-51.

Rovai M. 2000. Caracteres mamariosque afectan a la aptitud al ordeñomecánico de ovejas de razaManchega y Lacaune. Ph.D. Thesis.Universitat Autònoma deBarcelona.

Rovai M., X. Such, J. Piedrafita, G. Caja,M.R. Pujol. 1999. Evolution ofmammary morphology traitsduring lactation and its relation-ship with milk yield of Manchegaand Lacaune dairy sheep. In:Milking and milk production ofdairy sheep and goats. F. Barilletand N.P. Zervas (Eds.), EAAPPublication No. 95, WageningenPers., Wageningen, pp. 107-109.

Rovai M., X. Such, G. Caja and C.H.Knight. 2000. Interbreed differ-ences in cisternal and alveolar milkpartitioning in the udder accord-ing to yield in dairy sheep. J. DairySci., 83 (Suppl. 1): 166 (Abstr.).

Ruberte J., A. Carretero, M. Fernández,M. Navarro, G. Caja, F. Kirchner andX. Such. 1994a. Ultrasound mam-mography in the lactating ewe andits correspondence to anatomicalsection. Small Rumin. Res.,13:199-204.

Ruberte J., A. Carretero, M. Fernández,M. Pons, J.M. Giné and J. Sautet.1994b. Anatomía de la ubre de laoveja. Ovis, 32:9-16.

Sagi R. and M. Morag. 1974. Udder con-formation, milk yield and milk frac-tionation in the dairy ewe. Ann.Zootech., 23:185-192.

79


Such X., G. Caja and L. Pérez. 1999a.Comparison of milking abilitybetween Manchega and Lacaunedairy ewes. In: Milking and milkproduction of dairy sheep andgoats. F. Barillet and N.P. Zervas(Eds.), EAAP Publication No. 95,Wageningen Pers., Wageningen,pp. 45-50.

Such X., G. Caja, N. Fernández, M.P.Molina and A. Torres. 1999b. Theeffects of type of pulsator on therevolution of milk emission kineticsduring machine milking inManchega ewes. In: Milking andmilk production of dairy sheep andgoats. F. Barillet and N.P. Zervas(Eds.), EAAP Publication No. 95,Wageningen Pers., Wageningen,pp. 227-232.

Tenev S. and G. Rusev. 1989. Micro yultraestructura de la ubre deovejas en razas de distintorendimiento de leche. In: 4thInternational Symposium onMachine Milking of SmallRuminants. Kibbutz Shefayim,Tel-Aviv, Israel, pp. 47-65.

Turner C.W. 1952. The anatomy of theudder of sheep. In: The anatomy ofthe mammary gland. LucasBrothers Publishers, Columbia,Missouri, pp. 315-331

Wilde C.J., C.H. Knight and M. Peaker.1996. Autocrine regulation of milksecretion. In: Progress in dairyscience. C.J.C. Phillips (Ed.). CABInternational, Wallingford, Oxon,United Kingdom, pp. 311-332.

Williams J.M. and C.W. Daniel. 1983.Mammary ductal elongation:Differentiation of myoepitheliumand basal lamina during branchingmorphogenesis. Developm. Biology,97: 274-290.

Names in italics correspond to theLatin denomination of the NominaAnatomica Veterinaria (1994). WorldAssociation of Veterinary Anatomists,4th edition, Zurich.

80


There are a number of methods toimprove the quality and quantityof milk, some of which have been

neglected in the past 20 years. Muchwork has been dedicated to feedingand genetic improvement of milkyield and composition, but the milkingprocess has been considered second-ary because milking equipment tech-nology has not evolved sufficiently. InEurope, the extension of milk quotas(restriction to the right to produce) tosmall ruminants has encouragedfarmers to seek technical solutions toimprove the quality of the milk and atthe same time simplify working proce-dures. Some solutions lie in a betterunderstanding and use of milkejection mechanisms.

Review of milk syn-thesis and ejectionAfter a phase of mammary growth(secretory alveoli and ducts) con-trolled mainly by ovarian steroids, themilk surge, or lactogenesis, will neces-sitate stimulation of secretory cells bya number of pituitary hormones; pro-lactin and ACTH in particular. Notethat inducing lactation artificially onlyrequires the administration of steroidsto prepare the udder for these pitu-itary hormones before turning theanimals to milking.

Milking, through udder stimulation,induces the release of a hormonecompound necessary for the ultimatephase of mammogenesis and theinduction of lactation. To maintain lac-tation, other hormones that preferen-

tially act on mammary metabolism,such as growth hormone (GH) areneeded. It is worth noting that udderclearance is always followed by GHrelease.

This also explains the so-called lacta-tion maintenance reflex linked tomammary gland stimulation. Once themilk has been produced, it still has tobe drawn from the udder, otherwisedrying-out will occur very quickly.

This means that the accumulation ofmilk, adding to the lack of hormonesrequired for milk synthesis, will stop thecellular mechanism.Two reasons havebeen put forward to explain this. First,pressure in the secretory alveoli crushesalveolar cells and impedes secretionvesicles’ transfer and also slows downthe passive passage of elements fromblood to milk.The second cause isthought to involve one or several lac-toserum peptides (Feedback Inhibitorof Lactation: FIL) which, by accumulat-ing in the alveoli, have inhibitory effectson lactose synthesis.

This clearly demonstrates the impor-tance of thoroughly draining all themilk contained in the alveoli at eachmilking. But thorough drainingrequires the active participation of theanimal. Indeed, if between milkingsthe milk is partially discharged intothe cisterns in the lower part of theudder, some remains in the alveoli andin the small galactophores at the topof the udder. That milk contains muchmore fat because fat cells are largerthan the diameter of these smallducts.

81

Management for bettermilk yield and quality

Chapter 7

Solutions lie

in a better

understanding

and use of milk

ejection

mechanisms.

To be extracted, that milk must beexpelled from the alveoli by thepressure applied on the alveolar wallby myoepithelial muscle cells. Thesecells spontaneously contract (smoothmuscle cells), but the ejection of milkwill only be effective if their contrac-tions are synchronized, which can onlybe achieved if they are stimulated by aneuropituary hormone, oxytocin.Again, the release of that hormone inblood results from a neuro-humoralreflex initiated in the udder. So opti-mizing milk ejection comes down toretrieving the milk and usable matterthat was produced through geneticselection and feeding, and thus opti-mizing the animal’s potential. Themilk, by going down into the cisterns,increases the intramammary pressureand the pressure ratio between thecisterns and the mouth of the sucklinglamb or the machine vacuum nozzle.This also accelerates draining andmakes milking quicker.

Finally, if all information transitsthrough the central nervous system, itis likely that the CNS may act as amodulator of response to udder stim-ulation. For instance, the connectionsof the oxytocin-producing hypothala-mic nuclei (supre-aortic and paraven-tricular nuclei) to the limbic system,which is the emotion site, and thecortical areas which are the memorysites, explain why recognition of ananxiety factor (biting dog, stranger onthe farm, sudden replacementshepherd, bleating of lambs, undergo-ing treatment such as injection or foottrimming, shearing noise) may inhibitoxytocin release and hence milkejection.

Other factors, on the contrary, mayfacilitate milk ejection. In sucklingfarms, it is the sight and cry of theyoung and in dairy farming, when all iswell, the sight of the usual milker, thestarting of the vacuum pump and/orpulsation, entering the milking penand above all concentrate feeding inthe milking pen. It has to be notedthat there is a close relationshipbetween oxytocin and anotherpeptide: CCK (cholecystokinine).Although this has not been proven inruminants, the CCK released at theperipheral level when the feed bolusreaches the stomach is thought toinduce oxytocin release and mighttherefore promote milk ejection.However, CCK may also be released atthe central level, which controls rumi-nation. But oxytocin may in turninduce CCK release. This implies there-fore that rumination nearly alwaysfollows oxytocin release and milkejection.

Respecting the animal and stimulatingit as much as possible may appearcoarse (but not that easy), but is nec-essary to extract all the secreted milkas quickly as possible and to maintainlactation.

Milk ejectionMilk ejection, which in dairy cowsusually occurs during massage and inthe first minutes of milking, has totake place within only two milkingminutes in ewes. It therefore requirescareful animal selection and optimalsetting of the milking machine. Milkejection can be monitored duringmilking without using any invasivetechnique and without bothering theanimals.

Measuring the milk emission output atmilking is sufficient. The technique hasproduced very interesting results interms of the distribution of milk in theudder and is still a reference methodfor selecting animals according totheir milking easiness.

In ewes, milk generally flows in severalstages. The first outflow peak corre-sponds to cisternal milk discharge.Then a second outflow peak occursonly if the nervous connectionbetween the udder and the CNS isunimpaired. That outflow thereforedepends on oxytocin and representsthe volume of milk trapped in thesmall galactophores and the alveoli.That milk is called the alveolar milk.

Last, a third increase in outflow is notedat the time of stripping.That milkfraction represents the milk below theteat in the mammary gland pockets. Ifthe ejection of alveolar milk is incom-plete, the massage performed duringstripping and the tap stimuli applied bythe milker under the udder will help inretrieving all or part of it. Note that inthe early days of mechanization, thepoor performance of the machines, andthe large number of ewes that did notrespond to mechanical milking stimula-tion forced the milkers to perform handmilking to retrieve residual milk afterremoving the milking bundle.Thatoperation is now rarely performed.

82


An animal that does not

ruminate in the milking pen is

not in favorable conditions

and will not express its full

potential.

In 1982, almost all of French Lacauneewes were unresponsive and necessi-tated time-wasting and tedious strip-ping and manual re-milking opera-tions for all the milk produced to beretrieved and collected. In 1995, only7–8% of these remained, mainly ewelambs. Those ewes, which only emittheir cisternal milk have lower milkyield, less rich milk (up to 70% of thefat can be trapped in the alveolarfraction between milking) and poorlactation persistence. These ewestherefore are removed from the flocks.

It should be noted also that ewes withpoor reflex have a lower milk outflow,inducing protracted milking times. It istherefore important to carry on select-ing ewes according to their milkemission kinetics.

Today, because of the sharp increasein the volume of milk produced, cister-nal milk often does not finish flowingwhen the alveolar milk ejected by theaction of oxytocin reaches the teat. Asin cows and goats, it becomes difficultor impossible to distinguish betweenthe two emissions and to measuretheir respective outflow. At most, thereflex is known to have occurred if themilk emission kinetics lasts for morethan 40 seconds with a high outflow,which is the maximum time for effec-tive oxytocin release.

There is a good correlation betweenthe cisternal milk volume and milkyield. That volume currently representsas much as 38% of total milk yield onaverage. The alveolar milk volume issimilar (34%) and so up to 28% of totalmilk is represented by stripping milk.Stripping is therefore mandatory. But alarge part of that stripping milk islinked to the mammary gland mor-phology, not to a problem of effectivemilk ejection. Selection according tomilk production performance hasresulted in larger cistern volumes,partly due to the enlargement of thepockets at the base of the udder.

Consequently, the teats are higher andtheir position precludes completedrainage of the mammary gland.

Furthermore, that teat position makesthe fitting of nozzles more difficultand may induce air intake or bundledisconnection detrimental to theudder health (by the impact on teatsand increased risk of germ contamina-tion). It is therefore crucial, as in dairycows, to select ewes considering theirudder morphology and by choosinganimals with teats as vertical aspossible, properly draining the udder.Combined with a good oxytocinrelease, vertical teat placement willwarrant effective milking, which canbe simplified by automatic clusterremoval, a technique known to reduceovermilking and improve teat healthin cows. Among the various ewebreeds, some have milk emissionkinetics with a single flow peak, highvolume emission (a characteristic ofFriesian ewes).

If there are fewer of these ewes withoxytocin release at milking than themore highly selected Lacaune ewes, itis nonetheless true that these animalsoffer large cistern volume and theability to transfer alveolar milk intocisterns between milkings.This ensuresthat synthesis will not be hindered andthat the secretory potential will not bereduced throughout lactation. Inaddition, poor setting of the milkingmachine or the presence of milking-refractory animals will have less impactand milking will be simplified.

The effect of oxytocin releasebetween milkings on the distributionof milk in the udder and on milk yieldhas been verified. It appeared that ifblood oxytocin is maintained through-out the day at the same level asduring milking, the storage volumeincreases in proportion with total milkand the alveolar milk volume slightlydecreases and holds. The result is an18–25% increase in milk yield.

Whatever the reason, good milktransfer between milkings appears tobe as important a factor of better milkyield and easier milkings as milkejection during milking. It is worthnoting that luteal oxytocin could beamong the factors causing thattransfer, because milk transfer in thecistern increases when there is sexualactivity. Other milk ejection factorshave been evidenced in ovaries, whichled us to deepen our knowledge ofthe relationship that exists betweenthe ovarian sphere and the udder.

Oxytocin titration does not provideinformation about the occurrence ofmilk ejection because the importantfactor is the form of release ratherthan the amount of oxytocin released.Indeed, sustained oxytocin releaseresults in high intermammarypressure during milking and thus,quicker and complete draining of theudder. There is also a very smallnumber of cases when oxytocinrelease occurs and has no effect onthe udder. There are multiple reasonsfor that but the most likely ones arethe absence or deactivation of recep-tors on the mammary gland.Catecholamine release may also occurat the peripheral level, reducingmammary blood flow to a pointwhere the oxytocin level is no longersufficient to ensure effective alveolarcontraction. Considering the costs ofoxytocin assays and the necessity toperform several of these tests duringthe course of one milking, themethods should remain experimentalor at the most be used to select thebest breeding ewes in breeding units.

83

M A N A G E M E N T F O R B E T T E R M I L K Y I E L D & Q U A L I T Y C H A P T E R 7

Ewe managementat milkingThere are many different ways tomanage dairy ewes. In the very inten-sive Mediterranean systems, ewes aremanaged the same way as dairy cows.Weaning occurs immediately afterlambing, followed by exclusive milkingto the end of lactation (150–200 days).In that case, the lambs are artificiallyreared and difficult to train becausethey never learned from theirmothers. In many cases, however, avariable suckling period precedesexclusive milking.

In the largest flocks, lambs are suckledto weaning. In smaller flocks, the pointis to provide colostrum cover and toawait the seasonal opening of special-ized creameries such as those of theRoquefort region in France. In thatcase, however, milk production inLacaune ewes, which has doubled in20 years and largely exceeds theintake capacity of the lambs in theearly stage of growth, no longerpermits exclusive lamb sucklingwithout hindering lactation. For thatreason, milking has been combinedwith suckling for complete mammarygland drainage and to train the ewesto come to the milking pen. How tochoose between those systems?According to Labussiere’s results, itappears that the more the ewessuckle their lambs, the greater diffi-culty they have giving their milk to themachine while releasing oxytocin.

The drop in milk yield observed atweaning (23–35% according to breed)is explained mostly by the reducedfrequency of daily drainage (-20 to -25%) but also by the mother-lambseparation effect (estimated at -3 to -7%). Mixed management never reallyreduced the drop in milk yield atweaning.

This is clearly explained becauserecent studies have shown that aslong as the ewe has daily contact withher lamb, she refuses to releaseoxytocin at milking. However, she doesit without any problem when suckled(selectiveness). The proximity of thelamb in the milking pen (unfeasible inpractice) restores the milk ejectionreflex, which demonstrates the neces-sity of the lamb effect (most probablyolfactory and visual) for milk ejectionto occur. Ewes however get used tothe milking pen and passing to exclu-sive milking is made easier by theircalmness. As early as 48 hours afterlamb separation, the ewes beginreleasing oxytocin at milking, contraryto exclusively suckling ewes, a propor-tion of which will never adapt. Therate of adaptation to milking is alsothe same as that observed in ewesturned to exclusive milking uponlambing. This latter method howeverhas to be considered with caution. Ourstudies show that oxytocin release isless effective if the mother does notestablish her maternal instinct; that is,

if the first sucklings are not per-formed. So a 24-hour maximumcontact between ewe and lambs isbeneficial, and the lambs remain easyenough to train for artificial suckling.

Although this remains to be verifiedexperimentally, our results and thoseof “controle laitier” (EN: official milktesting) would tend to show milk yieldhigher as mixed management lastslonger. These results could be easilyexplained by the establishment andrepeated stimulation of a strongejection and secretion reflex, effectivein early lactation. Lastly, mixed man-agement permits functional selectionbased on the morphology of theudder and teats because as a rule theewes not capable of suckling theirlambs are removed from the flock.Consequently, flock homogeneity isgreater and milking is easier. Althoughno reliable data are available in thatrespect, the users of the variousmethods have not reported any signif-icant effect on udder pathology.

84


Apparatus used for half udder studies by Professor Marnet in Rennes, France.

Machine milking The milking machine must be stimu-lating enough to ensure strong milkejection during the very short milkingtime. Also, ewe milking includes time-consuming manual operation (strip-ping and possibly “re-milking” thatshould be reduced to a minimum, inparticular by selecting well formedudders).

Proper setting of the machinehowever may help increase the pro-ductivity of the milker and at thesame time simplify his task. Initially,choosing a pulsation rate as high as180 ppm was motivated by the needto emulate the natural condition oflamb suckling as best as possible.

All our experiments aimed at compar-ing pulsation rates from 60 to 180ppm have shown that milk yield isvery slightly higher when the rate isset above 120 ppm. The mechanicalmilking and stripping milk volumes donot vary significantly, but the mostspectacular effect is an increase in there-milking volume at 60 ppm, whetherthe pulsation ratio is 33% or 50%.Pulsation ratio trials tend to show thatratios below 50% would incompletelydrain the teats. Oxytocin assayselicited significantly lower release inthat case, and it can thus be con-cluded that a pulsation rate below 120ppm is too low to stimulate Lacauneewes and does not ensure totaldrainage and retrieval of all the usablematter. It should be noted also thatcup drop is more frequent (withrubber liners) when the pulsation rateis low.

The vacuum pressure chosen isbetween 36 and 53 kPa. The mostrecent tests we performed showedthat the vacuum effect is mainly sensi-tive on the milk retrieved after strip-ping. This may be due to a disruptionof mammary drainage induced by teatelongation, liner clambering and veryobvious congestion of the teats. Thiseffect therefore is more a physical one.However, considering that the vacuumpressure setting is a trade-off dictatedby the weight of the bundle and theneed to prevent it from falling off, thesolution could be to operate underlower vacuum pressure (36 kPa) withlighter bundles and better grippingliners (silicone). However, with no airintake at the clamp, the rated vacuumpressure under the teat may tran-siently exceed the regulator pressureand damage the teat while increasingthe leukocyte count. It is therefore rec-ommended to maintain some airintake, even if it means increasing thevacuum reserve slightly.

Note that with such a setting theleukocyte count will be higher thanwith a lower pulsation rate. There is noupper aggression on the mammarytissue. In fact that effect is only sensi-tive in animals with leukocyte countsabove 200,000 cells per milliliter(sterile controlled milk). So theincrease in leukocyte count is only theresult of the expulsion of the cells con-tained in the alveoli, through whichthey enter the udder. This clearlyconfirms better drainage induced byoxytocin and permits earlier detectionof possible udder infections.

Finally, the choice of liner is crucial forthe optimum application of themachine settings to the teat. There isno impact on milk yield if the milkerperforms proper stripping. This meansthat the teat liner has to be chosencarefully to facilitate physical drainageto the udder. Silicone liners appear toreduce cup drops and liner slippingand are therefore recommended.However, the most spectacular effectsare produced by the design of the cupand the flexibility of the liner body.Stripping is highly reduced when thecup diameter is increased to restrictteat squeezing at the end of milking;otherwise air intake is facilitated andcup falls are more frequent. A veryhard liner may increase stripping con-siderably because it moves moreslowly and remains open longer thana softer liner. Indeed, milk outflow canbe accelerated but the effect of thaton the teat is deleterious (upper con-gestion) and the liner clambering ismore marked. The flaring pressure forewe liners is thought to be close to 10 kPa.

A number of factors are yet to betested or re-tested because of theongoing standardization of milkingequipment for small ruminants.Further advances are still possiblethrough blood oxytocin assays, meas-urements of teat congestion andudder immunological conditionassessment, as indices of the physio-logical effect of the equipment and ofudder health.

85

M A N A G E M E N T F O R B E T T E R M I L K Y I E L D & Q U A L I T Y C H A P T E R 7

The best trade-off would be

low vacuum pressure (36 kPa)

and high pulsation rate

(180ppm) with a 50% ratio.

ConclusionAdding the losses in milk and usablematter to those in milker time and dis-comfort that can be endured unknow-ingly when operating under poor con-ditions, the losses can add up toimpressive figures (up to 20–25%). It istherefore necessary to use animalswith good mammary conformation(large cisterns, well drained by verticalteats at their base), good sensitivity tostimulation by the milking machine,and good and sustained oxytocinrelease during and possibly betweenmilkings.

High milk output at milking will ensueand working time will be reducedaccordingly. In more intensivesystems, ewes exhibiting low maternalinstinct will be preferred to facilitateweaning and adaptation to mechani-cal milking. The equipment will beadjusted so as to be stimulating (highpulsation rate) and optimize oxytocinrelease and increase drainage effec-tiveness and non-aggressiveness (lowvacuum pressure) to avoid tissue con-gestion and poor teat drainage, whichwould necessitate additional manualoperations. All these operations, bybetter draining the udder, will ensureand maintain better milk yieldthroughout lactation, all the more soas they are performed frequently inearly lactation. In that respect, mixedmanagement appears to be an addi-tional asset if the right to produce isnot restricted.

ReferencesLabussière J. 1988. Review of the phys-

iological and anatomical factorsinfluencing the milking ability ofewes and the organization ofmilking. Livestock ProductionScience, 18, 253-274.

Labussière J, P.G. Marnet, J.F. Combaud,M. Beaufils and F.A. de laChevalerie. 1993. Influence dunombre de corps jaune sur lalibération d’oxytocine lutéale, letransfert du lait alvéolaire dans laciterne et la production laitièrechez la brebis. Reprod. Nutr. Develo.33, 383-393

Le Du J. 1996. Milking animals otherthan cows. In: Proceedings of theSymposium on Milk Synthesis,Secretion and Removal inRuminants, April 26/27, Berne ,Switzerland. Blum J.W. and R.M.Bruckmaier (Edts). 49-52

Marnet P.G. and J. Labussière. 1994.Intramammary pressure and lutealoxytocin after PGF2a administra-tion in cycling and early pregnantewes. J. Dairy Res. 61, 345-353.

Marnet P.G., H. Volland, P. Pradelles, J.Grassi and M. Beaufils. 1994.Subpicogram determination ofoxytocin by an enzyme immunoas-say using acetylcholinesterase aslabel. J. Immunoassay 15 (1), 35-53.

Marnet P.G. and J.F. Combaud. 1995.Pharmacological evidence for anoxytocin-like hormone in earlypregnancy corpus luteum insheep. Short paper in “Oxytocin:Cellular and Molecular approachesin Medicine and Research.” Editedby R. Ivell and J. Russell, PlenumPublishing Corporation, 233 SpringStreet, New York.

Marnet P.G., J.A. Negrao and J.Labussière. 1998. Oxytocin releaseand milk ejection parametersduring machine milking. SmallRum. Res., 28:183-191

86


The presence of mastitis in a dairyherd, either cows or sheep, canhave an important impact on the

financial returns of the operation.Mastitis (clinical or sub-clinical) lowersthe total milk production, changes thecomposition of the milk, and affectsquality.

More importantly, the pathogenbacteria found in the milk of animalswith mastitis can pose a threat tohuman health.Therefore, it is importantto reduce or eliminate the incidence ofmastitis.This occurs by first detectinganimals with the disease, treating orculling them, and by respecting somehygienic rules to prevent the recur-rence of the disease in the flock.

Mastitis

Clinical mastitisPeracute mastitisThis is the most well-known form ofmastitis, easily recognized and gener-ally characterized by acute and rapidinflammation of one side of the udder.The udder becomes hard, red and hot.No milk is secreted but there is a smallamount of a clear, sometimes bloody,malodorous liquid discharge. Theanimal appears to be in pain andpresents a high fever (41-42°C).Morbidity is high. Ewes can respond tohigh level of penicillin but the udderoften becomes gangrenous. In case ofrecovery, the affected side stays non-productive because of extensivedamage.

The incidence of peracute mastitis isgenerally no higher than 5%. It iscaused by microbial agents such asStaphylococcus aureus and to a lesserdegree by Streptococcus uberis andSteptococcus suis or by other agentssuch as Pseudomonas aeruginosa andAspergillus fumigatus.

Peracute mastitis in dairy ewes causedby Staphylococcus aureus occursmostly during the suckling period orimmediately following the period ofexclusive milking. Compared to dairycows, the drying off period is not ahigh-risk period for peracute mastitis.

Acute or chronic mastitisAcute mastitis can often be detected bypalpation and observation of the udder,(asymmetric udder, nodules in one ortwo halves, hypertrophy of lymphnodules) and/or by the appearance ofthe milk, which may contain flakes,purulent material or be discolored.

Chronic mastitis may result from aspontaneous healing of a case ofperacute mastitis. Ewes with chronicmastitis should be removed from themilking herd because of their lowermilk production, but mostly becausethey represent a reservoir of infectiousagents.

87

Mastitis: detectionand control

Chapter 8

The pathogen

bacteria found in

the milk of animals

with mastitis can

pose a threat to

human health.

A milking ewe with peracute

mastitis should immediately

be removed from the milking

group.

Sub-clinical mastitisSub-clinical mastitis is an insidiousinfection of the udder characterized byan inflammation that is not easilydetected at its early stage.The animalmight appear physically normal and itslower production might go unnoticed.

Sub-clinical mastitis in dairy ewes (asin goats) is due to Coagulase NegativeStaphylocci (CNS) mostlyStaphylococcus epidermis that live onthe skin of the udder or on the teat.The proximity of the source of infec-tion makes it very easy for a largenumber of ewes to become infectedwith the bacteria. Rates as high as30–40% have been reported onManchega and Assaf ewes by LasHeras et al (1998); 20–30% in theRoquefort area of France (Bergonier etal. 1997); but also as low as 7–8.5% in aflock of Corriedale milking ewes inUruguay (Apollo et al. 1998).

Sources of infectionThe principal reservoirs ofStaphylococci are the infected uddersas well as the infected lesions of theteats. They can also be found on theskin of the udder and can survive along time in the milk lines of themilking machine even when they arecorrectly cleaned. Other causingagents of mastitis can be found inbedding, molded forage and water. Itis therefore easy to understand thatthe spread of bacteria can occur veryeasily from one animal to anotherwhen the hygienic conditions are lessthan satisfactory.

Treatment of mastitisIn case of peracute mastitis due tomicrobial infections, the objective is toprevent the animal’s death by admin-istering high doses of penicillin.Success is not always guaranteed. Ifthe animal recovers, the affected halfudder always presents lesions andbecomes non-functional, justifying theimmediate removal of the infectedanimal from the milking flock.

Treatment of sub-clinical mastitis ismore ambiguous because it necessi-tates the detection of the infection. Indairy ewes the spontaneous elimina-tion of the infection by natural biolog-ical defenses of the animal has beenestimated at 60% in a dry-off period offour months (Bergonnier et al. 1997).Therefore, a large reservoir of animalsthat can carry the infection from onelactation to the next and be a sourceof infection may still be present.

Antibiotic treatment by intramam-mary injection of antibiotic after dry-off is a viable solution in flocks withhigh incidence of sub-clinical mastitis.The procedure has to be performed inthe utmost hygienic conditions toavoid spreading infection to healthyanimals.

Prevention, detection and eliminationof problem animals should be apriority. The detection of animals withsub-clinical mastitis is fundamentalbut not easy. Bacteriological examina-tion of the milk can be done but it is along and expensive process. It cannotbe done to detect mastitis on aregular basis. Somatic cells in the milkon the other hand can be observedrapidly and counted with electronicdevices (Fossomatic type) which countall nucleated cells present in the milkincluding epithelial cells.

88


Reduce mastitis by observing

the following rules:

■ Detect infected animalsearly; then follow up witheither a treatment orculling.

■ Wash hands frequentlyduring milking. Milkersshould wear latex gloves todecrease the possibility ofspreading bacteria fromone udder to the other.

■ Shut off of the vacuum linewhen removing the teatcups to avoid possibleinfected milk dropletsreaching the teat openingof the next ewe.

■ Correct vacuum level andpulsation.

■ Do not over-milk, which cancause trauma to the teatand increase susceptibilityto infection.

■ Clean the milking machinethoroughly.

■ Clean air lines thoroughly.

■ Change teat cup liners andmilk lines periodically.

■ Provide abundant freshbedding for ewes in con-finement.

■ Clean the waterers.

■ Conduct a post dippingprogram.

The incidence of mastitis

varies with the level of hygiene

and the milking technique of

the producer.

Somatic cellsSomatic cells are an integral part ofmammary secretion and are foundcommonly in milk. According toRanucci and Morgante (1997) somaticcells in the milk of healthy sheep areconstituted of macrophages(55–70%), of PolymorphonucleatedNeutrophyl Leukocytes or PMNL(15–40%) which have the importantbiological function of phagocyticactivity, of lymphocytes (6–14%) andof other cell types (eosiniphils, epithe-lial cells and non identifiable cells) inlower percentage (0–5%).

When the mammary gland becomesinflamed, different cell types remainthe same but change markedly intheir distribution. PMNL become moreprevalent, as do lymphocytes in thecase of staphylococcal mastitis (themost common mastitis in dairysheep). Therefore, the main variationsof somatic cells as well as theirincreased number are caused by aninflammatory condition of themammary gland that may be clinicallyevident or sub-clinical. All studiesindicate that a significant increaseof SCC in milk represents an inflam-matory condition of the mammarygland.

Because of the strong relationshipbetween udder health and theamount of somatic cells in the milk,regulations in most countries have puta limit in the SCC above which milkcannot be marketed or above whichpenalties in terms of payment havebeen imposed by milk processors.

A rise in somatic cell count is notalways due to an infection. There aresome non-infectious factors thatmight have an impact on the detec-tion of infection.

Non-infectious factors ofvariation of SCCThe non-infectious variation factors ofsomatic cell counts have been wellreviewed by Bergonier et al. (1994).

Variation betweenmilking fractionsStripping milk, (milk obtained at theend of milking by massaging theudder) contains 1.7 times more SCCthan the foremilk, milk obtainedbefore applying the teat cups. Themilk sample taken for the counting ofsomatic cells should therefore betaken from the total milk obtainedduring milking.

Variation betweenmorning and eveningSomatic cell count is generally greaterin the evening milk than in themorning milk. The differences are par-ticularly important when the intervalsbetween milkings are irregular. Thegreater the interval between eveningand morning milkings, the larger thedifference. To minimize discrepanciesbetween samples it is important tokeep a constant milking interval.

Variation between daysor weeksVariations from one day or one weekto the other have been reported butnot always quantified in dairy sheep.In cows, daily variations of 20–30% arenot uncommon. Those variations aredue to unknown causes. Therefore it isnot possible to conclude that ananimal has an intramammary infectionbecause of higher SCC at the view ofonly one sample.

Variation during the lactationIn healthy dairy ewes put at exclusivemilking after a 30-day suckling period,the SCC is generally elevated at thestart of milking.This is due to the stresscaused by the change of several udderevacuations per day (suckling) to onlytwice a day (milking).The SCC rapidlydecreases after a week or two and staysstable for a good portion of the lacta-tion. At the end of lactation, SCCincreases either progressively orbrutally when close to dry off. AverageSCC of 50,000–200,000/ml in mid lacta-tion can rise to 250,000–500,000/mlduring the last month of lactation orclose to dry off.This increase in SCC isequivalent to an elevation due to aninfection with Staphylococci. Becauseof the variation of the SCC during thelactation (high-low-high) it is impossi-ble to determine if an infection existsbased on controls performed at the

89

M A S T I T I S : D E T E C T I O N & C O N T R O L C H A P T E R 8

A high somatic cellcount is generallyan indication of aninflamed udder.

Clinical or acute mastitis (sick animal)

(udder inflammation)

Death of eweDeath of lambs

Costly treatment

Damage to udderCulling of ewe

Lower productionMilk and cheeseQuality affected

High SCC

Sub-clinical mastitis(apparently

healthy animal)

RecoveryChronic mastitis

beginning or at the end of lactation.Yet, this is the period when decisionsmust be made about removing animalsthat present a risk of infection.

Variation due to theparity numberSeveral studies have shown that thesomatic cell count increases signifi-cantly from the first lactation to thefourth lactation (an average of 20,000/ml per lactation). However it has notbeen clearly demonstrated whetherthe increase in SCC corresponds to anincreased level of infection thatbecomes greater with age.

Variation due to nutritionNo studies exist that have shown arelationship between somatic cellcount and level of energy or protein inthe ration. However, it seems thatthere is a correlation between SCCand the level of Vitamin A, and b-carotene on one hand, and Vitamin Eand selenium on the other.

Variation between individual animalsThere is a definite genetic variationamong animals regarding mastitisresistance. In dairy ewes, few resultsare yet available but many studies arebeing performed to determine thegenetic parameters of resistance tomastitis as described in Chapter 3.According to preliminary results foundin dairy sheep, and based on what isknown in dairy cows, it could bepossible to select animals for resist-ance to infectious mastitis.

SCC and mastitisThere is no doubt that somatic cellcount is an indication of mammaryinflammation but the number ofsomatic cells is under the influence ofmany factors (infectious and non-infectious). It is therefore necessary todetermine the level of SCC for a eweto be declared infected with relativecertainty. Other questions such as therelationship between the SCC in themilk of the bulk tank and the individ-ual SCC are important for strategicplanning at the farm level.

Threshold of infectionAfter reviewing several studies,Bergonier et al. (1997) proposed thatthree classes be defined to predict thepresence of an infection with accuracygreater than 80%.

■ Healthy udder: If all SCC controlsbut one are inferior to 500,000cells/ml.

■ Infected udder (sub-clinicalmastitis): If at least two SCCcontrols are superior to 1,000,000cells/ml.

■ Doubtful udder (temporaryinfection): In all other cases.

The lower limit between “healthy” and“doubtful” is between 400,000 cells/mland 500,000 cell/ml. The higher limitbetween “doubtful” and “infected” isbetween 800,000 cells/ml and1,000,000 cells/ml.

Relationship betweenSCC in milk of bulk tankand individual SCCLagriffoul et al. (1998) report that thecorrelation between bulk tank SCCand individual SCC is 0.86 meaningthat an evaluation of the bulk tankSCC could be used to estimate theincidence of udder infection in theflock. Moreover, by considering indi-vidual SCC by the quantity of milk thateach ewe contributes to the bulk tank,the same authors have been able toestablish the proportion of ewes inthe flock contributing to the total SCCof the tank (table 1).

The percentage of ewes with high SCC(>1,000,000) increases with the SCC ofthe tank to the detriment of the per-centage of ewes with low SCC(<100,000) while the percentage ofewes with intermediate SCC (>100,000but <1,000,000) stays relativelyconstant. Although informative,Lagriffoul et al. (1998) warn that thefigures presented in table 1 do notreflect the infection status of the flock.

90


Table 1. Evolution of the percentage of ewes with less than 500,000 cells/mlaccording to the SCC of the tank.

Number of % of ewes with % of ewes withsomatic cells Average SCC < 500,000 > 1,000,000in the tank in the tank cells/ml cells/ml

0 to 400,000 257,000 90.1 5.1

400 to 600,000 492,000 83.5 9.2

600 to 800,000 693,000 78.9 12.5

800 to 1,000,000 895,000 74.8 15.5

1,000 to 1,400,000 1,160,000 68.2 20.6

> 1,400,000 1,865,000 56.7 30.9

Lagriffoul et al. 1998

To estimate the incidence of mastitisin the flock, it is necessary to use theregression equation shown in table 2with the average of all tank SCC takenduring the season. Table 3 gives anexample using different SCC values.

Though SCC is an excellent tool fordetecting sub-clinical mastitis, therepeated controls of SCC at the farmlevel can be cumbersome and expen-sive. However, it is indispensable todetect infected ewes, especially whenthe tank SCC (controlled monthly bystate agencies) rises above 500,000cells/ml. The limit allowed in theUnited States is 1,000,000 cells/ml butsome milk buyers can penalize milk ata much lower level. If controlling SCCis not possible, detection of inflamma-tion can be done with the CaliforniaMastitis Test.

California mastitis testWith the California Mastitis Test (CMT),a producer can evaluate the cellcontent of milk rapidly and cheaply.According to studies reviewed byBergonier et al. (1994), there is a goodrelationship between CMT and SCC,especially if a simplified grid is used.CMT gives 87% of right results for thenegative values (<250,000 cells/ml)and 92% of right results for positivevalues (>250,000 cells/ml). As for SCC,CMT has to be performed severaltimes through the lactation to takeinto account the non-infectiousfactors of variation, and the rule of thethree classes has to be respected.

The California Mastitis Test was devel-oped as a “cow-side” indicator ofmastitis. Essentially, it permits earlyindication of inflammation or poorudder health, and can be used reliablyin dairy ewes.

The test is easy to perform. Beforeputting on the teat cups each udderhalf is sampled and evaluated sepa-rately by squirting milk into a shallowpaddle device which contains aspecial reactive agent. The milk isswirled and the resulting clot forma-tion is subjectively graded to deter-mine the relative amount of inflam-mation (presumed infection) in theudder half. Results are generallyreported as:

■ Negative—no jelling seen at all

■ Trace—small amount of gel seenwhen tipped (<250000 cells/ml)

■ 1+—significant amount of gelseen when tipped

■ 2+—when paddle is swirled, geltends to clump in the middle

■ 3+—mixture completely jelled andclumps in middle when swirled.

91

M A S T I T I S : D E T E C T I O N & C O N T R O L C H A P T E R 8

■ When the percentage ofinfected ewes increases, thepercentage of healthy ewesdecreases, but the percent-age of doubtful ewes staysmore or less constant.

■ Although small in number,infected ewes contributegreatly to the elevation ofSCC in the tank.

■ The percentage of infectedewes increases in average of2.5–3% for each incrementof 100,000 cells/ml in thetank.

Table 2. Prediction of percentage of ewes in different classes of infectionaccording to the average SCC of the tank.

Regression equation R2 test

Healthy -0.028 X + 89.272 0.699 p<.001

Doubtful 0.004 X + 11.772 0.035 ns

Infected 0.024 X – 1.044 0.845 p<.001

X = SCC in milk of tank/1000 Lagriffoul et al. 1998

Table 3. Example of percentage of ewes with infection according to 3 differenttank SCC using table 2.

Average SCC of the tank during the season (103cells/ml)270 675 1,100

% healthy udders 81.7 70.4 58.4

% doubtful 12.8 14.4 16.2

% infected udders 5.4 15.2 25.4

ConclusionSub-clinical mastitis is mostly due toCoagulase Negative Staphylococcisuch as Staphylcoccus epidermis. Itsincidence can be fairly high when thehygienic conditions on a farm are notsatisfactory. Reducing mastitisdepends principally on good manage-ment techniques, good milking proce-dures, maintenance of milkingmachine and the elimination ofinfected animals. The detection ofinfected animals can be done throughSCC or CMT performed several timesduring the lactation.

ReferencesAppolo A., T. Bellizi, D. de Lima and M.

Burgueno. 1998. Subclinicalmastitis in dairy sheep in Uruguay.In: Proceedings of the 6thInternatinal Symposium on themilking of small ruminants, Athens,Greece, September 26-October 1,1998. EEAP Publication, No 95,1999, 168-170, Wageningen Press.

Bergonier D., G. Lagriffoul, X. Berthelotand F. Barillet. 1994. Facteurs devariation non-infectueux descomptages de cellules somatiqueschez les ovins et caprins laitiers. In:Proceedings of the InternationalSymposium on somaic cells andmilk of small ruminants. Bella, Italy,September 25-27, 1994. EAAPPublication No. 77, 1996.

Bergonier D., M.C. Blanc, B. Fleury, G.Lagriffoul, F. Barillet and X.Berthelot. 1997. Les mammites desovins et des caprins laitiers: étiolo-gie, épidémiologi, contrôle. 4èmeRencontres autour des Recherchessur les Ruminants. Paris, December4 and 5, 1997. Publication INRA-Institut de l’élevage.

Lagriffoul G., F. Barillet, D. Bergonier, X.Bethelot and M. Jacquin. 1998.Stratégies de contôle en ferme descomptages de cellules somatiquesdu lait de brebis et de chèvre”.Programme FAIR CT 95-0881.Objectif 4. Tache 11.1. Définition deseuils de CCS de troupeaux ovins.

Las Heras A., J.F. Fernandez-Garayzabal, E. Legas, I. Lopez and L.Domingues. 1998. Importance ofsubclinical mastitis in milkingsheep and diversity of etiologicalagents. In: Proceedings of the 6thInternatinal Symposium on themilking of small ruminants, Athens,Greece, September 26-October 1,1998. EAAP Publication, No 95,1999, 137-141, Wageningen Press.

Ranucci S. and M. Morgante. 1994.Sanitary control of the sheepudder: total and differential cellcounts in milk. In: Proceedings ofthe International Symposium onsomatic cells and milk of smallruminants. Bella, Italy, September25-27, 1994. EAAP Publication No.77, 1996.

92


Ewes can be milked by hand asthey stand on elevated platformsor by machine in parlors with

various levels of sophistication. Handand machine milking share the samegoals—keeping the milk clean andemptying the udder as completelyand quickly as possible without trau-matizing the udder or teats.

The conditions in which cows, goats

and sheep are milked in NorthAmerica are subject to many regula-tions to ensure that consumersreceive the highest quality product.The location of the milking parlor, itsdesign, the materials used for con-struction, the quality of the water usedto clean the milking equipment andmany other details need to berespected to qualify for a milkproducer’s license.

Types of milkingparlors

Hand milking (less than 20 ewes)Hand milking is still very popular inmany Mediterranean countries wheremanagement systems, labor andenergy resources (such as electricity)are quite different than in other coun-tries. A shepherd can hand-milkbetween 20 to 60 ewes per hour(sometimes more) depending on thebreed and the milk yield of the ewes.Until 1950, only about 20 Lacauneewes per hour could be milked(because the breed is difficult to milk)while 80 ewes per hour was possiblein Corsica where local breeds wereeasy to milk.

Because of the generally small size ofthe sheep dairy operations in NorthAmerica, hand milking can be a veryviable option. It is simple and does notrequire sophisticated equipment. Butthere are other considerations. For

93

Milking parlors and equipment

Chapter 9

Hand milking can

be a very viable

option.

It is very important to milk

rapidly in as clean an environ-

ment as possible. The area

must also be well-lighted and

designed for the maximum

comfort of both ewes and

milkers.

Before starting to build or

install any type of milking

system, producers should

contact the local dairy inspec-

tor to learn about existing reg-

ulations. For example, in the

United States, stanchions

cannot be built with porous

materials such as wood.

example, in a small-scale operationequipped with a milking machine,workers spend more time washing theequipment than actually milking.Hand milking also requires a certaintechnique that might be hard toacquire or might not appeal to themodern producer. Moreover, it is diffi-cult to get clean milk because ofpossible contamination caused byexternal agents.

Bucket milking andelevated platform(between 20 and 120ewes)Many sheep dairy producers in theU.S. and Canada have adopted thebucket-and-fixed-stanchion system onan elevated platform because of themodest investment it requires. Thestanchion is generally on an elevatedplatform with 6 or 12 fixed stalls with“cascading” yokes.

The first ewe on the platform goes tothe farthest end where a yoke is open.By putting her head through to get tothe feed, the animal locks herself inand releases the mechanism thatopens the next yoke and so on untilthe whole platform is occupied. Thesystem works fairly well and is easy tobuild at low cost. The feed is generallydistributed by hand between eachgroup of ewes milked.

The milking is performed in bucketsdeveloped for cows or goats. Thevacuum in the bucket is provided by avacuum pump located either next tothe bucket (as a wheel barrow system)or in an adjacent room, and thepulsator is fixed on the lid of thebucket, which is also equipped with afilter. Two ewes at a time can bemilked with the same bucket.

Some disadvantages arise with bucketmilking:

■ The vacuum level is not alwaysconstant. This can lead to anunusually high incidence of sub-clinical mastitis reflected by anelevated somatic cell count.

■ The pulsators are often old and notadjustable to the required speedfor ewes. This limits the stimulationnecessary for maximum evacua-tion of the udder.

■ The milk may not cool rapidlyenough.

■ The buckets can be heavy to haulby hand.

■ The speed at which the ewes aremilked may not be fast enough tobe profitable. One milker can milkonly 40 ewes per hour.

■ The milking process can lead tobad posture and resulting physicalproblems for milkers.

■ The time involved in cleaningequipment can be substantial.

The same type of stanchion can beused with a low line or high linepipeline which greatly facilitatesmilking, reduces the heavy lifting andpermits the use of equipment betteradapted to sheep milking.

Parlors for larger flocks(more than 120 ewes)The Casse SystemThe Casse parlor was born in 1961 inan experimental farm in the Roquefortarea of France called Casse farm. Thisparlor was developed for the Lacaunebreed and the special working routinerequired by the ewes’ poor milkingability. The typical milking routine atthat time involved:

■ Attaching clusters on teats withoutwashing.

■ Hand massaging after one minuteof milking.

■ Machine-stripping and detachingafter 180 seconds of milking.

■ Re-milking by hand for 10 or 20seconds.

94


Homemade six-stall elevated platform Small pump

Around 1960, only 80 ewes per hourcould be milked with this routine bytwo milkers in a 12-milking-unit, 24-stall parlor. Much progress has beenmade since then with the developmentof better equipment and a simplifiedmilking routine without massaging,stripping and re-milking by hand.

The Casse System is a side-by-sideparlor developed from the herring-bone parlor, which was in the earlystage of development in the latterpart of the 1950s. In a Casse parlor,ewes enter and walk to a mangerwhere a concentrate is distributedeither manually or automatically; theyare locked by their necks in specialyokes. The animals go to any headlockthey want; the other ewes can moveon the platform behind those that arealready locked in and eating concen-trates (figure 1). When the platform isfull, the milker moves the ewes backto the edge of the pit, manually with acrank or automatically with a pneu-matic device (figure 2). The through-put is generally 120 ewes/hour withone milker in a typical Casse System(24 ewes/12 milking units).

New Casse milkingparlorsToday, larger flocks are milked inmodern and highly efficient parlors.The new Casse system has fixed stallsinstead of movable stanchions (figures3 and 4). A gate moves on theplatform when ewes are entering. Itstops at the first place, then an auto-matic feeder distributes concentratesand the gate opens.

The first ewe enters the first spotequipped with an automatic headlock.When it is locked, the gate moves backto the next spot, the second eweenters the second headlock and so on.

The gate carries a special curtain thatrestrains ewes from going to an unoc-cupied headlock. This new systemallows for the possibility of distribut-ing the exact amount of concentrateto each ewe according to her level ofproduction. The movable gate isequipped with a transponder readingthe electronic ear tag of the ewe andgiving information to the feeder(through an interface system with acomputer) about the amount of feedto pour in the individual trough. Thissystem called ADC (automatic distri-bution of concentrate) solves thelogistical problem of properly feedinga large number of ewes that are in dif-ferent stages of lactation and levels ofproduction. As a general practice, feedwas distributed in an amount suffi-cient to cover the nutritional needs ofthe highest milking ewes leading to awaste of energy and proteins amongewes not requiring the same amount.

95

M I L K I N G P A R L O R S & E Q U I P M E N T C H A P T E R 9

Figure 1. Casse system with ewes entering the platform

Figure 2. Casse system with ewes ready to be milked

Six-stall elevated platform

crank pit

movable stalls

crankcrank pit

movable stalls

Most of these parlors, now verypopular in France, have 2x24 stalls with24 units and a high line pipeline.Twomilkers work in them except whenautomatic teat cup removal is used;then only one is needed. A dog usuallyhelps ewes entering the platform.

Other popular milking parlors inFrance are rotary parlors. These gener-ally have 30 units or more (from 30–48places, sometimes as many as 60) andare used only in big flocks of morethan 500–600 ewes with two milkers.Most are now equipped with ACR(automatic cup removal). Table 1 givesa survey of the different styles ofmilking parlors used in the Roquefortarea in 1997.

Other types of parlors The parlors described so far rely onfeeding animals in the parlor—with thefeed serving as the ewes’ reward forcoming in. All these parlor styles workwell but have the common disadvan-tage of being expensive and compli-cated.To reduce the initial cost of theinstallation, some North American pro-ducers have replaced the self-lockingstanchion system with the “crowdingsystem,” developed in New Zealand fordairy cows. A certain number of ewes(12, 18, 24) come in the parlor and aresqueezed side by side on the platform,stopped from moving forward by asimple bar.The feed concentrate is gen-erally distributed by hand.The feedingcan also be done outside the parlorafter milking.The milking parlorbecomes an obligatory passage for theewes to get to the feed and they there-fore enter willingly.

Throughputs in different parlorsToday, the most popular milkingparlors in France are Casse system,both old and new. These are designedwith 2x12 stalls with 6 or 12 milkingunits and 2x24 stalls with 12 or 24units. Producers with large flocks needequipment (and especially parlors)with a high degree of efficiency. Themain parameter to consider whenchoosing a new parlor is its potentialthroughput; that is, the number ofanimals coming efficiently in and outin a certain amount of time. Many fieldstudies and inquiries are regularlymade to give farmers guidelines asthey choose their parlors.

In old Casse systems, the averagethroughput observed in field studiesis between 100 and 350 ewes/hourdepending on the number of units,the number of milkers, the daily milkyield and the number of ewes perunit.

96


Figure 3. New Casse parlor with ewes exiting and entering

Figure 4. New parlor with ewes taking their place and ready to be milked

automatic feeder

24 units Bayle HL parlor (half)

curtain

entryexit

pit

fixed stalls

24 units Bayle HL parlor (half)

entryexit

pit

automatic feedercurtain

fixed stallsurtainurtain

Table 1. Number of milking parlorsused in the Roquefort area in 1997

Type Number %

Bucket 10 0.4

Classical Casse 1725 75

New Casse 230 10

Rotary 335 14.6

Total 2300 100

Field studies in the Roquefort areahave shown that parlors with a highline pipeline are more efficient thanparlors with a low line. Doubling thenumber of units only increases thethroughput by about 20–25%. This isthe reason most parlors in theRoquefort area have high pipelines,though low-line parlors also exist.

For small flocks, it is possible to buildonly one platform to limit costs. Theefficiency of such parlors is about 100to 200 ewes/hour with only onemilker (table 3).

Modern Casse parlors are more effi-cient. Table 4 shows that in a 2x24place with 24 units, average through-put could be anywhere between 320and 420 ewes/hour with two milkers.Most of the parlors with 2x24 placesand 24 units are now equipped withACR (Automatic Cluster Removal). Insuch parlors, one milker can milkbetween 350 and 400 ewes/hour(table 4).

Finally, rotary parlors with a largenumber of units are certainly the mostefficient parlors, but they are also themost expensive. Producers with morethan 500 ewes are the primary usersof rotary parlors. Table 5 shows that itis possible to milk 420–650 ewes perhour depending on the number ofunits, the number of milkers and thedaily milk yield of ewes.

Organization oflaborThe Casse system is based on thepremise that the number of unitsdepends on the time it takes themilker to attach clusters to all ewes,plus miscellaneous and idle time, andcome back to the first ewe withoutovermilking. Currently, the averagetime it takes to milk a Lacaune ewe isabout 3 minutes, depending on milkyield (2.5 minutes in mid-lactation and2 minutes at the end of lactation). This

means that a milker can work effec-tively with only 12 units. For parlorswith more than 12 units, a secondmilker or ACR (Automatic ClusterRemoval) is needed.

Each milker works in half the pit onone side of the parlor. For example,milker #1 attaches cluster numbers1–12. Simultaneously milker #2attaches cluster numbers 13–24. Thenreturning to the first ewe, the milkersmassage the udders in the sameorder.

97


Table 2. Average throughput in most popular Casse parlors

number number number number average of stalls of units milk line of milkers of pushers throughput

2 x 12 6 Low line 1 0 100–140

2 x 12 12 High line 1 0 180–250

2 x 12 12 Low line 1 1 140–200

2 x 24 24 Low line 2 0 220–300

2 x 25 24 High line 2 1 270–350

Table 3. Average throughput in one-platform Casse parlors

number number numberof stalls of units milk line of milkers throughput

1 x 12 6 High line 1 100–120

1 x 12 6 Low line 1 90–110

1 x 24 12 High line 1 140–200

1 x 24 12 Low line 1 120–180

Table 4. Average throughput in modern Casse parlors

number number number number average of stalls of units milk line of milkers of pushers throughput

2 x 24 24 HL 2 0 360–420

2 x 24 24 LL 2 0 320–400

2 x 24 24 HL 1* 1** 350–410

* with ACR, ** the pusher can be a dog

Table 5. Average throughput in rotary parlors

number number number averageof units of milkers of pushers throughput

32 2 1** 420-460

36 3 1** 450-500

48 2-3* 1** 600-650

* 1 milker less with ACR, ** the pusher is often a dog

Today, massaging is very rare and notusually necessary thanks to geneticimprovements. Therefore, milkers stripewes only if needed and alwaysdetach clusters in the same order.After detaching the cluster from thelast ewe, the platform is emptied andmilkers swing the milking units to theother side of the pit and repeat thesame routine. Then the pusher (whichcan be a dog) helps ewes enter theempty platform so they are readywhen the milkers have finishedmilking the other side. In these condi-tions, producers can milk more than350 ewes per hour with a steadythroughput of about 450 ewes/hour.

Importance of goodwork postureA person milking a large number ofewes in a very short time, twice a dayover 6 to 7 months, must have a goodwork routine and maintain goodposture. Poor posture leads to armand/or backaches, spinal problemsand other troubles that make the taskunpleasant. Some general rules ofthumb for milkers are:

1. Stand up as straight as possiblewhen working.

2. Avoid bending forward whenattaching and detaching clustersor working on udders.

3. Never work under the level ofelbows.

4. Never work above the level ofshoulders.

Maintaining good posture andworking conditions depends largelyon good parlor design. One of themost important elements is thedepth of the pit. In addition to therules just mentioned above, a milkermust know the average height of theewes’ teats to be milked. For example,

in the Lacaune breed, the distancebetween the floor and the base of theteat is an average 32 cm for ewes withtwo and more lactations, and 30 cmfor ewes during their first lactation.When ewes are standing on theplatform ready to be milked, uddersmust be within easy reach of themilker, taking into account ergonom-ics and comfortable working anglesfor body and arms. That means about10 cm above the level of elbows witha maximum variation of 20 cm. Forexample, if a milker is 1.7 m tall (5’9’’),his elbows are located at about 1 m(3’5’’) from the floor. Therefore theheight of the pit should be .85 m(2’10’’). Table 6 gives an idea of thedepth of the pit, which should alwaysbe calculated in relationship to theheight of the milker.

Figures 5 and 6 give examples ofdimensions (in meters) of classic andnew Casse parlors.

98


Figure 5. Parlor with 2x12 places and movable stalls (classic Casse System)

Figure 6. Parlor with 2x24 places and fixed stall (new Casse parlor)

entryexit

pit

platform

platform

movable stalls

movable stalls

5.75.9

7.8

1.2

2.25

entryexit

2.01.0

11.8

4.1

.45

.30

.80

Table 6. Depth of the pit in a milkingparlor

Depth Height of milker of pit

<1.5m (< 5') .75m (2'6'')

1.5m-1.62m (5'- 5'5'') .80m (2'8'')

1.62m-1.72m (5'5''-5'9'') .85m (2'10'')

1.72m-1.82m (5'9''- 6'1'') .90m (3')

1.82m-1.92m (6'1''- 6'5'') .95m (3'2'')

> 1.92m (6'5'') 1m (3'4'')

Milking machinesFour parts of the milking machine willbe covered:

1. Effective reserve and vacuumpump capacity

2. Size of milk lines

3. Pulsation characteristics

4. Vacuum level

1. Vacuum pump capacityand effective reserveVacuum pumpsThe vacuum pump should be capableof meeting the operating require-ments (milking and cleaning) of allequipment. This is true whether theequipment operates continuously orintermittently.

The vacuum pump should have suffi-cient capacity so that the vacuumdrop in or near the receiver doesnot exceed 2 kPa (kilo Pascal)during the course of normalmilking. This includes attaching andremoving the teat cups and linerslips.

Capacity should be measured inaccordance with ISO 6690: clause 5.3.

Effective reserveEffective reserve for milking machinesfor small ruminants should take intoaccount the special milking routinesused with these animals. In fact, mostfarmers do not shut off the vacuum atthe cluster or at the liner when theyattach or detach clusters from uddersso as to maintain high milking rates.

Effective reserve should compensateat least the total air admission of afully open cluster, evaluated at 600liters/minute when a milker puts theteat cups on or when a cluster falls off.

The air admitted depends particularlyon the type of cluster and the numberof milkers. If the clusters are equippedwith automatic shut-off valves, thetransient air admission is minimized,but may require extra air.

The installation requires a minimumeffective reserve determined in accor-dance with table 7. The extra airleakage is necessary for using ancillaryequipment. The manufacturer gener-ally states the maximum air leakage ofthe cluster equipped with an auto-matic shut-off device.

The effective reserve shall bemeasured in accordance with ISP6690, clause 5.2.

Influence of altitudeFor installations at altitudes of lessthan or equal to 300 meters, an atmos-pheric pressure of 100 kPa should beassumed for calculating effectivereserve.

To fulfill the requirements at altitudesgreater than 300 meters, a vacuumpump with increased capacity shouldbe installed.

Air demand for cleaningMilk and transfer lines are usuallycleaned by a mixture of air andcleaning solution transported andagitated by the vacuum difference toachieve effective cleaning by slugspeeds of 7 m/s to 10 m/s.

Where washing systems rely on highpump capacity to achieve the airspeed necessary to produce slugs forwashing this capacity, Qclean, in litresper minute, can be calculated from thefollowing formula:

Qclean = pd2 /4 *v*(pa -pw)/ pa.

Where:

d = internal diameter of the line, indecimetres,

v = air and slug speed in the milktube, in decimetres per minute,

pa = actual atmospheric pressureduring the test, in kPa,

pw = vacuum level when washingthe plant, in kPa.

Because of the low vacuum milkinglevel, milking installations for smallruminants can be washed at a highervacuum level to ensure a goodcleaning.

To estimate the minimal vacuumpump capacity, effective reserveneeds to be calculated from tables 7and 8.

99


Table 7. Minimum effective reserve (1) for different type of clusters (inlitres/minute of free air)

Type of cluster Number of units Pipelines Buckets

Conventional with n < or = 20 400 + 200 M + 20 n 200 + 100M + 20 nautomatic shut-off or clamp on the long milk tube

n > 20 800 + 200 M + 10 (n-20)

Automatic shut- n < or = 20 400 + 50 M + n AL* 200 + 25M + n AL *off valve

n > 20 500 + 50 M + n(AL - 5) *

Automatic n < or = 20 400 + 50 M + 10 nshut-off valve

+ automatic n > 20 500 + 50 M + 5 ncluster removal

(1): Plus addition for ancillary equipment in accordance with clause 17.

AL = extra air leakage at the cluster with automatic shut off valve necessary for workingM = number of milkers, n = number of units

Example: predicting vacuum pumpcapacity

If we have:

a. A parlor with 12 units automaticshut-off valve at the liner situatedat 300 m above sea level.

b. One milker.

c. Working vacuum level: 38 kPa.

d. Milk pipe diameter: 48 mm.

e. Air admission in the clusters: 10l/min.

f. Air leakage in the clusters: 20l/min.

g. Number of pulsators: 6

h. Air consumption for each pulsator:25 l/min.

i. Vacuum level for cleaning: 50 kPa.

Then:

1. According to table 8 the effectivereserve capacity for milking is:

400 l/min. + 50 l/min. + (12 x 20)l/min. = 690 l/min.

2. The air demand for cleaning at 50kPa and an altitude of 1000 mshould be 386 l/min., which islower than the effective reserve formilking.

3. The air consumption for themilking units (air admission andpulsators) is:

(10 x 12) l/min. + (25 x 6) l/min. =270 l/min.

4. The total air demand duringmilking is:

690 l/min. + 270 l/min. = 960 l/min.

5. The total air demand duringcleaning is:

386 l/min. + 270 l/min. = 656 l/min.

6. In this example the capacity formilking is the larger and thereforethe base for the pump dimensioning.

7. Leakage into the milk system:

10 l/min + (2 x 12) l/min. = 34l/min.

8. Total:

960 l/min + 34 l/min. = 994 l/min.

9. Regulation loss is 10 % of themanual reserve. The effectivereserve was 690 l/min. and issmaller than the manual reserve.Consequently:

Manual reserve = 690 l/min. x100/(100-10) = 767 l/min.

Regulation loss: 767 l/min.x 10/100 = 77 l/min.

Total: 994 l/min. + 77 l/min. = 1071l/min.

10. Leakage into the airlines are equalto 5% of the pump capacity; thatis, vacuum system leakage: 1071l/min. x (5/100-5) = 56 l/min.

Total: 1071 l/min. + 56 l/min. =1127 l/min.

11. With a pressure drop of 3 kPabetween pump and measuringpoint, the vacuum level at thepump is: 38 kPa + 3 Kpa = 41 kPa.

Correction for the altitude of 300m and a vacuum of 41 kPa will givea correction factor of 0.80, that is,for an atmospheric pressure of 100kPa and a vacuum level of 50 kPa, anominal pump capacity of: 1127l/min x 0.80 = 902 l/min.

12. The minimum nominal vacuumpump capacity must therefore be902 l/min.

2. Size of milk linesISO Standards 5707 for cows describesa new method for sizing milk lines.

Stratified or waved milk flow in milklines should be the normal flow of themilk. Slugged milk flow, which inducesvacuum fluctuations in milk linesgreater than 2 kPa should be avoided.

Research shows good relationshipsbetween large vacuum fluctuationsunder the teat and higher incidence ofmastitis. University of Wisconsin-Madison studies (G. Mein and D.Reinemann) have shown that avacuum fluctuation of 2 kPa or less ina milk line has no effect on vacuumbeneath the teats. These studies gavethe maximum milk flow rate to keepvacuum fluctuations no greater than 2kPa in the milk line.

It is also possible to predict themaximum milk flow rate through themilk line with some information aboutkinetics of ewes’ milk ejection. With afive-second attachment rate and apeak flow of 0.8 l/min and 200 l/mintransient air admission, the maximummilk flow rate can be easily predicted(figure 7).

100


25

20

15

10

5

0

nb units

flo

wra

te(l

/mm

)

0 20 30 40 6010 50

Figure 7. Maximum predicted milk flow rate in milklines (peak flow : 0.8 l/min)

101


Table 8. Minimum effective reserve for milking, in l/min. of free air: clusters with automatic shut-off valves at theliner (examples)

_________Pipeline milking machines_________ _________Bucket milking machines_________

nb Air leakage at Air leakage at Air leakage at Air leakage atunits cluster: 20 l/min. cluster: 40 l/min. cluster: 20 l/min. cluster: 40 l/min.

1 milker 2 milkers 1 milker 2 milkers 1 milker 2 milkers 1 milker 2 milkers

2 490 540 530 580 265 290 305 330

3 510 560 570 620 285 310 345 370

4 530 580 610 660 305 330 385 410

5 550 600 650 700 325 350 425 450

6 570 620 690 740 345 370 465 490

7 590 640 730 780 365 390 505 530

8 610 660 770 820 385 410 545 570

9 630 680 810 860 405 430 585 610

10 650 700 850 900 425 450 625 650

11 670 720 890 940 445 470 665 690

12 690 740 930 980 465 490 705 730

13 710 760 970 1020 485 510 745 770

14 730 780 1010 1060 505 530 785 810

15 750 800 1050 1100 525 550 825 850

16 770 820 1090 1140 545 570 865 890

17 790 840 1130 1180 565 590 905 930

18 810 860 1170 1220 585 610 945 970

19 830 880 1210 1260 605 630 985 1010

20 850 900 1250 1300 625 650 1025 1050

21 865 915 1285 1335

22 880 930 1320 1370

23 895 945 1355 1405

24 910 960 1390 1440

25 925 975 1425 1475

26 940 990 1460 1510

27 955 1005 1495 1545

28 970 1020 1530 1580

29 985 1035 1565 1615

30 1000 1050 1600 1650

Criteria to consider when sizing milk lines in parlors:

• Slope (1% or more if possible)

• Transient air admission (200 l/min)

• Milk line looped or deadlined (looped is better)

• Attachment rate (depending on the milker, milking routine and of the number of milkers : generally 5 secondswith two milkers and 10 seconds with one milker).

Table 9 gives examples of milk linediameters for dairy sheep parlorswhich could be calculated accordingto the new method of ISO 5707:(attachment rate: 5 seconds, maximummilk flow rate 0.8 l/min, transient airadmission: 200 l/min).

3. Setting of milkingmachine: Pulsation characteristicsIn many countries ewes are milkedat a high pulsation rate—from 120to 180 pulsation/min. French studiesshowed that ewes milked with a lowerpulsation rate have a lower milk pro-duction, more strip yield and probablymore mastitis problems. Pulsationratio has not been studied precisely(few results are available) but it seemsthat 50/50 is the most popular ratioalthough an inverse ratio 45/55 canalso be found.

4. Vacuum levelVacuum levels for dairy cows havebeen decreasing for the 20 last yearsdue to sanitary and mastitis problems.This is also true for dairy sheep. Today,most milking parlors have adoptedthe following adjusted vacuum levels:

■ Low line parlors : 34 to 36 kPa (10to 10.6 inches Hg)

■ High line parlors : 36 to 38 kPa(10.6 to 11.2 inches Hg)

Maintenance ofmilking machinesMilking machines have a great influ-ence on the speed of milking, the bac-teriological quality of the milk and onthe udder health as indicated by theoccurrence of mastitis or high somaticcell count. It is absolutely necessarythat the milking machine be installedproperly and maintained regularly.

Cleaning of the milkingsystemThorough cleaning of all equipmentused during milking is the mostimportant chore in a dairy opera-tion. A good cleaning and disinfectingroutine is one that, with a minimum oftime, effort and cost, results in visiblyclean equipment and milk that consis-tently meets the buyer’s requirementfor bacteriological safety.

For bucket and hand-milking equip-ment, there is no real alternative towashing by hand, although the mostlaborious part of brushing the clusterscan be partially replaced by flushwashing.

With a pipeline system, sanitizingbefore milking and washing after-wards is carried out easily with acleaning-in-place system. The cleaningprotocol will be outlined by themilking machine’s manufacturer andshould correspond to the regulationsof the individual country.

In a pipeline system, all elementsshould be periodically dismantled(teat cups, liners, milk tubes, claws)and cleaned by hand to removeresidue buildups not taken away bychemical solutions. If the total bacteriacount of the bulk tank increases signif-icantly it means the milking equip-ment is heavily contaminated.

102


Table 9. Diameter of milk lines:example of calculations for a 1% slope

nb units/ flow rateslope (l/min) diameter

6 4.8 2''

12 9.6 2''

24 19.2 2.5''

30 21.5 2.5''

36 22.8 2.5''

48 22.8 2.5''

This is not a standard. It is only a sample calculation.

Dismantle clusterevery 2 weeks.

Dismantle teatcups every 2weeks.

Change linersevery year.

Check cleaning-in-place periodically.

Figure 8. Producers should pay particular attention to these checkpoints in the milking system.

103

Check milk line.

Check hose forcracks.

Check sanitarytrap for cleanliness.

Check receiverjar for cleanliness.

Check bulk tankfor cleanliness.

Check coolingsystem.

Check pulsatorsetting.

Check milkhose. Changeevery 2 years.

Check cleaning-in-place.

Check waterqualitiy.

Check for suffi-cient hot water.

Double sink.

The wholesystem shouldbe sanitizedbefore eachmilking.

Figure 8. (continued)

104


Figure 8. Check points of the milking system

Vacuum settingtoo hightoo low• mastitis• high SCC

Air line trapIf not well cleaned:• high bacteria in milk

PulsatorWhen not set properly: • longer milking time• mastitis• high SCC

Milk lineWorn outCracked:• bacteria

ClusterInadequate Improper position: • longer milking time• mastitis• high SCC• fall-off

Milk lineInsufficient diameterInsufficient slope Improper cleaning: • mastitis• high SCC• high bacteria

Vacuum settinginsufficient capacity• falling off of clusters• longer milking time• imcomplete milking• high SCC

Cleaning-in-placeIf deficient • unsatisfactory cleaning• high bacteria

Co

ntr

ol

Receiver jar

double vat

New Casse system with automatic take-off andautomatic swing-over.

Ewes taking their places in a New Casse parlor.Ewes in place in the New Casse 2 x 24 high linepipeline.

Ewes entering a New Casse parlor.

105


One x 12 cascading yokes and elevated platform.

Eight-stall cascading yokes and pit.

Twelve-stall cascading yokes and buckets (Olivia Mills, England).

Two x 24 units high line pipeline (France).

Two x 24 indexing stanchion (Casse system).

Two x 24 indexing stanchion (Casse system) lowpipeline (Old Chatham, New York).

Two x 16 crowding system low line pipeline(Kieffer, Wisconsin).

Two x 12 indexing stanchion, 12 milking units,high line pipeline (Spooner Agricultural ResearchStation).

Cost of milkingsystemsWhen building a new parlor, theproducer pays for milking machines,the milk room, the sanitary room, theengine room and the bulk tank. Table11 gives some prices for recently builtFrench milking parlors (translatingFrench prices into U.S. currency). Sinceno milking parlors (or systems) arebuilt in the U.S. or Canada, most of theequipment needs to be imported,increasing the cost by 50%.

The cost of the 2 x 12, 6-milking unit(classic Casse System) parlor at theSpooner Research Station (Universityof Wisconsin–Madison) built in 1996,consisted of the following:

■ Building 36’ x 25’(without labor) $14,500

■ Stanchion andfeed hopper $ 4,800*

■ Milking equipment $10,200

■ Feed delivery system to parlor $5,000

■ Cleaning in place system $1,500

■ 4HP air compressor $329

■ 210 gallon bulk tank (used) and hook up $1,000

■ 80 gallon water heater $480

■ Miscellaneous items $500

TOTAL $ 38,309

* The real cost of stanchion and feed hopperis closer to $ 15,000 when purchased directlyfrom an equipment dealer.

106


Figure 9. Maintenance schedule of the milking system

Milking machine Milk tank

Before each Sanitizing with a Sanitizing with a milking chlorine solution chlorine solution

After each Thorough cleaning • Pre-rinsing with cold watermilking or according to dealer • Cleaning with hot water and emptying specification detergent with an adequate brush

• Rinsing with acid

Every day Checking admittance of air in the clusters

Every week Cleaning of the outside Control of temperatureof the clusters

Every two weeks Dismantling and cleaning Cleaning of air linesby hand of all elements

Every 6 months Cleaning of the clean-in-place system and control of the integrity of milking liners.

Once a year Conrol of the integrity Control by a technicianof the whole system by a technicianAnalysis of water

Every two years Replacement of milking liners and all milk lines

Table 11. Cost of milking parlors (US $)

Equipment Casse 24p 12u New Casse 48p 24u Rotary 36u

Building 21700 28300 31700

Milking system 23300 33300 65000 (34950 imported) (49950 imported) (97500 imported)

Automatic cleaning 1700 3300 4200

Self cleaning bulk tank 4700 7700 7700

Total $51,400 $72,600 $76,900($63,050 imported) ($89,250 imported) ($141,100 imported)

ConclusionThis chapter describes in detail severalmilking systems and discusses variousmilking rates. This is significantbecause a producer chooses a systembased on the number of ewes to bemilked and the number of milkers.

The greater the number of ewes themore efficient the system has to be sothat the producer does not spend allday in the milking parlor. Of course,the overall cost of the system is adetermining factor; often a producermust compromise between efficiencyand cost. However, by decreasing theefficiency of the operation theproducer might never achieve theproduction and profit goals setbeforehand.

The choice of the milking system hasto be a realistic compromise betweenefficiency and cost. No compromisecan be made on the quality of themilking equipment (pulsator, regula-tor, clusters, liners). Only equipmentspecifically designed to milk sheepshould be used.

ReferencesBillon P. 1998. Milking parlors and

milking machines for dairy ewes.In: Proceedings of the 4th GreatLakes Dairy Sheep Symposium,Spooner, Wisconsin, June 26-27,1998

Bosc J. 1974. Organisation et produc-tivité du travail de la traite desbrebis laitières. Importance duchoix d’une méthode de traite. In:Proceedings of the firstSymposium of milking machinesfor small ruminants. Millau. France.231-250.

Chambre d’Agriculture. Septembre2000. Groupe des EDE Midi-Pyrénées, Languedoc-Roussillon.

Chaumont L. 1998. L’installation detraite pour brebis 27 p.

Delmas Ch. and C. Poussou. 1984. Unenouvelle installation de traite pourles petits troupeaux de brebis. In:Proceedings of the 3rd Symposiumof milking machines for smallruminants. Valladolid. Spain.326-345.

Le Du J., J.F. Combaud, P. Lambion andY. Dano. 1984. Etude de la produc-tivité en salle de traite pour brebis:incidence du trayeur, de la race etde la taille de l’installation. In:Proceedings of the 3rd Symposiumof milking machines for smallruminants. Valladolid. Spain.303-325.

Roques J.L. 1997. Traite des brebis: deséquipements de plus en plus per-formants. Pâtre numéro spécialproduction ovine, Octobre 1997,29-32.

Sharav E. 1974. Comparison of varioussheep milking systems. In:Proceedings of the firstSymposium of milking machinesfor small ruminants. Millau. France.259-264.

Vallerand F. 1984. Les problèmes demécanisation de la traite dans lessystèmes laitiers extensifs. In:Proceedings of the 3rd Symposiumof milking machines for smallruminants. Valladolid. Spain.216-227.

107


108


Managing dairy ewes is verysimilar to managing most otherewes, but milking adds tasks

not involved in a meat and wool oper-ation. The number of ewes milkedtwice a day for 5–7 months will have atremendous impact on the producer’sworkload. Moreover, with respect tomilking, the producer is forced torethink certain aspects of manage-ment, and to reorganize and repriori-tize the work routine. Some aspects ofanimal health need to be reviewedsince many treatments with availablemedications cannot be performedconcurrently with milking. Therefore,important decisions made before,during and after milking will affect thewhole operation.

Choosing themilking seasonIndividual farms must determine thebest milking season according to feedresources, the availability of labor(family or hired help), the sale of milkor cheese, the goals of the producer(total or supplemental income) andthe breed’s sexual season.

Milking throughout the year Producers who process their milk intocheese or yogurt and sell to marketssuch as distributors, restaurants orupscale retail stores should considermilking throughout the year. Thesemarkets generally require a steadysupply of products. Given the limita-tions of dairy ewes in terms of lacta-tion length and the rapid decline ofmilk production throughout lactation,meeting the same level of production(quantity and quality) day after dayposes constraints that are not easilyresolved without using frozen milk.

Managing a year-round milking opera-tion works by splitting the flock intosix different groups that come intoproduction in two-month intervals. Onany given day, the producer hasroughly the same number of ewes atmilking with an equal number in early,mid- and late lactation. The fat andprotein content of the milk in the bulktank should stay fairly constant,allowing for the manufacture of a con-sistent product.

109

General managementof dairy ewes

Chapter 10

Important

decisions made

before, during and

after milking will

affect the whole

operation.

There may be seasonal variation in milkproduction and content due to differ-ences in dry matter intake as a result ofthe way ewes are fed (for example,pasture vs. complete confinement). Anexample of possible managementsystems is given in table 1.

Each of the six units has to be keptseparated and treated as a differentflock. Since each unit will always lambat the same time of year, replacementewe lambs should be kept from eachunit or from the unit immediately pre-ceding. Units that will be bred in thespring (April or June) should havemore ewes to compensate for lowerfertility at this time of the year.

A management system with only twounits, one lambing in February andthe other in September, would bemuch simpler in terms of productionbut the change that takes place inmilk composition affects the cheese-making process.

Seasonal milkingSeasonal milking is the most popularsystem with lambing concentratedover a few weeks and the majority ofewes milking at more or less the sametime. In this type of system ewe lambsgenerally lamb one or two monthslater than older ewes. Lambing andmilking can occur in winter or in springaccording to the objectives of theproducer: maximum milk production athigher cost, or lower input with maxi-mized green forage consumption.Producers should evaluate all consider-ations before making a choice.

Generally speaking, a Januarylambing, corresponding to a Februarymilking start, is more apt to sustain a5–6 month milking period than aspring lambing because most lacta-tion occurs during cooler tempera-tures. This better favors the animal’sfeed intake and comfort. However, thehighest feed demand (at the end ofgestation and early lactation) occursin winter and requires an amplesupply of expensive stored feed.

■ Winter lambing (and milking)makes more efficient use of hiredlabor while spring lambing (andmilking) is better for family labor.

■ Dry matter intake greatly influ-ences milk yield. High-producingewes arriving at peak yield whileon pasture need to have their dietssupplemented by either concen-trate or high quality dry forage.

■ High summer temperaturesdecrease ewes’ appetites, reducingthe feed intake and therefore themilk yield.

■ Milk produced in summer haspoorer cheese making perform-ance. It seems that warm tempera-tures do not affect the composi-tion of milk as much as the lengthof days.

Breeding of ewesManipulation of the ewe’s reproduc-tive cycle might be of interest if thesheep dairy producer plans to:

1. Synchronize estrus of ewes so thata sufficient number can be put atmilking on the same day. Iflambing is spread out over time,milking of the first ewes to lambmight be delayed, creating a signif-icant loss in milk income.

2. Breed out-of-season for milking allyear or just in the fall.

110


Do not plan for ewes to reach

their peak of production

during the hot summer

months.

Table 1. Possible management system for year-round milking

Breeding Lambing Milking Dry off Breeding

Unit 1 August January February June–July August

Unit 2 October March April Aug–Sept October

Unit 3 December May June Oct–Nov December

Unit 4 February July August Dec–Jan February

Unit 5 April* September October Feb–March April*

Unit 6 June* November December April–May June*

* Difficult months for breeding

SynchronizationSynchronization of estrus can befollowed by natural mating (one ramfor ten ewes) or by artificial insemina-tion with fresh or frozen semen.

The technique of synchronizationextends the luteal phase (with thehelp of progestagen) until all corporalutea (progesterone-secreting cells)have regressed and disappeared fromthe surface of the ovaries. A newestrus cycle with ovulation occurs atthe end of the treatment.

The progestagen is slowly administeredto the ewes over a period of 12–14days via an implant, a vaginal pessaryor a vaginal CIDR. At the removal ofthe device an injection of PMSG(pregnant mare serumgonadotropine) is injected to theewes to regulate ovulation. Estrusappears 48 hours later.

Out of season breedingThe ram effectWhen ewes are in estrus and isolatedfrom the rams for at least 30 days, theyovulate when the rams are reintro-duced. Isolation from sight and smellof the rams is recommended. Asuccess rate of 60–65% is reported inthe literature (Pearce and Oldham,1984). Success is improved when theram effect is accompanied by synchro-nization of estrus. However, the rameffect used for out-of-season breedingpermits the induction of only oneperiod of ovulation; that is, the sexualcycle is not maintained. The ram effectseems to have little or no effect onewe lambs.

Treatment with melatonin and manipulation of lightSheep from mild climates presentseasonal variations of breedingactivity that can be controlled byannual photoperiodic changes.Melatonin is a substance secreted bythe pineal gland during the short-length days that stimulates sexualactivity. In practical applications, mela-tonin can be delivered either in feedor via an implant. Treatment withmelatonin alone advances the sexualseason by 11⁄2 months. A melatonintreatment before the summer solsticedoes not appear to be efficient. Forgood control of out-of-seasonbreeding, melatonin treatment shouldfollow a light treatment.

By artificially manipulating the lengthof days (by use of artificial lighting inthe barn) it is possible to create shortdays during the natural long days,thus favoring the synthesis of mela-tonin. Chemineau et al. (1993) showedthat a “long day” light treatment isnecessary before a melatonin treat-ment to establish the ovulatory cycleand maximum expression of estrusbehavior. Light treatments are difficultand expensive because they requirecompletely enclosed barns. However,Chemineau et al. (1993) defined proto-cols that can be used either in closedbarns or open sheds (figure 1).

Habituating newewes to the parlorTraining new ewes or ewe lambs tocome willingly to the parlor can bestressful for the animals and is bestdone before the ewes lactate. Iftraining of the ewes is conductedsimultaneously with the weaning oflambs, expect a drop of 30–40% inmilk production (Labussière, 1988).

111

G E N E R A L M A N A G E M E N T O F D A I R Y E W E S C H A P T E R 1 0

24

20

16

12

8

4

0

Supplement light

Natural dusk

Natural dawn

long days

melatonin treatment

artificial dawn

introduction of rams

short days

January February March April May June

12-14 days 48-52 hours

PMSG

insertion of device

removal of device

mating or AI

Figure 1. Photoperiodic treatment in an open barn where extra light is givenas an artificial dawn and two hours of supplementary light 16 to 18 hours laterfor more than 60 days, followed by a melatonin treatment. Breeding with malesmanaged the same way is done 70 hours after the onset of the melatonintreatment (Chemineau et al., 1993).

Training of the ewes is best executed4–5 weeks after the end of breeding.At this time the resulting stress willnot affect embryonic developmentand the ewes will remember the feedin the parlor later at milking. Thetraining consists mostly of habituatingthe ewes to come willingly andwithout fear into the milking parlor,and is usually completed within aweek. At the beginning of the milkingperiod, training will consist mostly inthe milking per se without significantnegative effects on milk production.

Preparing the ewesfor lactation

Rearing of ewe lambsIn chapter 6, it was explained that theudder’s future milk capacity can beimpaired by excessive growth of thestroma (mainly adipose and connec-tive tissues) in comparison to theparenchyma (tubulo-alveolar epithe-lium). This critical development occursin sheep before puberty between twoand four months of age.

An excessive growth rate at thisperiod favors the development ofstroma over parenchyma. Therefore, arelatively low growth rate (50% ofhigh growth rate) from weaning at4–20 weeks of age will increase theparenchyma growth and the milk pro-duction in the first lactation. However,producers have to realize that suffi-cient growth of ewe lambs has to beattained for successful breeding at7–8 months. Milk and lamb produc-tion of ewe lambs at one year is a veryimportant economic component ofthe operation. A compromise wouldhave to be reached betweenmaximum milk production of ewelambs and age at first milking.

FeedingAs with non-dairy ewes, the monthpreceding lambing is critical in prepar-ing for milk production. Nutrientsgiven to ewes must support not onlythe rapidly developing feti but alsomammogenesis (development ofsecretory tissues in the mammarygland). Ninety-five percent of thedevelopment of the mammary glandtakes place during the last third ofgestation. Significant undernutritionduring this period can greatly reducemammogenesis (Treacher, 1970).

Bizelis et al. (2000a, 2000b) found thatdairy ewes receiving only 90% of theirmaintenance requirements during latepregnancy had much smaller uddersand significantly lower milk produc-tion, even though they were put on anadequate free choice diet in early lac-tation. Dairy ewes in late pregnancycan be fed the same as any other ewesat the same stage of production.Producers should refer to chapter 4 ongeneral nutrition of sheep.

ShearingShearing ewes before lambing is acommon practice in any sheep opera-tion. It provides a cleaner environmentat lambing time, gives the animalsmore room and makes lambing easierto supervise. In a dairy operation,shearing before milking is a mustbecause milking must be performedin a clean environment. Sanitary col-lection of milk is nearly impossiblewhen ewes are in full fleece. It isimperative that ewes be shornbefore milking starts. A secondshearing might be necessary in themiddle of lactation.

Weaning of lambs In dairy ewes, 25% of the total milkyield for the entire lactation isproduced during the first month(Folman et al., 1966; Ricordeau andDenamur, 1962). This is mainly due tothe fact that milk production increasesfrom parturition to about 24 days inlactation when peak milk productionis reached.

To complicate matters, ruminants havethe highest probability of mastitisduring the first 45 days post-partum(Hamman, 2000). Therefore, early lacta-tion management is critical to udderhealth and profit margins.

A wide variety of weaning systemsexists for dairy ewes that allow eitheroptimum lamb growth, commercialmilk production, or a combination ofthe two (Folman et al., 1966; Gargouriet al., 1993; Papachristoforou, 1990).

The weaning system that favors lambgrowth is the 30-day exclusivesuckling system (DY30) where ewesare not machine-milked during thefirst month of lactation and the lambsare weaned at about one month ofage. This is the most common systemused throughout the world and inNorth America.

112


The system that favors maximumcommercial yield is the day onesystem (DY1) where lambs areremoved from their dams within 24hours after birth and raised on artifi-cial milk replacers or with part of themilk collected from the ewes; theewes are machine-milked twice dailyfor the entire lactation. This system isparticularly common in NorthernEurope with the East Friesian breed(Flamant and Ricordeau, 1969).

Finally, a weaning system thatattempts to find a compromisebetween acceptable lamb growth andcommercial milk production is themixed system (MIX). The MIX systemallows for lambs to suckle their damsfor 8–12 hours per day, after whichthey are separated for the night, andthe ewes are machine milked the fol-lowing morning (McKusick et al.,2001a). Lambs are weaned at 28–30days and the ewes are exclusivelymachine-milked twice a day. The threesystems have been evaluated at theSpooner Agricultural Research Station(University of Wisconsin-Madison).

The weaning system most appropriatein terms of maintaining maximummilk production potential during thefirst 30 days of lactation is the MIXsystem. This allows for frequent udderevacuation during the day (lambssuckling) and one large evacuationevery morning (machine milking).

Because the MIX system permits thesale of at least some commercial milkduring the first 30 days of lactationand requires no artificial rearing of thelambs, it is more advantageouscompared to both traditional 30-dayweaning and the DY1 system(McKusick et al., 2001a). The DY1system is probably the least efficientin maintaining maximum milk produc-tion potential due to the fact that theudder is being emptied only twice perday. The traditional DY30 system reliesuniquely on the lamb to maintain milkproduction during early lactation. It isonly when milk requirements of thelambs are greatest that maximum milkproduction is met in DY30 ewes.

113


The DY30 system is not appro-

priate for ewes that rear only

one lamb.

Lambs separated from their dams for the once a day milking (MIX).

Although we see marked differencesbetween the three weaning systemsduring the first 30 days of lactation interms of milk production potential,they disappear after about 45 days inlactation (figure 2). This is to say thatDY1, DY30 and MIX ewes have similarmilk yield and lactation lengths fromseven weeks onward.

Milk composition and qualityIn addition to the significant differ-ences in milk production observed forthe three weaning systems, there arealso marked differences in milk fatcontent (figure 3) and somatic cellcount (figure 4) during the first 30days of lactation. Compared to DY1ewes, the commercial milk (totalmilk extracted with the machine) ofMIX ewes has significantly less milkfat content for as long as the ewesremain in contact with their lambs.This lower fat content can have aserious negative effect on the milk’svalue.

The reason for lower fat content isprobably due to failure of the milkejection reflex during machine milkingof MIX ewes, but could also result fromproblems associated with stress, milkfat storage and/or fat synthesis in theudder when the ewes are separatedfrom their lambs in the evening.

Oxytocin is a hormone released fromthe brain of mammals as a result of teatand udder stimulation, usually at thetime of suckling (Ely and Peterson,1941). Oxytocin is an integral part ofmilk ejection (see chapter 6). Duringmachine milking, if there is no releaseof oxytocin combined with otherfactors, milk remains in the alveolialong with large quantities of fat.Thisresults in incomplete udder evacuationas well as a less rich commercial milk.

114


3.5

3.0

2.5

2.0

1.5

1.0

.5

.0

milk

yie

ld, k

g/d

DY1MIXDY30weaning

weeks in lactation0 8 12 16 20 24 284

7.0

6.0

5.0

4.0

3.0

2.0

milk

fat,

%


DY1MIXDY30weaning

Figure 2. Average daily commercial milk yield per ewe for 3 weaning systems.

Figure 3. Percentage of milk fat for 3 weaning systems.

Somatic cell count (SCC) is often usedin monitoring udder health in dairyanimals. Although the probability ofinfection is higher as the number ofSCC increases (Billon and Decremoux,1998), it should be noted that SCC isnot a direct indicator of infection, butrather of inflammation.

Weaning systems can have markedeffects on SCC in dairy ewes.Observations at the SpoonerAgricultural Research Station indicatethat MIX ewes maintain significantlylower SCC during the first 30 days oflactation than DY1 ewes. This seems tobe related to more frequent udderevacuation when milk production isthe highest in lactation. When theudder is heavily distended and underhigh intramammary pressures, the

small junctions between cells in themammary gland begin to open. Thispermits an influx of SCC (white bloodcells and other cell types) into themammary gland (Stelwagen et al.,1997). Furthermore, if the mammarygland does get infected (via entry ofbacteria through the teat canal), morefrequent evacuation of the udderdecreases the chance of thosebacteria from colonizing the udderand establishing infection. DY30 ewestend to have significantly higher SCCcompared to both MIX and DY1 ewesaround the time of weaning andduring mid-lactation. It is, however, dif-ficult to say whether or not any of theweaning systems are beneficial inreducing the mastitis incidence indairy ewes.

Milk protein percentage (figure 5) wassimilar over the whole lactationbetween MIX ewes and DY1 ewes. Itwas highest during early lactation,decreased through mid-lactation, andthen increased for the remainder ofthe lactation.

115


7.0

5.5

5.0

4.5

4.0

SCC

, lo

g 1

0 u

nit

s


DY1MIXDY30weaning

Figure 4. Log-transformed Somatic Cell Count (SCC) for 3 weaning systems.

7.0

6.0

5.0

4.0

Milk

pro

tein

, %


DY1MIXDY30weaning

Figure 5. Percentage of milk protein for 3 weaning systems.

Choosing a weaning system is

not easy when all factors are

adequately considered. The

maximum amount of milk is

obtained with the DY1 system

because about 25–30% of all

milk is produced during the

first 30 days, but this system

calls for the artificial rearing of

lambs (see chapter 11). The MIX

system is the most economical,

but the percentage of fat in the

milk is reduced while lambs are

suckling, lowering the quality

and value of the milk. The DY30

system is economical but does

not favor maximum milk pro-

duction and should not be used

with ewes suckling only one

lamb.

Milking frequency In looking at different weaningsystems, it has been shown that for adairy ewe to maintain her maximummilk production potential, the uddermust be frequently and completelyevacuated, especially in early lactationwhen milk production is highest.

The effect of milking frequency ondaily milk production, fat percentageand protein percentage has beenstudied at the Spooner AgriculturalResearch Station (de Bie et al., 2000).Two groups of DY1 ewes (ewes exclu-sively milked beginning 24 hours afterlambing) with equal initial productionwere milked either two or three timesa day. Overall milk production (table 2)was increased by 15% during the first30 days of lactation but the responseto an increase in milking frequencyvaried according to the percentage ofEast Friesian breeding. Ewes withbetween 25–50% East Friesianbreeding produced 30% more milkwhen milked three times a day whileewes of 25% East Friesian breedingdid not respond at all.

The percentage of milk fat was signifi-cantly reduced in ewes milked threetimes a day, while protein percentagewas less sensitive to changes in thefrequency of milking.

In conclusion, more milk was obtainedby milking a third time, but theresponse varied greatly according tothe ewes’ genotype. The variableresponse was probably due to theewes’ genetic potential to producemilk and to the storage capacity oftheir udders. For a given level of milkproduction, the more milk the ewestores in the cistern, the less often shehas to be milked. The economicinterest of involving more labor andcost in a third milking, even for a shortperiod of time, cannot be ascertainedat this point. More research is neededto determine the exact reason whysome ewes respond better thanothers to higher milking frequencies.

Interval betweenmilkingsOne can think about the inside of theudder as having two compartments.One compartment is responsible forproducing and secreting milk (thealveoli), and the other is responsiblefor storing milk (the cistern).

In ewes, a dynamic relationship existsbetween these two compartmentsthat greatly affects milk yield.Immediately following evacuation ofthe udder, either by suckling or themachine, the pressure within theudder (intramammary pressure)decreases significantly (Labussière,1993). The removal of milk combinedwith the drop in pressure allows newlysecreted milk to accumulate naturallyin the udder. The alveoli begin tostretch as they accumulate newlysecreted milk, and eventually theyspontaneously contract in response tothe tension within the alveoli. Milkthen flows into a long system of smallducts, eventually traveling through asystem of larger ducts, to finally arrivein the cistern. The whole process isrepeated.

Eventually, because of the largevolume of milk that accumulates inthe cistern, the intramammarypressure of the cistern becomes greatenough to slow down the flow of milkfrom the small and large ducts. Milkbegins to distend the alveoli becauseit can no longer be expelled into thesmall ducts. In response to theincreased pressure within the alveoli,the neighboring secretory cells beginto shut down milk production.

116


Table 2. Daily milk production (Liters/ewe) during the first 30 days of lactationof ewes milked twice (2TM) or three times a day (3TM).

Group %EF N Total 30day Week 7

2TM All 72 82.6 ±2.8 a 2.1 ±.09 a

3TM All 53 95.2 ±2.3 b 2.1 ±.09 a

2TM 25% 20 88.3 ±5.0 a 2.3 ±.15 a

3TM 25% 16 89.1 ±5.1a 2.0 ±.15 a

2TM 37.5% 12 70.4 ±4.9 a 1.7 ±.16 a

3TM 37.5% 9 95.8 ±6.6 b 2.0 ±.16 a

2TM 50% 40 89.3 ±4.0 a 2.3 ±.12 a

3TM 50% 28 100.7 ±5.1 b 2.3 ±.12 ba,b For each group, means with a different letter differ significantly (P< .05)de Bie et al., 2000

For a dairy ewe to maintain her

maximum milk production

potential, frequent and

complete evacuation of the

udder is essential, especially in

early lactation when milk pro-

duction is highest.

Additionally, the feedback inhibitorlactation hormone (FIL) concentrationincreases when large volumes of milkremain in the alveoli. This hormoneessentially tells the secretory cells thatthere is too much milk beingproduced and that milk synthesisshould be slowed down. Thus, whenthe interval between milkings (orsucklings) surpasses around 16 hours,and cisternal milk storage capacity hasbeen reached, milk secretion may behampered (Davis et al., 1998).Prolonged periods of milk stasis in theudder, particularly at dry-off, are someof the factors that initiate apoptosis, or“programmed cell death.”

Oxytocin, a hormone produced by thebrain, influences the release of milkfrom the alveoli to the cistern duringsuckling or milking. Without oxytocin,milking is incomplete and the milk hasa lower fat content. In nature, sucklingis the normal stimulus for the releaseof oxytocin. During milking the stimu-lation is achieved just prior to milking(noise in milk room, the start of thevacuum pump) and at milking withthe attachment of the teat cups to theteats.

Another important factor condition-ing ewes to the machine milking is thetime of day. Ewes are creatures ofhabit and are time sensitive.

Since it appears that a milking intervalof 16 hours could be appropriate fordairy ewes because of their increasedcisternal storage capacity. McKusick etal. (2001)b looked at the possibility ofmilking only 3 times in 48 hours (6 am,10 pm, 2 pm) instead of 4 timesstarting in mid-lactation (90 days) tothe end. The authors concluded thatmilking every 16 hours apears to be areasonable compromise to normaltwice-daily milking routines for dairyewes, and does not result in any dele-terious effects on milk yield, milk com-position, somatic cell count or lacta-tion length. A longer milking intervalresults in a significant reduction inlabor and time spent in the milkingparlor.

Milking proceduresThe milking procedure should allowfor recovery of the maximum amountof milk possible, in the shortestamount of time, with the least amountof human intervention and withoutcausing harm to the ewe or udder.Time is essential because sheepdairying generally involves manyanimals. With an efficient milkingsystem, high throughput of animalsthrough the milking parlor can beachieved (300–350 per hour) as longas milking procedures stay simple andewes have uniform milk flow andudder conformation.

Washing the udderIt is a normal practice to start milkingewes without washing the udder. Thetime taken to properly wash and drythe udder increases tremendously thetime of milking. Moreover, if washingis carried out as a group (generally agroup of 12 ewes at a time), stimula-tion of the ewes can result in prema-ture release of oxytocin.

It has been shown that contaminationof the milk by external bacteria isgreater with improper washing thanwith no washing at all. The generalrecommendation is not to wash theudder but to make all possible effortsto keep the ewes clean. Fresh,abundant bedding should beprovided daily to ewes in confine-ment. Muddy pastures or muddyroads should be avoided as much aspossible. If management practices donot allow for ewes to be maintained ina sufficiently clean environment,udder washing may be necessary.

California Mastitis TestThe California Mastitis Test (CMT)gives an early indication of inflamma-tion or poor udder health, and can beused reliably in dairy ewes. The test iseasy to perform, but can significantlyincrease milking time.

Before putting on the teat cups, eachudder half is sampled and evaluatedseparately by squirting milk into ashallow paddle device that contains aspecial reactive agent. The milk isswirled and the resulting clot forma-tion is subjectively graded to deter-mine the relative amount of inflam-mation (presumed infection) in theudder half. CMT score actually has agood correlation with somatic cellcount; it is important because itprovides an early indication of mastitisthat can be confirmed with furtherveterinary diagnostics.

117


The interval between milkings

should not be more than 16

hours. Producers milking only

once a day, especially in early

lactation, should expect a

reduction in daily milk yield

and lactation length.

It is important that once a

milking routine has been

established, it should be

respected as much as possible

from one day to the next.

The California Mastitis Test should bedone when:

■ the somatic cell count in the milkof the bulk tank is abnormallyhigh. Problem ewes must then bedetected and removed from theflock.

■ the quality of some ewes’ milkappears doubtful. Just a few eweswith high somatic cell count areenough to rapidly elevate thecount in the bulk tank.

StrippingThe udder’s size, shape and form aredetermined genetically (Fernández etal., 1997), and play an important rolein storing and recovering milk(Labussière, 1988). In general, eweswith large udders produce more milkthan ewes with small udders. Eweswith taller cisterns (that is, more uddervolume below the teat canal exit) takesignificantly longer to milk than eweswith shorter cisterns (McKusick et al.,1999).

During milking the udders of eweswith tall cisterns often have to belifted and massaged during milking toallow for all the milk to drain from theudder. This is called stripping. Thisprocess takes time and thereforereduces the milking machine’s effi-ciency. Because the volume of milkgained by stripping is lower than thevolume of milk recovered by themachine (machine milk yield), it hasbeen proposed that machine strip-ping be eliminated, especially in eweswith low stripping percentages, togain in overall parlor throughput time.

At the Spooner Agricultural ResearchStation an experiment was conductedto see if stripping could be eliminated(McKusick, 2003). Commercial milkyield for non-stripped ewes was only15% less than for stripped ewes(average stripping percentage is 22%).Furthermore, machine milk yield for

non-stripped ewes increased and wassignificantly higher than strippedewes during the experiment. Theseresults imply that ewes become habit-uated to stripping, and that elimina-tion of the process could result in rela-tively more milk recovered by themachine. Therefore, stripped yieldmight be overestimated and could becloser to what has been found inFrance in ewes with excellent udderconformation (11 to 15%).

Speed of milkingAnother experiment at the SpoonerStation looked at machine milkingefficiency (volume of machine milkrecuperated / the total time requiredto milk the ewe). The experiment per-mitted the identification of five typesof ewes (McKusick, 2000).

1. The ”fast milker”— a ewe thatattains milk flow almost immedi-ately after the teat cups have beenplaced on the udder. She achievesone or two very high peak milkflow rates (1.5 to 2 liters/minute)and has usually emptied her udderin 60 seconds. If the ewe hascorrect teat placement, she typi-cally has only 5–10% strippingvolume. This is the type of ewe thatshould be kept and replacementschosen from.

2. The “average milker”—a ewe that,in one milking, during mid- to lategestation gives approximately 1liter of milk in 2 minutes. She gives80% of her milk in the first 1.5minutes and then requires about30 seconds for stripping to removethe rest of the milk (20% of thetotal yield). If the stripping yield isdue to habituation to massage, themachine yield will increase andstripping can be omitted. This typeof ewe is adequate for milking andwill not significantly decrease thespeed of milking.

3. The “slow milker”—a ewe that gen-erally requires a great deal ofmanual intervention in the parlor.Milk begins to flow 20–30 secondsafter the teat cups have beenplaced on the udder. Peak milkflow rates rarely exceed 0.5–0.6liter/minute. Milking proceduretimes often exceed 4 minutes, andthe milker has a tendency tospend significant amounts of timein udder massage and stripping.One slow milker in each batch ofewes is enough to greatly slowparlor throughput time. Moreover,because of the time spent on aparticular ewe, overmilking canoccur in other ewes, increasing thechance of mammary infection. Theslow milker should be removedfrom the flock.

4. The “poor udder conformation/teatplacement ewe”—a ewe with alarge amount of udder volumelocated beneath the teat canalexit. This is often seen in olderewes with relaxed medial suspen-sory ligaments. Milk flow rates andmilk yields might be acceptable;however, stripping percentages arehigh (30 to 40%) which signifi-cantly increase milking proceduretime. The age at which ewes’udders attain significant loss ofconformation due to stretching of

118


In some cases, stripping could

be eliminated because the

small amount of milk gained

during stripping is not worth

the labor input to extract it.

the middle suspensory ligamenthas to be determined. Select eweswith profound intramammarygrooves since deeper groovesimply stronger ligaments.

5. The “no milk-ejection ewe”—a ewethat does not release her alveolarmilk fraction during milking. This isprincipally due to a lack ofoxytocin release, implying that theewe does not possess any dairycharacteristics. Cud-chewingduring milking has been correlatedto oxytocin release which usuallyoccurs 30–45 seconds after theteat cups have been placed on theewe. Stripping these ewes some-times results in enough stimula-tion for milk ejection; however, thisis a very inefficient way to extractmilk. A ewe that never chews hercud in the parlor should beremoved.

Post dippingThe teat sphincter, a ring-like structureat the end of the teat composed ofsmooth muscle fibers, allows milk toflow from the teat canal duringsuckling or milking by relaxing andopening under the influence ofsuckling or milking stimuli. Soon aftermilking or suckling the sphinctercloses and keeps the milk fromflowing. The closing of the sphincter,however, is not immediate and there isa chance that environmental (mainlyCoagulase-Negative Staphylococci)bacteria can find their way throughthe teat canal and infect the udder.

When lambs suckle, the frequentemptying of the udder limits invasionby bacteria. With exclusive machinemilking, however, the long periodbetween two milkings favors the mul-tiplication of bacteria and increasesthe chance of udder inflammation.Dipping the end of the teat with anantibacterial agent immediately aftermilking reduces the risk of contamina-tion before the teat sphincter is com-pletely closed.

119


Speed of milking or number of

ewes milked per hour can be

increased significantly by

removing “problem” ewes.

These ewes can lead to over-

milking others, which may lead

to an increase of udder inflam-

mation.

Post dipping increases milking

time but appears to be a neces-

sary precaution.

In conclusion, the amount of timespent by the producer in the milkingparlor is directly linked to milking procedures.

■ Washing the udder is not neces-sary if ewes are kept in a cleanenvironment.

■ Detection of high somatic cellcount ewes should be done peri-odically but not systematically.

■ Stripping of some ewes should beomitted so as not to habituateewes to massages.

■ Post-dipping is essential to reduceudder infection.

■ Problem ewes (slow milker, poorudder conformation, high SCC)should be removed from themilking herd.

Dry offDry off is a natural process thatactually begins just after peak lacta-tion (three or four weeks afterlambing) and gradually continues foras long as the ewe remains in milk. Thecells in the udder that secrete milkundergo a process called “pro-grammed cell death” and are not“renewed” until the following lactation.

Dairy ewes lactate between 4–7months. The decline of milk produc-tion is estimated to be 15–20% permonth after reaching peak produc-tion. A small population of eweswithin the flock will continue to con-sistently produce a little bit of milklong after the rest of the flock hasbeen dried off. It will be up to theproducer to decide whether or not itis worth his or her time to keepmilking these ewes. According to datafrom the Spooner Station it appearsthat when milk production falls below0.5 liter per day, it is no longer eco-nomical to continue milking. However,other economic reasons such as ahigher milk price due to a higher fatcontent might push a producer to

keep milking.

When a producer decides to stopmilking because of low production, itis best to switch to once-a-day milkingfor 8–10 days, followed by one milkingevery other day for another 8–10 days.

Intramammary administration of

antibiotics at drying off is effective incows (Dossing, 1994) and in goats(Mecier et al, 1998). Most studiesfound that dry period antibiotictherapy is an efficient method tocontrol sub-clinical infections and todecrease milk somatic cell count insubsequent lactation, especially onanimals with high milk yield or highsomatic cell count at the end of lacta-tion. In dairy ewes Longo et al. (1994)found a cure rate of close to 95% inewes treated with spiramycin andneomycin after dry-off with only a3.1% infection rate during the dryperiod. Intramammary treatmentsafter dry-off must be performed withextreme care and scrupulous hygiene.

Treatment whilemilkingMilk intended for human consump-tion (fluid or processed) should befree of any drug. The Food and DrugAdministration (FDA) is responsible forapproving and enforcing the use ofdrugs in animals to ensure that foodobtained from treated animals doesnot contain illegal drug residues.

Such residues are detected duringregular tests conducted by state orfederal public agencies at milk andcheese manufacturing plants. If illegaldrug residues are found, the milkcannot be processed (reducing finan-

cial returns to the producer) and morefrequent testing of the producer’s milkwill follow. Producers caught repeat-edly with illegal drug residues in theirmarketed milk will have their milkproducer’s license revoked. Therefore,dairy sheep producers need to beaware of treatments that can belegally performed to avoid leavingillegal drug residues in milk.

AntibioticsAntibiotics are the most common illegaldrug residues found in milk. Manyantibiotics available in the United Statesare not approved for use in sheep.Theuse of antibiotics in farm animalsrequires a veterinarian-client-patientrelationship, and in the case of dairyewes, occasionally permits the use ofdrugs labeled for other species to beused under veterinary supervision.

120


Drying off should always be

accompanied by a drastic

reduction in the quantity and

quality of feed.

The use of antibiotics is

becoming increasingly contro-

versial and prophylactic

methods are preferred. Mastitis

is reduced by a properly func-

tioning milk machine (check

for inappropriate or unstable

vacuum levels, faults in milking

machine pulsation, teat cup

slippage), good milking man-

agement (avoiding over- and

undermilking), correct udder

hygiene (post-dipping), correct

hygienic bedding (abundant

and fresh bedding on a daily

basis), and correct flock man-

agement (culling of high SCC

ewes, which necessitates the

control of individual SCC on

regular intervals and culling of

all ewes showing hardness or

nodules in the udder after dry

off).

Each drug approved by the FDA has awithdrawal period, specific for theapproved species that must berespected before any tissue from thetreated animal can be sold for humanconsumption. Because of the lack ofapproved drugs for sheep, drugsapproved in cattle are often adminis-tered to sheep. However, the with-drawal times may not have been accu-rately established for sheep.

A good example of this situation ispresented in the study performed byRoncada et al., (2000). Following intra-mammary administration ofdicloxacillin in cows and in sheep, theauthors found that residues wereundetectable after 48 hours in cowsbut only after 72 hours in ewes of lowmilk production and 84 hours in ewesof high milk production. This presentsproblems, especially for dairy sheepproducers, and it is recommendedthat a veterinarian be consultedprior to the use of any medicationin dairy ewes. Many of the commondrugs and antibiotics used by lamband wool producers, are not allowedin dairy sheep, simply because thewithdrawal time is not appropriate forthe sale of milk (it is not uncommonfor many antibiotics to leave residuesin animal tissues for up to one month).

Antibiotics are commonly added tolivestock feed to enhance growth andprotect animals from possible infec-tion. When buying commercial feed, adairy producer should pay particularattention to the composition of thefeed.

As a general rule, antibiotics shouldnot be used in lactating dairy ewes.In case of acute mastitis, the eweshould be immediately removedfrom the milking group, treated,and not allowed back in the milkingparlor.

OxytocinAs previously noted, oxytocin is ahormone produced by the brain andreleased at milking (or suckling) as anintegral part of the milk ejectionreflex. During the transition fromsuckling to exclusive machine milking,it takes a few days for the ewe to getenough stimuli from the milkingprocess to release oxytocin. Thereforecomplete evacuation of the udder atthis time is impossible and yet neces-sary for the normal continuation oflactation.

To remedy this problem, producerscould rely on oxytocin injections forone or two milkings to ensure that theudder empties adequately. Oxytocin ,does not pose a drug residue problemat this time. Oxytocin injectionsshould be limited to only one or twomilkings because there is a risk thatthe ewe will not properly adapt tonormal machine milking due to theanimal’s becoming habituated to it.Finally, extended use of high doses ofoxytocin can actually decrease milkyield.

AnthelminticsDue mainly to intensive managementpractices, dairy ewes are extremelysusceptible to parasite infection andrequire treatment with anthelmintics(“de-wormers”). There are two timesduring the year when parasite infec-tion is important: during the spring(“spring rise”), when the combinationof increasing ambient temperatureand lush pasture growth are ideal forparasite development; and in the fall,just before the first frost when para-sites already inside the animal go intoa period of hibernation or “arresteddevelopment.”

To complicate matters, dairy sheepshould not be de-wormed during lac-tation when milk is being sold forhuman consumption. Therefore, strate-gic de-worming schedules need to beimplemented according to aproducer’s management system.Some anthelmintics can be adminis-tered during gestation (consult yourveterinarian), just prior to the firstfrost, which kills parasites before theyhave a chance to start the “arresteddevelopment” phase. This greatlyreduces the parasite load in the flockthe following spring. Depending on aproducer’s management system, de-wormers can be administered asecond time in the spring, approxi-mately a month prior to sale of milkfor human consumption (that is, amonth prior to lambing for ewes thatwill be exclusively milked in early lac-tation, or at lambing for ewes that willbe exclusively suckled in early lacta-tion). Many de-wormers used in sheephave officially been approved only foruse in other species, therefore it isalways important to consult your vet-erinarian before using anyanthelmintic or other drug in theflock.

121


As a rule, anthelmintics should

not be administered to dairy

ewes during the milking

season. Control of parasite

load should be performed

through careful rotational

grazing practices.

References Billon P. and R. Decremoux. 1998.

Mastitis of dairy ewes: etiology,detection, and control. In:Proceedings of the 4th Great LakesDairy Sheep Symp. Dept. Anim. Sci.,Univ. Wisc. Madison. 44-50.

Bizelis J.A., M.A. Charismiadou and E.Rogdakis. (2000a). Metabolicchanges during the perinatalperiod in dairy sheep in relation tolevel of nutrition and breed. I. Latepregnancy. J. Animal Physiologyand Animal Nutrition 83 :3-4 pp 61-72

Bizelis J.A., M.A. Charismiadou and E.Rogdakis. (2000b). Metabolicchanges during the perinatalperiod in dairy sheep in relation tolevel of nutrition and breed. II. Earlylactation. J. Animal Physiology andAnimal Nutrition 83 :3-4 pp 73-84

Bocquier F., M.R. Aurel, F. Barillet, M.Jacquin, G. Lagriffoul, and C. Marie.1999. Effects of partial milkingduring the suckling period on milkproduction of Lacaune dairy ewes.In: F. Barillet and N. P. Zervas (Eds.),Milking and Milk Production ofDairy Sheep and Goats. EAAP Publ.No. 95. Wageningen Pers,Wageningen, The Netherlands.257-262.

Chemineau P., X. Berthelot, B. Malpaux,Y. Guérin, D. Guillaume and J.Pelletier. 1993. La maîtrise de lareproduction par la photopériodeet la mélatonine chez les mam-mifères d’élevage. CahiersAgricultures 2:81-92

Davis S.R., V.C. Farr, P.J.A. Copeman, V.R.Carruthers, C.H. Knight, and K.Stelwagen. 1998. Partitioning ofmilk accumulation between cister-nal and alveolar compartments ofthe bovine udder: relationship toproduction loss during once dailymilking. J. Dairy Res. 65:1-8.

de Bie L., Y.M. Berger and D.L. Thomas.2000. The effect of three times aday milking at the beginning oflactation on the milk production ofEast Friesian crossbred ewes. In:Proceedings of the 6th Great LakesDairy Sheep Symposium, Guelph,Ontario, November 2, 3 and 4,2000.

Dossing F.. 1994. Clinical mastitis in thedry period. DanskVeterinaertidssrift 77 (8) 353-354,356-359.

Ely F. and W.E. Peterson. 1941. Factorsinvolved in the ejection of milk.J. Dairy Sci. 24: 211-223.

Fernández G., J.A. Baro, L.F. de laFuente and F. San Primitivo. 1997.Genetic parameters for linearudder traits of dairy ewes.J. Dairy Sci. 80:601-605.

Flamant J.C. and G. Ricordeau. 1969.Croisements entre les races ovinesPréalpes du Sud et Frisone(Ostfriesisches Milchschaf ). I. Labrebis laitiere de Frise Orientale.Elevage en race pure. Utilisation encroisements. Ann. Zootech., 1969,18 (2), 107-130.

Folman Y., R. Volcani, and E. Eyal. 1966.Mother-offspring relationships inAwassi sheep. I: The effect of differ-ent suckling regimes and time ofweaning on the lactation curveand milk yield in dairy flocks.J. Agric. Sci. (Camb.). 67: 359-368.

Gargouri A., G. Caja, X. Such, A. Ferret,R. Casals, and S. Peris. 1993.Evaluation of a mixed system ofmilking and suckling in Manchegadairy ewes. In: Proceedings of the5th Intl. Symp. on Machine Milkingof Small Ruminants. HungarianJ. Anim. Prod. (Suppl. 1): 484-499.

Hamann J. 2000. Teat tissue resistancemechanisms with special regard tomachine milking. In: ProceedingsIDF Symp. Immunol. Rum. Mamm.A. Zecconi (Ed.). Gland, Stresa, Italy.102-111.

Labussière J., J. F. Combaud, and P.Pétrequin. 1978. Influence respec-tive de la fréquence quotidiennedes évacuations mammaires et desstimulations du pis sur l’entretiende la sécrétion lactée chez la brebis.Ann. Zootech. 27(2): 127-137.

Labussière J. 1988. Review of physiolog-ical and anatomical factors influenc-ing the milking ability of ewes andthe organization of milking. Livest.Prod. Sci. 18: 253-274.

Labussière J. 1993. Physiologie de l’é-jection du lait: conséquences sur latraite. In: Biologie de la Lactation.INRA, Service de Publications,Versailles, France. 259-294.

Longo F., J.C. Béguin, G. Monsallier, P.Delas and P.J. Consalvi. 1994.Efficacy of spiramycin andneomycin combination in thecontrol of cell counts and udderpathogens in the dry ewe. In:Proceedings of InternationalSymposium on somatic cells andmilk of small ruminants. Bella, Italy,September 25-27, 1994. EAAPPublication No. 77, 1996.

McKusick B.C., Y.M. Berger and D.L.Thomas. 1999. Preliminary results:effects of udder morphology oncommercial milk production ofEast Friesian crossbred ewes. In:Proceedings of the 5th Great LakesDairy Sheep Symp. Dept. Anim. Sci.Univ. Wisc. Madison and Ctr. Sust.Ag. Univ. Vermont. 81-92.

122


McKusick B.C.. 2000. Physiologicalfactors that modify the efficiencyof machine milking in dairy ewes.In: Proceedings of the 6th GreatLakes Dairy Sheep Symposium,Guelph, Ontario, November 2, 3and 4, 2000.

McKusick B.C., P-G. Marnet, Y.M. Bergerand D.L. Thomas. 2000. Preliminaryobservations on milk yield, milkflow, and udder morphology traitsof East Friesian crossbred dairyewes. In: Proceedings of the 6thGreat Lakes Dairy SheepSymposium, Guelph, Ontario,November 2,3 and 4, 2000.

McKusick B.C., D.L. Thomas and Y.M.Berger. 2001a. Effects of weaningsystems on commercial milk pro-duction and lamb growth of EastFriesian dairy sheep. J. Dairy Sci. 84:1660-1668.

McKusick B.C., D.L. Thomas and Y.M.Berger. 2001b. Effect of reducingthe frequency of milking on milkproduction, milk composition, andlactation length in East Friesiandairy milk. Proceedings of the 7thGreat Lakes Dairy SheepSymposium. Nov. 1-3, 2001. EauClaire, Wisconsin.

McKusick B.C., D.L. Thomas and Y.M.Berger. 2003. Effect of omission ofmachine stripping on milk produc-tion and parlor throughput in EastFriesian dairy ewes. J. Dairy Science.86: 680-687

Mercier P., C. Baudry, D. Lenfant andM.P. Mallereau. 1998. Etude de l’effi-cacité d’un traitement antibiotiqueau tarissement chez la chèvre.Receuil de Médecine Vétérinaires174 (5/6) 7-14.

Papachristoforou C. 1990. The effect ofmilking methods and post-milkingsuckling on ewe milk productionand lamb growth. Ann. Zootech.39:1-8.

Pearce D.T. and C.M. Oldham. 1984. Theram effect, its mechanism andapplication to the management ofsheep. In : Reproduction of sheep.D.R Lindsay and D.T. Pearce (Edts).Cambridge University Press.

Ricordeau G. and R. Denamur. 1962.Production laitière des brebisPréalpes du Sud pendant lesphases d’allaitement, de sevrage etde traite. Ann. Zootech. 11:5-38.

Roncada P., L. Tomasi, G.L. Stracciari, L.Ermini and A. Strochia. 2000. Milkdepletion of dicloxacillin residuesin cows and in sheep followingintramammary administration.Journal of Veterinary, Pharmacologyand Therapeutics 23: 4 pp 237-241.

Stelwagen K., V.C. Farr, H.A. McFadden,C.G. Prosser, and S.R. Davis. 1997.Time course of milk accumulation-induced opening of mammarytight junctions, and blood clear-ance of milk components. Am. J.Physiol. 273: R379-R386.

Treacher T.T. 1970. Effects of nutritionin late pregnancy on subsequentmilk production in ewes. Anim.Prod. 12: 23-26.

Wilde C.J., D.T. Calvert, A. Daly and M.Peaker. 1987. The effect of goatmilk fractions on synthesis of milkconstituents by rabbit mammaryexplants and on milk yield in vivo.Biochem. J. 242:285-288.

123


124


IntroductionChapter 10 explained that themaximum milk production occurswhen ewes are milked twice a dayafter removing their lambs 24 hoursafter birth. It has also been demon-strated that high producing ewessuckling only one lamb during thefirst 30 days of lactation will have alower milk production and that25–30% of the total milk productiontakes place during the first 30 days oflactation.

But in a previous study, McKusick et al.(1999) showed that when all lambsborn (2.3 lambs per ewe) were raisedon milk replacer and all ewes were putat milking 24 hours after lambing, thetotal financial return was only 6%–7%more than the traditional system ofletting the lambs suckle their dams for30 days.

A weaning management plan combin-ing the two systems appears to be abetter solution for the time being torealize maximum milk production andthe highest profit. Therefore, in a dairysheep operation it is desirable that acertain percentage of lambs be raisedartificially either with part of the milkcollected during milking or with lambmilk replacer.

Many producers are reluctant to raiselambs artificially for several reasons:the increased workload, the numberof lambs that do poorly, a high mortal-ity rate and the expense. In manycases those reasons are valid, but theyare often linked to poor managementand organization.

The Spooner Agricultural ResearchStation (University of Wisconsin-Madison) developed a simple, low-cost lamb rearing system thatprovides ease of management,promotes good lamb growth andreduces lamb mortality to a minimum.

Materials neededLambs should feed themselves on afree-choice basis to minimize laborand maximize the amount of milkconsumed, promoting maximumgrowth. The self-feeding is done withthe use of a “lamb bar” system consist-ing of rubber nipples connected tothe source of milk by plastic tubing.The mixed milk replacer is put in con-tainers placed in a regular cooler. Thecooler’s function is to control the tem-perature of the milk. The amount ofmaterial needed to successfully raiselambs artificially depends on thenumber of lambs the producer isplanning to raise. However, the bestplan is to keep it as simple as possible.

For approximately 100 lambs, you willneed:

■ Two or three used plastic babybottles and used nipples in whichthe hole has been slightlyenlarged. It is not necessary topurchase fancy bottles and fancynipples.

■ One heat lamp.

■ Three coolers: one that holds 20liters (20 quarts), two that hold 55liters (55 quarts).

125

Artificial rearing of lambs

Chapter 11

Many producers

are reluctant to

raise lambs

artificially.

■ One cake pan small enough to fitinside the small cooler and 4–5cmdeep.

■ One plastic dishpan fitting inside abigger cooler and about 14 cmdeep

■ Two 8-liter (2-gallon) plastic pails.

■ One dozen nipples (lamb bar).Nipples with a valve system shouldnot be used because of cleaningdifficulties.

■ Clear plastic tubing to carry themilk from the source to the nipple.The tube is cut to desired length.The outside diameter of the tubeshould fit tightly inside the base ofthe nipple.

■ Three panels of 120 cm x 90 cm towhich a sheet of tin is attached.The tin is perforated with fourholes to receive the nipples.

■ Enough panels to build four pens.

■ Cleaning equipment: brush, bottle-brush, tube brushes.

■ Several deep-freeze plastic con-tainers to store frozen colostrum.

For a large number of lambs, one orseveral automatic milk mixingmachines might prove indispensable.They work well and save labor.

Products used

ColostrumSince a certain amount of maternalinstinct needs to be triggered in theewes to be milked, lambs are notremoved from their mothers until 24hours after birth, giving them suffi-cient time to consume colostrumdirectly. On their second day lambswill receive the colostrum collected atmilking.

Some diseases, such as Ovine PleuriPneumonia (OPP) also known as “HardBag,” is transmitted directly from thedams to their progeny through thecolostrum. The disease can be eradi-cated within a flock by rearing thelambs on milk replacer and notallowing them to suckle their mothers.In this event, cow colostrum can beused efficiently. Collect cow colostrumon a dairy farm ahead of lambing timeand freeze it in 16-oz. plastic contain-ers. This colostrum can be thawedwhen needed.

MilkMilk from the ewesPart of the milk collected from theewes at milking can be distributed tothe lambs raised artificially. However, itis recommended that lamb milkreplacer should be used because it isgenerally medicated, therefore pro-tecting the lambs against possibleinfection.

Lamb milk replacerUse only high quality lamb milkreplacer. Some people might achieverelative success with a few lambsusing goat or cow’s milk, but thesecannot be used on a large number oflambs. The fat content of sheep’s milkis much higher than cow’s or goat’smilk and the lactose content is lower.Modern lamb milk replacers are madeto meet the lambs’ exact require-ments. By using milk replacers accord-ing to the label, it is rare to see scoursin young lambs. Moreover, modernlamb milk replacers are easy to mixand stay in suspension for longperiods of time.

In a simple system, milk powder ismixed with water (one part drypowder to two parts water) with ahand beater. Fresh milk is madeseveral times during the day. When .Automatic milk mixing machines mixeonly a small amount of milk at a time,therefore providing fresh milk to thelambs at all time.

Starter feedA starter feed (19% CP) is provided tothe lambs at a very early age. Thestarter feed used at the SpoonerResearch Station has the followingcomposition:

■ Rolled shelled corn 47.8%

■ Rolled oats (can be 12.5%replaced by rolled corn)

■ Premix with Bovatec 16.5%

■ Soybean meal 17.2%

■ Molasses 5.0%

■ Sheep mineral .5%

■ Ammonium chloride 5%

100%

126


The milk (or colostrum)

produced during the first 3

days following parturition is

unsuitable for human con-

sumption or cheese making.

Do not make this milk avail-

able commercially.

Set upFour pens are set up: a small pen inwhich newborns are trained to thebaby bottle, a slightly bigger pen inwhich lambs are trained to the lambbar, an intermediate pen and agraduate pen (figure 1). All pens areset up in a heated area, which, in themiddle of winter stays at 2˚C to 5˚C.Pens are bedded with straw.

■ Bottle lamb pen (Area 1): After alamb is chosen to be raised onmilk replacer, it is placed in a smallpen (150 cm x 120 cm) with a heatlamp in a corner. In this pen, lambsreceive an adequate supply ofcolostrum (at least two feedings)and are trained to eat willinglyfrom the baby bottle. Lambs arefed approximately every 4–5 hours.This phase lasts generally 24 hours.The time spent with the lamb inthis pen depends on its behavior,which varies greatly between indi-viduals. In this pen, a heat lamp isprovided for the comfort of thevery young lamb.

■ First lamb bar pen (Area 2): As soonas the lamb takes the bottlegreedily, it is put into a slightlylarger pen (180 cm x 120 cm) inwhich a lamb bar has been set upusing the smaller cooler, the cakepan, two tubes and two nipples. Nomore than seven or eight lambs areput in this pen at the same time.The cake pan is placed inside thecooler resting in a block placed inthe bottom of the cooler. Thereforethe level of milk is kept highenough for lambs to receive themilk without sucking very hard.Warm milk is put in the cake pan. Inwintertime, the milk is kept warm

by placing a jug full of hot water inthe bottom of the cooler. In thispen, lambs are trained to find themilk by themselves. The first two orthree feedings, lambs are broughtto the nipple and held until themilk comes. It helps to let thelambs get slightly hungry beforethe first feeding. Some lambsunderstand the principle of the sur-rogate mother right away; othersare a little bit more reluctant, butall can feed themselves adequatelyin 48 hours. It is important toobserve the lambs from a distanceand take note of the ones eatingwell by themselves. Those lambscan be put in the intermediate pen.

■ Intermediate pen (Area 3):This pen isbigger (360 cm x 150 cm), where 15lambs can be put together. A four-nipple lamb bar is set up with abigger cooler. A plastic dishpan filledwith warm milk is placed inside thecooler that is raised on a block tokeep the level of milk high enoughfor easier sucking. At this point, nojug of hot water is placed in thecooler. It does not matter if the milkis allowed to cool.When lambs aredoing well in this pen they areadvanced to the graduate pen.

127

A R T I F I C I A L R E A R I N G O F L A M B S C H A P T E R 1 1

It is very important that the

rearing area be well-ventilated

and without any drafts.

There should be one nipple for

every five lambs in the training

and intermediate pen and one

for every 10 lambs in the

graduate pen.

feeder

cooler (cold milk in 2 pails)

small cooler (warm milk in cake pan)

cooler (warm milk in dish pan)

water heat lamp

15' 12'

4'

3'

4'

6'

Area 2

Area 4

Area 3

Area 1

5'

X

X

pail inside cooler

Cooler with 1⁄2” holes for plastic tubing.Nipples adapt to holes in tin in panel.

3’ x 5’ panel with tin insert, inwhich holes (9⁄16”) are drilledfor nipples.

3'

5'

Figure 1.

128


General set up of the lamb rearing operation at theSpooner Ag Research Station (UW-Madison).

One-day-old lambs.

Nipple attachment to panels.

Milk in a dish pan in the intermediate pen.

Older lambs in larger area.

Milk in a cake pan in the training pen.

129


Bottle feeding one-day-old lamb.Milk buckets in the graduate pen.

Lambs eating on their own.Training lamb on the lamb bar.

Feeding machine. Feeding machine.

■ Graduate pen (Area 4): this is amuch larger pen (450 cm x 360cm) where up to 30 lambs aretogether. A four-nipple lamb bar isset up. The cooler is big enough tohold two 8-liter (2-gallon) plasticpails filled with cold milk. Lambsdo very well on cold milk and itkeeps them from eating too muchat one time. In this pen, a starterfeed is put at the disposition of thelambs. At this stage, lambs are nolonger a source of intense work. Itis enough to bring them fresh milkat regular intervals and watchthem grow.

Daily routine1. In the early morning, completely

dismantle all lamb bars. Wash allelements thoroughly with deter-gent, brushes, and hot water. Thisis an essential step since cleanli-ness is the most important factorin keeping the lambs healthy.

2. Set up the lamb bars again. Mixfresh (either warm or cold, accord-ing to pen).

3. Put fresh bedding in all pens. Feedis put in pen four. Some lambsmove from pens three to four.

4. Lambs in pen one are fed milk inthe bottles. Some of the lambs willmove to pen two.

5. Lambs in pen two are trained atthe lamb bar. Some lambs aremoved to pen three.

6. New lambs are put in pen one.

7. During the day, the levels of milk inthe coolers are checked regularlyand more milk is mixed whenneeded.

WeaningLambs are weaned abruptlyanywhere from 17 to 45 days with anaverage of 28.9 days. Small lambs arealways weaned at an older age thanregular lambs. As a rule of thumb,lambs are weaned when they areclose to three times their birth weight.

The Spooner Research Station averageof 28.9 days for age-at-weaning issomewhat high (25 days would bebetter), although many lambs areweaned much earlier.

Age in days % of lambs

15 to 19 days 1.7%

20 to 24 days 20.5%

25 to 29 days 38.7%

30 to 34 days 22.4%

35 to 39 days 8.7%

40 to 50 days 7.9%

Generally, the smaller the lamb is atbirth, the older it is at weaning becauseof a slower growth rate. At weaning,lambs are removed from the nurseryarea. For the next two or three days,lambs bleat ferociously and lose someof their bloom. But soon the amount ofstarter feed consumed increases and avery decent growth rate is achieved.For the next few weeks, a high proteinration is essential.

The first lambs of the season arealways the most difficult to wean sinceno experienced lambs are around toshare their expertise. Thereafter, lambsthat are added to the already weanedgroup wean easier with less stress.

Performance oflambs raised onmilk replacerBetween 1989 and 2003, a total of2214 lambs have been raised on milkreplacer at the Spooner ResearchStation. All information pertaining tothese lambs are presented in the fol-lowing two tables. Table 1 gives theperformance of lambs according tothe year of lambing, the type of birth,and sex. Since the Spooner ResearchStation deals with many differentgenotypes of lambs, table 2 gives thesame type of data according to thebreeds of sire of the lambs.

Both tables reflect very good perform-ances of lambs raised on milk replacer.Mortality before weaning variesgreatly between breeds of lambs(from 0% to 10%) and growthbetween birth and weaning is good toexcellent although McKusick et al.(1999) reported a somewhat slowergrowth rate after weaning of EastFriesian crossbred lambs whencompared to the same type of lambsthat suckled their mother for the first30 days. Very small lambs at birth (2.0to2.5 kg.) do not do quite as well andit takes longer to wean them.However, most lambs grow well afterweaning with a gain of 300–330g/day.

130


At weaning, lambs are given a

booster of the C, D & T vaccine.

Lambs are vaccinated with

C, D & T as soon as they are

10 days old.

Economics ofraising lambs onmilk replacerThe apparent high cost of raisinglambs on milk replacer is the mainreason for the reluctance of sheepproducers to start an artificial rearingsystem. A brief analysis of theexpenses and income from the milkcollected follows.

Expenses—exampleMilk replacerFor the 2214 lambs raised on milkreplacer between 1989 and 2003, anaverage of 8.2 kg. of milk powder wasused per lamb. The average price ofmilk varies according to quality, brand,quantity purchased and retailer.Inlarge quantities the price can be aslow as US$ 1.94/kg but can also be a shigh as US$ 3/kg when purchased onebag at a time. In this study the price isUS$ 2/kg. Therefore, the cost of milkper lamb is $16.40.

LaborDuring the first three days, one lambgets approximately 10 minutes ofpersonal attention per day. Thereafter,the time spent per lamb up toweaning is very minimal, consistingmainly of cleaning equipment, makingmilk, and giving creep feed.

■ First three days 10 min./day

■ Remaining 25 days 2 min./day

■ Total 1 hr. 20 min.

x $8/hr.=$10.60

131


Table 1. Performance of lambs raised on milk replacer according to the type of birth, and sex.

Wean age ADG birth ADG wean MortalityNo. Birth wt. Wean wt. (days) weaning sale before weaning

All lambs 2214 4.6 13.0 29 286 324 3.9%

Type of birth

1 287 5.8 14.8 29 313 332 5.2%

2 932 5.0 13.9 29 306 326 4.8%

3 768 4.1 11.8 29 266 320 2.6%

4 197 3.5 10.7 30 243 329 3.0%

5 24 3.0 9.6 33 203 297 4.3%

6-7 6 2.6 9.2 36 203 104 0%

Sex F 1097 4.5 12.6 30 278 291 3.7%

Sex M 1117 4.8 13.3 29 294 341 4.1%

*partial results

Table 2. Performance of lambs raised on milk replacer according to genotype.

Wean age ADG birth ADG wean MortalityNo. Birth wt. Wean wt. (days) weaning sale before weaning

All lambs 2214 4.6 13.0 29 286 324 3.9%

Breeds of sires

Hamp-Suffolk 964 5.0 13.8 29 307 328 1.8%

Dorset 140 3.5 10.6 31 240 333 2.2%

Finnsheep 38 4.3 11.7 26 286 352 0%

Romanov 55 3.6 10.0 28 228 303 0%

Targhee 32 5.1 12.5 27 275 324 0%

Texel 150 4.3 11.6 29 265 300 2.7%

East Friesian 487 4.6 12.7 30 276 319 5.6%

Lacaune 348 4.5 13.1 31 277 332 10.4%


132

InvestmentAt total of $150 has been invested inequipment (coolers, pails, etc.). Afterten years of use, the coolers are stillsuitable for further use. The total costof investment per lamb is $0.15.

SuppliesA few supplies are needed, such as:

Nipples 10/year @ $1.10$0.10/lamb

Brushes $0.12/lamb

Detergent $0.12/lamb

Summary of costs Milk replacer $16.40

Labor $10.60

Investment $.15

Supplies $.34

Total cost $27.50

Therefore, the total cost of raisinglambs on milk replacer in 2003 frombirth to weaning is $27.50

IncomeThe income derived from the sale ofmilk collected during the first monthof lactation depends on the level ofproduction of the ewe. The expectedlevel of production can be obtainedby taking 30% of the total milk pro-duction at the previous milkingseason.

From this income, it is necessary tosubtract the cost of labor of milkingthis ewe twice a day for 30 days, thatis, $11.50 (see chapter 12). The resultsare best expressed in a table (table 3)according to the expected level ofproduction of a ewe for the first 30days and the number of lambs bornand artificially reared.

Table 3 shows that, for the best finan-cial return, a ewe having given birth to3 lambs should raise her own lambsrather than being milked twice a dayunless she is a very goodl animal ableto give 80 liters and more during thefirst month of lactation. On the otherhand an animal having given birth toonly one lamb should be milked andher lamb raised artificially even if sheis an average animal giving only 50liters of milk in the first month of lac-tation. Below an expected level of pro-duction of 50 liters it is best to havethe ewe raise her lamb(s). The breakeven point is very much dependenton the price of the milk sold and thecost of milk replacer.

ConclusionWith ewes producing more and moremilk it becomes difficult to keep thetraditional weaning system withoutsuffering a significant loss in financialreturn. The milking at 24 hours afterlambing of the best milking ewes andof the ewes producing only one lambbecomes necessary forcing producersto rear artificially a certain percentageof lambs born on the farm. Lambs canbe successfully raised artificially withgood management practices andgood working organization. The mostimportant considerations are:

■ Making sure lambs receive enoughcolostrum.

■ A high level of cleanliness.

■ A well-ventilated rearing area withno draft.

■ Providing fresh milk to the lambsat all times.

ReferencesBerger Y.M. and R.A. Schlapper. 1998.

Raising lambs on milk replacer. In:Proceedings of the 45th annualSpooner Sheep Day, University ofWisconsin-Madison, Spooner, 1998.

McKusick B.C., Y.M. Berger and D.L.Thomas. 1999. Effects on threeweaning and rearing systems oncommercial milk production andlamb growth. In: Proceedings ofthe 5th Great Lake Dairy SheepSymposium, Nov 4–6, 1999,Brattleboro, Vermont, USA.

Table 3. Potential return in US dollars during the first month of lactation according to the level of milk production and the number of lamb born and raised artificially (cost of milk powder US$2/kg,price of milk sold US$1.4/liter)

Number Expected level of production in the first 30 days (in liters)of lambs 50 60 70 80 90 100

1 +$31 +$45 +$59 +$73 +$87 +$101

2 -$3.5 +$17.5 +$31 +$45. +$59.5 +$73.5

3 -$24 -$10 -$+41 -$+18 +$32 +$46

Yves M. Berger

As in any enterprise, milking sheepand selling the milk (or processedproducts) is all about making a profit.Berger (1999) in a comparisonbetween a dairy sheep operation anda meat/wool operation has shownthat milking ewes can double aproducer’s return. Of course, the finan-cial return varies according to theproducer’s reasons for starting a dairysheep business (practical and/orphilosophical)—reasons that will influ-ence the type of operation.

No matter what the type of operation,it needs to be profitable while stillrespecting the principles of sustain-ability. There are practically as manytypes of operations as there are pro-ducers, but in North America mostsheep dairy enterprises are small-scalefamily businesses (up to 300 ewes)designed either to provide a supple-mental income or a full-time occupa-tion to at least one member of thehousehold, or to make sheep dairyingthe main source of income.

For most producers (and families), acertain way of life is the main motiva-tion for beginning a dairy sheep oper-ation. More often than not, thoseinvolved are fairly new to the sheepbusiness, able to look at a sheep oper-ation from a fresh perspective, andunderstand that there is much to gainby cultivating more products fromtheir sheep.

However, milking is not for everyone.It involves many constraints that aproducer must master before enteringthe business. In addition, plans to selland market the product and for sometype of financial analysis must be partof the overall operation.

What to considerbefore starting a sheep dairy operation

Labor resourcesSheep farming is a labor-intensiveoperation requiring a substantialnumber of hours per day. Milking caneasily double or triple the timeinvolved for an inexperienced or ill-equipped producer and the intensework can be overwhelming. Lambing,rearing lambs, weaning and milking,handling and/or processing milk oftencoincide at the beginning of theseason. Depending on the number ofanimals, the workload can be too muchfor one person to perform efficiently.Later in the season, when things calmdown, milking still must be performedeveryday, twice a day—without fail—for the next 5 to 7 months.

If milk processing is also being done,cheese (or other products) need to bemade on a regular basis. Entering thedairy business is, therefore, a long-termcommitment to be taken seriously byall members of the family, because attimes it will require that all membersbe involved in the operation.

133

Economics of raisingdairy sheep

Chapter 12

For most producers,

the appeal of the

lifestyle is the

main motivation

for beginning a

dairy sheep

operation.

A disciplined work routine andthorough understanding of themilking ewe are indispensable to getthe best production in the bestpossible conditions. Dairy ewes arecreatures of habit and changes in theirroutine create stress that affects milkproduction.

Forage resourcesIn contrast, to meat-only productionsystems where ewes’ lactation is oftencut short when their lambs areweaned, dairy sheep have a longerproduction period during which theyneed feed with high nutritional value.Feed reserves in terms of preservedforage, green forage and grain supple-ment must be calculated to cover thewhole lactation.

In many cases it is cheaper andsimpler to purchase feed than toproduce it on the farm. By not havingto produce the feed, the producer canput more effort on caring for andmilking the animals.

The season chosen for milking willaffect the overall cost of production.Winter milking favors milk productionand length of lactation but relies on agreater consumption of expensivepreserved forage. Spring milking reliesmore on cheaper green forage to thedetriment of milk yield because ofgenerally higher temperatures in Julyand August when ewes are still inmid-lactation. Deciding when to milkdepends on forage availability andcost, the producer’s ability to growhigh quality pastures, and on thedemand for milk or processedproducts.

If feed resources are based on hay,corn, soybean meal and pasture,approximate quantities of each feedare shown in table 1 according to thesystem used.

Choosing a milkingsystem The parlorThe different milking systems avail-able are described in chapter 9.

As a general rule, milking, cleaning andhandling milk should be completed inno more than two hours—whateverthe number of ewes involved. If it takeslonger, the milker, as described byOlivia Mills (1995) “gets tired, gets fedup, gets hungry, gets bad-tempered,gets problems.”Therefore, throughputof the ewes in the parlor is of theutmost importance.

The tables in chapter 9 can help pro-ducers decide on a system accordingto the number of ewes being milked.However, the number of ewes milkedper hour shown in these tables hasbeen calculated for Lacaune eweswhich give milk more rapidly than EastFriesian ewes (Bruckmaier et al., 1997).Lacaunes are also milked using asimpler milking procedure that doesnot include stripping. With the type ofewes actually milked in North America,the parlor throughput will be generallyless than reported in the literature.

The more ewes that are milked, themore sophisticated the systembecomes and consequently, the moreexpensive it becomes. Good planningcan help reduce expenses, but an initialinvestment of $170 per ewe in the U.S.is not farfetched. Installing a system for300 ewes if it will take 10 years to builda herd this size is not advisable.

134


In a larger (but still family size)

operation, hiring supplemental

labor might be necessary and

will need to be considered in

the financial plan.

A short supply of high quality

feed will have a negative effect

on the overall milk yield of the

flock and therefore on its prof-

itability.

The most important point to

consider in choosing a system

is the speed at which milking

can be performed in the best

possible conditions for both

the animals and the milker.

Table 1. Quantity of feed needed by a dairy ewe during a year.

Complete Semi- Semi-confinement confinement confinement year round winter lambing spring lambing

Alfalfa hay (kg) 730 410 350

Corn (kg) 148 120 80

Soybean meal (kg) 20 18 12

Pasture - 4 month high quality 6 month 2 months low quality high quality

The milking machineThe quality of the equipment and itssuitability for dairy sheep is also veryimportant. The incidence of mastitisand/or high somatic cell count in themilk, as well as the total milk yield, isdirectly linked to the milking machine.Sufficient vacuum reserve, constantvacuum level throughout the system,correct pulsation rate, correctly sizedteat cup liners, and ease of cleaning allensure complete evacuation of theudder, promoting udder health and along lactation.

There can be no possible compromiseon the quality of the equipment. Aproducer should not indulge inmakeshift equipment that is not espe-cially designed for milking sheep.

MarketingMarketing the milk should be a mainpriority for anyone considering sheepdairying. The “how” and “where” ofselling the products (milk, cheese,yogurt) should be settled before otherdecisions are made. The market fordomestic sheep dairy products is in itsinfancy and the products are not yetaccessible everywhere. Options toconsider are selling fluid milk to acheese plant (family or industrial) orprocessing the milk on the farm andselling the products (cheese, yogurt,ice cream, soap, etc.) directly to theconsumer.

Sale of fluid milkSelling fluid milk to an industrialcheese maker is certainly the moststraightforward way to distribute theproduct. The dairy sheep producer canthen concentrate on producing highquality milk without the burden ofmarketing more individual products.

However, there are several constraintsto note. Cheese makers, even smallones, generally use equipment thatallows for processing a significantamount of milk. Keeping just each day’sproduction (most equipment will holdabout three days’ worth) in the coolingtank is not efficient. In fact, an isolatedsheep dairy producer will find it diffi-cult to sell fresh fluid milk to a profes-sional cheese maker because a singleproducer cannot deliver sufficientquantity. To solve this problem, severalproducers in the area near a processingplant need to pool their milk.

The problem is made more compli-cated by the uneven seasonal produc-tion in which the daily quantity of milkproduced is determined by thenumber of ewes lambing on the sameday or week. An individual producer’svolume is low at the beginning of themilking season and reaches a peakwhen all ewes are at milking. Thevolume then decreases rapidly. Thus,delivering fresh milk during the peakof production could be consideredbut is almost impossible at both endsof the production period.

To help alleviate this problem, produc-ers can freeze milk that cannot bedelivered on the day of production.Many excellent cheeses and yogurtscan be made from frozen milk if it isfrozen correctly.

When planning to sell milk it is essen-tial to include the purchase and oper-ating cost of a commercial freezer inthe financial plan. Freezing the milkwill increase its cost of production.

Farmstead productsBecause of the many constraintslinked to the sale of fluid milk andfluctuations in buyers’ willingness topurchase it, many producers choose todevelop their own finished productand sell directly to consumers—indi-viduals, restaurants or specializedstores and markets.

The process can be very rewarding aslong as one understands that:

■ It requires a very high quality andoriginal product. Average or incon-sistent goods might sell once butwill not bring repeat buyers. Itmight take years and much trialand error before arriving at theperfect product. Income could bemarkedly reduced during thisprocess.

■ It requires many hours of work tomarket the product. Visits topotential consumers, farmers’markets, store promotions, etc. aretime-consuming. Generally it willtake a full-time person to care forand market the cheeses. There issimply not enough time for thesame person to also feed, milk andcare for the animals.

■ It requires some initial investmentsuch as cheese room equipment,pasteurizer, cold room and agingrooms. In the case of yogurtmaking, the investment could befairly substantial.

In some areas (such as Wisconsin),selling cheese other than directly fromthe farm requires completing an 18-month apprenticeship and test to geta cheese maker’s license. A license isalso required to pasteurize milk.

135

E C O N O M I C S O F R A I S I N G D A I R Y S H E E P C H A P T E R 1 2

Freezing milk correctly calls for

very low temperatures (see

chapter 1) that only commer-

cial freezers can reach. Home

chest freezers do not have the

capacity to freeze milk rapidly

enough. Freezing too slowly

leads to processing problems

later because milk proteins

degrade.

Potential financialreturn of a dairysheep operation:An exampleIt is important for a producer to have aprecise idea of the operation’s poten-tial financial return so as to be able toadopt techniques that allow for abetter return.

ExampleBackgroundThe operation consists of a flock of300 East Friesian crossbred ewesmilked in a 2 x 12 Casse parlor with 12milking units, high line, cooling tankand commercial freezer. All the milkproduced is frozen and sold to acheese plant at the price of $ 1.32 perliter. All feed is purchased outside thefarm, but improved pastures aregrazed for 6 months of the year.

Lambing occurs in February, with astart of milking 30 days after lambing.This is a conservative weaning systemthat does not maximize milk produc-tion (see chapter 10,“Weaning ofLambs”). Lambs are weaned at 30 daysand raised on high-energy ration to aslaughter weight of 55 kg. The price oflive lamb is $1.54 a kilogram. Thisoperation seeks an average return perewe with moderately high cost of pro-duction.

Animal parametersThe prolificacy of East Friesian cross-bred ewes is 220% with a fertility of95%, an annual mortality rate of 3%and a culling rate at milking of 3%. Thetotal number of lambs produced is627 with a mortality rate of 13%between birth and sale at 55 kg. Atotal of only 220 adult ewes will beavailable for milking.

Sixty ewe lambs are kept for replace-ment, 20 are sold for breeding as are 4ram lambs. Ewe lambs are bred at 7–8months of age with a success rate of95%. Because of some ewe lambs notadapting to milking, only 50 ewelambs will be available for milking.

Milk production of adult ewes is 200liters in a milking period of 180 days,while the milk production of ewelambs is 130 liters in 110 days. Theflock average is therefore 187liters/animal milked.

FeedEwes are fed alfalfa hay (1.8 kg beforelambing, 2.8 kg after lambing), shelledcorn (0.45 kg/day before lambing) anda 16%CP concentrate during milking(0.9kg/day). As soon as possible, ewesare put on improved pastures such asKura clover-grass mix. The cost ofpasture is evaluated at $2 per ewe permonth.

Lambs consume an average of 47 kgof a creep ration (21% CP) and 122 kgof a finish ration (13%CP) and are soldin June–July when the market is gen-erally at its highest point.

EquipmentThe total investment for the milkingsystem including the commercialfreezer is $50,000. Other equipmentfor general sheep management isevaluated at $15,000 and the build-ings at $30,000. Initial investment forlivestock is $50,000.

LaborLabor is provided by one personworking full-time and one part-timeperson working 900 hours and paid$8/hour. The hired labor is included asan expense while the possible salaryof the manager of the operation isrepresented by the return to labor andmanagement.

136


Producers in the same area

should consider collaborating

to reduce cost and labor. Milk

marketing, cheese making

and/or cheese aging coopera-

tives already exist in North

America. Potential sheep dairy

producers should contact them

for information (see Appendix

A: Useful Addresses).

It is impossible to describe all

possible types of operations

because, as already men-

tioned, there are practically as

many types of operations as

there are producers.

Nevertheless, the example

here can help establish a

framework, keeping in mind

that it does not consider the

interactions that might exist in

a real situation. The figures

presented are merely an

example and are not based on

the real conditions, resources,

management skills and philos-

ophy of the operation.

137


Income-expense reportprice

number unit ($) total

Receipts

Slaughter lambs 406 55kg 1.54 38,962

Sale of ewe lambs 20 200.00 4,000

Sale of ram lambs 5 500.00 250

Sale of culled ewes 50 75.00 3,750

Sale of older rams 2 500.00 1,000

Wool 300 4kg 22.00 264

Milk 270 187.1 1.40 70,686

Total receipts 121,162

Variable expenses

Ewe feed

pasture 300 6mo 2.00 3,600

3 month hay 1.8kg 300 160kg .08 3,840

3 month hay 2.8kg 270 250kg .10 6,750

1 month corn .45kg 300 16kg .13 624

3 month 16%CP .90kg 270 82kg .15 3,321

2 month 16%CP .45kg 270 28kg .15 1,134

Total 19,269

Lamb feed

Creep feed 21%CP 545 47kg .20 5,123

Finish ration 13% CP 545 122 kg .14 9,309

Hay replacement ewes 60 300kg .11 1,980

Total 16,412

Other feed

Salt and minerals 970

Milk replacer 1000

Total 1970

Total feed 37,651

138


In this type of system, excel-

lent management of the dairy

ewes is essential to ensure a

high return. It would be unrea-

sonable, however, to expect a

high milk yield during the first

year of production

(table 3).

Income-expense report, continued

total

Livestock expenses

Shearing 600

Marketing-trucking of lambs 1,300

Vet-Med 1,500

Suppliessheep 1,000milking 1,500

Bedding 1,800

Utilitieselectricity freezer 2,500other 1,000

Machine operation cost 1,500

Ram cost 2,000

Hired labor $8/hour 7,200

Maintenance and repairs 1,500

Operating loan interest 2,000

Dairy cooperative cost (.21 cts/liter of milk) 6,065

Total livestock expenses 36,003

TOTAL VARIABLE EXPENSES 73,654

Fixed expenses

Sheep equipment 8% of US$ 15,000 1,200

Livestock 4% of $50,000 2,000

Building 7% of $30,000 2,100

Milking equipment 8% of $50,000 4,000

Pickup truck 2,000

Property taxes 2,000

Insurance 2,000

Total fixed expenses 15,300

Returns

Total income 121,162

Less variable expenses 73,654

RETURN TO LABOR AND CAPITAL 47,508

Less fixed expenses 15,300

RETURN TO LABOR AND MANAGEMENT 32,208$107/ewe

This is a relatively high return to labor and management. The operator workingan average of 2500 hours a year could expect an hourly rate of roughly $13.Milk accounts for 50% of the total receipt. With a high cost of production thereturn will greatly depend on the total milk production of the ewes and on thesale price of milk. Table 2 shows the expected return according to the averagemilk production of the flock and the price of milk.

Cost of milk productionIn the example, the ewe feed cost rep-resents more than 31% of the variableexpenses. It is possible to reduce thisamount without affecting the totalmilk production. Marie et al. (1998)have shown that high-producing eweshave better feed efficiency than lower-producing ewes. Therefore, withhigher-producing ewes, the ewe feedcost/receipt from milk ratio would begreatly improved. Other means ofreducing the overall cost of produc-tion can be found by relying on lowinput systems—generally at theexpense of volume.

Since the cost of milk production isthe difference between receipts fromthe sale of milk and expenses incurredsolely for its production, it shouldroughly be similar in many types ofoperation. Expenses from milk produc-tion cut across all expense categories.

Ewe feedNot all ewe feed is required for theanimals to produce milk. Feed costs atthe end of gestation and early lacta-tion are the same for meat-only anddairy operations. The difference lies ina longer, high-quality feeding periodthat involves concentrates andpasture. In the example, one-third ofthe ewes’ feed expenses can be attrib-uted to milking ($6,423).

Lamb feedIn a dairy operation, lambs areweaned at 30 days or earlier whichtranslates into a higher consumptionof expansive creep feed (+ 20kg/lamb) and of finish ration(+32kg/lamb) for a total of $4,621.Lambs weaned at an early age need toreceive a high protein ration andcannot be put on forage. The extracost cannot easily be decreased.

SuppliesSupplies for milking such as deter-gent, acid, brushes, liners, milk tubes,milk filters and lab costs for milkanalysis are estimated at $1,500.

UtilitiesThe single most expensive item is theoperating cost of the commercial-grade freezer, estimated at $2,500 forthe milking season.

139


In any situation, however, it is

best to prioritize expenses,

choose what can be cut or

reduced without affecting the

performance of the animals or

of the operator, and to never

compromise on the animals’

welfare.

Table 2. Expected returns according to the price of milk and milk yield.

Price ofmilk in Average milk yield of the flock in liters$/liter 140 160 180 200 220 240

1.00 -263 4,003 8,269 12,535 16,801 21,067

1.20 9,297 12,643 17,989 23,335 28,681 34,027

1.40 14,857 21,283 27,709 34,135 40,561 46,987

Table 3. Evolution of milk production at the Spooner Research Station between 1996 and 2003.

1996 1997 1998 1999 2000 2001 2002 2003

Age of ewes (years) Ewe lambs Ewe lambs Ewe lambs Ewe lambs All ages All ages All ages All ages+2 +2 +3 +2+3+4

# of ewes milked(min. and max.) 23-130 49-193 12-225 18-215 27-267 30-261 16-326 20-293

Date started milking 4/2/96 3/18/97 2/11/98 2/13/99 2/11/00 2/13/01 1/18/02 1/14/03

Date stopped milking 9/11/96 9/14/97 9/8/98 9/16/99 8/31/00 9/19/01 9/27/02 9/16/03

Total production (liters) 10,000 23,903 36,550 43,257 59,661 50,254 64,559 70,389

Berger, 2000

The steady increase in milk production shown in table 3 is due to more mature ewes in production, better genetics andabove all better management of the milking ewes (introduction of the DYI weaning system).

Machine operation costHalf of the machine operation cost($750) can be imputed to milking.

LaborWith an efficient system, milking 270ewes should not take more than 4–5hours a day as an average over the160-day season. Considering that theproducer’s salary should be at least$13/hour, the cost of labor for milkingis $10,400. Moreover, about three-fourths of the hired labor cost isattributable to the dairy operation($5,400). The total labor cost for themilking and extra care of 270 ewes istherefore $15,800.

Operating loanHalf of the operating loan cost ($1000)is imputable to milking.

Milking systemThe cost of amortization of themilking system ($4,000) is imputableto milking.

The total cost of production of the50,490 liters of milk in the systemdescribed above is $34,844 or0.69/liter. The difference between thecost of production and the sale priceof milk represents the deficit or theprofit of the dairy operation in thesheep enterprise. The cost of produc-tion varies greatly according to theaverage milk yield of the flock asshown in table 4.

Cost of cheese productionThe economic potential for marketingsheep cheese products in the U.S. willdepend on the cost of producingthose cheeses. With sheep’s milk, theincreased solids and increased cheeseyield will impact favorably on produc-tion costs. However, the initial cost ofthe milk will be the greatest factoraffecting the cost of producingcheeses.

At the actual price (2003) of $1.40/literin the U.S., milk cost per kilogram ofcheese is very comparable to that ofgoat’s milk. The other significant costof producing cheese involves manu-facturing. Table 5 shows a comparisonof costs of producing a Cheddar-typecheese from cow’s, goat’s and sheep’smilk. Currently, goat and sheep’s milkprocessors do not have establishedmarkets for whey and processors arelosing potential revenue from thosecomponents.

Cheddar cheese is produced most effi-ciently from milk with a casein:fat ratioof .70. With a casein:fat ratio of .60 insheep’s milk, the processor would losesome of the excess fat in the wheyunless some separation of cream tookplace prior to cheese making. If theprocessor tries to maximize thecheese yield by designing a highmoisture cheese that uses thecasein:fat ration of .60, the estimatedcost of production of that cheesewould be as shown in table 6. The cost

140


Since the average milk yield of

the flock depends on the per-

formance of each individual

ewe and on the number of

ewes that stay healthy and in

lactation, the cost of produc-

tion is directly linked to the

management skill of the

operator.

Table 4. Cost of producing a liter of milk based on the average milk yield

Average milk yield of the flock in liters140 160 180 200 220 240

$.97/l $.85/l $0.75/l $0.68/l $0.62/l $0.56/ll

Table 5. Comparison costs of producing Cheddar cheese from cow, goat andsheep milk (38% moisture, 54.6-55.1% FDB)

Cowa Goata Sheepa

Raw milk costs/100kg 28.05 48.4 140

Cheese mfg. costs/100kg of milkb 7.7 7.7 7.7

Credit for excess cream (.64) — —

Credit for whey cream (.42) — —

Credit for whey solids (1.1) — —

Total cost/100kg of milk processed 33.59 56.1 147.7

Kg of cheese/100kg milkc 10.34 10.72 18.79

Cost per kg of cheese 3.24 5.23 7.86aMilk composition assumed to be: cow, fat = 3.95%, protein = 3.33 %; goat, fat = 3.9%,protein = 3.3%; sheep, fat = 6.9%, protein = 5.7%bSmall processors cost of $.77/kg production cost for Cheddarc Assuming 93% fat recovery, 96% casein recovery and a solids recovery factor of 1.09

of producing this sheep cheese isslightly lower than the Cheddar-typeprimarily due to more efficient use ofthe protein and fat in sheep milk.

In a farm situation, the manufacturingcost might be somewhat higherbecause of smaller volumes and lessefficient equipment. Add the costs ofexpenses related to the preparation ofthe cheese (curing, aging, etc.) andexpenses related to marketing, as wellas the profit of the producer. The finalretail price of the cheese could befairly high. To command such a highprice, the product must be the highestquality in terms of taste, character,texture and consistency.

Domestic cheese producers face thechallenge of producing cheese forwhich consumers are willing to paymore than they pay for subsidizedimported sheep cheeses, such asPeccorino.

ConclusionSheep dairying can provide a decentreturn to a producer with good man-agement skills if certain conditionsexist:

■ A well-thought-out financial plan

■ A readily available market

■ A long-term commitment

■ Good control of production costs

■ Ewes with good milking ability

■ An understanding of the impor-tance of a good milking system

■ Membership in a group, associa-tion, or cooperative that supportsthe marketing of the milk and aidsin solving in technical problems.

ReferencesBerger, Y.M.. 1998. An economic com-

parison between a dairy sheep anda non-dairy sheep operation. In:Proceedings of the 4th Great LakesDairy Sheep Symposium, June26–27, 1998, Spooner, Wisconsin.

Berger, Y.M.. 2000. As a sheepproducer, should I consider sheepdairying? In: Proceedings of the47th Spooner Sheep Day, Spooner,University of Wisconsin-Madison.

Marie C., Bocquier F. and F. Barillet.1996. Influence du potentiel laitiersur les composantes de l’efficacitéalimentaire de brebis Lacaune. In:Proceedings of Renc. Rech.Ruminants, INRA, Institut del’Elevage, 1996, 3, 297-300

Mills, O. 1995. Milking techniques andparlor management. In:Proceedings of the First GreatLakes Dairy Sheep Symposium,March 30, 1995, Madison,Wisconsin.

Wendorff W.L., 1995. Economic poten-tial for sheep dairy products in theU.S. In: Proceedings of the 1stGreat Lakes Dairy SheepSymposium. March 30, 1995.Madison, Wisconsin.

141


Domestic processors must

develop unique products that

are not in direct competition

with commodity cheeses or

imported sheep cheese.

Table 6. Comparison cost of producing a Manchego-type cheese from cow andsheep milk (45% moisture, 54.4% FDB)

Cowa Sheepa

Raw milk costs/100kg 28.05 140

Cheese mfg. costs/100kg of milkb 15.73 15.73

Added cream cost .79

Credit for whey cream (.73) —

Credit for whey solids (1.56) —

Total cost/100kg of milk processed 42.28 155.73

Kg of cheese/100kg milkc 12.96 20.74

Cost per kg of cheese 3.26 7.50a Milk composition assumed to be: cow, fat = 3.95%, protein = 3.33 %; sheep, fat = 6.9%,protein = 5.7%b Small processors cost of $1.21/kg production cost for Cheddarc Assuming 90% fat recovery, 96% casein recovery and a solids recovery factor of 1.12


142

Milking equipmentDeLaval, Inc.

11100 N. Congress AveKansas City, MO 64153-1296Phone: 816-891-7700www.delaval.com

Westfalia Surge1880 Country Farm DriveNaperville, IL 60563Phone: 877-973-2479Fax: 630-369-9875www.westfaliasurge.com

The Schlueter Company3410 Bell streetJanesville, WI 53545Phone: 608-755-5444Fax: 608-755-5440www.schlueter.com

The Coburn CompanyP.O. Box 147Whitewater, WI 53190Phone: 1-800-776-7042www.coburnco.com

Milk testingDHIA

National Dairy HerdImprovement AssociationSuite #102, 3021 E. Dublin GranvilleRoadColumbus, OH 43231Phone: 614-890-3630Fax: [email protected]

Performance recording softwareEWE BYTE Sheep Management

SystemP.O. Box 375FergusOntario N1M 3E2CanadaPhone: 519-787-0593Fax: [email protected]/associations/ewebyte/ewebyte.htm

Importation ofgermplasm and artificialinseminationUSDA-APHIS (importation regula-

tions, importation permits)USDA-APHISVeterinary Services-NationalCenter for Import/Export4700 River Road Unit 40Riverdale, MD 20737Phone: 301-734-3277 Fax: 301-734-4704www.aphis.usda.gov/vs/import_export.htm

Small ruminant geneticsR.R. 3MarkdaleOntario N0C [email protected]

Super Sire Limited (source of semen,embryos, artificial insemination)4000 University Rd.Hopland, CA 95449Phone/fax: [email protected]

GENELEX (source of Lacaune fromFrance)Upra-LacauneRoute de Moyrazès12033 Rodez Cedex 9, FrancePhone: (33) 565 73 78 14Fax: (33) 565 73 78 [email protected]

BRITBREED Ltd.Dr. James Mylne, Director1 Airfield FarmCousland, DalkiethMidlothian EH22 2PEScotlandPhone: 01875 320727Fax: 01875 320734www.britbreed.co.uk

AssociationsDairy Sheep Association of North

AmericaCarol Delaney, SecretaryUniversity of Vermont570 Main St., 200B Terrill HallBurlington, VT 05405Phone: 802-656-0915Fax: [email protected]

143

Resources

Appendix A

Wisconsin Sheep Dairy CooperativeN 50768 County road DStrum, WI 54770Phone: [email protected]

Ontario Dairy Sheep AssociationWooldrift farmRR3Markdale, Ontario NOC 1HOCanadaPhone: 519-538-2844Fax: [email protected]

Vermont Shepherd 875 Patch RoadPutney, VT 05346Phone: 802-387-4473Fax: [email protected]

The British Sheep DairyingAssociationBSDA Secretary The Sheep CentreMalvern, Worcestershire WR13 6PHENGLANDPhone: +44(0)1684 892 661 Fax: +44(0)1684 892 663 [email protected] www.sheepdairying.com

Research—Extension

University ofWisconsin–MadisonSpooner Agricultural Research

StationYves Berger (management)W6646 Highway 70Spooner, WI 54801Phone: 715-635-3735Fax: [email protected]

Center for Dairy ResearchBabcock Hall 226B1605 Linden driveMadison, WI 53706Phone: [email protected]/cheesedb.nsf/

Department of Animal SciencesDave Thomas (genetics)1675 Observatory driveMadison, WI 53706Phone: 608-263-4306Fax: [email protected]

Department of Food ScienceBill Wendorff (cheese, dairyproducts)Babcock Hall A203B1674 Linden driveMadison, WI 53706Phone: [email protected]

The Babcock Institute forInternational Dairy Research andDevelopment240 Agriculture Hall1450 Linden DriveMadison, WI 53706-1562Phone: 608-265-4169Fax: [email protected]://babcock.cals.wisc.edu

University of Vermont UVM Extension

Carol Delaney (management)Small Ruminant Dairy Specialist 570 Main St., 200B Terrill HallBurlington, VT 05405Phone: 802-656-0915Fax: [email protected]

Cornell UniversityDr. Michael L. Thonney

114 Morrison HallCornell UniversityIthaca, NY 14853Phone: 607-255-9829Fax: [email protected]

Dr. Pascal A. OltenacuB21 Morrison HallCornell UniversityIthaca, NY 14853Phone: 607-255-2852Fax: [email protected]

Journals, books,resource materialsJournal of the Dairy Sheep

Association of North AmericaPat Elliot, Editor23246 Clark Mountain RoadRapidan, VA 22733Fax: [email protected]

Sheep Dairy NewsSecretary, British Sheep DairyingAssociation The Sheep CentreMalvern, Worcestershire WR13 6PHENGLANDPhone: +44(0)1684 892 661Fax: +44(0)1684 892 [email protected]

Practical Sheep Dairying (out ofprint)Author: Olivia MillsPublisher: HarperCollinsISBN: 0722507313

Past Proceedings of The Great LakesDairy Sheep SymposiaThe Great Lakes Dairy SheepSymposium has been heldannually since 1995 in the U.S. andCanada. Past proceedings can beviewed at: www.uwex.edu/ces/animalscience/sheep.

144


145

Floor plans of milkingparlors & barn designs

Appendix B

Lacaune ewes in a typical barn in the Roquefort area

146

Floor plans of milking parlors.

147

A P P E N D I X B

Floor plans of milking parlors.

148

Barn for 150 ewes.

149

A P P E N D I X B

Barn for 250 ewes.

150

Barn for 300 ewes

151

A P P E N D I X B

Barn for540 ewes

Copyright © 2004 by the Board of Regents of the University of WisconsinSystem doing business as the division of Cooperative Extension of theUniversity of Wisconsin-Extension. All rights reserved. Send copyright inquiriesto: Manager, Cooperative Extension Publishing, 432 N. Lake St., Rm. 103,Madison, WI 53706.

Authors: Yves Berger, Pierre Billon, François Bocquier, Gerardo Caja, AntonelloCannas, Brett McKusick, Pierre-Guy Marnet, David Thomas.

Produced by Cooperative Extension Publishing, University of Wisconsin-Extension.

An EEO/AA employer, the University of Wisconsin-Extension, CooperativeExtension provides equal opportunities in employment and programming,including Title IX and Americans with Disabilities (ADA) requirements.

This publication is available from your Wisconsin county Extension office orfrom Cooperative Extension Publishing. To order, call toll-free: 1-877-WIS-PUBS(947-7827) or visit our web site: cecommerce.uwex.edu.

Fundamentals of Sheep Dairying in North America (A3767)

Principles of Sheep Dairying in North America (A3767)

Documents

Transcript of Principles of Sheep Dairying in North America (A3767)