A Summary of The Final Report The Ecology and Interactions ... · population was declining while...

The Ecology and Interactionsof White-tailed Deerand Eastern Coyotesas Influenced by HumanActivities in Nova Scotia

Prepared forThe Nova Scotia Departmentof Natural Resources

ByBrent R. PattersonBevan A. LockBruce A. Macdonald

December 1999

A Summary of The Final Report

Natural Resources

Foreword

Toward the end of the 1980s Nova Scotia’s deerpopulation was declining while its coyote popula-tion was increasing. This brought on publicconcern that coyotes were reducing deer num-bers, adding pressure on a deer herd they felt wasalready suffering due to clear cutting of winterhabitat. Wildlife biologists had little local infor-mation to explain what was happening.

To get scientifically defensible data on theseevents, the Wildlife Division of the Departmentof Natural Resources initiated a study of deerwintering behaviour and the effects of coyotepredation and forest harvesting on the deer herd.Three graduate students (two M.Sc. candidatesand one PhD candidate) from Acadia University’sCentre for Wildlife and Conservation Biology,were the principal researchers in the project.Funding was provided in part under the Canada/Nova Scotia Cooperation Agreement for ForestryDevelopment 1991–95.

This document is a summary of the researchers’final report to the Department. Opinions ex-pressed are those of the researchers and notnecessarily those of the Department of NaturalResources.

A. P. DukeManager Wildlife ResourcesDecember, 1999

Additional copies of this document are availablefrom:The Wildlife DivisionDepartment of Natural Resources136 Exhibition St.Kentville, NS B4N 4E5

Contents

Section 1Aspects of The Ecology of White-tailed Deer inNova Scotia 5

1.1 Introduction 5

1.2 Study Areas 5

1.3 Deer Population Trends 7

1.4 Deer Survival: Causes and Rates of Mortality 7

1.5 Deer Distributionand Movements 10

1.6 Deer Habitat Assessment 11

1.7 Habitat Use Analysis and Estimate of SuitableWinterHabitat Requirements 12

1.8 Deer Physical Conditionand Relationships with Habitat 15

Section 2Aspects of The Ecology of The Eastern Coyote inNova Scotia 17

2.1 Coyote Social Organizationand Spatial Distribution 17

2.2 Coyote Movements andActivity Patterns 20

2.3 Coyote Food Habits and Predation on White-tailed Deer 21

2.4 Factors Influencing Killing Rates of White-tailedDeer by Coyotes in Nova Scotia 25

2.5 Winter Condition of Coyotes in Relation to PreyDensity 26

Section 3Conclusion: Managing Deer, Coyotes,and Forestry in Nova Scotia 27

3.1 Deer 27

3.2 Coyotes 27

3.3 Forestry 27

ISBN 0-88871-619-2

© Crown copyright 2000Province of Nova Scotia

Section 1

Aspects of The Ecology of White-tailed Deer in Nova Scotia

hunters to taking males only. That same year, theDepartment of Natural Resources (DNR) began afour-year study to examine the effects of landscapepatterns, forest harvesting, winter severity, andcoyote predation on deer survival in the province.

This is a summary of that study. The full reportand other relevant documentation is available atthe DNR library, Halifax, Nova Scotia.

1.2 Study Areas

Two distinct geographic areas of the province werechosen for our investigation (Figure 1). Theseareas were selected to allow an examination of theeffects of winter severity and differing prey densi-ties on deer and coyote life histories, and to studycoyote-deer interactions in Nova Scotia.

The Queens County study area was located incentral southwestern Nova Scotia, and includedthe eastern half of Kejimkujik National Park(about 200 sq km) and 300 sq km of primarilyforested land east of the park. This area generallyreceives accumulations of less than 20 cm ofsnow in winter. For brevity, it is referred to as QCin this report.

The Cape Breton study area (CB) was centeredaround the 24 sq km Eden deer wintering area(DWA), which typically contains about 200 deerfrom January through March. The Cape Bretonarea has two distinct elevation zones: theCreignish Hills at about 300 m above sea level(CBH, Cape Breton Highlands), and the RiverDenys Basin (CBL, Cape Breton Lowlands).

The Ecology and Interactions of White-tailed Deer and Eastern Coyotes as Influenced by Human Activities in Nova Scotia 5

1.1 Introduction

When Europeans first arrived in Nova Scotia inthe 1600s, they did not encounter any white-tailed deer. In 1908, nine white-tailed deer werereleased in the province near Bear River, DigbyCounty. Food was abundant and their numberssoon swelled. In 1916, a hunting season for bucksonly was begun in some counties, and by 1940the season was opened province-wide for twoanimals of either sex to each hunter. Over thenext 50 years the number of white-tailed deer inNova Scotia continued to increase, except forbrief population crashes in the 1950s and early1970s. By the winter of 1986–87, their numbershad apparently exceeded the province’s carryingcapacity and crashed again – in spite of liberalhunting regulations and the rapid increase in thenumbers of eastern coyote in the province.

Deer in Nova Scotia are at the northern limit oftheir range, and are affected by a number offactors including losses due to winter malnutri-tion; predation by coyotes, bobcats, and domesticdogs; vehicle collisions; and legal and illegalharvest. Approximately 80 per cent of NovaScotia is forested, and forestry is an importantindustry in the province. For deer, the bestfeeding comes from small, scattered cuttingsrather than large cuts, which also provide lesscover in severe weather conditions.

The key to maintaining a stable deer population isin understanding how all these limiting factorsinterrelate, as well as being able to evaluate (orestimate) and manipulate one or more factors.Wildlife managers generally have the most controlover legal hunting, and in 1993 Nova Scotiainstituted a province-wide buck law limiting

6 The Ecology and Interactions of White-tailed Deer and Eastern Coyotes as Influenced by Human Activities in Nova Scotia

Figure 1Locations of the QC, CBH, and CBL study areas.

Nova Scotia

Queens CountyStudy Area

Cape BretonStudy Area

5 0 5 10 Kilometres

Kejimkujik National Park

Queens County Study Area

Route 8

Lake Rossignol

5 0 5 10 Kilometres


Highway 104

West Bay

St. Patrick'sChannel

Eden DeerWintering Area

Cape Breton Highlands Area

Cape Breton Lowlands Area


1.3 Deer Population Trends

Deer numbers in Nova Scotia have fluctuated sincetheir introduction at the beginning of the century.We estimated current deer numbers by countingtheir pellet groups along 30 survey lines, 1 kmlong and 2 m wide, in both study areas eachspring.

Although deer were evenly distributed thoughoutCB from May through November, most deer thenmigrated to their wintering grounds in CBL.

Based on the pellet group surveys, estimates ofwinter deer density averaged around 5.5 deer/sqkm in CBL, 2.2 deer/sq km in CBH, and 9.8 deer/sq km in the Eden DWA. However, fewer deer –only 4.7 deer/sq km – migrated to the Eden DWAduring the mild winter of 1995. Estimates for theQC area were consistently lower, at 2.5 deer/sqkm, with a relatively even distribution.

Figure 2The relative abundance of white-tailed deer inthe Queens County (QC) and Cape Breton (CB)study areas, 1992–1997.

To gain a clearer picture of the winter distribu-tion of deer in CB we also conducted aerial andground surveys during the winters of 1996 and1997. The results of these surveys are summa-rized in Figure 2, and show a clear difference inthe winter distribution of deer between thehighlands and the lowlands.

1.4 Deer Survival: Causes and Ratesof Mortality

A major difficulty in the analysis of wildlifepopulation dynamics is in obtaining reliableestimates of survival. Most studies have concludedthat hunting is the major cause of mortality inexploited deer populations. As a result, harvest byhunters is frequently targeted by deer managementplans. More recently however, studies conducted innorthern New Brunswick and on Vancouver Islandin British Columbia concluded that predation wasthe major source of mortality. At present, our lackof knowledge of non-hunting mortality rates ishindering the effective management of white taileddeer populations.

Deer capture and handlingBetween February 1994 and March 1997, wecaptured 124 deer and fitted them with radio-collars. Capture methods included box trapping,trapping with a ground-based rocket net, shoot-ing with drug-loaded darts, and netting from ahelicopter. The radio collars contained mortalityswitches that doubled the pulse rate of theemitted signal if the collar remained motionlessfor more than six hours.

Collared deer were checked for mortality signalsat least twice a week from April to mid-October,and four times a week through the rest of theyear. When a dead deer was located, the site andcarcass were examined to determine the cause ofdeath. A mortality was classified as a predator killif there was evidence of attack or chase. Weidentified the species of predator involved fromtracks, bite marks, and other related signs.0

2

4

6

8

10

12

1992 1993 1994 1995 1996 1997

Eden DWACB TotalQC Total

Deer

per

km

2


Causes of deathWe documented 45 deaths during a period ofmonitoring equivalent to 293 “deer-years,” whereone deer-year equals 365 days of telemetrycontact with deer. It could be the result of moni-toring a single deer for one year, or several deerfor a total of 365 contact days.

Sixteen deaths were due to predation (11 coyote,3 bobcat, 1 lynx, 1 unknown), 9 to registeredharvest, 12 to unregistered harvest, 4 to malnutri-tion, and 4 to other natural causes. Unregisteredharvest includes deer that were illegally harvested,lost due to wounding, or harvested but notreported by aboriginal people. Harvest by hunterswas the most significant cause of death for adultmales in both study areas. In spite of the fact thatlegal non-aboriginal hunting was restricted tobucks only starting in 1993, there was little effecton herd productivity, as does produced fawns atrates appropriate for their age throughout thisstudy. However, there was no noticeable increasein deer numbers until 1997. Given the mildwinters during this period, it seems likely thatpredation and illegal harvest of antlerless deerwere to blame.

Deer survival ratesBeginning with the birth of the new fawn crop,assumed to be June 1, we divided the biologicalyear into four intervals approximating to sum-mer, fall (which includes all days in bow andfirearm seasons), winter, and spring. The resultsare shown in Figure 3.

Only deer collared for more than 12 days wereincluded in survivorship estimates. Separate rateswere calculated for adult (more than one year)males, adult females, and fawns – although fawnswere grouped with adults when there were notenough data to consider them alone. In QCseparate comparisons were made for deer livinginside and outside Kejimkujik National Park.

Annual survival rates ranged from 41.9 per centfor adult males in CB to 94.6 per cent for adults inKejimkujik National Park (where only two deathsoccurred, both females). Survival rates weresignificantly lower for males than females in bothareas due to sex-biased harvesting, higher wintermortality, and higher predation rates. In QC,survival rates were significantly higher for deerliving within Kejimkujik National Park. As preda-tion was the only cause of death for deer inKejimkujik, and predation rates were virtually thesame within and outside the park, hunting-relatedmortalities outside the park were clearly responsi-ble for the lower survival rates we observed.

The annual survival rates of fawns were approxi-mately 44 per cent in CB, and approximately 47per cent in QC. These are higher than annualsurvival rates from studies in New Brunswick andMinnesota, but similar to those documented inQuebec. Many factors influence fawn survival,including harvest levels, predation, winter sever-ity, and habitat quality, so regional differences insurvival are expected. Survival rates of fawns inboth areas were lower than those of adult femalesbetween December and May.

We examine the effects of hunting and predationin more detail in section 2.3


Figure 3Survival and cause specific mortality rates, of radio-collared white-tailed deermonitored in the Queens County and Eden Study Areas, February 1994 – March 1998.

Dec 10 – March 31 April 1 – May 31 June 1 – Sept I5 Sept 16 – Dec 9 Annual

Cohort Rate No. Rate No. Rate No. Rate No. Rate No.Radiodays Radiodays Radiodays Radiodays Radiodays

CB Adult Females 0.943 7645 0.972 4330 0.975 8629 0.912 6460 .815 27064Predation1 0.057 0.014 0.012 0.013 (9.3%)Registered harvest 0.0 (0.0%)Unregistered harvest2 0.075 (6.7%)Natural mortality3 0.014 0.012 0.0 (2.4%)

CB Adult Males 0.701 948 1.0 771 1.0 1537 0.597 993 .419 4249Predation 0.I00 0.067 (14.7%)Registered harvest 0.268 (18.8%)Unregistered harvest 0.067 (4.7%)Natural mortality 0.199 (19.9%)

CB Fawns 0.634 492 0.936 923Predation 0.183Natural mortality 0.183 0.064

QC Adults (Park)4 0.970 3728 1.0 2179 0.975 4267 1.0 3393 .946 13192Predation 0.030 0.025 (5.4%)

QC Adult Fem. (Non Park)5 0.936 3436 0.974 2347 1.0 4793 0.851 2761 .778 13247Predation3 0.032 0.026 0.147 (5.6%)Registered harvestUnregistered harvest (13.4%)Natural mortality 0.032 (3.2%)

QC Ad. Males (NonPark)5 0.936 3436 0.974 2347 1.0 4793 0.544 741 .496 11317Predation3 0.032 0.026 (5.6%)Registered harvest 0.263 (24.0%)Unregistered harvest 0.175 (16.0%)Natural mortality 0.032 (3.2%)

QC Fawns (Non Park) 0.646 566 1.0 838Predation2 0.164Natural mortality 0.164

1 Of the 16 instances of predation observed during this study, 11 were attributed to coyotes, 3 to bobcats, 1 to a Iynx, and 1 unknown.2 Unregistered harvest includes deer which were known or suspected to have been illegally harvested, abandoned or lost (wounding loss), or harvested and

not reported by aboriginal people.3 Natural mortality includes deaths from malnutrition, old age, accidents, or other naturally occurring sources (excluding predation).4 Males and females were pooled for this analysis since we had little data on males in the Park and no males died within the Park. There was not enough

data to consider the survival rates of fawns in the Park.5 Data for males and females was pooled except during autumn because we observed no male moralities during other times of the year.


1.5 Deer Distributionand Movements

In this part of the study we examined the effectsof climate, geography, predation, and human useof the landscape on seasonal distribution of deer,their survivorship, activity, and winter physicalcondition in Nova Scotia.

Migrations are the seasonal movements of anindividual deer from a winter home range to asummer home range area and vice-versa. Disper-sals differ from migration, as they are “one-way”movements away from the animal’s place of birth.In many areas, male deer disperse before reachingsexual maturity, which may be a means of avoid-ing inbreeding.

MigrationsDuring the four-year study, radio-collared deerwere relocated mostly from the ground, usingportable receivers and antennae. The actuallocation of an animal was determined by takingthe mid-point of at least two bearings within a10-minute time interval. This technique is accu-rate to within about 90 m at a range of 1 km. Ahelicopter was used when deer could not belocated from the ground.

In CB between 14 per cent of collared deer (4 of29 animals, winter 1997) and 67 per cent ofcollared deer (12 of 18 animals, winter 1994)migrated to the Eden DWA. In QC we only notedmigrations during the winter of 1994, when 36per cent of collared deer (4 of 11 animals) mi-grated to feed in the vicinity of forest harvestingoperations. Migration distances in both areasaveraged just under 10 km.Our observations indicate that deer in NovaScotia mostly migrate in response to snow that ismore than 20 cm deep.

YardingYarding is a behaviour learned by deer in regionsthat normally have severe winter weather. Theanimals migrate to their wintering areas (“deeryards”) before deep snow hampers their move-ments and makes them vulnerable to predators.Most deer migrated from CB to the Eden DWAby early January, and returned in April. Thetrigger to return to summer ranges seemed to bemean weekly temperature highs of +5°C and lowsof -5° C. Deer may benefit from delaying thereturn to their summer ranges until the risk ofsnow storms has diminished. However, we notedthat some deer made several trips between theirsummer and winter ranges if the weather wassuitable, even though they had already migratedto the Eden DWA. Overall, it seems that deer inNova Scotia congregate in winter yards onlywhen conditions warrant it.

Dispersal movementsDispersals were noted primarily among yearlingmales. Of 18 males monitored to the end of theirsecond fall, nine dispersed an average of 19 km,mostly in late summer or early fall. By compari-son, of 15 older males that were monitored forjust over a year, none dispersed. We noted onlyfive dispersals by female deer during 137 deer-years of monitoring.

Seasonal home range useDeer home ranges in CB were smaller than thosein QC during winter, but slightly larger in sum-mer. About twice as much winter food is availablein CB than in QC, whereas during the summer,food was plentiful in both areas.


The influence of forest harvesting on deermovements and activity patternsExtensive clear cutting has caused deer to aban-don many wintering areas throughout northeast-ern North America. New guidelines now limit thesize of clear-cuts throughout the region. In NovaScotia, no single clear-cut in any area designatedas a DWA should exceed 10 hectares. With theexception of very small DWAs, a disturbance ofthis size may do no more than locally displacesome deer while providing an abundance ofwinter food. Many authors have noted that deerare attracted to forest harvesting operations inwinter as a source of food. In spite of this, har-vesting operations create disturbances that maycause deer to expend excessive energy.

We tested the effects of controlled harvesting ondeer movements in winter by cutting within theknown home ranges of collared deer, then moni-toring their subsequent movements. Ten hectaresof forest were harvested in the Eden DWA – onCrown land and adjacent private land – inaccordance with the Forest/Wildlife Guidelines andStandards for Nova Scotia (1989). To properlyassess the extent of our disturbance, we moni-tored deer activity patterns before, during, andfollowing harvest operations.

During felling, deer moved out of the harvestarea temporarily, although many of these sameanimals would return to feed there “after-hours.”Though it seems our forestry operations did notcause a permanent displacement of deer outsidetheir home range, and did not result in an in-crease in deer activity, the harvest took placeduring a particularly mild winter when theanimals had little need for thermal cover andwere able to move about freely to forage. Theresults might have been different had we beenable to conduct this experiment during a severewinter. Current standards in the Guidelinesshould meet the needs of deer, but the planningof forest harvest operations in and around DWAsmust consider their needs during severe winters,even if these occur infrequently.

1.6 Deer Habitat Assessment

Carrying capacity – the maximum number ofanimals that a particular area of habitat cansupport for an extended period of time – isaffected by factors such as the composition anddensity of vegetation, the preferences of deer fordifferent types of forage, and the abundance ofpreferred forage.

In this part of the study we determined thepotential amount of forage available per hectarein each study area – ie, the total amount ofwoody forage within the physical reach andcapability of deer – as well as the amount actuallybrowsed by deer. In both study areas typicalwinter food for deer (stems of woody species)from sample plots was tallied, gathered, thendried and weighed, during spring and fall of1994–95 and 1995–96. Total available browse wasrepresented by stems located between 0.3 and 2.0metres above the ground.

The QC study area had an average of 10,660 stemsper hectare, whereas the CB area averaged morethan twice this amount, 22,790 stems per hectare.In QC the preferred species included red maple,witch hazel, wild raisin, and red oak. In CBL redmaple, wild raisin, and aspen were preferred.

In QC the majority of available stems werelocated in the regenerating type of cover, ratherthan in mature cover types. In contrast, in CBthere were rather more stems available in allmature cover types combined (hardwood,softwood, and mixed-wood), but there weresignificantly more stems available in regeneratingcover types than in mature softwood cover.

The mean mass of browsed stems was signifi-cantly higher in CBL than in QC, probablyreflecting the higher density of overwinteringdeer in CBL. In CBL red maple was browsedmore than aspen. In QC red oak was browsedmore than red maple and witch hazel, which wereboth browsed equally.


Since the type of forest is very different betweenthe two study areas, it is not surprising thatbrowse availability also differed between them.Recent research suggests that winter food sourcesother than browse – such as litterfall (lichens andbranches) – may also be important. As litterfall ismore abundant in unharvested areas, it appearsessential to maintain these areas.

We used results published elsewhere to determinethe potential energy and protein content ofwoody stems, and converted this to the numberof deer the area is capable of supporting. Wesuggest 11 deer per square kilometre could besupported in CBL during a typical yardingperiod. However, this figure is a maximum thatwould be reduced by poor accessibility, high deerdensities, or prolonged severe winter weather.

1.7 Habitat Use Analysis andEstimate of Suitable WinterHabitat Requirements

Effective wildlife management depends on a goodunderstanding of animal habitat selection, as wellas the ability to predict habitat needs. The firststep involves observation of a species in itsnatural habitat, and surveying the biologicalcomponents of that habitat in detail. The nextstep involves studying the relationship betweenthe biological data and the physical attributes ofthe habitat. Advances in technology – notably thedevelopment of Geographic Information Systems(GIS) – have greatly enhanced our ability toundertake this second stage.

Much previous research has been dedicated toshowing how deer in the northern part of theirrange seek out overwintering areas (yardingbehaviour, section 1.5) We sought to complementthis research by developing statistical models thatidentify preferred wintering areas. Using GIS wewere able to overlay the locations of radio-collared deer collected during three winterseasons on detailed habitat maps. This providedus with a visual representation of real scenarios.

We then applied these models to the study areasto predict where deer were likely to be found.When verified, we hope our results will enablewildlife managers to predict suitable winterhabitat for deer, thus allowing the province’ssilviculture demands to be met as well as main-taining critical deer winter habitat.

We located radio-collared deer several timesduring each 24 hour period to ensure that loca-tions were representative of the animals’ 24 hourroutine. For QC, 998 usable winter locations wereobtained using 18 deer, and for CBL, 617 loca-tions were obtained from 33 deer.

Vegetative samplingBased on telemetry locations, browse data, andground and aerial surveys, we constructed a mapof overwintering deer concentration sites (DCs).We also selected and mapped seemingly suitablestands of comparable cover-type, height, and agethat exhibited little or no use by deer to representno deer concentration sites (NDCs). One DC siteand one NDC site, each of about 100 hectares,were selected within each study area for intensivevegetative sampling. We took measurements ofseveral characteristics of the overstory, includingtotal number of stems, number of trees byspecies, and stand age and height. In theunderstory we sampled a total of 16 variables,including presence of deer pellets or browse,distribution of moss, and total number of shrubspecies in three levels of understory.

From this data we generated maps showing theprobability of deer being present in all parts ofboth study areas. The accuracy of the modelgenerated was then tested by comparing it toactual deer locations not used in the developmentof the model.


Figure 4Probability of deer occurrence in CB.


Figure 5Probability of deer occurrence in QC.


Preferred deer wintering areasThe models we developed consist of a set ofequations. Overall, it appears that deer in CBprefer low elevations (because higher elevationsare colder, receive more snow, and have longerwinters); high site quality (larger overstory treesand diverse and plentiful understory vegetation);absence of a second story under the main canopy(typically made up of regeneration 2 to 4 m inheight intertwined with fallen dead wood – thisprovides very little warmth or forage for theanimals, and may impede deer movement); andproximity to, or interspersion with high conifer-ous canopy (for cover). They tend to avoidnortherly slopes.

By comparison, deer in QC seem to select areasof diverse cover type with easterly exposure and10 to 20 per cent slope. They are less influencedby proximity of stands with a high percentage ofconiferous crown closure. Interestingly, they werenot affected by an increasing distance to clear-cuts or partial cut edges (which are prime deerfeeding areas). This is understandable in light ofthe relatively mild winters in southern NovaScotia, which mean that the animals have moreenergy to forage further afield.

Our models proved to be very precise, accuratelyidentifying 26 per cent of the Cape BretonLowlands and 44 per cent of Queens County asareas with a high probability of deer occurrence.We believe that this represents a significantimprovement over past wildlife models.

1.8 Deer Physical Conditionand Relationships with Habitat

Understanding the relationships between winterhabitat, deer population density, survival, andoverall fitness is one of the central aims of thisreport. Here we attempt to find the relationshipbetween age specific fertility, survival, and overalldeer fitness to winter severity and populationdensity, and make recommendations for assign-ing optimal population goals for deer herds.

UrinalysisThe analysis of urine-soaked snow from free-ranging deer can be used to assess the animals’physical condition. We collected urine samples inboth study areas during six two-week periods inthe winter of 1995. Unfortunately, there was verylittle snowfall that winter, resulting in low samplesizes. Each sample was analyzed for the presence ofurea nitrogen, creatinine, potassium, and sodium.

Based on analysis of these samples, it seems thatdeer in QC were more nutritionally stressed thandeer in CB. However, it appears that any differ-ences between the mean condition of deer in QCand those in the Eden DWA were minor.

Malnutrition and age-specific fertility in rela-tion to winter severity and deer densityWe analyzed estimates of winter deer density, sex,age, age specific fertility, physical condition, andwinter and spring severity collected by DNR staffbetween 1983 and 1997. They estimated deerdensity from pellet group surveys. DNR staff alsocollect deer carcasses (mostly road-killed) year-round, from which they estimate age from toothwear and eruption. All female deer carcassesfound between February 1 and the middle ofMay (between 100 and 300 each year) arechecked for the presence, number, and sex offetuses. The amount of fat in the femur bone wasestimated visually as an indicator of physicalcondition. Winter severity (based on snowfall andtemperature) and spring severity (based onrainfall and temperature) were calculated forwinters between 1983 and 1995.


The proportion of individuals in the populationwith low levels of femur bone fat was not relatedto winter severity, but to high winter populationdensity. It appears that at moderate to high deerdensities, food competition during winter inNova Scotia can be substantial. There may be alag in the recovery of vegetation after severebrowsing by deer because, during the mid-1990s,a substantial portion of Nova Scotia’s deer herdcontinued to experience malnutrition duringwinter despite low deer densities and mild winterconditions. This also indicates that Nova Scotia’swinter carrying capacity is low.

Although other food sources (such as litterfall)may be significant, it appears likely that – undercurrent conditions – overwintering herds inexcess of 70,000–80,000 deer will probablyexhibit reduced fertility and physical conditionowing to food competition. If overbrowsing orsub-optimal forest management has affected deerhabitat quality in the past, carrying capacity fordeer in Nova Scotia may increase in the future asthe browse recovers, and forest managementbecomes more ecologically aware.

Winter mortalityOther researchers have estimated winter mortal-ity in southern New Brunswick at an average of5.7 per cent between 1988 and 1996. Given theclimatic similarities between southern NewBrunswick and Nova Scotia, it is apparent thatduring most years winter-kill (other than preda-tion) may not be a significant mortality factorthroughout most of Nova Scotia.

Estimates of fawn survival to the age of ninemonths in our study ranged from 40–85 per cent.Coyote predation is a major factor affecting fawnsurvival (section 2.3), and cold, wet springweather can also influence survival during thefirst few weeks of life.

FertilityAge specific fertility data was quite variable,however fertility in both yearlings and adults wasmost closely related to winter density three yearsprior. Therefore, during the increase in deernumbers from 1982 to 1985, the fertility of deerin Nova Scotia decreased annually, and continuedto decline for four years following the populationpeak of 1986. Such a time lag is probably due tothe slow recovery of over-browsed vegetation.

Population goalsMaximum sustained yield for northern deerherds is usually achieved at densities of about 55per cent of summer carrying capacity. However, itis often difficult to convince the public andwildlife managers that sustainable harvestsactually decline for deer populations exceedingthis figure. Summer range quality has a majoreffect on antler development in bucks, thereforeyearling antler beam diameters can be used as areliable indicator of deer population relative tosummer carrying capacity. In general, we suggestthat the herd should be held below 40 per cent ofsummer carrying capacity, the point above whichbreeding in fawns often ceases. This is equivalentto a mean yearling antler beam diameter ofapproximately 16 mm.


Section 2

Aspects of The Ecology of The Eastern Coyote in Nova Scotia

Monitoring coyotesWe captured 51 coyotes – representing 14 differ-ent family groups and 12 transient or dispersingindividuals – and fitted them with radio-collars.Coyotes were classified as breeding residents,resident associates, juveniles (offspring of thatyear), or transients. They were primarily moni-tored from the ground using triangulation,however a helicopter was used when they couldnot be found from the ground. Most radio-collared coyotes were relocated five times a weekfrom December to March, and twice a week fromMay to November. More than 3500 relocationswere collected. These showed that coyotes in bothstudy areas were territorial. We were able toclearly define twenty-four annual territories(Figures 6 and 7), which were significantly largerin QC than in CB.

Group formation and cohesionSociality in coyotes is a strategy that allows themto hunt larger prey while preserving the advan-tages that smaller body size gives in being able tohunt small prey. Pup survival may also be in-creased through group assistance in feeding andrearing.

Until now we did not know whether coyotesform groups as a direct response to the presenceof larger prey, or if larger prey merely strengthencohesion of groups formed for other reasons. Ifgroups are formed primarily to exploit large prey,we should expect to find smaller, less-cohesivegroups using smaller prey as a primary foodsource.

We found that group formation and cohesion wasnot limited to packs that used deer – rather thansnowshoe hare – as their main winter food source.

2.1 Coyote Social Organizationand Spatial Distribution

For most medium to large-sized carnivores,distribution and abundance of food is probablythe single most important factor influencingdistribution and social structure. The socialstructure of the eastern coyote seems to revolvearound resident adult pairs and their offspring.These family groups maintain non-overlappingbut adjoining home ranges of between 30 and 50square kilometres – two to three times larger thantheir western counterparts. Solitary, transientcoyotes may live on large areas encompassingparts of several different coyote territories. Theyalmost always travel alone, and do not breedunless a vacant territory can be found. Winterpacks of coyotes generally consist of adult pairswith young coyotes that have not dispersed, andare typically made up of three or four animals.

White-tailed deer and snowshoe hare are thestaple food items of the eastern coyote. It hasbeen suggested that the larger home range size ofthe eastern coyote is a result of decreased preydiversity and abundance throughout much of thenortheast. However, there is no consensus as tohow the size of areas used by coyotes varies withprey abundance.

In many areas of the northeast, deer migrateseasonally into wintering areas. As a result, somecoyote territories have access to a large number ofdeer, yet others contain few or none during thewinter months. Under such conditions, coyotesmay trespass into neighbouring territories. How-ever, the effects of varying densities of deer andhare on the social and feeding ecology of coyotesin the northeast remains largely unknown.


Figure 6Distribution of coyotes in CB.

Figure 7Distribution of coyotes in QC.

5 0 5 10 Kilometres


EdenSeptember 1994 –March 1997(7) 38

River Denys Mt. Sept. 1995 –March 1997(2) 102

Iona (newly formed)May 1996 – March 1997(1) 130

RoseburnMay 1995 – March 1997(1) 106

ExampleRoseburn: Pack IDMay 1995 – Nov. 1997: Tracking period(3): Number of coyotes radio-tracked108: Number of independent relocations

Skye MountainMay 1995 – March 1997(1) 186

Breeding group of 3-4 present during winter 1997. No indication of presence duringprevious winters.

Marble MountainJune 1995 – March 1997(1) 38

Maple BrookApril 1994 – September 1996(4) 801

5 0 5 10 Kilometres

Kejimkujik National Park

Queens County Study Area

KejiSept 1992 – March 1997(5) 371

PeskoweskNov 1992 – July 1996(4) 110 Devonshire

June 1994 –Dec 1996(2) 290

Grassy LakeNov 1992 –July 1994, Sept 1996 –Nov 1996, (4) 87

NorthfieldDec 1994 –Oct/Nov 1996(3) 49

Mt. MerritOct 1994 – Feb 1996, (4) 173

Mt. MerritMarch 1996 –March 1997(1) 176


This suggests that cooperative foraging (hunting asa pack) is not the main factor driving groupformation and cohesion, but merely an additionalbenefit. We believe that territoriality and groupformation by eastern coyotes improves the survivalof pups and juveniles prior to dispersal. We are notaware of any non-territorial eastern coyotes raisingpups. We also believe the reason solitary coyotes(transients) do not maintain territories is becauseterritories are not necessary for non-breedingcoyotes, not because they are unable to defend aterritory against pairs or groups.

Prey distribution and abundanceWe determined the relative abundance of white-tailed deer within each study area using pelletgroup counts. Hare pellets were also counted. Anaerial survey was conducted in mid-February 1997to provide further information on the relativewinter distribution and abundance of deer in CB,and to define the limits of the Eden DWA. From1995–97, observations from many less formalaerial surveys during January to March alsoprovided information. Ground surveys conductedalong a trail network passing through all spring-fall coyote territories following fresh snowfallsfrom December to March 1996–97 provided dataon the distribution of both deer and coyotes.

Estimates of deer and hare densities were consist-ently higher in CB than in QC. Overall the haredensities were high in CBH, with only a fewscattered pockets of deer being present during thewinter. CBL contained moderate hare densitiesand relatively higher deer densities year round.QC was typified by considerably lower, moreuniform densities of both deer and hare all year .

Except for temporary excursions, coyotes used thesame general territory areas during winter andsummer, and from year to year, although theaverage size of winter areas was larger than that ofsummer areas. Although food is generally mostrestricted for coyotes in the summer, the presence ofimmobile pups at this time probably limits move-ments. Extended movements during winter mayalso result from an increase in the need to defendand mark territories during the breeding season.

Coyote–deer spatial relationsOverall we found that coyotes did not use areasof higher deer density any more than other areaswithin their territories. In fact coyotes in CBL(excluding the Eden DWA) used areas containingfew or no deer more than expected. This may berelated to the higher vulnerability of deer in lowdensity areas (section 2.3)

Territories were generally very stable, and at leasttwo maintained the same approximate bounda-ries throughout the five years of our study. In QCwe noted one confirmed territory shift, whichmay have been related to prey availability. In CBwe did not record any shifts in the area used byentire breeding groups.

Coyotes in CBH made more excursions outsidetheir territories than those in CBL. Most excur-sions took place in the summer in both studyareas, which suggests that prey availability mayhave been a significant factor, as this is the timewhen food is most restricted for eastern coyotes.In CB coyotes generally travelled to areas ofhigher deer density, often trespassing into neigh-bouring territories. Although coyotes in CBH hadaccess to the highest hare densities, they made themost frequent excursions during winter. This maybe related to the need for a higher fat diet duringpregnancy – snowshoe hare is relatively low in fatcompared to white-tailed deer. This theory issupported by the fact that coyotes in CBL andQC, which had access to deer all year round,made relatively few excursions during winter.

Excursions were generally less than 10 km, andrarely lasted more than three days. The shortduration is probably related to the risk of aggres-sive encounters while trespassing, and of losing aterritory to other coyotes while absent. Territorialbehaviour may prevent coyotes from concentrat-ing in DWAs, and may keep the ratio of coyotesto deer relatively low.


Survival, reproduction, and dispersalWe recorded the deaths of 32 radio-collaredcoyotes during this study, the majority betweenOctober and April. Twenty-nine of these were theresult of human activities (14 snared, 10 shot, 3car collisions, and 2 trapped), one coyote waskilled by other coyotes, one died of infection, andone from unknown causes. None died withinKejimkujik National Park.

Packs in CB reared pups successfully in 83 percent of attempts, compared to 78 per cent in QC.Failures were due to human exploitation (killingof breeding females) in CB, and to malnutritionand the old age of a breeding female in QC.

Packs are formed primarily as a result of delayeddispersal of juvenile coyotes. Most juvenilecoyotes in our study had dispersed by the end oftheir first winter. Dispersal distances averaged 53km in CB, and 39.6 km in QC. We observed twoinstances of delayed dispersal in QC and one inCB. In all three cases, the juveniles maintained anassociation with the breeding pair and appearedto assist in pup rearing prior to their departure.

Coyote densitiesEstimates of mid-winter coyote densities changedmarkedly during the course of our study. The sizeof packs were halved in QC from 1993–97,whereas there was a substantial increase indensity in CB during 1996–97 (the only periodwhen estimates were made for that area). Wesuggest that prey abundance and human harvestwere primarily responsible.

In QC coyote numbers continued to declinefollowing the deer population crash of the late1980s and early 1990s. Low prey abundance (deernumbers), mild winters (low deer vulnerability),and increased coyote harvests probably all contrib-uted to the decline. In CB rapidly increasing harenumbers and some reduction in deer harvestingappear to be responsible for the recent increase.

The potential influence of human harvest oncoyote population dynamics remains unclear. Theoverall reproductive success of coyotes is high,and current harvesting levels are unlikely to causesignificant population declines. When breedingadults are killed, they are nearly always replacedvery quickly – probably from remote areas orreserves. We estimate that during the year ofhighest provincial coyote harvest, 1994–95, only25 per cent of the provincial population wasremoved. Intense harvests during years whencoyote numbers are already declining may bemore effective in reducing coyote numbers. Sincecoyote predation on deer is most serious whendeer numbers are already low, localized coyotecontrol efforts at these times may enhance deersurvival in some areas.

2.2 Coyote Movements andActivity Patterns

Most studies of coyote activity have concludedthat coyotes are largely nocturnal, althoughothers report little difference in coyote activitywith respect to time of day. Activity patterns ofeastern coyotes remain poorly described, and theeffects of decreased prey availability and diversitythroughout much of the northeast remain un-known. We tried to determine the daily distancestravelled by coyotes in Nova Scotia, to describecoyote activity patterns in relation to season andreproductive status, and to determine the amountof time spent active and resting.

We monitored coyotes both opportunistically andduring intensive monitoring sessions, relocatingradio-collared animals using standard methods oftriangulation. Their daily activity patterns weredetermined by pooling all observations in each offive time periods: early morning, late morning,afternoon, early evening, and night. We estimatedcoyote movements by summing the total distancetraveled between successive locations. We alsocompared daily distances traveled, and the meanduration of activity and resting periods.


Daily and seasonal activity patternsBased on 1400 records, we found that coyotesspend approximately the same amount of timeresting and active, with several periods of rest andactivity interspersed throughout the day and night.However, they were generally most active at duskand least active during late morning. Coyotes weremore active during summer than winter.

Most previous studies that show coyotes to bemainly nocturnal were conducted in areas ofintensive agriculture or other human develop-ment. We suspect that activity throughout theday is typical of undisturbed coyote populations,particularly in areas providing abundant forestcover, such as Nova Scotia.

Increased activity levels near daybreak and duskmay be related to hunting, as they correspondroughly with the activity patterns shown bysnowshoe hare and white-tailed deer. Yet, coyoteselsewhere hunt for deer near midday duringwinter, possibly because it is easier to catch deerthat are sleeping than to catch those that are active.We agree with other researchers that coyoteactivity patterns are linked to those of their prey.

Daily movementsCoyotes traveled an annual average of 20.2 kmevery 24 hours. Breeding males travelled thegreatest distances of all, during the pup rearingseason, although they were reluctant to foragemore than 5 km from the den. At this time theytraveled significantly further than breedingfemales during the same period, further than allcoyotes during winter, and further than non-breeding coyotes during summer. All coyotestraveled their least distances during winter.

Coyotes travelled shorter daily distances duringthe pair formation and breeding season (winter)than at all other times of year, so mate seekingwas probably not a significant influence onmovements during this study. We suggest thatthese decreased movements during winter wererelated to increased vulnerability of major prey(deer and hare).

2.3 Coyote Food Habits andPredation on White-tailed Deer

There has been widespread public speculationthat predation by the eastern coyote is an impor-tant factor in the prolonged suppression of deernumbers in Nova Scotia. Some studies suggestthat coyote predation does indeed limit deerdensities. Our purpose here was to determine theextent and significance of coyote predation as amortality factor for deer in Nova Scotia. Theterm killing rate refers strictly to the number ofdeer killed; predation rates also take predator andprey densities into account.

Seasonal food habitsWe determined the seasonal food habits of coyotesprimarily from scat (droppings) analysis. Scatswere collected systematically along the trails androadways of each study area during all months ofthe year, and their contents identified as to species.Winter feeding habits were also determined moredirectly through snowtracking. Where possible, ajawbone and femur were collected from anycoyote-killed or scavenged deer discovered, toprovide information on age (from tooth wear anddevelopment), and on physical condition (frombone marrow fat content).

Coyotes consumed at least 35 different preyspecies: 18 wild mammals, 3 reptiles, 1 amphib-ian, 4 birds, domestic livestock, cats, dogs, 6 typesof wild berry, and other vegetation. Diet wasmost restricted during the winter, reflectingseasonal changes in the availability and abun-dance of common food items.

Together, deer and hare comprised the majorityof scat content (66–81 per cent), although smallmammals and fruits were important food itemsfrom late summer to early fall. In QC the use ofdeer was highest between December and May,and lowest during late summer. In CB the use ofdeer was highest during June and July and lowestduring spring and fall.


In QC the use of deer declined and hare in-creased in early 1995. Prey densities changed onlyslightly at this time, but there was much moresnow accumulation during the winters of 1993and 1994 than in 1995–97.

Fawns composed a significant part of the diet ofcoyotes during the summer. Indeed, consumptionof fawns exceeded that of hare in all areas in Juneand July, despite very high densities of hare in CBH.Most fawns were killed rather than scavenged.

Fruit as a potential food bufferIn both study areas, the use of fruits declined ashare and deer densities increased, althoughwildberries remained readily available. Althoughwildberries have a high calorific content, they areonly about 50 per cent as digestible as mamma-lian prey. High fruit use appeared to be associatedwith decreased availability of prey, and probablydid little to buffer predation on deer or hare.

Estimation of predation rates on deerWe estimated total deer losses to coyotes bycalculating the energy requirements of the coyotepopulations in each study area and then convert-ing this into “deer equivalents.” We also deter-mined the proportions of adult deer and fawnsingested by coyotes from scat analysis.

All coyote packs in this study killed deer duringwinter. Winter killing rates averaged between 2.8and 6.2 deer by each pack every 100 days. Theserates appear to be influenced more by winterseverity and the availability of snowshoe hare thanby pack size. However, single coyotes may kill asignificant number of deer (section 2.4), so it maybe inappropriate to estimate winter predation ratesbased solely on deer killed by breeding groups.

Carcasses of 102 deer were consumed by coyotesduring the winter periods. Sixty nine were victimsof predation, the remainder died of other causesand were scavenged. Fawns were over-representedin the sample of coyote-killed deer we examined inQC during 1992-94, and under-represented thereduring 1995-97. Deep snow may have increased

their vulnerability during the winters of 1993 and1994, but we do not know why they were under-represented subsequently. We found no evidencethat deer eight years old or more when killed weredisadvantaged by their age.

In general, bone marrow fat reserves of deerkilled by coyotes appeared to be as good or betterthan those of road-killed deer in each study area.Overall, we conclude that the majority of deerkilled by coyotes were prime-aged, and were notsuffering from severe malnutrition.

Estimates of annual predation rates in QC de-clined from 20.5 per cent in 1992 to 8.1 per centin 1997. In CB they remained at about 10 percent between 1995 and 1997. Predation on fawnsaccounted for a large portion of the total esti-mates during summer, but predation on adultdeer was substantial in QC from 1993–95. Highuse of fawns by coyotes in QC suggests thatsummer predation may have had a substantiallimiting effect on deer populations. We estimatedpredation rates on deer fawns by other predators,such as bobcat, lynx, and black bears at 10 percent in QC and 5 per cent in CB. Figure 8 sum-marizes the birth and death rates for deerthroughout the study.

The low availability of alternate prey, low deerdensities, and unusually severe winter conditionswere probably responsible for the high predationobserved in QC during 1992–93.


Figure 8Estimated mortality factors (expressed as percentages) for white-tailed deer in the Queens County (QC)study area, June 1992 – May 1993, June 1995 – May 1996, and June 1996 – May 1997; and the CapeBreton study area (CB), June 1995 – May 1996, and June 1996 – May 1997.

QC 1992 QC 1995 QC 1996 CB 1995 CB 1996

Registered harvestMales>l yr. old 42.61 25 25 20 20Antlerless deer 251 0 0 0 0

Unregistered harvestMales > I yr. old 15 15 15 5 5Antlerless deer1 5 13 13 7 7

Winter killMales >I yr. old 5 3 3 8 8Females >1 yr. old 3 3 3 3 3Fawns 10 5 5 10 10

Coyote Predation2

Adults 12.7 8.3 5.2 6.1 6.8Fawns 31.6 26.8 12.1 14.6 I5.5

Other Predators3

Adults 7.9 5.7 3 1.5 2Fawns 10 10 10 5 5

Other mortality4

Fawns 5 5 5 5 5Adults 3 3 3 3 3

Recruitment (%)5 I5 22.8 34.8 28.5 30.5

Growth Rate6 .67 .93 1.08 1.04 1.06

Comments Continued to decline until winter 1996, currently increasing. Population increasing at close to predicted rate.

1 Harvest rates for this period were calculated based on regional population estimates and registered harvest (NSDNR, unpublished data). Harvest rates(registered and unregistered) for other periods are based on radio-telemetry data.

2 Estimated rates for adults are based on the average of rates predicted from radio-telemetry and scat analysis. Since limited telemetry data was available forfawns, predation rates on fawns are based on scat analysis.

3 Four of 15 radio-collared deer killed by predators were killed by either bobcats or Iynx so we estimated this predation rate on adult deer as(.267) x (predation rate attributable to coyotes). Predation rate on fawns by predators other than coyotes was estimated crudely from the literature (see text).

4 Includes deaths attributable to vehicle collisions, old age, and natural accidents or sickness.5 Calculated as the total number of fawns surviving to I year divided by the total number of deer alive immediately prior to the birth of this fawn cohort.


Coyote response to changes in prey density, andevidence of prey switchingThe number of prey consumed by a predatorvaries with the density of its prey. We examinedhow the relative densities of deer and hare affectthe number of each prey type consumed percoyote within each territory during winter. Wealso examined the influence of deer and haredensities on deer consumption during June andJuly, when predation on fawns is greatest. As itshould be easier for coyotes to capture hare thandeer, we hypothesized that coyotes would eatmore deer than hare when deer were moreabundant than hare, and switch to hare as theybecame more plentiful.

During winter, the number of deer consumed percoyote increased when the density of deer washigh, and when hare density was low. Thenumber of hare consumed per coyote increasedas hare density increased, but did not alter inrelation to changes in deer density. Although theuse of hare increased in winter and the taking ofdeer decreased significantly as hare densityincreased, coyotes showed a higher preference fordeer than hare at high hare densities and /or lowdeer densities than would be expected from ourhypothesis.

During June and July, the number of hare con-sumed per coyote increased significantly as harebecame more abundant. At this time the numberof fawns consumed per coyote may have decreasedas hare abundance increased, but did not changewith deer density. Our results suggest an overallpreference for fawns over hare in June and July.

The eastern coyote has been described as ageneralist predator, and generalists are expectedto feed non-selectively. In areas where both deerand hare were readily available, coyotes fedpredominantly on hare, and the use of deerdeclined as hare density increased. However, wecould not identify a traditional switch in preyselection in relation to the relative abundance ofeach prey species. Our data support the conclu-sion that – although coyotes should be consid-ered generalist feeders – they prefer to feed ondeer (when available) rather than hare, presum-ably because it is more profitable.

Coyotes in CBH appeared to be satiated on hareduring winter, and hare may have providedsufficient energy for coyotes to pursue larger prey,namely deer. We suspect that during most win-ters, coyotes in QC are forced to focus theirhunting efforts on hare and other small mam-mals, despite low hare densities, due to lowvulnerability of deer. However, when winterconditions are severe, coyotes switch to feedingmainly on deer.

Management implicationsOur study reveals the importance of legal harvestas a tool for managing deer populations. It isimportant to prevent populations from erupting,as the subsequent crashes are nearly impossible toprevent. But continued high harvests following apeak in deer numbers may accelerate and prolongthe subsequent decline in numbers. It is unlikelythat managers can prevent excessive coyotepredation under certain conditions. At such timesmanagers must be ready to impose rapid restric-tions on the hunting of antlerless deer.


2.4 Factors Influencing Killing Ratesof White-tailed Deer by Coyotesin Nova Scotia

Predation is well recognized as a major factoraffecting the dynamics of north American ungu-lates (hoofed mammal species), but has beenlittle studied. We sought to determine whichfactors affect deer killing rates by coyotes in NovaScotia. Deer that had been killed by coyotes werelocated and examined as described in section 2.3.In addition, DNR staff snowtracked coyotesopportunistically during the winters of 1989–94.

Effects of coyote social group size on deerkilling ratesApproximately 38 per cent of the 87 deer chaseswe documented were successful. Groups of threeor more coyotes initiated proportionally morechases than single coyotes. The mean distance ofsuccessful chases (269 m), was not significantlyless than the mean distance of unsuccessfulchases (330 m).

Many authors have reported that groups ofcoyotes are more successful at killing large prey,and that solitary coyotes are more apt to scavengelarge carcasses and kill smaller prey. Contrary toour expectations, the success of chases did notvary with group size, although larger groups weremore reluctant to give up a chase – perhaps dueto a greater expectation of success, or higher foodrequirements for groups as opposed to solitaryanimals. Single coyotes killed at least 16 deer,which suggests that predation by solitary coyotesmay be more significant than previously thought.

Effects of snow depth and density on deerkilling ratesOur prediction that chases would be moresuccessful and shorter when thick, dense snowconditions inhibited deer movements provedcorrect. Kills occurred significantly more oftenwhen the sinking depth of snow was more than30 cm than when snow was shallower.

Effects of deer distribution and abundance onvulnerability to predationWe found that coyotes killed more deer in lowerdensity areas than in higher density areas. It ismore difficult for small numbers of deer to detectpredators and create hard-packed trails to escapepredators. We conclude that individual deer weremore vulnerable to predation where their densi-ties were low (section 2.3).

Effects of forest harvesting on deer vulnerabilityto coyote predationForest harvesting may affect coyote killing rateson deer by removing valuable cover for deer, andmay cause deer to congregate around a predict-able food source. We found that deer kills in bothstudy areas took place closer to clear-cuts thanexpected by chance alone. Relatively deepersnows and less escape cover probably increasedthe vulnerability of deer to predation in theselocations. Kills did not appear to take place closerthan expected to stands with low crown closure.

Prey switching during winterDeer kills increased monthly from Januarythrough March. This increased use of deer aswinter progresses seems to occur despite noperceptible change in alternate prey (snowshoehare) availability. Previous suggestions that preyswitching resulted from increased travel andsociality of coyotes during the breeding seasonare not supported by our data. Daily movementswere actually reduced during winter, and breed-ing pairs are very stable throughout the winter. Infact family groups are generally largest duringearly winter – before prey switching occurs.

We conclude that eastern coyotes do exhibitprey switching, and that switching may beinfluenced by changes in prey diversity, abun-dance, and vulnerability. In areas where deerand hare are the principal prey items, we suggestthat predation on deer may increase sharplywith increased snow depths, or when hareand/or deer numbers decline.


2.5 Winter Condition of Coyotes inRelation to Prey Density

Several researchers have noted that easterncoyotes are typically emaciated during earlysummer, whereas carcasses collected duringwinter generally have moderate fat reserves. Thissuggests that it may be easier for coyotes toobtain adequate food during winter than at othertimes of year. Yet few studies have attempted torelate relative nutritional condition of coyotes tothe abundance of major prey species.

UrinalysisThe analysis of urine collected from snow can beused to compare the nutritional status of free-ranging animal species during winter – researchwith wolves has shown that the ratio of urea tocreatinine in urine samples is generally lower instarved animals than in those that are well fed.

We used urinalysis to compare the relative nutri-tional condition of coyote family groups living interritories containing different densities of white-tailed deer and snowshoe hare. Coyote urinesamples were collected opportunistically whilesnowtracking radio-collared coyotes betweenJanuary and March, 1996 and 1997. We analyzed525 samples for urea and creatinine using spec-trophotometry.

Overall we found that coyotes were in betternutritional condition in CBH than in QC.Coyote packs that fed primarily on snowshoehare during winter were in better condition thanpacks using both deer and hare as major foodsources. We believe this is because coyotes areable to feed more frequently when taking smallerprey, although it may be an artefact introducedby our sampling procedures. Coyote conditiondid not change significantly as winter progressed,but did worsen in January and February (thepeak time of breeding) in both QC and CBL.However, coyotes remained consistently in goodcondition in CBH.

Deer density exerted little influence on coyotecondition, probably because deer vulnerabilityrather than just abundance was more closelyrelated to the ability of coyotes to kill or scavengedeer. Although there was little difference in therate at which groups of two to four coyotes killeddeer, increasing coyote travelling group size mayhave had a positive effect on coyote nutritionalcondition during winter. Larger packs using deeras a major food source may do so more effi-ciently than smaller packs because of less loss toscavengers

Management implicationsAlthough there is no baseline data from coyotesyet, we suggest that urinalysis can be used tomeasure the relative fitness of coyotes occupyingdifferent areas during winter. However, samplingis labour intensive, and the technique will prob-ably remain largely restricted to situations wheresamples can be collected in adjunct with otherresearch activities.


Section 3

Conclusion: Managing Deer, Coyotes,and Forestry in Nova Scotia

3.1 DeerDeer numbers in Nova Scotia have seldomremained stable for more than a few years at atime since the 1930s. We have demonstrated that– under certain conditions – predation, harvest,food competition, and winter severity can se-verely affect the growth of deer populations.Carrying capacity for deer in Nova Scotia duringwinter appears to be lower than in other areas ofthe northeast, and care must be taken to preventanother major decline in deer numbers.

Temporary over-populations of deer that occurredin the past were in no way sustainable undertypical winter conditions. An overabundant deerpopulation may be followed by a prolongeddecline in deer numbers. Recent improvements indeer management in the province – such as zonebased allocation of antlerless permits, which areissued in accordance with the need to reduce thedensity of deer – should allow managers to re-spond quickly and adequately to expected changesin the growth of deer populations. Biologically,there is no longer an excuse for such a serious andlengthy decline in deer numbers as occurredprovince-wide during the last decade.

3.2 CoyotesThere appears to be little we can do to preventcoyote predation from occurring other thansupplying deer with optimal winter habitat toimprove their physical condition and enhance theprobability of avoiding predators. In general, weexpect the effects of predation to be most seriouswhen deer numbers are declining or low.

3.3 ForestryWe believe that the maintenance of quality DWAsis important for several reasons, in particular:1 Deer living in low deer density areas were

more vulnerable to winter predation than deerliving in wintering areas (this was particularlyevident in CB).

2 Wintering areas appear to serve a summer areamore than 10 times their size. The use ofwintering areas is highly traditional, so thesuccess of deer overwintering in any area hasan effect on a much greater surrounding area.Over time, even a slight improvement inwinter survival will result in an increasednumber of deer using the area.

There is little doubt that extensive clear cuttinghas decimated many DWAs. However, guidelinesare now in place that limit harvesting in andaround these areas. In Nova Scotia, no singleclear-cut in any area designated a DWA shouldexceed 10 ha. Quality DWAs enhance deer sur-vival during winter, even though they will have tosupport some forest harvesting when managed.

We believe that DWAs should be managed on alarger scale, so there is always an adequate mix offood and cover within their boundaries, as deerare unlikely to relocate to another wintering area.If deer become numerous enough to damagetheir winter habitat, increased harvests should beapplied throughout the summer range served bythe DWA in question. In general, we believe thatmanaging large wintering areas will be morebeneficial than small, scattered winter pockets.


We suggest that controlled harvesting within areasdesignated as DWAs be allowed and even encour-aged. Our critical recommendation is that no areawithin any DWA be more than one kilometre from astand of at least one square kilometre in size, whichhas more than 70 per cent canopy closure. Excessiveforest harvest can severely limit the availability ofquality cover for deer, and may decrease physicalcondition and increase vulnerability to predation.Current standards outlined in the Forest/WildlifeGuidelines and Standards for Nova Scotia (1989)should meet the needs of deer.

Unfortunately, many DWAs have not been offi-cially identified at present and thus are notmanaged. In addition, the guidelines are onlyenforceable on crown lands. Efforts should bemade to enter partnerships with private land-owners to ensure that DWAs on private lands areadequately managed. The planning of forestoperations in and around DWAs must considerthe needs of deer during severe winters, eventhough these are infrequent.

Successful management will require an integra-tion of habitat and harvest management, and anability to predict when natural conditions arelikely to severely limit population size.

Managers currently have the information andtools required for biologically sound manage-ment of our deer populations. The public shouldbe aware that compliance and cooperation withharvest regulations are necessary for successfulmanagement. Efforts should be made to encour-age both the public and political support neces-sary to execute sound management policies.

A Summary of The Final Report The Ecology and Interactions ... · population was declining while...

Documents

Transcript of A Summary of The Final Report The Ecology and Interactions ... · population was declining while...