The Decline of the Raven, Corvus corax, in Relation to Afforestation ...

The Decline of the Raven, Corvus corax, in Relation to Afforestation in Southern Scotlandand Northern EnglandAuthor(s): M. Marquiss, I. Newton, D. A. RatcliffeReviewed work(s):Source: Journal of Applied Ecology, Vol. 15, No. 1 (Apr., 1978), pp. 129-144Published by: British Ecological SocietyStable URL: http://www.jstor.org/stable/2402925 .Accessed: 24/01/2012 22:15

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

British Ecological Society is collaborating with JSTOR to digitize, preserve and extend access to Journal ofApplied Ecology.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=briteco

http://www.jstor.org/stable/2402925?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

Journal of Applied Ecology (1978), 15, 129-144

THE DECLINE OF THE RAVEN, CORVUS CORAX, IN RELATION TO AFFORESTATION IN SOUTHERN

SCOTLAND AND NORTHERN ENGLAND

BY M. MARQUISS*, I. NEWTON* AND D. A. RATCLIFFEt

* Institute of Terrestrial Ecology, 12 Hope Terrace, Morningside, Edinburgh EH9 2AS and t Nature Conservancy Council, 19 Belgrave Square, London SWI

SUMMARY

(1) The breeding raven population of southern Scotland and Northumberland was formerly (pre-1960) characterized by year-to-year stability in both numbers and distribution. The 123 known pairs occupied traditional nest-sites on cliffs or trees. Breeding density was correlated with altitude (greatest on high ground) and land productivity (greater on base- poor granite than on sedimentary rocks). This was probably linked with a greater availability of sheep carrion on high ground and base-poor ground.

(2) The breeding population has been declining since the 1960s, and in 1974-5 only 55% of former regular nesting areas were still occupied, with breeding pairs reduced to 44%. This was mainly associated with the afforestation of former sheepwalk. The argument is based on a geographical parallel (greatest decline in areas with most afforestation), and on a temporal parallel (desertion of particular nesting areas coincident with planting in the area surround- ing the nest sites).

(3) The level of afforestation at which ravens deserted varied between nesting areas. Probably it depended on the overall quality of the original habitat, and the alternative food sources available: good habitat could take more afforestation before it became untenable than could poor habitat.

(4) Not all desertion of nesting areas could be attributed to afforestation: four raven pairs were dispossessed from cliff nest-sites by golden eagles which recolonized southern Scotland over the period considered, and at least five other pairs by rock climbers. Sheep management also improved over the years, and may have contributed to the desertion of some marginal nesting areas, by leading to a reduction in the amount of carrion available. Organo-chlorine compounds and persecution are unlikely to have been involved in the decline.

(5) In the Lake District, a comparable hill area with practically no afforestation, there was no decline in the numbers of breeding ravens over the same period.

(6) Among the pairs studied in 1974-6, those occupying the most heavily afforested ground showed more non-breeding, and produced later, smaller broods than did those on less afforested ground. Those nesting areas still occupied and where more than 25% of former sheepwalk within 5 km had been afforested, produced later and smaller broods than they did before afforestation.

(7) To judge from pellets, sheep carrion formed a major part of the diet, but a great variety of other foods available from grassland was also eaten. The percentage frequency of sheep remains in pellets declined significantly with increasing afforestation in the area around the nest-sites.

(8) The prediction is made of a further decline in raven breeding numbers if blanket afforestation continues to expand over former upland sheepwalk.

INTRODUCTION

Ravens (Corvus corax) have long been associated with the uplands of southern Scotland and northern England. They are year-round residents, occupying traditional territories

0021-8901/78/0400-0129$02.00 ? 1978 Blackwell Scientific Publications

129

130 Decline of the raven in Britain

based on particular cliffs or trees which provide nest-sites. Their breeding populations were formerly stable in numbers and distribution from year to year (Ratcliffe 1962), but have been declining in the areas concerned, mostly since 1960. The decline coincided with the steadily increasing afforestation of former sheepwalk, and with improvements in sheep husbandry. Sheep carcases are an important source of food for ravens in these areas, as are many other small animals associated with sheepwalk (Ratcliffe 1962). In 1974-6, therefore, an attempt was made to (a) document the extent of decline in breeding ravens, and (b) assess its connection with afforestation and declining sheep stocks.

STUDY AREAS

The Cheviots and the Southern Uplands may conveniently be regarded as a single upland region occupying much of northern Northumberland and southern Scotland. The region is divisible into several sectors, in three of which individual summits exceed 800 m: the Merrick (845 m)-Corserine ranges of Galloway in the west; the Broad Law (841 m) -White Comb (822 m) groups of the Moffat-Tweedsmuir Hills in the centre; and the Cheviot (816 m) in the east. The geology is varied, but mainly sedimentary, with large areas of Ordovician-Silurian greywackes and shales in the Southern Uplands, and of Carboniferous gritstone and shales in the Cheviots. There are localized igneous com- plexes, mainly granites in Galloway and andesite in Cheviot. Except on the granite and gritstone, the soils are relatively base-rich, and the good fertility of these uplands has for centuries sustained large sheep populations. Grouse moors under heather have been maintained at several places, but these usually carry some sheep as well.

After centuries under sheep, these uplands presented an essentially open grassy or heathery landscape, the earlier native forests having been almost totally destroyed. In some places one can still walk several kilometres without seeing a single tree or bush; in others scattered bushes occur on some hillsides and fringes of trees along streams, while in yet other places are small shelter clumps at various stages of growth or decay. Crags are locally plentiful in the western Galloway area, sparser in the central Moffat-Tweedsmuir area, and few in the eastern Cheviot area. Ravens nest for the most part on crags but also in isolated trees, shelter clumps or on the edges of mature forest.

The Forestry Commission planted some of its earliest forests in the study area. At first, only small areas were involved, but after 1945, there was an enormous expansion in the afforestation of unenclosed hill land. After 1960 several private forestry concerns were established which also bought and planted most sheep farms which came onto the market. The principal trees planted were all coniferous, mainly non-native species.

Afforestation entails the removal of sheep stocks, and the fencing of the forest margin from ground still carrying sheep. The ground is ploughed and the tree seedlings planted in the upturned sod. In the absence of sheep, the original short-cropped vegetation soon becomes luxuriant and develops a thick litter. This in turn leads to changes in the fauna and in its accessibility to ravens and other predators. After about 10 years the tree canopy closes, the ground vegetation is shaded out, and the young forest develops into a dense impenetrable thicket; herbaceous vegetation remains chiefly on roadsides and rides and on areas which are too wet, too high or too sterile for trees.

METHODS

Ravens were much sought after by early naturalists and egg collectors, so many of the

M. MARQUISS et al. 131

traditional nesting areas of these birds have been known for decades. A nesting area is here defined as the area containing all the nest sites known to have been associated with particular territories, that is, with particular pairs or their successors. Making use of early information, together with other local knowledge and searching other possible nest sites systematically, D.A.R. covered much of the study area several times in the period 1945-70. Ravens are conspicuous birds; their bulky nests are built in easily-recognized sites and often last for years. In general, the pairs are also regularly distributed (Ratcliffe 1962), so by plotting the positions of known pairs on maps, gaps became apparent which were then checked for extra pairs. Over the years, a knowledge of their distribution was built up, and it was unlikely that, by 1970, any regularly used sites were unknown to us except perhaps for a few tree nests, away from other breeding areas.

All places where breeding had previously been recorded were visited at least twice each season in 1974 and 1975, and details were noted of presence and breeding performance. Pellets were collected from around nest and roost sites and analysed to give an indication of diet. Each pellet was broken up by hand and all parts were separated and identified. The hair and feather remains were identified microscopically using the text of Day (1965), supplemented by local reference material. In 1976, areas occupied in the 2 previous years were again visited to obtain more data on breeding performance and to collect more pellets.

The pattern of afforestation was examined from standard 1:63360 Ordnance Survey maps, which were updated to show all the land under trees, and the dates that particular areas were planted. It was then possible to (a) investigate the use of nesting areas, breeding performance and food of ravens in relation to the extent of local afforestation; and (b) compare the years that ravens were known to have deserted particular areas with the years of planting.

RESULTS Former numbers and distribution

In the past, 123 raven nesting areas were recorded in the study area, with the lowest density in the eastern Cheviot hills and the highest in the western Galloway hills. We used the distance to nearest neighbour as an index of spacing between nesting areas. For pairs which apparently alternated between two or more nest-sites, we took the central point between these sites from which to measure the distance to the nest-site(s) in the next nesting area. Two main trends were apparent. First, nearest neighbour distances were negatively correlated with altitude (Fig. 1); nesting areas were closer together on high hills and further apart on low hills and moorland. Secondly, for any given altitude, nesting areas were closer together in the Galloway and Moniaive hills than in the South Ayrshire or the Moffat-Tweedsmuir hills (Fig. 1). They were even further apart in the Leadhills and further still in Northumberland.

Former stability of population and recent decline From 1946 onwards, thirty-six raven nesting areas in Galloway were examined fre-

quently. Of these, twenty-eight were used every year until desertion, and eight were used only in 1, 2 or 3 years. The twenty-eight regularly used areas were examined in a total of 245 nesting area-years (between 1946 and the final year of use for each nesting area). The nest in any one nesting area-year was not found on only five occasions. On this basis, by


12 I(a) ( b)

8- 0

0 -

o 4- % .

.0 0-

* 0-O~~~~~~~~~~~ pp ~~~00

12 >' | ' | ''-

(C) (d)

a3 8

a) ~~~~~~~~~~~~~~0

4-4 0 _-O a)S

C

0'~~~~~~~~~~~Attd .. . . I . (m)I I

1 2 (e) (f)

0~~~~~~~~~ *4 6

4--

0~~~~~~~

0 0 0 0 0O 4060 00

Altiude a..sS1.

FIG. 1. Nearest-neighbour distances between nesting areas plotted in relation to altitude for six parts of the study area. Each nesting area held one raven pair. Some nesting areas were closer to one another than to any other nesting area, so the same nearest-neighbour distance held for both. The two measurements cannot therefore be considered statistically independent, and in the figure these points are connected by lines. (a) S. Ayrshire, (b) Galloway, (c) Moniaive, (d) Leadhills, (e)

Northumberland, (f) Moffat and Tweedsmuir. -

examining all known nesting areas we should have located 98wr of the regularly nesting raven pairs in the first year, and all of them in the 2 years of the survey.

For the purposes of our survey, it was essential to separate those areas that were regularly used in the past from those that were only occasionally used. By questioning local observers, shepherds and gamekeepers, it was established that at least eighty-one of the 123 nesting areas were in regular use. The rest were either used sporadically (twenty- four) or were of uncertain status (eighteen).

In 1974 and 1975, forty-four of the former regular areas, one of the occasional areas and three of those whose past use was uncertain were still occupied. The situation was the same in both years and no nesting area was deserted between the first year and the second. Thus at least 45% of the former raven nesting areas were no longer occupied in 1974-5.


The situation varied between regions, however, with a 750 decline in the eastern Cheviot area, a 38% decline in the central Tweedsmuir-Moffat Hills, and a 42% decline in the western Galloway Hills. There was no possibility that some ravens merely shifted to breed elsewhere in the study counties, with no overall decline in numbers. As some pairs disappeared however, others sometimes extended their range, and in one instance two cliffs formerly used by two pairs became alternative nest-sites of the same pair; i.e. two nesting areas became one. Nor during 1974-5 were any non-breeding groups of ravens seen. Such birds were rare in the study area even in former years, but occurred as large flocks in other parts of Britain (Ratcliffe 1962) and still do so.

The relationship between the population decline and afforestation Most raven pairs were seen within 3 km of known nest sites, but a few were seen on

sheepwalk up to 5 km away, a distance greater than half the average distance between the nest sites of adjacent pairs in the same area. Afforestation should therefore have had its greatest effects on ravens where it occurred within 3 km of the nest-sites, but might also have had some effect at up to 5 km. We therefore calculated the percentage of former sheepwalk now afforested (up to 1975) within 3 km and 5 km respectively of the nest sites in every nesting area.

Only sixteen nesting areas (13%) out of the total of 123 had no afforestation within 3 km and only one (088%) had none within 5 km. In general, for former regularly used nesting areas the greater the amount of afforestation, the less likely the area was still to be occupied (Fig. 2). As predicted, this tendency was stronger using the figures for afforestation within 3 km (P<0 001) than within 5 km (P=0 05). Some nesting areas continued to be occupied despite extensive afforestation (in the most extreme instance 76% of the former sheepwalk within 3 km, and 63% within 5 km, had been planted), and one totally unafforested nesting area had been deserted.

The only former irregularly used nesting area that was occupied in 1974-5 was unafforested. Similarly, of the areas whose former status was unknown, the three occupied in 1974-5 were on relatively little afforested ground (all less than 20%). Nineteen of the thirty-six unoccupied nesting areas in these two categories had been afforested by more than 20% within 5 km. The probability that occupancy of these areas was unrelated to afforestation was low (P= 0066, Fisher's Exact Test, in Siegel 1956).

For twenty-two nesting areas, it was known exactly in which year occupation ceased. (These were mostly in the last 10 years, though many other areas fell into disuse in the previous 10 years, with no record of the precise date.) Around some nesting areas, sheep were removed the year before planting and around others during the year of planting. For all except one nesting area, the year of desertion was immediately before, immediately after, or during a year of tree planting within 3 km of the nest site(s) (Fig. 3). Around some of these nesting areas tree planting also occurred long before and after cessation of occupancy, but the probability of the coincidence of the two events by chance alone is low (P=0 002).

Breeding performance in relation to afforestation Non-breeding

At eight nesting areas, raven pairs were present but did not nest in at least one of the 3 years of survey. The afforestation figures for these areas were compared with those of eighteen areas where pairs bred in all 3 years. The non-breeding pairs were on areas that


.- e

U)o4 Co0

z 8-

o % sheepwalk C5 0 20 40 60 80 100 afforested

0-r O-_-- A _ - 4 . within 3km

''4 C

z

FIG. 2. Occupation of different nesting areas in relation to extent of afforestation within 3 km and 5 km of the nest-sites.

Year

1945 1950 1955 1960 1965 1970 17

A - B C _ - _

F G -_ H __ 0__ -0 I -

SR _ T

0 M wihn5k

VU -il

CY CL

FIG. 3. Thecdesetion of dforerregula nesting areas in relation toexnto afforestationdae within 3 kman ~~~~~~~~~~~ mof the nest-sites.Lie shwyaso plnig an afildcrlsso steltyar hnte

nesting area w

A

B U V

FG3.Te dsrino omrrglrnsigaesi eaint foetto ae ihn3k oftensFie.Lnsso er fpatnan ildcrlssostels erwe h

G~~~~nsigae a tl nw ob cuid


were on average more heavily afforested (Fig. 4). The difference was more marked using the figures for afforestation within 5 km than those within 3 km. The eight nesting areas were once regularly occupied by breeding ravens, so that inclusion of these with the thirty-seven desertions increases the decline in breeding population of regular pairs from 450 to 56%.

Laying date, clutch size and brood size The date on which the first egg of a clutch was laid was calculated by back-dating from

partially laid clutches and partially hatched broods (incubation period assumed constant at 21 days and a 1 day interval between the laying of eggs). The approximate laying date (to the nearest week) for some clutches was calculated from the ages of small young (less than a week old). Thus the date of first egg in different clutches varied from 19 February to 23 March, with most in the first week of March. Clutch size varied from four to seven (overall average 5 1), and fledged brood size from zero to six (overall average 2 7). The only broods of nil included in this calculation were those which started with eggs and were not destroyed by human agency (see later), i.e. those where failure was attributable to 'natural' causes.

Laying date was inversely correlated with clutch size using exact laying dates (r= -0 70; d.f.= 10; 0 02>P>0 01) and using dates estimated to the nearest week (r= -0 56; d.f. = 15; 0 02 > P > 0 01). The earliest laid clutches were the largest.

Laying date was strongly correlated with the extent of afforestation within 5 km but less so within 3 km (Table 1). Clutch size was not significantly correlated with the extent of afforestation, even when data from all three years were grouped. Brood size was inversely correlated with the extent of afforestation in 1974 and 1976 but not in 1975. All six correlation coefficients for brood size were negative, and in all 3 years the correlation with

a) 4z

Z 0 % sheepwalk

0 20 40 60 80 100 afforested c0, O- *-_ s S | s s ffi Iwithin 3 km

. _

FIG. 4. Incidence of non-breeding in different occupied nesting areas in relation to the extent of afforestation within 3 km (P= 0082) and 5 km (P=0 0017) of the nest sites.


afforestation within 5 km was better than that within 3 km. Summarizing, ravens showed more non-breeding and produced later, smaller broods in areas that were heavily afforested than in areas lightly afforested.

TABLE 1. Correlation coefficients ('r' values) between breeding performance and the extent of afforestation

Breeding performance data Percentage afforestation data (n) Within 3 km Within 5 km

Laying dates 1975 0 254 0 495** 19 1976 0.512* 0 602** 12

Clutch sizes 1975 -0 414 -0 354 12 1974, 5 and 6 (combined) -0 129 -0 184 26

Brood sizes 1974 -0 330 -0.482*** 23 1975 -0 295 -0 300 21 1976 -0714**** -0772**** 18

* 0 10> P>0 05, ** 005> P>0 02, *** 0 02> P>0 01, **** 0-001 > P.

For some nesting areas there were enough data to compare the mean laying dates (to the nearest week) and brood sizes, before and after some of the sheepwalk within 5 km was planted (Fig. 5). At those sites where less than 25% of former sheepwalk was planted there was no significant difference in mean laying date (t =1 03; d.f. = 7; P> 0 10) or mean brood size (t = 089; d.f. = 8; P> 0 10) between the two periods. At those nesting areas where over 25% of the former sheepwalk was planted, the ravens laid eggs later (t = 4 51; d.f. =6; 0005 > P> 0 001, one-tailed test) and reared smaller broods (t= 1b98; d.f. =7; 0-05 > P> 0 025, one-tailed test) than formerly.

Eagles Other factors possibly affecting raven numbers

Golden eagles (Aquila chrysaetos) were probably eliminated from the study area by human persecution, but from 1945 were allowed to recolonize. The appearance of the first

0 10 20 30 40 50 60

%/ of former sheepwalk within 5km afforested FIG. 5. The mean laying dates and brood sizes (young at least 18 days old) before (o) and after (u)

afforestation around particular nesting areas.


pair of eagles in this year was at once followed by non-breeding of the ravens which formerly used the same cliffs. The arrival of a second eagle pair in 1952 coincided with the non-breeding of another raven pair in the same vicinity, closely followed by non-breeding of a third raven pair at a cliff about 1.5 km from the eagles' nesting cliff. Ravens have never bred at these nesting areas since then and were therefore excluded from the analysis comparing occupancy with afforestation.

A third eagle pair settled in about 1960, displacing a fourth raven pair, but these eagles moved to another site 5 km away in 1970. The ravens did not return and the site is included in the above-mentioned analysis. No other eagles were present in 1974-5 at sites previously occupied by ravens, and so eagles alone were not responsible for the overall decline in ravens.

Rock-climbers Five cliffs (where formerly ravens nested) are now regularly used by rock climbers.

Three were known to be once regular breeding sites. Ravens have been seen frequently in recent years at two of these cliffs but no nesting has occurred. None of these nesting areas was included in the analyses of occupancy and breeding performance with afforestation. None of them was afforested by more than 30%.

Egg robbing and persecution Four first and two repeat clutches of eggs were known to have been taken by egg-collec-

tors in the 3 years of the study. Egg robbing was therefore negligible in 1974-6 compared with the large scale robbing that occurred into the 'fifties (about 42% of first clutches in the extensive unpublished data of D.A.R.). In former years ravens were persecuted at times by shepherds and almost always by gamekeepers, who shot or poisoned the adults or destroyed the nest contents, but gamekeepers operated over only small parts of the study area. When land was bought for forestry, destruction of adults ceased. In the present survey there were two cases of poisoning on non-afforested land, and one case of nest destruction.

Other human disturbance One other nesting area was affected by human disturbance. The cliff nest-site was once

regularly used, but not since the building of a small reservoir and pumping house on the river below.

Pollutants Persistent pesticides have been implicated in the decline of some raptor populations,

partly through causing egg-shell thinning (Ratcliffe 1970). In 1975 eggs were-taken from six raven nests in the study area for chemical analysis. The levels of organochlorine compounds were very small (Table 2) and, by analogy with birds-of-prey (Newton & Bogan 1974; Peakall 1976), would have had no influence on breeding performance or population. No significant shell-thinning occurred. This agreed with earlier data on organochlorine levels in four eggs taken from the Southern Uplands in the 1960s, and with the lack of thinning in 205 shells taken from various parts of Britain after the introduction of DDT (Ratcliffe 1970).


TABLE 2. Organochlorine levels in undeveloped eggs from six raven nests in the Southern Uplands, 1975. Results are given as ppm in wet weight (and ppm in lipid)

Nest DDE PCB HEOD

1 0023(28) 1 4(17) 002(02) 2 0-09 (1-3) 1 4 (18) 0 04 (0 6) 3 0 11 (2 7) 0 7 (19) 0 05 (1 2) 4 0.23 (45) 1.2 (27) 006(09) 5 0 29 (4 8) 0 8 (13) 0003 (004) 6 0 16 (1 7) 0 5 (5) 0 06 (0 7) DDE is the main terminal metabolite of DDT, PCB is from industrial polychlorinated biphenyls, and HEOD

is from aldrin and dieldrin.

Food Neither the incidence nor the volume of various remains in raven pellets were directly

proportional to the quantities of various foods ingested, so we could not assess the exact diet from pellets alone. However, pellets indicated quite accurately the qualitative compo- sition of the diet and enabled broad comparison to be made between the diets of different pairs.

A total of 697 pellets were examined from twenty-nine different nesting areas, but ten of these were represented by less than ten pellets (Appendix 1). Food-items were assessed in terms of their percentage frequency in a sample.

Overall, sheep wool occurred more frequently than any other animal remains, and was recorded in pellets from all nesting areas examined. About 3% of the pellets containing sheep wool also contained the rubber rings used to castrate lambs and to remove their tails. Fine particles of vegetable debris sometimes accompanied sheep wool and probably represented the remains of sheep dung. Pellets containing the hair of large mammals usually contained some slivers of bone. Many of the lagomorph bones were also broken bits. Small mammals were frequently recorded and were particularly prevalent in one sample. This pair nested in a recently afforested area where voles were abundant, and was seen on several occasions to rob short-eared owls (Asioflammeus) of voles.

Bird remains were largely feather traces, and the bones of the neck, head, wings and feet of duck, pigeon and grouse. These are the pieces that remain after peregrines (Falco peregrinus) have killed and eaten their prey. The percentage frequency of bird remains in the pellets of raven pairs was negatively correlated with their distance from the nearest peregrine pair (r = - 0496; d.f. = 22; 0 05 > P> 002). Identifiable fragments of eggshell were chiefly from curlew (Numenius arquata) and red grouse (Lagopus lagopus scoticus) eggs. The chitinous exoskeleton of insects and arachnids was most often present as small particles; where larger particles were present, these were usually the legs or elytra of beetles. Pellets containing chitin fragments also contained large amounts of vegetable material. The latter appeared in nearly all pellets but it is unlikely to have contributed much to nutrition because it was cast up largely intact. The same was true of cereal grain, which was recovered from pellets containing pigeon or duck remains and so was probably ingested incidentally with the digestive tracts (commonly discarded by peregrines). Articles of refuse were scarce but occurred in five samples. Raven feathers were probably swallowed during preening. Raven eggshell in castings could be the result of either eating the shells after hatching or eating the eggs themselves. On several occasions, single eggs disappeared from raven nests during incubation and these may well have been eaten by the adult.


The percentage frequency of sheep remains in pellets was negatively correlated with the percentage of afforestation within 3 km (r=-0616; d.f.=17; 001>P>0001) and within 5 km (r= -0 668; d.f. = 17; 0 01 > P > 0 001) (Fig. 6). Where afforestation levels were high, the castings contained remains from a variety of foods, but largely goat, deer, lagomorph and bird carrion and, in recently planted areas, voles. Both high and low levels of chitin and eggshell were found in samples from areas of low afforestation, whereas only low levels were found in samples from highly afforested areas.

100-

0 80- _ 0~~~~~~~~~~~~~ 3 00

CL

4) 60 * 0

- 0

40-

o U - 20- 0 *0

0 10 20 30 40 50 60 % sheepwalk afforested within 5 km

FIG. 6. Relationship between frequency of sheep wool in pellets and the level of afforestation within 5 km of the nest-sites of particular raven pairs.

The general picture of raven feeding habits gained from our analysis is similar to that gained from smaller samples of pellets given by Bolam (1913) for Wales, Ratcliffe (1962) for the English Lake District, and Harlow et al. (1975) for Virginia, United States of America. In all the British areas, sheep formed a staple food, a finding borne out by observations of sheep/raven interactions made by M. A. Ridpath (in Murton 1971).

DISCUSSION Whilst raven pairs tended to be most numerous in the most rugged parts of the study area, their former density could not be explained solely by the availability of nest sites. In general, crags were more abundant on higher hills and in areas of igneous rock. However, in many districts ravens breed readily in trees and tree sites were abundant throughout most of the study area, particularly at lower elevations where ravens were sparser. Also in Northumberland (a low density area) some pairs used, as alternative nesting places, crags several kilometres apart, but in other parts of the study area (with high density), separate pairs nested within this distance of one another. In a few parts of the study area, the lack of breeding ravens was associated with the lack of any suitable nest site, but this factor had much more influence on presence/absence than on nearest neighbour distances.

The most reasonable explanation of the dispersion and density of breeding ravens in the study area is one based mainly on food supply. There has long been a government subsidy on hill sheep, tending to maximize their numbers, often above the carrying capacity of the grazing. The grazing is poorest at high altitudes and in areas overlying acidic rock, so under these conditions one would expect the highest sheep mortality. This may explain the


higher densities of breeding ravens at higher altitudes, and in middle Galloway, the only region where the underlying rock is largely granite.

Other factors would tend to accentuate any differences in the availability of sheep carrion between areas. Thus some of the best hill ground in the study area was managed primarily for red grouse (e.g. in the Leadhills and parts of Northumberland) and supported far fewer sheep than other ground managed for sheep alone. Secondly, the most acidic substrate (granite) within the study area was associated with a more precipi- tous terrain, and some sheep were killed accidentally when they became 'cliff-bound' whilst foraging on the relatively luxuriant vegetation of cliff ledges. (On one occasion D.A.R. found as many as nine dead sheep below a Welsh precipice.) Thirdly, the ground at lower altitudes was usually less remote, so sheep husbandry was possibly better and the sheep that died were more likely to be buried by the shepherd than those on more distant ground.

Most bird species that have been studied show a positive relationship with land productivity, with the densest populations on the richest ground (Jenkins, Watson & Miller 1967 for red grouse; Ratcliffe 1969 for peregrine; von Haartman 1971 for various song-birds; Newton et al. 1977 for sparrowhawk Accipiter nisus). In these species, density probably adjusts not to land-productivity as such but to available food-supply, the two often being closely correlated. In contrast, the raven is a species whose food (through human intervention) improves with deterioration in land productivity. Similarly Lockie & Stephen (1959) thought that golden eagle populations in some impoverished areas were also maintained largely by the abundance of sheep carrion.

The former occupancy of raven nesting areas may also have been related partly to food-supply, in that many irregularly occupied areas were on relatively low ground, or on grouse moor where carrion was probably scarcer. However, another factor may have been the poor quality (in terms of height and remoteness) of many of the low ground cliffs used for nesting. As a parallel, the frequency with which particular peregrine territories were occupied in eastern North America was directly related to the height and remoteness of the nesting cliff; the reduced use of 'poor quality' cliffs apparently arose from the reduced protection they offered against human and other predators (Hickey 1942; Hagar 1969). Hence, in an era when ravens were heavily persecuted, they may have been generally reluctant to use the less safe and more disturbed sites. We suggest this as a possible factor influencing the former occupancy of nesting areas and not their dispersion; including irregularly used ones, nesting areas were as evenly spaced on low ground as on high ground.

A small number of raven pairs have been dispossessed by eagles and rock climbers, but the larger part of the recent decline in numbers was probably linked with the afforestation of sheepwalk. In 1974-5 the occupancy of nesting areas was associated with the extent of afforestation chiefly within 3 km of the nest sites. Ravens are ground foragers, feeding mainly on large items of carrion but also eating any animal material down to the size of a beetle or snail. Items smaller than 100 g are probably often taken alive. Afforested ground had no sheep carrion and in addition the high vegetation obscured other carrion and made it more difficult for ravens to forage for eggs, ground insects, frogs, lizards and small mammals. Those pairs still present in 1974-6 despite heavy afforestation were feeding on goat, deer and lagomorph carrion or bird-remains, the latter apparently scavenged from peregrines. On some recently afforested ground there was an abundance of voles, and these provided a temporary sustenance for ravens, sometimes for 2 or 3 years after sheep had been removed.


The stage at which nesting areas were deserted probably depended on the balance between (a) the original quality of the area for sheep carrion, (b) the amount of afforestation and (c) the amount of alternative food. If sheep carrion was scarce, due to low sheep stocks, high quality grazing or good sheep husbandry, even small amounts of afforestation may have removed a critical amount of food. This probably explained the lack of a very consistent association between afforestation and occupancy by ravens. In addition, those areas of sheepwalk which had the highest sheep mortality were those most likely to be sold for forestry. The greater the amount of existing forest, the more likely was a nesting area to be deserted on subsequent afforestation. Some nesting areas were deserted even the year before tree planting, but coincident with the removal of sheep. Mostly, however, desertion was coincident with planting.

Breeding performance was associated chiefly with the extent of afforestation within 5 km of the nest sites, whereas occupancy was most closely associated with afforestation within 3 km. Ravens probably used a large area for their general foraging, even though the tenancy of a nesting area seemed to have been mainly dependent on the amount of food immediately around the nest-sites. Ravens bred early in the year, probably to coincide with the period of major sheep mortality at the end of the winter and during lambing in early spring. Feral goats also had their kids in early spring. Ravens on more heavily afforested ground bred later. This could have been a response either to a generally poorer food supply or to an alternative food supply: one late breeding pair fed largely on lagomorph carrion which was most abundant later in the spring than sheep and lamb carrion.

Another factor complicated the pattern of raven decline, namely the general improve- ment in sheep husbandry. Over the years, some areas were drained, fertilized and reseeded to improve the forage for sheep, the provision of supplementary winter feed became more prevalent and drugs were increasingly used to prevent disease. The net result is likely to have been less carrion. The extent of this sort of food depletion for ravens was impossible to determine within the limited scope of our study, but seven or eight unoccupied nesting areas were still less than 20% afforested in 1974-6, and may have been rendered untenable through lowered sheep mortality.

Immediately to the south of the Southern Uplands and Cheviots, the English Lake District represents a control area for this study, because its hills are still largely sheepwalk. There has been little afforestation, and this mainly around the edges. The Lake district has held a stable raven breeding population of about sixty-five to seventy pairs since 1900. During 1974-6, forty-eight known nesting areas in this region were examined: of these, forty-six held breeding pairs, although six of these nesting areas were formerly regarded as irregular in occupation. One nesting area was occupied in 1974 but deserted in 1975 and 1976. This nesting area and one of the other two deserted ones were in localities where there had been substantial recent afforestation within 3 km of the nest sites. The third nesting area was deserted for unknown reasons. While the raven can probably survive at low density in afforested uplands, this will be likely to depend on the extent of sheep ground left unplanted. Almost certainly the species will decline further if blanket afforestation continues to expand over former upland sheepwalk.

- ACKNOWLEDGMENTS

For permission to work on their ground, for placing every facility at our disposal and for


full co-operation and help in other ways, it is a pleasure to thank the staff of the Forestry Commission, South Scotland Conservancy. Mr S. Petty of Forestry Commission, Northeast England Conservancy, kindly provided us with up-to-date planting maps for Northumberland. For historical information on raven distribution, we are grateful to the late E. Blezard, the widow of the late Captain R. Cross, to J. Hutchinson, W. R. Laidler, B. Little, E. Meek, W. Murdock, the late W. Murray, R. Nelson, D. Watson, R. Roxburgh, G. Shaw, the late G. W. Temperley and T. Todd. For helping with the survey in 1974-6, we are especially grateful to B. Little, E. Meek and other members of the Northumbria Ringing Group, G. Carse, G. Horne, R. Mearns, the late W. Murray, and R. Roxburgh. We are also grateful to various shepherds, forest rangers, gamekeepers and other local observers, too numerous to mention individually. For drawing the figures we thank Mrs S. Adair, for statistical help P. Rothery, and for helpful criticism of the manuscript Dr D. Jenkins and Miss J. Rowe.

REFERENCES

Bolam, G. (1913). Wildlife in Wales. London. Day, M. G. (1965). Identification of hair and feather remains in the gut and faeces of stoats and weasels. Journal

of Zoology, 147, 201-1 7. Hagar, J. A. (1969). History of the Massachusetts Peregrine Falcon population. 1935-57. Pp. 123-31. In: Pere-

grine Falcon Populations, their Biology and Decline (Ed. by J. J. Hickey). Univ. of Wisconsin Press, Madison.

Harlow, R. F., Hooper, R. G., Chamberlain, D. R. & Crawford, H. S. (1975). Some winter and nesting season foods of the common raven in Virginia. Auk, 92, 298-306.

Haartman, L. von (1971). Population dynamics. Avian Biology, Vol. 1 (Ed. by D. S. Farner and J. R. King) Pp. 391-459. Academic Press, London and New York.

Hickey, J. J. (1942). Eastern population of the Duck Hawk. Auk, 59, 176-204. Jenkins, D., Watson, A. & Miller, G. R. (1967). Population fluctuations in the red grouse (Lagopus 1. scoticus).

Journal of Animal Ecology, 36, 97-122. Lockie, J. D. & Stephen, D. (1959). Eagles, lambs and land management on Lewis. Journal of Animal Ecology,

28, 43-50. Murton, R. K. (1971). Man and Birds. Collins, London. Newton, I. & Bogan, J. (1974). Organochlorine residues, eggshell thinning and hatching success in British

sparrowhawks. Nature, London, 249, 582-3. Newton, I., Marquiss, M., Weir, D. N. & Moss, D. (1977). Spacing of sparrowhawk nesting territories. Journal of

Animal Ecology, 46, 425-41. Peakall, D. B. (1976). The peregrine falcon (Falco peregrinus) and pesticides. Canadian Field-Naturalist, 90,

301-7. Ratcliffe, D. A. (1962). Breeding density in the peregrine Falco peregrinus and raven Corvus corax. Ibis, 104,

13-39. Ratcliffe, D. A. (1970). Changes attributed to pesticides in egg breakage frequency and eggshell thickness in

some British birds. Journal of Applied Ecology, 7, 67-115. Siegel, S. (1956). Non-parametric Statistics for the Behavioural Sciences. McGraw-Hill, New York.

(Received 25 August 1977)

M. MARQuiss et al. 143

9' P1~ > C C) 9n 9 >C-C-I-C> ~~~~~~~~~~~~~~~~~~~~~~00 CAC C-C

00~ ~~~0

10

, 0 9 9 0

4.0 - ~~~~~~c7NC -- Cl 00-0 11.0 0~~~~~6~1 ..)O ~ b C) 1. 0 6

-- 009 99 99 9 99

00 ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~ ~~~~

Cl C - Cr- - s-l0 CA 7-4~~~~~~~~~~~~~~~~~~~~~~~~~~~~4

lo 0 0 C4 0CI - 4 IC 0C 4C C

6.) ~~~~~C-IC-In

0-.oM 00 C--C4 D 0(D en r 0CA

.fl 0~~~~~~~~~~~0 0 D 40 I)C) <

00 C 6 c oo en t00 r 0.

~~~~~~~~ 0 0.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t o0

'WA~

C r.4C -~-CI~ C

1-0 ~ -

~~~~~ 0 4.0~~~~~~~~~~~~~~~tg 140 '~ 4CCZ

o4 O.0.9~f4 ~.C 4.f4 ~ 404 9.0 4 0,~~~~~~~~~~~~~~~~~~~~~~~~~~ 6 -0 co U


00 tn o I I I I I I _ I I I I

0 O-

m1 ^ t ? ? ? _ ^ - ?? ^ _~~7171 7

09 _N 1. N do -, do _

N 0 0 - _ F _ s_ t _ 8 _ N~~~~~00 0 0

? ~ ~ ~ ~ ~ ~ ~ ~~~r 00 (14 CD r- 't V )

0 0 - N Nt_

00 I M n I IN

o Nt^

_ O N s ~~~~~~ ~~ _ ~~~ _ N s o oob o r C

? CI T|| cI

Cn 17 _ 00N?

oo oM t oo o

m W I _ I I I I _ I _ I I I I I I N I I N N I I I t- xCX (N I mP I -~~~~~~~~~~~C

i mP ? _ _ t _ ~~~~~~~~~~~o 0

mI ceX oo - ?? b l- F (s] F t 0? t N F FU -~~~~~~~~~~~~~~, ' , O+

xs- -s -0 Cd Cd-- s ss -

O U4 zs w .; C4C C

0 r;Cd 3QtBi& ;3rS 3 X

The Decline of the Raven, Corvus corax, in Relation to Afforestation ...

Documents

Transcript of The Decline of the Raven, Corvus corax, in Relation to Afforestation ...