Some results from trapping rooks

Transcription

1 Ringing & Migration ISSN: (Print) (Online) Journal homepage: Some results from trapping rooks Paul T. Green To cite this article: Paul T. Green (1981) Some results from trapping rooks, Ringing & Migration, 3:4, , DOI: / To link to this article: Published online: 11 Apr Submit your article to this journal Article views: 155 Citing articles: 6 View citing articles Full Terms & Conditions of access and use can be found at

2 Some Results from Trapping Rooks 203 by Paul T. Green INTRODUCTION The catching of crows, Corvus spp., can be a time consuming and expensive process (Rowley 1968), and during a study of the social behaviour of the Rook Corvus frugilegus ringing and wing-tagging of large numbers of this species was necessary. Several trap designs were tried and a cheap, effective one developed which could be of use to ringers. This paper investigates the effects of seasonal factors and design of the trap on trapping success. Over 500 Rooks were caught over two years, and the results of morphometric analysis are given and discussed. All trapping was done within 100 m of a rookery situated 20 miles south of Edinburgh in the Borders Region of southern Scotland, 55 44'N., 2 54'W. SEASONALITY AND TRAPPING Methods. Feare, Dunnet and Patterson (1974) found that in NE Scotland Rooks could be caught only during spring and again in early summer. In this study, trapping was attempted with automatic cage traps (see Appendix) between January and August 1979, and again from November 1979 to July 1980, using bread and oats as bait, although other items were tried (see below). Results. Rooks could be caught only from mid-march, with a peak during April falling throughout May to zero in July (Fig. 1). Jackdaws Corvus monedulan could be caught from June onwards with a peak during July and August. Considered in relation to meteorological data, no clear trend emerges, with differences between the two years (Fig. 1). In 1979, trapping success was good when the minimum temperature dropped below zero and precipitation (some of which was snow) was high. In 1980, trapping was good when the minimum temperature remained above zero, but precipitation was low. Discussion. Trapping success may have reflected the availability of the Rooks' natural food. During the spring and early summer, Rooks feed mainly on earthworms, which peak in availability in April (Heppleston 1971), and leather jackets which reach peak biomass in June (Dunnet 1955). Earthworms may become unavailable in spring during periods of cold weather when they burrow deeply, and again in dry periods of the summer (Gerard 1967), the latter coinciding with emergence of leather jackets as winged adults, and thus no longer available as grubs obtained by probing. These events could all lead to potential food shortages and could arguably cause Rooks to enter traps. It is possible that successful trapping in 1979 was caused by inactivity of invertebrates during the early morning after a freezing night, which could in theory be tested by collecting samples from collared nestlings during different times of the day. However, whereas the cold-induced food shortage, which coincides with a high food demand of the growing young, resulted in successful trapping, that resulting from dry, warm weather did not. This latter dryness coincides with an increase in food demand due to the onset of moult (Feare et al 1974). The latter finding is surprising as there is evidence of a food shortage; lack of rain and higher temperatures resulted in a predictable gathering of Rooks into large flocks on a few damp fields (Green in prep). The common factor for both years was that Rooks were caught during the breeding season, when food demand is high. [Ringing and Migration 3: , 1981) (Received 4 November 1980, accepted 24 June 1981)

3 204 RINGING AND MIGRATION Trapping during the moult, which lasts from late May until September (Seel 1976), was unsuccessful. This is in spite of a food shortage, caused by aestivation of earthworms, emergence of leather jackets and absence of grain, coupled with measures of poor body condition (Beshir 1970; Feare et al 1974). This can possibly be explained by food quality. Amino acid requirements for feather growth are very specific, and the requisite ones are acquired from thinly dispersed invertebrate prey, and not from the grain put down as bait in the traps (Newton 1968). Acquisition of the amino acid cystine is important during moult; it is about the fifth most common amino acid of feathers, but one of the four least common in most other animal proteins. Thus much protein has to be gathered in order to concentrate the amounts required for new feathers (Newton 1968). Adult mortality during the summer is not high (Holyoak 1970) and it would appear that Rooks are not at risk from starvation at this period. As cystine is fifteen times more common in animal-derived food than in grains such as oats and barley (MAFF 1974) it is probable that Rooks are more selective in their food intake and therefore not tempted into the traps after grain during moult. This trend was anticipated, and mealworms were offered as an alternative bait. However, Rooks -.-.-X r ntaan daily rainlalt for 7 day parted = no trapping * * * = weak with complata enow cow i i I r* *. 15 waafc no. A. M Figure 1. Trapping success in relation to season and precipitation. is

4 TRAPPING ROOKS 205 bill 2 Figure 2. Bill measures of Rooks. appeared not to recognize these as food, and the mealworms were not taken until the end of the study. Other high protein food was tried, such as eggs and raw meat, but this met with little success. Trapping was successful during the breeding season, with traps located near the rookeries. Post-breeding trapping may have been successful if the traps had been located on the upland pasture, where the Rooks feed in the summer months before the grain harvest. MORPHOMETRICS AND SEXUAL DIMORPHISM Methods. All Rooks were weighed, and wing-length (maximum length), three bill measures (Fig. 2), and tarsus length were measured. Adults were sexed on the basis of wing and bill length using discriminant function analysis (Green in press). Of the 288 adults used in Table 1,156 were sexed by behaviour at the nest (only the female incubates, and is fed on the nest by the male, Coombs 1960), and 132 were sexed to 99.9% probability using discriminant analysis of measurements. Results. The frequency distributions for wing and bill measures for each sex are given in Fig. 3 and 4. The differences between the sexes are significant for both the wing and the bill measure (bill, t = 20.3, d.f. =285, P <0.001; wing, t = 22.2, d.f.= 287, P <0.001). The mean sex difference in size for all skeletal measures taken is 8.3% with an overlap of the ranges of 51 % for the long bill measure, and 64% for the wing measure. The degree of sex difference in ' size is similar to that found in the Carrion Crow, for which Holyoak (1970) showed the difference to be 8.05% and Picozzi (1975) 7.6%, expressed as the mean sexual difference using various measures (Table II). TABLE I. MORPHOMETRICS OF ADULT ROOKS IN THE STUDY POPULATION. sex M F M F M FM F M F M F parameter weight weight bill 1 bill 1 bill 2 bill 2 bill 3 bill 3 wing wing tarsus tarsus mean S.D range : Notes: a) weights in grams, other measurements in millimetres, b) see Figure 2 for definition of bill measures. N

5 206 RINGING AND MIGRATION J...J,, 310 wing langth nun. Figure 3. Frequency histogram of maximum wing length of adult Rooks. Discussion, Many factors exist which may have caused this difference in size between the sexes. Sexual dimorphism results from the different selection pressures acting upon each sex, and their relative sizes will depend on the sum of all pressures acting both to increase and to decrease size (Rails 1976). Various aspects of competition may operate to cause these size differences. Increase in male size may result from male-male competition for females, selection will favour those traits which aid in winning fights (intra-sexual selection) and hence females. Thus males increase in size over a number of generations, and the prominence of those features used in display also increases. This is supported by evidence derived from looking across a spectrum of species, for male rivalry is more important in those species with polygamous mating systems. These show much greater sex dimorphism than do monogamous ones due to more intense male-male competition (Brown 1975) O- r 20- r 1 J i 10- L r--l" i 0 r '.'0 ^. "1 7 max. bill langth mm. Figure 4. Frequency histogram of bill length (bill 2 of Fig. 2) of adult Rooks.

6 TRAPPING ROOKS 207 TABLE II. SIZE DIFFERENCES BETWEEN SEXES IN ADULT R00K5> AND CARRI CROWS. species derivation Rook mean all " mated pairs " mean all " mated pairs " mean all " mated pairs " Carrion mean all mean all Crow mated pairs " " mean all mated pairs " mean all " mated pairs " mean all " mean all parameter bill length bill length bill depth bill depth wing length wing length tarsus bill length bill length bill depth bill depth wing wing bill length wing length % difference N NOTE: mean length of males was bigger than that of females in all cases. authority Green this study Hoiyoak (1972) Picozzi (1975) In the Carrion Crow, bill size is a good predictor of success in all types of competitive encounters. Charles (1972) found astrong positive correlation (>0.9) between all aspects of bill size and success in encounters between males. This is important when defending a territory, for only territorial Carrion Crows can breed, with the male most active in maintaining the territory. In times of critical food shortage, success at encounters is important again; Houston (1977) and Picozzi (1975) have noted a predictable male bias in winter flocks possibly as a result of this. So a large bill, or a factor accompanying it, is advantageous to the Carrion Crow. My own data suggest that this is true for Rooks also, for in twelve of fifteen encounters observed between marked birds over a food item on grassland during spring, the bird with the larger bill was successful (Green, in prep.). Being a colonial breeder, the pay off from encounters at the nest are not so large for a Rook. Defeat of a Carrion Crow may mean inability to breed in that year, but this is not so for a Rook, for it can build elsewhere in the colony and breed successfully. So pressures for increase in size are not so large for a Rook as they are for a Carrion Crow. This general increase in body size will also be accompanied by an increase in the size of the feeding apparatus, which may in turn affect the size of food items the sexes take. Differences in food size taken by each sex have been variously interpreted as being a side effect of a sex-specific change in size for some other reason, or as a reason in itself. Given the many other means of reducing inter-sexual feeding competition which have been demonstrated, e.g. various aspects of niche displacement and allopatry outside the breeding season, then it is possible that selection has caused divergence in size as a result of the advantages accrued from feeding niche differentiation (Selander, 1966,1972). Only in cases where bill si2e alone is dimorphic is the situation clear, as is so in some species of woodpecker and the extinct Huia, Heteralocha acutirostris. In those species, such as the Rook, where the feeding apparatus is dimorphic as a result of body size differences it is a matter of conjecture as to whether niche displacement is the primary advantage or merely an incidental one. Whereas crows of the genus Corvus are sexually monomorphic for plumage characteristics (except for the Jackdaw, Voipio 1968), in all aspects of body size there is a

7 208 RINGING AND MIGRATION difference of around 8%, with males the larger sex. The effect of this upon the size of food items has been shown by Holyoak (1970) for the Carrion Crow. Invertebrates fed to the young by the male averaged 63% longer than those given by the female, where the difference in bill size is around 10%. The sexes also differed in the proportions of different feeding actions used. This bill size difference may be of importance in reducing competition between the sexes for food, but it may, on the other hand, be only a by-product of another pressure which gave rise to sex dimorphism in overall size. Bioenergetics may be another important factor when considering reasons for sexual dimorphism. The mass of an animal influences the rate at which it uses energy, so that such measures as resting metabolism, energetic cost of flight, egg weight, heart rate and conductance decrease exponentially as body mass increases, whereas storage capacity, digestive ability and lipid reserves decrease linearly (Downhower, 1976). These relationships will affect the ability of a bird to respond to conditions of deprivation or surfeit. A large bird can survive for a longer period on its reserves than a small one, but also takes longer to replenish them. Such relationships affect birds differently depending on their life history pattern. Thus for a bird resident in its breeding range which has a greater reproductive output the earlier in the season it starts to breed, selection should favour small females who can come into reproductive condition rapidly with the spring flush of food (e.g. Great Tit, Parus major Jones 1973).iLarger size in females would be selected for in those species where eggs are produced from reserves laid down before migration to the breeding grounds, and here females are generally the larger sex (e.g. Dunlin Calidris alpina, Downhower 1976). Other factors may override this; for instance in the leking Ruff Philomachus pugnax (Jehl 1970) sexual selection appears to have led to males being the larger sex in a species which migrates to its breeding grounds. In a population of the Hooded Crow which is resident, however, larger females breed first (Loman 1980), so the problem is unresolved. Despite the overlap in ranges between the sexes in the Rook in all aspects of size, in all' mated pairs the male was always larger than his female, which has led some workers to suggest that mates are seleaed on the basis of size (Holyoak 1970). From my data on sizes of mated pairs of Rooks, the size dimorphism between mated pairs does not differ significantly from that predicted by taking males and females at random from the adult population t = , d.f. = 140, P >0.2). Although Table II suggests that mated pairs of Carrion Crows show greater size dimorphism than that of the population means, it is possible that this difference would disappear if the data were analysed in the same way as above for Rooks. CHANGES IN MEASUREMENTS OVER TIME Retrapping of marked birds enabled an assessment of changes over time during the March-June period in several aspects of size of birds. A total of 24 birds were caught on more than one occasion (mean 3.7 times, S.D. 1.8, range 2-10), all but three being males. Results. To reveal whether a trend of increase or decrease in certain measures occurred over the breeding season, the sign test (Siegel 1956) was applied. Measures of tarsi and of wing-length showed no significant trend. Weight changes have been discussed by Lockie (1954) and at length by Beshir (1970). My results do not conform to the trend of decrease which these authors show during this period. Beshir obtained his samples by shooting, my birds were taken from a trap, well fed with bait, perhaps obscuring any apparent trends, and perhaps selecting a non-random sample. However, both measures of bill-length showed a significant decrease over time (Sign Test, P <0.001) as revealed in Fig. 5.

8 TRAPPING ROOKS 209 f J Figure 5. Change in bill length (bill 2 of Fig. 2) of adult male Rooks during the breeding season. Each line joins the measurements for a retrapped bird. Discussion. The only two birds to show an increase in bill length were two females, known to have bred successfully, and who were incubating and brooding during the intertrap period. The decrease in length occurred in the distal part of the bill, as the short bill measure ('Bill 1' of Fig. 2) also decreased. The distal portion of the top mandible is of cartilage which appears to be continually growing, and the many reports of malformed Rook bills in the literature (e.g. King and Rolls 1968) and from my own observations, appear to result from the crossing over or overgrowth of the top mandible. Wear is therefore prevented, and unchecked growth follows. I suggest that the decrease in bill length at this time occurs because wear on the bill tip excedes growth. This is supported by the fact that Rooks probe for earthworms and leather jackets (Lockie 1954), often digging with a hammering action, at the time of year that the measures were taken. Large amounts of food are collected by the males for themselves, their mate and their growing young, and so the bill is subject to a large amount of wear. After the breeding period Rooks spend more of their feeding time on field types where probing is far less frequent, for instance stubble fields, livestock troughs etc. (Feare el al 1974), where presumably wear is much less. In zoological collections, where Rooks and other corvids are unable to perform such bill wearing activities, the incidence of overgrowth of the top mandible is very common (pers. obs.). A more extreme case of feeding technique shortening the bill has been shown by Heppleston (1970) for the Oystercatcher Haematopus ostralegus. ACKNOWLEDGEMENTS I extend my thanks to Kathy Bogan for help in the field, and to John Weir of Cortleferry for giving free access to his land. Anthony Collins, John Deag, Richard Waite and an anonymous referee helped to improve the script, and the Science Research Council funded the study of which this is a part, and I thank them also.

9 210 RINGING AND MIGRATION SUMMARY Using automatic cage traps near a rookery in S.E. Scotland, Rooks could be caught between March and June. This is discussed in terms of the food requirements of the birds at different times of the year, coupled with its changing availability and those factors causing these changes. Measures taken of the adult birds showed that males average 8.3% larger than females, but with considerable overlap in all measures. The significance of the sex difference is dicussed in relation to reduction in competition within a pair, inter-pair encounters and energetics. During the breeding season, the bill length of males decreases, probably as a result of wear caused by a greater than normal amount of probing. APPENDIX Trap Types. All traps were basically cages with ground entrance tunnels as described by Hollom and Brownlow (1955). In 1978 a cylindrical trap of 1.25m height and 1.5m diameter was tried, built to the design of Allsager et al (1972). It was made from 20 mm x 40 mm rigid mesh as it was required to be portable. This failed to catch any birds, and observation revealed Rooks and Jackdaws could leave the trap through the tunnel. Addition of a trap door sprung by the entering bird later achieved a few catches of individual birds. Three traps were built in Trap 1 was a portable rectangular box 2mxl.5m and 1.5 m tall, made of the same rigid mesh, with one entrance tunnel at ground level on each of the long sides. Trap 2 was a 2 m square 1 m tall cage with two ground tunnels, all made of 25 mm mesh 'chicken' wire netting. Trap 3 was a 4 m square and 2 m tall variant of trap 2, built to a plan supplied by Dr Ian Patterson (pers. comm.). A sketch of traps 2 and 3, which are built on site, is given in Figure 6a. All traps had ground tunnels constructed according to instructions in Hollom and Brownlow (1955) scaled to give an entrance of 500 mm span and 250 mm height, tapering to give an inner arch of 150 mm span and 250 mm in height (Fig. 6b). The tunnels projected about 500 mm into the trap cavity. An important innovation fitted to all three traps in 1979 was a trap door fitted to the inner end of the tunnel (Fig. 6b). This was constructed of 3.5 mm gauge wire for the frame, covered by 25 mm mesh wire netting, and of such a size as to be slightly larger than the inner end of the tunnel. Being hinged to the top of the inner end of the tunnel it formed a one way trap door. When the trap was set, these doors were propped open by means of a piece of wife secured to the door by means of a piece of Blu-Tack, a proprietary adhesive putty. As a result, a number of birds entered the trap before the wires supporting the trap doors were knocked away. Those birds inside were prevented from leaving by the closed doors and this attracted other birds down which, as observations proved, entered by pushing open the doors from the outside. It is likely that the presence of a number of Rooks in the cage provides a strong attraction which encourages others to enter (Green, in press; Waite, in press). Trap 1 is fairly expensive to build as the cost of the mesh is high, and it is also a difficult material to work. Trap 3, a fairly traditional design, is expensive to build because of the large amount of netting and the substantial corner poles required, which are difficult to drive into the ground. Trap 2, on the other hand, is very cheap and simple to build, being smaller. It can be built by one person in less than three hours and will give long, maintenance-free service. All three traps were in simultaneous use during 1979, so that their comparative effectiveness could be assessed. They were spaced 50 m apart, out of sight of each other and all within 100 m of a rookery. They were all baited with bruised oats and bread. Trap 2 is by far the most effective design, catching (a) Figure 6. a) Trap of type 2 & 3, as described in appendix, b) Detail of entrance tunnel, with propped trap door at inner end.

10 TRAPPING ROOKS 211 just 11% fewer birds than the large trap 3 despite being only one third the volume (Fig. 7). Trap 1 was only half as big as trap 2 and the catch was reduced by 50%. Thus trap 2 is by far the most cost effective, and details for its construction are given below. The use of stuffed or artificial decoys had no apparent effect on trapping success. Construction of the recommended trap. Four 20 mm diameter bamboo poles of approximately 1.5 m length are driven into the ground at the corners of a 2 m square. Lengths of 1 mm binding wire are used as guy wires, which are secured to the ground with tent pegs. A length of 1.5 mm wire is no trapping Jl r~^ H i trap 2 J T Ivl t.yx-x*:*:r-h P Ei id is 1 id w*«k no. F. M, A. M, J «: : : : : :!! : Figure 7. Numbers of Rooks trapped by each of the three designs of trap, placed equally close to a rookery.

11 212 RINGING AND MIGRATION attached to the top of a pole and fed from pole to pole to make a supporting wire from which the sides are hung. The sides are made from 25 mm mesh, 1 m tall wire netting bound to the supporting wire with 1 mm wire and to the ground with tent pegs at 500 mm intervals. The tunnels are constructed as described above with strengthening arches of 3.5 mm wire (see Fig. 6b), and bound to the walls with 1 mm binding wire and to the ground with tent pegs. The roof of 25 mm mesh wire netting is then attached, a side door is made and the trap is complete. Such a trap situated near to a rookery and baited with bread and bruised oats can be expected to catch Rooks in moderate numbers from March to June, thereafter large numbers of Jackdaws can be caught until August. If the trap is emptied early in the morning, one or two repeat catches can often be made in one day. Other species caught in moderate numbers included Herring Gulls Larus argentatus, Blackheaded Gulls Larus ridibundus, Pheasants Phasianus colchicus and PartridgesPerdixperdix. No Carrion Crows were ever caught, although many ted near to the traps. Addition of a roof funnel might help to catch this species. ALLSAGER, D. E., STENRUE, J. B. and BOYLES, R. L Capturing Black-billed Magpies with circular live traps. J. Wildl. Mgmt. BESHIR, S. A., Seasonal variations in some aspects of condition in the Rook. Unpublished M.Sc. Thesis, University of Aberdeen. BROWN, J The Evolution of Behaviour. New York: Norton. CHARLES, J. K Territorial behaviour and the regulation of population size in the Carrion Crow Corvus corone & Corvus cornix. Unpublished Ph.D. Thesis, University of Aberdeen. COOMBS, C. J. F Observations of the Rook Corvus frugilegus in Southwest Cornwall. Ibis 102: DOWNHOWER, J. F Darwin's finches and the evolution of sexual dimorphism in body size. Nature 263: DUNNET, G. M The breeding of the Starling in relation to its food supply. Ibis 97: FEARE, C. J., DUNNET, G. M. & PATTERSON, I. J Ecological studies of the Rook in N.E. Scotland; food intake and feeding behaviour. J. appl. Ecol. 11: GERARD, B. M Factors affecting earthworms in pasture. J. Anim. Ecol. 36: GREEN, P. T. (in prep) Coloniality in the Rook. Ph.D. thesis, University of Edinburgh. GREEN, P. T. (in press) Sexing Rooks by discriminant function analysis. Ibis. HEPPLESTON, P. B. Anotatomical observations on the bill of the Oystercatcher Haematopus ostralegus. J. Zool. (Lond.) 161: HEPPLESTON, P. B Anatomical observations on the bill of the Oystercatcher Haematopus ostralegus. J. Zool. (Lond.) 161: HEPPLESTON, P. B The feeding ecology of the Oystercatcher in North East Scotland. J. A nim. Ecol. 40: HOLLOM, P. A. D. & BROWNLOW, H. G Trapping methods for bird ringers. BTO Field Guide, No. 1. HOLYOAK, D Movements and mortality in the Corvidae. Bird Study 18: HOUSTON, D The effect of Hooded Crows Corvus cornix on hill sheep farming in Argyll, Scotland. The food supply of Hooded Crows, J. appl. Ecol. 14: JEHL, J. R Sexual selection for size differences in two species of sandpipers. Evolution 24: JONES, P Some aspects of the feeding ecology of the Great Tit Parus major. Unpublished D. Phil, thesis, University of Oxford. KING, B. &ROLLS, J. C Feeding methods of Rook with malformed bill. Brit. Birds. 61: LOCKIE, J Breeding and feeding of three species of Corvidae. D. Phil. thesis, University of Oxford. LOMAN, J Reproduction in a population of the Hooded Crow Corvus cornix Ph.D. Thesis, University of Lund. MAFF, Poultry nutrition. HMSO Bulletin no NEWTON, I Temperatures, weights and body composition of moulting Bullfinches. Condor 70: PICOZZI, N A study of the Hooded/Carrion Crow in N.E. Scotland. Brit. Birds. 68: ROLLS, K Mammals in which females are larger than males. Q. Rev. Biol. 51: ROWLEY, I The ABC of Crow Catching. Australian Bird Bander 6: SEEL, D. C Moult in five species of Corvidae in Britain. Ibis 118: SELANDER, R Sexual dimorphism and differential niche utilization in birds. Condor 68: SELANDER, R Sexual dimorphism in birds. In Campbell, B. (ed.) Sexual Selection and the descent of Man. London: Heinemann. SIEGEL, S Non parametric statistics. New York: McGraw. VOIPIO, P On sexual dimorphism in the Jackdaw (Corvus monedula). Ornis. Fenn. 45: WAITE, R. K. (in press). Local enhancement for food finding in Rooks Corvus frugilegus foraging on grassland. Z. Tierpsychol. Paul T. Green, Zoology Department, The University, West Mains Road, Edinburgh EH9 3JT.