Sunteți pe pagina 1din 8

397

HUNGER-DEPENDENCY OF HIDING BEHAVIOUR AFTER A


PREDATOR ATTACK IN DOMINANT AND SUBORDINATE
WILLOW TITS
KARl KOIVULA, SEPPO RYTKONEN & MARKKU ORELL

ABSTRACT We studied rank-specificity of predation risk taking in captive


Willow Tits Parus montanus. Level of risk taking was expressed by dura-
tion of hiding after a simulated predator attack. By manipulating available
food, we separated the effects of hunger and other rank associated factors
on accepted predation risk. As expected, hungry birds took more risks than
satiated ones. When both the dominants and the subordinates were hungry
or satiated, no rank-related differences existed in the duration of hiding.
When the dominants had consumed more food than the subordinates, they
also hid longer. Similarly, the subordinates hid longer when they were sati-
ated and dominants hungry. We conclude that incautiousness of subordi-
nates, reported earlier with regard to variety of antipredatory behaviours of
flocking birds, probably reflects a strategy to cope with high risk of starva-
tion resulting from restricted access to food.
Department of Biology, University of Qu1u, Linnanmaa, FIN-90570 Qu1u,
Finland. Fax: +358-81-5561227; E-mail: ktkoivu1@cc.ou1u.fi. Correspon-
dence to Karl Koivu1a.

INTRODUCTION such as foraging. When the tits Parus spp. detect


a predator, they usually rapidly fly into a hide,
Obtaining food and avoiding predators are often dense bushes or trees where they freeze (e.g.
considered as conflicting goals for animals. In Morse 1973, Ficken & Witkin 1977). After the
many instances this trade-off has been converted predator has disappeared, birds have to decide
into a time allocation problem. Especially in when to resume feeding. The probability that the
flocking birds the relationship between group size predator could still be in the vicinity obviously
and time devoted to foraging and scanning for decreases as time passes. Therefore, hesitation
predators has been studied a lot (see e.g. Elgar can increase chances of survival i.e. be less risky
1989 and references). Optimality models predict than ceasing from hiding earlier. On the other
that when compromising between predation and hand, starvation risk increases in the course of
starvation risks, hungry animals should take gre- time used in hiding.
ater risks in terms of predator avoidance than the Regner (1985) found that subordinate captive
ones near satiation (e.g. McCleery 1978, Krebs Blue Tits Parus caeruleus returned to a feeder
1980, Krebs & Kacelnik 1991). earlier than dominants after presentation of Spar-
Dominance secures free access to essential re- rowhawk Accipiter nisus model. De Laet (1985),
sources such as food or shelter (e.g. Kaufmann Waite & Grubb (1987) and Rogstad (1988a) have
1983). Because of restricted access to food, prob- obtained similar results from other Paridae.
ability of starvation can be higher among subordi- There are several alternatives, although not mutu-
nates. To cope with increased starvation risk sub- ally exclusive, how this kind of rank-dependence
ordinates may be forced to decrease time devoted could evolve. First, it is possible that because of
to antipredator activities. their free access to food dominants could have
One aspect of predation risk emerges when a better nutritional status than subordinates, who
predator appears and interrupts other activities therefore could be willing to take more predation

Received 29 December 1993, accepted 2 March 1995 ARDEA 83:397-404


398 ARDEA 83(2), 1995

risks to avoid starvation. Second, if dominants lengths (for details see Koivula & Orell 1988).
can exclude subordinates from food sources, sub- Two age-classes were separated: adults (born be-
ordinates could be forced to feed during unsafe fore the previous summer) and juveniles (born in
periods when dominants are not using the feeding the previous summer). Ageing was based on dif-
sites. A third alternative is that the higher risk tak- ferences in tail feather shape and abrasion (Laak-
ing among subordinates might be caused by rank- sonen & Lehikoinen 1976).
correlated factors not attached with food access. The birds were habituated to captivity condi-
For example, subordinates could be younger and tions for at least one week before the experiments.
less experienced, and hence, less cautious than They were kept separately in wire-meshed cages
dominants. (size: 50*80*50 cm). The cages were covered
If risk taking in terms of predation is hunger with white clothes to avoid disturbance and visual
mediated, subordinates should take more risks contact to other individuals. Furthermore, the
than dominants only when they are hungrier than birds were maintained on ad libitum diet of sun-
dominants. On the other hand, if accepted high flower seeds and water.
predation risk reflects a strategy to cope with re-
stricted time in access to food, subordinates could The experiments
risk more regardless of their hunger levels. Simi- The experimental cages consisted of two
larly, if the factors that determine the social status parts: inner and outer compartments. The inner
are correlated with risk taking, differences in hun- cage was similar to the ones in which the birds
ger levels should not have an influence in them. were housed during habituation. On the other
To test these predictions, we measured risk side, there were two perches and a thick hiding
taking and manipulated hunger levels of captive place that was built using small birch and spruce
Willow Tits Parus montanus, which is a forest- branches. Food and water dishes were situated on
living, year around resident passerine that forms the open side. The floor was wire-meshed and did
dominance-structured flocks during the non- not touch the platform surface. Therefore, once
breeding season. dropped, the seeds were not available for the birds
any more and therefore the food dish was the only
available seed source. The amount of food offered
METHODS was controlled by an observer who supplied food
through a plastic tube which pierced the walls of
The birds both cages and ended in the food dish. The walls
The experiments were conducted in Oulu, of the outer cage were cloth-walled and higher
northern Finland (65°N 25°30'E) during two suc- than those of the inner cage. Therefore, the birds
cessive winters 1989-1991. We used captive birds, could not see the observer sitting outside. A loose
because it is nearly impossible to individually piece of cloth formed a ceiling which was re-
manipulate the hunger levels of birds in natural moved on the day before the predator presenta-
conditions. The birds for the experiments were tions. The birds were observed through a small
captured at seed-baited feeders using mist-nets. hole in the wall of the outer cage.
To get the birds habituated to sunflower seed diet, Two birds of the same sex were introduced at
the feeders were maintained for several weeks be- the same time into the experiment cage on the day
fore catching the birds. Netting always took place before the first trial. The birds came from differ-
before noon. The birds were immediately trans- ent winter flocks and were therefore not familiar
ported indoors. They were marked and their wing, to each other. In the evening the food dishes were
tail and tarsus lengths were measured using meth- emptied and the lights were switched off. The
ods recommended by Svensson (1984). The birds darkness lasted for 17 hours (16.00 - 09.00),
were sexed by dimorphism in wing and tarsus which represents the approximate light conditions
Koivula et al.: RISK TAKING IN WILLOW TITS 399

of midwinter in northern Finland. The tempera- both birds were satiated by giving them 40 seeds
ture of the room was constantly kept near +20 dc. at the same time; (3) the dominant bird got more
In the morning, after the lights were switched on, seeds than the subordinate. Last of the treatments
the birds were deprived of food for two hours be- was arranged by offering the seeds to the birds
fore the experiments were conducted. one by one as when measuring dominance. In this
treatment the subordinate sometimes succeeded
Dominance to get some (1-4) seeds. In such cases we gave as
When measuring the dominance order, sun- many additional seeds as needed, so that finally
flower seeds were given to the birds one by one. the dominant had got 10 seeds more than the sub-
The next seed was provided only after the bird ordinate. We altered the sequence of treatments to
that picked the former one had consumed it. The avoid errors due to possible habituation to preda-
individual which first had picked and consumed tor model. After the trial the birds got food in ex-
10 seeds was considered to be the dominant one. cess. Only one nutritional treatment and predator
Usually, subordinate individuals got no seeds. presentation trial was conducted in one day for
The dominants, according to this method, were each dyad.
the winners in all the observed agonistic interac- A total of 13 dyads were tested. One pair,
tions, too. however, was excluded due to the death of the
subordinate counterpart on the day the last trial
Predator presentation and risk variables was conducted.
A mounted male Tengmalm's Owl Aegolius
funereus, a realistic predator for Willow Tits Experiment 2 Cage arrangements of the first ex-
(Korpimaki 1981), was used as a predator model. periment did not allow a treatment where the sub-
The owl was mounted to a standing position with ordinates would have been more satiated than the
its head facing downwards. It was placed by the dominants. Therefore, the design was changed by
observer on the upper frame of the outer cage at building a cage consisting of two compartments.
the same end where the food dishes were situated A tunnel with a trap door was built between them.
in the inner cage. The birds did not see the ob- Closing the door when the birds were in different
server during the placement. Model presentation cages enabled us to control the hunger levels of
lasted for 15 seconds. Food dishes that had been both birds. Two treatments were performed: dom-
empty for the preceding two hours were filled inants satiated and subordinates hungry and vice
during the predator presentation. When the model versa. Also here the birds were deprived of food
emerged, the birds rapidly moved into the hide, before the experiments and the sequence of treat-
froze and occasionally gave alarm calls. mentswas altered. The bird that was made sati-
Two time variables were measured after the ated got 10 sunflower seeds and the other one got
removal of the owl: the time that passed before none. After this the trap door was opened and clo-
the bird started to move inside the hiding place sed again immediately when the birds were in the
("scanning period", SP) and the time between the same compartment. Predator presentation and the
removal and the moment when the birds left the variables measured were the same as in experi-
thicket ("hiding period", HP). After leaving the ment 1. A total of 12 dyads were tested here.
cover, the birds usually came to the food dish im-
mediately.
Statistics
Treatments Despite the transformation attempts, durations
Experiment 1 Three kinds of nutritional situa- of scanning and hiding periods were not normally
tions were arranged for each dyad: (1) neither distributed. Therefore, non-parametric methods
bird got food before the model presentation; (2) were used when comparing risk-taking between
400 ARDEA 83(2), 1995

hunger levels and dominance classes. All the tests


were two-tailed. Critical values were obtained
180 o
from Rohlf & Sokal (1981). 150

~ 120
Ol
<:
RESULTS "c 90
<:
Cll
0
(J) 60
The duration of motionless scanning and time
30
spent inside the cover varied greatly between the
individuals and treatments. The shortest scanning 0
period lasted only for 3 seconds while the most
hesitating bird spent 11 minutes inside the hide.
180 ®
150
However, the frequency distributions of both
scanning and hiding were skewed towards short
periods. Most of the birds started to move and left
~
.e
Ol
120
/***
<: 90
the cover within the first minute. The birds scan-
ned on the average for 60.5 seconds (SE = 6.5,
~
..c:
60 / /***
n = 120). The mean duration of hiding period was 30 ~ 0 expenment 1
76.6 seconds (SE = 8.4, n = 120). • experiment 2
0
hungry satiated

Predation risk and hunger levels Figure 1. Time stayed motionless (SP) and time
The birds had significantly shorter scanning spent inside the cover (HP) in different stages of hun-
and hiding periods when they were hungry than ger. Means with SE-bars. n = 24 in all the cases. Wit-
when satiated. When no seeds were given before hin-individual comparisons between treatments with
the predator presentation, the average duration of Wilcoxon's matched-pairs signed-ranks tests (2-tailed).
both scanning and hiding periods was approxi- *** = p < 0.001.
mately 30 seconds (Fig. 1). When the birds were
satiated, scanning lasted on the average for 1 min- cantly longer than the subordinates in both ex-
ute and hiding nearly for 2 minutes (Fig. 1). In ex- periments (experiment 1: SP: T = 8, n = 12,
periment 1, one individual scanned equally long p = 0.012, HP: T = 8, n = 12, p = 0.012, Fig. 2c;
after both treatments, but 20 birds out of 24 (83%) experiment 2: SP: T = 10, n = 12, p = 0.021, HP:
scanned more and stayed inside the cover longer T = 8, n = 12, p =0.012, Fig. 3a). In experiment 2,
when they were satiated (Wilcoxon's matched- however, dominants took more risks than subor-
pairs signed-ranks. tests; SP: T = 30, n = 23, dinates when they were hungrier; in 11 out of 12
p =0.001, HP: T = 29, n = 24,p =0.0005, Fig. 1). dyads they had shorter scanning and hiding peri-
Similarly in experiment 2, 83% of the birds had ods than subordinates (SP: p = 0.023, HP:
longer SP's and HP's when they were satiated p = 0.023, Fig. 3b). No such differences existed in
(SP: T = 23, n = 24, p = 0.0003, HP: T = 28.5, experiment 1 when the hunger levels were equal.
n = 24, p = 0.0005, Fig. 1). When both the birds were hungry, dominants
were the first to move and leave the cover in 6 out
of 12 cases (SP: T = 37, n = 12, p = 0.88, HP:
Predation risk and dominance T = 33.5, n = 12, p = 0.67, Fig. 2a). Similarly,
When the dominants got more food they were when both the birds were satiated dominants had
more cautious than the subordinates. They both shorter scanning and hiding periods in 50% of the
scanned and remained inside the hide signifi- trials (SP: T = 39, n = 12, P = 1, HP: T = 37,
Koivula et al.: RISK TAKING IN WILLOW TITS 401

180 180
150 150
120 120
~ ~
Ql 90 Ql 90
E
:;::; ~
60 n.s. n.s. 60
30 30
0 0

180 ® 180
®
150 150
120 120
~ ~
Ql 90 Ql 90
E E
:;::;
:;::; 60 60
30 30 dominants
subordinates
0
O:=~~~=============~ scanning hiding
180
DOMINANT SATIATED
©
Figure 3. Lengths of scanning and hiding periods in
150 experiment 2 when (a) dominant was satiated and sub-
120 ordinate hungry and when (b) dominant was hungry
~ and subordinate satiated. Means with SE-bars. n = 12
Ql 90 for each treatment. Difference between dominance cat-
.§ egories tested with Wilcoxon's matched-pairs signed-
60
ranks test (2-tailed). * = p < 0.05.
30 dominants
subordinates
0

DISCUSSION
Figure 2. Mean lengths (with SE-bars) of scanning
and hiding periods of dominants and subordinates in Recent research suggests that varying predation
experiment 1 when (a) both the birds were hungry, (b) pressure is the key factor behind the rank-depen-
both were satiated, and (c) dominants were satiated and dent survival observed in certain birds, among
subordinates hungry. n = 12 for each treatment. Differ- others in the Willow Tit (e.g. Ekman et al. 1981,
ences between dominance categories tested with Wil-
Ekman 1987, Desrochers 1989). Additionally, for-
coxon's matched-pairs signed-ranks test (2-tailed).
aging time seems to be critical for survival in
n.s. = p > 0.10, * = p < 0.05.
temperate tits (e.g. Jansson et al. 1981, Ekman
1987). In fact, rank-dependent time allocation pat-
n = 12, p = 0.88, Fig. 2b). The results show that tern forms an evident link between rank and sur-
dominance without differences in satiation has no vival (e.g. Ekman 1990).
impact on a taken predation risk. Subordinates It is known that because of presence of domi-
can take more risks than dominants in terms of nants, subordinate Willow Tits use more open or
predation, but only when they are hungrier than otherwise unsafe microhabitats than dominants
the dominants. (Ekman & Askenmo 1984, Rogstad 1988 a,b). As
402 ARDEA 83(2), 1995

with foraging habitat, exposure to predation may levels. In our study the absence of differences in
vary with foraging time. We assumed that the risk taking, when the dominant and subordinate
time passed after the appearance of the predator Willow Tits were both hungry or both satiated,
predicts its probability to be nearby, which in tum suggests that dominance per se does not cause
equals to a predation risk. In this sense the risk any differences in risk taking. Thus, it is not prob-
taking of Willow Tits was consistently dependent able that subordinates in natural flocks risk more
on their hunger levels. Thus, Willow Tits seem to to acquire rank-limited foraging time in the ab-
be capable of adjusting their foraging behaviour sence of dominants. In fact, this possibility is re-
to balance with predation and starvation risks. alistic only if the food sources are so highly clum-
The results are in line with Krebs (1980) who ped that simultaneous feeding by many group
found risk taking in Great Tits Parus major to be members is not possible. Food at the feeders can
state-dependent: tits allocated more time for scan- provide such circumstances, but as far as natural
ning when approaching satiation. Similar hunger- food is concerned, this is hardly the case in boreal
dependency of predation risk taking has also been tits.
reported with regard to other aspects of foraging Our experiments consisted of unisexual dy-
and antipredator behaviour (Milinski & Heller ads. Therefore, we could exclude the possible
1978). sex-specificity in antipredatory behaviour. Simi-
Subordinate Willow Tits were less cautious larly, with one exception, the birds in dyads of ex-
than the dominants, but only when being more periment 1 were matched with respect to their
hungry than them. There are results of four differ- age, although in captivity age does not correlate
ent Parus species showing that subordinates are with rank among birds of matching status of resi-
less cautious and resume feeding earlier than dency (Koivula et at. 1993). Nevertheless, there is
dominants after the appearance of a natural pred- evidence for sex- or age-related predation risk
ator (De Laet 1985) or a model of it (Hegner 1985, taking, for example with respect to scanning for
Waite & Grubb 1987, Hogstad 1988a). The obser- predators (e.g. Heinsohn 1987, Elgar 1989), giving
vation that subordinates take more risks than alarm calls (Alatalo & Helle 1990), or responding
dominants has also been reported in studies con- to alarming (Breitwisch & Hudak 1989). There-
cerning the openness of foraging habitat. Subor- fore, we cannot rule out the possibility that corre-
dinates use more feeding sites far from protective lates of rank such as sex and age are partly re-
cover (Schneider 1984, Hogstad 1988a, Piper sponsible for the reported associations between
1990, Koivula et at. 1994). Hunger levels, how- rank and predation risk. Further experiments are
ever, were not controlled in these studies. In fact, needed to solve their role. Our results, however,
De Laet (1985) reported that dominant Great Tits show that hunger level can be a correlate of dom-
visited the feeders more often than subordinates inance playing an important role in shaping those
before the predator appearance. Therefore, their associations. The fact that the dominants were
nutritional status could also have been better. less cautious than the subordinates when they
Similarly, dominant Blue Tits in Hegner's (1985) were more hungry, strengthens this interpretation.
laboratory study tended to spend more time in The type of risk taking that we handled can
feeders before the presentation of the predator have direct survival consequences. In addition,
model. In captive Tufted Titmice Parus bicotor there is a hypothetical indirect mechanism that
Waite & Grubb (1987) tried to minimize the effect could augment the rank-related differences in pre-
of hunger levels, although they did not equalize dation pressure. When the subordinates resume
the food provision in a quantitative manner. Re- feeding first, the predator can attack them if it is
sults of Waite & Grubb (1987) suggest that domi- still in the vicinity. Consequently, the dominants
nants could behave more cautiously than subordi- can have a possibility to get information whether
nates even if there were no differences in hunger it is safe for them to resume feeding or not. In the-
Koivula et al.: RISK TAKING IN WILLOW TITS 403

ory, using subordinates as decoys, dominants can Hegner, R.E. 1985. Dominance and anti-predator beha-
by small loss of foraging time reach a low level of viour in Blue Tits (Parus caeruleus). Anim. Be-
predation risk, which without incautiousness of hav. 33:762·768.
Heinsohn, R.G. 1987. Age-dependent vigilance in win-
subordinates would have required a substantially ter aggregations of cooperatively breeding White-
longer hiding time, and hence, a higher risk of winged Choughs (Corcorax melanorhamphos).
starvation. Behav. Ecol. Sociobiol. 20:303-306.
Hogstad, O. 1988a. Social rank and antipredator beha-
viour of Willow Tits Parus montanus in winter
flocks. Ibis 130:45-56.
ACKNOWLEDGEMENTS
Hogstad, O. 1988b. Rank-related resource access in
winter flocks of Willow Tit Parus montanus. Ornis
We thank Kimmo H. Kumpulainen and Kimmo Lahti Scand.19:169c 174.
for helping us in catching the birds and maintaining Jansson, C., Ekman J. & von Bromssen A. 1981. Win-
them in the laboratory. The study was financially sup- ter mortality and food supply in tits Parus spp. Oi-
ported by the Finnish Cultural Foundation, the Acad- kos 37:313-322.
emy of Finland and the University of Oulu. Kaufmann, J.H. 1983. On the definitions and functions
of dominance and territoriality. BioI. Rev. 58: 1-20.
Koivula, K & Orell M. 1988. Social rank and winter
survival in the Willow Tit Parus montanus. Ornis
REFERENCES Fennica 65: 114-120.
Koivula, K, Lahti K., Orell M. & Rytkonen S. 1993.
Alatalo, R.Y. & Helle P. 1990. Alarm calling by indi- Prior residency as a key determinant of social
vidual Willow Tits Parus montanus. Anim. Behav. dominance in the Willow Tit (Parus montanus).
40:437-442. Behav. Ecol. Sociobiol. 33:283-287.
Breitwisch, R. & Hudak J. 1989. Sex difference in risk- Koivula, K, Lahti K., Rytkonen S. & Orell M. 1994.
taking behavior in foraging flocks of House Spar- Do subordinates expose themselves on predation?
rows. Auk 106:150-153. Field experiments on feeding site selection of Wi1-
De Laet, J.E 1985. Dominance and anti-predator beha- low Tits. J. Avian BioI. 25:178-183.
viour of Great Tits Parus major: a field study. Ibis Korpimiiki, E. 1981. On the ecology and biology of
127:372-377. Tengmalm's Owl (Aegolius funereus) in southern
Desrochers, A. 1989. Sex, dominance and microhabitat Ostrobothnia and Suomenselka western Finland.
use in wintering Black-capped Chickadees: a field Acta Univ. Oul. Ser. A 118:1-84.
experiment. Ecology 70:636-645. Krebs, J.R. 1980. Optimal foraging, predation risk and
Ekman, J. 1987. Exposure and time use in Willow Tit territory defence. Ardea 68:83-90.
flocks: the cost of subordination. Anim. Behav. Krebs, J.R. & Kacelnik A. 1991. Decision-making. In:
35:445-452. J.R. Krebs & N.B. Davies (eds) Behavioural Ecol-
Ekman, J. 1989. Ecology of non-breeding social sys- ogy. An Evolutionary Approach. Blackwell Scien-
tems of Parus. Wilson Bull. 101 :263-268. tific Publications, Oxford.
Ekman, J. 1990. Alliances in winter flocks of Willow Laaksonen, M. & Lehikoinen E. 1976. Age determina-
Tits; effects of rank on survival and reproductive tion of Willow and Crested Tits Parus montanus
success in male-female associations. Behav. Ecol. and P. cristatus. Ornis Fennica 53:9-14.
Sociobiol. 26:239-245. McCleery, R.H. 1978. Optimal behaviour sequences
Ekman, J. & Askenmo C.E.H. 1984. Social rank and and decision making. In: J.R. Krebs & N.B. Da-
habitat use in Willow Tit groups. Anim. Behav. vies (eds) Behavioural Ecology. An Evolutionary
32:508-514. Approach. Blackwell Scientific Publications, Ox-
Ekman, J., Cederholm G. & Askenmo C. 1981. Spac- ford.
ing and survival in winter groups of Willow Tits Milinski, M. & Heller R. 1978. Influence of a predator
Parus montanus and Crested Tits P. cristatus - a on the optimal foraging behaviour of sticklebacks
removal study. J. Anim. Ecol. 50:1-9. (Gasterosteus aculeatus). Nature 275:642-644.
Elgar, M.A. 1989. Predator vigilance and group size in Morse, D.H. 1973. Interactions between tit flocks and
mammals and birds: a critical evidence of the em- Sparrowhawks Accipiter nisus. Ibis 115:591-593.
pirical evidence. BioI. Rev. 64: 13-33. Piper, W.H. 1990. Exposure to predators and access to
Ficken, M.S. & Witkin S.R. 1977. Responses of Black-cap- food in wintering White-throated Sparrows Zono-
ped Chickadee flocks to predators. Auk 94: 156-157. trichia albicollis. Behaviour 112:284-298.
404 ARDEA 83(2), 1995

Rohlf, EI. & Sokal R.R. 1981. Statistical Tables. W. R. SAMENVATTING


Freeman and Company, San Francisco.
Schneider, K.I. 1984. Dominance, predation, and opti-
Van Matkopmezen werd in gevangenschap bekeken
mal foraging in the White-throated Sparrow
flocks. Ecology 65:1820-1827. hoe individuen van verschillende sociale status in een
Svensson, L. 1984. Identification guide to European groep reageerden op het aanbod van voedse1 in re1atie
passerines. Stockholm. tot de aanwezigheid van een predator. Ret nemen van
Waite, T. A. & Grubb T.C. 1987. Dominance, foraging risico werd hierbij uitgedrukt als de duur waarin ge-
and predation risk in the Tufted Titmouse. Condor schuild werd na een aanva1 van een gesimuleerde pre-
89:936-940. dator. De aangeboden hoeveelheid voedsel werd gema-
nipuleerd, om uit te zoeken tot welk niveau risico aan-
vaardbaar werd geacht door de vogels. Zoals verwacht
namen uitgehongerde individuen meer risico dan ver-
zadigde soortgenoten, ongeacht hun socia1e status. Er
werd geen verschil gevonden in schui1gedrag tussen
dominante en ondergeschikte vogels bij een gelijk aan-
bod van voedsel. De auteurs concluderen dan ook dat
het tijdens eerdere onderzoeken gevonden roeke1oze
gedrag van ondergeschikte vogels in een groep in re1a-
tie tot diverse antipredator-gedragingen te maken heeft
met het hoge risico van uithongering als gevolg van een
beperkt voedselaanbod. - GOK

S-ar putea să vă placă și