Behavioral Ecology Vol. 12 No. 4: 375-380
© 2001 International Society for Behavioral Ecology
The influence of predation risk on threat display in great tits
Department of Zoology, Stockholm University, S-106 91 Stockholm, Sweden
Address correspondence to H. Lange. E-mail: henrik.lange{at}zoologi.su.se .
Received 1 July 2000; revised 14 July 2000; accepted 1 August 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In wintering birds, conflicts over food are often resolved by threat displays. For displays to be effective, there ought to be a cost associated with displaying. We investigated whether increased vulnerability to predators due to reduced vigilance could be such a cost. Conflicts ought then to be resolved using fewer or less intense displays in conditions of high risk. We also looked for differences between dominants and subordinates in their reaction to risk. Because there is considerable evidence that subordinate wintering birds forage in riskier places than dominants, one might expect dominants to be less successful in conflicts under high predation risk. In our experiment, nine flocks of four or five wintering male great tits were kept in outdoor aviaries. In the predation risk treatment, a stuffed pygmy owl was briefly shown before birds were allowed access to a feeder. In the control treatment the owl did not appear. The predator presentation caused a reduction in the amount of aggression shown by subordinates, whereas for dominants there was no statistically significant change. Dominants were at least as successful in subduing subordinates under high risk as under low risk. A possible interpretation is that our experiment reflected a natural foraging situation for great tits, where ephemeral resources can appear unpredictably. In such situations, dominants may need to be bold to gain priority of access even under increased risk of predation, whereas a subordinate would gain little by risking a conflict with small chances of winning.
Key words: great tit, Parus major, predation risk, threat display, social dominance.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The agonistic behavior of many group-living animals, such as wintering passerines, ranges from overt aggression to more or less ritualized threat displays. Overt aggression in a group is often more frequent during an initial stage of group formation, before a dominance hierarchy has emerged, and later decreases so that displays become the most common form of aggressive behavior (Chase, 1980
;
Verbeek et al., 1999).
Displaying tends to indicate an individual's motivation; for instance, in
conflicts over food, a hungry individual is more likely to display. Beyond
this basic premise, there are different possibilities for the interpretation
of the nature of the behavior. First, threat displays could function as
threats in the ordinary sense of the word, signaling a high probability of
imminent attack. The feasibility of this interpretation has been debated
(Caryl, 1979;
Hinde, 1981;
Maynard Smith, 1979), but it
is now generally thought to be logically consistent
(Enquist, 1985). It is less
clear that threat displays actually function in such a manner i.e., that
exchanges of display behavior frequently escalate to physical fighting). It
has been argued that the correlation between threat display and attack might
be quite weak (Andersson, 1980;
Caryl, 1979;
Popp, 1987;
Senar, 1990).
A second possibility is that displaying entails other types of costs than
the possibility of injury in subsequent physical fighting. One such cost might
be increased vulnerability to predators for a displaying dyad, since
individuals engaged in a conflict tend to direct a larger part of their
attention toward each other, lessening their vigilance. This kind of
relationship between aggressive behavior and vigilance has been demonstrated
in fish and birds by Jakobsson et al.
(1995) and by Brick
(1998). Our purpose in this
study was to investigate whether changes in predation risk affect interactions
between great tits in a manner consistent with such a risk-taking
hypothesis.
Assuming that a function of threat display is to subject an opponent to
higher risk by reducing its vigilance, it follows that a given display becomes
more effective, but also more costly, with higher background risk of
predation. One would then predict conflicts to be resolved using fewer or less
intense display behaviors in conditions of higher risk (cf.
Slotow, 1996). In an extreme
case of very high risk, individuals might avoid aggressive behavior
altogether.
In a dominance-structured group one could also expect predation risk to
affect dominant and subordinate behavior differently. There is considerable
evidence that subordinate wintering birds forage in riskier places than do
dominants (e.g., Ekman, 1987;
Koivula et al., 1994;
Slotow and Rothstein, 1995). A
suggested explanation for this phenomenon is that subordinates have a more
limited choice of where and when to forage and are forced to take higher
risks, whereas dominants can avoid exposing themselves to increased predation
risk (Ekman, 1987;
Ekman and Askenmo, 1984;
Koivula et al., 1994). Also,
if dominants are stronger fighters than subordinates, but are no better at
escaping predator attacks, the advantage for dominants ought to be smaller
when predation risk is an important cost of aggressive interactions. Thus, one
might expect subordinates to be relatively more successful under high risk of
predation. On the other hand, there is the possibility that dominants are not
simply characterized by having higher fighting ability, but that they are also
bolder and generally more active (cf.
Verbeek et al., 1994). One
might then expect dominants to be even more successful under high risk of
predation.
In this study, we looked for an influence of predation risk on threat display in an experiment with flocks of wintering great tits in outdoor aviaries, using a stuffed owl as predator. We tested the prediction that there will be fewer or less intense threat displays with higher risk, and we also investigated if dominants and subordinates respond differently to higher risk.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study animals and experimental setup
The experiment was carried out at Tovetorp Zoological Research Station in south-central Sweden between 10 December 1998 and 12 March 1999. During this time, the outdoor temperature varied from 7° to -18°C. Nine flocks of three to five great tits (Parus major) were used (permit Dnr. 58-98 by Linköpings djurförsöksetiska nämnd). All the birds were male, and the majority were in their first winter (age was scored according to Svensson [1994
]). The birds were
wild-caught in the vicinity of the research station and subsequently released
in the same area. After capture, we color banded birds and grouped them into
flocks; each flock spent its entire period of captivity in an outdoor aviary,
where the experiment was performed. We used two similar roofed outdoor aviaries, with dimensions 3.5 x 7 x 3m (W x L x H), placed approximately 7 m apart. Each contained two remote-controlled seed feeders, a remote-controlled experimental feeder, and some branches and bushes for shelter. In each aviary, a tube measuring 20 cm wide and 60 cm high was mounted, from which a stuffed pygmy owl (Glaucidium passerinum) could be displayed by remote control. We made observations using two video cameras (Panasonic NV-DX100 digital video) that were set up in fixed positions outside the aviaries, facing the experimental feeders. The feeders allowed access to only one bird at a time, but had a 20-cm long array of perches attached, along which intruders could approach the feeding bird. The experimental food consisted of a mixture of mealworms and animal fat melted into cakes, which proved to be quite attractive to the great tits.
After being introduced into the aviary, birds were kept for 3-5 days with food available ad libitum, including occasional access to experimental food, during which time dominance ranks could establish. After the initial period, we recorded dominance interactions by preventing access to food and then allowing access to the experimental feeder. This was performed twice, with food available ad libitum for one day in between. After one more day the experiment started; the birds were fed ad libitum for one day between experimental trials.
In the morning of the trial days, all feeders were closed for a period of between 0.5 and 3 h, depending on weather conditions and temperature, after which the experimental feeder was opened. In the predator treatment the stuffed pygmy owl was displayed for 30 s, 5 min before the experimental feeder was opened. In the control treatment, the owl was not shown. We videotaped the ensuing interactions during 15 min, after which time the trial was stopped and all feeders were opened. The order of control and predator treatment days for each flock was determined by randomization.
Data recording and analysis
The videotaped behaviors were transcribed onto Microsoft Excel worksheets
as sequences of behaviors. A row of such a worksheet referred to the behavior
of a single bird at a given instant and contained entries for time, color band
identity, and behavior performed. Birds present at the feeder were recorded in
this way in an alternating fashion, with new rows of the worksheet being added
when behavior changed or individuals arrived or left. However, scanning
behavior was too frequent to form new rows for each scan; instead, we recorded
the number of scans between two instants. On average, there were 1.7 rows/s
during interactions. A Visual Basic program was used to extract data from the
worksheet. Threat displays and other aggressive behaviors were classified
according to Blurton Jones
(1968)
(Table 1). For the analyses we
grouped supplanting attack, lunge, and full attack into a single category
referred to as "attacks." Concerning displays, we mainly analyzed
the occurrence of the wings-out threat display because this behavior was
typically used in the most intense sequences of displaying. We also divided
the worksheet sequences of behaviors into dyadic bouts. A bout started with
the first aggressive behavior since the ending of the previous bout and ended
when either the initiator or recipient of this first aggression escaped or
stopped displaying completely. The bird remaining at the feeder was considered
the winner of the bout.
|
We assigned dominance ranks in the flocks using a matrix of dyadic wins and losses. For the analyses we only used two dominance classes: the top-ranked bird in each flock and the remaining subordinates.
Statistical analysis
All statistical tests were performed as pairwise comparisons at the flock
level (N = 9), either comparing predation and control treatments or
sometimes comparing time periods or dominance categories within flocks.
Throughout, we used the Wilcoxon signed-ranks test. When reporting a
comparison we indicate the number of nonzero differences
(Ndiff) out of the total (N = 9) because only
those differences are actually used by the test. Mean values are reported with
their standard errors, where the flock again is the sampling unit.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
General observations
The experimental trials usually developed in the same way: a single bird approached the feeder at first and was soon thereafter joined or supplanted by one or several others. During bouts, one bird was usually sitting at the feeder and the other at a varying distance on the attached array of perches. Full attacks often caused both individuals to leave the feeder briefly, after which at least one of them returned. A considerable number of bouts were observed (total 318), containing both threat displays and attacks. In a few cases the aggression escalated to the level of prolonged grappling, but these interactions took place out of camera view.
Vigilance
Displaying the stuffed owl shortly before making food available resulted in
an increased vigilance in the birds. The number of scans per second per
individual during the first 5 min was higher in the predation treatment than
in the control (0-5 min: T = 0, Ndiff = 9,
p <.005; Figure 1).
The effect of the owl appeared to fade over time, with successively smaller
differences in scanning rate between the treatments (5-10 min: T =
10, Ndiff = 9, p = 0.2; 10-15 min: T =
9, Ndiff = 6, p >.5) and a tendency toward a
lower scanning rate in the owl treatment in the 10- to 15-min interval than
during the first 5 min (T = 4, Ndiff = 8,
p =.055; Figure
1).
|
In addition to the higher average scanning rate, one might expect that the birds should avoid long interscan intervals during the owl treatment because such pauses in vigilance could allow a predator to attack. This pattern was in fact found. During the first 5 min of the owl treatment, there were essentially no interscan intervals longer than 2 s, whereas a considerable number were observed in the control treatment (T = 4.5, Ndiff = 9, p =.033; Figure 2). The birds were either eating or displaying during these pauses in vigilance.
|
2 s), for all individuals
present at the feeder during the first 5 min of control treatment (shaded bar)
and predation treatment (open bar). *p <.05.
Feeding
Even though the birds clearly behaved more cautiously in the owl treatment,
they still managed to keep up their rate of foraging. Comparing the total
amount of time (in seconds) spent eating at the feeder during the first 5 min
showed no difference between treatments (control: 220.6 ± 19.0 s; owl:
188.6 ± 31.4 s; T = 18, Ndiff = 9,
p >.5). Similarly, considering on the one hand the time spent
eating by the dominant male (control: 79.2 ± 24.5 s; owl: 80.2 ±
22.5 s), and on the other hand the total time spent eating by all the
subordinates (control: 141.3 ± 17.1 s; owl: 108.4 ± 29.6 s),
there were no significant differences between the treatments (p
>.4 for both comparisons).
Agonistic behavior by dominants and subordinates
For the analysis, we considered two classes of behavior patterns. The first
group consisted of the three kinds of attacks (full attack, lunge, and
supplanting attack) and the second of threat displays involving wings-out (WO)
as one component. Looking at the number of behaviors of these types performed
by all individuals present at the feeder during the first 5 min, there was no
significant difference between the treatments (attacks: T = 9,
Ndiff = 7, p >.4,
Figure 3a; wings-out displays:
T = 8, Ndiff = 8, p >.1,
Figure 3b). However,
considering instead long threat displays, uninterrupted by scans for more than
one second, there were fewer such sequences in the owl treatment (T =
2, Ndiff = 8, p = 0.02,
Figure 3c).
|
Next, we split the data into behaviors by the dominant and the remaining subordinates. For the dominant there were no significant differences for either attacks (T = 17,5, Ndiff = 8, p > 0.5; Figure 4a) or WO displays (T = 15,5, Ndiff = 8, p >.5; Figure 4b), whereas the subordinates both attacked less (T = 0, Ndiff = 7, p =.02, Figure 4a) and used fewer WO displays (T = 2.5, Ndiff = 8, p =.03; Figure 4b) in the owl treatment. Thus, it could be that subordinates reacted more strongly than dominants to the predator presentation, although our data leave this question open. There was no significant difference between dominance classes in the change between treatments in the number of attacks or displays (attacks: T = 12, Ndiff = 8, p >.4, WO displays: T = 8, Ndiff = 8, p >.2).
|
The reduction in the amount of aggression by subordinates could be explained in two different ways. First, it could be that the dominant individual participated more often in aggressive bouts in the owl treatment and, accordingly, that the subordinates participated less often. Because the dominant does not participate in all interactions, bouts between two subordinates could be less common under high predation risk. In fact, at least part of the reduction in aggression by subordinates appears to have this cause. On average, during the first 5 min the dominant participated in 85% of the bouts in the owl treatment and 65% in the control, although this difference is not quite significant (T = 4, Ndiff = 8, p =.055; Figure 5a).
|
Second, it could be that increased predation risk caused the subordinates to lower their rate of aggression during conflicts to a greater extent than the dominants. This seems not to have been the case regarding attacks: neither the subordinates (T = 14, Ndiff = 8, p >.5; Figure 5b) nor the dominants (T = 22.5, Ndiff = 9, p >.5; Figure 5b) changed their rate of attack between treatments. For WO displays, however, we found that subordinates had a lower rate in high predation risk (T = 0, Ndiff = 8, p =.008; Figure 5c), whereas there was no significant difference for the dominants (T = 16, Ndiff = 9, p >.4; Figure 5c).
Our results on the pattern of displaying in relation to dominance and predation risk were obtained by looking specifically at the WO element. Considering instead all display elements in Table 1 led to the same conclusion: the subordinates showed a significant difference between treatments, but the dominants did not. Because several display elements often occur in close association, it is reasonable to use the rate of WO display as a measure of display behavior.
Regardless of treatment, dominants appeared to have a higher rate of aggressive behavior (Figure 5b, c). Comparing dominance classes during the first 5 min, and averaging over treatments, there was a tendency toward difference in attack rate (dominants: 0.20 ± 0.07/s, subordinates: 0.06 ± 0.03/s; T = 7, Ndiff = 9, p =.08) and a significant difference for the rate of WO displays (dominants: 0.42 ± 0.07/s, subordinates: 0.12 ± 0.03/s; T = 0, Ndiff = 9, p =.008).
Wins and losses by dominants and subordinates
We did not find that subordinates were relatively more successful under
high risk of predation. Instead, the dominant bird won a somewhat larger
proportion of his conflicts in the owl treatment (90%) than in the control
(82%), although this difference was not statistically significant (T
= 4, Ndiff = 5, p >.2). Evidently, the
dominant won considerably more than half of the interactions (control:
T = 1, Ndiff = 8, p =.02; owl treatment:
T = 0, Ndiff = 8, p =.008), indicating
that the dominance ranks had remained stable during the study period.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
It seems clear that our experiment succeeded in manipulating the perceived level of predation risk because the birds responded to the owl treatment by temporarily adjusting their vigilance behavior (Figures 1 and 2). To a large extent the birds kept pursuing the same kinds of activities in the predation treatment as in the control, although these activities were more frequently interrupted by scanning behavior in the predation treatment. A higher rate of scanning can, at least to some degree, act as a behavioral compensation for increased predation risk (Lima and Dill, 1990
). Perhaps as a consequence of such behavioral adjustment, the
total amount of time spent eating at the feeder was similar in both
treatments. We also failed to find significant differences between treatments
in the total amount of aggression (Figures
3a, b). Instead, the birds changed the time structure of their
aggressive behavior and avoided long, uninterrupted sequences of displays
(Figure 3c). Nevertheless, it would be wrong to conclude that the birds simply compensated for the increased predation risk through a higher rate of scanning. The amount of aggression by subordinates decreased rather sharply in the owl treatment (Figure 4a, b). No similar drop in aggression by the dominant bird was observed, which ought to be the reason we did not find significant differences between treatments in the overall amount of aggression. However, from our present data we cannot rule out an effect on dominant behavior.
The reduction in subordinate aggression was partly a result of a tendency toward a lower proportion of bouts between two subordinates in the owl treatment and thus a higher degree of participation by the dominant (Figure 5a), and partly a result of a lower rate of displaying by the subordinates in the owl treatment (Figure 5c). As could be expected from their lower rate of aggression, subordinates did not win more bouts against the dominant bird in the owl treatment, but rather tended to lose more.
Considering then the prediction that conflicts should be settled with fewer or less intense display behaviors in conditions of high predation risk, we can conclude that subordinates responded in this manner, whereas the dominants appeared not to, at least not for the level of risk induced by our owl treatment. Thus, it seems that dominant and subordinate great tits might differ in how they respond to predation risk. From the idea that dominants will avoid situations of high predation risk, one would have expected subordinates to be more successful under high risk, but we did not find this effect. If anything, the subordinates appear to be less successful under high risk.
The foraging behavior of dominant and subordinate wintering birds in
relation to predation risk has been investigated several times (e.g.,
Ekman, 1987;
Ekman and Askenmo, 1984;
Hogstad, 1988;
Koivula et al., 1994;
Krams, 1996;
Lahti, 1998;
Lahti et al., 1997;
Slotow and Rothstein, 1995).
Most of these studies show that subordinates tend to forage in riskier places.
An exception is the study by Slotow and Paxinos
(1997), which showed that
dominant white-crowned sparrows preferred to forage from a high-quality feeder
placed away from cover, forcing subordinates to use a low-quality feeder
closer to cover, where presumably the perceived predation risk was lower.
Thus, the commonly observed pattern of subordinate risk-taking could well be a
consequence of dominants having acceptable opportunities to forage in safety.
But if risk taking would be needed for high-quality foraging, one might
instead expect dominants to preferentially take these risks. The situation in
our experiment, where a high-quality food source appeared for a brief period
of time, might have been such that the dominant bird considered it worthwhile
to attempt to exclude the subordinate birds from this resource also under
increased risk of predation. In such a situation, it is perhaps not so
surprising that subordinates reduced their rate of displaying under high risk
because a subordinate would have to take the double risk of predation and
attacks by the dominant for a small chance of winning the interaction. Our
setup might reflect a natural situation rather well. In winter, the great tit
can be regarded as an opportunist when it comes to foraging; the flock members
roam widely during the day looking for any kind of food source, including
carrion. Thus, to claim such suddenly appearing high-value resources could be
an important benefit of dominance in the great tit.
The general issue of how aggressive behavior might be influenced by
predation risk has been dealt with by Jakobsson et al.
(1995) and by Brick
(1998). These authors argued
that escalated forms of fighting behavior could be more detrimental to
vigilance than ritualized displays and verified this idea for contests between
cichlid fish (Brick, 1998;
Jakobsson et al., 1995).
Consistent with these effects, Brick
(1998) showed that the
structure of conflicts was influenced by predation risk, with more
low-intensity display and fewer and shorter bouts of escalated fighting under
high risk.
An important difference between the great tit interactions in our
experiment and the cichlid fish contests studied by Brick
(1998,
1999) is that the protagonists
in our study were well acquainted and had already developed a dominance
relationship. Aggressive behavior in such a situation ought to be indicative
of an individual's level of motivation or dominance status, rather than
achieving an assessment of fighting ability. This difference could have a
number of consequences. First, it may well be possible for the great tits to
keep up a rather high level of vigilance during brief threat interactions,
which they would have been unable to do during prolonged bouts of grappling
(cf. Jakobsson et al., 1995).
Second, it is not certain that more escalated behavior, in the form of quick
attacks, should be relatively more dangerous than displays under high risk of
predation. We found that subordinates reduced their rates of displaying under
high risk, but they did not reduce the rate of attacks, which might be because
displaying was perceived as relatively more risky. Third, one might expect
that a higher cost of interacting should cause a highly motivated individual
to more quickly defeat a less motivated opponent, so that displaying in effect
becomes more efficient when the cost of interacting is higher. Predation risk
could be a component of such a cost, both because of reduced vigilance of
contestants and because displaying could attract the attention of predators.
The cautious behavior by subordinates observed in our owl treatment could
reflect an increased effectiveness of the dominant's aggression.
More generally, if food sources are unpredictable and ephemeral, a dominant cannot afford to always play safe, but has sometimes to act in order to gain priority of access, even under increased risk of predation. A subordinate, on the other hand, would gain little by risking an extended bout of aggressive display with small chances of winning. In such cases, it could be that dominants are characterized by greater boldness and shorter latency to agonistic behavior, which appears to be corroborated by our observations.
|
ACKNOWLEDGEMENTS |
|---|
We thank Sven Jakobsson for fruitful discussions about great tit social behavior and Olle Brick for comments on the manuscript. We are also grateful to AnnaKarin Lange and Cilla Kullberg for bird-catching and Anders Bylin and Nils Andbjer for animal care. This study was supported by grants from the Swedish Natural Science Research Council (to O.L.)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Andersson M, 1980. Why are there so many threat displays? J Theor Biol 86: 773-781.[Medline]
Blurton Jones NG, 1968. Observations and experiments on causation of threat displays of the great tit (Parus major). Anim Behav Monogr 1: 75-158.
Brick O, 1998. Fighting behaviour, vigilance and predation risk in the cichlid fish Nannacara anomala. Anim Behav 56: 309-317.[Web of Science][Medline]
Brick O, 1999. A test of the sequential assessment
game: the effect of increased cost of sampling. Behav Ecol
10: 726-732.
Caryl PG, 1979. Communication by agonistic displays: what can games theory contribute to ethology? Behaviour 68: 136-169.
Chase ID, 1980. Social process and hierarchy formation in small groups: a comparative perspective. Am Sociol Rev 45: 905-924.
Ekman J, 1987. Exposure and time use in willow tit flocks: the cost of subordination. Anim Behav 35: 445-452.
Ekman J, Askenmo C, 1984. Social rank and habitat use in willow tit groups. Anim Behav 32: 508-514.
Enquist M, 1985. Communication during aggressive interactions with particular reference to variation in choice of behaviour. Anim Behav 33: 1152-1161.
Hinde RA, 1981. Animal signals: ethological and games-theory approaches are not incompatible. Anim Behav 29: 535-542.
Hogstad O, 1988. Rank-related resource access in winter flocks of willow tit Parus montanus. Orn Scand 19: 169-174.
Jakobsson S, Brick O, Kullberg C, 1995. Escalated fighting behavior incurs increased predation risk. Anim Behav 49: 235-239.
Koivula K, Lahti K, Rytkonen S, Orell M, 1994. Do subordinates expose themselves to predation—field experiments on feeding site selection by willow tits. J Avian Biol 25: 178-183.
Krams IA, 1996. Predation risk and shifts of foraging sites in mixed willow and crested tit flocks. J Avian Biol 27: 153-156.
Lahti K, 1998. Social dominance and survival in flocking passerine birds: A review with an emphasis on the willow tit Parus montanus. Ornis Fenn 75: 1-17.
Lahti K, Koivula K, Orell M, 1997. Dominance, daily activity and winter survival in willow tits: detrimental cost of long working hours? Behaviour 134: 921-939.
Lima SL, Dill LM, 1990. Behavioural decisions made under the risk of predation: a review and prospectus. Can J Zool 68: 619-640.
Maynard Smith J, 1979. Game theory and the evolution of behaviour. Proc R Soc Lond B 205: 475-488.[Medline]
Popp JW, 1987. Risk and effectiveness in the use of agonistic displays by American goldfinches. Behaviour 102: 141-156.
Senar JC, 1990. Agonistic communication in social species: what is communicated? Behaviour 112: 3-4.
Slotow R, 1996. Aggression in white-crowned sparrows: effects of distance from cover and group size. Condor 98: 245-252.
Slotow R, Paxinos E, 1997. Intraspecific competition influences food return-predation risk trade-off by white-crowned sparrows. Condor 99: 642-650.
Slotow R, Rothstein SI, 1995. Importance of dominance status and distance from cover to foraging white-crowned sparrows: an experimental analysis. Auk 112: 107-117.
Svensson L. 1994 Identification guide to European passerines. Stockholm: Lars Svensson.
Verbeek MEM, De Goede P, Drent PJ, Wiepkema PR, 1999. Individual behavioural characteristics and dominance in aviary groups of great tits. Behaviour 136: 23-48.
Verbeek MEM, Drent PJ, Wiepkema PR, 1994. Consistent individual differences in early exploratory behavior of male great tits. Anim Behav 48: 1113-1121.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Lange and O. Leimar Social stability and daily body mass gain in great tits Behav. Ecol., July 1, 2004; 15(4): 549 - 554. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||