Behavioral Ecology Vol. 12 No. 4: 475-481
© 2001 International Society for Behavioral Ecology
Dominance and feeding interference in small groups of blackbirds
a Ornithology Group, Graham Kerr Building, Institute of Biomedical and Life Sciences, Glasgow University, Glasgow G12 8QQ, UK b Edward Grey Institute, Zoology Department, Oxford University, South Parks Road, Oxford OX1 3PS, UK
Address correspondence to G. Ruxton. E-mail: g.ruxton{at}bio.gla.ac.uk .
Received 10 September 1999; revised 27 April 2000; accepted 5 November 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Dominance and/or interference parameters play a pivotal role in most ideal free distribution models, but there has been scant empirical study of the exact manner in which they jointly operate. We investigate how foraging effort and success varied amongst individuals of different dominance rankings in groups of 1-3 wild blackbirds (Turdus merula) attracted to patches of hidden food. Foraging effort (number of feeding movements per unit time), as opposed to vigilance tradeoffs, was greater when an individual fed with a subordinate conspecific than when it fed alone, but tended to be less when it fed with a dominant individual. Within dyads, changes in foraging effort were associated with the direction of the dominance relationship, but not the relative difference in dominance rank between the two individuals. Similarly, amongst threesomes, top-ranked birds (but not the lowest-ranked individual) showed higher foraging effort compared to when foraging alone. Top-ranked birds also profited from a greater increase in foraging success (food items per unit effort) than bottom-ranked birds when feeding in threesomes than when feeding alone. Dominant birds showed increased foraging success, but not effort, after displacing a subordinate. Our results suggest that an individual's foraging effort is determined by the interplay of group vigilance benefits and interference costs, the latter being more expensive for subordinate individuals. The foraging success of dominant birds may further increase if they use subordinates as food-finders. We discuss the implications of our findings for interference parameters in current Ideal Free Distribution models.
Key words: blackbird, dominance, foraging, group size, Ideal Free Distribution, interference.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A key theoretical tool in many branches of behavioral and conservation biology is the Ideal Free Distribution and its many elaborations (see Tregenza, 1995
for an
overview). This describes the distribution of individuals across a range of
habitats in situations where the success rate of one individual in an
environment is affected by the number (and possibly identity) of those sharing
the habitat with it. Dominance relations are common in the natural world and
have been included in several such models (e.g.,
Houston and Lang, 1998;
Parker and Sutherland, 1986;
Stillman et al., 1997,
Sutherland and Parker, 1985;
van der Meer, 1997). These
models make different assumptions about how the foraging performance of an
individual will be based on its own competitive ability relative to the
competitive abilities of its competitors, and hence how prey encounter rates
will vary with group size and composition (reviewed by
van der Meer, 1997;
Weber, 1998). For example,
Sutherland and Parker's (1985)
model incorporates an individual's competitive ability by comparing it to the
average competitive abilities of individuals in a patch. An
individual's relative competitive ability is calculated as its competitive
ability divided by the average competitive abilities of all animals in the
patch. The prey encounter rate E of individual j is
proportional to
 |
The likelihood of such an outcome has been challenged in recent work, and
has sparked several alternative methods of modeling competitive differences
(van der Meer, 1997). The main
difference between these models and that of Sutherland and Parker
(1985) is that an animal's
feeding rate Ej is calculated not in terms of the number
of competitors present in the patch, but the number of `effective competitor
units' present in the patch. Ej is proportional to
 |
) describes a number of
general models which make assumptions about how the dominance relationship
between individuals j and k influence
vjk. They illustrate examples where
vjk is dependent on the ratio between the competitive
abilities of j and k, or has unit value only if k
is dominant over j. However, Sutherland and Parker
(1998) suggest that if such
general rules cannot be supported, it may be necessary to establish
empirically the whole matrix of vjk values for all
individuals or phenotypes visiting patches.
So far, the mechanism by which dominance affects interference and hence
foraging rate has received only very limited comparisons with empirical data
(Ruxton, 1999;
Sutherland and Parker, 1998;
Weber, 1998). Hence, here we
will use a wild population of European blackbirds Turdus merula
feeding on seminatural resource patches to explore the relationship between
dominance and feeding rates. We present data from small feeding groups where
the presence of others (social foraging) may allow an individual to spend a
greater fraction of its time foraging, because of the antipredatory advantages
of risk dilution (Hamilton,
1971) or collective detection (see
Elgar, 1989 for review;
Lima, 1990). However, this is
a particularly appropriate study species because interference affects on
feeding rates are known to reduce the benefits of social foraging, even in
small groups where dominance effects are likely to be most marked (Cresswell,
1997,
1998a,
b).
The rate of food finding considered in these models is a combination of the
intensity of foraging activity and the success rate per unit effort. It seems
quite possible that the presence of other birds could have differential
effects on these two processes. For example, risk dilution and collective
predator detection may allow an individual to spend a greater fraction of its
time foraging, but its success rate per unit effort might decline because its
use of the full extent of a resource patch may be limited by interference with
conspecifics (e.g., Cresswell,
1997). The potential costs and benefits that each individual
accrues from group foraging will vary, however, because individuals will
differ in foraging ability, susceptibility to interference and dominance.
Individuals that have high absolute foraging abilities (foraging rate when
alone) but that are highly susceptible to interference competition will have
much lower feeding rates in groups. Any benefits of increased time available
for feeding because of vigilance/dilution effects may not be sufficient to
outweigh any interference-mediated losses of feeding time. In contrast, a
dominant individual may suffer little interference competition and may
actually increase its success rate in a group because of its ability to
displace or kleptoparasitize subordinates, as well as gaining extra feeding
time from the association due to dilution/vigilance trade-offs (e.g.,
Ens and Goss-Custard, 1984;
Goss-Custard and Durell,
1988). Most previous empirical and theoretical works have simply
subsumed foraging intensity and success rate per unit effort into overall
intake rate. However, in view of the argument above, we will treat these
processes separately, investigating the effects of group size changes for
birds of contrasting dominance in relation to both foraging effort and
success.
Last, the effects of "public information" on foraging success
have attracted recent theoretical (e.g.,
Beauchamp et al., 1997;
Valone, 1993), and empirical
study (Livoreil and Giraldeau,
1997; Smith et al.,
1999; Templeton,
1998; Templeton and Giraldeau,
1995,
1996). The suggestion here is
that individuals can use information on the position and behavior of other
individuals in order to improve their own intake rate. While intuitively
appealing, the importance of this effect has not previously been demonstrated
in a situation which closely mimics natural foraging conditions. Hence, as our
study fits this criterion, we will also examine the effects (if any) of this
on the foraging success of our blackbirds.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We studied the feeding behavior of a population of wintering blackbirds in the grounds of Hopetoun House, near Edinburgh between 23 January and 6 February 1998. This population was individually color-marked and had been the subject of previous foraging research in the previous three winters (Cresswell, 1997
,
1998a,
b). Birds were offered patches
of shredded suet covered in a layer of autumn leaves. The locations of our
experimental suet patches, as opposed to natural leaf litter, were made
distinctive by clearing the surrounding area of fallen leaves (up to
approximately 5 m), and inserting white pegs into the ground at the corners of
the patches. After a short period of training, birds appeared to
preferentially recognize these patches, and tossed the leaves aside to uncover
the suet. Both our subjects, and other blackbird populations
(Cramp, 1988), have been noted
uncovering items from below natural leaf litter in the same way. We spread suet shreds uniformly throughout each 1 x 2 m rectangular patch, at two different densities, either 150 shreds or 30. Five sites were used in an area of approximately 1 ha of parkland, comprising areas of mixed woodland, mown lawn, and dense mature bushes. Although the sites were only 30-110 m apart they each consistently attracted different individual blackbirds, perhaps dependent on each bird's range. Our five sites were chosen from a larger number of training sites because they suffered only low levels of overt aggression from territorial resident blackbirds, and consequently tended to attract larger numbers of birds per presentation. Our blackbirds were therefore "free" to visit these neutral sites, and were only occasionally driven away by despotic behavior. As well as the territorial residents, the patches attracted neighboring territory holders and nonterritorial individuals and was typical of blackbird populations in other winters and other parts of our site. Forty individually identifiable birds fed at the patches during the experiment, 35 of these feeding at only 1-2 patch sites, and the remaining five birds fed at three. Each patch attracted up to 12 feeders during a given trial, and between eight and 19 birds over the entire set of experimental trials.
Individual blackbirds would typically take between 10 and 30 pieces of suet before retiring to cover, and up to seven birds were noted foraging simultaneously, although 1-4 birds were most frequent. Although the number of foragers was usually highest early in the trial, there were often breaks in feeding activity and subsequent arrival of "new" individuals. We continued the trial until feeding activity had stopped, most of the expected resident feeders had fed, or the feeding rates of the remaining feeders suggested that the densities of suet remaining were much reduced. Presentations lasted between 13 and 69 min (mean 35.1 min ± 12.6 SD). Few individuals fed more than once during a given trial. At least 1 h passed before we carried out a succeeding trial at the same patch location. Up to four trials were conducted at the same site each day, and up to nine trials at all sites in a day.
All trials were filmed on a portable camcorder, allowing us to ascertain which blackbirds fed simultaneously, and when they found suet. Although it was not uncommon for birds to curtail feeding movements for several seconds or more while visiting a patch, these interruptions were usually associated with periods of vigilant head-raising or conspecific aggression rather than comfort behaviors. Before and after feeding, blackbirds moved swiftly between our patches and protective cover, and rarely moved outside the patches unless competition at the patches was high. This suggests that our birds felt vulnerable in open feeding situations, and that the sole purpose of visiting our patches was to feed. We therefore calculated feeding rates throughout the entire time that a bird was present in a leaf patch. However, we obtained similar results in an analysis that omitted periods when birds did not feed for 5 s or more.
Our birds fed using a series of similar head movements into the leaf litter, which functioned to either toss leaves aside, or peck at (i.e., consume) exposed suet. Although a bird's head and bill were sometimes obscured, preventing us from seeing suet or directly counting swallowing motions, we were still able to distinguish pecks from tosses by the lack of leaf movement, and the slightly shorter duration of pecks compared to leaf tosses (averages of 0.9 s versus 1.4 s). Suet was swallowed as the head was lifted, so handling time was considered to be negligible. The accuracy of our observed number of suet pecks was confirmed by counting the number of suet shreds remaining after a patch presentation was completed.
Pecks and tosses were combined as total head movements when calculating feeding rates. Our patches were small and prone to patchy depletion as birds fed, and therefore feeds were not always long enough to allow proper compensation of periods of good and bad luck when searching for suet. Therefore, to avoid including the added variability of whether or not birds actually find suet while foraging, our main measure of foraging rate was foraging effort, that is, the number of head movements per minute (HM min-1) that the bird was present in a patch. However, we also calculated foraging success (the number of suet pecks per hundred peck/toss head movements; P HM-1 x 100). Although intake rates are the traditional currency of foraging models, intake rate is actually the product of foraging effort and foraging success. Our analyses are able to examine how these separate feeding components may be modified when feeding in different social situations. We calculated feeding rates for birds which spent at least 10 s feeding at constant flock size, resulting in a total of 1621 time budgets, lasting between 11 and 461 s (median = 38 s).
Inevitably, foraging success increased with suet density. To minimize the variance in suet peck rates associated with pecking at low suet densities, we confined our analysis of foraging success to time budgets starting when at least 20% of the initial suet density was still present. Foraging success was linearly related to the density of suet remaining in a patch at the start of a time budget (mean of 8.3 P HM-1 x 100 when 30 suet shreds present, 30.6 P HM-1 x 100 when 150 shreds present, r1444 =.73, p <.001, quadratic term not significant). As suet handling times were short, and feeding pecks closely resembled leaf tosses, the relationship between foraging effort (pecks plus tosses per minute) and suet density was less predictable. The mean foraging effort across all time budgets was 44.9 ± 10.1 SD HM min-1, decreasing slightly from 48.1 HM min-1 in patches where 150 suet shreds were initially present, to 43.8 HM min-1 when there were 30 suet shreds initially (r1619 =.18, p <.001). Therefore, as suet density affected both feeding measures, we calculated the values for foraging effort and foraging success after controlling for suet density, and used these residual values throughout this article.
As expected, despite controlling for the effects of suet density, variation in foraging success was still much larger than variation in foraging effort (standard deviation of each residual measure expressed as a percentage of the original mean: foraging success 54%, foraging effort 22%). This was partly because feeds were short in relation to the time taken to find successive suet shreds, and partly because successive suet pecks were likely to occur in the same part of the patch as an individual foraged, causing suet distribution within the patch to become uneven. Only the first blackbird to arrive in a new patch would encounter a relatively even suet spread. The foraging success of its successors would depend on how long they were able to spend in relatively undepleted parts of the patch. Our measure of feeding effort does not suffer this additional stochasticity, so we were more likely to detect differences in feeding effort within individuals in different social situations. An additional problem with foraging success, but not foraging effort, was that its variance increased (at a declining rate) with suet density (correlations of absolute magnitude of feeding residual versus suet density: foraging effort, r1619 = -.01, p =.75; foraging success, r1444 =.34, p <.001). We therefore stabilized the variance of foraging success in relation to suet density by dividing it by the square root of suet density, and used transformed residuals for foraging success throughout this article.
All other things being equal, increases in either foraging effort or foraging success will be reflected by increases in intake rate. Our measure of foraging effort was least affected by sampling limitations. Moreover, it is likely that animals searching for hidden prey will have more control over their foraging effort than their ultimate success, so it is this variable which will suffer directly from increased conspecific or predator vigilance. Consequently, we have concentrated on foraging effort throughout much of this article.
We also recorded rates and outcomes of aggression in each patch, allowing us to establish individual dominance. On average, birds were involved in 2.5 ± 1.5 SD (n = 35 individuals) aggressive encounters per competitor minute. Most encounters were simple noncontact displacements. Most dominance relationships were unambiguous and linear, permitting calculation of dominance hierarchies. Individuals in triangular relationships (where A succeeded against B and B beat C, yet C won encounters with A) were given equivalent rankings unless their relative success with other individuals indicated that they merited higher or lower ranking. A bird's dominance in a patch hierarchy was then calculated as the number of other individuals which it ranked higher than in a patch divided by the number of birds which it either ranked higher or lower than (i.e., codominants were excluded from the ratio). Unless otherwise stated, the influence of dominance rank and group size on feeding rates was examined by comparing mean feeding rates of individuals in contrasting situations (e.g., as dominants versus as subordinate in dyads) using matched pairs t tests. All probabilities quoted are two-tailed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We recorded the outcome of a total of 842 encounters between 39 individually identified blackbirds. Our within-patch dominance rankings correctly classified the outcome of 89% of encounters. A further 6% of encounters were between birds which were given codominant status, and 42 (5%) were reversals. For 24 of these reversals, there were a similar or larger number of victories in the dominant birds favor, indicating that the assigned direction of dominance was correct. Dominance was unrelated to age (t test: adult versus first winter birds, t24 = 1.24, p =.23) or sex (t test: t30 = 0.93, p =.36), but, within individuals, was similar in different patches (r11 =.48, p =.097).
Feeding rate versus dominance in dyads
The residual feeding rates of each bird were compared when they fed
alongside a dominant competitor as opposed to a subordinate competitor.
Foraging effort, but not foraging success, was higher when feeding with a more
subordinate individual than with a more dominant individual
(Table 1).
|
However, does the magnitude of the difference in dominance between two
competitors further affect foraging effort? Subtracting the dominance of the
competitor from the dominance of the focal bird gave a dominance difference
score between -1 (when the competitor dominated all birds in the patch and the
focal bird was entirely subordinate) to 1 (focal bird dominant, competitor
entirely subordinate). The dominance difference score was not constant for
each focal individual, but varied according to the dominance of its paired
competitor. Hence, the individual which was most subordinate in a patch could
have dominance difference scores between -1 and 0, depending on the
competitive ability of its paired feeding partner, while the top dominant bird
could have dominance difference scores between 0 and 1. Intermediate dominance
birds had intermediate scores, with a maximum range of 1. We looked at the
effect of dominance difference score on residual foraging effort in a general
linear model (SPSS Inc, 1997)
which also controlled for the focal bird's identity. We selected individuals
for which there were at least five time budgets. There was no evidence for
individual variation in the manner in which residual foraging effort was
influenced by dominance difference score (interaction between bird identity
and score, F23,381 = 1.18, p =.26). After
removing this term from our model, both bird identity
(F23,404 = 2.96, p <.001) and dominance
difference score (F1,404 = 6.07, p =.014)
explained significant variation in overall foraging effort.
However, such a result could also be obtained if foraging effort increased when feeding with a subordinate and declined when feeding with a dominant, but is not influenced by the magnitude of the feeding partner's dominance or subordinance. Hence, we repeated our analysis, splitting our data into whether the focal bird was dominant or subordinate. Again, within dominance classes, there was no evidence for individual variation in the relationship between residual foraging effort and dominance difference score (interaction term for subordinate focal birds, F10,148 = 1.23, p =.28; for dominants, F11,148 = 1.44, p =.16). After removing this term, there was still significant individual variation in residual foraging effort (for subordinate birds, F10,158 = 2.98, p =.002; for dominants, F11,159 = 2.23, p =.015), but this was unaffected by dominance difference score (subordinates, F1,158 = 0.13, p =.72; for dominants, F1,159 =.46, p =.50). This indicates that, after determining which bird was dominant, the magnitude of the difference in dominance between two individuals explained little additional variation in foraging effort.
Feeding rates in dyads compared to solitaires
The residual feeding rates of focal birds feeding with one other bird were
then compared to the same bird's feeding rate when feeding alone. Interference
could cause solitary feeders to suffer reduced feeding rates when joined by a
competitor, while group foraging benefits could enhance their feeding
rates.
Overall, restricting analysis to birds which were time budgeted at least three times both when alone and with one other bird, foraging effort was higher when with one other bird than when alone (paired t test, difference = 1.9 HM min-1, t25 = 2.30, p =.030). However, foraging effort was not independent of a competitor's dominance status. Birds feeding with a more subordinate competitor still had a higher foraging effort than when they fed alone, but birds feeding with a more dominant competitor tended to have lower foraging effort than when feeding alone (Figure 1). We did not detect significant differences in foraging success of birds feeding with either a dominant or subordinate competitor compared to when they fed alone (Figure 1).
|
How do feeding rates change when a third bird is present?
Three or more time budgets were obtained from nine individuals when
solitary, when in groups of two and when in groups of three. Although average
residual feeding effort of these birds were slightly higher in groups of three
than when solitary (paired t test, difference = 3.4 HM
min-1, t8 = 1.50) or when in pairs (paired
t test, difference = 2.0 HM min-1, t8
= 1.52), the differences, although exceeding those in the previous section,
did not reach significance (both p =.17).
Again we compared foraging effort of birds in groups of three with the residuals of the same birds when solitary in relation to dominance rank. Top dominants had higher effort when feeding in groups of three than when feeding alone, but bottom subordinates had slightly but not significantly lower effort when foraging in groups of three (Figure 1). Repeating this analysis for foraging success gave similar but subtly different results (Figure 1). Residual foraging success was slightly but not significantly greater in groups of three than when solitary in dominants, but subordinates had substantially lower foraging success when feeding in tryads (Figure 1). In contrast to the results for foraging effort, the change in foraging success of middle-rank birds was slightly lower than that of dominants.
To look at whether interference affects were multiplied or diluted on the addition of birds of different status, we compared the residual feeding rates of a bird when in a group of two with those when it was with the same competitor plus one other. One hundred time budgets were available where the feeding rate of a bird in a group of three could be compared with its feeding rate with the other individuals when in groups of two; the majority (84%) of these budgets had unique combinations of individuals undertaking the three roles (focal bird, second competitor, and third joiner). As the number of potential comparisons was large (each of three ranks joined by each of the other two, i.e., six), and as we considered some birds to have the same dominance rank (effectively making 12 comparisons, thereby reducing sample sizes), we analyzed the data using four different groupings (Table 2).
|
The foraging effort of any dominant (i.e., the top dominant or middle-rank individual) with a subordinate was higher when another dominant joined the system (Table 2), possibly due to the shared advantage of feeding in groups. Such shared increases in residual effort were not obvious in the other three groupings. As with residual effort, the foraging success of any dominant with a subordinate, was significantly higher if another dominant was present. In contrast to foraging effort, the foraging success of two other groupings showed significant changes when a third bird was present. First, top dominants had higher foraging success when any other subordinate was present. Second, and somewhat complementarily (though based on an independent set of time budgets), bottom subordinates had lower foraging success when any other dominant was present.
Such a difference in success, but not effort, might be expected if dominants displaced subordinates from less depleted areas within our patches. We tested this in birds that displaced a competitor during a time budget, by comparing their feeding rates before and after the displacement. We selected data where at least 10 s were spent in the patch before and after the displacement, and, for birds which made more than one displacement during a time budget, selected only the first such displacement in the time budget. Foraging effort was similar before and after displacements (paired t test, mean increase of 1.5 HMmin-1, t28 = 0.49, p =.63), indicating that dominants did not change effort following an interaction. However, the foraging success of dominants increased after they had displaced a subordinate (mean increase of 6.0 P HM-1 x 100, t28 = 2.48, p =.019), despite the slightly reduced suet densities, supporting the idea that dominants use subordinates as food finders.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
It is clear that there was a dominance hierarchy in our birds, birds could identify whether a foraging partner was higher or lower in the hierarchy than them, and that birds modified their behavior in the light of this information. It is perhaps unsurprising that our birds were able to do this as blackbirds show age, sex, and individual variation in plumage (Cramp, 1988
) and they interact
with a relatively small number of other blackbirds repeatedly throughout the
winter, affording them plenty of opportunity to assess the characteristics of
others. This information has a strong influence on their behavior: birds feed
at a greater rate with subordinate individuals than they would do if alone,
but, when in the presence of an individual that is dominant to them, feed at a
lower rate than when alone. The first result could occur in two ways. First,
group vigilance may allow dominants to increase foraging effort without a
corresponding increase in predation risk. Blackbirds in the study area are
attacked by both avian (sparrowhawk Accipiter nisus) and terrestrial
(Mustelid) predators. Our artificial food patches required them to feed away
from cover, in situations where they were likely to perceive an increased
threat of predation (Bednekoff,
1996). There are many observational studies of reduced vigilance
rates in passerines as group size increases
(Elgar, 1989). This reduced
vigilance requirement could therefore liberate time for increased foraging
effort. Second, dominant birds may be able to use subordinates as food-finders
(Rohwer and Ewald, 1981;
Wiley, 1991), taking advantage
of information gleaned from subordinates to forage in parts of the patch where
depletion is least.
It is likely that some antipredatory benefits of foraging with other
individuals holds for all birds regardless of their place in the dominance
hierarchy. It might therefore seem surprising that foraging effort decreases
for subordinate birds in the presence of a dominant individual. There is
clearly an interference cost paid by the subordinate which supercedes any
antipredatory vigilance advantages which the dominant might bring. Our
understanding of how dominance relationships might affect feeding rates
through interference has largely depended on studies of oystercatchers
Haematopus ostralegus feeding on mussels Mytilus edulis and
cockles Cerastoderma edule. When mussel-feeding, dominant
oystercatchers may increase their feeding rate with increasing competitor
density because kleptoparasitism opportunities increase
(Ens and Goss-Custard, 1984;
Goss-Custard and Durell, 1988;
Stillman et al., 1996).
However, dominant oystercatchers did not have higher feeding rates than
subordinates when feeding on cockles at lower competitor densities
(Sutherland and Koene, 1982),
perhaps because subordinates were able to make behavioral adjustments to
reduce the probability of kleptoparasitism without reducing their intake rate
greatly (Norris and Johnstone,
1998). Interference is, however, found in cockle-feeding
oystercatchers at higher densities, especially when cockles are rare
(Triplet et al., 1999). In
comparison to these studies, group-feeding subordinate blackbirds showed
reduced foraging effort compared to dominants, but were never seen to be
kleptoparasitized. Therefore, the cost of interference in blackbirds may
largely result from a need to monitor the aggressive behavior of more dominant
individuals (see Cresswell,
1997). Our data show that dominance is a very good predictor of
the outcome of aggressive interactions and that such interactions occur
frequently. Hence, in contrast to dominant individuals, subordinates may have
to reduce foraging effort in order to be vigilant for possible attack by
dominants.
However, whatever the mechanisms involved, our results show that the
presence of other conspecifics at a feeding site does affect foraging effort,
and that the effect on an individual will depend both on its own place in the
dominance hierarchy and those of its competitors. As with oystercatchers, this
gives important empirical support to ideal free distribution models which
allow for individual differences in competitive ability, such as Sutherland
and Parker (1985) and van der
Meer (1997). However, an
important caveat to this is that in the simple situation when birds fed in
pairs, there was no evidence that an individual responded to the
magnitude of the difference in dominance between it and a competitor
(only on whether they were higher or lower in the hierarchy). This stands in
direct contrast to almost all the commonly used models of the effect of
interference in systems where individuals differ in dominance (but see
Stillman et al., 1997; linear
hierarchy model of van der Meer,
1997). This may be a real effect. Alternatively, additional costs
involved with feeding with a highly dominant competitor may increase
comparatively slowly in comparison to the initial cost of feeding with a
dominant competitor of any magnitude. As this relationship is key to a great
deal of the existing literature on the effects of interference on foraging
decisions, there is a clear need for further experimental and theoretical
effort focused on this assumption.
When two foraging birds are joined by a third, the affects on foraging are
complicated. We might predict foraging effort should increase due to the
antipredatory benefits of collective vigilance. However, mean food intake rate
commonly increases at a reducing rate with group size, and sometimes drops in
larger groups (Beauchamp,
1998). Our results were mixed. An original subordinate individual
did not increase its effort, regardless of the newcomer's place in the
hierarchy. However, there was some evidence that original dominant birds did
increase their foraging effort when a newcomer arrived, although the effect
was not universal (Table 2).
Foraging success followed a more intuitively appealing pattern. If the
newcomer was subordinate, the dominant bird's foraging success (but not that
of the original subordinate) increased. In contrast, if the newcomer was a
dominant bird, then the success rate of the original subordinate, but not the
original dominant, decreased. This suggests that dominant individuals benefit
from the presence of subordinates, perhaps by supplanting them from good
foraging areas within the patch (as found, e.g., in several seed-eating
passerines, Baker et al., 1981;
Rohwer and Ewald, 1981;
Theimer, 1987;
Wiley, 1991). To examine this,
we looked at a dominant bird's foraging pattern before and after supplanting a
subordinate. There was no detectable change in foraging effort after
displacing a subordinate, but an increase in foraging success. This suggests
that dominants do obtain useful public information (sensu
Valone, 1989) to improve their
ability to exploit a food resource.
Recently, there has been considerable interest
(Ruxton, 1999;
Sutherland and Parker, 1998)
in comparing experimental data to the underlying assumptions of the models of
Sutherland and Parker (1985)
and van der Meer (1997). The
latter make the prediction that an individual's return from foraging will be
reduced as more competitors arrive, dependent on the number of effective
competitor units they contribute to its interference. This contrasts with
Sutherland and Parker's (1985)
model which predicts that although an individual's foraging rate will usually
decrease as more competitors arrive, there will be times when the feeding
rates of the original foragers will increase as more subordinates arrive. Such
circumstances arise when the difference between the competitive abilities of
dominant and subordinate birds is high, and the increase in an original
forager's relative competitive ability more than compensates for the increased
number of foragers competing for food. Although this outcome was not expected
when the original model was put forward, Sutherland and Parker
(1998) suggest that it is
plausible that additional subordinates could divert dominant animals'
attention from the original subordinates, diluting the frequency of attacks or
the vigilance invested in avoiding attacks. It is not, however, so easy to
explain why the feeding rates of the original dominants might also increase
without invoking benefits arising from group size, such as predation risk
dilution or faster food finding. Our data cannot directly address the question
of what happens when additional competitors are present without first
accounting for the beneficial effects of social foraging. Furthermore, we have
not measured relative competitive ability, only the linear dominance
hierarchy, where increases in dominance between adjacent members of the
hierarchy may not always reflect similar-sized increases in relative
competitive ability. However, our observation that feeding effort depended
more on the direction of dominance, rather than on the magnitude of the
dominance difference, suggests that in blackbirds feeding encounter rate does
not increase with an individual's relative competitive ability (as modeled by
Sutherland and Parker, 1985).
Instead, encounter rate may depend more on how many competitors are dominant
(as in the linear hierarchy matrix, V3,
Van der Meer, 1997). However,
the high yet variable values of foraging effort and foraging success seen in
birds of intermediate dominance in tryads
(Figure 1) suggests that in
some cases the presence of a subordinate may shelter the
intermediate-dominance individual from interference from the dominant bird.
Indeed, the increase in foraging effort with group size observed in
intermediate birds may exceed that seen in dominant birds, perhaps because
intermediate birds are prepared to take larger risks in profitable foraging
situations. Intermediates may locate themselves in parts of the patch which
take advantage of the position of subordinates to dilute the possibility of
aggression from dominants. If so, it is possible that such opportunities would
disappear in larger groups containing dispersed dominants, so that encounter
rate would again increase with an individual's competitive ability, perhaps in
the fashion employed by Van der Meer's
(1997) V4
interference matrix (interference constant m acts on the sum of the
ratios between competitive abilities of competitors in the patch).
Clearly, average foraging effort did increase as group size increased from one to at least three birds, and may have increased further in large groups. Yet, in our study, larger groups were rare. It is possible that this was driven by the departure of subordinate birds, whose foraging effort declined in groups in comparison to when solitary. Departure of subordinates from tryads may increase the level of interference experienced by intermediate-dominance birds, instigating their ensuing departure (perhaps to join subordinates). Indeed, in more general terms, when or where groups occur, as well as maximum group sizes or densities, may depend on the shape of the cost-benefit curve for subordinates rather than dominants. Reduced feeding rates in groups may be offset against greater predation risk when feeding alone, especially if alternative patches are of lower quality. Nevertheless, it is likely that as group size antipredation benefits are hindered by interference costs, switching to forage solitarily in less rich patches may become a viable option for animals in large groups, especially if the probability of being joined quickly by other animals is high. Such benefits may give rise to small groups of foragers taking advantage of lower predation risk (than when solitary) but low interference costs (than when in a larger group). Invasion of such groups by dominants may give rise to a dynamic distribution of small groups of animals. Such situations could be explored by including group size benefits in future ideal free models.
In summary, our results suggest that foraging on a shared resource is
constrained in a complex fashion by conspecific competitors, dependent on the
relative dominance and aggression of the individuals involved. Further, when
we divided foraging into effort and success we found that both were affected
by dominance relations but not in exactly the same way. Our results therefore
suggest that some assumptions of models which seek to describe the
distribution of a group of individuals of different competitive abilities
between a range of resources in a very simple way (e.g.,
Sutherland and Parker, 1985)
are too simplistic. However, our results also suggest that more complex models
(such as that of van der Meer,
1997 which requires input of matrices of interindividual
interference effects) can be somewhat generalized if intraspecific
dominance-interference patterns are measured. Current models also generally
assume that interference is the main influence of group size on foraging,
whereas we have shown that, at least for our system, the effects of safety in
numbers and public information may also be important. If we are to improve the
predictive ability of these models, then a more detailed description of such
effects will be required. This represents a major challenge to empiricists and
theoreticians, who must be aware of all the difficulties of disentangling a
large range of dominance effects which together act to modify foraging
behavior.
|
ACKNOWLEDGEMENTS |
|---|
We thank Paul Normand, Tom Weaver, the Marquis of Linlithgow, and all at Hopetoun House for practical help. This manuscript was greatly improved by the contributions of three anonymous referees. Our work was carried out under a Natural Environment Research Council (NERC) research grant to Glasgow University.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Baker MC, Belcher CS, Deutsch LC, Sherman GL, Thompson DB, 1981. Foraging success in junco flocks and the effects of social hierarchy. Anim Behav 29: 137-142.
Beauchamp G, 1998. The effect of group size on mean food intake rate in birds. Biol Rev 73: 449-472.
Beauchamp G, Bélisle M, Giraldeau L-A, 1997. Influence of conspecific attraction on the spatial distribution of learning foragers in a patchy habitat. J Anim Ecol 66: 671-682.
Bednekoff PA, 1996. Translating mass dependent flight
performance into predation risk: an extension of Metcalfe, Ure. Proc R
Soc Lond B 263:
887-889.
Cramp S, 1988. Handbook of the birds of Europe, the Middle East and North Africa: the birds of the Western Palearctic, vol. 5. Tyrant flycatchers to thrushes. Oxford: Oxford University Press.
Cresswell W, 1997. Interference competition at low competitor densities in blackbirds Turdus merula. J Anim Ecol 66: 461-471.
Cresswell W, 1998a. Relative competitive ability changes with competitor density: evidence from feeding blackbirds. Anim Behav 56: 1367-1373.[Web of Science][Medline]
Cresswell W, 1998b. Variation in the strength of interference competition with resource density in blackbirds, Turdus merula. Oikos 81: 152-160.[Web of Science]
Elgar MA, 1989. Predator vigilance and group size in mammals and birds. Biol Rev 64: 13-33.[Medline]
Ens BJ, Goss-Custard JD, 1984. Interference among oystercatchers, Haemotopus ostralegus, feeding on mussels, Mytilus edulis, on the Exe estuary. J Anim Ecol 53: 217-231.
Goss-Custard JD, Durell SEA le V dit, 1988. The effect of dominance and feeding method on the intake rates of oystercatchers, Haemotopus ostralegus, feeding on mussels. J Anim Ecol 57: 827-844.
Hamilton WD, 1971. Geometry for the selfish herd. J Theor Biol 31: 295-311.[Web of Science][Medline]
Houston AI, Lang A, 1998. The ideal free distribution with unequal competitors: the effects of modeling methods. Anim Behav 56: 243-251.[Web of Science][Medline]
Lima SL, 1990. The influence of models on the interpretation of vigilance. In: Interpretation and explanation in the study of animal behavior, vol. 2. Explanation, evolution and adaptation (Bekoff M, Jamieson D, eds). Boulder, Colorado: Westview Press; 246-267.
Livoreil B, Giraldeau L-A, 1997. Patch departure decisions by spice finches foraging singly or in groups. Anim Behav 54: 967-977.[Web of Science][Medline]
Norris K, Johnstone I, 1998. Interference competition and the functional response of oystercatchers searching for cockles by touch. Anim Behav 56: 639-650.[Web of Science][Medline]
Parker GA, Sutherland WJ, 1986. Ideal free distributions when individuals differ in competitive ability: phenotype-limited ideal free models. Anim Behav 34: 1222-1242.
Rohwer S, Ewald PW, 1981. The cost of dominance and the advantage of subordination in a badge-signaling system. Evolution 35: 441-454.[Web of Science]
Ruxton GD, 1999. The differing effects of interference on individuals of different rank. Ecol Lett 2: 4-5.
Smith JW, Benkman CW, Coffey K, 1999. The use and
misuse of public information by foraging red crossbills. Behav
Ecol 10:
54-62.
SPSS Inc, 1997. SPSS advanced statistics 7.5. Chicago: SPSS Inc.
Stillman RA, Goss-Custard JD, Caldow RWG, 1997. Modeling interference from basic foraging behaviour. J Anim Ecol 66: 692-703.
Stillman RA, Goss-Custard JD, Clarke RT, Durell SEA le V dit, 1996. Shape of the interference function in a foraging vertebrate. J Anim Ecol 65: 813-824.
Sutherland WJ, Koene P, 1982. Field estimates of the strength of interference between oystercatchers Haematopus ostralegus. Oecologia 55: 108-109.[Web of Science]
Sutherland WJ, Parker GA, 1985. Distribution of unequal competitors. In: Behavioural ecology (Sibly RM, Smith RH, eds) Oxford: Blackwell Scientific Publications; 225-274.
Sutherland WJ, Parker GA., 1998. Interference with ideal-free models. Trends Ecol Evol 13: 410.
Templeton JJ, 1998. Learning from others' mistakes: a paradox revisited. Anim Behav 55: 79-85.[Web of Science][Medline]
Templeton JJ, Giraldeau L-A, 1995. Patch assessment in
foraging flocks of European starlings: evidence for the use of public
information. Behav Ecol 6:
65-72.
Templeton JJ, Giraldeau L-A, 1996. Vicarious sampling: the use of personal and public information by starlings in a simple patchy environment. Behav Ecol Sociobiol 38: 105-114.
Theimer TC, 1987. The effect of seed dispersion on the foraging success of dominant and subordinate dark-eyed juncos, Junco hyemalis. Anim Behav 35: 1883-1890.
Tregenza T, 1995. Building on the Ideal Free Distribution. Adv Ecol Res 26: 253-307.
Triplet P, Stillman RA, Goss-Custard JC, 1999. Prey abundance and the strength of interference in a foraging shorebird. J Anim Ecol 68: 254-265.
Valone TJ, 1989. Group foraging, public information and patch estimation. Oikos 56: 357-363.[Web of Science]
Valone TJ, 1993. Patch information and estimation: a cost of group foraging. Oikos 68: 258-266.[Web of Science]
Van Der Meer J, 1997. The ideal free distribution when predators differ in competitive abilities. Oikos 80: 301-310.[Web of Science]
Weber TP, 1998. News from the realm of the ideal free distribution. Trends Ecol Evol 13: 89-90.
Wiley RH, 1991. Both high- and low-ranking white-throated sparrows find novel locations of food. Auk 108: 8-15.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  I. M. Smallegange and J. van der Meer Interference from a game theoretical perspective: shore crabs suffer most from equal competitors Behav. Ecol., January 1, 2007; 18(1): 215 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. K. Vahl, T. Lok, J. van der Meer, T. Piersma, and F. J. Weissing Spatial clumping of food and social dominance affect interference competition among ruddy turnstones Behav. Ecol., September 1, 2005; 16(5): 834 - 844. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. K. Vahl, J. van der Meer, F. J. Weissing, D. van Dullemen, and T. Piersma The mechanisms of interference competition: two experiments on foraging waders Behav. Ecol., September 1, 2005; 16(5): 845 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. T. Hansen and T. Slagsvold Early learning affects social dominance: interspecifically cross-fostered tits become subdominant Behav. Ecol., March 1, 2004; 15(2): 262 - 268. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||