Selection by passerine birds is anti-apostatic at high prey density

Most of the results from past experiments with wild birds and green and brown pastry 'baits' have suggested that disproportionately more of the rare forms are eaten when bait density is high (i.e. selection is anti-apostatic). In two separate series of experiments we presented birds with dishes containing 270 baits of one colour and 30 of another. In series I, five different pairs of colours were presented simultaneously to wild birds at two sites. One colour of each pair was common at one site and the same colour was rare at the other site. After 35 days the ratios of the colours were reversed and the dishes were presented for another 35 days. There was a statistically significant tendency for the colours to be at a higher risk when rare. In series II, three caged blackbirds were offered green and brown baits in two dishes simultaneously; in one dish green was rare and in the other brown was rare. Selection over 6 days was anti-apostatic for all three birds combined but the data proved heterogeneous both between and within individuals. At any one time, each bird tended to concentrate on one colour, irrespective of whether that colour was rare or common. We believe that this behaviour leads to anti-apostatic selection, as has been observed in these and other experiments with pastry prey.     

The influence of density on frequency-dependent food selection: a comparison of four experiments with wild birds

We compare the results of four experiments, conducted at different times and with different protocols, that explored the relationship between frequency-dependent selection and prey density in wild birds feeding on artificial populations of coloured baits. One (experiment 4) used pastry baits that differed only in the presence or absence of a red stripe, and this experiment provided no evidence for any kind of selective behaviour. The other three experiments used green and brown baits, and they all provided evidence for a trend towards increasing anti-apostatic selection with high densities (>100 baits m^–2). However, one of these (experiment 3) provided no evidence for frequency-dependent selection at low densities (0.5–20 baits m^–2), while the other two experiments concurred in suggesting a trend towards increasing apostatic selection with low densities (down to 2 baits m^–2). Together, these experiments both support and qualify the published findings of experiment 1 that frequency- dependent selection by wild birds on bait populations is modified by density. Experiment 4 indicates that frequency-dependent selection may break down entirely if bait types are too similar, while experiment 3 indicates that some details of this trend with density will depend either on the protocol used or on exogenous changes in the birds’ feeding behaviour. Received: 1 September 1999 / Accepted: 22 March 2000     

Further evidence for apostatic selection by wild passerine birds -9 :1 experiments.

Apostatic selection occurs when predators concentrate disproportionately on the common varieties of non-mimetic polymorphic prey species. This has been tested in 14 experiments by presenting populations of green and brown lard-and-flour "baits" to inexperienced wild passerine birds in their normal surroundings. In seven experiments a 9 green : 1 brown population was presented for a number of days, followed by a 1 green : 9 brown population for a similar period. in the remaining seven experiments the populations were presented in the reverse order. The birds often had strong "natural" colour preferences (for example, blackbirds and songthrushes preferred browns) which were not caused by the relative conspicuousness of the two colours. The data within most of the experiments were very heterogeneous, but in every experiment there was good evidence that the birds tended to concentrate on the common colour. The consistency of the replicated experiements gives strong reason to believe that apostatic selection is a widespread phenomenon among avian predators, and provides an explanation for many of the non-mimetic colour and pattern polymorphisms found among their prey.     

Apostatic selection on prey that match the background

There is good experimental evidence that predators often remove disproportionately more of the common prey types. This 'apostatic selection' could maintain colour polymorphisms within prey species. In nature, morphs of many species appear to match components in the background, but most of the experiments that have tested apostatic selection have used prey that were conspicuous. In work described in this paper, wild birds at one site were presented a random order of 51 experiments. Seventeen frequencies of orange and grey pastry prey were presented on each of three types of background: a hessian sheet scattered with either orange and grey stones (the 'matching' background), or lilac and yellow stones, or green stones (two sorts of 'control' background). Each experiment consisted of four trials in succession and the numbers of the two colours eaten in each trial were recorded when about half the total prey had been eaten. Two methods of graphical analysis revealed that apostatic selection occurred on the prey on all three backgrounds, but was strongest in the grey/orange one. This last result must have been caused by some effect of the match between prey and background colour, and behavioural explanations are suggested. It is unclear whether the prey were exhibiting 'crypsis' or 'masquerade'.     

Anti-apostatic selection by wild birds on quasi-natural morphs of the land snail Cepaea hortensis: a generalised linear mixed models approach

Anti-apostatic selection occurs when predators feed disproportionately on rare prey, a process that leads to a decrease in variation within prey populations. Most of the evidence for this phenomenon comes from experiments using artificial prey items distinguished by simple artificial colour differences. We describe an experiment performed at 10 different sites in which we tested whether selection by wild birds is anti-apostatic when presented with high density populations of shells of the polymorphic land snail, Cepaea hortensis , stuffed with pastry. At each site we presented two sorts of populations in sequence: 45 yellow unbanded with 5 yellow banded, and 5 yellow unbanded with 45 yellow banded. Selection was measured using an analysis based on generalised linear mixed models (GLMMs) that has broad applicability to all resource selection studies where extraneous variance is a problem. Using this analytic method, we found that birds altered their behaviour to produce significant anti-apostatic selection in direct response to changes in frequency of the quasi-natural Cepaea morphs.     

Does prey dispersion affect frequency-dependent predation by wild birds?

Despite the fact that the vast majority of natural prey items are dispersed in a non-random manner, few studies of frequency-dependent selective predation have explicitly examined the effect of prey dispersion on selectivity. We examined the effect of prey dispersion on the direction and strength of frequency-dependent selection by wild birds feeding on artificial prey (green or brown pastry baits). In a series of four experimental manipulations, we tested for the occurrence of frequency dependence with two different dispersion patterns (random or clumped). Manipulations were carried out at one of two absolute densities (25 prey m⁻² or 100 prey m⁻²), and were repeated at different sites in Southampton, England and Aljarafe, Spain. Our results suggest that prey dispersion has no effect on either frequency-dependent or -independent preferences. One possible explanation for this is that the birds had virtually complete information about prey frequencies at the relatively high densities used and based their preferences on their overall perception of availability. However, we cannot rule out the possibility that dispersion may influence selectivity when prey are cryptic or available at lower absolute densities. Although there was no effect of dispersion, frequency-dependent selection was, overall, significantly anti-apostatic (i.e. rare baits were more preferred than common baits). This anti-apostatic effect was stronger in Southampton than Aljarafe and stronger at 100 prey m⁻² than 25 prey m⁻². The differences in the strength of selection between the two locations was interpreted in terms of (i) whether the avian predators foraged in flocks or not, and (ii) the number of different species present (and, consequently, the variation in preference among individual birds). Received: 21 December 1996 / Accepted: 26 February 1997     

Optimal diet selection,frequency dependence and prey renewal

This paper extends existing models of frequency-dependent diet selection by considering the optimal diet selection of a predator feeding upon prey populations which can be depleted but are also capable of renewal (e.g. immigration, growth, or reproduction). This model and existing models which include prey depletion, predict partial-preference and a generic diet preference for the commonest prey types (apostatic selection). Unlike previous diet selection models, it is found that the optimal diet selection of an individual predator can be to favour the rarest prey type (anti-apostatic selection) when encounter rates are high, even if the individual prey do not differ in their nutritional value. Studies have demonstrated that predators generally show apostatic selection, even when all prey have the same nutritional value. Anti-apostatic selection has also been observed when prey are crowded, and therefore at high density, consistent with the idea of high encounter rates. This anti-apostatic diet selection has previously been proposed as evidence for the use of prey search images by a predator, or variation in individual prey preference. In this paper it is suggested that prey renewal is a further factor, often confounded in experiments, which could favour anti-apostatic selection.     

Wild birds prefer the familiar of striped and unstriped artificial prey

Previous work has shown that wild birds can become trained to search for a prey type on the basis of its colour. The experiments presented in this paper extended this work to two types of green artificial prey that were identical except for the presence or absence of a red stripe. Wild birds at six widely separated sites were trained on populations of one type and then were offered a choice between equal numbers of the familiar type and a second type. All the populations were presented on lawn backgrounds. The experiments were then repeated at each site with the birds being trained on the second prey type. The results showed that there was a consistent tendency for the familiar prey type to be overpredated and this was statistically significant. The behaviour described in this paper, if it occurs in nature, could lead to frequency-dependent (apostatic) selection and the maintenance of polymorphism in prey species in which the morphs are distinguished by colour patterns.     

Fast movement of densely aggregated prey increases the strength of anti-apostatic selection by wild birds

'Anti-apostatic' selection occurs when predators preferentially remove rare forms of prey, and has been demonstrated in both static (artificial) prey and moving (natural) prey. We performed 24 experiments at separate sites to test the effect of prey mobility on the strength of anti-apostatic selection by wild passerine birds. The prey were yellow and red Calliphora maggots and were presented in 9: 1 and 1: 9 ratios on a specially designed bird table. The temperature of the maggots, and hence their speed of movement, was adjusted by a combination of the ambient temperature and a candle under the bird table. Selection was anti-apostatic at all three classes of temperature and was strongest at the highest. We conclude that anti-apostatic selection on static prey is enhanced when they are made to move–possibly because the 'confusion effect' caused by the moving prey makes the birds concentrate more strongly on the rarer colours.     

The burning peanut

Visual selection occurs when predators preferentially remove conspicuous varieties from prey populations and thus confer a selective advantage on inconspicuous varieties. In previous papers we have described a general method for simulating such natural selection, and we here give details of an improvement that again uses wild birds as predators, green and brown pastry ‘baits’ as prey, and trays containing coloured stones as the backgrounds. We used backgrounds of three different colours: green (on which brown baits were conspicuous), brown (on which browns were inconspicuous), and white (on which both prey types were conspicuous). The important difference from the previous design was that these backgrounds were presented simultaneously on the same bird table. We placed equal numbers of green and brown prey on each background and recorded the numbers eaten by wild birds. We did six experiments to test the design. Two different methods of measuring selection showed that the results were generally in the direction predicted from the hypothesis that conspicuous prey are more likely to be eaten by predators.     

Frequency-dependent selection by predators

Sometimes predators tend to concentrate on common varieties of prey and overlook rare ones. Within prey species, this could result in the fitness of each variety being inversely related to its frequency in the population. Such frequency-dependent or 'apostatic' selection by predators hunting by sight could maintain polymorphism for colour pattern, and much of the supporting evidence for this idea has come from work on birds and artificial prey. These and other studies have shown that the strength of the observed selection is affected by prey density, palatability, coloration and conspicuousness. When the prey density is very high, selection becomes 'anti-apostatic': predators preferentially remove rare prey. There is still much to be learned about frequency-dependent selection by predators on artificial prey: work on natural polymorphic prey has hardly begun.     

Frequency-dependent predation by birds at edges and interiors of woodland

Structural variations between edge and interior areas within forest fragments may bring about differences in food availability that may influence the selective behaviour of predators and prey population dynamics. The purpose of this paper was to assess patterns of artificial prey selection by wild birds (House Sparrow Passer domesticus and Rock Dove Columba livid) between edge and interior areas of woodland, taking into account differences in prey frequency (10% of one food type and 90% of the other) and density (30 baits/m^-2 and 50 baits/m-²). Experiments were conducted at 24 plots in 3 forest fragments in the city of Madrid, Spain. Selectivity did not vary among parks or between densities. However, selectivity did vary with the frequency and location of baits, showing an anti-apostatic trend (baits were preferred at low rather than at high frequencies) that was more pronounced at interiors than at edges. Two possible factors that may account for stronger anti-apostatic selection at edges are the higher densities of predators and pedestrians found there. However, there are many other possible explanations, and no specific conclusion can be supported with the current data. The results of this study also point out that site heterogeneity should be taken into account in the experimental design of future studies on frequency-dependent food selection by wild birds, particularly in fragmented landscapes.     

Can the dietary conservatism of predators compensate for positive frequency dependent selection against rare,conspicuous prey?

The evolution of conspicuous colouration in prey is puzzling because such coloration attracts the attention of predators. Anti-apostatic selection, in which rare prey forms are predated disproportionately often, is a second potential obstacle to the evolution of conspicuous colouration in prey, as bright novel prey forms are likely to be very rare when they first appear in populations. It has recently been postulated that dietary conservatism in predators, an extended feeding avoidance of novel prey, would allow novel conspicuous prey to survive and multiply despite anti-apostatic and conspicuousness effects. We tested this hypothesis for a novel prey type arising in an otherwise cryptic population, providing a direct test of whether anti-apostatic selection or the predators’ wariness to attack the novel prey type is the more important force acting on the novel conspicuous prey. We conducted our experiment in the “Novel World”; an experimental system designed to test predators’ foraging decisions in a large landscape. We found that the conspicuous, novel prey suffered high initial costs of conspicuousness compared with cryptic prey, since most of these prey were attacked during the first “generation”, with no opportunity to “reproduce”. However, a subset of the 17 birds (24%) were following a dietary conservative foraging strategy and they were reluctant to eat the novel prey. Interestingly these birds were not more neophobic or less explorative. Our data demonstrate how difficult it is for the novel conspicuous prey to survive in cryptic populations, but they also highlight the importance of the predator’s foraging strategies in helping to promote the evolution and maintenance of aposematism.     

Wild birds prefer the familiar morph when feeding on pastry-filled shells of the landsnail Cepaea hortensis (Müll.)

Colour polymorphisms in prey could be maintained if predators concentrate on common morphs and confer a selective advantage on rare morphs. We describe experiments to test whether wild birds feed on pastry-stuffed shells of Cepaea hortensis in a manner that might lead to such apostatic selection. The birds were first given a 'pre-training' choice test of a shell population with equal numbers of yellow unbandeds and yellow five-bandeds; they were then trained on one morph alone, given a second choice test, trained on the other morph and, finally, given a third choice test. The birds preferred five-bandeds in five of the six pre-training tests. In all six experiments the first training session increased the birds' preferences for the morph that was familiar. The results were less clear-cut when selection during pre-training was compared with selection after the second training session. However, a comparison between selection after each of the two training sessions showed that in all six experiments the results were in the direction predicted from the hypothesis that familiar morphs are preferred. This set of experiments is one of the few in which behaviour which could lead to apostatic selection has been tested with morphs that differ in pattern. The findings support the idea that polymorphism in Cepaea could be maintained by apostatic selection.     

The evolution of colour polymorphism in British winter‐active Lepidoptera in response to search image use by avian predators

Phenotypic polymorphism in cryptic species is widespread. This may evolve in response to search image use by predators exerting negative frequency‐dependent selection on intraspecific colour morphs, ‘apostatic selection’. Evidence exists to indicate search image formation by predators and apostatic selection operating on wild prey populations, though not to demonstrate search image use directly resulting in apostatic selection. The present study attempted to address this deficiency, using British Lepidoptera active in winter as a model system. It has been proposed that the typically polymorphic wing colouration of these species represents an anti‐search image adaptation against birds. To test (a) for search image‐driven apostatic selection, dimorphic populations of artificial moth‐like models were established in woodland at varying relative morph frequencies and exposed to predation by natural populations of birds. In addition, to test (b) whether abundance and degree of polymorphism are correlated across British winter‐active moths, as predicted where search image use drives apostatic selection, a series of phylogenetic comparative analyses were conducted. There was a positive relationship between artificial morph frequency and probability of predation, consistent with birds utilizing search images and exerting apostatic selection. Abundance and degree of polymorphism were found to be positively correlated across British Lepidoptera active in winter, though not across all taxonomic groups analysed. This evidence is consistent with polymorphism in this group having evolved in response to search image‐driven apostatic selection and supports the viability of this mechanism as a means by which phenotypic and genetic variation may be maintained in natural populations.     

Apostasy and selection for crypsis in the marine snail Littoraria filosa: an explanation for a balanced colour polymorphism

The marine snail, Littoraria filosa, is polymorphic for shell colour, with yellow, brown, and pink morphs that correspond in both appearance and frequency to the predominant background colours of its habitat. Previous research on this polymorphism has found indirect evidence of apostatic selection and selection for crypsis by unknown agents, probably crabs, and direct evidence of selection for crypsis by the parasitoid fly Sarcophaga megafilosia. In the present study, we report on field experiments to investigate whether S. megafilosia and shell‐crushing predators exert apostatic selection on L. filosa. For S. megafilosia, seven experimental treatments containing yellow and brown snails in the proportions of 0.1, 0.2, 0.4, 0.5, 0.6, 0.8, and 0.9 of each colour were established on mangrove trees and used to separately quantify the proportions of each colour attacked on grey/brown trunks and yellow/green leaves. The results obtained confirmed an earlier finding of selection for crypsis, but only showed slight, but significant, anti‐apostatic selection by S. megafilosia. For shell‐crushing predators, seven experimental treatments containing yellow and brown snails in the proportions of 0.08, 0.17, 0.33, 0.50, 0.66, 0.83, and 0.92 were established on two types of trees that differed in their background proportions of brown and green: (1) trees which had been pruned of approximately 90% of their foliage and (2) unpruned trees. The results obtained showed both selection for crypsis and apostatic selection. Furthermore, a selectively neutral frequency for yellow L. filosa was found for each background, and was less on pruned trees than unpruned ones (and vice versa for brown L. filosa), which can therefore account for the maintenance of a colour polymorphism where the proportions of each morph tend to resemble and correspond in frequency to the colours of the background. © 2008 The Linnean Society of London, Biological Journal of the Linnean Society, 2008, 95 , 62–71.     

Multichannel feeding by spider functional groups is driven by feeding strategies and resource availability

Multichannel feeding, whereby consumers feed across resource channels such as upon herbivore and detritivore resources, acts to link discrete compartments of a food web with implications for ecosystem functioning and stability. Currently however, we have little understanding which feeding strategies of consumers underlie multichannel feeding. We therefore link spider functional group and resource density‐dependent or density‐independent feeding strategies to multichannel feeding by quantifying not only consumer diet, but also the relative availability of resources. Here we analysed herbivore (green) and detritivore (brown) prey use by spider communities in grasslands, and tested if available prey biomass proportions were linked to observed spider dietary proportions. Different spider functional groups each linked green and brown resource channels, but while green prey were always consumed in proportion to their relative biomass, brown prey were consumed independently of proportion by some functional groups. Additionally, we found greater intraguild predation by cursorial spiders when green resources were relatively scarcer, suggesting green prey was preferred, and needed to be compensated for when rare. Overall, we observed a stronger consumer connection to the green than brown resource channel, yet this green connection was more variable due to greater range in green resource availability across grasslands and density‐dependent consumption on green prey. Consequently, multichannel feeding by spiders was determined by density‐dependent and density‐independent feeding strategies that varied by spider functional group and across resources channels. Our results demonstrate that the role of multichannel feeding by spiders in linking separate food web compartments is a dynamic component of food web structure in these wild grasslands.     

Some population consequences of variation in preference among individual predators

A number of foraging studies have demonstrated that populations of predators rarely consist of individuals with identical preferences for particular types of prey. Variation among predators can lead to frequency-dependent changes in population preference, because those predators mat prefer the rarer type of prey generally have the greatest influence on population preference. In this study we develop a series of theoretical models which demonstrate how anti-apostatic selection (i.e. selection against the rare form) can arise out of (a) bimodal and (b) normal variation in preference among individuals of the same species. We show that population level anti-apostatic selection can occur even when individual predators show pro-apostatic selection (i.e. selection against the common form). Furthermore, patterns of population prey selection that arise out of variation in preference can potentially be pro-apostatic over one range of relative densities and anti-apostatic over another range of relative densities. Finally, we examine a case study involving predation by female waterboatmen Notonecta glauca and show that the variation in preference in this species is large enough to generate higher anti-apostatic selection than would be expected from the diet selected by the average individual.     

Reversible frequency-dependent predation of a puffer, Takifugu niphobles (Pisces: Tetraodontidae), related to spatial distribution of colour-polymorphic prey

The puffer Takifugu niphobles is a top predator of hard-shelled prey such as molluscs; its predatory tactics may affect the evolution of prey coloration. Two hypotheses concerning its foraging were tested: (1) T. niphobles shows frequency dependence in foraging colour-polymorphic prey, and (2) such dependence reverses in response to changes in prey distribution. Captive fish were provided with 70 artificial prey, coloured either dark brown or pale brown, at four frequencies (1 : 4, 2 : 3, 3 : 2, 4 : 1) and in two distribution patterns (uniform and aggregated). When prey were uniformly distributed, frequency and feeding rate significantly influenced colour preference: the common morph was consumed more. When prey were aggregated, frequency significantly affected preference only when the feeding rate was low, in which case the rare morph was consumed more. Thus both hypotheses were supported. The impact of T. niphobles 's frequency-dependent predation and its reversal on the colour evolution of prey species, especially molluscs, is discussed.  © 2004 The Linnean Society of London, Biological Journal of the Linnean Society , 2004, 81 , 197–202.     

Frequency-dependent Food Selection by Domestic Cats: A Comparative Study

Preferences for common food types (‘apostatic selection’) have been demonstrated in a wide variety of vertebrate predators, yet there are few examples of preferences for rare food types (‘anti-apostatic selection’). Anti-apostatic selection is predicted to occur when, among other things, there are nutritional benefits to be gained from the consumption of a mixed diet. We tested this hypothesis by examining the frequency-dependent food preferences of domestic cats (Felis silvestris catus) with different nutritional histories. Subjects were classified as being either nutritionally ‘experienced’ (farm and rescue shelter cats, with a history of scavenging for nutritionally variable foods) or nutritionally ‘inexperienced’ (cats reared indoors on high-quality, nutritionally complete diets). We tested for frequency dependence by allowing individuals and groups of cats from the two groups to select from high-density mixtures of two types of artificial food pellet. In experiments on individual cats, nutritionally experienced subjects showed significant anti-apostatic selection, whereas inexperienced cats produced only a weak anti-apostatic trend. In experiments on groups of cats, both inexperienced and experienced groups showed significant anti-apostatic selection. The apparent inconsistency between individual and group results could be explained in terms of the additional anti-apostatic effects that result from variation among individuals in group foraging situations (i.e. when the effects of individuals are pooled). Because other behavioural explanations, such as perceptual contrast and sampling effects, were unlikely to have influenced our results, we conclude that the differences in selection between experienced and inexperienced individuals were probably due to the differing extent to which the consumption of a mixed diet was beneficial. These experiments may offer some insight into the success of the domestic cat in urban areas: although obligate carnivores, they appear to possess flexible feeding strategies which will tend to allow them to select a reasonably balanced diet from nutritionally variable resources in, for example, refuse bins.