Differential predation on sexes affects colour polymorphism of the isopod Idotea baltica (Pallas)

Variable selection, including spatio-temporal variation, frequency-dependent selection and differential selection due to habitat choice, may maintain polymorphism in heterogeneous environments. We studied predation as a selective agent on colour polymorphism of the aquatic isopod I baltica. Variable predation on this species can arise from at least three sources. First, apostatic selection was studied by testing the formation of preferences on colour morphs in the perch, a common predator of I baltica. Such acquired preferences should induce apostatic selection. While our results indicate some acquired preferences, there was significant heterogeneity in the behaviour of predator individuals. Second, temporal variation in selection can arise due to habitat shift from the green algae juvenile habitat to the bladderwrack adult habitat, and the consequent change in the crypsis of the morphs. Different crypsis between sexes probably promoted high predation mortality among females in the juvenile habitat. The high rate of male mortality during the breeding period, on the other hand, was presumably due to their high mate-searching activity. Third, the sex-dependent habitat choice of I baltica leads to sexual differences in the susceptibility of morphs to predation. Predators preferred the white-spotted morph over the uniform one in males but not in females, supporting the 'dimorphic niche' hypothesis as an explanation of sexual differences in morph frequencies. Finally, no evidence was found that the colouration patterns were under sexual selection. We therefore conclude diat variable predation is the most promising explanation for the maintenance of polymorphism in I. baltica.     

Selection by wild birds on artificial dimorphic prey on varied backgrounds

Our aim was to test the effects of prey frequency and background composition on selection by free-ranging birds. We did three series of experiments with populations of grey and orange pastry prey scattered among coloured stones that made the prey either conspicuous or inconspicuous. Series 1 tested whether the predicted equilibrium frequency of the two prey types was influenced by the frequency of matching grey and orange stones. Birds at a single site were given a random sequence of combinations of prey frequency and stone frequency. Selection was dependent on background and the effect of prey frequency also varied with background. In series 2, we explored the frequency-independent effect of background: birds at five sites were given equal numbers of the two prey in three frequencies of matching stones and two of non-matching. There was a higher risk of predation for prey that matched rarer stones. In series 3 we attempted to measure, at a single site, the actual equilibrium prey frequencies in three different backgrounds: two extreme stone frequencies and one intermediate. Each experiment started with a population of equal numbers of grey and orange prey. After half the prey had been eaten we calculated the frequencies of the survivors and presented a new population of the original size but with the new prey frequencies; each experiment ran for 25 such 'generations'. The results suggested that at equilibrium the commoner 'morph' was the one that resembled the commoner colour of stone. Overall, our findings support the idea that visual selection can result in morph frequencies becoming related to the proportions of their matching background components and that this equilibrium will 'track' temporal or spatial changes in the background.     

Frequency-dependent predation and maintenance of prey polymorphism

In positive frequency-dependent predation, predation risk of an individual prey correlates positively with the frequency of that prey type. In a number of small-scale experiments individual predators have shown frequency-dependent behaviour, often leading to the conclusion that a population of such predators could maintain prey polymorphism. Using simulations, I studied the dynamics of frequency-dependent predation and prey polymorphism. The model suggests that persistence of prey polymorphism decreases with increasing number of predators that show frequency-dependent behaviour, questioning conclusions about polymorphism based on experiments with few predators. In addition, prey population size, prey crypsis, difference in crypsis between prey morphs and the way the behaviour was adjusted affected the persistence of polymorphism. Under some circumstances prey population remained polymorphic for a shorter time under frequency-dependent than under frequency-independent predation. This suggests that although positive frequency-dependent predator behaviour may maintain prey polymorphism, it is not a sufficient condition for persistent prey polymorphism.     

Natural selection for rare and mimetic colour pattern combinations in wild populations of the diadem butterfly, Hypolimnas misippus L

Mark recapture and morph frequency data, gathered during a population irruption of Hypolimnas misippus in southern Ghana, provide evidence for apostatic and mimetic selection. During a period of low adult survival, both the recapture rate and the frequency of the commonest morph ( misippus ) were significantly reduced. Selection against this form increased phenotypic diversity and generated significant disequilibrium in the combinations of unlinked fore- and hindwing phenotypes. There was also evidence for selection against those forms (weak alcippoides ) which most closely resemble misippus . Other morphs, including both good mimics of Danaus chrysippus and rare non-mimics, showed no reductions in recapture rate during the period of low survival, but only the good mimics increased significantly in frequency. The results provide a predictive ecological model for density-dependent selection by predators which is consistent with field data from previous studies of H. misippus in Ghana and Tanzania. Their evolutionary implications are discussed, and it is suggested that anomalies in the mimicry of this species may be partly due to lack of predation when it is scarce.     

Does a polymorphic species have a ‘polymorphic’ diet? A case study from a lacertid lizard

Lizards are ideal for studying colour polymorphism, because some species are polymorphic and the morphs often have different ecological or reproductive strategies. We studied the feeding habits of six polymorphic populations of Podarcis muralis to test whether morphs differed in their diet. Some taxa were selected in a similar way by all morphs, but selection on other taxa varied and was characteristic of each morph. Diet was most different for the red and yellow morphs. Two hypotheses could explain these differences: active segregation in the trophic niche or active segregation in space dependent on spatial heterogeneity in prey availability. The former is improbable because P. muralis is considered an opportunistic feeder, whereas the latter could occur if the morphs adopted alternative territorial strategies with consequent spatial segregation.     

Mimetic butterflies support Wallace's model of sexual dimorphism

Theoretical and empirical observations generally support Darwin's view that sexual dimorphism evolves due to sexual selection on, and deviation in, exaggerated male traits. Wallace presented a radical alternative, which is largely untested, that sexual dimorphism results from naturally selected deviation in protective female coloration. This leads to the prediction that deviation in female rather than male phenotype causes sexual dimorphism. Here I test Wallace's model of sexual dimorphism by tracing the evolutionary history of Batesian mimicry-an example of naturally selected protective coloration-on a molecular phylogeny of Papilio butterflies. I show that sexual dimorphism in Papilio is significantly correlated with both female-limited Batesian mimicry, where females are mimetic and males are non-mimetic, and with the deviation of female wing colour patterns from the ancestral patterns conserved in males. Thus, Wallace's model largely explains sexual dimorphism in Papilio. This finding, along with indirect support from recent studies on birds and lizards, suggests that Wallace's model may be more widely useful in explaining sexual dimorphism. These results also highlight the contribution of naturally selected female traits in driving phenotypic divergence between species, instead of merely facilitating the divergence in male sexual traits as described by Darwin's model.     

An overview of the relationships between mimicry and crypsis

There have been many different and conflicting definitions of mimicry. Some of the definitions of mimicry include crypsis and others do not. Each definition includes different groups of phenomena and uses different criteria to distinguish mimetic from non-mimetic phenomena. The confusion is eliminated by a consideration of the criteria of all definitions. This shows that there are in fact three major criteria dividing six phenomona, rather than a single dichotomy between mimicry and crypsis (Table 2). The criteria are defined by the results of a mistake in discrimination between the model and mimìc: (a) the mistake does or does not depend upon relationship between mimic and background; (b) the mistake has or has no effect on the population dynamics or evolution of the model and (c) the mistake affects dynamics or evolution of one or of many models. The main reason for the contusion about mimicry and crypsis is that each author's definition includes differing and partially overlapping subsets of the six classes: crypsis; masquerade; Batesism; Müllerism; polymorphism and convergence.     

An overview of the relationships between mimicry and crypsis

There have been many different and conflicting definitions of mimicry. Some of the definitions of mimicry include crypsis and others do not. Each definition includes different groups of phenomena and uses different criteria to distinguish mimetic from non-mimetic phenomena. The confusion is eliminated by a consideration of the criteria of all definitions. This shows that there are in fact three major criteria dividing six phenomona, rather than a single dichotomy between mimicry and crypsis (Table 2). The criteria are defined by the results of a mistake in discrimination between the model and mimìc: (a) the mistake does or does not depend upon relationship between mimic and background; (b) the mistake has or has no effect on the population dynamics or evolution of the model and (c) the mistake affects dynamics or evolution of one or of many models. The main reason for the contusion about mimicry and crypsis is that each author's definition includes differing and partially overlapping subsets of the six classes: crypsis; masquerade; Batesism; Müllerism; polymorphism and convergence.