Protective coloration and frequency-dependent selection : responses of birds to prey colour patterns

Abstract

Animal coldration in defence against predators has been discussed for over a hundred years. Early works were mainly descriptive and often assumed that any difficulties experienced by humans in detecting animals were also experienced by visual predators. Subsequent evidence for cryptic coloration was surmised from convergent evolution of colours and patterns and from correlations between backgrounds and particular colours and patterns. There is strong circumstantial evidence but relatively little experimental evidence that the colours and patterns of animals actually afford protection against visually hunting predators. Many prey species that are polymorphic for coloration also appear to be cryptic, and it has been suggested that predators may select colour patterns in a frequency-dependent manner and thereby maintain such colour pattern polymorphisms. The experiments described in this thesis used striped colour patterns to investigate protective coloration and frequency-dependent selection. Birds were offered choices between prey differing in their colour patterns. The ultimate effects on the prey selection were studied. Artificial prey made from coloured pastry were used for all experiments except one in "which pastry-filled Cepaea bortensis shells were used. The pastry prey were flat and striped coloured patterns could be applied, Predators were wild birds in all but one of the experiments, where domestic chicks were used. Three basic experimental designs were involved: simple choice tests to investigate colour patterns and crypsis; training experiments where birds were trained on one prey type and then given a choice between two or more prey types; experiments with populations of two prey types at different densities and frequencies to test for frequency-dependent selection. The first experiment used flat artificial backgrounds and results indicated that striped prey were at a slight advantage on a matching striped background. Subsequent experiments were carried out on grass, which presented a more complex and three-dimensional background. Choice tests showed that prey with a ground colour that matched the grass were at an advantage over unmatching prey and that a matching stripe gave an added advantage. Observations showed that some birds avoided white prey types with a red stripe but not green prey types with a red stripe'. Experiments were undertaken with plain and striped brightly coloured prey types. Birds tended to avoid the plain brightly coloured prey types and 'the possible reasons for this are discussed, Training experiments showed that wild birds can become trained to search for a prey type on the basis of the presence or absence of a coloured stripe or on banded and unhanded Cepaea hortensis shells. The effect of training decreased when the prey types became more similar and when the training period was shorter." Overpredation of the familiar prey type could result in apostatic selection and the maintenance of polymorphism in prey species in which the morphs are distinguished by colour patterns. Further experiments with prey types that differed only very slightly in their colour patterns found that, in one series where the training prey type was the same throughout, as the birds became more familiar with one prey type they took proportionally more of the unfamiliar prey type, suggesting that training effects or overpredation of common prey found in experiments may only be short-term. Chicks, which were trained on one prey type and then given a choice between prey types that did not include the training prey type, preferred the prey type that was most similar to the training prey type. Experiments involving a range of prey densities and frequencies found anti-apostatic selection at the prey "density of SO/m---, both apostatic and anti- apostatic selection at the prey density of 385u/m^ ('maximum') and no frequency-dependent selection at two lower prey densities. The experiments reported in this thesis show that birds can discriminate between very similar colour patterns, that they can select on the basis of colour patterns and that this selection can be frequency-dependent.

More information

Identifiers

Catalogue record

Export record

ASCII CitationAtomBibTeXData Cite XMLDublin CoreDublin CoreEP3 XMLEndNoteHTML CitationHTML CitationHTML ListJSONMETSMODSMPEG-21 DIDLOpenURL ContextObjectOpenURL ContextObject in SpanRDF+N-TriplesRDF+N3RDF+XMLRIOXX2 XMLReferReference ManagerSimple Metadata