Coping with habitat heterogeneity: the story of Mediterranean blue tits

A prerequisite for understanding adaptation is to understand how populations respond to environmental heterogeneity. We chose the blue tit Cyanistes caeruleus and Mediterranean habitat mosaics which exhibit a large diversity of habitats for analysing the effects of environmental heterogeneity on phenotypic variation. Three main factors of heterogeneity have been considered: ( 1) whether dominant tree species are deciduous or evergreen, ( 2) the geographic configuration of habitats and landscapes, and ( 3) the degree of infestation by parasites, which considerably varies in space. Several study sites equipped with nest-boxes and traps for collecting the droppings of caterpillars falling from the leaves of trees have been monitored over several decades in a series of habitats in mainland and insular ( Corsica) landscapes. Depending on the geographic configuration of habitat patches within landscapes, the large phenotypic variation observed in many demographic, morphometric and behavioural traits has been shown to result either from a plastic response to habitat variation or from genetically determined specialisation to local habitats. Blue tits in deciduous habitats started to breed ca. 1 month earlier than in evergreen habitats, but patterns differed between the mainland and Corsica. On the mainland, populations may be locally maladapted because of gene flow across habitat patches, which results in a low supply/demand ratio of food, poor breeding performance and a source-sink population structure. In Corsica, higher phenotypic variation resulted from lower dispersal ranges in islands, a component of the insular syndrome. Genetically-based habitat-specific specialisation to local habitats on this island is a demonstration that adaptive responses of suites of life history traits to habitat-specific selection regimes may operate on a scale which is much smaller than the scale of potential dispersal and gene flow. Adaptive responses of blue tits to two constraints, i.e. high levels of parasitism and low amounts of food, have been studied in detail. The very small amount of measured genetic divergence between populations contrasts with the large inter-population phenotypic variation which is observed in many traits. This is one more example that natural selection can produce rapid and sometimes strongly adaptive morphological divergence in the absence of discernable differentiation at neutral DNA loci, and that weak genetic differentiation does not necessarily mean phenotypic resemblance.