The ecological causes of evolution

Natural selection is the process that results in adaptive evolution, but it is not the cause of evolution. The cause of natural selection and, therefore, of adaptive evolution, is any environmental factor (agent of selection) that results in differential fitness among phenotypes. Surprisingly little is known about selective agents, how they interact or their relative importance across taxa. Here, I outline three approaches for their investigation: functional analysis, correlational analysis and experimental manipulation. By refocusing attention on the structure and consequences of ecological variation, a better characterisation of selective agents would improve understanding of natural selection and evolution, including adaptive radiation, coevolution, the niche, the evolutionary ecology of the ranges of species and their response to environmental change. What causes natural selection? When one contemplates how understanding of evolution has improved over the 150 years since Darwin first published his ideas about natural selection [1], it might come as a surprise to realise how little is understood about the causes of adaptive evolution in natural populations. Recent developments are causing evolutionary ecologists to think more explicitly about the nature of natural selection: how it varies spatiotemporally [2-5], the ecology that drives it Darwin obviously thought about such things. In The Origin of Species [1] he reflected on 'checks to increase': that is, the factors that curb population growth and result in the struggle for existence. Darwin was aware that 'What checks the natural tendency of each species to increase in number is most obscure', but he considered that biotic interactions were more important than were abiotic ones, and he regretted 'our ignorance on the mutual relations of all organic beings'. Although there is now some understanding of the importance of individual selective agents in specific systems (see below), there is still a lack of systematic understanding of the gamut of agents for any one system, let alone the relative importance of agents across traits or taxa. Thus, Darwin's assertion that 'We know not exactly what the checks are in even one single instance' is almost as true today as it was 150 years ago