It’s the genotype, stupid!
Article first published online: 21 JAN 2009
© 2009 Blackwell Verlag GmbH
Journal of Animal Breeding and Genetics
Volume 126, Issue 1, pages 1–2, February 2009
How to Cite
Merilä, J. (2009), It’s the genotype, stupid!. Journal of Animal Breeding and Genetics, 126: 1–2. doi: 10.1111/j.1439-0388.2008.00795.x
- Issue published online: 21 JAN 2009
- Article first published online: 21 JAN 2009
Evolution is about genetics and selection. In order to understand evolution, we need to understand genetics and forces of selection acting on traits of interest. Yet in much of the current evolutionary biology research, actual details of inheritance as well as selection processes are sometimes viewed as nuisances: genetic variances and covariances are hard to estimate in the wild, and the same is often true with selection coefficients. Hence, strong and far-reaching evolutionary inferences are sometimes made by making questionable, if not unjustified, assumptions about the genetics of observed phenotypic changes.
A recent interest in fisheries-induced evolutionary changes provides a case in point. While it is probably true that fisheries are highly selective with respect to many phenotypic traits, imposing strong directional selection, e.g. for earlier maturation at smaller size, the actual forces of this selection have seldom been estimated. The same goes for identifying the genetic basis of the postulated selection responses: the still relative small, but rapidly exploding literature of fisheries-induced ‘evolution’, is filled with claims of evolutionary responses, yet none of the case studies from the wild published so far – and this is a fact – has demonstrated a genetic basis for the claimed selection responses.
As animal breeders and evolutionary biologists working with quantitative genetics methods know, phenotypic changes in a population can – and often do – come about due to changes in environmental conditions only. Likewise, constraints imposed by genetic covariances among traits can limit selection responses, and genetic variances and covariances can change as a function of environmental conditions and the very process of selection (e.g. fishery or climate change) either for genetic or environmental reasons. Hence, while basing evolutionary inference solely on phenotypic measurements is obviously questionable science for geneticists (no self-respecting genetic journal would publish such an inference), this view is not shared among all evolutionary biologists and journals publishing evolutionary biology research, including the highly respected and weekly publishing glossy science magazines.
Part of the problem here might be that many evolutionary biologists are ecologists turned into evolutionary biologists without insight and deep understanding of the complexities of quantitative trait inheritance and dynamics. Another issue may reside with the fact that there is tradition of ‘forgiveness to details’ in evolutionary biology: fossil record does not allow separation of genetic and plastic effects, but evolutionary inference based on fossils was for long – and in many cases still is – the best game in town. However, as the genetic analyses of long-term studies of individually marked and pedigreed animal populations have now repeatedly demonstrated, evolutionary inference based only on phenotypic data can lead us astray.
Having raised a critical voice against my fellow evolutionary biologists, I should also say that I am the first to admit that there is also lot to improve and explore in the current practice of employing quantitative genetics in evolutionary biology research. One area where animal breeders could make significant contributions to the toolbox of evolutionary biologists is in the analytics pertinent to the comparison of the degree of population differentiation in quantitative traits (cf. QST index). Methodological and conceptual developments in this area have so far progressed largely without input from animal breeding science.
Likewise, application of ‘animal model’ analyses and breeding value inferences in evolutionary biology is a rather new and increasingly popular development among people having access to long-term data sets of pedigreed individuals from the wild. Here, the tools have been adopted directly from the animal breeding community which has not always been unproblematic. One particular issue is obvious: methods developed for animal breeding purposes are not fine-tuned to work smoothly with the data sets from the wild, which are typically orders of magnitudes smaller than those commonly used by animal breeders.
In conclusion, evolutionary biology needs genetics: phenotypic evolution refers, by definition, to genetic change in the mean trait value in the population. However, a phenotypic change alone is not sufficient to postulate the occurrence of evolution. Likewise, genetic changes are not necessarily reflected on the phenotype if opposed by environmental effects. Hence, quantitative genetic tools have a central role in contemporary evolutionary biology. Therefore, animal breeders – having strong and solid understanding of the genetic models and analytical tools – are well placed to contribute to current debates and research in evolutionary biology.