Interpreting local adaptation studies

Adaptation through natural selection is the basis for evolutionary change. At the micro-evolutionary scale, population differentiation is the path from which species eventually form. For this reason, researchers have a long history of studying local adaptation within species. Tests of local adaptation usually involve reciprocal transplants of individuals between populations and com-paring some kind of performance/fitness measure of the individuals. In general, local adaptation is defined as when local individuals do better in their local habitat than individuals transplanted from other environments ('local vs. foreign', Kawecki and Ebert 2004). Alternat-ively, local adaptation can also be defined as individuals having higher fitness at their home site compared with other sites ('home vs. away', Kawecki and Ebert 2004). Of course, not all comparisons of reciprocal transplants meet these criteria, and Kawecki and Ebert (2004) suggest that the ‘local vs. foreign’ criteria should be used as a diagnostic for local adaptation, especially when the ‘home vs. away’ criterion is met, but not the ‘local vs. foreign.’ In these cases, further studies could reveal why some genotypes do better than the local ones. Of course, local adaptation is not predicted to be or found in all cases (Leimu and Fischer 2008, Hereford 2009). However, Vesakoski and Jormalainen (2013) suggest we might be ignoring a signal of local adapta-tion from reciprocal transplant studies. Similar to the ‘home vs. away’ criterion, their ‘allopatric site advantage’ hypothesis (naming is mine) suggests some genotypes are superior in all conditions. However, it differs from the ‘home vs. away’ criterion because the ‘home’ site is not necessarily the best for all populations. Vesakoski and Jormalainen suggest that individuals may locally adapt to the level of stress to which they are exposed. If the ‘allopatric site advantage’ operates in populations, than Vesakoski and Jormalainen lay out a particular pattern that one would expect from reciprocal transplant studies. Here, a ‘low’ stress population should do better in its home site than away sites, meeting the ‘home vs. away’ criterion. In a ‘high’ stress population, individuals from the high stress environment should do better than individuals from the low stress environment, meeting the ‘local vs. foreign’ criterion. However, at the low stress site, individuals from the high stress environment should do equally well or better than the low stress individuals, meeting none of the local adaptation criteria. The ‘allopatric site advantage’ hypothesis shares similarities with the idea that the opportunity for selection is related to the interaction strength (e.g. Vanhoenacker et al. 2013). In general, the opportunity for selection on traits is expected to increase with the interaction strength (although perhaps not linearly). Similarly, if a population is tolerant of the local stress, than one would not expect further selection for having even higher tolerance to that stress. Presumably a population would need to be exposed to increased levels of the stressor to have further selection to increase tolerance. Thus, in the hypothetical example given by Vesakoski and Jormalainen (2013), herbivores adapt to tolerate the local level of a plant resistance chemical. Once a population of herbivores has adapted to the local defense levels, the interaction strength is diminished (for the herbivore) and there is little/no opportunity for selection on the traits associated with tolerance. Of course, for illustrative purposes, this example simplifies plant-herbivore interactions to a moment in time and