Evidence of phenotypic correlation between exploration activity and resting metabolic rate among-populations across an elevation gradient in a small rodent species

Behavior and metabolism are frontline reactions to environmental challenges that can covary in their response through at least two mechanisms. First, natural selection can generate correlation in phenotype among distinct populations if they are exposed to a common selective force. Thus, metabolism and behavior can exhibit phenotypic correlation among populations when responding (independently from each other) to co-varying selective forces. Second, because behavioral responses are energy demanding, variation in energy acquisition or allocation among individuals of the same population can also generate, respectively, a positive or negative correlation within populations. To address this issue, we investigated among- and within-population (co-)variations in exploration activity (EA) and resting metabolic rate (RMR) of adult common voles (Microtus arvalis) issued from four high elevation populations (>1400m a.s.l.) and five low elevation populations (<520m a.s.l.). Individuals were acclimatized for at least one month to the same laboratory conditions before being tested for EA and RMR. Voles from high elevation populations were more explorative and they had higher RMR than their counterparts from low elevation populations. The similar effects of elevation on EA and RMR accounted for a correlation of 0.28 [0.064; 0.658] between EA and RMR across low and high elevation populations. We found no evidence of a withinpopulation correlation between EA and RMR. More work relying, for instance, on repeated sampling or experimental selection is nonetheless needed to confirm a lack of integration between metabolism and behavior at the individual level. Our results highlight the importance of co-varying selective forces in generating among-population phenotypic correlation between EA and RMR in this small rodent species.