Abstract

The opportunity for selection, I, defined as the variance in relative fitness, has been called an estimate of the ‘total amount of selection’. However, while a non-zero I is a necessary condition for selection, it is not a sufficient one. We investigated the relationship between I and the magnitude of standardized linear and non-linear selection gradients for body size in the waterstrider Aquarius remigis, measured over three episodes of selection. Male I exceeded female I for daily reproductive success, but the difference was not statistically significant and had little impact on net adult I. Linear selection gradients were only weakly correlated with I, while non-linear gradients were uncorrelated with I. Partitioning I among the three episodes of selection revealed that variance in net adult fitness was largely generated by variance in prereproductive survival. The patterns of selection across the adult life stage suggested by analysis of the opportunity for selection differed qualitatively and quantitatively from those revealed by selection gradient analysis. In particular, the former identified pre-reproductive survival as the key component of net adult fitness, even though there is little selection on total length in this life stage. We conclude that I is a useful adjunct to selection gradient analyses, but is perhaps most useful in the analysis of life-history evolution where the traits themselves are direct estimates of fitness.