The evolution of threshold traits: effects of selection on fecundity and correlated response in wing dimorphism in the sand cricket

The quantitative genetic basis of traits can be determined using a pedigree analysis or a selection experiment. Each approach is valuable and the comb ined data can contribute more than either method alone. Analysis using both sib analysis and selection is particularly essential when there are likely to be nonlinearities in the functional relationships among traits. A class of traits for which this occurs is that of threshold traits, which are cha racterized by a dichotomous phenotype that is determined by a threshold of sensitivity and a continuously distributed underlying trait called the liab ility. In this case, traits that are correlated with the liability may show a nonlinear relationship due to the dichotomy of expression at the phenoty pic level. for example, in wing dimorphic insects fecundity of the macropte rous (long-winged) females appears in part to be determined by the allocati on of resources to the night muscles, which are almost invariably small or absent in the micropterous (short-winged, nightless) females. Pedigree anal ysis of the cricket Gryllus firmus has shown that wing morph, fecundity and the trade-off between the two have additive genetic (co)variance. It has a lso been shown that selection on proportion macroptery produced an asymmetr ic correlated response of fecundity. The present paper details the results of direct selection on fecundity and the correlated response in proportion macroptery. Selection for increased fecundity resulted in increased fecundi ty within both wing morphs and a correlated decrease in proportion macropte ry. Similarly, selection for decreased fecundity resulted in a decrease wit hin morphs and a correlated increase in the proportion of macropterous fema les. This provides additional evidence that the trade-off between fecundity and wing morphology has a genetic basis and will thus modulate the evoluti on of the two traits.