Abstract
Mosquitofish (Gambusia affinis) embryos from two controlled laboratory experiments and five wild populations were individually weighed to: 1) document the magnitude and pattern of variation in propagule size; and 2) test theoretical predictions regarding optimality vs plasticity in propagule size. A large amount of variance in embryo size was documented in all three samples. Among- and within-female variance in dry weight was high for every stage of embryonic development. Less than one-half of the overall variance was attributable to developmental stage, maternal size, clutch size or treatment effects. Because of the high variance and the relatively weak tradeoff found between embryo size and number, selection for an optimal size was not supported by these data.