Extrinsic effects on fecundity-maternal weight relations in capital-breeding Lepidoptera

"Capital-breeding Lepidoptera depend for reproduction on metabolic resources assembled either entirely or primarily by their larvae, the former termed 'perfect' the latter 'imperfect'. Empirical evidence suggests that maternal size determines capital-breeder fecundity. The fecundity-maternal size relation is usually formulated as F = bW + a, where F is fecundity, W is final maternal size in units such as weight of newly transformed pupae, b is the slope, and a the intercept. Exhaustive search yielded 71 fecundity-maternal pupal weight relations for 41 capital breeders in 15 families, 58 of which, including 2 previously unpublished, were based on individual specimens, and 13 on grouped specimens. In 22 individual-specimen relations, cohorts divided into 2 or more subgroups were reared simultaneously at different temperatures, on different diets, or exposed to other extrinsic factors. These 22 'multiform' relations were compared with 36 'uniform' relations, and where possible cohort subgroups were compared. Pupal weights of cohort subgroups were affected much oftener than underlying slopes and intercepts. Individual-specimen slopes based on transformed data ranged 0.52-2.09 with a mean and standard error of 1.13±0.04, and slopes did not differ significantly among perfect, imperfect, multiform, and uniform categories. Despite the evident similarity, one relation does not apply to all capital breeders. Tradeoffs sometimes occur between fecundity, F, and mean egg weight, E. Reaction norms of fecundity and pupal weight across extrinsic-factor ranges were overwhelmingly congruent, which supports axiomatic status for the dependence of fecundity on capital-breeder maternal size. Cooler rearing temperatures usually produced heavier female pupae and greater fecundities, a phenomenon of population dynamics interest. The two sides of practically all fecundity-maternal weight regressions are not statistically independent, in effect stating F = b(W + [F × E]) + a, which artificially inflates test statistics. Where desirable, the fully independent relation R = b(W - [F × E]) + a can be used, where R is reproductive bulk, the mathematical product of F × E."