Sex-Investment Ratios in Ants: Has Female Bias Been Systematically Overestimated?

Abstract

The claim by Trivers and Hare (1976) and Nonacs (1986a) that the ratio of sexual investment in monogynous ants is as much female-biased as 3:1 on the average is argued to be untenable because of (1) bias in the gyne-to-male dry-weight cost ratios, as a result of increasingly diverging rates of respiration and fat accumulation with increasing sexual dimorphism; and (2) bias in the numerical sex ratios estimated from small samples. Partial regressions of the numerical gyne-to-male (f/m) sex ratio (S) on the individual dry-weight ratio (D) and sample size (N) appeared to be consistently present in the data sets for monogynous and polygynous ants, although significantly so only for the monogynous ants. These relationships were primarily interpreted as methodological artifacts, but possible biological explanations are also discussed. The alternative explanations, however, are less well supported by data and/or apply to only parts of the data set. Under the assumption that dry-weight cost ratios are unbiased only for species without sexual dimorphism, an approximate formula was derived to transfer dry-weight f/m cost ratios (D) into approximately energetic f/m cost ratios (C). The formula C = D0.7 was found to give a reliable quantitative prediction of the difference between the individual dry-weight ratio and the energetic-cost ratio in Lasius niger, which was earlier estimated empirically (Boomsma and Isaaks 1985). After adjustment for the assumed artifacts, the geometric-mean investment ratio across monogynous ants was estimated to be 1.82:1 in favor of gynes and to differ significantly from both 1:1 and 3:1. The adjusted pattern is consistent with the hypothesis that workers control the sex-allocation process throughout monogynous ants, but in a context of variable degrees of relatedness asymmetry because of the frequent occurrence of multiple mating and worker reproduction.