Sperm Competition in Fishes: The Evolution of Testis Size and Ejaculate Characteristics

TLDR

Fishes display a broad spectrum of sperm competition intensity, and theoretical models predict that sperm length should either increase or remain constant depending on fertilization mode. The study compares species to test how sperm competition intensity influences spermatogenic investment and sperm number and size. The authors conduct a comparative analysis across species measuring gonadosomatic index, sperm counts, and sperm dimensions. Across species, higher sperm competition intensity is linked to larger gonadosomatic index and sperm numbers but shorter sperm, supporting a raffle‑based competition model, while sperm longevity declines with sperm length and the negative sperm length–competition relationship contradicts both continuous and instant fertilization predictions.

Abstract

Fishes show one of the widest ranges of sperm competition intensity of any animal group. Here we present a comparative study whose aim is to investigate the effect of relative intensity of sperm competition on investment in spermatogenesis and the number and size of sperm produced. We find that both the gonadosomatic index (GSI = [gonad weight/body weight] x 100) and sperm numbers increase with intensity of sperm competition across species but that sperm length decreases. These new findings are consistent with a raffle-based mode of sperm competition in fishes. Most of these results (positive correlation of the GSI and sperm number with sperm competition intensity) concur with the predictions of current sperm competition theory. However, we also find that sperm longevity decreases with sperm length across species. Current models for continuous fertilization suggest that if length increases a sperm's speed but decreases its longevity, sperm length should increase with sperm competition intensity, whereas models for instant fertilization suggest that sperm length should remain constant. The negative relationship found between sperm competition and sperm length therefore does not fit predictions of either model.