Sex matters for defence and trophic traits of threespine stickleback

Background: The extent and importance of sexual dimorphism in non-sexually selected traits has received considerable attention in studies of birds and reptiles but similar studies in fishes are limited. Aim: Quantify the extent and direction of sexual dimorphism in both defence and trophic traits of adult threespine stickleback (Gasterosteus aculeatus). Methods: We amassed data from research done on 103 ecologically pristine stream and lake localities on the Haida Gwaii archipelago, western Canada. Using these data, we tested many hypotheses of ecological correlation with variability in the amount of dimorphism. We used a nested ANOVA with principal components and individual size-corrected morphometric traits as the dependent variables, with sex (male, female), water clarity (stained, clear) and water body (stream, small lake, large lake) as fixed factors, and locality as a random factor. Results: Adult body length (SL) is mainly female-biased but this is accentuated in large clear water lakes. Dorsal and anal fin ray numbers are male-biased, the dimorphism being greater in clear water than stained lakes. The number of bony lateral plates is male-biased in most habitats except for large clear water lakes, which are female-biased. Relative size of the pelvis (spine length, ventral plate length) is female-biased in most habitats. Relative head length, head depth, jaw length, and eye diameter are male-biased in most populations. Dual isotopic signatures of muscle tissues of adult male stickleback in 33 populations show a higher trophic level (delta N-15) but no consistent differences in delta C-13. In a single locality where multiple size classes were located, males show a steeper ontogenetic slope of delta N-15 against SL than do females, resulting in their signatures being depleted relative to sub-adult females but enriched relative to adult females. Stickleback from two localities sampled over several decades exhibit substantial temporal variability in the extent of the isotopic sexual dimorphism. Conclusions: Sexual dimorphism in stickleback is widespread in multiple morphometric traits not directly associated with sexual selection. Population differences in the extent and direction of dimorphism are related to niche separation across predator and trophic gradients within localities and reflect general diversifying selection within populations.