PREDATION, FOOD LIMITATION, PHENOTYPIC PLASTICITY AND LIFE-HISTORY VARIATION INGAMBUSIA HUBBSI

Abstract

Langerhans et al. (2007) present molecular data, morphological variation of both males and females, and female mate preferences as evidence that ecological speciation, mediated by differences in exposure to predation by piscivores (Gobiomorus dormitor as well other piscivorous fish), occurred among populations of Gambusia hubbsi occupying blue holes on Andros, Commonwealth of the Bahamas. Their conclusion rests on two critical assumptions, namely, that previous studies of G. hubbsi (Krumholz 1963; Sohn 1977; Downhower et al. 2000) provide “. . . evidence for strong divergent selection between predator regimes.” and that “observed differences are unlikely to merely reflect environmentally induced phenotypic variation as morphological differences . . . between populations within species exhibit a strong genetic basis.” The first assumption is essential to their claim that differences in body shape are due to differences in exposure to predation and hence predator avoidance. The second is essential if their observed mate preferences are to have evolutionary consequences. Previously published studies show these claims to be false, at least for G. hubbsi on Andros Island (Downhower et al. 2000, 2002), and invalidate Langerhans et al.’s conclusions regarding predation and its role in presumptive ecological speciation in this species. With regard to “strong divergent selection between predator regimes,” Krumholz collected fish from just two localities, a “tidal pool” in which piscivores were present and a man-made, brackish water pond with no predators but where phytoplankton blooms reduced visibility to “as little as one foot.” As a consequence, any conclusions regarding the role of predation in shaping differences between these two populations are confounded by other edaphic and ecological factors (see Hurlbert 1984). We note that other piscivore-present, piscivore-absent comparisons cited by Langerhans et al. are similarly confounded. Sohn’s laboratory study used fish collected “. . .from a pond on South Bimini. . .” No further details are given, and no piscivores were used in his experiments. Rather, he showed that a juvenile male reared in the presence of an adult male delayed maturation until it is larger than the adult, and when juvenile males are raised together, the second male to mature is larger than the first male to mature. Because male poeciliids grow little after their gonopodium has matured (Turner 1941), Sohn’s findings would be expected to apply to any poeciliid in which male size affects the outcome of competition with other males for access to females (Borowsky 1973; Hughes 1985). Hence, neither Krumholz nor Sohn provides evidence that predation shaped the differences they reported. More seriously Langerhans et al. also claimed that our work provides “evidence for strong divergent selection between predator regimes.” In fact, we rejected that hypothesis and demonstrated remarkable phenotypic plasticity in this species. Our dataset,