Stream drift, size-specific predation, and the evolution of ovum size in an amphibian.

A stream-breeding race of small-mouthed salamanders (Ambystoma texanum) in central Kentucky produces ova that are twice as large as those of a pond-breeding race found nearby. Embryos of stream-breeders also hatch at a more advanced developmental stage than those of pond-breeders. Morphological evidence indicates that stream-breeders were derived from pond-breeding stock. Assuming that differences between stream and pond-breeders reflect evolutionary change, and that the ancestral pond stock that invaded streams was similar to extant pond-breeders, we examined three hypotheses that might explain changes in ovum size and stage at hatching following the invasion of streams. (1) Larger ovum size evolved indirectly as a consequence of selection for rapid development which minimizes mortality risk from stream drying. (2) Increased ovum (hatchling) size and stage at hatching of stream-breeders are adaptations to resist stream current. (3) Increased ovum (hatchling) size and stage at hatching are adaptations to reduce predation on hatchlings from stream invertebrates. The results of field and laboratory studies only support hypotheses (2) and (3). Hatchlings that were relatively large or at a more advanced developmental stage had slower drift rates and were less vulnerable to predation by Phagocata gracilis, a flatworm abundant in streams in central Kentucky. Developmental and growth parameters were not correlated significantly with ovum size in populations of either geographic race. Differences in degree of parental care among races also cannot explain variation in ovum size since both races abandon their eggs immediately after oviposition.

Effect of microhabitat on fitness components of larval tiger salamanders, Ambystoma tigrinum nebulosum.

Survivorship and growth of larval tiger salamanders, Ambystoma tigrinum nebulosum, in the White Mountains of east-central Arizona were compared in six microhabitats using field enclosures during summers, 1983-85. Microhabitats were vegetated and nonvegetated shallows, surface, middle, and bottom horizontal layers of limnetic areas, and the vertical limnetic water column. Initial enclosure densities (0.025 larvae per 1) were identical among microhabitats. Three enclosures were placed in each microhabitat in two ponds. Larval survivorship and growth were usually greatest in vegetated shallows in lowest in middle and bottom limnetic enclosures, despite several population dieoffs. Lower fitness, as reflected in survivorship and growth, in these latter enclosures was correlated with lower food levels, temperatures, and oxygen concentrations in deeper limnetic areas. Relative fitness varied greatly between years while food levels, temperatures, and oxygen concentrations within microhabitats remained relatively constant indicating additional factors affected fitness. Disparities in fitness between microhabitats apparently affect habitat use patterns of larvae.