SEMIVOLTINISM, SEASONAL EMERGENCE, AND ADULT SIZE VARIATION IN A TROPICAL STREAM MAYFLY (EUTHYPLOCIA-HECUBA)

Life history characteristics of the stream mayfly Euthyplocia hecuba ( Hagen) (Polymitarcyidae:Euthyplociinae) were studied over a 4-yr perio d in two tropical streams (Rio Tempisquito and Quebrada Marilin) drain ing primary evergreen forest in northern Costa Rica. Larvae burrow in the substratum of the stream, living under small to large cobbles that are firmly embedded in the stream bottom. Euthyplocia hecuba is sexua lly dimorphic and one of the largest mayflies of Central America, with maximum larval size of 149 mg (dry mass) for females and 35 mg for ma les. Mature female and male larvae are about 40% larger in Q. Marilin relative to R. Tempisquito. Subimagoes emerge about 1 h before dawn (s imilar to 0400 h); the molt to the image (or true adult stage) occurs similar to 20 min later. Adult emergence and reproduction occurs mainl y between June and November. Females mate and then oviposit in riffle areas just before dawn, with the time period gradually changing from 0 446 and 0507 h in June to 0508 and 0529 h in October. We found that si ze of adult males and females gradually declined during the emergence period. Fecundity averaged 984 in October (range 364-2851), with eggs being very large for a mayfly(0.46 x 0.4 mm; 0.017 mg dry mass). The p roportion of adult biomass allocated to eggs (reproductive effort) ave raged 0.51 but varied according to female size. Median egg development time was 113, 55, and 31 d at 15, 20, and 25 degrees C. The hatch suc cess of eggs was >90% at 15 and 20 degrees C, only 68% at 25 degrees C , and 0% at 10 and 30 degrees C. First instar larvae are relatively la rge for a mayfly, having a head width of 0.11 mm and a body length of 1.30 mm. Seasonal changes in larval size distribution suggest that lar val development takes similar to 22 mo to complete and may involve a d evelopmental quiescence or diapause during the first year. A 2-yr life history appears to be the most parsimonious hypothesis given a 2-mo e gg development time, the occurrence of two size cohorts of larvae thro ughout most of the year, and the temporal pattern of maximum and avera ge larval sizes over the course of the sampling period. Our data sugge st that seasonality in the developmental dynamics of larvae underlies both the seasonal emergence and decline in dry mass of both adult male s and females during the emergence period. The data provide the first evidence for semivoltinism in a tropical mayfly.