LIFE-HISTORIES, SECONDARY PRODUCTION, AND MICRODISTRIBUTION OF HEPTAGENIID MAYFLIES (EPHEMEROPTERA) IN A TROPICAL FOREST STREAM

The ecology of five species of heptageniid mayflies was investigated i n Tai Po Kau Forest Stream (TPKFS), New Territories, Hong Kong, over a two-year period (1977-79). Quantitative benthic samples were taken fr om different microhabitats in the stream on 47 occasions and, in combi nation with light-trap catches of adult mayflies, were used as a basis for investigating life-history patterns and estimating secondary prod uction. Electrogena sp. (49.8 individuals 0.5 m(-2); 60% of total hept ageniid densities) and Cinygmina sp. (15.7 individuals 0.5 m(-2)) were the most abundant heptageniids in TPKFS, followed by Epeorus sp. (9.1 individuals 0.5 m(-2)). Iron sp. and Paegniodes cupulatus, which were present in relatively low densities, made up < 10% of total heptageni id abundance. In terms of mean biomass (1977-79), Electrogena sp. and Cinygmina sp. were equally important (10.3 vs. 10.5 mg ash-free dry we ight (AFDW 0.5 m(-2)), while Epeorus sp. was ranked third (6.8 mg AFDW 0.5 m(-2)). There was, however, considerable inter-year variation in heptageniid standing stocks: mean densities and biomass of Cinygmina s p. in 1977-78 were (respectively) eight and nine times greater than in 1978-79, while Iron sp. and P. cupulatus exhibited the converse trend , with densities more than twice as high and biomass more than three t imes greater in 1978-79. Heptageniid abundance in TPKFS seemed to foll ow an annual pattern of wet-season decrease and dry-season increase, b ut this tendency was more apparent in some species (e.g. Iron sp.) tha n others (e.g. Cinygmina sp.). The life histories of TPKFS heptageniid s were unclear. All species showed rather stable larval size-frequency distributions throughout the study period, which was indicative of as ynchronous growth and continuous recruitment. Adult flight periods did not exhibit strong seasonality. Calculations of secondary production were based upon the conservative assumption that each species was univ oltine, but multivoltine life histories may have been possible for the se tropical mayflies. Total heptageniid annual production over the per iod 1977-79 was 212.0 mg AFDW 0.5 m(-2); and was slightly higher durin g 1977-78 than in 1978-79 (225.0 vs. 202.1 mg AFDW 0.5 m(-2)). Inter-y ear variations in production reflected variations in annual mean bioma ss: production of Cinygmina sp. was 7.9 times greater in 1977-78 than in 1978-79, but production of Iron sp. was 2.7 times higher during 197 8-79, and that of P. cupulatus was 4.6 times greater. As a consequence of these inter-year variations, species-specific production decreased in the order Cinygmina sp. > Electrogena sp. > Epeorus sp. > Iron sp. > P. cupulatus in 1977-78, but the sequence was Electrogena > Iron sp . > Epeorus sp. > P. cupulatus > Cinygmina sp. in 1978-79. For the stu dy period as a whole (1977-79), the order was Electrogena sp. > Cinygm ina sp. > Epeorus sp. > Iron sp. > P. cupulatus. Production:biomass (P :B) ratios varied from 4.8 (Cinygmina sp., 1977-78) to 11.1 (Iron sp., 1977-78), and P:B ratios (1977-79) decreased in the order lion sp. > Epeorus sp. > Electrogena sp. > P. cupulatus > Cinygmina sp. Productio n calculations and P:B ratios are highly sensitive to the conservative assumptions made about heptageniid voltinism in TPKFS, and therefore may be significant underestimates. Epeorus sp., Iron sp., and P. cupul atus were most abundant in midstream microhabitats but Cinygmina sp. w as more numerous close to the stream banks. The microdistribution patt ern of Electrogena sp. was less distinct, but larvae seemed to favour midstream microhabitats. Multivariate statistics revealed that the mic rodistribution of all heptageniids (except Electrogena sp.) was influe nced significantly by sediment grain-size characteristics rather than by algae or detritus, but the proportion of variation (4-14%) in the a bundance of each heptageniid species accounted for by sediment charact eristics was low.