Field Experiments on the Relationship between Drift and Benthic Densities of Aquatic Insects in Tropical Streams (Ivory Coast). II. Cheumatopsyche falcifera (Trichoptera: Hydropsychidae)

Abstract

(1) Gutters containing natural substrata colonized by benthic macroinvertebrates were exposed in streams in the Ivory Coast (West Africa) and drifting organisms were captured in 790 samples over a period of more than 600 h. (2) Drift was related to density and to abiotic factors for instars of the hydropsychid caddis Cheumatopsychefalcifera (Ulmer), of which about 55 000 were caught. (3) Drift of larvae peaked at different times at night. Emerging adults and pupal exuviae appeared briefly after sunset. Exuviae of larval instars I-IV were present throughout 24 h. (4) The relationship of drift (y: drift out of 0.1 m2 per unit time or individuals per 1 m3) of a larval instar to its own benthic density (x) differed in its significance at various times and between larval instars. (5) No consistent relationship was found in instar I. In later instars it was best described by y = a + blxb2x4 (instar II); y = axb (instar III); y = aebx (instars IV and V). Since instars IV and V predominated in our trials, an exponential relationship was also found for total larvae of the species. (6) Self-regulation of an upper benthic density by emigration through drift was statistically not evident. (7) Models based on benthic densities explained up to 99% of the variation in the drift; the drift of a larval instar was often defined more precisely by the benthic density of another instar than its own. (8) We conclude that a net of behavioural interactions in the benthic population had a large effect on the loss of individuals through drift. (9) Drift increased with increasing water temperature and velocity in the gutter; moonlight had no depressant effect on the drift. (10) Minor manipulations of the gutter, less drastic than in laboratory experiments, resulted in artifacts in the drift. (11) The average percentage decrease of density through drift over 24 h was 5.6, 9.9, 17-0, 14-3. 5.1, and 9.2 in instars I, II, III, IV, V, and total larvae. (12) Moulting rates, estimated from larval exuvial drift, indicated that the development period for the instars I-IV took more than a month. (13) Drift was reduced in some cases in streams previously treated with insecticides. Drift as determined in the West African Onchocerciasis Control Programme to monitor