Meta-analyses have shown that benthic feeding fish tend to exert strong neg
ative effects on densities of stream-living chironomids, but it is not know
n whether this pattern reflects variation in consumption rates or prey move
ment rates. We reanalysed data from an experimental study that showed stron
g effects of a benthic Feeding fish (Cottus gobio) On chironomid densities.
Gut content data suggested that chironomids were rarely consumed by sculpi
ns. Path analysis suggested that indirect interactions mediated by net-spin
ning polycentropodid caddis larvae (fish-net-spinners-chironomids) were muc
h more important than the direct interaction (fish-chironomids).
The effects of the net-spinning caddisfly larva Polycentropus flavomaculatu
s on chironomid densities were examined in a manipulative field experiment.
Artificial substrates, stone-filled plastic baskets, were exposed in a str
eam for 5, 14 or 26 d. P. flavomaculatus had a pronounced positive effect o
n the densities of most chironomid taxa. The experimental design allowed us
to separate the effects of P. flavomaculatus nets from the effects of the
larvae. Observed effects were primarily due to the presence of the nets. Ne
gative effects of P. flavomaculatus larvae, which could be caused by direct
predation, were not found. We conclude that sculpins reduced chironomid de
nsities through an indirect interaction involving a negative effect of scul
pins on net-spinning caddis larvae and a positive effect of caddis larval n
ets on chironomid densities.
To examine the potential generality of this indirect interaction, we review
ed experiments examining effects of benthic feeding fish on chironomid dens
ities. A majority of the experiments were performed in systems where net-sp
inning caddis larvae either were absent or occurred in low numbers. However
, one third of the experiments reported negative effects on net-spinners an
d was performed in systems with reasonably high densities of net-spinners,
suggesting that this mechanism may be more common than is presently believe
d.