Outdoor stream channels were treated with a commercial neem formulation, Ne
emix(R) 4.5, and a neem extract (no formulation ingredients) to determine t
he effects on aquatic insect communities. An exposure period of 5 h was cho
sen to simulate the transient nature of pesticide residues in streams and r
ivers after aerial applications. Multivariate analytic procedures based on
Bray-Curtis similarity matrices were used to compare community structure am
ong the replicate channels 9 d after treatment. No significant differences
in community structure were found among controls and channels treated with
Neemix at an azadirachtin concentration of 0.28 mg/L, but a multivariate me
asure of community stress indicated an increase in variability among treate
d channels. Significant differences in community structure were found among
controls and channels treated with Neemix at 0.84 and 2.54 mg/L, and this
resulted from reductions in several hey taxa. During subsequent experiments
with a neem pow der extract. the formulation ingredients of Neemix were at
least partially responsible for the significant effects on community struc
ture at 0.84 mg/L azadirachtin. No significant differences were found among
controls and channels treated with the extract at 0.9 mg/L, whereas the co
mmunity structure of aquatic insects in channels treated at 3.0 mg/L differ
ed significantly from controls. In a Canadian forest pest-management contes
t, the expected environmental concentration in water bodies of areas spraye
d with azadirachtin at 50 g/ha is 0.035 mg/L.