Effects of discharge/substrate stability relationships on the abundance, diversity, and distribution of aquatic insects were studied from 1986 to 1988 in three reaches of a small stream in the Manitoba escarpment. Substrate stability (percentage of streambed paving material at incipient motion) ranged from 80 to 20% at bankfull discharge. Eleven taxa of mayflies, 15 taxa of caddisflies, and 10 taxa of stoneflies were collected. All but a few taxa were present at all stations. Benthic insect densities were negatively correlated with increasing discharge and substrate particle movement at all stations. Regression models describing insect density in terms of substrate stability usually accounted for more variation than models describing density in terms of discharge. Insect density decreased up to 94% at the most unstable stream reach following summer and autumn spates when 8 and 12%, respectively, of the streambed paving material was at incipient motion wheres no reduction occurred in a stable riffle constructed nearby. Most insect species were highly resilient; densities and diversities returned to prespate levels during periods of low flow. Temporal and spatial aspects of substrate stability should be considered in stream insect studies, particularly in relation to disturbance.

None of several samplers compared in limited tests in Lake Ontario gave realistic estimates of benthos in all substrates encountered. The Ponar and Shipek grabs, the only samplers that functioned in gravel, came closest to all-sediment samplers. In sand, however, the Franklin–Anderson grab appeared to be much more efficient than either of these two samplers. In mud the Ekman grabs gave the highest mean numbers of animals per m2.In trials in Lake Winnipeg profundal mud, hand-taken diver's core samples being used as a quantitative standard, only the Fisheries Research Board multiple corer and the standard Ekman grab gave quantitative results for total macrobenthos. However, the multiple corer collected significantly fewer chironomids, and the standard Ekman grab significantly fewer oligochaetes, than did the diver's cores. There were indications that the tall Ekman grab was either not tall enough or too heavily weighted for use in soft sediments and that a small improvement in design could make both this and the standard Ekman more efficient. Most of the samplers apparently sampled some groups of organisms much better than others. Neither the Ponar grab nor the tall weighted Ekman grab sampled the chironomid or oligochaete populations satisfactorily. However, both grabs indicated densities of sphaeriids not significantly different from the diver's samples.

No abstract

A total of 3999 Trichoptera adults, represented by 8 families, 17 genera and 33 species was collected in emergence traps in 1983 and 1984 from five stations on the Ochre River, Manitoba (50°04' N, 99°48' W). Species composition for the two years was comparable, but as the result of a summer spate in 1984, abundance was only 40% of that in 1983.Species diversity by station was negatively correlated with substrate instability of the reach, whereas density per trap was negatively correlated with substrate instability and local factors such as sedimentation in some reaches following peak discharges. Analysis of historical peak discharge records indicated that relatively infrequent mid-summer spates had a detrimental effect on subsequent emergence of the Trichoptera fauna. The combination of spates and unstable stream bed substrate resulting from land use practices in the drainage basin have resulted in an impoverished caddisfly fauna in the Ochre River in comparison with other rivers in Manitoba.

A single 0.3 ppm injection of methoxychlor into the Athabasca River, Alberta on 4 June 1974 for 15 min caused catastrophic drift for a distance of over 400 km, and a subsequent large decrease in the drifting population. This decrease, when expressed as a percentage reduction from pretreatment drift, is in close agreement with percentage reduction of standing crop recorded by other sampling methods. The time required for the pesticide to affect different species varied considerably but was not related to the mode of feeding. Methoxychlor residues above ambient levels in water were recorded in all the invertebrate populations sampled. Caged animals had significantly different residues than the natural populations. The use of caged animals as indicators of environmental damage is therefore questioned.