A field experiment showed that benthic invertebrates in running water exhibit preferences for different substrate particle sizes. Maximum numbers and biomass occurred on medium gravel (24.2‐mm mean diam), whereas diversity was greatest on large gravel (40.8 mm). Individual species fell into four groups: upper, medium, and lower size preference, and no preference. The addition of a limited quantity of sand to medium gravel affected only a few species.

Responses of stream periphytic algae and benthic insects to increases in dissolved inorganic phosphorus (P) were assessed in a streamside mesocosm. Controls and treatments were colonized continuously in summer by biota from the stream. P was maintained in the treatments at 10 μg∙L−1, [Formula: see text] times the concentration in the controls. In the treatments the biomass of chlorophyll a reached 3.5 times that of the controls, accompanied by an increase in Cyanophyta relative to diatoms. No difference was detectable in the numbers of insects drifting from controls and treatments. Numbers of individuals emerging (> 40 species) from the treatments over 7 wk were 2.2 times those from the controls. In both controls and treatments, 77% or more of emerging insects were Chironomidae. After week 7 the density of benthic insects in the treatments, determined without size selection, was 1.75 times that of the controls; size distributions in treatments and controls were similar. Addition of P, therefore, increased the food of insects and resulted in a doubling of their survival to emergence.

The minimum unit of the environment containing the essential processes of smolt production is identified as a riffle–pool sequence at a meander. Optimization for smolt rearing is seen to lie through control of discharge, temperature, food production, and cover. The pathways of food production are traced. Ways of increasing the stock of fish-food organisms include adjustment of the ratio of riffle to pool area, choice of streamside vegetation, control of light to the stream, and inorganic and organic enrichment. Ways of making food available to fry include collecting drifting invertebrates at night and releasing them in the day, dislodgement of benthos, attracting aerial insects by lights, and supplementary feeding with artificial foods. None of these procedures, when applied to a natural stream, seems economically feasible. When applied, however, to channels made alongside streams and stocked by the parent stream, they should combine some desirable features of rivers with the productive capacity, but not the costs, of hatcheries.

Stream bed materials, both biotic and abiotic, in the size range 50 μ–ca. 200 mm can be sampled unselectively, in shallow streams, with a simple inexpensive apparatus consisting of a box provided with an adjustable upstream inlet, and, downstream, two nets, one within the other. Collected materials are wet-sieved and the volumes of inorganic material passed by successively finer sieves are plotted as cumulative curves against a logarithmic size scale. Curves are given for materials from three contrasting habitats: a riffle, and pool, of a coastal stream, and an artificial spawning channel. Examples are also given of the densities, size distribution, and vertical stratification of invertebrates from these habitats. Applications of the method to studies on fish biology, invertebrate ecology, and geomorphology are indicated.

No abstract