Benthic invertebrate faunas were compared to two fifth‐order streams, the Atigun River flowing northward and the Dietrich River flowing southward. Sixty‐eight taxa were collected, forty‐nine from each stream. Aquatic insects comprised 88% of the taxa and 97% of the individuals from the Dietrich River and 73% of the taxa and 97% of the individuals from the Atigun River. Diptera, especially Chironomidae, were most abundant. Plecoptera, Ephemeroptera, Oligochaeta, Acarina, and Collembola were significant. In both streams the headwaters were dominated by the subfamily Diamesinae which was replaced by Orthocladiinae downstream. Diversity seemed to increase with stream order. Cluster analysis showed a high degree of resemblance between the benthic faunas of the rivers. Faunal resemblance decreased with increasing distance between stations, both within and between the streams. Although some taxa may occur in only one stream, the evidence for faunal resemblance is stronger than for faunal differences. Other studies have shown that differences in total radiation associated with valley aspect affect local climate, hydrology, and distribution of terrestrial plants and animals. However, the benthic faunas of Atigun and Dietrich Rivers were remarkably similar. Factors which operate independently of aspect, possibly freezing solid in winter, may control the occurrence of species in these streams.

Simultaneous hourly net collections in a meadow and canyon reach of a mountain stream determined die1 and spatial abundances of drifting Chironomidae larvae. Sixty-one taxa were identified to the lowest practical level, 52 in the meadow and 41 in the canyon. Orthocladiinae was the most abundant subfamily with 32 taxa and a 24 h mean density of 294 individuals 100 m-3 (meadow) and 26 taxa and a mean of 648 individuals 100 m-3 (canyon). Chironominae was the second most abundant subfamily. Nonchironomid invertebrates at both sites and total Chironomidae larvae (meadow) were predominantly night-drifting.Parakiefferiella and Psectrocladius were day-drifting (meadow) whereas 8 other chironomid taxa (meadow) and 2 taxa (canyon) were night-drifting. All others were aperiodic or too rare to test periodicity, Stempellinella cf brevis Edwards exhibited catastrophic drift in the canyon only. The different drift patterns between sites is attributed to greater loss of streambed habitat in the canyon compared to the meadow as streamflow decreased. Consequent crowding of chironomid larvae in the canyon caused catastrophic drift or interfered with drift periodicty. This study adds to knowledge of Chironomidae drift and shows influences on drift of hydrologic and geomorphic conditions.

Nested nets of three different mesh apertures were used to study mesh-size effects on drift collected in a small mountain stream. The innermost, middle, and outermost nets had, respectively, 425 pm, 209 pm and 106 pm openings, a design that reduced clogging while partitioning collections into three size groups. The open area of mesh in each net, from largest to smallest mesh opening, was 3.7, 5.7 and 8.0 times the area of the net mouth. Volumes of filtered water were determined with a flowmeter. The results are expressed as (1) drift retained by each net, (2) drift that would have been collected by a single net of given mesh size, and (3) the percentage of total drift (the sum of the catches from all three nets) that passed through the 425 pm and 209 pm nets. During a two day period in August 1986, Chironomidae larvae were dominant numerically in all 209 ,um and 106 pm samples and midday 425 pm samples. Large drifters (Ephemerellidae) occurred only in 425 pm or 209 pm nets, but the general pattern was an increase in abundance and number of taxa with decreasing mesh size. Relatively more individuals occurred in the larger mesh nets at night than during the day. The two larger mesh sizes retained 70% of the total sediment/detritus in the drift collections, and this decreased the rate of clogging of the 106 pm net. If an objective of a sampling program is to compare drift density or drift rate between areas or sampling dates, the same mesh size should be used for all sample collection and processing. The mesh aperture used for drift collection should retain all species and life stages of significance in a study. The nested net design enables an investigator to test the adequacy of drift samples.

Artificial substrates were designed using rock filled polyethylene bags which were perforated with holes. The substrates trapped waterborne sediment and detritus which enhanced microhabitat complexity. Colonization was compared in side‐by‐side tests with multiple plate samplers in mountain streams ranging from second to seventh order. After 41 days the bag samples contained more sediment and detritus and more animals than did multiple plates. Plastic bags exceeded multiple plate samples by a factor of nearly 8 for individuals and 1.5 for taxa expressed as numbers/sampler. Although detritus amounts differed significantly between samplers, catch composition was similar in habitat preference and functional groups. Most taxa were “lotic erosional” or “lotic erosional‐depositional” detritivores. The plastic bags better represented the streambed fauna judged by their greater similarity to dip net samples. Bag samplers had 4.5 × the colonization area of multiple plates, hence would be expected to support more species. Catch/m2 of colonization area was not significantly different between samplers. Functionally the plastic bags act as detritus retention devices, offering a diverse, highly dynamic microhabitat for colonization. Results are interpretable in terms of research on microdistribution of stream benthos and the river continuum model. This study supports the conclusion that stream benthos abundance and diversity are related to the amount of detritus. Maximum diversity and numbers of individuals occurred in samples from third and fourth order streams. Grazers reached peak abundances in the same streams where the continuum model predicts P>R Shredders reached maximum abundances in third and fourth order streams where the riparian canopy was greatest. Predator abundance changed little with stream size. Although bag samples required more sorting time, the samplers are catch effective, inexpensive, and adaptable.