- Species interactions and their effect on the use of habitat by benthic fishes were examined in a guild composed of Cottus perplexus, C. rhotheus, and C. beldingi in the Marys River above Philomath, Oregon, and in a constant environment provided by an artificial stream. In the Marys River system, the pattern of longitudinal succession in this guild was one of species additions downstream rather than sequential longitudinal replacement. Cottus perplexus occurred throughout the river system, C. rhotheus was found only in the middle and lower reaches, and C. beldingi was restricted to the lower reaches within the study area. Longitudinal distributions appeared to be most strongly related to substrate characteristics in riffles. Cottus perplexus occurred on all substrates, C. rhotheus was found only where rock was present, and C. beldingi was captured only in riffles dominated by loose, cobble-sized rock with many interstitial spaces. Local use of habitat was strongly related to the breadth of environmental tolerance of each species arid the presence or absence of other cottus species. The species broadly overlapped in their use of different substrates, depths, and distances from shore, but segregated in an orderly manner with respect to current velocity. Where it existed alone, C. perplexus utilized both pools and riffles
and was found in greatest densities near cover. Where it coexisted with C. rhotheus, C. perplexus was primarily confined to pools and declined significantly in abundance. The degree to which C. perplexus was excluded from riffles varied in relation to seasons and the size composition of the C. rhotheus population. In areas where there were low densities of large C. rhotheus, the exclusion of C. perplexus was not as pronounced as in areas where there were many large C. rhotheus, and was apparently due mainly to competition. In these areas overlap between the species was also greater in spring, the time of peak food abundance, than in summer or winter. In areas where there were greater numbers of large C. rhotheus, the exclusion of C. perplexus from riffles was more pronounced and did not vary seasonally. In these areas predation was apparently the most important mechanism causing the interactive segregation. Cottus rhotheus was generally
confined to riffles, although some large individuals did wander into pools during times of food scarcity. Cottus beldingi was exclusively found in riffles and had a greater maximum current velocity tolerance than C. rhotheus. In summer, all riffle areas in the Marys River flowed slowly enough for C. rhotheus, and C. beldingi apparently was forced, by predation and competition exerted by C. rhotheus, to burrow into the substrate. In winter and spring, C. rhotheus was excluded from areas of some riffles by the faster currents. tn these areas C. beldingi had a refuge from competition and predation and was found atop the substrate to a greater degree.
Experiments with the three species in the artificial stream supported the field data and suggested that the competition contributing to the habitat shifts was interference competition for
space. Cottus perplexus selected riffle areas when alone, but was displaced partially to pools by C. beldingi through interference competition. The addition of C. rhotheus caused C. perplexus to be more rigidly confined to pools, but C. beldingi was able to coexist in riffles with C. rhotheus apparently by utilizing the interstitial spaces in the substrate. Comparison of these results with those of other studies suggests that segregation in stream fishes should be more commonly based on habitat than on food or time. A model was constructed that could be used to predict patterns of habitat utilization in other guilds of benthic fishes. Species interactions were also suggested as having a major influence on the evolutionary patterns of Cottus in western
North America. Large, dominant species will obtain their preferred habitat throughout their geographical range. They will therefore not need to adapt to a wide range of conditions and thus may exhibit low morphological variability. Contrastingly, less dominant species may be forced into many different habitats, depending on which dominant species are also present, and may thus exhibit greater morphological variability.