7405 - Laserfiche WebLink

Home Browse Search

16 BIOLOGICAL REPORT 90(5) Fish Community Structure and Stability in Warmwater Midwestern Streams by William J. Matthews University of Oklahoma Biological Station Kingston, Oklahoma 73439 Impact assessment generally implies a search for ways to measure and predict changes in stream systems that ensue from cultural alterations. In order to really know the degree to which a cultural activity results in change in stream systems, some- thing must be known~of the magnitude of change or variation that occurs naturally. Even if managers are constrained from long-term studies of each stream of interest, the long-term data sets that do exist can be used to predict the kinds and degree of changes that may naturally occur in a variety of stream systems. Over the last two decades, a relatively long-term data base has been accumulated from our collecting activities assessing fish assemblages and faunas in three midwestern streams: Piney Creek, Izard County, Arkansas; Brier Creek, Marshall County, Oklahoma; and Kiamichi River, southeastern Okla- homa (Ross et al. 1985; Matthews 1986; Matthews et al. 1988). These data sets have been evaluated at two levels: stability of whole-stream faunas (based on pooled collection data at multiple sites on each stream) and stability of fish assemblages at each individual site over time. The data set spans 1969-86 for Brier Creek, 1972-86 for Piney Creek, and 1981-86 for Kiamichi River. Each of these time intervals allows for one to two complete generations of at least the small fish species that are most abundant. The whole-stream faunas of all three streams have been highly persistent (in terms of presence and absence of species) and stable (based on abundance data) in all years of study. No common species has been lost from any of the streams, and no "rare" species has become common at any of the sites. Both similarity indices (Morisita's and percent similarity) and statistical rank correlation of species abun- dances show that all three streams have been stable, and that the most environmentally stable stream (Piney Creek of the Ozark uplands) has had the greatest degree of fish population stability across all years. At individual sites on each of the streams, the local fish assemblage has had a tendency to be stable, although exceptions exist in all three systems. Where individual collecting sites have shown a marked change in the composition of the fish assem- blage, this change can usually be attributed to known events, such as deposition of large woody debris by floods, death offish in intermittent headwaters due to drought, or other such disturbances. However, recolonization of intermittent headwaters is not a random phenomenon. Matthews (1987) showed that existence and persistence of fish species in harsh headwaters of at least one of the streams was directly related to their ability to withstand low oxygen conditions, which are common. Thus, there is both a chance component and a directed compo- nent to dynamics of assemblage composition in head- water streams. The long-term study of Piney Creek included a flood event that was clearly of 100-year magnitude. Vertical water levels were elevated as much as 12 m at some sites, and physical destruction and distur- bance of habitat along the creek was extreme. Im- mediately after the flood, fish assemblages showed some differences from those that existed before the flood, but after 8 months, both the whole-stream fauna and most local assemblages were statistically indistinguishable from their composition before the flood. This well-documented event-and other ac- counts of flood effects on fish in midwestern streams (Gelwick 1990; Harvey 1987)-indicates that while floods may have an immediate effect on composition of midwestern fish communities, these changes are largely transient, with the systems returning to preflood conditions rather rapidly. Some methodological recommendations are pos- sible from collections made on these streams. There is a longstanding question in fish community assess- ~