Stream sites in the Willamette Basin have been categorized on the basis of relative abundance of fish species (percentage of total fish collected at a given site) by Waite and Carpenter (in press). They distinguished four categories of sites:
Although land use influences upstream from the highly impaired sites were primarily agricultural, the pie chart shows that fish communities at these sites were quite different from those at other agricultural sites.
A total of 29 fish species, including 10 non-native (introduced) species, were collected from the 24 sites during 1993-95. The most abundant species, in decreasing order of relative abundance, were reticulate sculpin, redside shiner, torrent sculpin, speckled dace, yellow bullhead, and cutthroat trout.
Forested reference sites were characterized by high abundances of trout (cutthroat, rainbow) and sculpin (mottled, Paiute, torrent, reticulate); no external anomalies were found on fish from these sites. Fish communities from the other three categories consisted of few or no trout and generally low abundances of pollution sensitive sculpin species (including mottled, Paiute, torrent). As seen in the bar graph, forested sites exhibited the largest riffle areas, most extensive canopies with best riparian quality (greatest species and diversity of vegetation and largest riparian zone), lowest maximum water temperatures, highest minimum dissolved oxygen (DO) percent saturation, lowest nutrient concentrations (example, total nitrogen), and no detections of pesticides (sum of atrazine, metolachlor and simazine).
Small agricultural/urban sites were characterized by high abundances of pollution tolerant fish species (including minnows and reticulate sculpin), low levels of external anomalies, and low abundances of introduced species, such as bullheads, carp, and sunfish. These sites exhibited small riffle areas, moderately open canopies of moderate riparian quality, relatively high maximum water temperatures, lowest minimum DO percent saturation, and highest nutrient and pesticide concentrations.
Fish communities varied significantly among site categories.
Large agricultural sites were characterized by relatively high abundances of torrent and reticulate sculpin, minnows, and introduced species, with moderate levels of external anomalies. These sites exhibited relatively small riffle areas, open canopies of low riparian quality, high maximum water temperatures, and intermediate nutrient and pesticide concentrations.
The highly impaired sites--Long Tom River, Bear Creek (tributary to the Long Tom River), and Little Muddy Creek (located north of Eugene and east of the Willamette River)--were characterized by high abundances of introduced species and tolerant reticulate sculpin, and high levels of external anomalies. The Long Tom and Little Muddy sites have been channelized; discharge of the Long Tom River is regulated by an upstream dam; and Bear Creek had poor quality instream habitat. All three sites exhibited relatively low concentrations of nutrients and pesticides; however, the highest maximum water temperatures and percent open canopies, poorest riparian quality, smallest riffle areas, and low minimum DO percent saturation were found at these sites.
Site categories derived from fish relative abundances reflect average habitat conditions and water chemistry at the sites.
From a basinwide perspective, relative abundances of native fish species were most correlated with the quality of instream and riparian habitat (Waite and Carpenter, in press). Trout and sensitive sculpin species were common at forested reference sites characterized by abundant riffles, closed canopies, and lowest water temperatures. Tolerant reticulate sculpin and minnows dominated fish communities at small agricultural/urban sites, where riffles were rare, canopies were relatively open, and water temperatures were higher than at forested reference sites.
Within the small agricultural and urban sites, water chemistry became more important than habitat in controlling fish communities. For example, nutrient and pesticide concentrations and DO percent saturation were critical for understanding fish distributions (Waite and Carpenter, in press). Also, at highly impaired sites, high water temperatures and low DO percent saturation were correlated with high relative abundance of introduced species and with external anomalies, such as anchor worms and lesions.
External anomalies were nonexistent at forested reference sites, greatest at highly impaired sites, and intermediate at remaining sites.
On the basis of other studies in Oregon and throughout the United States, occurrences of introduced species and external anomalies at rates greater than 2 percent are considered an indication of impaired conditions (Hughes and Gammon, 1987; Karr, 1991). The large agricultural and the highly impaired sites were characterized by introduced and tolerant fish species with relatively high percentages of external anomalies (9 and 25, respectively). Although small agricultural/urban sites exhibited smaller percentages of introduced species and lower numbers of anomalies (2 percent each), these sites also showed higher numbers of tolerant fish and generally were dominated by one or two species--another indication of impairment.
An Index of Biotic Integrity (IBI) (Hughes and others, in press), which expresses fish community structure numerically, has been developed for wadable streams in the Willamette Valley and was applied to the small agricultural/urban sites in the current study. The results suggest that all sites were at least moderately impaired; the five sites with the highest percentages of introduced species also reflected the poorest community structure (lowest IBI scores).
Gales Creek in the Coast Range
Little Pudding River in the Willamette Valley
Instream and riparian habitat differences are apparent when comparing a forested reference site, such as Gales Creek in the Coast Range with an agricultural site, such as the Little Pudding River in the Willamette Valley.
Assessment of fish communities revealed different patterns of impairment among sites and helped develop hypotheses to explain these differences. It also supported the conclusion that identification of all fish species, not just game fish, was needed to allow a complete assessment of ecological health and sound comparisons among all sites. This is demonstrated by the distributions of the four sculpin species commonly collected. Forested reference and small agricultural/urban sites exhibited similar percentages of total sculpins, but the small agricultural/urban sites were dominated by tolerant reticulate sculpin, whereas forested reference sites supported sensitive species that were either less abundant or absent from the other sites.
During spring and summer of 1993-95, fish communities, habitat, and water chemistry were evaluated at 6 intensive assessment and 18 synoptic assessment sites (p. 25). Site locations are shown on the ecoregion map on page 24. Fish were not collected at the most downstream (northernmost) Willamette River site on the map.
Fish were collected within a minimum reach length of 500 feet using backpack and boat electrofishing procedures, and all fish were identified to species (Meador and others, 1993a ).
Habitat measurements were made along four to six transects aligned perpendicular to the stream channels. Thirty-eight instream variables (including discharge, velocity, depth, substrate size, percent riffles), and 40 stream channel and bank variables (including channel dimensions, bank height and stability, riparian quality, percent open canopy, tree size and density) were measured (Meador and others, 1993 b ).
Twenty-four water chemistry variables, including stream temperature, specific conductance, and concentrations of dissolved oxygen (DO), nutrients (nitrate, ammonia, total nitrogen, soluble reactive phosphate, total phosphorus), and dissolved pesticides (sum of atrazine, metolachlor, and simazine) were evaluated during high and low flow.