Community structure of the aquatic biota—including fish, macroinvertebrates (especially the aquatic larvae of many insects), and algae—is a useful indicator of water quality. Aquatic-community data often are used to assess and monitor environmental quality in an approach termed “bioassessment” or “biomonitoring” (Plafkin and others, 1989). Metrics or numerical indices, based on the distribution and abundance of species, have been used for decades by agencies and citizen-monitoring groups to provide a time-integrated picture of water quality and the responses of aquatic biota. Use of multiple trophic levels and metrics in any evaluation is important because no single metric can adequately reflect the many possible types of effects that can result from changes in water quality. Collection of associated habitat information is important to help differentiate effects due to water-chemistry change from effects due to habitat degradation.

Fish Results

The use of fish-community data for bioassessment and biomonitoring techniques has been shown to be a useful way to detect and quantify environmental degradation in aquatic systems (Lyons, 1992). An Index of Biotic Integrity (IBI) for warmwater streams in Wisconsin (Lyons, 1992) was used for the data collected at the 14 wadeable streams sites (table 26). The Wisconsin version of the IBI was largely derived from the Ohio Environmental Protection Agency “wading sites” version (Ohio Environmental Protection Agency, 1988) of the IBI. The Ohio version in turn was a modified version of the IBI developed during the late 1970s and early 1980s to assess biotic integrity and environmental quality in small streams in Indiana and Illinois (Karr, 1981; Karr and others, 1986). The IBI consists of a series of fish community attributes, or metrics, that reflect basic structural and functional characteristics of biotic assemblages: species richness and composition, trophic and reproductive function, and individual abundance and condition (Lyons, 1992). The Wisconsin version of the IBI consists of 10 basic metrics and 2 correction factors. The IBI is predicated on the assumption that the number of species in a community declines with increasing environmental degradation.

Table 26. Biotic integrity ratings for Index of Biotic Integrity score (modified from Karr and others, 1986).

Of the 14 wadeable sites sampled, IBI scores for fish could not be reliably computed at 3 sites because only a small number of fish were collected (table 27). If fewer than 50 fish are collected at a site, the Wisconsin warmwater IBI should not be used in assessing the fish-community data (Lyons, 1992). IBI scores were computed for the remaining 11 sites: 6 scored very poor; 2 were poor; and 1 each was fair, good, and excellent (table 27).

Table 27. Fish-community information from one-time surveys conducted during July, August, and October 2004, for 14 Phase II stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis.

[*, fewer than 50 individual fish were collected, so the Wisconsin warmwater Index of Biotic Integrity was not used (Lyons, 1992)]

The two Milwaukee River main-stem sites had the highest IBI scores, with all other sites scoring poor or very poor. Two of the individual metrics included in the Wisconsin IBI are “percent of fish tolerant to low dissolved-oxygen levels” and “percent of fish tolerant to disturbed habitat.” All but four sites (Milwaukee River at Milwaukee, Milwaukee River near Cedarburg, Willow Creek, and Jewel Creek) had more than 50 percent of the fish community in 1 of these 2 metrics, indicating a high percentage of fish in the population that are tolerant of degraded stream conditions.

Fish IBI scores indicated a negative relation to urban land use, where fish IBI scores decreased as urban land use increased (fig. 59). Phase I IBI scores were computed on data collected from 1990 through 2002. Data were available for Phase I and Phase II fish IBI score comparisons at six sites (appendix 5). Of these, four sites remained in the same category as was determined by the Phase I data (all very poor). However, Kinnickinnic River and Honey Creek both went from poor in Phase I to very poor in Phase II.

Figure 59. Fish Index of Biotic Integrity (IBI) scores plotted against percent urban land use in site drainage basins for 15 stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

return to top

Macroinvertebrate Results

Selected metrics were computed based on macroinvertebrates found at each site. These metrics included the number of macroinvertebrate species and genera, Shannon index of diversity (Odum, 1971), the percentage of macroinvertebrate individuals or genera in the orders Ephemeroptera-Plecoptera-Trichoptera (EPT), and the Hilsenhoff Biotic Index (HBI) and 10-Max BI (HBI-10). Taxa (species and genera) would be expected to decrease with degrading water quality. Shannon index of diversity scores generally decrease with degrading water quality; however, pristine headwater streams may have low diversity and excellent water quality. EPT invertebrates are generally considered to be relatively intolerant of water-quality degradation. Intolerant organisms tend to dominate in streams with good water quality, whereas tolerant organisms dominate in polluted streams; therefore, the percentage of EPT individuals and genera tend to decrease with decreasing water quality. The HBI was designed to assess oxygen depletion in streams resulting from organic-matter pollution (Hilsenhoff, 1987); however, the HBI also may be sensitive to other types of pollution, such as that from some chemicals. The HBI represents the number of arthropod macroinvertebrates in certain families or species, multiplied by their respective pollution-tolerance score, divided by the number of arthropods in the sample. HBI values can range from 0.00 (excellent water quality) to 10.00 (very poor water quality) (table 28). A modification of the HBI (HBI-10) was used in the Phase II analyses because it limits the number of individuals per taxa to 10 for computation of the index and is thought to be more accurate than the HBI because it is less affected by dominance of a single taxon (Hilsenhoff, 1998).

return to top

Table 28. Water-quality ratings for Hilsenhoff Biotic Index (HBI) and HBI-10 values (from Hilsenhoff, 1987 and 1998).

[≤, less than or equal to]

The Milwaukee River near Cedarburg and Milwaukee River at Milwaukee sites had the highest numbers of macroinvertebrate species and genera; these sites also had the highest Shannon index of diversity scores (greater than 3.6), along with Willow Creek, a much smaller stream (table 29). The Lincoln Creek, Menomonee River at Menomonee Falls, and Menomonee River at Wauwatosa sites had the fewest species and genera. The lowest Shannon index of diversity scores were found at the Menomonee River at Menomonee Falls and the Little Menomonee River sites. This finding suggested that the water quality at these sites was more degraded than at other sampled sites.

Table 29. Macroinvertebrate community information from one-time surveys conducted during August and September 2004, for 14 Phase II stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis.

[EPT, Ephemeroptera, Plecoptera, and Trichoptera; HBI-10, modified Hilsenhoff Biotic Index; see table 28 for explanation of HBI-10 water-quality ratings]

Sites with the lowest EPT percentages (EPT individuals < 30 percent and EPT taxa <10–20 percent) were the Kinnickinnic River (lowest overall), Honey Creek, Root River at Grange Avenue, and Lincoln Creek sites (table 29). Historically, the Middle and Lower Root River Phase I subwatersheds had much higher median EPT percentages than the Upper Root River or East Branch Root River subwatersheds (fig. 60), suggesting that the lower subwatersheds had less-degraded water quality. The August 2004 sample for Root River near Franklin (Middle Root River subwatershed) had a much lower percentage of EPT taxa (28 percent) compared to historical percentages (median of 50 percent EPT taxa). Further investigation of the macroinvertebrate communities in the Middle Root River may be warranted to determine whether water quality has declined notably. On the other hand, historical data indicated that EPT taxa were few or absent from the Lincoln Creek Phase I subwatershed, but samples from August 2004 indicated a higher percentage of EPT taxa (18 percent) possibly indicating improved water quality. The Little Menomonee River subwatershed also indicated a higher percentage of EPT taxa (23 percent) in the Phase II sample; however, this result may be misleading since the sample contained a high proportion of a pollution-tolerant Ephemeroptera species (Baetis intercalaris). Percent EPT taxa indicated a negative relation to urban land use, where percent EPT taxa decreased with increasing urban land use (fig. 61).

Figure 60. Sites sampled for macroinvertebrates with percent Ephemeroptera, Plecoptera, and Trichoptera (EPT) taxa in the Milwaukee Metropolitan Sewerage District (MMSD) planning area, Wis. Site abbreviations listed in table 1.

Figure 61. Percent Ephemeroptera, Plecoptera, and Trichoptera (EPT) taxa plotted against percent urban land use in site drainage basins for 15 stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

HBI-10 scores for August 2004 samples ranged from fairly poor water quality at the Kinnickinnic River site to fair or good at other sampled sites (fig. 62). Sites with fair HBI-10 scores were Lincoln Creek, Menomonee River at Menomonee Falls, Menomonee River at Wauwatosa, Underwood Creek, Jewel Creek, Oak Creek, Root River at Grange Avenue, and Root River near Franklin. Sites with good HBI-10 scores were those in the middle and upper part of the planning area, specifically Willow Creek, Little Menomonee River, Honey Creek, Milwaukee River near Cedarburg and Milwaukee River at Milwaukee. Scores for Little Menomonee River and Honey Creek were anomalously high when compared with data for other biological metrics (table 29). High HBI-10 scores at Little Menomonee River are likely the result of the same predominant pollution-tolerant Ephemeroptera species that affected the EPT percentages; therefore, high HBI-10 scores at this site may not be indicative of true water quality. Honey Creek’s HBI-10 scores may be misleading, as all other macroinvertebrate metrics indicate a decline in water quality. HBI-10 scores indicated a positive relation with urban land use, where HBI-10 scores increased (indicating decreasing water quality) with increasing urban land use (fig. 63).

Figure 62. Sites sampled for macroinvertebrates with Hilsenhoff Biotic Index (HBI) in the Milwaukee Metropolitan Sewerage District (MMSD) planning area, Wis. See table 28 for more information on HBI water-quality ratings. Site abbreviations listed in table 1.

Figure 63. A modified Hilsenhoff Biotic Index (HBI-10) plotted against percent urban land use in site drainage basins for 15 stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis (Hilsenhoff, 1998). Site abbreviations listed in table 1.

return to top

Algae Results

Attached benthic algae (periphyton) are often the dominant primary carbon producers and energy source for food chains in small- to medium-sized streams. The term “periphyton” refers to the collection of the attached benthic algae and other heterotrophic bacteria or microbes that are affixed to the submerged substrata in freshwater systems. The abundance and species diversity of periphyton provides valuable information on water quality in a particular stream reach. Periphyton can be used to develop indicator indices in a manner similar to those for macroinvertebrates and fish.

The percent relative abundance (PRA) of each algal group (blue-green algae, diatoms, green algae, red algae) is the number of cells present of the algal group divided by the total number of algal cells. Blue-green algae had the highest PRA at Menomonee River at Wauwatosa (90.98 percent) and diatoms were dominant at Willow Creek (PRA of 58.74 percent) (table 30). Green algae were found at only eight sites, with Lincoln Creek having the highest PRA (3.65 percent). Red algae were also found at eight sites, with Root River at Greenfield and Root River near Franklin having the highest PRA (86.92 percent and 51.72 percent, respectively). Algal group PRAs indicated no appreciable relation to urban land use.

Table 30. Algal-community results from one-time surveys conducted in August and September 2004, for 14 Phase II stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis.

Percent biovolume of an algal group is determined by multiplying the number of algal cells by the volume of space each cell occupies, divided by the total biovolume of the algal cells. The percent biovolume for a group of algae can be drastically different than the PRA because of the size of the cells. Blue-green algae may have been the most abundant in the count; however those algae occupied very small amounts of volume compared to other groups such as diatoms or green algae and therefore may not have had a high-percent biovolume. The most notable difference in biovolume composition from PRA occurred at Milwaukee River near Cedarburg, where blue-green algae composed 83.46 percent of the relative abundance but only 1.76 percent of the biovolume and green algae composed 0.41 percent of the relative abundance and 86.46 percent of the biovolume (table 30). Percent biovolumes of algal groups indicated no appreciable relation to urban land use.

Nuisance, bloom-producing algae were found only at Milwaukee River near Cedarburg (Cladophora glomerata) and Underwood Creek (Stigeoclonium lubricum). Both of these taxa are green algae that are common in Great Lakes-area streams and lakes with high nutrient loading, especially phosphorus (Prescott, 1962; Wehr and Sheath, 2003). On the basis of occurrence of nuisance algae and the pollution classes, these two streams were identified as possible areas of high nutrient loading and high concentrations of other pollutants.

Pollution-tolerance classes for diatoms (“Most pollution tolerant” and “Pollution sensitive” in figure 64) have been used as water-quality indicators in streams (Lange-Bertalot, 1979; Bahls, 1993). The classes are based on several variables such as nutrient concentration, saprobic conditions (organic rich, oxygen poor), biochemical oxygen demand (BOD), and toxics that each taxon can tolerate and are based on the percent relative abundance of each taxon in the sampled streams. Fewer pollution-sensitive taxa at a site indicate that at least one of the variables that make up the pollution-tolerance classes is elevated and not suited for pollution-sensitive diatoms (Bahls, 1993). Jewel Creek, Willow Creek, Milwaukee River near Cedarburg, and Root River near Franklin had the highest percentages of diatom-normalized relative abundances of pollution-sensitive diatoms all over 60 percent (fig. 64). Lincoln Creek and Kinnickinnic River had the lowest percentages of diatom-normalized relative abundances of pollution-sensitive diatoms both below 20 percent.

Figure 64. Percentage of diatoms in pollution-tolerance classes in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

Streams with low percentages of pollution-sensitive diatoms indicate that at least one of the variables used to determine pollution sensitivity is elevated and that those streams cannot sustain a large community of pollution-sensitive taxa. Streams that have less than 30 percent pollution-sensitive diatoms are of concern because these streams may contain nutrients, such as phosphorus, that lead to high oxygen demand and eutrophication. The streams that had less than 30 percent pollution-sensitive diatoms were Lincoln Creek, Honey Creek, Menomonee River at Wauwatosa, and Kinnickinnic River. Pollution-sensitive diatom percentages indicated a negative relation to urban land use, where percentages decreased with increasing urban land use (fig. 65). Percentages of other pollution-tolerance classes for diatoms indicated no appreciable relation to urban land use.

Figure 65. Percent pollution-sensitive diatoms plotted against percent urban land use in site drainage basins for 15 stream sites in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

return to top

Relations among Habitat Variables and Biotic Communities

The streams in the MMSD planning area exhibited considerable variation in their physical habitats, mainly because of the wide range in size of streams sampled. For example, mean wetted width of the sampled reaches ranged from less than 3 m at the site with the smallest drainage area (Willow Creek, 16.39 km²/6.33 mi²) to almost 70 m at the site with the largest drainage area (Milwaukee River at Milwaukee, 1,787 km²/690 mi²). Mean depth ranged from 0.15 m at Willow Creek to 0.56 m at Milwaukee River at Milwaukee.

Correlations among biological metrics and habitat variables (table 31) were evaluated using Spearman’s rank correlation (Iman and Conover, 1983). Results indicated that the larger streams had relatively higher-quality fish and macroinvertebrate communities than small streams; however, algal community metrics did not correlate with stream-size characteristics. Fish and macroinvertebrate communities were significantly correlated (p < 0.05) with the stream-size variables: width-to-depth ratio, low flow volume, wetted channel area, bankfull channel area, and drainage area. Unexpectedly, streams with a high amount of bank erosion and a high percentage of pools had low HBI scores, indicating the potential of confounding relations among multiple measured and unmeasured environmental characteristics.

Table 31. Correlations among metrics of aquatic biology and selected habitat metrics for Phase II of the Milwaukee Metropolitan Sewerage District Corridor Study.

[EPT, Ephemeroptera, Plecoptera, and Trichoptera; HBI, Hilsenhoff Biotic Index; IBI, Index of Biotic Integrity; WIHAB, Wisconsin Department of Natural Resources habitat index; m³, cubic meter; m², square meter; %, percent; m, meter; <, less than; r-values in bold indicate significance at the p < 0.05 level]

¹ (modified from Simonson and others, 1994)
² (modified from Simon and Hupp, 1992)