Species Composition

A total of 25 different fish taxa were found in the Underwood Creek reaches (fig. 1; table 1), an increase of four species from the total reported in Bell and others (2021). Creek chub (Semotilus atromaculatus) were found during every survey at all reaches. The creek chub is a native Wisconsin species that has no preference for streambed type, water clarity, or water temperature (Becker, 1983), and is found in medium and small streams across the state. The next two most abundant species at all reaches but missing in two individual sampling events, were western blacknose dace (Rhinichthys obtusus) and white sucker (Catostomus commersonii). Both species are ubiquitous in Wisconsin lakes and rivers and are tolerant to a wide range of temperature fluctuations, dissolved oxygen concentrations, substrate types, and pollution conditions (Becker, 1983). Although native to Wisconsin, western blacknose dace and white sucker are often abundant in disturbed urban streams such as Underwood Creek because of their ability to thrive in a variety of stream conditions (Becker, 1983). Other species that were collected at all reaches and in most samples include central stoneroller (Campostoma anomalum) and johnny darter (Etheostoma nigrum).

Fish community data collected in spring 2021 showed that Underwood Creek is meeting some of the criteria for success in the Underwood Creek restoration plan. The Underwood Creek restoration plan (US Army Corps of Engineers, 2015, p. 5) states:

White sucker play an important role in streams in early spring by transporting nutrients that in turn feed the algae and insects that anchor the food chain. White sucker have been collected every year except 2010 in Underwood Creek reach A, but in low numbers (less than 10 percent of total catch in terms of individual fish). Restoration of reach A was completed in 2017 but even when it was partially complete during the 2016 sampling, abundance of white sucker increased substantially to 155 individuals, which comprised about 67 percent of the total catch. During 2021 spring sampling, all reaches had lower fish abundance than previous fall surveys, but white suckers still made up about 70 percent of total catch in reach A. White suckers were found both above and below the drop structure that separates reaches B and D, which are both sections that were reconstructed. White suckers were not found in reach C, but fish abundance there was the lowest of any survey in Underwood Creek (24 individuals).

The longnose dace (Rhinichthys cataractae) is similar in appearance to the western blacknose dace but was not found in Underwood Creek during USGS sample collection in any year. The longnose dace is found in medium-sized fast-moving streams in northern Wisconsin and in small-sized, fast-moving streams in southwestern Wisconsin (Becker, 1983). However, according to descriptions by both Becker and Smith, the longnose dace that occur in southeastern Wisconsin and northeastern Illinois are commonly found in “windswept shoals” (Becker, 1983, p. 474) and “along the turbulent pebble beaches” (Smith, 2002, p. 86) of Lake Michigan. Smith (2002, p. 87) notes that “longnose dace is common along the shoreline of Lake Michigan but extremely rare in inland streams.” Based on the published distribution, an occasional longnose dace may be found in streams in the area; however, the gravel and cobble streambed of the restored and unrestored reaches of Underwood Creek will not likely house a sustainable population of longnose dace despite enhanced habitat conditions.

Northern pike have not been collected since a single specimen was collected in 2013, although these fish may find the restored stream potential suitable habitat in the future. Both species are found in the Menomonee River that Underwood Creek flows into (Scudder Eikenberry and others, 2020b) and it is possible that these larger sport fish will be found in Underwood Creek in the future.

An objective of the Underwood Creek reconstruction was to increase connectivity to headwater spawning habitat. Although steelhead are considered an exotic species (Lyons, 1992), they provide insight into the ability of fish to migrate in Underwood Creek. Steelhead were observed but none collected in reach A. No steelhead were observed or collected in either reach C or reach B. Reach C is a shallow and wide reach just before the confluence with the Menomonee River and lacks spawning habitat for these large fish, but it does not prevent them from swimming to reach suitable habitat upstream. Lack of steelhead at reach B indicates that either the fish were unable to traverse the concrete section of the stream between reach D and reach B, or the habitat in reach B was unsuitable and if any fish did swim through the concrete section, they kept going past reach B in search of spawning habitat further upstream. Of the four restored reaches of Underwood Creek, reach D, which contains a newly restored channel and riparian area with diverse habitat, supported the greatest fish abundance and species richness. Reach D is also directly below a drop structure. Species observed here in 2021, but not in previous surveys of Underwood Creek, include black bullhead (Ameiurus melas), blackside darter (Percina maculata), brook trout (Salvelinus fontinalis), and steelhead. Many steelhead were observed in the reach in spawning condition. Most were observed but not collected to minimize disruption of their spawning.

Community Traits and Index of Biotic Integrity

Bell and others (2021) lists the species attributes to calculate the warmwater IBI for Underwood Creek. We include here an updated table that includes species newly observed in 2021 and attributes applicable to the warm-transition IBI (table 2).

Sampling in 2021 took place in early April to target spring migrating fishes, a key metric of restoration success for Underwood Creek (U.S. Army Corps of Engineers, 2015). Data indicate that the restoration has resulted in suitable habitat for at least some migrating fishes, primarily white sucker and steelhead. White suckers were found at all reaches except reach C (table 1). Steelhead were observed at reaches A and D but collected at only reach D to avoid disruption to spawning (table 1). IBI scores from the spring 2021 sampling effort are included for information but are not directly comparable as the IBI is designed to be used with midsummer-to-early-fall sample collection (table 3).

Fish tolerance to pollution, low dissolved oxygen, and disturbed habitat can be used to make inferences about the water quality of a stream. Lyons (1992, p. 11) describes intolerant species as those that are “intolerant of environmental degradation, particularly poor water quality, siltation and increased turbidity, and reduced habitat heterogeneity.” Tolerant species are those adaptable to degraded water quality and habitat. Non-tolerant species include intolerant species and species that are neither designated tolerant or intolerant. Pollution-tolerant species comprised 57–75 percent of the species collected at each reach in spring 2021, representing a small decline from the percentage of tolerant species collected in spring 2019 (71–75 percent). Brook trout was the only pollution-intolerant species observed; one individual was collected from reach D and was the first intolerant fish collected since surveying began in 2004 at Underwood Creek. Fishes tolerant to low dissolved oxygen were observed at low abundance (1–17 percent) in all surveyed reaches in spring 2021. Fishes tolerant to pollution and disturbed habitat were more abundant and accounted for 58–88 percent and 42–86 percent of the fish abundance at Underwood Creek in spring 2021, respectively. Spawning traits can also provide information about the quality of physical habitat. Simple lithophilous spawners lay eggs in clean coarse gravel and can be impacted by excess sedimentation (Lyons, 1992). The 2021 surveys indicated that 20–29 percent of species in each reach are lithophilous spawners.

Categorizing fish into feeding groups can provide insight into the structure of the food web. When a stream undergoes reconstruction, it may take several years to stabilize and reestablish the food web. In general, rivers that have a higher percentage of invertivore fish (greater than 30 percent) have a more diverse food web and can support the higher trophic level fish such as northern pike and largemouth bass (Goldstein and others, 1999). The reaches sampled in the 2021 survey contained 38–50 percent invertivore species and 3–38 percent invertivore fishes (table 3).

In 2021 sampling, only reach D contained game or sport fish. Most of these were steelhead, but a brook trout and a pumpkinseed (Lepomis gibbosus) were also collected (table 1). Northern pike and largemouth bass are game fish that were found in surveys prior to 2021 in Underwood Creek, but not during the 2021 sampling. Other species of game fish include top carnivores such as muskellunge (Esox masquinongy), smallmouth bass (Micropterus dolomieu), and sturgeon family (Acipenseridae), among others, but these have not been found in Underwood Creek since surveys began in 2004. Sport fish, in general, are species that are targeted for recreational fishing; this category includes game fish and includes some members of the sunfish family, perch family (Percidae), and bullhead and catfish family (Ictaluridae). In contrast, forage fish are those that are not considered sport fish. The forage fish account for most of the abundance across all samples (96 percent or greater). These forage fish provide food for the top carnivores and are mostly invertivores and omnivores. The largemouth bass (Micropterus salmoides), a top carnivore and sport fish, was not collected in the spring 2021 sampling at reach A despite being collected during previous surveys 2004–19. The habitat alterations in reach A reduced the size of the large plunge pool that existed at the bottom of the former concrete channel, which may have affected the habitat available for fish that are less riverine including sunfish and largemouth bass.

Bray-Curtis Similarity Comparison

In this report, communities in each reach survey are compared based on their Bray-Curtis Similarity score. Two reaches are considered similar if their similarity score is 50 percent or greater. As of this report (2026), reach A has been sampled seven times, with sampling dates between April and October. When reach A surveys were compared to each other, communities were similar about 50 percent of the time, based on a similarity score of 50 percent or greater (table 4). Reach B has been sampled four times since its reconstruction, twice in April and twice in July. The Bray-Curtis Similarity scores always indicated similarity within reach B, indicating the community there is stable. A drop structure separates reach B from reach A and the confluence with the Menomonee River. Prior to 2019, reach A and reach B were similar about 50 percent of the time. In 2019 and 2021, after reach A was reconstructed, reach A and B Bray-Curtis Similarity scores were always similar to each other.

All four samples from reach B have similarities of 58 percent or greater to the Elm Grove reach sample, which was surveyed in 2019 and is upstream from reach B (table 4; fig. 1). Only two surveys in reach A indicated similarity to the Elm Grove reach community. Reach C was generally dissimilar to reach A, despite their proximity. These reaches differ in their history and habitat; reach A was reconstructed in 2017, whereas reach C was not lined with concrete and is low gradient with soft substrate and little structure other than the deep pool at the bottom of the reach. Reach D was only similar to three out of seven surveys in reach A. Reach D underwent similar reconstruction as reach A, but reach D has a higher gradient. Both reaches C and D were similar to Elm Grove reach. Reaches C and D were also similar to all but the earliest survey in reach B.