Purpose and Scope

The purpose of this report is to synthesize previous studies and evaluate monitoring data to understand the distribution of fish and invertebrates and the ecological response to streamflow, water temperature, and water quality within the Fountain Creek Basin, and to help identify opportunities for aligning data collection that could maximize information gained from additional cooperative studies. The report focuses on USGS ecology-based studies in the Fountain Creek Basin performed between the mid-1980s and 2022; selected studies carried out by other entities were also included. To help guide additional monitoring, selected ecological datasets from ongoing monitoring in the basin were evaluated to further support previous results or to potentially fill selected information gaps in the previously published studies.

Study Area Description

The study area was mostly encompassed by the Fountain Creek Basin, which was described in detail by Edelmann and others, 2002; Mau and others, 2007; and Zuellig, Bruce, and others, 2007, and consists of sites on the Arkansas River just upstream and downstream from the confluence with Fountain Creek (fig. 1, table 1). In general, the Fountain Creek Basin drains approximately 926 square miles (mi²) of the eastern slope of the Rocky Mountains in south-central Colorado and confluences with the Arkansas River near Pueblo, Colorado (fig. 1).

Within the study area, elevation ranges from 4,509 feet (ft) at the lowest site on the Arkansas River to 14,109 ft at the summit of Pikes Peak. Fountain Creek and Monument Creek are the two main drainages located in the transition of two distinctive physiographic regions: the Southern Rocky Mountains and the Colorado Piedmont (Hansen and others, 1982). These landforms correspond to two ecoregions, the Southern Rockies and the Southwestern Tablelands (Omernik, 1987). The primary area of interest for this report consists of the transitional area between these main landforms, and ecoregions in the Fountain Creek Basin to just below the confluence of Fountain Creek and the Arkansas River (fig. 1).

In general, historical records documenting observations of early explorers from the 1840s indicate that several changes occurred to physical attributes of regional streams (Fausch and Bestgen, 1997). For example, plains streams, such as in the Fountain Creek Basin near the eastern edge of the Front Range of the southern Rocky Mountains, that flowed clear over cobble-gravel streambeds now consist of much finer particles with noticeable turbidity (Fausch and Bestgen, 1997). Farther east on the plains, sinuosity and pool frequency increased as riparian areas were encroached by woody vegetation in response to reductions in overbank flooding after the advent of irrigated agriculture in the mid-1800s (Fausch and Bestgen, 1997). Streamflow alteration evolved with the construction of small ditches (1840s) followed by large canals (1860–85), large main-stem reservoirs (1885–1940s), and more recently (1940s to the early 21st century) additional reservoirs, transmountain water diversions, and wastewater-treatment facility discharge (Fausch and Bestgen, 1997; Eschner and others, 1983). Changes in water quality likely also occurred concurrently with the characteristics described previously (streamflow alteration, in-stream habitat) as the landscape changed due to regional mining, industry, and pressures from urbanization. These early changes in hydrology, water quality, and the physical template of regional streams, including the Fountain Creek Basin, shaped the present-day biota, which are markedly different from those documented in historical records (Fausch and Bestgen, 1997).

Describing Changes in Fish and Invertebrate Communities

Changes in aquatic communities were documented by calculating the percent frequency of occurrence of selected fish species or invertebrate taxa based on observational patterns in the sampling data. Percent frequency of occurrence was calculated by compiling presence-absence data sorted into groups of 3 years for each site, where the number of times a species or taxon occurred was calculated as the percentage of years it was present within a group. Data indicating percent frequency of occurrence were then examined for changes at a site, in which a species or taxon showed obvious downward or upward patterns through time were retained for presentation and discussion. Fish data consisted of samples collected between 2003 and 2022, and invertebrate data consisted of samples collected between 1985 and 2022, dependent on site.

Additionally, invertebrate multimetric index (MMI) values were calculated for each site following Jessup and Stribling (2017) and plotted against year to detect obvious decreasing or increasing patterns in MMI values through time. In general, MMI values are calculated by selecting metrics that best discriminate between predefined reference and disturbed sites within a given basin, region, ecoregion, or other spatial stratification. The final set of metrics typically represents various aspects of community composition, richness, tolerance, abundance, function, and life history or other trait-type characteristics. Once a final set of metrics is selected, values are scored on a continuous scale, summed, and then rescaled to range between 0 and 100. The invertebrate MMI used herein was developed for the CDPHE by Tetra Tech, Inc., to evaluate the biological condition of streams in Colorado (Jessup and Stribling, 2017). Calculation of MMI values was automated using a tool packaged in a Microsoft Access database as the Ecological Data Application System (EDAS) for Colorado (Jessup and Stribling, 2017). A working version of EDAS was unavailable at the time of this publication, so MMI values were calculated by CDPHE staff in June 2023 using selected invertebrate data compiled from Zuellig and others (2022). Refer to Jessup and Stribling (2017), Zuellig and others (2014), and Bruce and others (2018) for details regarding the EDAS MMI for Colorado. After MMI values were calculated, values were plotted against year, and a loess smooth line (span=0.90) and 90-percent confidence intervals were added using the ggplot2 package (Wickham and others, 2023) in R (R Core Team, 2023) to aid interpretation. Data at sampling sites that showed obvious changes in MMI values were retained for presentation and discussion.

Corroborating the Association Between Streamflow and Age-Zero Flathead Chub

Haworth and Bestgen (2017) proposed that growth and survival of age-zero (young-of-the-year) flathead chub were greatly reduced in the Fountain Creek Basin during years when daily streamflow exceeded 20 cubic meters per second (706.29 cubic feet per second). Roberts and others (2018) used this information to suggest catch per unit effort of age-zero flathead chub consistently decreased in years with more than 25 days between May and August having streamflows at or greater than the threshold suggested by Haworth and Bestgen (2017) (fig. 6 in Roberts and others, 2018). To substantiate the proposed thresholds of these two studies, streamflow of 706.29 cubic feet per second was determined at site 10 as the 98th percentile of the average daily streamflow from 2003 to 2022. The 98th-percentile streamflow value was then calculated at site 8, which was 618.95 cubic feet per second. For each year at these two USGS streamgages (table 1; USGS, 2023), the number of daily maximum streamflow occurrences (hereafter referred to as “peak streamflow days”) were summed from May through August and represent the number of peak streamflow days above the threshold each year. The results were then plotted with the data used to create figure 6 of Roberts and others (2018) and evaluated for continuity.

Exploring Changes in Invertebrate Community Tolerance to Water Temperature

Compliance with State-regulated water temperature standards has become an area of concern among wastewater treatment facilities within the Fountain Creek Basin (Annie Berlemann, Colorado Springs Utilities, oral commun., 2023). Ideally, continuous water temperature data paired with ecological samples during the period of record of available ecological samples would be used to evaluate an ecological response to changes in water temperature. However, very little continuous water temperature data were available at ecological monitoring sites in the basin in 2023 (USGS, 2023). Therefore, a surrogate measure of ecological response to water temperature was necessary to evaluate if invertebrate community tolerance to water temperatures changed through time. Carlisle and others (2007) derived taxa-specific tolerance indicator values (TIVs) for water temperature from a nationwide dataset; values were estimated from abundance-weighted averages of each taxon in relation to concurrently collected water temperature data. Those values could then be tested with independent data to confirm the applicability of the estimated TIVs and to discern specific stressors (Yuan, 2004; Carlisle and others, 2007; Meador and others, 2008). Herein, average community tolerance to water temperature was estimated by applying taxa-specific temperature TIVs of Carlisle and others (2007) to invertebrate samples collected within the Fountain Creek Basin between 2005 and 2022 (Zuellig and others, 2022). Increasing TIVs indicate invertebrate communities are becoming more tolerant to warmer water temperatures through time, whereas decreasing TIVs indicate they are becoming less tolerant to warming temperatures. After values were calculated, average TIVs for each sample were plotted against year for each sampling site to detect obvious decreasing or increasing patterns in community tolerance through time. A loess smooth line (span=0.90) and 90-percent confidence intervals were added using the ggplot2 package (Wickham and others, 2023) in R (R Core Team, 2023) to aid interpretation. Data at sites that showed obvious changes in water temperature TIV values were retained for presentation and discussion.

Exploring the Association Between Multimetric Index Values and Documented Trends in Water Quality

Bern and others (2024) examined trends in available water quality data from 1999 to 2022 at eight sites in the Fountain Creek Basin that overlapped with sites included herein (table 1). Trends reported by Bern and others (2024) as highly likely related to ammonia, nitrate plus nitrite, unfiltered phosphorus, orthophosphate, and chloride were selected to determine if there were corresponding patterns in changes in MMI values at overlapping sites. Highly likely trends were those having the greatest statistical likelihood that they occurred in the stated direction as indicated by a two-sided attained p-value (Hirsch and others, 2015) ranging between 0.95 to 1.0. In general, increases in concentration in the selected constituents could result in decreases in MMI values if the changes in concentration were great enough to be biologically relevant. Sites with decreasing MMI values that corresponded to highly likely trends in the selected water quality data reported by Bern and others (2024) were included for discussion.

Understanding the Fish Assemblage

Fausch and Bestgen (1997) provide a thorough history of early collections of plains stream fishes in Colorado. Before 1900, fish were collected from only seven sites within the Arkansas River Basin near the cities of Cañon City, Colorado Springs, and Pueblo (Fausch and Bestgen, 1997). Basinwide fish surveys of the Arkansas River Basin, including Fountain Creek, did not occur until after 1958 (Fausch and Bestgen, 1997). Two additional plains-fish surveys of the Arkansas River Basin were completed by the Colorado Division of Wildlife (now Colorado Parks and Wildlife) in the 1980s and 1990s (Loeffler and others, 1982; Nesler and others, 1999). Since 2008, Colorado Parks and Wildlife has extensively sampled fish in the basin, although no formal document reporting findings of this ongoing effort was available at the time of this publication. However, data associated with this more recent sampling effort are available, as is historical information, by making a special request to Colorado Parks and Wildlife.

It is difficult to determine exactly how current (2023) patterns of fish distribution compare to historical patterns, because the distribution of fishes in the Arkansas River Basin and Fountain Creek was not documented until well after major changes in hydrology, stream water quality, and habitat had already occurred. However, Fausch and Bestgen (1997) used literature collection records, museum records, and unpublished data to estimate that there were possibly 20 fish species native to the Arkansas River Basin in Colorado (refer to table 6.1 in Fausch and Bestgen, 1997, for details). Of these 20 species, three (Macrhybopsis aestivalis, speckled chub; Hybognathus placitus, plains minnow; and Carpiodes carpio, river carpsucker) have not been collected in more than 100 years; one (Notropis girardi, Arkansas river shiner) was presumed native to the Arkansas River Basin because of its occurrence in the Cimmaron River upstream and downstream from Colorado in New Mexico and Kansas (Fausch and Bestgen, 1997). Of these four species, plains minnow was collected in 1913 from the Arkansas River at Cañon City upstream from the Fountain Creek confluence and at Pueblo in 1889; speckled chub was collected in 1889 from the Arkansas River at Pueblo, suggesting that they occurred within the study area. Zuellig and others (2007) indicated that plains minnow and speckled chub were likely associated with downstream reaches of the Arkansas River and not Fountain Creek but were unaware of the additional nearby historical records, so they are included herein as historical occupants of the study area. Fausch and Bestgen (1997) also included Oncorhynchus clarkii stomias (greenback cutthroat trout) as native to the Arkansas River Basin, but this was later corrected with advanced genetic analysis by Metcalf and others (2007, 2012)67, who determined Oncorhynchus clarkii macdonaldi (yellowfin cutthroat trout) was the only salmonid native to the Arkansas River Basin.

As many as 18 native species may have historically occurred within the study area based on historical records. However, the most recent published plains-fish inventories of the Arkansas River Basin indicate nearly half of the native species are missing. For example, Loeffler and others (1982) reported 11 native species captured from the Fountain Creek Basin whereas Nesler and others (1999) reported 9. Roberts and others (2018) reported 23 species, of which 11 were considered native to the study area. Note that the results of Roberts and others (2018) were not based on inventory data as were the results of the other two studies (Loeffler and others, 1982; Nesler and others, 1999), but rather from routine monitoring that occurred between 2003 and 2016. Nonetheless, the fish fauna of Fountain Creek is markedly different from its original condition, because more than half of the 23 species reported by Zuellig and others (2022) are considered nonnative (Loeffler and others, 1982; Fausch and Bestgen, 1997; Nesler and others, 1999; Roberts and others, 2018) and because of the extirpation of several species that occurred historically.

Of the native species reported from the Fountain Creek Basin in 2023, three are State listed and occur east of the Front Range of the southern Rocky Mountains: Chrosomus erythrogaster (southern redbelly dace, State endangered), Etheostoma cragini (Arkansas darter, State threatened), and flathead chub (State species of concern); one (yellowfin cutthroat trout) has been considered extinct since 1904 (Wiltzius, 1985) but possibly historically occurred in the upper Fountain Creek Basin based on figure 4 in Metcalf and others (2012) and figure 1 in Bestgen and others (2019). Additionally, greenback cutthroat trout occurs in Bear Creek in the upper Fountain Creek Basin. Although it is not native to the Arkansas River Basin, the Bear Creek population has played a major role in the recovery of greenback cutthroat trout, which is currently (2023) State and federally listed as threatened (Metcalf and others, 2012).

It is apparent that fish distribution has changed, especially with the introduction of nonnative species compared to historical patterns, with several examples of site-specific changes in fish distribution occurring more recently (Zuellig and others, 2022). The following discussions in this paragraph represent selected examples in which a fish species either became more common or less common at a site through time. For example, despite annual sampling since 2003, Notropis stramineus (sand shiner) has not been observed at sites 5, 6, and 10 since 2005, 2003, and 2017, respectively, whereas Lepomis cyanellus (green sunfish) appears to be more common at sites 5 and 9 since 2015 and 2010 (table 2). Fundulus zebrinus (plains killifish) is a habitat specialist that prefers slow-moving backwaters and side channels (Cashner and others, 2020) and has not been collected at sites 9 and 10 since 2009 and 2012 (table 2). Cyprinella lutrensis (red shiner), Rhinichthys cataractae (longnose dace), and Arkansas darter occur at other locations but have been less frequently detected at a few sites where they were previously found (table 2). Additionally, occurrence of the State-listed flathead chub has become more common or less common through time depending on the site (table 2). Finally, nonnative Semotilus atromaculatus (creek chub) and Catostomus catostomus (longnose sucker) have occurred less frequently at sites 10 and 12, and the occurrence of Gambusia affinis (western mosquitofish) has apparently increased at site 12 (table 2).

Understanding the Invertebrate Assemblage

Little is known about the historical distribution of aquatic invertebrate species in streams of the eastern plains of Colorado. Only a few sporadic early collections of aquatic invertebrates from the region were made, and most of those are from the South Platte River Basin, where several species have been locally extirpated (Edmunds and McCafferty, 1984; McCafferty and others, 1993; Cordeiro, 1999; Kondratieff and Baumann, 2002; Zuellig and others, 2012). Since these early collections, there have been few organized field efforts to document species distribution of stream invertebrates from the plains part of the Arkansas River and Fountain Creek Basins. This type of effort requires compiling museum and literature records in addition to collecting new specimens often using specialized methods that differ from the typical benthic invertebrate sampling used in biomonitoring studies. This, combined with sparse historical information, makes it difficult to fully understand the likely extent of changes in species distribution relative to historical patterns before the human-induced changes that occurred in hydrology and habitat (Kondratieff and Baumann, 2002; Zuellig, Sprague, and others, 2007; Zuellig and others, 2012).

Although it is difficult to determine exact changes in the distribution of invertebrates relative to historical patterns, there is evidence that the occurrence of some invertebrate taxa has changed, or in some cases, invertebrate taxa may have been eliminated from stream locations within the Fountain Creek Basin (von Guerard, 1989a; Zuellig and others, 2022). The following discussion in this paragraph represents selected examples in which invertebrate taxa have either become more common, less common, or have been apparently eliminated. Apparent elimination of multiple invertebrate taxa in the Fountain Creek Basin occurred between the late 1990s and early- to mid-2000s, especially at upstream monitoring sites (table 3). For example, species associated with the mayfly taxa Drunella, Ephemerella, and Rhithrogena; the stonefly taxa Capniidae, Prostoia, Pteronarcella, Sweltsa, and Isoperla; and the caddisfly taxa Arctopsyche and Rhyacophila were frequently collected in various combinations at several sites (sites 2, 3, 4, 6, and 7) between 1985 and 1999 but were rarely encountered after 2000 (table 3). These results represent a substantial change in community structure, especially at the two most upstream locations (sites 2 and 3). These results are likely biased low as there is more than one species associated with each listed taxon that was possibly eliminated at these locations (Ward and others, 2002; Zuellig and others, 2012). Similar changes in invertebrate community structure have occurred in other transition-zone streams along the Front Range of the Rocky Mountains in Colorado that have been pressured by urbanization and water management (Zuellig, Sprague, and others, 2007; Zuellig and others, 2010). Theoretically, these taxa could return if local conditions improved, assuming they are still present at upstream locations. Further investigation could be beneficial in determining if these taxa still occur upstream near the affected sites and better understanding the possibility of their return under an effort to improve current (2023) environmental conditions.

Although several taxa have been apparently eliminated from upstream locations, there are several taxa that have been detected more frequently. For example, various combinations of the mayfly taxa Acentrella, Fallceon, Paracloeodes, or Heptagenia and caddisfly taxa Cheumatopsyche have been occurring more frequently at 10 (table 3) of the 14 sampling sites (fig. 1, table 1). Historically, these taxa occurred in the transition zone between the mountains and plains but are mostly associated with locations farther out on the eastern plains of Colorado (Zuellig and others, 2012), where water temperatures are warmer, coarse substrate is often highly embedded or absent, and the presence of woody debris is more frequent than at sites closer to the plains-mountain interface. Additionally, several of these taxa are tolerant to disturbance and harsh conditions, as they frequently occur in other urbanized areas of transition-zone streams along the Front Range of the Rocky Mountains in Colorado (Sprague and others, 2006; Zuellig and others, 2012).

Changes in Invertebrate Multimetric Index Values

Invertebrate MMI values were calculated for Fountain Creek Basin samples collected between 2006 and 2022 and plotted against year to detect obvious changes in MMI values through time. A total of 6 locations (sites 2, 4, 5, 6, 13, and 14) revealed decreasing values through time, indicating that biological condition has decreased at these sites during the sampling period (fig. 3A–F). Some sites showed rapid decreases in values relative to other locations (sites 6 and 14 compared to sites 2, 4, 5, and 13). The remaining eight sites did not show any clear patterns and therefore are not reported herein. Invertebrate MMI values were based on the most current version of CDPHE’s index (Jessup and Stribling, 2017), which is slated for revision in 2025 (Chris Theel, CDPHE, written commun., 2023). Understanding why these changes may have occurred is outside the scope of this report but could benefit from further investigation when the revised invertebrate MMI is available.

Fish and Invertebrate Response to Streamflow

The relation between streamflow characteristics and aquatic communities has received substantial attention, primarily because the association has often been modified from what it would be in natural conditions (Carlisle and others, 2019). This is an important distinction, because stream hydrology of a specific region is directly linked to the life history traits and behavioral response of the aquatic communities that evolved in the region (Lytle and Poff, 2004). For example, the rising or falling limb of a hydrograph in a snowmelt-based system is often related to spawning cues for fish (Yarnell and others, 2010) or emergence timing for invertebrates (Harper and Peckarsky, 2006). As a result, disruption of life history occurrences can prevent the persistence of native fauna (Brown and Bauer, 2010; Freeman and others, 2022) and can permanently alter communities within a stream (Gido and others, 2013; Wrona and others, 2016). Additionally, streamflow characteristics before sampling affect sampling results, interpretation of trend analysis, and often explain variability in ecological endpoints (Zuellig and Carlisle, 2019).

In the Fountain Creek Basin, numerous studies have identified various hydrograph streamflow characteristics that are important for describing variation in spatial patterns and temporal changes in invertebrate and fish community structure. Initially, von Guerard (1989a) reported a decrease in invertebrate densities in response to flooding 30 days before sampling, with flooding most prevalent in spring and early summer samples. Zuellig, Bruce, and others (2007) indicated that the number of annual high-flood pulses and average annual maximum daily streamflow contributed to explaining spatial patterns in invertebrate community similarity. Temporal invertebrate response to annual maximum mean daily streamflows were also reported by Zuellig and others (2010). Variation and magnitude in antecedent streamflow also explained temporal patterns in invertebrate and fish richness and community similarity among samples (Roberts and others, 2018).

Corroboration Between Streamflow and Age-Zero Flathead Chub

Studies examining population characteristics of early life stages of flathead chub reported lower growth and recruitment in years with more annual peak streamflow days compared to years with more stable streamflow conditions (Haworth and Bestgen, 2016, 2017). Roberts and others (2018) later validated findings of the Haworth and Bestgen studies (2016, 2017) with independent data in which they found greater catch per unit effort of age-zero flathead chub in years with fewer than 25 peak streamflow days. These patterns were corroborated with more recent data (since 2017; fig. 4) in which a similar pattern was observed between catch per unit effort of age-zero flathead chub and the number of peak streamflow days. As in Roberts and others (2018), a lower proportion of age-zero flathead chub were caught when the number of peak streamflow days was greater than 25 days (fig. 4). The extensive data on the relation between aquatic communities and streamflow documents the importance of streamflow in the Fountain Creek Basin in potentially regulating fish and invertebrate communities (von Guerard, 1989a; Zuellig, Bruce, and others, 2007, 2010; Roberts and others, 2018) and recruitment and population dynamics of individual fish species (Haworth and Bestgen, 2016, 2017; Roberts and others, 2018).

Fish and Invertebrate Response to Water Temperature

A tremendous amount of research has been dedicated to understanding the association of water temperature with the ecology and life history traits of aquatic organisms (Connor and others, 2002; Smith and others, 2003; Pankhurst and Munday, 2011; Bonacina and others, 2023). Fish and invertebrate activity and growth are directly correlated with the ambient water temperature. Seasonal changes in water temperature can cue specific behavior, such as reproduction or migratory behavior of fish and the emergence of aquatic insects (Tyus, 1990; Finn and others, 2022). Additionally, aquatic species have unique optimal thermal preferences, such that body function is hampered in temperatures below or above each species’ threshold (Coutant, 1987). Exposure to undesirable temperatures can force individuals to vacate a location to seek refuge or even result in mortality if exposure is chronic. The elimination of species from a location because of intolerable water temperature can greatly alter local fish and invertebrate community structure (Jacobsen and others, 1997; Daufresne and others, 2004; Haase and others, 2019).

In the Fountain Creek Basin, most available water temperature data are from intermittent discrete measurements made during water quality or ecological sampling and therefore have been of limited use to explain variability in invertebrate and fish community data. Nonetheless, discrete water temperature has been noted as important for fish and invertebrates by Bruce (2002) and Roberts and others (2018). The Bruce (2002) study indicated invertebrate communities in the basin changed along a water-temperature gradient from cooler to warmer as water moved downstream away from the plains-mountain interface. Although these findings are factual, this is expected in transition zone streams in Colorado. The difficulty in these systems is separating the natural water temperature gradient (as well as other natural gradients) from human-induced changes. Additionally, Roberts and others (2018) found invertebrate community similarity and total fish catch were associated with discrete water temperature measurements at a few upstream sites (refer to table 6 in Roberts and others, 2018).

Other studies used the limited continuous water temperature data available in the Fountain Creek Basin. For example, Haworth and Bestgen (2016, 2017)40 determined spawning behavior of flathead chub was initiated after 15 degrees Celsius (°C), and individuals captured in warmer water temperatures had faster growth rates than individuals from cooler temperatures. The continuous water temperature data paired with continuous streamflow measurements at site 10 (table 1) made it possible to specify temperature-dependent early life development and growth characteristics of flathead chub in the Fountain Creek Basin (Haworth and Bestgen, 2016, 201740).

Invertebrate Community Tolerance to Water Temperature

The results of plotting average TIV values against year for each site indicate invertebrate community tolerance to warmer water temperatures increased since 2006 at 10 of 14 sites in the Fountain Creek Basin (fig. 5). The shift in communities toward tolerating warmer water could be explained by a general warming trend in water temperature, but the lack of continuous water temperature data made the explanation difficult to determine for this report. Bern and others (2024) reported upward trends in air temperature at Colorado Springs and Pueblo between 1980 and 2022. Air temperature is often correlated with water temperature (Isaak and others, 2010), but verifying this correlation in the Fountain Creek Basin could benefit from further investigation. Nonetheless, invertebrate communities in the Fountain Creek Basin are apparently shifting to tolerate warmer water temperatures since 2006 (fig. 5A–J).

Fish and Invertebrate Response to Selected Water Quality

Changes in water quality can be detrimental to aquatic organisms and have been linked to shifts in the structure and function of entire ecosystems (Appelberg and others, 1993; Kratina and others, 2012; Chen and others, 2019). Degradation of water quality dictates which species persist by filtering out intolerant taxa, altering species interactions, affecting ecosystem function (Sawyer and others, 2004; Bilotta and Brazier, 2008), and degrading biological condition (Violin and others, 2011).

Few studies in the Fountain Creek Basin have directly associated water quality constituents with temporal shifts in aquatic communities, because of inconsistencies in water quality monitoring at ecology locations that have restricted data analysis (Zuellig, Bruce, and others 2007, 2010108; Roberts and others, 2018). Nonetheless, there are a few examples in which water quality has explained spatial and temporal variability in aquatic communities. Zuellig, Bruce, and others (2007) observed dissolved nitrate plus nitrite explained spatial variation in invertebrate and fish community structure at a few sites in the Fountain Creek Basin between 2003 and 2005 (refer to table 7 in Zuellig, Bruce, and others, 2007). Specific conductance was correlated with temporal trends in fish communities at one site (refer to table 6 in Roberts and others, 2018). Others have explored the distribution of mercury (Hg) and selenium (Se) concentrations in fish and invertebrate tissues within the study area (Nimmo and others, 2016, 201870). In general, Hg in fish tissue was higher at upstream sites in the Fountain Creek Basin but was present at all sites studied, and it was detected in tissue even when not detected in water (Nimmo and others, 2016). Conversely, Se concentration in tissue was higher at downstream sites, where several exceedances of the Environmental Protection Agency Se standards occurred (Nimmo and others, 2016). Later, Nimmo and others (2018) investigated the effect of Se concentration in water on midge distribution in the Fountain Creek Basin and found some taxa were associated with higher and others with lower Se concentrations. Although neither study was designed to show direct body function or performance effects of Se or Hg on fish and invertebrate communities or individuals, Se and Hg are known contaminants with negative effects on aquatic life (Boening, 2000; Hamilton, 2004; Lemly, 2004).

Fish and Invertebrate Response to Habitat

Instream habitat is typically described by measuring numerous characteristics of the stream channel, streambank, and streambed along a stream reach to explain spatial or temporal patterns in stream communities. In general, the presence of specific habitat types has been directly linked to fish and invertebrate community structure and population dynamics (Death and Winterbourn, 1995; Angermeier and Winston, 1998) where changes in habitat condition led to population declines or even the elimination of certain species (Rahel and others, 1996). Additionally, habitat diversity within a location has been directly related to local and regional species diversity (Boulinier and others, 1998) as the variety of habitat allowed numerous taxa to occupy a stream reach (Brown, 2003; Lepori and others, 2005). In contrast, locations with fewer habitat types often support fewer taxa (Rabení and others, 2005).

In the Fountain Creek Basin, habitat has contributed to spatial and temporal patterns in fish and invertebrate community structure. Spatial patterns in invertebrate communities are related to characteristics of the streambed (von Guerard, 1989a; Bruce, 2002) and stream wetted width (Bruce, 2002), which is expected because streambed and stream width naturally change as streams transition from mountain to plains environments. Typically, coarser particles, such as cobble, change to finer substrate, such as sand, as the stream widens and moves downstream across the plains. The proportion of substrate particle size classes (sand, gravel, or cobble) has a direct implication for the structure of aquatic communities. In the Fountain Creek Basin, substrate dominated by cobbles, with a mix of gravel and sand, contains invertebrates characterized by a higher proportion of disturbance- and contaminant-intolerant taxa (Bruce, 2002), whereas locations dominated by sand typically consist of fewer or no disturbance- and contaminant-intolerant taxa (Bruce, 2002). Although these associations are important and expected, it is difficult to separate human-induced changes in the streambed from naturally occurring gradients of transition zone streams. Zuellig, Bruce, and others (2007) attempted to lessen the effect of natural gradients and isolate human-induced effects by selecting habitat variables that were strongly related to urbanization but weakly related to elevation. In doing so, they found streambank characteristics, in part, contributed to spatial patterns in fish and invertebrate communities (Zuellig, Bruce, and others, 2007). Habitat was important in describing temporal changes in fish and invertebrate communities in the Fountain Creek Basin. For example, bank characteristics (lower bank capacity, maximum bank height) have contributed to site-specific temporal patterns in fish and invertebrate communities as well as substrate characteristics and the proportion of geomorphic units (percent of run, riffle, or pool) present at a site (Zuellig and others, 2010; Roberts and others, 2018).

Changes in substrate characteristics are often associated with natural or human-caused changes to the channel that result in the redistribution of substrate size classes along the streambed. When finer sediments (smaller particles) are redistributed along the streambed, it is often referred to as sedimentation, which is considered a primary factor affecting habitat quality, and results in decreases in fish and invertebrate production and changes in community structure (Waters, 1995). Often, finer sediments fill streambed pools and interstitial spaces between larger particles as well as directly abrade, dislodge, or bury benthic organisms and gravel spawning habitat (Jones and Clark, 1987; Wang and others, 2001; Shepherd and others, 2011). Certain families of caddisflies, mayflies, and stoneflies are highly susceptible to sedimentation and can easily be extirpated from systems that experience episodic sediment loads resulting from disturbance within the basin (McKenzie and others, 2020; Zuellig and others, 2002). High sediment loads can also alter fish behavior through a reduction in feeding success and growth rates; in addition, fine entrained sediments can damage gills, diminishing individual fitness (Kemp and others, 2011; Kjelland and others, 2015). Furthermore, high sedimentation during spawning showed delayed or decreased hatching rates in salmonids (Bowerman and others, 2014) and other fishes (Sutherland and others, 2002).

The direct effects of sedimentation on stream communities have not been well studied in the Fountain Creek Basin; however, characteristics of the streambed such as estimates of embeddedness, Wolman pebble counts (Wolman, 1954), and measures of sediment transport, have been shown to contribute to spatial and temporal changes in community structure (von Guerard, 1989a; Bruce, 2002; Zuellig and others, 2010; Roberts and others, 2018). Earlier studies in the Fountain Creek Basin related the spatial distribution of invertebrate communities to sediment transport and characteristics of the streambed (von Guerard, 1989a; Bruce, 2002). Locations with higher sediment transport and mobile benthic substrate (mostly sand) were often dominated by Oligochaetes (aquatic worms), and most insect species were less common or nonexistent. Locations with more stable substrate (mostly large gravel and cobble) had invertebrate communities with more diverse and contained higher proportions of mayflies, stoneflies, and caddisflies (von Guerard, 1989a). Several other studies found characteristics of the streambed (percentage of sand, gravel, and cobble) were also important in describing temporal patterns in fish and invertebrate community structure in the Fountain Creek Basin (Zuellig and others, 2012; Roberts and others, 2018).