Introduction

The Mohawk River is approximately 257 kilometers (km) long (McBride, 1994) and flows south from its headwaters near Boonville, New York, to Rome, where it turns and flows eastward, joining the Hudson River just north of Albany. The Mohawk River and New York State Canal System run together as a series of permanent and seasonal impoundments for most of the distance between Rome and Albany. Fish communities of the Mohawk River and associated sections of the New York State Canal System have been well documented (McBride, 1994, 2009; Carlson, 2015; George and others, 2016), but until 2019, comparatively less information was available regarding the status of fish communities in the extensive network of tributaries that feed the Mohawk River. This information gap was problematic because long-term shifts in species distributions or abundances from climate change or other anthropogenic stressors may go unnoticed in the absence of baseline data. In recognition of this, the 2012–16 Mohawk River Basin Action Agenda identified “expanding the river-wide fisheries survey to the tributaries” as a research priority for the watershed (New York State Department of Environmental Conservation, 2012).

The lack of information on the status of fish communities in tributaries to the Mohawk River has been made increasingly urgent by the approach of a high-profile invasive fish species, the round goby (Neogobius melanostomus). The species was first documented in North America in the 1990s in the Great Lakes region and was identified in the New York State Canal System near Utica in 2014 (U.S. Geological Survey, 2021a). Introductions of round goby have had profound effects on fish communities in the Great Lakes and other areas (Kornis and others, 2012). Round goby can outcompete native benthic fish species, reduce the recruitment of desirable gamefish through egg predation, and increase biomagnification of contaminants to higher trophic levels (Corkum and others, 2004). They can also carry the viral hemorrhagic septicemia virus (VHSV), which has been linked to significant fish kills in parts of New York (Farrell and others, 2017; Getchell and others, 2019), and some evidence indicates round goby are an important vector in avian botulism outbreaks (Ruffing, 2004; Hannett and others, 2011). Round goby are also an abundant food source for predators, however, and their establishment has been linked to increased health and growth rate of smallmouth bass (Micropterus dolomieu; Steinhart and others, 2004; Crane and others, 2015; Crane and Einhouse, 2016).

The effects of round goby on ecosystems in small and medium sized streams have not been investigated as thoroughly, and the overall effect of the species in lotic habitats remains poorly understood (Šlapanský and others, 2017). A number of studies have identified the theoretical potential for adverse effects on native fish assemblages through studies of diet overlap or other criteria (Poos and others, 2010; Abbett and others, 2013; Firth and others, 2021), yet few studies have attempted to directly quantify the ecological effects of round goby in streams. There is some circumstantial evidence indicating that round goby have displaced native benthic fish or altered the composition of macroinvertebrate communities (Krakowiak and Pennuto, 2008; Abbett and others, 2013; Raab and others, 2018), but a larger subset of research has shown subtle or negligible effects of round goby in lotic habitats (Kornis and others, 2013; Janáč and others, 2016; Šlapanský and others, 2017; Krabbenhoft, 2019). The relative abundance of round goby in streams appears highly variable, with estimates ranging from less than 10 percent to as high as approximately 80 percent of the entire fish community (Phillips and others, 2003; Krakowiak and Pennuto, 2008; Kornis and others, 2013). There is also mounting evidence that in some systems round goby exhibit seasonal migrations in and out of stream habitats (Pennuto and others, 2010; Blair and others, 2019), potentially further complicating efforts to assess population abundance and ecological effects. Thus, the high degree of uncertainty surrounding the effects of round goby in stream habitats, recent establishment of round goby in the Mohawk River watershed, and the lack of information on fish communities in streams of this region have created an urgent need to quantify the distribution and abundance of fishes in tributaries to the Mohawk River so that the effects of round goby invasion can be assessed and understood.

In response, the U.S. Geological Survey (USGS) and the New York State Department of Environmental Conservation, through its Mohawk River Basin Program, initiated a study in 2019 to characterize fish communities in tributaries to the Mohawk River. The objective of this study was to collect quantitative baseline information on fish communities and stream habitats, which could be used in the future to determine the effects of round goby on local fish assemblages, identify substrate and other habitat characteristics that facilitate or inhibit colonization by round goby, and evaluate the effects of other stressors or changes in watershed management. Obtaining information on the current distribution and abundance of darter species (Percidae: Etheostomatinae) was of particular interest. Darters are small bottom-dwelling fish (fig. 1) that have been shown to be highly vulnerable to displacement by round goby (Lauer and others, 2004; Kornis and others, 2012).

Equipment and Methods

Fish communities were surveyed at 19 tributaries to the Mohawk River between June and August 2019 (fig. 2; table 1). A single study reach was sampled on each tributary, with the exception of the Schoharie Creek where, because of its large size, two separate study reaches were sampled, for a total of 20 study reaches. Reaches ranged from 33 to 79 meters (m) in length and typically contained one or two complete geomorphic channel-unit sequences (Simonson and others, 1994; Fitzpatrick and others, 1998; Meador and others, 2003). Reaches were usually downstream from any major barriers and within the downstream-most 5 km of each stream (table 1) in order to collect data from areas that were at high risk for round goby colonization. The entire stream width was sampled during all surveys, with the exception of Cayadutta Creek, Ninemile Creek, West Canada Creek, and the two reaches on Schoharie Creek where the channel width and stream discharge were prohibitively large. At the Cayadutta Creek, Ninemile Creek, Schoharie Creek near Wellsville, and West Canada Creek sites, a braided section of the channel was surveyed in which the entire width could be sampled, and at the Schoharie Creek at Burtonsville site, a single near-shore reach was isolated using three blocking seines.

Fish communities were sampled using multipass depletion electrofishing surveys. During each survey, fish were collected from seine-blocked reaches in three consecutive passes, with one person operating a Smith-Root LR–24 backpack electrofisher and three to five people netting fish. All fish were identified to species, measured for total length, weighed, and returned to the stream after all passes were completed. For some small and highly abundant species, lengths and weights were recorded from a subsample of 30 fish across their length distribution, after which mean length and pooled weights were recorded in batches of up to 30 fish.

Information on the dimensions and substrate composition of the sampled reaches was also collected during each survey. The total length of the reach and the widths of 10 evenly spaced transects were measured and used to calculate mean reach width and total area sampled. Substrate size was characterized using a reach-averaged, modified Wolman pebble count (Wolman, 1954) in which 10 particles were measured across each of the 10 perpendicular transects (100 total particles). The intermediate axis or “b-axis” of each particle was measured to the nearest millimeter using a meter stick. Fine particles less than 2 millimeters (mm) in size were not measured and instead were classified as sand if they were grainy to the touch or as silt if they were smooth to the touch. Bedrock or boulders with an intermediate axis greater than (>) 1 m were classified as >1,000 mm. The pebble count data were used to estimate commonly reported measures of substrate size, including the 16th percentile (D₁₆), 50th percentile or median (D₅₀), and 84th percentile (D₈₄), using linear interpolation with the PERCENTILE.INC function in Microsoft Excel. All data from the fish surveys and habitat characterization are available in George and others (2021b).

The number of fish captured during each pass was used to generate population estimates, with associated 95 percent confidence intervals for each species at each site, using the Carle-Strub method (Carle and Strub, 1978) with the FSA package (Ogle and others, 2021) in R (R Core Team, 2021). Population estimates and associated confidence intervals were standardized by the area of the sampled reach to produce estimates of fish density per 1,000 square meters (m²) of stream. Estimates of population biomass and associated 95 percent confidence intervals were generated by multiplying the density estimates and associated confidence intervals by the mean weight for each species at each site. In rare instances where the lower bound of the confidence interval around an estimate was less than zero, the negative value was replaced with a zero. Estimates of total density and biomass for the entire community (all species combined) were also generated for each site using the same process described above. Additionally, the diversity of the community was assessed for each site using two indices: species richness and Shannon’s index of diversity (Shannon, 1948). Species richness was calculated as the total number of species captured at a site and Shannon’s index of diversity was calculated as:

Results

A total of 11,794 individual fish and 37 fish species were captured during the 20 surveys. Longnose dace (Rhinichthys cataractae), white sucker (Catostomus commersonii), blacknose dace (Rhinichthys atratulus), fantail darter (Etheostoma flabellare), and creek chub (Semotilus atromaculatus) were the most frequently encountered species, occurring at 18, 18, 17, 17, and 16 of the 20 sites, respectively (table 2). Eight species were present at more than half the sites, whereas the remaining 29 species were encountered infrequently, occurring at less than half the sites. Species richness ranged from 3 at Sandsea Kill to 19 at Oriskany Creek, and Shannon’s index of diversity ranged from 0.33 at Sandsea Kill to 2.61 at Sauquoit Creek. The total density of all fish species (community density) ranged from 162 fish per 1,000 m² at Sauquoit Creek to 7,182 fish per 1,000 m² at Crane Creek (fig. 3), and the total biomass of all fish species (community biomass) ranged from 498 grams per 1,000 m² at Revine Creek to 10,588 grams per 1,000 m² at Crane Creek (fig. 4; table 3).

Six darter species were captured during the 20 surveys, and at least one darter species was captured at every site except for Sandsea Kill. Fantail darter was the most frequently encountered darter, found at 17 of the 20 sites, followed by tessellated darter (Etheostoma olmstedi) at 11 sites, rainbow darter (Etheostoma caeruleum) and logperch (Percina caprodes) each at 9 sites, blackside darter (Percina maculata) at 4 sites, and greenside darter (Etheostoma blennioides) at 3 sites (table 2). Fantail darter was widely distributed throughout the watershed, while tessellated darter and logperch were found less commonly but were also widely distributed (fig. 5). Rainbow darter was concentrated exclusively at the eastern end of the study area, occurring at 9 of the 10 easternmost sites. In contrast, blackside darter was found exclusively at the western end of the study area, occurring at four of the six westernmost sites. No discernable pattern was evident in the distribution of greenside darter, as the species was found at only three sites that were geographically distant from one another.

Alplaus Kill at Glenridge, N.Y

There were 12 fish species present at the Alplaus Kill site (fig. 6). The total community density was 1,482 fish per 1,000 m², and rainbow darter was the most abundant species, composing 43 percent of the community. The total community biomass was 1,423 grams per 1,000 m², and smallmouth bass had the greatest biomass, composing 43 percent of the community. There were five darter species present—fantail darter, greenside darter, logperch, rainbow darter, and tessellated darter—which together composed 68 and 47 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 31 mm, D₅₀ of 64 mm, and D₈₄ of 122 mm (table 3).

Auries Creek at Auriesville, N.Y

There were nine fish species present at the Auries Creek site (fig. 7). The total community density was 488 fish per 1,000 m², and blacknose dace was the most abundant species, composing 40 percent of the community. The total community biomass was 915 grams per 1,000 m², and blacknose dace had the greatest biomass, composing 34 percent of the community. There were two darter species present—fantail darter and rainbow darter—which together composed 24 and 20 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 20 mm, D₅₀ of 79 mm, and D₈₄ of 210 mm (table 3).

Caroga Creek Near Fort Plain, N.Y

There were 12 fish species present at the Caroga Creek site (fig. 8). The total community density was 305 fish per 1,000 m², and longnose dace was the most abundant species, composing 28 percent of the community. The total community biomass was 913 grams per 1,000 m², and brown bullhead (Ameiurus nebulosus) had the greatest biomass, composing 16 percent of the community. Fantail darter was the only darter species present, which composed 2 and less than 1 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 51 mm, D₅₀ of 118 mm, and D₈₄ of 314 mm (table 3).

Cayadutta Creek Below Commons Road at Berryville, N.Y

There were 14 fish species present at the Cayadutta Creek site (fig. 9). The total community density was 3,149 fish per 1,000 m², and longnose dace was the most abundant species, composing 78 percent of the community. The total community biomass was 6,525 grams per 1,000 m², and longnose dace had the greatest biomass, composing 57 percent of the community. There were three darter species present—logperch, rainbow darter, and tessellated darter—which together composed less than 1 percent of both the community density and biomass. The pebble count yielded a particle size distribution with a D₁₆ of 38 mm, D₅₀ of 97 mm, and D₈₄ of 215 mm (table 3).

Crane Creek Below Old River Road at Careys Corners, N.Y

There were 10 fish species present at the Crane Creek site (fig. 10). This site was noteworthy because it produced the greatest estimates of community density and community biomass in the entire study. The total community density was 7,182 fish per 1,000 m², and blacknose dace was the most abundant species, composing 72 percent of the community. The total community biomass was 10,588 grams per 1,000 m², and blacknose dace had the greatest biomass, composing 59 percent of the community. There were two darter species present—fantail darter and tessellated darter—which together composed 12 and 10 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 27 mm, D₅₀ of 50 mm, and D₈₄ of 113 mm (table 3).

Crum Creek Near Indian Castle, N.Y

There were 15 fish species present at the Crum Creek site (fig. 11). The total community density was 4,123 fish per 1,000 m², and blacknose dace was the most abundant species, composing 60 percent of the community. The total community biomass was 6,614 grams per 1,000 m², and blacknose dace had the greatest biomass, composing 32 percent of the community. Fantail darter was the only darter species present, which composed 12 and 7 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 41 mm, D₅₀ of 85 mm, and D₈₄ of 173 mm (table 3).

East Canada Creek Near St. Johnsville, N.Y

There were eight fish species present at the East Canada Creek site (fig. 12). The total community density was 499 fish per 1,000 m², and longnose dace was the most abundant species, composing 44 percent of the community. The total community biomass was 1,534 grams per 1,000 m², and stonecat (Noturus flavus) had the greatest biomass, composing 39 percent of the community. There were two darter species present—fantail darter and tessellated darter—which together composed 25 and 19 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 67 mm, D₅₀ of 126 mm, and D₈₄ of 220 mm (table 3).

Flat Creek at Sprakers, N.Y

There were nine fish species present at the Flat Creek site (fig. 13). The total community density was 595 fish per 1,000 m², and blacknose dace was the most abundant species, composing 64 percent of the community. The total community biomass was 925 grams per 1,000 m², and blacknose dace had the greatest biomass, composing 46 percent of the community. There were three darter species present—fantail darter, logperch, and rainbow darter—which together composed 7 and 14 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 17 mm, D₅₀ of 85 mm, and D₈₄ of >1,000 mm (table 3).

Fulmer Creek at Mohawk, N.Y

There were eight fish species present at the Fulmer Creek site (fig. 14). The total community density was 803 fish per 1,000 m², and white sucker was the most abundant species, composing 62 percent of the community. The total community biomass was 708 grams per 1,000 m², and longnose dace had the greatest biomass, composing 30 percent of the community. There were four darter species present—blackside darter, fantail darter, logperch, and tessellated darter—which together composed 15 and 43 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 16 mm, D₅₀ of 43 mm, and D₈₄ of 94 mm (table 3).

Lasher Creek at Mouth Near Randall, N.Y

There were 12 fish species present at the Lasher Creek site (fig. 15). The total community density was 258 fish per 1,000 m², and longnose dace was the most abundant species, composing 19 percent of the community. The total community biomass was 673 grams per 1,000 m², and white sucker had the greatest biomass, composing 46 percent of the community. There were two darter species present—fantail darter and rainbow darter—which together composed 5 and 3 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 11 mm, D₅₀ of 33 mm, and D₈₄ of 141 mm (table 3).

Lisha Kill Northwest of Niskayuna, N.Y

There were 11 fish species present at the Lisha Kill site (fig. 16). The total community density was 3,525 fish per 1,000 m², and central stoneroller (Campostoma anomalum) was the most abundant species, composing 30 percent of the community. The total community biomass was 5,755 grams per 1,000 m², and creek chub had the greatest biomass, composing 46 percent of the community. There were three darter species present—fantail darter, rainbow darter, and tessellated darter—which together composed 10 and 4 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 6 mm, D₅₀ of 25 mm, and D₈₄ of 46 mm (table 3).

Moyer Creek at Mouth at Frankfort, N.Y

There were 14 fish species present at the Moyer Creek site (fig. 17). The total community density was 2,819 fish per 1,000 m², and longnose dace was the most abundant species, composing 70 percent of the community. The total community biomass was 3,422 grams per 1,000 m², and longnose dace had the greatest biomass, composing 65 percent of the community. There were three darter species present—blackside darter, fantail darter, and logperch—which together composed 2 and 7 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 23 mm, D₅₀ of 63 mm, and D₈₄ of 125 mm (table 3).

Ninemile Creek Near Oriskany, N.Y

There were 15 fish species present at the Ninemile Creek site (fig. 18). The total community density was 974 fish per 1,000 m², and longnose dace was the most abundant species, composing 49 percent of the community. The total community biomass was 7,769 grams per 1,000 m², and smallmouth bass had the greatest biomass, composing 46 percent of the community. There were three darter species present—fantail darter, logperch, and tessellated darter—which together composed 19 and 3 percent of the community density and biomass, respectively. Additionally, round goby were present in the study reach with an estimated density of 20 fish per 1,000 m² and biomass of 68 grams per 1,000 m² (fig. 19). The pebble count yielded a particle size distribution with a D₁₆ of 28 mm, D₅₀ of 51 mm, and D₈₄ of 81 mm (table 3).

Oriskany Creek Near Ballfield at Oriskany, N.Y

There were 19 fish species present at the Oriskany Creek site (fig. 20). The total community density was 769 fish per 1,000 m², and longnose dace was the most abundant species, composing 34 percent of the community. The total community biomass was 1,959 grams per 1,000 m², and green sunfish (Lepomis cyanellus) had the greatest biomass, composing 27 percent of the community. There were five darter species present—blackside darter, fantail darter, greenside darter, logperch, and tessellated darter—which together composed 21 and 9 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 24 mm, D₅₀ of 51 mm, and D₈₄ of 84 mm (table 3).

Revine Creek at New York State Thruway at Auriesville, N.Y

There were eight fish species present at the Revine Creek site (fig. 21). The total community density was 369 fish per 1,000 m², and white sucker was the most abundant species, composing 21 percent of the community. The total community biomass was 498 grams per 1,000 m², and creek chub had the greatest biomass, composing 23 percent of the community. There were three darter species present—fantail darter, rainbow darter, and tessellated darter—which together composed 14 and 15 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 12 mm, D₅₀ of 40 mm, and D₈₄ of 99 mm (table 3).

Sandsea Kill at Pattersonville, N.Y

There were three fish species present at the Sandsea Kill site (fig. 22). The total community density was 2,298 fish per 1,000 m², and blacknose dace was the most abundant species, composing 92 percent of the community. The total community biomass was 2,510 grams per 1,000 m², and blacknose dace had the greatest biomass, composing 83 percent of the community. There were no darter species present. The pebble count yielded a particle size distribution with a D₁₆ of 30 mm, D₅₀ of 70 mm, and D₈₄ of 170 mm (table 3).

Sauquoit Creek at Yorkville, N.Y

There were 18 fish species present at the Sauquoit Creek site (fig. 23). The total community density was 162 fish per 1,000 m², and longnose dace was the most abundant species, composing 12 percent of the community. A single brown trout (Salmo trutta) with a length of 448 mm and mass of 1,214 grams, the largest fish captured in the entire study, was responsible for nearly half of the biomass in the study reach (fig. 24). The total community biomass was 2,652 grams per 1,000 m², and brown trout had the greatest biomass, composing 46 percent of the community. There were four darter species present—blackside darter, fantail darter, logperch, and tessellated darter—which together composed 19 and 3 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 4 mm, D₅₀ of 21 mm, and D₈₄ of 71 mm (table 3).

Schoharie Creek Below Bridge at Burtonsville, N.Y

There were five fish species present at the Schoharie Creek at Burtonsville site (fig. 25). The total community density was 476 fish per 1,000 m², and smallmouth bass was the most abundant species, composing 83 percent of the community. The total community biomass was 4,227 grams per 1,000 m², and smallmouth bass had the greatest biomass, composing 88 percent of the community. Rainbow darter was the only darter species present, which composed 3 and less than 1 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 47 mm, D₅₀ of 178 mm, and D₈₄ of 575 mm (table 3).

Schoharie Creek Near Wellsville, N.Y

There were 14 fish species present at the Schoharie Creek near Wellsville site (fig. 26). The total community density was 1,281 fish per 1,000 m², and smallmouth bass was the most abundant species, composing 51 percent of the community. The total community biomass was 3,360 grams per 1,000 m², and smallmouth bass had the greatest biomass, composing 61 percent of the community. There were four darter species present—fantail darter, greenside darter, logperch, and rainbow darter—which together composed 33 and 18 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 56 mm, D₅₀ of 111 mm, and D₈₄ of 300 mm (table 3).

West Canada Creek at Herkimer, N.Y

There were 14 fish species present at the West Canada Creek site (fig. 27). The total community density was 313 fish per 1,000 m², and longnose dace was the most abundant species, composing 24 percent of the community. The total community biomass was 500 grams per 1,000 m², and longnose dace had the greatest biomass, composing 25 percent of the community. There were two darter species present—fantail darter and tessellated darter—which together composed 12 and 10 percent of the community density and biomass, respectively. The pebble count yielded a particle size distribution with a D₁₆ of 21 mm, D₅₀ of 56 mm, and D₈₄ of 154 mm (table 3).

Findings and Conclusions

The findings from this study have several important implications. First, it appears that round goby had not extensively colonized tributaries to the Mohawk River at the time the reaches were surveyed in summer 2019. Round goby were only captured in one tributary, and the density of that population was low. The near absence of round goby from tributaries to the Mohawk River in 2019 was expected given the existing information on the sparse distribution of round goby in the mainstem of the Mohawk River at that time. At the end of 2019, the easternmost capture of round goby was near Utica, although environmental DNA data suggested the species may have been present further east than fish capture records indicate (George and others, 2021a). Overall, this information indicates that fish communities in tributaries to the Mohawk River remained largely unaffected by round goby in 2019, and the data described in this report represent the preinvasion baseline condition, thus fulfilling a primary objective of this project.

Second, although fish communities were generally dominated by minnow species such as blacknose dace and longnose dace, darters were an important component of communities, averaging 15 and 12 percent of the density and biomass, respectively. The distribution and abundance of darter species were of particular interest in this study because darters and round goby occupy similar niches, and darter populations are often negatively affected by the presence of round goby. The darter community in tributaries to the Mohawk River comprised six species: blackside darter, fantail darter, greenside darter, logperch, rainbow darter, and tessellated darter, and some evidence indicates the distribution and abundance of these species may be changing. Of this group, fantail darter, greenside darter, and tessellated darter are believed to be native to the Mohawk River watershed, blackside darter and rainbow darter are believed to be nonnative to the Mohawk River watershed, and there is not a consensus regarding the status of logperch (Carlson and others, 2016a; U.S. Geological Survey, 2022). Rainbow darter, for example, is only native to the western portion of New York State and was first identified in the Mohawk River watershed in 2009 in Fox Creek, a tributary to Schoharie Creek (Carlson and others, 2016a). The surveys described in this report, which were conducted only 10 years later (2019), indicated that rainbow darter were present in nine tributaries in the eastern half of the study area and was the dominant darter species at most of these locations. This suggests that the species has expanded rapidly since its recent introduction to the Mohawk River watershed and may be outcompeting other resident darter species.

Lastly, data from future fish community surveys in tributaries to the Mohawk River could be compared with the 2019 data provided in this report to address a number of important research and management questions. First, it would be instructive to determine if round goby indiscriminately colonizes all the tributaries included in this study or if the species shows affinity for certain habitat characteristics, such as substrate size or water velocity. In turn, this information could be used to predict the invasion susceptibility of streams in other adjacent watersheds, such as those in the lower Hudson River and Lake Champlain. It would also be valuable to obtain estimates of how quickly round goby populations are expanding upstream in tributaries to the Mohawk River. A few studies have estimated rates of upstream expansion in tributaries to the Great Lakes and other areas with values ranging from 0.5 to 4 kilometers per year (Bronnenhuber and others, 2011; Kornis and others, 2012; Šlapanský and others, 2017), but it is not known if these estimates are applicable to streams in central and eastern New York. Most importantly, future fish community data could be used to evaluate the effects of established round goby populations on the composition of resident fish assemblages. In particular, future analyses could compare quantitative estimates of darter population density and biomass to determine if round goby have extirpated some of these species, reduced their density, restricted their distribution, or caused no discernable effect. Similarly, future studies might assess changes in populations of stonecat, a small native catfish species that was present at five of the study sites and may also be adversely affected by round goby (Poos and others, 2010; Kornis and others, 2012). Finally, temporal changes in streams of this watershed during the past century could be analyzed more in depth using historical data from the New York State Fish Atlas Database (Carlson and others, 2016b). Thus, it may be prudent to repeat fish community surveys at the same locations and using the same methods 5 to 10 years after the surveys described in this report. Together, this new information and the information from the 2019 surveys could be used to identify the ecological effects of round goby in tributaries to the Mohawk River and to project future effects on aquatic ecosystems in adjacent watersheds.