Species Status Assessment (SSA) Report for the ( harryi) Version 1.2

Ozark chub (Photo credit: Dustin Lynch, Natural Heritage Commission)

August 2019 U.S. Fish and Wildlife Service - Arkansas Ecological Services Field Office

This document was prepared by Alyssa Bangs (U. S. Fish and Wildlife Service (USFWS) – Arkansas Ecological Services Field Office), Bryan Simmons (USFWS— Ecological Services Field Office), and Brian Evans (USFWS –Southeast Regional Office). We greatly appreciate the assistance of Jeff Quinn (Arkansas Game and Fish Commission), Brian Wagner (Arkansas Game and Fish Commission), and Jacob Westhoff (Missouri Department of Conservation) who provided helpful information and review of the draft document. We also thank the peer reviewers, who provided helpful comments.

Suggested reference: U.S. Fish and Wildlife Service. 2019. Species status assessment report for the Ozark chub (Erimystax harryi). Version 1.2. August 2019. Atlanta, GA.

CONTENTS Chapter 1: Executive Summary 1 1.1 Background 1 1.2 Analytical Framework 1 CHAPTER 2 – Species Information 4 2.1 and Genetics 4 2.2 Species Description 5 2.3 Range 6 Historical Range and Distribution 6 Current Range and Distribution 8 2.4 Life History Habitat 9 Growth and Longevity 9 Reproduction 9 Feeding 10 CHAPTER 3 –Factors Influencing Viability and Current Condition Analysis 12 3.1 Factors Influencing Viability 12 Sedimentation 12 Water Temperature and Flow 14 Impoundments 15 Water Chemistry 16 Habitat Fragmentation 17 3.2 Model 17 Analytical Units 18 Watershed Parameters 18 Population Parameters 20 3.3 Current Condition—3Rs 21 Resiliency 21 Representation 21 Redundancy 22 CHAPTER 4 – Future Condition Analysis/Species Viability 27 4.1 Parameters 27 Urban Cover 27 Riparian Forest Cover 28

Water Quality 28 Abundance 28 4.2 Scenarios 32 Scenario 1 (positive habitat change) 32 Scenario 2 (increased urbanization) 32 Scenario 3 (increased urbanization, other parameters slightly decreased) 32 Scenario 4 (increased urbanization, other parameters decreased) 32 4.3 Summary 33 REFERENCES 34 Appendix A. Ozark Chub Ecological Management Units and Watersheds 41 Upper White River Ecological Management Unit 42 Middle White River Ecological Management Unit 44 Middle Black River Ecological Management Unit 45 St Francis River Ecological Management Unit 47 Little Red River Ecological Management Unit 48

Chapter 1: Executive Summary 1.1 Background This report summarizes the results of a species status assessment (SSA) conducted for the Ozark chub (Erimystax harryi), which includes relevant information about the species’ life history characteristics, how those characteristics are affected by stressors, and conservation measures to address those stressors. The U.S. Fish and Wildlife Service (Service) was petitioned to list 404 aquatic, riparian, and wetland species, including the Ozark chub, as endangered or threatened under the Endangered Species Act of 1973, as amended (Act) on April 20, 2010, by the Center for Biological Diversity (Center for Biological Diversity 2010, p. 417-418). In September of 2011, the Service found the petition presented substantial scientific or commercial information indicating that listing 374 species, including Ozark chub, may be warranted. Thus, we conducted a SSA to compile the best scientific and commercial data available regarding the species biology and factors influencing species viability. 1.2 Analytical Framework The SSA framework (USFWS 2016) is intended to be an in-depth review of the species’ biology and threats, an evaluation of its biological status, and an assessment of the resources and conditions needed to maintain long-term viability. The intent is for the SSA report to be easily updated as new information becomes available and to support all functions of the Endangered Species Program from candidate assessment to listing to consultations and recovery. As such, the SSA report will be a living document used to inform decisions under the ESA.

The Ozark chub SSA is intended to provide the biological support for the decision on whether to propose listing the species as threatened or endangered and, if so, to determine whether it is prudent to designate critical habitat areas essential to its conservation. This report is not a decisional document by the Service; rather, it provides a review of available information strictly related to the biological status of Ozark chub. A listing decision will be made by the Service after reviewing this document and all relevant laws, regulations, and policies. The results of a proposed decision will be announced in the Federal Register with appropriate opportunities for public input. Using the SSA framework (Figure 1), we consider what the species needs to maintain viability (the species’ ability to sustain populations in the wild over time) by characterizing the status of the species in terms of its redundancy, representation, and resiliency (USFWS 2016, entire; Wolf et al. 2015, entire). For the purpose of this assessment, we generally define viability as the ability of the species to sustain populations in natural stream ecosystems within a biologically meaningful timeframe. Resiliency, redundancy, and representation are defined as follows: ● Resiliency is assessed at the population level and reflects a species’ ability to withstand stochastic events (arising from random factors). Demographic measures that reflect population health, such as fecundity, survival, and population size, are the metrics used to evaluate resiliency. Resilient populations are better able to withstand disturbances such as random fluctuations in birth rates (demographic stochasticity), variations in rainfall (environmental stochasticity), and the effects of anthropogenic activities.

● Representation is assessed at the species’ level and characterizes the ability of a species to adapt to changing environmental conditions. Metrics such as a species’ adaptive potential and genetic and ecological variability can be used to assess representation. Representation is directly correlated to a species’ ability to adapt to changes (natural or human-caused) in its environment. In the absence of species-specific genetic and ecological diversity information, we evaluate representation based on the extent and variability of habitat characteristics within the geographical range.

● Redundancy is also assessed at the species level and reflects a species’ ability to withstand catastrophic events (such as a rare destructive natural event or episode involving many populations) by having a sufficient number of populations. Redundancy is about spreading the risk of such an event across multiple, resilient populations. As such, redundancy can be measured by the number and distribution of resilient populations across the range of the species.

The decision whether to list a species is based not on a prediction of the most likely future for the species, but rather on an assessment of the species’ risk of extinction. Therefore, to inform this assessment of extinction risk, we describe the species’ current biological status and assess how this status may change in the future under a range of scenarios to account for the uncertainty of the species’ future. We evaluate the current biological status of Ozark chub by assessing the primary factors negatively and positively affecting the species to describe its current condition in terms of resiliency, redundancy, and representation (together, the 3Rs). We then evaluate the future biological status of Ozark chub by describing a range of plausible future scenarios representing a range of conditions for the primary factors affecting the species and forecasting the most likely future condition for each scenario in terms of the 3Rs. As a matter of practicality, the full range of potential future scenarios and range of potential future conditions for each potential scenario are too large to individually describe and analyze. These scenarios do not include all possible futures, but rather include specific plausible scenarios that represent examples from the continuous spectrum of possible futures. Consequently, the results of this SSA do not describe the overall risk to the species. Recognizing these limitations, the results of this SSA nevertheless provide a framework for considering the overall risk to the species in listing decisions. This report includes: a description of Ozark chub resource needs at both individual and population levels (Chapter 1); a characterization of the historical and current distribution of populations across the species’ range (Chapter 2); an assessment of factors that contributed to the current and future status of the species and degree to which various factors influenced viability (Chapter 3); and a synopsis of factors characterized in earlier chapters as a means of examining future biological status of the species (Chapter 4). This document is a compilation of the best available scientific information (and associated uncertainties regarding that information) used to assess Ozark chub viability.

CHAPTER 2 – Species Information 2.1 Taxonomy and Genetics Ozark chub (Erimystax harryi) belongs to the family of fishes. It is currently recognized as a valid taxon by the American Fisheries Society (Page et al. 2013, p. 70) and is listed as such in the Integrated Taxonomic Information System (ITIS) database (http://www.itis.gov). The currently accepted classification is (ITIS 2018): Phylum: Chordata Class: Teleostei Order: Family: Cyprinidae Genus: Erimystax Species: Erimystax harryi

Ozark chub was initially recorded as a disjunct population of (Erimystax dissimilis, then dissimilis), now restricted to east of the River (Figure 2). In the late 1800s and early 1900s, what was likely Ozark chub were described as H. watauga (Jordan 1888, p. 355), but that name was never accepted for this species. Hubbs and Crowe (1956, p. 6) elevated the Ozark mountain populations to subspecies level, as Hybopsis dissimilis harryi, based on a specimen collected in 1940 from the White River in Barry County, Missouri. Other specimens from northern Arkansas and southern Missouri were also used to describe the species. Research by Jenkins and Lachner (1971, entire) suggested Ozark chub was distinct enough to warrant elevation to species level, and it was formerly elevated to H. harryi based on genetic and morphological data in 1986 (Harris 1986, entire). The Hybopsis genus “had a chaotic taxonomic history” (Dimmick 1993, p. 173), but genetic and morphological research in the late 20th century led to the reclassification of the Ozark chub to Erimystax (Mayden 1989, entire; Coburn and Cavender 1992, entire). Ozark chub is a member of the dissimilis species group with the Streamline Chub and the (Erimystax insignis) (Harris 1986; Mayden 1989; Dimmick 1993, p. 181). The Streamline Chub is found in the River drainage whereas the Blotched Chub is endemic to the Cumberland and Tennessee River drainages. Both Streamline Chub and Blotched Chub are found east of the ; it is likely that historically there was habitat in southern that connected the eastern and Ozark populations of the parent species (Conant 1960, p. 33). Figure 2. The range of Streamline Chub (Erimystax dissimilis striped) and Ozark Chub (Erimystax harryi, dotted). From Conant 1960, p. 33.

2.2 Species Description Ozark Chub are small (43-106 millimeters (mm) (3.0-4.5 inches(in)) standard length [SL; the length measured from the tip of the snout to the last vertebra, which excludes the length of the tail fin]), slender freshwater fish that tend to occupy runs and riffles (i.e., shallow areas of fast, turbulent flow) over gravel in clear, cool water streams (Pflieger 1997, p. 121). Individuals are yellow green, tan, or gray dorsally with dark blotches along the midline. Breeding males develop small tubercles on the head and forward part of the body but no nuptial coloration. Ozark Chub is the only Erimystax species to have a long, looped gut, up to twice SL. Some populations tend toward the loss of lateral spots, and individuals in the St. Francis and Little Red Rivers display morphology intermediate between other Ozark Chub and Streamline Chub, including a shortened gut and enlarged or absent lateral spots (Figure 3; Harris 1986, p.45).

Figure 3. From top to bottom, Streamline Chub side view, Ozark Chub (Current River) side view, Ozark Chub (Current River) top view, Ozark Chub (Little Red River) side view. From Harris 1986, p. 52-53.

2.3 Range Historical Range and Distribution The first scientific collections of fishes in Arkansas were in the second half of the 19th century (Meek 1894, entire; Black 1940, p. 4-8). While early (pre-1900) survey data are sparse, a few researchers surveyed many of the Ozark and Ouachita streams during this period (Jordon 1888, entire; Meek 1894, entire). Arkansas at the time was only “thinly settled” and the White River basin was “one of the clearest and most beautiful streams in the Mississippi Valley” (Meek 1894, p. 69). The majority of fish collections in the state come from the 1930s or later. Ozark Chub is endemic to the White River basin, including the Black River and its tributaries, the Little Red River, and the St. Francis River in Arkansas and Missouri (Figure 5). These drainages were likely part of an extensive highland drainage area later fragmented by glaciation and other geologic events (Mayden 1988, entire; Strange and Burr 1997, entire). The St. Francis River may have been part of a preglacial drainage across southern Illinois that allowed the movement of the E. dissimilis ancestral species from the east to the Ozark Highlands (Mayden 1988, p. 341; Strange and Burr 1997, entire). The St. Francis and Little Red River populations are morphologically and meristically somewhat intermediate between other populations of Ozark Chub and Streamline Chub, supporting this theory (Figure 4; Harris 1986, p. 45). The earliest recorded collections of this species, identified as Hybopsis dissimilis and H. watauga, were in the late 1800s from the White River and South Fork Little Red River in Arkansas (Figure 5; Jordan 1888, p. 355; Meek 1894, p. 78, 82). The next recorded collections were in the late 1930s from the Black and Current rivers in south-eastern Missouri. The holotype, as subspecies Hybopsis dissimilis harryi, was collected and described from the White River in Missouri in 1940 (Hubbs and Crowe 1956, p. 6). By the 1950s, surveyors had documented the species in the St. Francis River, the upper White River (main stem, War Eagle Creek, and Richland Creek), King’s River, James River, Swan Creek, Buffalo River, and lower main stem White River. Surveys after 1951 continued collecting the species in new (Bull Creek, Beaver Creek, Sylamore River, Piney Creek, Long Creek, Strawberry River, Spring River, and Eleven Point River) and previously documented waterways throughout the White River basin, and included a few collections in the Archey Fork Little Red River. Ozark Chub was collected for the first time in Flat Creek after 2000. Multiple collectors commented that this species was uncommon in collections throughout the White River basin but was often collected in large numbers when found; others suggested that it was rare to find > 50 Ozark Chub at a single site (Keith 1964, p. 22, 29; Harris 1986, p. 245; Pflieger 1997, p. 121). The species was described as “scarce” in the White River (Meek 1894, p. 78), and from “rare” to “abundant” in the headwaters of the White River (Keith 1964, p. 72-72). In the Buffalo River it was “numerous in its preferred habitat and…second in numerical abundance to [Central Stoneroller] Campostoma anomalum” (Harris 1986, p. 245). Martin and Campbell (1953, p. 46, 53) found Ozark Chub represented 0.7% of small fishes collected in the Black River and was the most numerous of bottom feeding species in slow riffles and a common species in the deeper water along riffle margins.

Figure 4. Population centroids with 95% confidence circles for Streamline Chub (left) and Ozark Chub (right) in canonical space determined from a meristic and morphometric data set. STF: St. Francis, LRR: Little Red, CUR: Current, SPR: Spring, BLK: Black, WRA: White, ELP: Eleven Point, BRA: Buffalo (from Harris 1986, p. 26-27).

Removed records – There are records of Ozark Chub, primarily as Hybopsis dissimilis, collected in the Neosho River of and , the Saline and Ouachita Rivers in the Ouachita Mountains of south-central Arkansas, and several northern-flowing rivers in Missouri. These records are all likely misidentified collections of similar species, probably the (Erimystax x-punctatus), which is known to occur in these watersheds (Cross 1954, p. 308; Wildhaber et al. 2000, p. 248; Westhoff, pers.comm. 2019). The Saline and Ouachita River collections were identified as H. dissimilis by Jordan (1888, p. 356) rather than his newly described White River species H. watauga, (very likely Ozark Chub), which suggests that these southern Arkansas records are not Ozark Chub. Although there is some evidence for a pre- glaciation connection between these watersheds and the White River (Matthews and Robison 1988, p. 368; Mayden 1988, p. 341), these are sporadic older records in areas that have been repeatedly surveyed; if Ozark Chub was present it is likely that we would have more recent collections of the species.

Current Range and Distribution

Figure 5. Range of the Ozark Chub (Erimystax harryi) with collection data. Colored watersheds are Ecological Management Units and gray areas are presumed historically occupied habitat.

Ozark Chub occurs in 21 watersheds (primarily classified at the HUC 8 level) within the White (HUC 6) and St. Francis River (HUC 8) basins. The White River is approximately 1,210 km long with a drainage basin of approximately 72,520 km2. There are eight dams on the main stem White River, built between 1913 and 1966. Five of these dams (Sequoyah, Beaver, Table Rock, Taneycomo, and Bull Shoals) are high-head hydroelectric dams. These reservoirs also provide flood control and drinking water. Approximately 40% of collection records for this species were made since 2000, and all historically occupied watersheds have recent collections (collected after 2000) except for those within the Upper White River and tributaries upstream of Beaver Lake, the main stem White River, Black River, and Archey Fork Little Red River. A few recent collections in the Buffalo and Kings River documented >100 individuals at a site; the maximum number of Ozark Chub collected at one site was 241. However, most collections are < 50 individuals.

We have split the watersheds making up the current range of the Ozark Chub into five Ecological Management Units (EMU) for data management and analysis purposes. These EMUs are discussed in Appendix A. 2.4 Life History Habitat The general habitat of Ozark Chub is moderately large, clear streams with continuous strong flow and clean gravel and rocky bottoms. They prefer larger streams of lower gradients (<1.52 m/km) rather than headwater or higher gradient upstream areas like other Erimystax species in the Ozark Mountains (Keith 1964, p. 22, 72-73). Adults prefer runs and riffles approximately 45 – 60 cm deep over gravel, habitat directly below riffles, or shallow pools with noticeable current (Pflieger 1971, p. 121; Harris 1986, p. 245). Adults rarely occur in side channel habitats or torrential raceways with bedrock or boulder, but have been collected in low numbers in all available microhabitats (Harris 1986, p. 245). Young Ozark Chub occupy backwater and shoreline or side channel habitats with low velocity, such as the shallow marginal areas of pool headwaters (Martin and Campbell 1953, p. 53; Harris 1986, p. 245). Peterson (2004, p. 44) found that Ozark Chub (along with Bigeye Shiner [ boops] and Yoke Darter [Etheostoma juliae]) is associated primarily with sites on the main stem of the Buffalo River with large basins and low levels of agriculture. Ozark Chub capture rates are lower in winter, likely due to different habitat preferences in cooler months, such as areas of deeper water and pools rather than riffles (Keith 1964, p. 38, 41, 53).

Growth and Longevity Initial juvenile growth rate is relatively high, with young of the year (YOY) reaching 20 mm or more in just a few months (Harris 1986, p. 260). Males grow faster than females for the first twelve months, and are generally larger than females for the first 20 months. However, females are larger by the end of the second year, reach a larger maximum size and have higher survival rates overall; although the initial female to male ratio is 1:1, by the fourth year the ratio is 9:1 (Harris 1986, p. 264). Ozark Chub maximum age is slightly < four years, although it is uncommon for males to live beyond two years (Harris 1986, p. 264-266); the oldest female recorded was 42 months and the oldest male was 40 months (Harris 1986, p. 264).

Reproduction Rising water temperatures and increased water discharge in spring likely trigger spawning. Spawning was recorded at >15⁰ C (Harris 1986, p 252). Males develop tubercles in late winter and early spring and peak male development occurs in late April and early May. Most males have lost tubercles by July. Neither sex develops nuptial coloration (Harris 1986, p.250-255). Females are found in or near spawning condition by mid- to late April (Harris 1986, p. 250, 253). Ova number in females is correlated with standard length, and ranges from approximately 375 ova at 84 mm SL to 1200 ova at 110 mm SL (Harris 1986, p. 254-255). The smallest observed reproductive female was 63.35 mm SL, and the smallest reproductive male was 76.92 mm SL. Although the proportion of females in each age class varies by season, year one females make up

approximately 43% of the population annually (Harris 1986, p. 258). The proportion of females is progressively lower for older age classes (Y2: 40%, Y3: 11%, Y4: 7%). Adults breed over clean gravel substrate during a two- to four-week period in April and May (Harris 1986, p. 277). Ozark Chub is a non-guarding lithophilic spawner, meaning they deposit eggs in the substrate and have no parental care (Harris 1986, p. 277). No data exists on Ozark Chub eggs and larvae, but information is available for other, similar species. In other Erimystax, eggs are smooth, non-adhesive, and demersal (sinking); they likely settle into interstitial spaces in the gravel substrate (Shute and Rakes 2001, p. 16). Eggs hatch between 1 and 4 days after laying, and young fish stay near the substrate after hatching (Bottrell et al.1964, p. 398; Shute and Rakes 2001, p. 16). When threatened, young larvae swim downwards into the interstitial spaces of the substrate (Shute and Rakes 2001, p. 16). After a short time, they begin to swim in short bursts into the water column and likely drift with the current until deposited in slackwater areas, although larvae remain primarily benthic (Shute and Rakes 2001, p. 16). YOY rarely occupy adult habitats, but instead occur in side channel, shoreline, or backwater areas with low water velocity (Harris 1986, p. 278). No information is available on recruitment, although based on life history data collected by Harris (1986) and other literature, we’ve modeled the effects of varying recruitment levels on population persistence as part of future condition modeling in Section 4.1.

Feeding This species relies primarily on sight for feeding and generally stays just above the substrate, swimming continuously to maintain position (Pflieger 1997, p 121). Individuals graze the tops of rocks and in large interstitial spaces, although they will also feed on drifting particulate matter within ten centimeters of the substrate (Harris 1986, p. 246) Individuals feed almost exclusively on periphytic detrital aggregate (PDA) and algae. The proportion of each varies with the season; during colder months (September through April), their diet is up to 90% PDA, and plants and algae make up a larger portion of the diet in warmer water months (May - August) (Harris 1986, p. 248-249, 274). This is likely due to seasonal availability of plant matter (Harris 1986, p. 248-249, 274). Almost half of the macroinvertebrates found in Ozark Chub stomachs were Chironomidae, but insects make up a small, possibly opportunistic or accidental portion of their diet (Harris 1986, p. 274).

Life Stage Resource/Circumstance needs References Eggs (April- ● Water temperature >15° C and Harris 1986, p.253- May) increased discharge 256. ● Appropriate substrate (clean gravel, interstitial crevices) ● Sexually mature males and females; gravid females.

Larvae (late ● Slower side channel and backwater Harris 1986, p.278 spring-early areas downstream of spawning habitat summer) to accommodate larval drift. ● Appropriate clean substrate (low sedimentation; interstitial spaces) Juveniles (< ● Clear, flowing water Harris 1986, p.256 ~60mm SL) ● Appropriate substrate ● Adequate food availability ● Connectivity to adult habitat Adults (> ● Clear, flowing water Harris 1986 ~65mm SL) ● Appropriate substrate (clean gravel and cobble) ● Adequate food availability (detritus and algae) ● Appropriate flow and temperature range ● Presence of other breeding adults Table 1. Life stages and associated needs for Ozark Chub

CHAPTER 3 –Factors Influencing Viability and Current Condition Analysis 3.1 Factors Influencing Viability Sedimentation Excessive stream sedimentation (or siltation) results from soil erosion associated with upland activities (e.g., agriculture, forestry, mining, unpaved roads, road or pipeline construction, and general urbanization) as well as activities that can destabilize stream channels themselves (e.g., dredging or channelization, construction of dams, culverts, pipeline crossings, or other instream structures). Excessive sediment can cover the stream bottom and fill interstitial spaces between substrate particles (i.e., sand, gravel, and cobble) and in severe cases cause stream bottoms to become embedded (large substrate features including cobble, rocks, and boulders surrounded by or buried in sediment). Increased sediment can also increase suspended particle loads, increasing turbidity, decrease dissolved oxygen levels, and alter channel form (i.e. depth and width) and stability, which can lead to modification in habitat parameters such as flow, water temperature, light availability, and substrate. This can affect fish species directly by limiting sheltering or breeding habitat, causing shifts in plant community structure and photosynthesis, reducing respiratory efficiency through clogged gills, and disrupting metabolic processes and reducing growth rates (Bruton 1985, entire; Berkman and Rabeni 1987, 291–293; Chambers and Messinger 2001, p. 50–51; Sutherland et al. 2002, entire; McGinley et al. 2013, pp. 223–226). Pesticides and other contaminants can also bind sediment particles; increased sedimentation in streams can also lead to increase contaminant accumulation within aquatic habitats and organisms (Widenfalk 2002, p. 4-6). As a benthic feeder, Ozark Chub may be at high risk of ingestion and absorption of contaminants (Widenfalk 2002, p. 25).

Sedimentation is likely a major stressor on this species, with wide-ranging effects across all life stages (Table 1). Historically, prior to White River impoundments, the entirety of the main stem White River was gravel or sand-bottomed and clear, as were the other watersheds within the Ozark Chub range (Meek 1894, p. 73-74). Reviews of Ozark Chub life history stress the importance of clear water and clean substrate for survival and reproduction. Adults are visual feeders, and rely on clear water and the availability of plant material and interstitial spaces in the substrate for feeding (Harris 1986, p. 246; Pflieger 1997, p 121). Clean gravel needs to be present for spawning to occur in related species (Shute and Rakes, 2001, p. 16), and it likely provides refugia for eggs and larvae. Siltation also may increase egg mortality in a number of freshwater fish species through smothering (Dahlberg 1979, p.4).

The watersheds containing the Ozark Chub are primarily rural, and agriculture and logging comprise a large portion of land use in the area. Threats include habitat modification from certain types of logging practices that do not utilize best management practices to protect streamside management zones, agriculture (primarily livestock), gravel and heavy metal mining, and various other point and nonpoint pollution discharges (Miller and Wilkerson 2001). Activities can have varying effects on water quality depending on the type of practice, and conservation practices exist that can reduce negative effects of many of these sources.

Some logging practices can degrade water quality, primarily through increased sedimentation and turbidity, changes in nutrient cycling, and increased water temperature during and after a logging operation (Swank et al. 2001, p. 167-176). Disruption of the ground surface by tree

removal, skid trails or roads allows for soil erosion at greater than normal rates, which can be a primary producer of off-site sedimentation (Corbett et al. 1997, entire; Swank et al. 2001, p. 167- 176). Most states with a commercial timber industry, including Arkansas and Missouri, have created Best Management Practices (BMPs) specific to timber harvest activities to protect water quality (Koirala 2009, p. 3). These BMPs often include a streamside management zone (SMZ), an undisturbed buffer around all waterways that acts as a filter to sediment and slows surface runoff (Bunger 2005, entire). Studies across the eastern United States have demonstrated that timber harvest causes an increase in sediment, discharge, and nutrients, but the implementation of BMPs with SMZs seems to be effective in reducing the effect of the harvest on water quality (reviewed in Boggs et al. 2016). Although these BMPs are voluntary in both states, certification groups such as the American Tree Farm System and the Sustainable Forestry Initiative often require forest owners to apply state BMPs. According to the Arkansas Forestry Commission, a survey of randomly selected silviculture sites across the state, including 37 sites in the Ozark Mountain region, showed that BMP use overall is high in the Ozark region (Table 2). However, SMZ is the lowest scoring category, followed by roads, which together have the greatest potential to negatively impact water (Hameister and Fox 2017, p. 9-11). Public land and industry/corporate land all have high rates of BMP implementation, as industrial timber lands are often subject to certification requirements. However, small, private forest tracts, which make up approximately 46% of forestland in the state, had a significantly lower rate of BMP implementation (86% overall, 76.59% SMZ; Hameister and Fox 2017, p. 19-22). In Missouri, over 84% of the forest land statewide is owned by private landowners (MDC 2014, p. 1). Missouri does have voluntary programs such as the Missouri Managed Woods program to encourage management and protection of privately-owned forested lands, but does not track BMP use on private, non-certified lands.

Table 2. BMP implementation rate within the Ozark Region in Arkansas, taken from Hameister and Fox 2017, p. 17.

The Ozark Mountain ecoregion contains some of the primary cattle-producing areas for Missouri, and predictions show increases in density of cattle (Miller and Wilkerson 2001, p. 60). Certain practices, primarily unrestricted cattle access to streams and riparian areas, are sources of nonpoint source pollution due to soil compaction, reduced riparian vegetation, increased in- stream disturbance through increased suspended sediment and associated contaminants, and actively contribute to bank erosion (Owens et al. 1996, entire; Vidon et al. 2008, entire). BMPs to limit cattle access to streams and riparian areas (including exclusion fencing, off–stream water sources, and seasonal or rotational grazing) can greatly reduce sedimentation, soil loss, and sediment-bound pollutants such as nitrogen and phosphorus, and allow for revegetation of riparian buffers (Miner et al. 1992, entire; Owens et al. 1996, entire; Sheffield et al. 1997, entire; Clary 1999, p. 225). However, large scale implementation of fencing can be costly and difficult

(Wilson and Clark 2002, entire), and many landowners are unwilling to implement BMPs and maintain associated structures without funding assistance. Dirt and gravel roads provide transportation for rural communities, economic sectors such as timber and agriculture, and recreation. However, eroding unpaved roads can channel rainfall runoff, which increases erosion and sediment delivery to waterways. Unpaved road BMPs can reduce road erosion and sedimentation transport by as much as 95% by stabilizing erodible areas and providing better road drainage. In Arkansas, over 85% of county roads are unpaved.

Water Temperature and Flow For the Ozark Chub, certain water temperature and flow levels are associated with spawning period; spawning in Hybopsis and Erimystax is generally associated with temperatures between 12 and 17°C and flows beginning to reduce from peak spring flood (Stallsmith et al 2015, p.70- 71;Tarver and Stallsmith 2019). Water temperature is an important factor for aquatic species affecting all aspects of their life history. At a large scale, stream temperature is an integrator of complex climatic, hydrologic, and land-use factors in watersheds, and stream temperatures across the United States have been rising in the past 50 years (Kaushal et al. 2010, entire). Fish are generally adapted to the fluctuations in their normal habitat. Based on lab and field studies, heat death is relatively uncommon compared to death through exposure to cold or other biotic or abiotic factors, although this could be due to underreporting (reviewed in Beitinger et al. 2000, p. 238-241). In freshwater fish in general, the laboratory-derived critical thermal maximum (CTM) that leads to heat death is generally higher than the warmest weekly mean water temperatures within their range, and freshwater fish can acclimatize to warmer conditions to a certain extent (Beitinger et al. 2000, p. 238-241). Very low rates of temperature increase (1°C/day or slower) likely create a conflict between acclimatization and overall accumulation of heat stress; in some cases, this can lower the estimated CTM as the heat stress outpaces the ability to acclimatize (Fields et al. 1987, entire; Beitinger et al. 2000, p. 246-247). Within Cyprinidae, CTM ranges from 28° to above 40⁰ C. There is no direct estimate of CTM for Ozark Chub; however, the (Hybopsis amblops), also found in the , has a measured CTM of 31°C. Fish generally associated with Ozark Chub (Harris 1986, p. 245; Matthews et al. 1992, p. 305) have CTM from approximately 28 to 34°C (Eaton and Scheller 1996, p. 1111; Beitinger et al. 2000, p. 248-263). As these fish are found within the same habitat as Ozark chub, and temperatures over 34⁰ C are extremely uncommon in the White and St. Francis River basins, it is likely that the thermal maxima for Ozark Chub is within this range. Currently, deforested areas (i.e., urban areas, agricultural fields and pastures, timber harvests, non-buffered riparian areas) likely contribute directly and indirectly to warming water temperatures. Water temperatures do not exceed 34°C every year, but most large, occupied watersheds have had water temperatures > 34°C recorded at least once since 2000, usually in July. Although temperatures above 34°C have been recorded prior to 2000, primarily in the Kings and Buffalo River in the late 1990s, the majority of high temperature records were collected in the last 20 years.

Predicting the long-term effect of warming water temperatures is made more complicated by the fact that stream temperatures are not homogenous; localized stream temperature is a range affected by a variety of habitat parameters and fish can move to utilize this variability (Westhoff et al. 2014, entire). However, air temperature has been shown to be a very strong predictor of water temperature in streams and rivers (Webb and Nobilis 2007, entire; Westhoff and Paukert 2014, entire), and there is a projected warming trend of approximately 0.2°C per decade (IPCC 2007). Based on temperature projections, water temperatures within the White River watershed are likely to increase by approximately 1.2°C by 2049. Combined with the slight predicted increase in precipitation during the same time period, the spawning period of Ozark Chub may be shortened by both extended spring high flow and more rapid spring warming (Figure 6; NCCV 2017).

Figure 6. Predicted relative change in max temperature (°C, top) and precipitation (mm/month, bottom) for the month of April in the White River watershed between 2010 and 2100.

Impoundments Impoundments cause short- and long-term changes in fish community structure, decreasing species richness and composition through the loss of natural flow regimes, habitat fragmentation, alteration, and loss, introduction of non-native species and changes in water quality (reviewed in Vogt 2013, p. 3-4). Studies on the early effects of dam construction and water impoundment on the White and Black Rivers suggests that Ozark Chub and other non-migratory fish upstream of a dam are only affected in areas specifically modified by increased water levels. There is little change to habitat upstream of the impoundment, but most riffle species rapidly decline in abundance within the reservoir (Martin and Campbell 1953, p. 56; Keith 1964, p. 433). However, isolation from other populations due to dam construction effects demographic, environmental, and genetic stochasticity of populations. Fish in isolated populations more likely to be affected by stochastic

events and, therefore, and demonstrate decreased genetic diversity and low growth rates (reviewed in Lande 1998). Population isolation and habitat fragmentation has a negative effect on population persistence, and local extirpations of freshwater fish and lowered species richness occur in areas isolated by dams (Morita and Yamamoto 2002, p. 1319). There is a strong correlation between watershed area upstream of a dam, isolation period, and presence of historically present fish species (Morita and Yamamoto 2002, entire). The Ozark Chub, although not migratory, still requires some up- and downstream habitat connectivity to allow for larval drift and movement between juvenile and adult habitat. Dam releases create areas of cold water and low dissolved oxygen downstream of dams called tailwaters (Bayless and Vitello 2002). It is unlikely that Ozark Chub can survive in the lowered temperature, lower dissolved oxygen, and with the presence of trout in these areas. In total, over 241 km of the main stem White River is classified as tailwater, the majority of which is downstream of Bull Shoals dam (Vogt 2013, p. 7, 14). However, Ozark Chub were collected in the cool-water transitional zone between the tailwater and the natural warm-water habitat (Vogt 2013, p. 7), and small numbers of Streamline Chub were found in two warm-water tailwaters in Kentucky (Robinson 1980, p. 10, 25, 32), suggesting that some parameter of the cold-water tailwaters is precluding Ozark Chub from these habitats within its range.

Water Chemistry There is little information regarding the Ozark Chub’s tolerance of specific water quality parameters. We can infer from available occurrence data and scientific literature that the species is probably adapted to waters that are well oxygenated, have circumneutral or slightly high pH, low sedimentation levels and total dissolved solids, and are free of contaminants at levels likely to cause toxicity to native aquatic fauna (Matthews et al. 1992, p. 301). Within the Ozark Chub range, the common water quality constituents that can cause toxicity issues include heavy metals, ammonia/nitrates, and pesticides (discussed in Appendix A). In general, ammonia nitrate levels are rarely measured at levels toxic to co-occurring species, but multiple stream segments within the Ozark Chub range have been 303(d) listed by the EPA for heavy metals (Appendix A). One study on Erimystax species in the mid-Atlantic Highlands found that the extirpation concentration (the conductivity at which 95% of the individuals of a species will not be able to survive) of grouped Erimystax species was 744 µS/cm. This ranked 17th lowest out of 101 species, although this was a genus-level measurement and so may be skewed towards the most tolerant Erimystax species. All measured fish species were above the benchmark for macroinvertebrates (300 µS/cm), and at about 500 µS/cm, many species become less likely to occur because this value exceeds their optimal specific conductivity range (Griffith et al. 2018, p. 877, 880). In a study of eighteen Ozark Mountain large stream reaches, median specific conductivity for a two year period ranged between 63 and 331 µS/cm (Peterson 1998, p. 8). The Black River, St. Francis River, James River have had collections with a specific conductivity of higher than 744 µS/cm, but these are all at sites downstream of lakes, mines, or wastewater treatment plants. There were no collections near any recent Ozark Chub sites with specific conductivity above approximately 550 µS/cm. This suggests that conductivity is not currently a major stressor on Ozark Chub unless the species extirpation concentration is much lower than

that of other Erimystax. Also included in the water chemistry stressor category are spills or releases of chemicals, petroleum, or other substances toxic to aquatic organisms. The risks from these types of events are difficult to predict, but many waterways are crossed by transportation infrastructure (roads and railways) and many commercial and industrial facilities are adjacent to streams and rivers, increasing the risk that a spill or release could affect the aquatic habitat. No gas or oil spills or leaks have been recorded within the range of the Ozark Chub.

Habitat Fragmentation Little is known regarding the minimal habitat patch size or degree of habitat connectivity necessary to support persistent Ozark Chub populations. However, larger and more connected populations contribute to the long-term viability of a species and smaller isolated populations are more at risk of decline or extirpation as a result of genetic drift, demographic or environmental stochasticity, and catastrophic events (Gilpin and Soulé 1986, pp. 32–34; Morita and Yamamoto 2002, p. 1319). Although we discussed large-scale habitat fragmentation (see Impoundments), smaller-scale habitat fragmentation within watersheds can also affect population persistence. Ozark Chub require a variety of habitats to complete their life cycle, and the construction of barriers and flow modification caused by road crossings, water withdrawal, urbanization, climate change, and dams may limit movement by various life stages. Cyprinids have been found travelling across riffles and between pools more often than game fish, but shallow riffles can still act as a barrier for pool-to-pool movement (Matthews et al. 1994, p. 393-395; Schaefer 2001, p. 383-386). Few studies on the home range size of small, non-game fishes exist, and none on Erimystax, but previous studies suggest home range size is approximately 50 m, although movements of 100 – 500 m have been observed in water-column cyprinids (Hill and Grossman 1987, p. 377; Goforth and Foltz 1998, p. 52). Populations may have a small “mobile” proportion that strays further than the small home range, but almost all data have focused on game fish (reviewed in Gerking 1959, 222-227). Studies on non-game fish are seldom set up to detect complex or long-range movements, but Gorforth and Foltz (1998, p. 54) found no movement of marked Notropis lutpinnis outside of their 600 – 700 m study areas. Home range can vary among systems based on stream characteristics such as food and habitat availability and water quality parameters, but it is likely that the majority of Ozark Chub have similar home ranges. If so, this means that Ozark Chub habitat connectivity within watersheds is of vital importance, as the majority of watershed-level populations are isolated.

3.2 Model Because no consistent, rangewide assessment of the Ozark Chub is available, we developed a semi-quantitative model that produced a “condition score” for each Ozark Chub watershed (Table 5). The model relies on two categories of interrelated metrics, including habitat parameters (i.e., water quality and forest cover) and estimates of the Ozark Chub demographic status within each unit (Table 4). The individual metrics were then combined to produce a unitless condition score for each watershed. Because empirical data relating some of these metrics directly to Ozark Chub life history needs are sparse, we consulted species experts who

generally agreed that, for the purpose of this SSA report, the selected metrics were appropriate for assessing the viability of Ozark Chub populations across the species’ range. To aid in the comparison of watersheds (with each other and under various future scenarios) and assess the species’ viability under the 3Rs, we categorized the final condition scores as “high” (population likely secure), “medium” (population likely somewhat secure), or “low” (population likely insecure). We based these categories primarily on our understanding of Ozark Chub habitat needs, known stressors, and the principles of conservation biology (Table 4). We acknowledge that there is uncertainty associated with this model and some of the supporting data but consider the methodology suitable for assessing Ozark Chub status across its range.

Analytical Units There is little information available regarding the spatial structure of Ozark Chub populations. Therefore, for the purpose of analyzing the status of Ozark chub in this report, we defined populations based primarily on river and watershed connectivity. These analytical units generally conform to a Hydrologic Unit Code (HUC), which are geographical units used to define watersheds at various scales. For this analysis, Ozark Chub populations are defined by a larger watershed, generally a HUC 8 or HUC 10, which has internal connectivity but limited or no connectivity with other river watersheds. Ozark Chub occurrences, current or historical, within a river system are considered a population. Populations are categorized into five Ecological Management Units (EMUs) to assist with data analysis and are discussed in Appendix A.

Watershed Parameters Urban and Riparian Landcover Maintaining forest on a large proportion of a watershed can help protect water quality and maintain high quality aquatic habitat. Watersheds with over 80% forested landcover generally have streams with a high IBI (index of biological integrity) and maintain aquatic habitat (Wang et al. 1997, entire). Stream stability can be maintained at lower levels of forest cover (60-80%) when combined with low levels of impervious surfaces (<10%) and/or high levels of riparian forest cover (Brabec and Richards 2002, p. 505-507). Impervious surfaces such as parking lots, roads, and roofs prevent water from soaking into the ground. Instead, water runs off of (during wet periods) and evaporates from (dry periods) impervious surfaces and can lead to dramatic fluctuations (more frequent and higher magnitude flooding) in water flow (Ferguson and Suckling 1990, entire; May et al. 1998, p. 2-6; Wang et al. 2001, p. 255). Low impervious surface levels allow water to soak into the ground and be released over a longer period of time, leading to more stable base flows. A review of the effect of impervious surfaces on stream quality found that, in general, the impact threshold for biotic health ranges from 3.6 – 15% watershed impervious surface (Brabec and Richards 2002, p. 505- 507). Sensitive fish species can show population declines when as little as 2% of the watershed is impervious surfaces, and at ≥ 10% fish and macroinvertebrate health drastically declines. Connected imperviousness between 8 and 12% is a threshold zone in which minor increases in urbanization are associated with sharp declines in biotic measurements of stream quality (Wang et al 2001 p. 264). Abiotic watershed health measurements such as water quality and habitat

characteristics are inconsistently related to impervious surface percentage, with an impact threshold of up to 50%. However, biotic factors are likely a better representation of overall watershed health, as biota reflect the long-term impact of a combination of abiotic factors, rather than abiotic changes that may be short-lived (Brabec and Richards 2002, p. 507). Although watershed-wide metrics are associated with water quality, the influence of riparian forest cover has a disproportionate effect on habitat and biological factors (Richards et al. 1996, entire; Wang et al. 2001, p. 263). In predominately non-forested watersheds, riparian buffer zones reduce sedimentation through streambank stabilization and erosion control, filter nutrients, shade and cool streams, and provide organic material for in-stream habitat and food resources (Wang et al. 2001, p. 263). Percent impervious or forested area within the riparian buffer zone has a higher influence on stream quality than the same amount spread across a watershed (Wang et al. 2001, p. 264; Rios and Bailey 2006, 158-159). Agricultural lands contribute more nutrients to watersheds than any other land use, but nutrient levels seem less critical to IBI scores than runoff volume and high temperatures found in urban environments (Brabec and Richards 2002, p. 508). More than 50% agricultural land was required to reduce IBI scores, and implementation of riparian buffer zones and other best management practices (BMPs) allowed for relatively healthy stream communities even in watersheds that are majority agricultural (Wang et al. 1997, entire). The presence of forest buffers on unbuffered row-crop fields and pasture could reduce total stream bank soil loss by 72% and sediment from overland flow by up to 90% (Lee et al. 2003, p. 6-7; Zaimes et al. 2004, p. 26-27) We used National Land Cover Database (2011) 30-m land cover layers in ArcGIS to analyze and calculate watershed and riparian cover parameters. We developed a 30-m buffer around all perennial streams within each watershed to determine riparian cover, based on average effective buffer widths (Wenger and Fowler 2000, p. 9-11; Table 3). Watersheds were categorized as High, Medium, Low, or Very Low quality based on percentage of the total watershed area and total riparian buffer, respectively, that was developed and forested (See Table 5).

Table 3. Range of buffer widths for specific riparian functions (from USFWS 2006, p.22).

Land Ownership Although land conservation and restoration efforts on private lands can have a large influence on water quality, public lands are protected and managed in perpetuity to maintain environmental quality and biological diversity. Federal lands specifically are required to analyze the impact of projects on trust species and the environment, and must put in place BMPs to maintain water and habitat quality. Additionally, the status of private lands are uncertain when looking into the future; although public lands will be maintained, private lands may shift in land use type and quality over time. For this analysis, we quantified the percent of each Ozark Chub watershed that is held by federal or state agencies. The majority of public land within the range of the Ozark Chub is held by the U.S. Forest Service as part of the Ozark-St. Francis and Mark Twain National Forests (Appendix A).

Water Quality To analyze water quality, we looked at water quality data collected by state and federal agencies as well as the states’ 303(d) Impaired Waterbodies lists (see Appendix A). Although we have no water quality parameters for Ozark Chub specifically, we do have quantitative data for specific conductivity and temperature tolerances for Erimystax and other co-occurring species. The three lower ranking categories are split between 303(d) Category 4 and Category 5 watersheds. Category 4 includes watersheds in which at least one designated use is impaired or threatened, but pollution control requirements are in place that are expected to result in the attainment of water quality standards in a reasonable period of time. Category 5 include waterbodies that currently have impaired biotic communities or are impaired by one or more pollutant. Category 5 waterways are generally listed by stream segment, but a few rivers within the Ozark Chub range are listed in their entirety (Appendix A).

Population Parameters Isolation As discussed previously, the risk of extirpation for a species is strongly associated with the size of the isolated watershed. To analyze watersheds for their isolation ranking, we categorized each watershed into “small”, “medium”, or “large” size category (Table A1). We then categorized the level of isolation based on connections to other watersheds.

Species Presence As the majority of records for this species are presence/absence data and not full population counts, we cannot estimate total abundance or population size for our watersheds. Instead, as a proxy for species abundance, we used the presence of recent collection data in each watershed (Table 5; Figure 5). There have been surveys in most of the watersheds within the Ozark Chub range; the Missouri Resource Assessment and Monitoring (RAM) program and Arkansas Department of Environmental Quality (ADEQ) have ongoing fish sampling in the region and the Ozarks is an area of interest for many aquatic researchers. Watersheds in which we have verified presence after 2000 are ranked high, while watersheds with the most recent collection prior to 2000 are ranked lower.

3.3 Current Condition—3Rs The results of the population condition model provide the basis for our analyses of the species’ current status using the 3Rs (Table 6). The population condition scores allow us to directly assess and compare the resiliency of each Ozark Chub population, which then supports our analysis of the species’ redundancy (within and among the various populations) and representation (across its environmental settings). We emphasize that this portion of the assessment is a “snapshot in time” of the Ozark Chub’s current condition and does not consider future trends. Chapter 4 assesses the species’ potential condition under several future scenarios.

Resiliency No long-term studies have focused on the resiliency of Ozark Chub. It is unlikely that they are highly resilient to extreme habitat modification; although 15 individuals were collected in the shallows of Bull Shoals Reservoir, it is unlikely that they can survive and reproduce in that habitat (Buchanan 2005, p. 34, 37). The collection of this species in the lower main stem White River suggests that they can persist in somewhat impacted cool-water habitats, but not within cold tailwaters (Vogt 2013). No other Ozark Chub collections exist in any lakes or dam tailwaters, although we have records of the species in these areas prior to dam construction (Figure 5; Keith 1964). The continued presence of this species in some watersheds for decades suggests that they are likely resilient to natural stochastic events and low-level anthropogenic impacts. For example, the earliest recorded collection of Ozark Chub within the Current River was in 1947, and individuals were collected just downstream and upstream of that location in 2016. Even in small watersheds such as Bee Creek and Bear Creek (Upper White River EMU), collections have recently been made of this species, which has likely persisted in isolation since the construction of Table Rock Lake. Because of these factors, we rate the current overall resiliency of the Ozark Chub as moderately high.

Representation There are no studies that identify genetic populations in this species. However, the difference in morphology between Little Red River and White River basin populations suggests that the Little Red River should be counted as a distinct representative group (Harris 1986, p. 26-27). It is likely that the Ozark Chub occurred throughout the Little Red River, but after the construction of Greers Ferry Lake in 1962 the only suitable habitat remaining is the isolated upper reaches of the four headwater streams (Mitchell et al. 2002, p. 129). It is uncertain whether the Ozark Chub is extant within the Little Red River basin. However, sampling targeting the species has not occurred in the basin. In the absence of recent species-specific sampling, we interpret Ozark Chub occurrence in collections made between the 1880s and 1990s (>30 years after dam construction) as an indication that the species may still be present. Additionally, there has been a focus on water quality in the basin due to the endemic and federally listed Yellowcheek Darter (see discussion in Appendix A).

Similarly, the St. Francis River is not within the White River drainage. Although the St. Francis River has impoundments and evidence of water quality degradation in some areas, there have been many recent collections of this species within the watershed. It is likely that mixing within the White River basin occurred historically, but currently dam construction has isolated a majority of occupied watersheds. We have split the White River basin into three representative units based on current and likely historical connectivity (see Figure 5). Within the White River Basin, there are current records (post-2000) in all but one historically occupied watersheds. Because of these factors, we rate the current overall representation of the Ozark Chub as moderate. Redundancy Historically, Ozark Chub occurred throughout the White River basin and the upper St. Francis River. There were no large barriers within the White River basin or its tributaries, and it is likely that there was some level of gene flow between watersheds within each occupied river basin. Of our three representative groups within the White River basin, we have current records in all historically occupied watersheds except for the North White River upstream of Beaver Lake and the main stem White River upstream of Bull Shoals Lake. There are many recent collections of this species in the upper St Francis River. As discussed above (see Representation), there have been no collections of Ozark Chub in the Little Red River since 1999; we have very few records of this species in the Little Red River basin overall and it is unknown if Ozark Chub still persist in the watershed. However, they did occur in at least two of the four headwater streams. The Ozark chub occurs in at least 22 of 23 management units that were occupied historically (Table 6), although there has been some curtailment of the species historical range. The condition of the populations within these units is predominantly medium-high and predominantly medium for the units themselves. Therefore, we rate the overall current redundancy of the Ozark chub as moderate

Needs Associated Threat Effect Sedimentation; water Reduced survival at Clear, Flowing Water use; climate change all life stages

Reduced survival at Clean rocky/gravel Sedimentation all life stages; reduced substrate reproduction

Appropriate temperature range: Reduced survival at all Loss of riparian areas; <35C for survival life stages; reduced climate change; water use >15C for reproduction reproduction (April) Appropriate water parameters Reduced survival at all Water quality DO < 5ppm, low TDS, life stages etc.

Maximum discharge in Water use; climate Reduced spring for reproduction change reproduction

Appropriate/adequate food availability: Water quality; riparian Reduced survival for Detritus (winter) and areas juveniles and adults plant matter (spring/summer) Isolate populations, Connectivity: reduced recruitment, Impoundments, dams, and reduced Accommodate larval road crossings; water use reestablishment of drift extirpated Genetic drift populations Table 4. Life history needs of Ozark Chub (Erimystax harryi), threats to those needs, and the effect the threat has on the life history need.

Watershed Parameters Population Parameters

Urban Land Quality Forest Cover- Cover- Ownershi Water Quality Isolation Abundance Ranking Riparian Watershed p Connectivity with multiple watersheds or Recent surveys Primarily >50% Meets designated many High 0-2% (>2000) show forested (>80%) public land use standards occurrences presence throughout large isolated watershed. Connectivity with another 25-50% 303(d) listed watershed or Last collection Medium 3-8% 60-80% forested public land Category 4 large isolated 1981-1999 watershed with few occurances 303(d) listed Medium-size 50-60% forested 15-25% watershed Last collection Low 9-14% isolated riparian zone public land segment Category 1965-1981 watershed 5 Less than 50% Last collection forested/habitat >15% Entire main stem Small isolated pre-1965 or Very Low 15-100% is no longer public land 303(d) listing watershed negative survey present data Table 5. Ranking criteria for estimating current condition of each watershed within the range of Ozark Chub. Definitions for each of the six ranking criteria are given as high, medium, low, and very low quality.

Table 6. Current condition of each occupied watershed based on four physical parameters and two population parameters. See Table 5 for definitions.

Population Total Watershed Score Physical Parameters Parameters EUs MUs Urban Cover Riparian Forest Land Ownership Water Quality Isolation Presence Upper White River Flat Medium Low Medium High Low High M Kings Medium Medium Very Low Medium High High M Swan Medium High High High Low High H Beaver Crk. Medium Medium Medium High Medium High M Long Creek Medium Low Very Low High Low High L N. White/Beave r Res Medium Medium Low Low Low Very Low L Upper James Low Low Very Low Low Medium High L Lower James Very Low Very Low Very Low Low Medium High L Bull/Bear Medium High Medium Low Low High M Bee Medium High Low High Very Low High M Black River Strawberry Medium Medium Very Low Medium Medium High M Spring Medium Medium Very Low Medium High High M Eleven Point Medium Medium Medium Low Medium High M Current Medium Medium High Very Low High High M Black Medium Medium High Very Low Medium High M

Population Total Watershed Score Physical Parameters Parameters EUs MUs Urban Cover Riparian Forest Land Ownership Water Quality Isolation Presence Middle White River Sylamore Medium High Medium High Medium High H North Fork Medium Medium Medium High Medium High M Piney Medium Medium Very Low High Medium High M Buffalo Medium High High Medium High High H Lower White Medium Medium Very Low Low Medium High M Crooked Medium Low Very Low Low Medium High L St. Francis St Francis Medium Medium Medium Very Low Medium High M Little Red River Archey Fork Little Red Medium High Low High Very Low Low M Table 6 cont’d. Current condition of each occupied watershed based on four physical parameters and two population parameters. See Table 5 for definition

CHAPTER 4 – Future Condition Analysis/Species Viability To predict species viability into the future, we created a semi-quantitative model with the methodology used to determine current condition. To predict species viability into the future, we chose parameters that predict a calculable effect on Ozark Chub populations (Table 7). We modeled 4 scenarios. For each scenario, we attempt to make plausible predictions for population conditions 10, 25, and 50 years into the future (Tables 8, 9). Based on the best available information, we expect that increased sedimentation and other habitat and water quality issues will be the most likely stressors affecting Ozark Chub throughout its range. As habitat change and the resultant effects to Ozark Chub populations occur gradually over time, we estimated effects as a rate of change per decade.

Table 7. Future condition scenarios for four factors likely to affect Ozark Chub in the future.

Riparian Water Scenario Urban Cover Forest Cover Quality Abundance 1% 1% increase in Water Increases with Scenario 1 increase/decad riparian forest quality increased habitat e cover/decade improves parameters 1% Scenario 2 increase/decad No change No change No change e 1% decrease Slight 1% in riparian decrease Slight decrease with Scenario 3 increase/decad forest in water habitat parameters e cover/decade quality

Decreases with 2% decrease 2% Reduced decreased habitat in riparian Scenario 4 increase/decad water parameters; smallest forest e quality populations/watershed cover/decade s extirpated

4.1 Parameters Urban Cover Based on U.S. census data, Northwest Arkansas has recently been growing at an average rate of approximately 2.5%/year. In the recent past, the annual growth rate has reached almost 6% in city areas, and average population density has been decreasing, implying a higher level of urbanization associated with population increase. However, a large majority of the area under review is rural, and rural areas have been showing a low average annual growth rate in recent years; land cover analysis shows an average growth of development of 0.1% from 2001 – 2011. Growth is uneven throughout the species range. For example, the James River watershed in

Missouri had development increase by 1%. Similarly, the Buffalo River watershed saw 0.7% increase in urban landcover, while the Upper White River, Beaver Creek, Lower White, and St. Francis watersheds experienced 0.1 – 0.2% urban growth. For this model, we chose to use a 1% and 2% increase in urbanized landcover per decade to account for the shift from forest to urban land types at a watershed level. One weakness of land cover analyses datasets, because of scale, is the difficulty estimating the proliferation of suburban and rural development. The construction of one home or farm would likely not show up on land cover databases, and the proliferation of these throughout the watershed could have an effect on waterways. For example, within the overall Ozark Chub range, lands in farmsteads, buildings, ponds, and roads increased by over 20% between 2007 and 2012, but the landcover analysis would likely not include these types of development. To cover uncertainty and to make sure we are including the effects of urbanization, we chose urban growth rates that are slightly higher than expected for some watersheds. Riparian Forest Cover As discussed previously, riparian forest cover is an important factor in stream health. The importance of riparian buffers is well known, and the restoration of riparian vegetation is one of the most common management practices for restoring stream banks and improving water quality. From 2001 – 2011, riparian forest cover decreased an average of 0.8% throughout Ozark Chub’s range, primarily due to conversion of forest to pasture along the riparian corridor in all EMUs. We rounded this up to 1% per decade for Scenarios 1-3 to maintain similar levels of growth, and increased it to 2% for Scenario 4. As discussed in Section 3.1, cattle production is expected to continue increasing in the Ozark Mountain region, along with other agricultural operations. Additionally, forest to pasture conversion can combine with other factors such as row cropping, logging without BMPS, and urbanization to create a higher rate of overall riparian forest loss. Water Quality To analyze overall water quality, we analyzed how predicted changes in water flow, temperature, and water chemistry may combine to affect the Ozark Chub in each watershed. We chose these parameters because we have data to suggest they affect Ozark Chub and to predict how they will change in the future (see Section 3.1). For example, an increase or maintenance of current flow levels is likely required for the species to maintain healthy populations; modifications in flow level or pattern, such as increased flashiness or increased drought, are likely to have negative effects on the population mimicking the severity of the flow modification. Temperature, as discussed in Section 3.1, can have a direct and indirect effect on Ozark Chub, and increased temperatures, even if they do not reach CTM, will likely have negative effects on the species. Abundance To estimate abundance, we utilized a survival model (described below) and the number of recent collections within each watershed to determine probable abundance and probability of persistence into the future for each population under the four scenarios. An example of the probability tables output by the survival model is shown in Figure 7. Daily survival rates of Great Plains cyprinids through the first year of life range from 0.00296 to 0.00479, and are likely determined by events in the first few months of life (Wilde and Durham 2008, p. 831). Proportional changes in age-0 survival and age-1 fecundity in several cyprinid

species can account for up to 90% of the growth rate response to variation in age-specific survival and fecundity rates (Wilde and Durham 2008, p. 832); understanding the threats to fish species during the first year of life can be critical to fish conservation. In freshwater fish, broadcast spawning with no egg protection and a quick development period are associated with high egg and larval mortality, and stressors during this critical period may have greater consequences than during other times (Dahlberg 1979, p. 8-9). For Ozark Chub, we utilized life history data from Harris (1986) to predict population trends over time. Although populations between 100 and 241 individuals have been recorded in the Kings and Buffalo Rivers, the majority of records logged < 50 individuals/collection and in many cases < 10 individuals/collection. We determined a base average egg/larvae survival rate of 0.0035, which is within the range found by previous cyprinid studies (Wilde and Durham 2008, p. 831-832) and is based on the assumption that there is an average of 50 breeding individuals within a healthy population. We used RISKAmp in Microsoft Excel to run Monte Carlo simulations on a variety of population sizes to determine probable abundance and probability of persistence into the future.

Figure 7. An example of the probability of persistence for a single occurrence at 10 and 25 years, with a 0.0035 egg survival rate and 30 breeding females.

.

Table 8. Predicted condition of Ozark Chub watershed-level populations at 10, 25, and 50 years for the four scenarios described in Table 7. Green is high condition, yellow is medium, red is low, and grey is likely extirpated.

Table 9. Predicted condition of Ozark Chub watershed-level populations at 10, 25, and 50 years for the four scenarios described in Table 6. Green is high condition, yellow is medium, red is low, and grey is likely extirpated.

4.2 Scenarios Scenario 1 (positive habitat change) Under the positive habitat change scenario after 50 years, the model predicts three populations will increase in condition from current conditions (Long Creek, Upper James and Bee Creek), five will decrease in condition (Swan Creek, Sylamore Creek, Piney Creek, Buffalo Creek, and St. Francis River), and the others will remain relatively unchanged. Under this scenario, we have growth in both watershed level urban development and forested riparian buffers, water quality, and abundance. All MUs but four (North White, Lower James, Bee Creek, and St. Francis River) shift to the medium condition category after 50 years, including the three currently in high condition (Swan, Buffalo, and Sylamore). In some watersheds (Crooked Creek, North White, and Upper James), there is a fluctuation between medium and low condition over the 50 year period. These are all watersheds that currently have higher rates of urbanization, and increases in forested riparian buffers and water quality cannot entirely negate the negative effects of urbanization. The model predicts no risk of population extirpation under Scenario 1. Scenario 2 (increased urbanization) Under this scenario, urbanization increases at 1% per decade and percent of forested riparian buffer, water quality, and Ozark Chub abundance remain stable. We see little change from current condition, with only four MUs declining in condition score in 50 years. Although two currently medium condition populations are predicted to shift to low condition (Eleven Point and Lower White), the model predicts no risk of population extirpation under this Scenario 2. Scenario 3 (increased urbanization, other parameters slightly decreased) Under this scenario, urbanization increases 1% per decade, percent forested riparian buffer decreases 1% per decade, and water quality and Ozark Chub abundance decrease slightly. Under this scenario, 57% (13 of 23) of the populations are at a low condition in 50 years, 61% (8) of which were previously at high or medium condition. Because they are now at a low condition score, the smaller populations (Bee Creek, Long Creek, and Flat Creek) are at an increased risk of extirpation due to environmental or demographic stochasticity (see Section 3.1, Isolation). Additionally, our future predictions cannot predict catastrophic events; in a small watershed, a single catastrophic event such as a drought or spill could wipe out even a healthy isolated population, and at a low condition score, a single year of failed or reduced recruitment could be enough to extirpate the population. The loss of any one of these populations would slightly decrease the representation and redundancy of the species as a whole as they are all part of the Upper White EMU; the loss of all three populations would represent a loss of a third of the currently occupied MUs within the Upper White EMU. Under this scenario, although the Ozark Chub would likely maintain extant populations over the next 50 years, the species would lose resiliency in over half of the currently occupied watersheds. Scenario 4 (increased urbanization, other parameters decreased) Under this scenario, urbanization increases at 2% per decade, percent forested riparian buffer decreases by 2% per decade, and water quality and Ozark Chub abundance decrease. Under this scenario, 57% (13 of 23) of the populations are at a low condition in 10 years, and all our currently occupied populations are at low condition or extirpated in 50 years. The model predicts

extirpation of 43% (10 of 23) of the populations in 50 years, including 70% of the Upper White EMU and 30% of the Middle White EMU. The Ozark Chub loses resiliency in almost all of its populations within 25 years. Under this scenario, resiliency, redundancy, and representation would all be significantly decreased. 4.3 Summary Based on the Ozark Chub life history and habitat needs, and in consultation with species experts, we identified the potential stressors and their sources that are likely to affect the species’ current condition and viability. We evaluated how these stressors may currently be affecting the species and whether, and to what extent, they would affect the species in the future. Sedimentation, water quality (including chemistry, temperature, and flow), isolation, and habitat destruction and fragmentation likely influenced the current condition of the species and will likely continue to affect the Ozark chub in the future. We considered what the Ozark Chub needs to maintain viability by characterizing the status of the species in terms of species resiliency, redundancy, and representation. For the purpose of this assessment, we define viability as the ability of the species to sustain populations within their natural stream ecosystems for a biologically meaningful timeframe. At some point between 25 and 50 years (the maximum time frame we assessed), uncertainty likely becomes too large to make predictions concerning population viability. The best available data indicate that the majority of Ozark Chub populations are still extant, with recent collections of the species in 91% (21 of 23) of historically occupied watersheds. Many of these populations still persist even though dams and impoundments have partially or entirely isolated them for ≥50 years.

Under the four scenarios, the rate of landscape-level and riparian-level change has a major effect on the species. In 50 years, under the three scenarios with low rates of negative landscape change, the Ozark Chub will likely remain extant in all watersheds, although with increasingly low population resiliency over time. No EMU is predicted to lose a population at the end of 50 years under scenarios 1, 2, and 3. If extirpation does occur, the species will lose some redundancy and representation, but no EMU has a majority of the extirpation risk. Under the fourth scenario, with a high rate of negative landscape change, condition score drastically declines in 10 years, extirpations occur by 50 years in almost half of the current populations, and resiliency is low. All EMUs may be represented, but redundancy is also low.

REFERENCES Bayless, M, and C. Vitello. 2002. White River inventory and assessment. Missouri Department of Conservation. Beitinger, T. L., W. A. Bennett, and R. W. McCauley. 2000. Temperature tolerances of North American freshwater fishes exposed to dynamic changes in temperature. Environmental Biology of Fishes 58: 237-275. Berkman, H. E. and C. F. Rabeni. 1987. Effect of siltation on stream fish communities. Environmental Biology of Fishes 18(4): 285-294. Brabec, E. and P. L. Richards. 2002. Impervious surfaces and water quality: a review of current literature and its implications for watershed planning. Journal of Planning Literature 16(4): 499-514. Bruton, M. N. 1985. The effects of suspensoids on fish. Hydrobiologia 125: 221-241. Black, J. D. 1940. The distribution of the fishes of Arkansas. PhD dissertation, University of . 228 pages. Boggs, J., G. Sun, and S. McNulty. 2016. Effects of timber harvest on water quantity and quality in small watersheds in the Piedmont of North Carolina. Journal of Forestry 114: 27-40 Bottrell, C. E., R. H. Inersol, and R. W. Jones. 1964. Notes on the embryology, early development, and behavior of Hybopsis aestivalis tetranemus (Gilbert). Transactions of the American Microscopical Society 83(4): 391-399. Buchanan, T. M. 2005. Small fish species of Arkansas reservoirs. Journal of the Arkansas Academy of Science. 59(6): 26-42. Bunger, C. A. 2005. Instrumentation for monitoring timber harvest effects. 141 pp. PhD diss. University of Missouri—Columbia. Center for Biological Diversity. 2010. Petition to list 404 aquatic, riparian and wetland species from the Southeastern United States as threatened or endangered under the Endangered Species Act. Chambers, D. B. and T. Messinger. 2001. Benthic invertebrate communities and their responses to selected environmental factors in the Kanawha River basin, West Virginia, Virginia, and North Carolina. Charlston, WV: U.S. Geological Survey. Water-Resources Investigations Report 01-0421. Clary, W. P. 1999. Stream channel and vegetation responses to late spring cattle grazing. Journal of Range Management 52: 218-227.

Coburn, M. M., and Cavender, T. M. 1992. Interrelationships of North American cyprinid fishes, p.328–373. In: Systematics, historical ecology, and North American freshwater fishes, R. L. Mayden (ed.). Stanford Univ. Press, Stanford, CA.

Conant, R. 1960. The queen snake, Natrix septemvittata, in the interior highlands of Arkansas and Missouri, with comments upon similar disjunct distributions. Proceedings of the Academy of Natural Sciences of Philadelphia 112: 25-40. Corbett, C. W., M. Wahl, D. E. Porter, D. Edwards, and C. Moise. 1997. Nonpoint source runoff modeling: A comparison of a forested watershed and an urban watershed on the South Carolina coast. Journal of Experimental Marine Biology and Ecology 213.1: 133-149. Cross, F. B. 1954. Fishes of Cedar Creek and the South Fork of the Cottonwood River, Chase County, Kansas. Transactions of the Kansas Academy of Science 57(3): 303-314. Dahlberg, M. D. 1979. A review of survival rates of fish eggs and larvae in relation to impact assessments. Marine Fisheries Review March 1979: 1-12. Dimmick, W. W. 1993. A molecular perspective on the phylogenetic relationships of the barbeled , historically assigned to the genus Hybopsis (Cyprinidae: Cypriniformes). Molecular Phylogenetics and Evolution 2(3): 173-184. Eaton, J. G. and R. M. Scheller. 1996. Effects of climate warming on fish thermal habitat in streams of the United States. Limnology and Oceanography 41(5): 1109-1115. Ferguson, B. K. and P. W. Suckling. 1990. Changing rainfall-runoff relationships in the urbanizing Peachtree Creek watershed, Atlanta, . Water Resources Bulletin, American Water Resources Association 26 (2): 313-322. Fields, R., S. S. Lowe, C. Kaminski, G. S. Whitt, and D. P. Phillipp. 1987. Critical and chronic thermal maxima of northern and largemouth bass and their reciprocal F1 and F2 hybrids. Transactions of the American Fisheries Society 116 856-863. Frazier, G. C. and J. K. Beadles. 1977. Fishes of Sylamore Creek, Stone County, Arkansas. Journal of the Arkansas Academy of Science 31: 38-41. Gerking, S. D. 1959. The restricted movement of fish populations. Biological reviews 34(2): 221-242. Gilpin, M. E. and M. E. Soule. 1986. Minimum viable populations: Processes of species extinction. Conservation Biology: The Science of Scarcity and Diversity, M.E. Soulé (ed.). Sinaur Associates, Sunderland, Mass. Goforth, R. R. and J. W. Foltz. 1998. Movements of the yellowfin shiner, Notropis lutipinnis. Ecology of Freshwater Fish 7: 49-55. Griffith, M. B., Zheng, L., and S. M. Cormier. 2018. Using extirpation to evaluate ionic tolerance of freshwater fish. Environmental Toxicology and Chemistry 37(3): 871-883. Harris, J. L. 1986. Systematics, distribution, and biology of fishes currently allocated to Erimystax (Jordan), a subgenus of Hybopsis (Cyprinidae). PhD diss, University of Tennessee – Knoxville.

Hill, J. and G. D. Grossman. 1987. Home range estimates for three North American stream fishes. Copeia 1987: 376-380. Hubbs, C. L., and W. R. Crowe. 1956. Preliminary analysis of the American Cyprinid fishes, seven new, referred to the genus Hybopsis, subgenus Erimystax. Occasional Papers of the Museum of Zoology University of Michigan 578: 1-8.

IPCC. 2007. AR4 Synthesis Report, Intergovernmental Panel on Climate Change. https://www.ipcc.ch/report/ar5/syr/synthesis-report/ ITIS. 2018. Integrated Taxonomic Information System on-line database. http://www.itis.gov. Retrieved August 2018. Jenkins, R. E., and E. A. Lachner. 1971. Criteria for analysis and interpretation of the American fish genera Nocomis Girard and Hybopsis Agassiz. Smithsonian Contributions to Zoology 90: 1-15. Jordan, D. S. 1888. Descriptions of fourteen species of fresh-water fishes collected by the U.S. Fish Commission in the summer of 1888. Proceedings of United States National Museum 351-362. Keith, W. E. 1964. A pre-impundment study of the fishes, their distribution and abundance, in the Beaver Lake Drainage of Arkansas. Master’s Thesis. University of Arkansas, Fayetteville, Arkansas. Koirala, A. K. 2009. An evaluation of pre- and post-timber harvest water quality in low-order streams in the Missouri Ozarks. 287 pp. PhD dissertation, University of Missouri – Columbia. Kaushal, S. S., G. E. Likens, N. A. Jaworski, M. L. Pace, A. M. Sides, D. Seekell, K. T. Belt, D. H. Secor, and R. L. Wingate. 2010a. Rising stream and river temperatures in the United States. Frontiers in Ecology and the Environment 8: 461-466. Kaushal, S. S., M. L. Pace, P. M. Groffman, L. E. Band, K. T. Bell, P. M. Meyer, and C. Welty. 2010b. Land use and climate variability amplify contaminant pulses. Eos, Transactions American Geophysical Union 91: 221-222. Lande, R. 1998. Anthropogenic, ecological and genetic factors in extinction and conservation. Researches on Population Ecology 40 (3): 259-269. Lee, K. H., T. M. Isenhart, and R. C. Schultz. 2003. Sediment and nutrient removal in an established multi-species riparian buffer. Journal of Soil and Water Conservation 58: 1-8. Martin, R. G. and R. S. Campbell. 1953. The small fishes of Black River and Clearwater Lake, Missouri in J. L. Funk, E. M. Lowry, M. H. Patriarche, R. G. Martin, R. S. Campbell, and T. R. O’Connell, Jr. eds. The Black River Studies, University of Missouri Studies 26(2): 45-66. Matthews, W. J. and H. W. Robison. 1988. The distribution of the fishes of Arkansas: a multivariate analysis. Copeia (2): 358-374.

Matthews, W. J., D. J. Hough, and H. W. Robison. 1992. Similarities in fish distribution and water quality patterns in streams of Arkansas: congruence of multivariate analyses. Copia 1992 (2): 296-305 Matthews, W. J., B. C. Harvey, and M. E. Power. 1994. Spatial and temporal patterns in the fish assemblages of individual pools in a midwestern stream (U.S.A.). Environmental Biology of Fishes 39: 381-397. May, C. W., R. R. Horner, J. R. Karr, B. W. Mar, and E. B. Welch. 1998. The cumulative effects of urbanization on small streams in the Puget Sound Lowland Ecosystem. Proceedings of the Puget Sound Research. Mayden, R. L. 1988. Vicariance biogeography, parsimony, and evolution in North American freshwater fishes. Sytematic Zoology 37: 329-355. Mayden, R. L. 1989. Phylogenetic studies of North American minnows, with emphasis on the genus Cyprinella (Teleostei: Cypriniformes). Miscellaneous Publications of the University of Kansas Museum of Natural History 80: 1-189. McGinley, E.J., R.L. Raesly, and W.L. Seddon. 2013. The effects of embeddedness on the seasonal feeding of mottled sculpin. The American Midland Naturalist 170(2): 213-228.

Meek, S. E. 1894. Report of investigations respecting the fishes of Arkansas conducted during 1891, 1892, and 1893, with a synopsis of previous explorations in the same state. Bulletin of the United States Fish Commission: 67-94. Miller, S. M. and T. F. Wilkerson, Jr. 2001. North Fork River watershed inventory and assessment. Missouri Department of Conservation. West Plains, Missouri. Miner, J. R., J. C. Buckhouse, and J. A. More. 1992. Will a water trough reduce the amount of time hay-fed livestock spend in the stream (and therefore improve water quality)? Rangelands 14: 35-38. Missouri Department of Natural Resources. 2018. Section 303(d) Listed Waters. https://dnr.mo.gov/env/wpp/waterquality/303d/documents/1-2018-303d-list-cwc-10-18- 2018.pdf

Mitchell, R. M., R. L. Johnson, and G. L. Harp. 2002. Population structure of an endemic species of yellowcheek darter, Etheostoma moorei (Raney and Suttkus), of the Upper Little Red River, Arkansas. The American Midland Naturalist 148(1): 129-137. Morita, K., and Yamamoto, S. 2002. Effects of habitat fragmentation by dammin on the persistence of stream-dwelling charr populations. Conservation Biology 16 (5): 1318- 1323. NCCV. 2017. U.S. Geological Survey –National Climate Change Viewer. https://www2.usgs.gov/landresources/lcs/nccv/viewer.asp. Accessed April 2019. Owens, L. B., W. M. Edwards, and R. W. Van Keuren. 1996. Sediment losses from a pastured watershed before and after stream fencing. Journal of Soil and Water Conservation 51: 90-94.

Page, L. M., H. Espinoza-Perez, L. T. Findley, C. R. Gilbert, R. N. Lean, N. E. Mandrak, R. L. Mayden, and J. S. Nelson. 2013. Common and scientific names of fishes from the United States, Canada, and Mexico. 7th ed. Bethesda, Maryland: American Fisheries Society. Peterson, J. C. 1998. Water-quality assessment of the Ozark Plateaus Study Unit, Arkansas, Kansas, Missouri, and Oklahoma – fish communities in streams of the Ozark Plateaus and their relations to selected environmental factors. U.S. Geological Survey Water Resources Investigations Report 98-4155, 34 pp. Peterson, J. C. 2004. Fish communities of the Buffalo River basin and nearby basins of Arkansas and their relation to selected environmental factors. U. S. Geological Survey Scientific Investigations Report 2004-5119, 93 pp. Pflieger, W. L. 1997. The fishes of Missoui. Jefferson City: Missouri Department of Conservation. 372 p. Richards, C., L. B. Johnson, and G. E. Host. 1996. Landscape-scale influences on stream habitats and biota. Canadian Journal of Fisheries and Aquatic Science 53: 295-311. Rios, S. L. and R. C. Bailey. 2006. Relationship between riparian vegetation and stream benthic communities at three spatial scales. Hydrobiologia 553: 153-160. Robinson, W. 1980. An ichthyofaunal survey of three Kentucky tailwaters. Master’s Thesis, Western Kentucky University, Bowling Green, Kentucky, USA. Schaefer, J. 2001. Riffles as barriers to interpool movement by three cyprinids (Notripis boops, Campostoma anomalum, and Cyprinella venusta). Freshwater Biology 46: 379-388. Shute, J. R. and P. L. Rakes. 2001. Captive propagation of chubs of the genus Erimystax. American Currents 27(4): 15-16. Sheffield, R. E. , S. Mostaghimi, D. H. Vaughan, E. R. Collins, Jr., and V. G. Allen. 1997. Off- stream water sources for grazing cattle as a stream bank stabilixation and water quality BMP. Stallsmith, B, J. Mann, C. Allen. 2015. Reproductive timing of Erimystax insignis (Blotched Chub) in the Flint River of North . Southeastern Fishes Councel Proceedings 55: 62-73. Strange, R. M. and B. M. Burr. 1997. Intraspecific phylogeography of North American highland fishes: a test of the Pleistocene vicariance hypothesis. Evolution 51: 855-897. Sutherland, A. B., J. L. Meyer, and E. P. Gardiner. 2002. Effects of land cover on sediment regime and fish assemblage structure in four southern Appalachian streams. Freshwater Biology 47: 1791-1805. Swank, W. T., J. M. Vose, and K. J. Elliott. 2001. Long-term hydrologic and water quality responses following commercial clearcutting of mixed hardwoods on a southern Appalachian catchment. Forest Ecology and Management 143: 163-178.

Tarver, C. L. and B. Stallsmith. 2019. Reproductive biology of Hybopsis amblops, the Bigeye Chub, in the Flint River of Alabama. PeerJ Preprints. https://doi.org/10.7287/peerj.preprints.26719v2 U.S. Fish and Wildlife Service. 2016. USFWS Species Status Assessment Framework. Version 3.4 dated August 2016. U.S. Fish and Wildlife Service. 2018. Recovery Plan for the Yellowcheek Darter (Etheostoma moorei). Atlanta, Georgia. 5 p. Vidon, P., M. A. Campbell, and M. Gray. 2008. Unrestricted cattle access to streams and water quality in till landscape of the Midwest. Agricultural Water Management 95: 322-330. Vogt, N. E. 2013. Spatial transition of the fish community and environmental characteristics in the white river below Bull Shoals Dam. Master’s Thesis. Arkansas Tech University, Russellville, Arkansas. Wang, L., J. Lyons, P. Kanehl, and R. Gatti. 1997. Influences of watershed land use on habitat quality and biotic integrity in Wisconsin streams. Fisheries 22: 6-12. Wang, L., J. Lyons, P. Kanehl. 2001. Impacts of urbanization on stream habitat and fish across multiple spatial scales. Environmental Management 28: 255-266. Webb, B. W. and F. Nobilis. 1997. Long-term perspective on the nature of the air-water temperature relationship: a case study. Hydrological Processes 11: 137-147. Widenfalk, A. 2002. Pesticide bioavailability in aquatic sediments – a literature review. IMA Rapport 2002: 11. Department of Environmental Assessment, Swedish University of Agricultural Sciences – Uppsala. Wenger, S. J. and L. Fowler. 2000. Protecting stream and river corridors: creating effective local riparian buffer ordinances. 68pp. Carl Vinson Institute of Government, University of Georgia, America. Westhoff, J. T. and C. Paukert. 2014. Climate change simulations predict altered biotic response in a thermally heterogeneous stream system. PLoS ONE 9: e111438. https://doi.org/10.1371/journal.pone.0111438 Westhoff, J. T., C. Paukert, S. Ettinger-Dietzel, H. Dodd, and M. Siepker. 2014. Behavioral thermoregulation and bioenergetics of riverine smallmouth bass associated with ambient cold-period thermal refuge. Ecology of Freshwater Fish 25: 72-85. Wilde, G. R. and B. W. Durham. 2008. Daily survival rates for juveniles of six Great Plains cyprinid species. Transactions of the American Fisheries Society 137: 830-833. Wildhaber, M. L., A. L. Allert, C. J. Schmitt, V. M. Tabor, D. Mulhern, K. L. Powell, and S. P Sowa. 2000. Natural and anthropogenic influences on the distribution of the threatened Neosho madtom in a Midwestern warmwater stream. Transactions of the American Fisheries Society 129: 243-261.

Wilson, R. K. and R. T. Clark. 2002. EC830 costs of cattle fencing for grazing areas. Historic Materials from University of Nebraska-Lincoln Extension 75: 1-8. Wolf, S., B. Hartl, C. Carroll, M. C. Neel, D. N. Greenwald. 2015. Beyond PVA: Why recovery under the Endangered Species Act is more than populations viability. BioScience 65: 200-207. Zaimes, G. N., R. C. Schultz, and T. M. Isenhart. 2004. Stream bank erosion adjacent to riparian forest buffers, row crop fields, and continuously grazed pastures along Bear Creek in central . Journal of Soil and Water Conservation 59(1): 19-27.

Appendix A. Ozark Chub Ecological Management Units and Watersheds

Figure A1. Map of the Ozark Chub range with collection data and watersheds identified by Ecological Management Unit.

EMU MU Size (km2) Upper White Flat 845 Kings 1,536 Swan 516 Beaver Creek 1,102 Long Creek 683 N. White/Beaver Res 3,734 Upper James 698 Lower James 1,992 Bull/B 475 Bee 95 Black Strawberry 2,004 Spring 3,151 Eleven Point 3,129 Current 5,660 Black 3,221 Middle White Sylamore 568 North Fork 2,449 Piney 451 Buffalo 3,471 Lower White 627 Crooked 1,196 St. Francis St. Francis 2,816 Little Red S. Fork/Archery Fork 708 Table A1. Size of watersheds (in KM2) within the range of the Ozark Chub.

Upper White River Ecological Management Unit The Upper White River EMU is composed of 10 watersheds flowing into the White River above Bull Shoals Dam. This EMU contains some of the largest urban areas within the range of the Ozark Chub; Springfield, Missouri is situated in the James River watershed, and the North White River drains some of the exurban development in the quickly growing Fayetteville-Bentonville metro area. The watersheds in this EMU are isolated by Beaver Lake, Lake Taneycomo, and Table Rock Lake, all constructed prior to 1965. All of the watersheds except the North White River/Beaver Reservoir have documented records of Ozark Chub since 2000.

Figure A2. Average percent of land-cover use (±SD) across the 10 watersheds of the Upper White EMU. The watersheds within the Upper White River EMU average 5% urban development, 33% pasture/hay, and 57% forest. However, this EMU has considerable variation between watersheds, ranging from Bee Creek (approximately 117,400 acres, 2 % urban development, 14% pasture, and 80% forested) to the Lower James River (492,300 acres, 15% urban development, 32% forested, and 49% pasture) (Figure A2). In Arkansas, the Kings River is designated an Extraordinary Resource Waterbody, a beneficial use ranking based on a combination of the chemical, physical, and biological characteristics of a waterbody and its watershed which is characterized by scenic beauty, aesthetics, scientific values, broad scope recreation potential, and intangible social values. Additionally, a segment in Madison County is designated as a Natural and Scenic Waterway, a legislatively accepted beneficial use ranking. Eight miles of Bull Creek is designated as Outstanding State Resource Waters. The Kings River, North White River/Beaver Resevoir, Flat Creek, Bull Creek, Swan Creek, and Beaver Creek watersheds contain property managed by the U.S. Forest Service as part of the Ozark-St. Francis and Mark Twain National Forests. Several state-owned wildlife management areas and NGO-owned properties are also in this EMU. Six of the watersheds currently have waters listed on the 303(d) Impaired Waterways list in either Arkansas or Missouri. In 2016, Flat Creek was listed for elevated mercury levels, and three tributaries of the James River were listed due to unknown non-point source pollution. A portion of Beaver Creek was listed in 2008 for sedimentation, and in 2014 for excess E. coli. Three tributaries in the North White River/Beaver Reservoir watershed were also listed for sulfates and total dissolved solids (TDS). In the Arkansas 2018 draft 303(d) impaired waterway report, two segments of the Kings River equaling 61.2 miles were listed: 38.2 miles for elevated TDS and the remainder for low dissolved oxygen (DO). Temperatures exceeding the presumed thermal tolerance (>34°C) and specific conductivity tolerance (>744µS/cm) of Ozark Chub have also been recorded in the Kings and James Rivers, almost entirely since the late 1990’s.

Middle White River Ecological Management Unit The Middle White River EMU is composed of five watersheds and the mainstem White River below Bull Shoals dam. The confluences of all five watersheds with the White River were historically geographically close and there was likely movement of individuals between the five watersheds and the mainstem White River. Only one of the watersheds, the North Fork White River, is isolated; the construction of Norfork Lake in the early 1960s inundated the lower portion of this river and isolated upstream populations. The Norfork Lake and Bull Shoals cold- water tailwaters likely restrict the distribution of the species for kilometers downstream of the dams.

Figure A3. Average percent of land-cover use (±SD) across the 6 populations of the Middle White EMU.

The watersheds within the Middle White River EMU average 4.5% urban development, 67% forest, and 14% pasture (Figure A3). Crooked Creek is the most impacted of the watersheds in this EMU, with 6% urban development, 45% pasture, and only 44% forested, followed by the Lower White, which has the second highest percentage of row-crop land cover of all Ozark Chub watersheds (6%). Although the North Fork White River has urban development on par with the Sylamore, Piney, and Buffalo River watersheds, it has a much higher pasture to forest ratio (27% pasture to 67% forest). The Sylamore, Piney, and Buffalo Rivers are more than 70% forested and less than 20% pasture. The Buffalo River, North Fork White River, and Sylamore River contain the majority of protected public land within the Middle White River EMU. A large portion of the North Fork White River watershed is part of the Mark Twain National Forest, and there are also two large conservation areas within the watershed. The North Fork White River has never been listed for water quality issues, and a 5.5 mile segment is categorized as Missouri Outstanding State Resource Waters. The highest water temperature recorded in the watershed in manual samples

between 1969 and 2018 was 29°C, while the highest specific conductance was 460 µS/cm – both below the maximum for Ozark Chub. The entire length of the main stem Buffalo River is protected by the National Park Service, and it is designated a Natural and Scenic Waterway and an Extraordinary Resource Waterbody. Additionally, much of the southern portion of the watershed, including the headwaters of four of its main tributaries are within the Ozark-St. Francis National Forest. Approximately half of the Sylamore watershed, including the entire North Sylamore Creek, is within the Ozark-St. Francis National Forest. There are a few wildlife management areas and conservation areas within the remaining watersheds, but they make up a very small proportion of the watershed areas. Three of the watersheds have listed water impairments. The upper half of Crooked Creek and Bear Creek, a tributary of the Buffalo River, are both listed as a 303(d) Impaired Waterways in Arkansas for TDS, and temperatures >36°C have been recorded in both rivers in summer months. Hicks Creek and Greenbrier Creek, tributaries of the main stem White River, are 303(d) listed for Nitrates and DO, respectively. No Ozark Chub were collected in extensive surveys in Piney Creek between 1972 and 1973, and surveys in the same time period found only one Ozark Chub in the Sylamore River, near the confluence with the White, although more recent collections of the species have been made in both watersheds (Frazier and Beadles 1977, p. 39). Even at this point in time, sedimentation was noticed as a major limiting factor to fish distribution and the lower portion of the stream was characterized by silty sand (Matthews and Harp 1974, p. 42). Although Ozark Chub have been collected in the Piney more recently, it is likely that sedimentation is still a large issue in this watershed and the EMU as a whole. As current records in both watersheds are further upstream than Frazier and Beadles (1977) collection, the Ozark Chub population is likely either expanding or excessive sedimentation in the lower portions of both streams are pushing populations to clearer upstream waters. Middle Black River Ecological Management Unit The Middle Black EMU contains the Strawberry, Spring, Current, Eleven Point, and Black Rivers, and is the largest EMU by watershed area. All five watersheds flow into the Middle Black River; it is likely that there was gene flow between the watersheds historically and some movement probably still occurs between the watersheds, although the last record from the main stem Black River is from 1984. Clearwater Lake in the Upper Black River likely restricts gene flow between the upper and lower portions of the river.

Figure A4. Average percent of land-cover use (±SD) across the 5 watersheds of the Middle Black EMU. Within the Middle Black EMU, the watersheds range from the Strawberry (59.5% forested, 5% development, and 29% pasture) to the Upper Black (85% forested, 3% development, and 6% pasture). The Middle Black EMU is on the edge of the Ozark Plateau, and the row-crop land cover common to the Mississippi Delta region is evident in the Upper Black and Strawberry River watersheds (approximately 1%) (Figure A4). The Eleven Point, Spring, Strawberry, and Current rivers are all designated as Extraordinary Resource Waterbodies in Arkansas. In Missouri, the headwaters of the Current and Eleven Point rivers and 3 miles of the East Fork Black River are designated as Outstanding National Resource Waters. The headwaters of the Strawberry River are classified as a Natural and Scenic Waterway in Arkansas. The majority of public land in this EMU is located in the Eleven Point, Current, and Upper Black Watersheds. Three different units of the Mark Twain National Forest lie within these watersheds, and a there are also a large number of state-owned conservation areas. All the watersheds within this EMU are listed on either Arkansas’ or Missouri’s 303(d) Impaired Waterways List. Sedimentation is likely a major stressor for the Ozark Chub and approximately half of the Strawberry River is listed as a Category 4 303(d) waterway for sedimentation. Interestingly, although we have recent collections of Ozark Chub in the Strawberry River, none are within the 303(d) listed area. High pH impairs approximately 20 miles of the Strawberry River. Impaired waters also occur in the Warm Fork Spring River (E. coli), portions of the Eleven Point (mercury), and the entire Current and Black Rivers (mercury). Additionally, there are a number of heavy metal mines in the upper reaches of the Black River, and short segments downstream of mines are 303(d) listed in Missouri for elevated levels of lead, nickel, and cadmium. Temperatures between 36 and 40°C occasionally occur in the main stem Black River, and specific conductance exceeding the assumed Ozark Chub limit have been recorded in the Black

and Current Rivers, most often directly downstream of mines, lakes, and wastewater treatment facilities. St Francis River Ecological Management Unit The St. Francis River EMU contains only one population, the Upper St. Francis River. This watershed is isolated from other Ozark Chub populations, as the St. Francis River travels through the eastern portion of Missouri and Arkansas and merges with the Mississippi River south of Memphis, Tennessee. As discussed in Section 2.3, the St. Francis River was originally part of a larger pre-glacial river system with the White River, and it is likely the ancestor of the Ozark Chub traveled to the Ozark Plateau through the St. Francis River.

Figure A5. Percent watershed-wide and riparian (30m buffers) land-cover within the Upper St. Francis River. The Upper St. Francis River watershed is approximately 5% urban development, 75% forested, and 16% pasture (Figure A5). It is the only watershed within the range of the Ozark Chub with a smaller proportion of forest within its riparian buffers than the watershed as a whole. Approximately 41% of the St. Francis River watershed is in public ownership, the majority of which is managed by the U.S. Forest Service as part of the Mark Twain National Forest in the central and lower portions of the watershed. 93.1 miles of the Upper St. Francis River is listed on the 303(d) list of impaired waters for failing to meet temperature standards. Additionally, southeastern Missouri has been a primary producer of lead and other heavy metals, and the Upper St. Francis River has historically had a large number of lead mines. Because of this, two tributaries are on the 303(d) list of impaired waters for lead and cadmium levels above water quality standards.

Little Red River Ecological Management Unit The Upper Little Red River EMU contains one known population comprised of the Archey Fork Little Red River and the South Fork Little Red River. These two headwater streams converge near the town of Clinton, Arkansas, and are isolated from the other two headwater tributaries and the White River system by Greers Ferry Lake, which was completed in 1964. As discussed in Section 2.3, Ozark Chub from the Little Red River system are morphologically intermediate between Ozark Chub from the White River system and the eastern Streamline Chub, and it is likely the Little Red River population was isolated from other Ozark Chub populations prior to dam construction. However, it is likely that this species was once wider spread throughout the Little Red River but was restricted to the headwater forks with the construction of Greers Ferry Lake, as happened with the endemic Yellowcheek Darter. Although the last collection in the watershed was in 1999 (Figure A1), scientific collections and studies within the watershed have primarily focused on the endangered Yellowcheek Darter, and no large-scale or Ozark Chub- specific collections have been implemented; the status of the Ozark Chub in the upper Little Red River is uncertain

Figure A6. Percent watershed-wide and riparian (30m buffers) land-cover within the Little Red River EMU. The Little Red River EMU is primarily forested (81%), with approximately 3% urban development and 10% pasture; most development is localized in Clinton, Arkansas near the confluence of the two forks and the lake. This watershed has been the focus of many riparian and waterway restoration efforts to conserve the endemic Yellowcheek Darter, and riparian buffers are primarily forested and scrub/herbaceous cover (84% combined) with very little urban development or pasture (both 1.8%) (Figure A6). Public land in the watershed includes a portion of the Ozark National Forest as well as state and NGO-owned conservation areas. The South Fork Little Red River was listed on the 303(d) list of impaired waterways in 2008 for mercury levels in fish tissue higher than the water quality standard. Additionally, chemicals used

in the hydraulic fracturing process for natural gas extraction and other potentially detrimental chemical are routinely transported across streams in the upper Little Red watershed, and potentially catastrophic spills could occur. The Yellowcheek Darter requires many of the same habitat parameters as the Ozark Chub, such as clear water, permanent flow, and gravel, cobble, and boulder substrate, although it is a riffle- inhabiting species (USFWS 2018). Recovery actions for the Yellowcheek Darter include protection of water quality and quantity and implementation of best management practices (BMPs) throughout the region to prevent or reduce pollution and sediments (USFWS 2018). As such, many conservation efforts for the Yellowcheek Darter will also benefit extant Ozark Chub populations