Southeastern Naturalist 259 P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 22001166 SOUTHEASTERN NATURALIST 1V5o(2l.) :1255,9 N–2o6. 82 The Influence of Streamflow on the Recruitment of Rock Bass in the New River Pearce Cooper1, John R. Copeland2, Sean C. Lusk3, and Donald J. Orth4,* Abstract - Few studies have addressed the effect of stream discharge on Ambloplites rupestris (Rock Bass). We investigated the effects of spawning-season discharge on Rock Bass recruitment in the New River at 2 sites, 5 and 55 miles downstream from Claytor Dam, VA. We back-calculated length at age 1 and 2 using otoliths in order to estimate average lengths and used these values to identify age-1 fish from 14 years of fall electrofishing data on the New River. We determined a catch per unit effort (CPUE) of age-1 fish at each site and correlated it to spawning-season discharge data from the previous year using Spearman’s rho rank-correlation. The CPUE at the upstream site was negatively correlated with average and maximum discharge in August, while the CPUE at the downstream site was negatively correlated with average and maximum discharge in July. August and July directly precede spawning season of Rock Bass in Virginia. We surmise that high levels of discharge strongly influence mortality in juvenile Rock Bass. Introduction Fish species exhibit a variety of currently unquantified ecological responses to changes in flow in the fluvial environment. Researchers have determined that high levels of stream flow during the spawning season negatively affect the recruitment of Micropterus dolomieu (Lacépède) (Smallmouth Bass) due to the disruption of spawning, as well as egg and larval mortality (Graham and Orth 1986, Lukas and Orth 1995, Smith et al. 2005). Several studies have addressed this topic in Virginia alone (Graham and Orth 1986, Lukas and Orth 1995, Smith et al. 2005). However, very few studies have investigated the effects of stream flow on the recruitment of Ambloplites rupestris (Rafinesque) (Rock Bass), and none of this research has been done in the southeastern US. Though Rock Bass are an important game fish in Virginia, they are less studied than Smallmouth Bass because they are smaller and less sought-after by anglers. Evidence that high flow negatively effects the recruitment of 2 important game fish in Virginia would be an incentive to regulate the release of water by hydroelectric dams during the spawning periods of these species, which overlap in Virginia (Cashner and Jenkins 1982, Surber 1970). The Rock Bass is native to the upper and middle Mississippi River drainages, and parts of the eastern US and southeastern Canada. The species has been 1Department of Marine Sciences, University of South Alabama, Mobile, AL 36688. 2Virginia Department of Game and Inland Fisheries, 2206 South Main Street, Blacksburg, VA 24060. 3School of Fisheries, Aquaculture, and Aquatic Sciences, Auburn University, Auburn, AL 36849. 4Department of Fish and Wildlife Conservation, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061. *Corresponding author - dorth@vt.edu. Manuscript Editor: Max Nickerson Southeastern Naturalist P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 260 introduced throughout other drainages across North America because they are a popular sport fish (Cashner and Jenkins 1982). Similar to Smallmouth Bass, Rock Bass prefer clear water with rocky substrate and occupy a variety of stream sizes and lakes (Probst et al. 1984). In much of their range, Rock Bass occupy a similar niche to the Smallmouth Bass, feeding on a variety of fishes, insects, and crustaceans (George and Hadley 1977). In Virginia, the spawning periods of both species overlap in May and June; Rock Bass typically spawn from April to June (Surber 1970, Wallus and Simon 2008). Similar to Smallmouth Bass and other centrarchids, fertilized eggs are guarded by the male in a gravel depression (Cashner and Jenkins 1982, Surber 1970). Life-history similarities between Rock Bass and Smallmouth Bass suggest that high flow during the spawning season of Rock Bass may have a similar negative effect on Rock Bass recruitment. Noltie and Keenleyside (1986) correlated minimum-flow rates during the spawning season, among other factors, to high reproductive success of Rock Bass in streams in Ontario, Canada. In addition, a study of a fish assemblage in an Illinois stream showed that the apparent reproductive success of Rock Bass was significantly reduced by high flow (Schl osser 1985). In the current study, we evaluated 14 years of discharge data and fall electrofishing data in the New River below Claytor hydroelectric dam and farther downstream in a section of stream where flow is less regulated by the dam to assess the effect of flow rate during and directly after the reported spawning season on Rock Bass recruitment. We correlated our measure of recruitment to age 1 with an average and maximum daily flow by month in both locations for the April–August period of the previous year using data from USGS gaging stations. We employed Spearman’s rho rank-correlation to correlate these variables for each month. Based on studies associated with Smallmouth Bass and Rock Bass spawning season, we hypothesized that Rock Bass recruitment would be negatively correlated with average and maximum discharge in the months of May and June. Field-Site Description We chose Radford and Glen Lyn, VA, as our sites for discharge measurement because the US Geological Survey (USGS) has a gaging station at each location, and both are within the Virginia Department of Fish and Game (VGDIF) sampling area. The Radford gaging station is located just over 8.0 km (5 mi) downstream of Claytor Dam, a hydroelectric dam on the New River that regulates the water level of Claytor Lake. Discharge in this section of stream is effectively regulated by dam outflow due to the site’s proximity to the dam. The Glen Lyn gaging station is about 88.5 km (55 mi) downstream of Claytor Dam and represents a less-regulated stream where flows from Claytor Dam are attenuated by distance from the dam. Major tributaries such as Walker Creek and Wolf Creek enter the river between the sites and discharge heavily during storm events, contributing to the less-regulated nature of the downstream site. Though the Radford site was sampled every year for our study, few CPUE data were available from Glen Lyn because it had not been routinely sampled; however, Southeastern Naturalist 261 P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 we instead used data collected from alternative sites close to the Glen Lyn gaging station. These alternative sites included Rich Creek (data for 1998–2008 and 2012), Shumate’s Falls (data for 2009–2011), and Bluff City (data for 1996). Shumate’s Falls, Bluff City, and Rich Creek are within 4.83 river km (3 river mi) of Glen Lyn. These 3 sites are collectively referred to as the downstream sites and Radford is referred to as the upstream site. Both the upstream and downstream sites have electrofishing catches of age-1 Rock Bass for all years included in the study, indicating that these sites provide suitable spawning habitat. Methods Back-calculation and determination of an age-1 length interval We back-calculated length-at-age to identify age 1 Rock Bass from the dataset. In October 2012, we captured 80 Rock Bass at 3 New River sampling sites near the towns of Radford, Whitethorne, and Pembroke. Radford, Whitethorne, and Pembroke are 8 river km (5 river mi), 32.19 river km (20 river mi), and 56.33 river km (35 river miles) downstream from Claytor Dam, respectively. We sacrificed and subsequently aged all Rock Bass specimens captured. We measured the total length (TL) of the fish in the lab prior to otolith extraction. We removed the saggital pair of otoliths from each fish and stored them dry in labeled coin envelopes. We determined fish age by counting otolith annuli under an Olympus SZ-61 dissecting scope with a digital camera on 2x magnification equipped with a Fiber-Lite M-150 High Density Illuminator (Dolan-Jenner Industries, Boxborough, MA) as a fiber-optic light source. We used Image-Pro Plus 6.1 (MediaCybernetics, Rockville, MD) to adjust the image until the annuli were most visible. Two independent readers read each pair of otoliths and recorded the number of annuli. If there was agreement between readers, we recorded the age and included the otolith in back-calculation analyses. If the readers disagreed, both readers and a 3rd reader re-read the pair. If there was agreement during the 2nd reading, we retained the pair for back-calculation analysis and if not, we removed the otoliths from analysis. We used the calibrated-measurement tool in Image Pro Plus 6.1 to calculate the length from the otolith nucleus to the outer edge of the first and second opaque annuli and the distance to the outer edge of the otolith of our 74 aged otoliths. We took measurements along the same axis at the point where the edges of the annuli were most clear. We then calculated the length at age 1 and 2 using the direct proportional method: Lx = (Sx/ST) × LT , where, Lx is the TL at age x, Sx is the distance from the focus to the xth annulus, ST is the distance from the focus to the edge of the otolith, and LT is the TL of the fish at capture (Schramm et al. 1992). We employed this method to determine TL at age 1 and 2 and then calculated the arithmetic mean and standard error of the lengths at age 1 and 2. We tested for significant differences in average lengths at age 1 and 2 using a 2-sample t-test assuming equal variance and a separate t-test assuming unequal variance. In order to check the accuracy of our calculations, we compared Southeastern Naturalist P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 262 them to Rock Bass age data from the same sample. We calculated the average length of age-1 fish from our 2012 sample in order to determine whether they fell between the calculated length at age 1 and 2. The age-1 Rock Bass that we collected in 2012 and in previous fall surveys should have fallen between the back-calculated lengths at age 1 and 2 because they likely hatched during April–June the previous year, and hence, their age should have been ~1.5 y. Measuring recruitment In this study, we calculated Rock Bass recruitment to age 1 for 14 year-classes using data collected in late September and throughout October during the annual fall boat electrofishing survey of the New River. We used lengths of captured Rock Bass included in data from the years 1996–2012 (excluding 1997, 2002, and 2003). We calculated recruitment by determining the number of Rock Bass in the age-1 length-interval captured each year at the upstream site and at 1 of the downstream sites. We used data from age-1 fish as opposed to young-of-the-year fish because boat electrofishing has an inherent sampling bias of capturing reduced numbers of very small fish, which fail to recruit to the electrofishing gear (Jackson and Noble 1995). We recorded the number of fish that were between the lengths at first- and second-annulus formation and classified them as age 1. We calculated an average age-1 CPUE in fish/h for each transect at each site and used these values to determine an overall average age-1 CPUE for each site to be used as a measure of recruitment to age 1. Measuring discharge We used data from the USGS New River gaging stations at Radford and Glen Lyn to account for discharge from April to August 1995 to 2011 (years prior to sampling years). We used daily discharge data from the USGS website (http://waterdata. usgs.gov/va/nwis/rt) to calculate a maximum and average daily discharge with standard error for each month. We included the months April–August in our analysis to account for Rock Bass spawning and early life-history seasons. Determining correlation We made separate comparisons of average and maximum discharge during each month at both sites to the next year’s CPUE of Rock Bass in the age-1 length interval in order to determine which month’s discharge had the greatest effect on Rock Bass recruitment. We analyzed both the CPUE and discharge data with the Shapiro-Wilk test for normality in the statistical program SAS JMP. We used the non-parametric Spearman’s rho rank-correlation test in SAS JMP to determine correlation between the variables of CPUE and discharge because the Shapiro-Wilk test indicated that their distributions were both non-normal even after transformations. We tested the correlation of CPUE with average discharge and maximum discharge for each month from April to August for each site. We considered P-values significant if they were ≤0.05. We created scatter plots showing the relationship between the CPUE of age-1 fish for each site and the average and maximum discharge the previous year for the months April–May for each site. Southeastern Naturalist 263 P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 Results Back-calculation The average TL of New River Rock Bass was 70 mm at age 1 and 111.7 mm at age 2 (Table 1). We observed a significant difference between the sets of lengths at annulus 1 and 2 (P < 1x10-32). The average lengths of age-1(~age-1.5) Rock Bass collected in the fall during the New River sampling survey in 2012 fell between our calculated lengths of age 1 and 2 (Table 1). Recruitment CPUE of age-1 fish ranged from 0.8 to 28.5 fish/h at the upstream site and 1.0 to 98.5 fish/h at the downstream sites. In 2008, the downstream site had the highest CPUE of 98.5 fish/h—more than double the next-highest CPUE. The average CPUE of individuals in the age-1 length interval at the upstream site in 1996 was less than 1 fish/h and the CPUE of the downstream site was less than 5 fish/h. In 2004, average age-1 CPUEs at both the upstream site and the downstream site were less than 1 fish/h (data not shown). Discharge Daily discharge at the less-regulated Glen Lyn site varied to a greater degree than at the more-regulated Radford site, with standard deviations of 109 m3/s and 76 m3/s, respectively. Average discharge for both sites ranged from ~28 m3/s to 339.8 m3/s, and maximum discharge went as high as 1379 m3/s in April 1998. For both sites, the average discharge decreased each successive month from April to August (data not shown). Correlation Significance values for correlations varied by month (Table 2). The negative values of Spearman’s rho indicate that as discharge increased, Rock Bass recruitment decreased. At the Radford Site, there was a significant negative relationship when we compared the maximum or average discharge in the month of July with the CPUE of individuals captured in the age-1 length interval the subsequent year (Table 2). At the downstream site, there was a significant negative relationship when we compared the maximum or average discharge in the month of August with the CPUE of captured individuals in the age-1 length interval the subsequent year (Table 2). When we graphed the significant relationships between discharge and CPUE, we observed a steep negative relationship that varied in the level of linearity depending on the month (Fig. 1). Table 1. The average back-calculated length (mm) at age 1 and 2 of Rock Bass captured from the New River in 2012 and the average length of ~1.5-y-old Rock Bass captured and aged in 2012. SD = standard deviation of each sample and n = the number of observations or sample size. Year Method Age Length (mm) SD n 2012 Back-calculated 2 111.7 15.4 62 2012 Back-calculated 1 70.0 40.0 74 2012 Aged 1+ 96.0 14.7 12 Southeastern Naturalist P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 264 Figure 1. The relationship between average and maximum discharge (m³/s) in the previous year and the catch per unit effort (CPUE, # fish/h) of age-1 Rock Bass at the upstream and downstream sites during the months the relationship was found t o be significant. Table 2. Spearman’s rho values and associated P-values analyzing the relationship between average and maximum discharge (m3/s) in different months and the catch per unit effort (CPUE, fish/h), of individuals in the age-1 length interval caught the subsequent year. We considered P-values less than 0.05 as statistically significant denoted by *. Upstream site Downstream site Spearman’s rho P-value Spearman’s rho P-value April average 0.0945 0.7479 -0.1648 0.5733 April max -0.1209 0.6806 -0.1077 0.7140 May average -0.0198 0.9465 -0.1385 0.6369 May max -0.0396 0.8931 -0.1385 0.6369 June average -0.3802 0.1799 -0.3802 0.1799 June max -0.4198 0.1351 -0.2396 0.4094 July average -0.7319 0.0029* -0.4198 0.1351 July max -0.7714 0.0012* -0.0637 0.8286 August average -0.0154 0.9584 -0.8901 less than 0.0001* August max 0.033 0.9109 -0.6703 0.0087* Southeastern Naturalist 265 P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 Discussion Several studies of Smallmouth Bass, a species similar to Rock Bass, have shown a negative effect of high discharge during their spawning season (Graham and Orth 1986, Lukas and Orth 1995, Smith et al. 2005). Noltie and Keenlyside (1986) showed a negative effect of high discharge on Rock Bass recruitment during the species’ spawning season in Canada. We hypothesized that we would observe a negative effect of discharge on Rock Bass in the New River during their reported spawning season. However, our results indicate that a significant effect of discharge on Rock Bass recruitment in this section of the New River occurred only in the months after the reported spawning season. This delayed effect may be due to either a later Rock Bass spawning season in this section of New River or effects of discharge on juveniles after they left the nest. Based on temperature data and the reported spawning season for the species, we conclude it is more likely that discharge affects juvenile Rock Bass after they left the nest. Rock Bass recruitment was negatively affected by discharge in July at the upstream site and in August at the downstream site (Table 2)—a period of the year after which the spawning season of the species in Virginia has likely ended. Jenkins and Burkhead (1994) stated that Rock Bass can spawn from April to July depending on location and weather. Rock Bass in the Shenandoah River system in Virginia were observed nesting from April to early June (Surber 1970). According to Tyus (1970), female Rock Bass remained gravid only into late June in the upper reaches of the New River near Sparta, NC. Wallus and Simon (2008) reported that Rock Bass spawning seasons in multiple states all end before July with the exception of Michigan, where Rock Bass can spawn as late as mid-July. If Rock Bass are most likely spawn from April to June in Virginia, why do they appear to be most vulnerable to discharge in July and August? Two possible explanations are that Rock Bass in this section of the New River are spawning later than average, or juvenile Rock Bass that have recently left the nest are particularly vulnerable to increases in discharge. The first possibility is that Rock Bass in this section of the New River are simply spawning later than reported in the rest of Virginia and many other states. There are no published data on the timing of Rock Bass spawning in this section of the New River. However, Tyus (1970) reported on the timing of Rock Bass spawning in the New River in North Carolina. Female Rock Bass contained mature ova until 28 June 1968 at a water temperature of 21.1–22.2 °C, but by 6–7 July (22.2–24.4 °C), fish contained no mature ova, and it was evident they had already spawned. We examined 2 years of unpublished VDGIF temperature-logger data from this section of the river. Average temperatures reached above 23 °C in July near the upstream and downstream sites. Records of long-term average temperatures are needed, but when we considered this baseline temperature data in conjunction with published Rock Bass spawning seasons, we concluded that it seems unlikely that Rock Bass in the New River spawn in July or August. Rock Bass larvae can be affected by discharge as they develop in their nest (Noltie and Keenlyside 1987). Therefore, a brood that was fertilized in late June could still be affected by discharge in early July. Noltie (1986) found that Rock Southeastern Naturalist P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 266 Bass eggs hatch in about 4 d and disperse from the nest as free-swimming larvae in about 10 d. Noltie and Keenlyside (1987) found that during this period, high flow causes nest failures due to the physical removal of the larvae from the nest or by siltation of the nest. If the Rock Bass in the section of river we studied spawn as late as 28 June like they do in NC, eggs and larvae would still be vulnerable during a period of about 14 days before they leave the nest. Therefore, Rock Bass recruitment could be affected by discharge in July before young leave the nest. However, it would be extremely unlikely for recruitment to be affected in August if mortality of larvae in the nest was the mechanism for the observed effect of discharge. Though scouring of the nest, sediment deposition, and reduction of suitable substrate have been documented as common mechanisms by which increased flow can reduce recruitment (Graham and Orth 1986, Lukas and Orth 1995, Smith et al. 2005), it seems unlikely that these factors are affecting recruitment at our sites if Rock Bass are following a predictable spawning pattern. Lastly, the relationships we observed indicate that CPUE is affected by moderate changes in discharge. Rock Bass nests are built in sheltered locations; thus, we would not expect that these moderate changes in discharge would produce a noticeable effect on larvae within the nest (Noltie and Keenlyside 1987). We conclude that juveniles are being negatively impacted by discharge after they have left the nest. Further research is needed to confirm this conclusion and determine the causative mechanism. The simplest explanation is that young-of-the year Rock Bass are simply being swept downstream during periods of high discharge, and any significant input from upstream is being blocked by Claytor Dam. However, there are many other possible scenarios and the real mechanism may involve one or more of these scenarios. Also, more research is needed to determine why recruitment at the upstream site was most affected by discharge in July, yet recruitment at the downstream site was most affected by discharge in August. This finding may suggest different spawning times for Rock Bass in these locations, but contradictory to our results, it would seem that Rock Bass in Radford should spawn later because the mean temperature of the New River at Radford is consistently lower than the New River at the downstream site (Thomas R. Payne and Associates et al. 2009). An observational study is needed to clarify our results. If the timing of spawning and the dispersal of Rock Bass from the nest were observed in this section of stream, then we could determine with more certainty when young Rock Bass are most affected by discharge and if spawning times differ between sites. If juvenile Rock Bass are being affected after they leave the nest, then determining the mechanism would be difficult, but it would be useful to examine a dataset that tracked water temperature and discharge. Though the mechanism of the effect is not quite clear, our study shows that Rock Bass recruitment is negatively affected by discharge during the late summer, and that it is most likely juvenile Rock Bass that are being affected. We hope that the results reported here and those from future studies will help provide a rational basis for decisions about the management of these fish populations and the controlled release of water from dams. Southeastern Naturalist 267 P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 Acknowledgments We thank George Palmer and other VDGIF employees who helped on the annual fall electrofishing survey. We are grateful to the Frimpong Lab in the Virginia Tech Department of Fish and Wildlife Conservation for letting us use their microscopes and imaging software. Equipment and software was provided by Virginia Tech, all electrofishing data was provided by the VDGIF, and all discharge data was provided by the USGS website. Literature Cited Cashner, R.C., and R.E. Jenkins. 1982. Systematics of the Roanoke Bass, Ambloplites cavifrons. Copeia 1982:581–594. Jackson, J.R., and R.L. Noble. 1995. Selectivity of sampling methods for juvenile Largemouth Bass in assessments of recruitment processes. North American Journal of Fisheries Management 15:408–418. George, E.L., and W.F. Hadley. 1979. Food and habitat partitioning between Rock Bass (Ambloplites rupestris) and Smallmouth Bass (Micropterus dolomieu) young-of-theyear. Transactions of the American Fisheries Society 108:253–261. Graham, R.J., and D.J. Orth. 1986. Effects of temperature and streamflow on time and duration of spawning by Smallmouth Bass. Transactions of the American Fisheries Society 115:693–702. Jenkins, R.E., and N.M. Burkhead. Freshwater Fishes of Virginia. 1994. American Fisheries Society, Bethesda, MD. 1079 pp. Lukas, J.A., and D.J. Orth. 1995. Factors affecting nesting success of Smallmouth Bass in a regulated Virginia stream. Transactions of the American Fisheries Society 124: 726–735. Noltie, D.B. 1986. A method for measuring reproductive success in the Rock Bass (Ambloplites rupestris), with applicability to other substrate-brooding fishes. Journal of Freshwater Ecology 3:319–323. Noltie, D.B., and M.H.A. Keenleyside. 1986. Correlates of reproductive success in streamdwelling male Rock Bass, Ambloplites rupestris (Centrarchidae). Environmental Biology of Fishes 17:61–70. Probst, W.E., C.F. Rabeni, W.G. Covington, and R.E. Marteney. 1984. Resource use by stream-dwelling Rock Bass and Smallmouth Bass. Transactions of the American Fisheries Society 113:283–294. Schlosser, I.J. 1985. Flow regime, juvenile abundance, and the assemblage structure of stream fishes. Ecology 66:1484–1490. Schramm, H.A., Jr., S.P. Malvestuto, and W.A. Hubert. 1992. Evaluation of procedures for back-calculation of lengths of Largemouth Bass aged by otoliths. North American Journal of Fisheries Management 12:604–608. Smith, S.M., J.S. Odenkirk, and S.J. Reeser. 2005. Smallmouth Bass recruitment variability and its relation to stream discharge in three Virginia rivers. North American Journal of Fisheries Management 25:1112–1121. Surber, E.W. 1970. Smallmouth Bass stream investigations. Virginia Commission of Game and Inland Fisheries, Federal Aid in Sport Fish Restoration, Project F-14-R, Job 2-Shenandoah River Study, 1 January1964–30 June 1969. Final report. Richmond, VA. Thomas R. Payne and Associates, and the Louis Berger Group, Inc. 2009. Claytor IFN study: Stream-temperature model. Claytor Hydroelectric Project FERC No. 739-018. Available online at http://www.claytorhydro.com/documents/studyReportsDocs/ClaytorIFNStudy- StreamTemperatureModel_13Jan09.pdf. Southeastern Naturalist P. Cooper, J.R. Copeland, S.C. Lusk, and D.J. Orth 2016 Vol. 15, No. 2 268 Tyus, H.M. 1970. Spawning of Rock Bass in North Carolina in 1968. Progressive Fish- Culturist 32:32–25. Wallus, R., and T.P. Simon. 2008. Reproductive biology and early life-history of fishes in the Ohio River drainage: Elassomatidae and Centrarchidae. Vol. 6. CRC press, Boca Raton, FL. 430 pp