Under the Direction of Dr

ABSTRACT

COPE, WILLIAM ROBERT. Status, Trends, Habitat, and Genetics of the Endemic Carolina Madtom. (Under the direction of Dr. Thomas J. Kwak)

Nongame fishes contribute to diversity and important ecological functions in freshwater ecosystems, but many are imperiled, and their status and ecology are poorly understood.

Instream habitat loss and degradation are major threats among nongame species as degraded and fragmented habitat affects fish directly, but also divides species into multiple small populations, which may be at greater risk for loss of genetic variation or extirpation.

One such imperiled, nongame species is the Carolina Madtom (Noturus furiosus), a small catfish endemic to the Tar and Neuse river basins of North Carolina. Systematic surveying has shown declines in Carolina Madtom populations, and as such, the species is listed as State

Threatened. Although populations are declining, the Carolina Madtom has been sparsely studied, with only three major surveying events describing and assessing extant population status, and no genetic research has been conducted on the remaining populations.

The objectives of this research were to assess the population status, microhabitat use, and genetic structure of the Carolina Madtom to inform protective listing and management decisions for this understudied species. We snorkel surveyed for Carolina Madtoms at 75 sites in the Tar and Neuse river basins during 2016 and 2017. Microhabitat data were collected at all surveyed sites and at points-of-capture for all Carolina Madtoms. Additionally, artificial cover units were constructed and deployed at 8 sites to determine if they could be an effective passive sampling technique for Carolina Madtoms. All captured Carolina Madtoms were fin-clipped for genetic structure and diversity analyses.

We collected 59 Carolina Madtoms during snorkel surveys in the Tar River basin, whereas no fish were collected from the Neuse River basin. Occupancy modeling estimated

Carolina Madtom occupancy probability at 0.35 and detection probability at 0.81, with dominant substrate particle size affecting occupancy and mean-column water velocity affecting detection probability. Analysis of microhabitat use and available microhabitat among sites found that

Carolina Madtoms nonrandomly select microhabitat. Habitat suitability functions were developed, and we determined the most suitable ranges of microhabitat parameters for Carolina

Madtom occupancy. Comparison of available suitable habitat in the Tar and Neuse river basins determined that adequate suitable habitat was available in the Neuse River basin.

Thirty Carolina Madtoms were collected from artificial cover units at 2 sites in the Tar

River basin. Occupancy modeling estimated Carolina Madtom detection probability using artificial cover units at 0.92. Compared to other standardized survey methods, artificial cover units were found to be an efficient, passive sampling technique for detecting Carolina Madtoms.

Observations also revealed that artificial cover units were used in reproduction by Carolina

Madtoms.

Using 10 microsatellite primers developed for the related Yellowfin Madtom (Noturus flavipinnis), we successfully identified genetic structure of the Carolina Madtom. Resulting analyses quantified low genetic diversity in the species. Mean M-ratios for the Tar and Neuse river basin populations indicated that both populations have experienced demographic bottlenecks, and effective population size (Ne) estimates for the respective populations were small, indicating low genetic diversity within populations. However, multilocus population differentiation metrics G’EST and DEST were significantly different from zero indicating significant genetic variation between the Tar and Neuse river basin populations.

The application of these results may inform natural resource managers on the status of the extant populations, habitat use, and genetic structure of the Carolina Madtom and guide planning

toward informed protective listing and management decisions to maintain the viability of this important endemic species.

Status, Trends, Habitat, and Genetics of the Endemic Carolina Madtom

by William Robert Cope

A thesis submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the degree of Master of Science

Fisheries, Wildlife, and Conservation Biology

Raleigh, North Carolina

2018

APPROVED BY:

______Dr. Thomas J. Kwak Dr. Tyler R. Black Chair of Advisory Committee

______Dr. Krishna Pacifici

DEDICATION

To my parents and family. Thank you for the support, encouragement, and love — I wouldn’t be where I am without you.

BIOGRAPHY

I was born in La Crosse, Wisconsin, in 1993, but I moved to North Carolina at the age of four and have lived here ever since. My interests in aquatic fauna began long ago, as I grew up fishing with my family all over the place, in ponds, lakes, and rivers. As I grew older, my love for fish and other aquatic life only became stronger. With the aid of my father, I had numerous opportunities to explore the field. At the age of 14, I was already working with freshwater mollusks as part of a North Carolina Wildlife Resource Commission research project. From there, my experience only grew. I participated in summer research internships with Dr. Damian

Shea and Dr. Tom Kwak at North Carolina State University, assisting graduate students and fellow researchers on a variety of freshwater fish and mollusk ecology and toxicology projects while attending Apex High School.

I graduated from Apex High School in 2011 and was accepted into the Honors College at

East Carolina University. I already knew fisheries science was my passion, but my time at East

Carolina was invaluable in furthering my knowledge of aquatic systems and helping prepare me for a future in fisheries. A requirement of the Honors College was the completion of an undergraduate thesis, and with the help of Dr. Joe Luczkovich, I researched and wrote a report on the diet composition of two co-occurring, closely-related Silverside species in the Pamlico and Albemarle sounds of North Carolina. This project also gave me a chance to get a glimpse into the professional fisheries world, as I was able to present this research at an annual Tidewater

Chapter of the American Fisheries Society meeting.

I graduated from East Carolina in 2015 with a B.S. in Biology, and I accepted this M.S. position with Dr. Kwak at NC State. My time here has been an amazing experience. I have spent the past three years learning much about fisheries science and how to achieve as a fisheries

iii professional through taking classes, conducting research, and presenting research at multiple

American Fisheries Society meetings, as well as serving as President of the NC State Student

Fisheries Society.

I hope to continue my education after completing this degree and am positive my experiences at NC State have helped me become a better researcher and scientist so that I may pursue a professional career in aquatic sciences after I finish my academic career.

ACKNOWLEDGEMENTS

First, I would like to thank my advisor, Dr. Tom Kwak, for his continuous guidance and support. I have grown as a person and a biologist as a result of your teachings and encouragement over the many years of working and researching with you. I would also like to thank Dr. Krishna

Pacifici and Dr. Tyler Black who have been fantastic mentors and committee members. Thank you both for the guidance and assistance in the background research, long field work hours, methodology, and analyses to ensure we completed the most accurate research possible. All your mentorship and advice has been invaluable and has helped me become a better fisheries researcher.

I would also like to acknowledge numerous individuals who provided immense support on this project. Thanks to Dr. Eric Hallerman and his students, Shelia Harris and Caitlin Miller, at Virginia Tech University for their support with the genetic analyses and guiding me through the murky waters of conservation genetics. Thanks also to the North Carolina Wildlife Resources

Commission biologists who assisted in making this project successful. I am grateful to Chris

Wood for the insight and prior knowledge on the Carolina Madtom, ensuring I had all the necessary information to successfully survey the species and to Tom Fox and his field technicians who spent many long hours on the river with me searching for a seemingly invisible fish.

Thank you to my two awesome field technicians, Will Wood and Joseph McIver, for putting up with my nonsense for months on end and keeping a positive attitude and hardworking mentality even with the long hours, uncooperative weather, and a lack of fish that would drive anybody insane. Thanks to Ruby Valeton for the support and for keeping things moving smoothly on the administrative front, and also to Spencer Gardner, Tiffany Penland, Gus Engman, Sean Buczek,

Jennifer Archambault, and the rest of the NC State University professors, staff, graduate, and undergraduate students that helped and supported me throughout my time here at NC State. A

v special thanks to Stephen Parker and Emilee Briggs for being the best friends and colleagues a person could want and for supporting and challenging me to be the best researcher possible, while keeping me sane during the arduous process of writing a thesis — you two are the best.

I would like to thank my family for being a constant source of support, encouragement, and love. Having a support group just a short drive away has been a blessing during the trying times during research and writing and I would not be the person I am today without your guidance. I thank you all for always pushing me to do my best, be the best person possible, and encouraging me to take risks and shoot for the stars.

Finally, I would like to acknowledge the various agencies that made this project possible.

I would like to thank the North Carolina Wildlife Resources Commission for funding the project under the State Wildlife Grants program and specifically Todd Ewing for grant administration. I would also like to thank NC State University and the North Carolina Cooperative Fish and

Wildlife Research Unit for providing all the gear and supplies to make this project successful, and the U.S. Fish and Wildlife Service for their collaboration in research and help protecting this little fish that has grown near and dear to my heart.

TABLE OF CONTENTS

LIST OF TABLES ...... ix LIST OF FIGURES ...... xii CHAPTER 1: Distribution, Occupancy, and Habitat of the Carolina Madtom ...... 1 Abstract ...... 1 Introduction ...... 3 Study Area ...... 7 Field Methods ...... 9 Snorkel Survey ...... 9 Microhabitat Use, Availability, and Suitability ...... 10 Data Analysis ...... 11 Occupancy Modeling ...... 11 Habitat Analyses ...... 12 Results ...... 14 Snorkel Surveys ...... 14 Occupancy Modeling ...... 15 Microhabitat Use, Availability, and Suitability ...... 16 Discussion ...... 20 Conservation Implications ...... 25 References ...... 28 Tables ...... 37 Figures...... 47 CHAPTER 2: Evaluation of Artificial Cover Units as a Sampling Technique for Madtoms in Rivers ...... 51 Abstract ...... 51 Introduction ...... 52 Methods...... 55 Construction ...... 55 Deployment ...... 56 Monitoring ...... 56 Statistical Analyses ...... 57

vii

Results ...... 58 Discussion ...... 59 References ...... 63 Tables ...... 67 Figures...... 72 CHAPTER 3: Genetic Structure and Diversity of the Endemic Carolina Madtom and Conservation Implications ...... 74 Abstract ...... 74 Introduction ...... 76 Study Area ...... 78 Methods...... 79 Collections ...... 79 DNA Markers ...... 80 Data Analysis ...... 81 Results ...... 82 Discussion ...... 85 Conservation Implications ...... 88 References ...... 92 Tables ...... 100 Figures...... 107 APPENDIX ...... 109 Appendix A: Supporting Information for Chapter 3...... 110

viii

LIST OF TABLES

CHAPTER 1

Table 1. GPS coordinates (latitude, °N /longitude, °W, Lat/Lon) of areas surveyed for Carolina Madtoms in the Tar and Neuse river basins during the summer 2016 and 2017. Areas surveyed in 2016 were 150-m reaches with GPS coordinates starting at the downstream end of the reach. Areas surveyed in 2017 were 30-m sites along 4-km reaches with GPS coordinates starting at the downstream end of each site ...... 37

Table 2. Number of Carolina Madtoms collected during the summer 2016 and 2017. Areas surveyed in 2016 were 150-m reaches, and areas surveyed in 2017 were 30-m sites along 4-km river reaches ...... 39

Table 3. Number of observations of Carolina Madtoms, three co-occurring ictalurid fish species, and the endemic salamander, Neuse River Waterdog, associated with Carolina Madtom surveys during summer 2016 and 2017 ...... 41

Table 4. Occupancy models estimating occupancy probability (Ψ) and detection probability (p) for Carolina Madtom in the Tar River basin, including microhabitat covariates (in parentheses). Models presented are within 10% AIC weight (AICw) of the top performing model. Dominant = dominant substrate; mean velocity = mean-column water velocity; depth = water depth; subdominant = subdominant substrate; distance to cover = distance to nearest cover object; large woody = large woody debris as cover; cobble = cobble as cover; .(dot) = constant (null) ...... 42

Table 5. Carolina Madtom habitat variable rankings for model covariates influencing occupancy (Ψ) and detection (p) probabilities. Weights of all models including each respective variable were summed and averaged to calculate mean AIC weight (AICw) of each variable. Bold values indicate the variables with the highest mean AICw ...... 43

Table 6. Temporal comparison of Carolina Madtom microhabitat use between Midway et al. (2010a) and the current study ...... 44

Table 7. Comparison of the percentage by area of available microhabitat at surveyed sites, suitable for Carolina Madtom occupancy in the Tar and Neuse river basins ...... 45

Table 8. Top performing Generalized Linear Models showing microhabitat correlation to Carolina Madtom capture locations ranked by AIC values. Individual covariates in bold had significant (p < 0.05) correlation with Carolina Madtom capture locations ...... 46

CHAPTER 2

Table 1. Location of artificial cover units deployment sites in the Tar and Neuse river basins during 2016 and 2017 ...... 67

Table 2. Comparison of Carolina Madtom detection results using a snorkel survey method and artificial cover units. Site marked with an asterisk represents a detection found while supplemental snorkeling, not following standardized methods ...... 68

Table 3. Comparison of occupancy modeling results using snorkel survey methods from Wood and Nichols (2008) and Chapter 1 of this thesis and artificial cover units .... 69

Table 4. Length, weight, and microhabitat characteristics from Carolina Madtoms (N = 21) collected in artificial cover units ...... 70

Table 5. Observation of co-occurring ictalurid species and other fishes of interest captured in artificial cover units during summer 2016 and 2017 ...... 71

CHAPTER 3

Table 1. Collections of Carolina Madtoms tissue used for genetic analysis from North Carolina Museum of Natural Sciences, North Carolina Wildlife Resources Commission, and this study ...... 100

Table 2. Results of MICROCHECKER analysis of genotype frequencies at 10 microsatellite loci in Carolina Madtoms, estimated frequency of null alleles using the van Oosterhout methods ...... 101

Table 3. Metrics of genetic variation in Tar and Neuse river basin populations of Carolina Madtom: A = number of alleles per locus, HO = observed heterozygosity, HE = expected heterozygosity. All departures of HO from HE were statistically significant (p < 0.001) ...... 102

Table 4. A. Analysis of molecular variance (AMOVA) results. B. F-statistics for Carolina Madtom populations. C. Results of significance testing for F-statistics. D. Results of significance testing for population specific F-statistics ...... 103

Table 5. Matrices of two key metrics of genetic differentiation among stated subpopulations. A. G’ST (Hedrick 2005). B. DEST (Jost 2008). For each matrix, the metric is shown above the diagonal and significance (p-value) of its difference from zero below ...... 104

Table 6. Locus-specific and overall M-ratios (Garza and Williamson 2001) for Carolina Madtom populations in the Tar and Neuse river basins ...... 105

Table 7. Estimated effective population size, Ne (mean + 95% CI) for stated groups of Carolina Madtoms using the linkage disequilibrium method ...... 106

APPENDIX A

Table SI 1. Allele frequencies at 10 microsatellite loci for Carolina Madtoms ...... 111

Table SI 2. Results of STRUCTURE analysis for K = 1-15 clusters of multilocus genotypes. Each value of LnP(X/K) represents the mean of 5 independent runs ...... 118

LIST OF FIGURES

CHAPTER 1

Figure 1. Map of Carolina Madtom (CMT) survey results combined over both summers sampled (2016 and 2017). Anecdotal collections came from the mainstem Tar River and Little River ...... 47

Figure 2. Frequency distributions of Carolina Madtom microhabitat use and availability for (a) depth, (b) bottom velocity, (c) mean-column velocity, (d) dominant substrate, (e) subdominant substrate, and (f) cover type from the Tar River basin ...... 48

Figure 3. Carolina Madtom microhabitat suitability for (a) depth, (b) bottom velocity, (c) mean-column velocity, (d) dominant substrate, (e) subdominant substrate, and (f) cover type based on use and availability data collected from the Tar River basin ...... 49

Figure 4. Frequency distributions of available microhabitat in the Tar and Neuse river basins for (a) depth, (b) bottom velocity, (c) mean-column velocity, (d) dominant substrate, (e) subdominant substrate, and (f) cover type from the Tar River basin ...... 50

CHAPTER 2

Figure 1. (Top) Artificial cover unit constructed and deployed in Swift Creek in 2007 by Midway et al. (2010a). (Bottom) Same artificial cover unit recovered in Swift Creek during summer 2016 ...... 72

Figure 2. Artificial cover unit sampling results. Sites marked with stars had Carolina Madtom detections from both artificial cover units and snorkel surveys. Site marked with white diamond had no detection from artificial cover units, but did have detection from snorkel surveying ...... 73

CHAPTER 3

Figure 1. Collection sites of Carolina Madtoms for genetic analysis. Collections were from the North Carolina Museum of Natural Sciences, North Carolina Wildlife Resource Commission, and this study ...... 107

Figure 2. STRUCTURE histogram plots for K = 1-4 genotypic clusters for Carolina Madtom populations. Each vertical bar represents one individual, and each color indicates affiliation with an inferred genetic cluster of multilocus genotypes. Numbers below each bar indicate geographic collections: within the Tar River basin, 11 = Swift Creek, 12 = Fishing Creek, 13 = Little Fishing Creek, 14 = Tar River, and 15 = Town Creek; within the Neuse River basin,

xii

21 = Little River, 22= Contentnea Creek, and 23= Sandy Creek. No geographic pattern of genetic differentiation was apparent at these or higher levels of K ...... 108

xiii

CHAPTER 1

Distribution, Occupancy, and Habitat of the Carolina Madtom

Abstract

The Carolina Madtom (Noturus furiosus) is a small nongame catfish endemic to the Tar and Neuse river basins of North Carolina. Systematic surveying has shown declining occurrence and abundance in the Neuse River basin, with stable populations in the Tar River basin, and as such, the species carries a State Threatened protective listing. To assess current Carolina

Madtom population status and habitat use, we surveyed 75 sites in both basins to (1) determine the current extant range and distribution of the Carolina Madtom in the Tar and Neuse river basins, (2) develop occupancy models to estimate probability of detection and occurrence of the species throughout its range, and (3) determine Carolina Madtom current instream habitat use and suitability and compare findings to previous studies. In total, we collected 59 Carolina

Madtoms during snorkel surveys in the Tar River basin, whereas no fish were collected from the

Neuse River basin, indicating Carolina Madtom populations are decreasing in both occurrence and abundance throughout their historical and current range. Occupancy modeling estimated

Carolina Madtom occupancy probability at 0.35, and detection probability at 0.81, with dominant substrate particle size affecting occupancy and mean-column water velocity affecting detection probability. Generalized Linear Modeling of the microhabitat covariates revealed that dominant substrate and mean-column velocity were positively related to detection of Carolina Madtoms, similar to occupancy modeling results. Carolina Madtoms occupied slow-to-moderate velocity water over sand and gravel substrate, using cobble and woody debris as cover. Analysis of microhabitat use and available microhabitat among sites found that Carolina Madtoms nonrandomly select microhabitat. Habitat suitability functions were developed, and we

1 determined the most suitable ranges of microhabitat parameters for Carolina Madtom occupancy.

Comparison of available suitable habitat in the Tar and Neuse river basins determined that adequate suitable habitat was available in the Neuse river basin, suggesting that alternative factors, such as the presence of nonnative Flathead Catfish (Pylodictis olivaris) or poor water quality may be affecting the drastically declining populations in the basin. The application of these results may inform natural resource managers on the status of the Carolina Madtom and guide planning toward informed protection listing and management decisions to maintain the viability of this important endemic species.

Introduction

North America supports the greatest temperate freshwater biodiversity globally, and the southeastern United States is especially rich with freshwater fauna (Jelks et al. 2008). Sixty-two percent of United States fishes (493 species) occur in the southeast, and 91% (269 species) of the

Nation’s freshwater mussels are found in the southeastern United States (Neves et al. 1997;

Warren et al. 1997). The southeastern United States also has the greatest number of imperiled freshwater species in North America (Jelks et al. 2008). Through anthropogenic factors such as habitat degradation, fragmentation and loss, flow modification, and pollution, many southeastern fishes are experiencing population declines. Most of these imperiled fishes are nongame species.

Nongame fishes are critical to freshwater ecosystems, as they contribute to biodiversity and provide important ecological functions; however, many are greatly in need of conservation effort and are understudied and data deficient. Nongame species may be overlooked in conservation planning because conservation efforts are frequently allocated to more economically or recreationally important species, and in some systems may be a lower management priority than nonnative fishes (Cooke et al. 2005, Clarkson et al. 2005). Thus, nongame fishes remain poorly understood and minimally managed globally (Kwak et al. 2011).

One such imperiled nongame species is the Carolina Madtom (Noturus furiosus), a small

(less than 200 mm total length) catfish endemic to the Tar and Neuse river basins of North

Carolina (Burr et al. 1989). The species occurs in free-flowing streams in riffles, runs, or pools in shallow areas (Burr et al. 1989). Carolina Madtoms are benthic-associated organisms, and as such, substrate composition is an important habitat component. They most commonly occupy areas with substrates consisting of a mixture of sand or gravel, with leaf litter and small cobble included for cover (Midway et al. 2010a). The Carolina Madtom spawning season begins in

May, and reproduction may last through July (Burr et al. 1989). Carolina Madtoms seek out spawning areas with low water velocity and adequate cover to nest. Males guard nests consisting of natural cavities such as leaf litter packs, small logs, under small rocks, inside empty native unionid mussel shells or in discarded beverage containers (Burr et al. 1989; Midway et al.

2010a). Artificially constructed and deployed cover units have also been shown to enhance

Carolina Madtom detections, and they have shown preference to such artificial cover units over natural cover items (Midway et al. 2010b).

Carolina Madtoms have experienced a drastic decline in distribution since the 1960s, and are now protected with a State Threatened listing (LeGrand et al. 2008; NCNHP 2016). The

International Union for the Conservation of Nature also lists the Carolina Madtom on its Red

List of Threatened Species as near threatened after changing its classification from data deficient in 2014 (NatureServe 2014). Populations of Carolina Madtoms have been relatively stable in the

Tar River basin, but drastically declined in the Neuse River basin (Wood and Nichols 2008;

Wood and Nichols 2011). The Neuse River is generally considered a degraded river basin, and the North Carolina Department of Environmental Quality concluded that 14% of the total stream distance in the basin was imperiled according to their water quality thresholds (NCDEQ 2009).

Impacts from urban wastewater, fertilizer, industrial development, and commercial animal operations contribute to pollution and eutrophication of the system. However, Midway et al.

(2010a) found that suitable Carolina Madtom physical habitat existed in the Neuse River basin

(depth, velocity, substrate, and cover), leaving poor water quality or predation from non-native

Flathead Catfish (Pylodictis olivaris) as possible threats contributing to declines in the species.

Carolina Madtoms have not been widely studied and were until recently considered data deficient (NatureServe 2014). In the past 50 years, only three intensive sampling events have

4 been undertaken for Carolina Madtoms. In the 1960s, Smith and Bayless conducted basin-wide rotenone sampling and found the Carolina Madtom was common in both basins (Bayless and

Smith 1962; Smith and Bayless 1964). In the 1980s, Burr et al. (1989) re-sampled the Tar and

Neuse river basins and found slight declines in the species’ distributions in both basins. Wood and Nichols (2008; 2011) sampled Carolina Madtoms in 2007 and found drastic decreases in their distribution in the Neuse River, while Tar River distributions were stable, relative to those found in the 1980s. No population distribution research has occurred since 2007 on the Carolina

Madtom.

In a population status study, the proportion of sites occupied by the species of interest is vital. Field-based sampling is the optimal approach to determine population status; however, these data may be variably biased and inaccurate. Through gear bias or environmental challenges, detection of the target species is not perfect (Mackenzie et al. 2002). Such imperfect detection probabilities can bias results when researchers erroneously label an occupied site as unoccupied. Occupancy modeling produces an unbiased estimate of occupancy and can also involve related environmental covariates. In a species such as the Carolina Madtom, which lives in physical cover, imperfect detection is almost certain, and an unbiased measure of occupancy is greatly needed. Mackenzie et al. (2002) proposed a method in which imperfect detection could be incorporated into occupancy estimates so that the occupancy at the proportion of sites sampled would be more accurate than estimates coming from field data alone. Further research by Mackenzie and others has led to occupancy estimate models for many sampling scenarios

(Mackenzie et al. 2006). One such scenario is the single-species, single-season occupancy model, which allows for unbiased occupancy estimates of a species over a single sampling season.

During the sampling season, the sites must be assumed closed with respect to occupancy,

5 meaning that detection or non-detection will not be altered throughout the season by immigration or emigration (Bailey and Adams 2005). Through repeated sampling events or having multiple independent surveyors during a single sampling event, detection histories can be created and analyzed to estimate true occupancy probabilities and their influence by environmental covariates (Bailey and Adams 2005; MacKenzie and Royle 2005; MacKenzie et al. 2006). Using this approach Wood and Nichols (2008) found Carolina Madtom occupancy to be highly variable; however, probability of detecting the species was uniformly high. Variable occupancy is expected when sampling endemic, rare populations (Magoulick and Lynch 2015).

Habitat use has been studied for multiple madtoms (Noturus spp.); however, only one study extensively analyzed microhabitat use and suitability for Carolina Madtoms. Habitat use and suitability is an important factor when assessing benthic species (e.g., madtoms and darters,

Ammocrypta, Crystallaria, Etheostoma, and Percina) as substrate composition, flow, and instream cover are some of the first factors to become altered in impacted streams (Angermeier

1995; Warren et al. 1997). The Neuse and Tar river basins are both impacted basins, and as such, have experienced alterations that may have led to reductions in suitable Carolina Madtom habitat

(NCDWR 2001; NCDWR 2010). Spatial and temporal comparison of suitable instream habitat and Carolina Madtom habitat use is vital to examine the effects basin alterations may exert on freshwater fauna.

Our study was designed to provide a timely assessment on the extant population status and trends of the Carolina Madtom and its habitat. Our objectives were to (1) determine the current extant range and distribution of the Carolina Madtom in the Tar and Neuse river basins,

(2) develop occupancy models to estimate probability of detection and occurrence of the species

6 throughout its range, and (3) determine Carolina Madtom current instream habitat use and suitability and compare findings to previous studies.

Study Area

This research was conducted in the Neuse and Tar river basins of the Piedmont and

Coastal Plain physiographic provinces of North Carolina, USA. The Neuse River flows approximately 325 km through North Carolina from its headwaters originating in the Piedmont at the confluence of the Eno and Flat rivers to its mouth at Pamlico Sound near the city of New

Bern (NCDWR 2010). The basin covers an area of 10,034 km2 and spans 18 counties.

Approximately 1.7 million people currently live in the Neuse River basin, with populations expected to reach 3 million by the year 2050 (NCDWR 2010). The associated human activities in the river and over its watershed impact the habitat and water quality of the Neuse River.

Currently, 13% of the basin is considered urban, 45% forested, and 29% crop and pasture land

(NCWRC 2005). In 2007, the American Rivers Foundation listed the Neuse River as one of the ten most endangered rivers in the United States (American Rivers 2007). Non-point source pollution from agriculture and forestry has degraded water quality and habitats throughout the basin. Commercial farming inputs, such as animal waste and fertilizers, contribute 60% of the nitrates and phosphates in the system (NCWRC 2005). Due to the dense human population in the basin, many municipalities have constructed dams and withdraw water from the created impoundments for human use, affecting river flow. Habitat loss is also an issue in the Neuse

River basin, as increasing human population results in the loss of natural areas and increases in impervious surfaces (NCDWR 2010).

The Tar River runs through North Carolina from its origin in Person County to the town of Washington, where it becomes the Pamlico River and flows 65 more kilometers to its mouth at Pamlico Sound (NCDWR 2001). The basin covers 8,755 km2 and spans 16 counties (NCDWR

2001). The Tar-Pamlico basin is more rural and less impacted by human activities than the Neuse

River basin, with a human population of only 415,000. Currently, 55% percent of the basin is classified as forest and wetland, 28% crop and pasture, and 7% urban (NCDEQ 2010). The primary habitat problems affecting the basin are erosion and sedimentation due mainly to channel dredging for crop and livestock irrigation purposes (NCWRC 2005). Urbanization is a concern, but is not considered a serious cause of water degradation at present.

In 2016, we studied 11 reaches in the Tar River basin and 9 reaches in the Neuse River basin for a total of 20 reaches to encompass previously extant Carolina Madtom population distributions (Table 1). These reaches included the main waterways of each basin: the mainstem

Tar, Fishing Creek, and Swift Creek in the Tar River basin and Contentnea Creek and Little

River in the Neuse River basin. Reaches surveyed included areas of historical Madtom occurrence, recent Madtom occurrence (post-2006), and exploratory areas where Carolina

Madtoms have never been detected. Each reach was 150 m in length and followed the GPS coordinates corresponding to the same reach sampled by Wood and Nichols (2008). In 2017, we sampled 11 reaches in the Tar River basin and 5 reaches in the Neuse River basin for a total of

16 reaches (Table 1). Once again, we focused effort on the mainstem Tar River, Fishing Creek, and Swift Creek in the Tar River basin, and Contentnea Creek and Little River in the Neuse

River basin. However, due to low catch-rates of Carolina Madtoms and a low amount of river area surveyed in 2016, surveyed reaches in 2017 were areas of recent madtom occurrence

(≥2007) or areas immediately upstream or downstream of recent occurrence locations and each

8 reach was extended to 4 km in length. Three to four 30-m sites were chosen to be surveyed within each reach in areas deemed to have suitable Carolina Madtom habitat, with a random survey (30 m) added to reduce positive sampling bias (55 total sites in 2017).

Field Methods

Snorkel Survey

We surveyed Carolina Madtoms from May through October in 2016 and 2017. Each survey was conducted following a snorkeling protocol similar to that of Wood and Nichols

(2008). In 2016, surveys were conducted on 150-m reaches, with two to five snorkelers participating, and 2017 surveys were conducted at 30-m sites, with two to four snorkelers. Each snorkeler covered no more than 5 m wetted width to standardize effort and ensure thorough searching, with multiple passes being made by each snorkeler if the river width (in meters) was five times greater than the total number of snorkelers. Surveys were conducted during daylight hours (between 0800 and 1800). Each snorkeler started at the downstream portion of the reach and slowly surveyed upstream, upturning any leaf litter, woody debris, cobble, or litter present in the river to maximize detection potential. All detected Carolina Madtoms were collected by dipnet and were placed in buckets along the shore for further processing at the conclusion of the snorkel survey. Collected individuals were weighed to the nearest 0.1 gram and total length was measured in millimeters. A right-side pelvic fin clip was taken for genetic analysis (Moyer and

Williams undated). Point-of-capture was designated by a weighted marker for microhabitat use data collection. After data collection was finished each individual was returned to their point-of- capture and released alive.

Microhabitat Use, Availability, and Suitability

Microhabitat use data were collected during both the 2016 and 2017 seasons at all

Carolina Madtom occurrence locations. Depth (m), bottom velocity (m/s), mean-column velocity

(m/s), dominant substrate composition, subdominant substrate composition, distance to nearest cover, and cover type were measured at the point-of-capture of each Carolina Madtom. Depth, bottom velocity, and mean column velocity were measured using a top-set wading rod and a

Marsh-McBirney Model 2000 flow meter. Bottom velocity was measured directly on the substrate surface and mean-column velocity was measured at 60% of the total depth from the surface. Dominant and subdominant substrate compositions were determined by greatest and second greatest percentage of substrate type at the location, according to a modified Wentworth particle scale (Bovee and Milhous 1978). Distance to nearest cover was measured as the distance from point-of-capture to the nearest cover object (zero if the fish was directly associated with cover). Cover type was classified as small woody debris, large woody debris, cobble, leaf litter, artificial, and none.

Available instream microhabitat data were collected at all Carolina Madtom survey sites during both 2016 and 2017 under base-flow conditions determined by consulting USGS streamflow monitoring gauges to ensure that rivers were not influenced by precipitation events prior to survey. Instream microhabitat data were collected by sampling cross-sectional transects.

For 150-m reaches in 2016, 10 cross-sectional transects were spaced 15 m apart, starting from a random point within the first 15 m moving upstream through the reach. The same transect spacing was used in 2017; however, each 30-m site encompassed two cross-sectional transects for a total of six to eight transects per reach. Each transect was sampled at a minimum of 10 evenly spaced points, usually 1-3 m apart, depending on stream width. At each transect point we

10 measured depth, bottom velocity, mean column velocity, dominant substrate, subdominant substrate, distance to nearest cover, and cover type using the methods above.

Data Analysis

Occupancy Modeling

Snorkel survey results from 2017 were modeled to estimate Carolina Madtom occupancy and detection probabilities using a single-species, single-season occupancy model (Mackenzie et al. 2002) in the software program PRESENCE (USGS Patuxent Wildlife Research Center,

Laurel, MD; https://www.mbr-pwrc.usgs.gov/software/presence.html). Detection histories were created using spatial replication with a 4-km reach serving as a sampling location and each 30 m site within each reach serving as a repeat visit to the location. In total, 16 reaches were sampled with the number of sites ranging from three to four within each reach, resulting in 16 sampling locations with detection histories with three to four repeat visits each. Microhabitat covariates were also incorporated into models to identify microhabitat effects on Carolina Madtom occupancy and probability of detection. Covariates were categorized as either affecting occupancy or detection probability, based on assessment of which variables were most ecologically sensible for placement in each group. Due to the low number of sites surveyed and

Carolina Madtoms detected, we restricted the complexity of the models fit by investigating models with only one covariate influencing occupancy and one influencing detection. We compiled all models and ranked them following an information-theoretic approach (Akaike information criterion, AIC). Individual estimates of occupancy (Ψ) and detection (p) probabilities were provided for all models fit. All models within 10% of the AIC weight of the top-performing model were included as confidence models as described by Ruiz and Peterson

(2007). Individual covariates were compared according to their summed AIC weights across the models fit to determine which variables had the most influence on occupancy and detection probability estimates.

Habitat Analyses

Using the cross-sectional transect and point-of-capture microhabitat data collected from sites snorkel surveyed in 2016 and 2017, we analyzed the available instream microhabitat and microhabitat use data to determine current habitat use, habitat selectivity, and suitable habitat ranges for Carolina Madtoms. To determine Carolina Madtom microhabitat use, we calculated the mean depth, bottom velocity, and mean column velocity measurements of occupied microhabitats, and the mode of substrate and cover type. Habitat selectivity (or nonrandom habitat selection) was assessed using the percentage of habitat use by Carolina Madtoms compared to the available instream habitat from our surveyed sites. To determine these percentages, we developed separate frequency distributions for Carolina Madtom microhabitat use and available instream microhabitat for each variable. The two frequency distributions were then compared using a Kolmogorov-Smirnov (K-S) two-sample test for the continuous variables

(depth, bottom velocity, mean column velocity, dominant substrate, subdominant substrate) and a log-likelihood ratio G-test for independence for the categorical variable (cover type), with P- values < 0.05 (α) revealing significant differences between available instream habitat and

Carolina Madtom habitat use (i.e., non-random habitat use). Carolina Madtom habitat suitability was calculated as the proportion of habitat use divided by that for available habitat for each habitat variable range in corresponding frequency distributions. The resulting values were standardized to a maximum of 1.0, and the range of the habitat variable with a suitability value of 1.0 was deemed the most suitable. If multiple ranges had values approximate to 1.0, the entire

12 combined range was deemed most suitable. Habitat suitability analysis was based on fish locations from the Tar River, as no Carolina Madtoms were found in the Neuse River basin during snorkeling surveys.

We also compared the available habitat between the Tar and Neuse river basins to determine if the drastic decrease in Neuse River basin populations may be attributed to a lack of suitable instream physical habitat for Carolina Madtoms. All available instream habitat measurements were pooled over both seasons separately for the Tar and Neuse River basins.

Available instream habitat frequency distributions were then developed for each microhabitat variable, and the two distributions were compared using the K-S two-sample test and G-test as above. Using the ranges of most suitable Carolina Madtom habitat from our calculations, we determined if there was available habitat in the suitable ranges for Carolina Madtoms in the

Neuse River basin and compared the proportion of each basin’s available instream habitat to determine if there were differences between the amount of suitable habitat in the Tar and Neuse river basins.

We also used a Generalized Linear Model (GLM) as another method to assess microhabitat characteristics. Using the continuous microhabitat variables (depth, mean-column velocity, dominant substrate, and subdominant substrate), we developed 11 models to determine which microhabitat variables were significantly correlated to Carolina Madtom capture locations.

Each Carolina Madtom point-of-capture data point was paired with a random cross-sectional transect data point to create a set of microhabitat use and availability data. The data was then run through 11 GLMs using combinations of the microhabitat covariates listed above. Covariates from the GLMs with P-values < 0.05 (α) were significantly correlated (positive correlation for

13 positive values, negative correlation for negative values). Overall model fit was also assessed by ranking the entire GLM suite following the same AIC approach used above.

Results

Snorkel Surveys

We captured a total of 59 Carolina Madtoms in standardized snorkeling surveys during

2016–2017. During the 2016 season, we captured 15 Carolina Madtoms from May to September.

Carolina Madtoms occurred at only one of the 20 surveyed 150-m reaches. The one reach with observed specimens was Swift Creek in the Tar River basin (site Tar 8a). No individuals were found in the Neuse River basin. During the 2017 season, we captured 44 Carolina Madtoms from

May to October. Carolina Madtoms were captured at 5 of the 16 surveyed 4-km reaches, or 14 of the 55 total 30-m surveying sites (Table 2, Figure 1). Carolina Madtom were found at three reaches on Fishing Creek and two reaches on Swift Creek in the Tar River basin. Two individuals were captured from the Little River in the Neuse River basin in 2017; however, these individuals were captured during supplemental sampling and were not detected following our standardized snorkel survey method. Therefore, they were not counted in our final results and were marked as anecdotal collections (Figure 1). We also documented occurrences of other ictalurid catfishes and additional species of interest during our snorkel surveys as indicators of habitat quality and to suggest possible effects of other co-occurring species on the Carolina

Madtom (Table 3). Over the two sampling seasons, we observed 213 Margined Madtoms

(Noturus insignis) a native congeneric competitor to the Carolina Madtom. We also observed

143 Channel Catfish (Ictalurus punctatus), a non-native competitor and possible predator, and 19

Flathead Catfish, a voracious non-native predator (Pine et al. 2005; Kwak et al. 2006; Baumann

14 and Kwak 2011). Finally, we counted occurrences of Neuse River Waterdogs (Necturus lewisi), a rare aquatic salamander that is also endemic to the state of North Carolina (Braswell and

Ashton 1985; NCNHP 2016).

Occupancy Modeling

Prior to developing occupancy models, we assessed our microhabitat covariates and placed them into two groups, variables influencing occupancy and variables influencing detection. We began with a pool of 10 microhabitat covariates that could be used to either influence occupancy or detection: depth, mean column velocity, dominant substrate, subdominant substrate, distance to nearest cover, and the five individual cover type options; large woody debris, small woody debris, cobble, leaf pack, and artificial. Using information from the most commonly inhabited microhabitats by madtom species (Taylor 1969; Burr and Stoeckel

1999), we classified seven microhabitat covariates as most likely to influence Carolina Madtom occupancy at a site: depth, mean column velocity, dominant substrate, subdominant substrate, distance to nearest cover, and two cover types; large woody debris and cobble. Based on concerns over impaired visibility and sampling efficiency (Ensign et al. 1995; Thompson 2003;

Weaver et al. 2014), we classified three variables likely to influence detection of Carolina

Madtoms: depth, mean column velocity, and distance to nearest cover. Our most plausible model estimating detection probability with habitat covariates was Ψ(dominant)p(mean velocity), selected by its lowest AIC value (34.00) and highest AIC weight (0.2531) (Table 4). This model estimated Carolina Madtom occupancy with an effect of dominant substrate on occupancy and mean-column velocity effect on detection probability. The modeled occupancy probability using our top model was Ψ = 0.35 and our modeled detection probability was p = 0.81. Our naïve occupancy value was Ψ = 0.31, which is based on our survey results of 5 out of 16 reaches with

Carolina Madtom occurrences. Because of limitations in the occupancy modeling dataset and inherent uncertainty in fish modeling (Ruiz and Peterson 2007), nine other occupancy models were selected as confidence models to help determine which microhabitat covariates influence

Carolina Madtom occupancy and detection. All confidence models had an AIC weight within

10% (>0.02531) of our selected model and covariates included in the confidence models were depth, subdominant substrate, large woody debris, and cobble for occupancy, and depth and distance to nearest cover for detection. All confidence models yielded similar occupancy and detection probability estimates (Table 4). Individual covariate analysis identified both dominant substrate and mean-column velocity as the most explanatory variables affecting occupancy and detection of Carolina Madtoms (Table 5).

Microhabitat Use, Availability, and Suitability

From our 65 collections (59 standardized snorkeling surveys and 6 anecdotal Carolina

Madtom collections), we found that Carolina Madtoms occupied areas with means of water depth of 0.47 m, bottom velocity of 0.10 m/s, and mean-column velocity of 0.22 m/s (Table 6).

The most commonly used substrate was sand, and they were predominantly found under cobble cover. These findings are similar to those of Midway et al. (2010a) who originally described

Carolina Madtom microhabitat use from sampling the Tar and Neuse river basins during 2007–

2008 (Table 6). All microhabitat use and availability data came from collections in the Tar River basin in our study, as the Neuse River basin suffered from an extremely small collection sample size (n = 2), which would not be useful for suitability comparisons.

After compiling and comparing available instream habitat with habitat use of Carolina

Madtoms, we found that Carolina Madtoms nonrandomly select instream habitat. For all six habitat parameters measured (depth, bottom velocity, mean-column velocity, dominant substrate,

16 subdominant substrate, and cover type), we found that Carolina Madtoms inhabited a specific subset of the total available microhabitat in the basin. A wide range of depths were available in the Tar River basin from extremely shallow to over 1 m deep. However, Carolina Madtoms microhabitat use followed a somewhat normal distribution with peak usage between 0.30 and

0.49 m deep (Figure 2). The distribution for bottom velocity availability was positively skewed, with a majority of the velocity values being either negative (i.e., upstream flow) or near 0.00 m/s.

Carolina Madtoms occupied some of the same slow-moving areas; however, bottom velocity use was recorded up to 0.39 m/s, indicating selection for slightly faster water than the slow, stagnant velocities that dominate the available microhabitat. The distribution of available mean-column velocities followed the same pattern as bottom velocities with a distribution skewed positive with low velocities prevalent. Mean-column velocity use was approximately normal in distribution, and again Carolina Madtoms were found occupying areas with slightly faster moving water than the slow, stagnant velocities that were common throughout the surveyed reaches. Dominant substrate composition was predominantly sand and silt, with lower occurrences of gravel and cobble. Carolina Madtom most frequently occupied habitats over sand; however, gravel was used almost as frequently, and cobble was occupied less frequently. Carolina Madtoms were never observed over silt substrate, even though silt was prevalent in over 16% of the total basin area surveyed. Subdominant substrate habitat associations were similar to dominant substrate with sand and gravel over the majority of the available habitat, and Carolina Madtoms occupying substrates of sand, gravel, and cobble, with strong selection for gravel subdominant substrates.

Available cover type in the basin was equally distributed among cobble, small woody debris, and large woody debris. Artificial habitat (i.e., beverage containers or other artificial litter) and leaf packs were much less commonly available cover. Nearly 40% of the surveyed available habitat

17 in the Tar River basin had no adequate cover for Carolina Madtoms to inhabit. Carolina

Madtoms were found most frequently under cobble cover, although both large and small woody debris were used as well, and 20% of our captures came from Carolina Madtoms occupying artificial cover.

We developed Carolina Madtom habitat suitability distributions, based on our microhabitat use and availability data, to determine the most suitable ranges of microhabitat for each of the six microhabitat parameters. The most suitable range of water depth was between

0.30 and 0.49 m (Figure 3). The most suitable bottom velocity was 0.30 – 0.39 m/s, although this may be biased by collections of Carolina Madtoms (6 individuals) at a site with particularly fast velocity, as compared to relatively slow water elsewhere in the basin. Mean column velocity had multiple suitable ranges with modal peaks at 0.20 – 0.29, 0.40 – 0.49, and 0.70 – 0.79 m/s.

Again, these multiple peaks are strongly influenced by the Carolina Madtom captures at the site with particularly high water velocity. For both dominant and subdominant substrate composition, gravel and cobble were the most suitable substrate types (Figure 3). The most suitable cover type was artificial. This is because we placed constructed artificial cover units in the stream for another component of the study and treated them as existing instream cover if they were in place while we conducted our snorkel surveying. Carolina Madtom occurrence was high inside the cover units; however, artificial habitat, either placed purposefully by us or other instream litter, availability was extremely low throughout the basin, leading to high suitability. Availability of the suitable habitat ranges was generally low (Table 7). Most suitable ranges accounted for less than 15% of the available instream habitat for each parameter measured.

We also compared the instream habitat between the Tar River basin and Neuse River basin to indicate if the drastic decline in Carolina Madtom populations in the Neuse River basin

18 might be attributed to a loss of suitable instream habitat. Comparisons of the two basins showed that there were significant differences in the availability of habitat for each parameter (Figure 4).

The Tar River basin was shallower than the Neuse, as the availability in the Neuse peaked at

0.20 – 0.29 m and had higher percentages at depths of 0.70 and above. The Tar River basin reaches were also faster moving in general. Although both basins had right-skewed distributions, the Neuse River basin experienced slower velocities and exhibited lower percentages of available habitat at ranges of 0.20 m/s and above. Substrate composition was similar between the basins, with the Neuse River basin having more cobble and boulder as dominant substrate, while the Tar

River basin was composed of a sand and gravel mixture. The Tar River basin also had a higher amount of total available cover for all cover types except for artificial cover, which could be explained by the more degraded, urban nature of the Neuse River basin, which may be more susceptible to artificial litter. Suitability analysis between the two basins revealed variability in the amount of available suitable habitat. Using the ranges calculated above, the Neuse River basin generally had more habitat available in the suitable ranges for Carolina Madtoms (Table 7).

The Neuse River basin had higher percentages of available habitat in the suitable ranges for depth, bottom velocity, dominant substrate, and cover type, while the Tar River basin had higher percentages of suitable habitat for mean column velocity and subdominant substrate. As such, lack of suitable instream physical habitat for Carolina Madtoms cannot wholly explain the lack of Carolina Madtom occurrences in the Neuse River basin.

Generalized Linear Modeling results showed that dominant substrate and mean-column velocity were both positively correlated with Carolina Madtom capture locations. Of the 11 models developed, 7 contained significant variables (p < 0.05) (Table 8). The most plausible model ranked by its lowest AIC value (148.43) identified dominant substrate (p = 0.004) as the

19 most explanatory variable, meaning that Carolina Madtom capture locations tended to have larger dominant substrate sizes. In lower performing models, mean-column velocity showed significant positive correlations, meaning that once again Carolina Madtom capture locations tended to have faster mean-column velocities.

Discussion

Carolina Madtom populations are decreasing in both occurrence and abundance throughout their historical and current range. The Neuse River population may be approaching extirpation, while the Tar River population is experiencing less drastic populations losses. We found a decrease in the extant range of populations in the Tar River basin, and we captured only two individuals from a single site in the entire Neuse River basin. As well as declining in occurrence, Carolina Madtom populations have also decreased in abundance. We captured only

84 specimens (from standardized surveys, anecdotal collections, and artificial cover units) during the 2016 and 2017 field seasons, even with substantial sampling effort among 75 sites. Previous surveys have shown steady declines in Carolina Madtom populations over the past 30 years, and our research supports those findings and reveals a precipitous decline in the Neuse River basin.

Wood and Nichols (2008) found only two sites in the Neuse River basin supported Carolina

Madtoms, accounting for a 92% decrease in populations since the 1960s. Wood and Nichols

(2008) captured 208 Carolina Madtoms during the 2007 field season, and a temporally overlapping study by Midway et al. (2010a) observed 274 Carolina Madtoms over the 2007 and

2008 field seasons. Although occurrences and abundances were low, our modeling results suggest that the visual snorkel surveying technique is efficient as a means to study Carolina

Madtom populations.

Detection probability of Carolina Madtoms during snorkel surveying was quite high (p =

0.82), validating our snorkel technique as an effective method to capture Carolina Madtoms.

Even though Carolina Madtoms occur in and under cover, employing a thorough search in snorkeling surveys allowed us to effectively find and capture specimens. Wood and Nichols

(2008) also estimated high detection probabilities when snorkel surveying for Carolina Madtoms, and although no occupancy modeling results were developed, Midway et al. (2010a) found similar Carolina Madtom abundances, further validating the sampling method. In wadeable streams, snorkeling is a common and effective visual technique to sample fishes with minimal mortality. Although snorkel surveying raises concerns over sampling bias and impaired visibility

(Ensign et al. 1995; Thompson 2003; Weaver et al. 2014), studies on related madtoms and other benthic fish species have shown visual detection through snorkeling is as effective as traditional methods such as electrofishing or seining (Hankin and Reeves 1988; Persinger et al. 2004).

Snorkel surveying has also been recommended as a non-destructive method to sample rare species, as in our study with Carolina Madtom (Dunham et al. 2001; Thurow et al. 2011).

Our modeled estimate of Carolina Madtom occupancy probability was low (Ψ = 0.35).

Low occupancy with high detection demonstrates the patchy spatial distribution of the remaining

Carolina Madtoms. If an area encompassing a small patch population is sampled, it is likely that

Carolina Madtoms will be detected in nearly every survey of the sampling area. Low occupancy is expected among rare, endemic species sampling, and occupancy estimates below 0.6 are common (Collier et al. 2012; Mogoulick and Lynch 2015). Our modeling analysis also found that microhabitat covariates partially explained occupancy and detection probabilities. Our top performing model included the influence of dominant substrate on occupancy and mean-column velocity on detection. Substrate composition is an important microhabitat variable for all

21 madtoms. Most species are found in riffles with sand and gravel substrates (Taylor 1969; Burr and Stoeckel 1999), including the Carolina Madtom. As such, our site-specific occupancy estimates showed that occupancy was highest at sites with predominantly sand substrate composition. Mean-column velocity detection results were similar, confirming that madtom species are found in free-flowing streams with adequate velocity to provide oxygenated water

(Taylor 1969; Burr and Stoeckel 1999). Detection probability decreased at sites with exceptionally high velocities. High water velocities can negatively affect detection while snorkel surveying, as such fast moving water can cause problems with mask leakage, surveyor grip, or increased debris in the water column. Studies of similar benthic species have also noted difficulties sampling in areas of high water velocity (Fisher 1987; Peterson and Rabeni 2001).

Carolina Madtom microhabitat use is similar to most of its congeners, as nearly all madtom species are found in riffles or pools in free-flowing streams with adequate cover (Taylor

1969; Burr and Stoeckel 1999). Microhabitat use observed during our study is in accord with all known documented sources, including general, anecdotal, and quantitative descriptions, of

Carolina Madtom habitat occupancy (Burr et al. 1989; Wood and Nichols 2008; Midway et al.

2010a). Substrate, flow, and instream cover are important microhabitat features for madtom occupancy; however, they are among the first affected in impacted streams (Angermeier 1995).

Increased urbanization and related land development is a major cause of sedimentation, and increased fine sediment and silt load in a river can be detrimental to aquatic life (Waters 1995;

NCDEQ 2009). Suspended sediment decreases the ability of light to penetrate the water, leading to decreases in growth and biomass of primary producers and macrophytes, and impacts the visibility, respiratory, and feeding function of fishes and invertebrates (Van Nieuwenhuyse and

LaPerriere 1986; Aldridge 1987). Deposited sediment associated with land development also has

22 adverse effects on freshwater organisms. Fine silt sediment deposition fills in interstitial spaces between larger substrate particles and increases their embeddedness (Waters 1995). This not only reduces abundance and biomass of primary fish food sources, such as invertebrate larvae, but also decreases suitable habitat and areas of cover for benthic fishes. Sedimentation problems from bridge construction, mining operations, and clearcutting riparian vegetation have been documented for the Ouachita Madtom (N. lachneri; Robison and Harp 1985), Frecklebelly

Madtom (N. munitus; Piller et al. 2004), and Neosho Madtom (N. placidus; Wildhaber et al.

2000), and is a potential threat to the Carolina Madtom in the heavily impacted Tar and Neuse river basins.

Dams are also a major detriment to many freshwater species. Reduced river continuity and altered flows from dams have also been shown to negatively affect the intensively studied

Neosho Madtom (Wildhaber et al. 2000; Tiemann and Gillette 2004). There are multiple dams along both the Tar and Neuse river basins that have the potential to cause population fragmentation and impact the Carolina Madtom. Although habitat degradation is common throughout most areas inhabited by madtoms, they continue to inhabit specific microhabitats.

Whether due to specific reproductive constraints or obligate cavity nesting common to all madtom species (Burr and Stoeckel 1999), we found Carolina Madtoms nonrandomly select microhabitat. These findings support those of Midway et al. (2010a) and identify important specific stream microhabitats and reaches, especially relating to substrate, flow, and cover, that support successful madtom occurrence. These findings provide guidance for maintaining or restoring these habitats that are necessary to conserve Carolina Madtom populations.

We developed suitability functions to determine current suitable habitat parameter ranges for Carolina Madtoms. Monitoring instream flow and habitat for modeling and management

23 generally relies on either a single or multi-species suitability functions, in which suitable or preferred microhabitat ranges are determined for the target species, allowing for management of streams to maximize such suitable habitat (Bovee 1986; Annear et al. 2004). Determination of suitable habitat is especially important for madtom species, as their obligate cavity nesting and benthic association does not allow them to adapt to changing instream conditions as well as other more generalist species. Habitat use has been widely studied for almost all madtom species

(Taylor 1969; Burr and Stoeckel 1999), and habitat suitability criteria have been developed for the Freckled Madtom (N. nocturnus; Orth and Maughan 1982) and Orangefin Madtom (N. gilberti; Simonson and Neves 1992). Habitat suitability criteria are especially useful in impacted basins, such as the Neuse River basin, where flow and substrate composition can change rapidly and drastically. Midway et al. (2010a) originally determined suitable habitat ranges for Carolina

Madtoms, and the suitable habitat ranges that we developed were similar to those, indicating that microhabitat use and availability distributions had not drastically changed, even though over 850 river km of the Neuse River basin have experienced habitat change and degradation (NCDEQ

2009). This suggests that the species’ habitat affinities have remained stable where it occurs, but the area and number of sites where the species is extant have decreased over time. Because the

Neuse River basin is a highly degraded and urbanized basin (NCDEQ 2009), a logical hypothesis is that Carolina Madtom populations are declining because of the many negative effects of urbanization, land development, pollution, and deforestation, which are common drivers of population declines in other madtom species. However, two consecutive studies on Carolina

Madtoms have shown that suitable instream physical habitat exists in the both the Tar and Neuse river basins, so alternative stressors and threats that contribute to decline warrant investigation, such as predation by nonnative Flathead Catfish or poor water quality in impacted river basins.

Conservation Implications

The Carolina Madtom is listed as Threatened by the state of North Carolina and the

IUCN Red List currently lists the species as Near Threatened. An updated Species Status

Assessment from the US Fish and Wildlife Service is due to be published in 2018, and our findings on Carolina Madtom occurrence, abundance, and habitat will help inform future protective listing decisions. Reintroduction efforts are common among federally listed species as viable population recovery goals (Tear et al. 1993). However, concerns have been raised over reintroduction efforts as species are often reintroduced into habitat that retains problematic factors which caused the original population collapse (Conant 1988). However, multiple madtom species have been successfully reintroduced in the southeastern United States. The Smoky

Madtom (N. baileyi) and Yellowfin Madtom (N. flavipinnis) have both been successfully propagated in captivity, returned to stream habitat in their native ranges in Tennessee, and successful reproduction and population increases in the wild post-reintroduction have been documented (Rakes et al. 1999; Shute et al. 2005; Throneberry 2009). Such goals are unattainable for the Carolina Madtom until the drivers of population decline are identified and ameliorated.

A major threat that may be influencing Carolina Madtom populations is the presence of the nonnative Flathead Catfish. Flathead Catfish have been introduced into the Neuse River basin and more recently, the Tar River basin (Kwak et al. 2006), and are common in the mainstem and larger tributaries, where historical Carolina Madtom populations no longer exist. Our study was the first to document co-occurrence of Carolina Madtoms and Flathead Catfish at the same site.

We found both species occupying a site on the Little River in the Neuse River basin (site Neuse

12a) and a site on Fishing Creek in the Tar River basin (Tar 7a). Both of which are concerning

25 findings, as Flathead Catfish are voracious predators that have been shown to suppress native fish populations by up to 50% and positively select for ictalurid prey, including native bullhead catfish and madtom species in North Carolina rivers. (Pine et al. 2005; Pine et al. 2007;

Baumann and Kwak 2011). Future studies focusing on Flathead Catfish ranges and diet analysis may be warranted if they continue range expansion into smaller tributaries that have been previous refugia for Carolina Madtoms.

Another potential stressor influencing Carolina Madtom decline is poor water quality.

Although physical habitat degradation has been widely studied in madtom species, water quality and toxicology studies are less common. However, metals and other aquatic contaminants have been identified as potential factors leading to population decrease in the Neosho Madtom

(Wildhaber 2000; Allen 2001). Such pollutants may also be affecting the Tar and Neuse river basins and their biota, which are experiencing an increase in urbanization and land development, with four counties in the Neuse River basin expected to have population growth rates in excess of 35% by the year 2020 (NCDEQ 2009). Population growth and increased anthropogenic land use can lead to increased run-off from artificial impervious surfaces, and a greater number of point and non-point sources pollutants entering the water system (e.g., sewage, industrial effluent, and road runoff). Future research may include water quality analysis throughout both basins in relation to Carolina Madtom occurrence. Little is known of Carolina Madtom chemical sensitivity; however, given their restricted range and low fecundity (hundreds of eggs) that is common to most madtom species (Burr and Stoeckel 1999), water quality impacts are of concern. However, we observed fairly abundant numbers of another sensitive endemic species, the Neuse River Waterdog (Braswell and Ashton 1985), in reaches where the Carolina Madtom was absent, possibly signifying that water quality may not be a widespread concern.

Overall, the Carolina Madtom is an important component of stream ecosystems.

Functioning both as an important ecological link and as part of the network of endemics that make North Carolina and the southeastern United States freshwater systems biologically diverse and unique. Swift Creek and Fishing Creek within the Tar River basin have been called the most biologically diverse waterways in the state, and Swift Creek may be the most important lotic system remaining on the Atlantic Seaboard (Alderman 1993; NCNHP 1997). Conservation of these faunas and ecosystems is vital, not only to maintain a unique, endemic species, but to also maintain diversity and overall health of these waterways. The application of our results will further inform managers on the status of the Carolina Madtom and guide them toward informed protection listing and management decisions to maintain the viability of this important endemic species.

References

Alderman, J. M., A. L. Braswell, S. P. Hall, A. W. Kelly, and C. McGrath, 1993. Biological

inventory: Swift Creek subbasin. North Carolina Wildlife Resources Commission; North

Carolina Natural Heritage Program; North Carolina State Museum of Natural History.

Raleigh, North Carolina.

Aldridge, D. W., B. S. Payne, and A. C. Miller. 1987. The effects of intermittent exposure to

suspended solids and turbulence on three species of freshwater mussels. Environmental

Pollution 45:17-28.

Allen, G. T., S. H. Blackford, V. M. Tabor, and M. S. Cringan. 2001. Metals, boron, and

selenium in Neosho Madtom habitat in the Neosho River in Kansas, USA. Environmental

Monitoring and Assessment 66:1-21.

American Rivers Foundation. 2007. America’s most endangered rivers of 2007. American Rivers

Foundation, Washington, DC.

Angermeier, P. L. 1995. Ecological attributes of extinction-prone species: loss of freshwater

fishes of Virginia. Conservation Biology 9:143-158.

Annear, T., I. Chisholm, H. Beecher, A. Locke, and 12 other authors. 2004. Instream flows for

riverine resource stewardship, revised edition. Instream Flow Council, Cheyenne,

Wyoming.

Bailey, L., and M. Adams. 2005. Occupancy models to study wildlife fact sheet. United States

Geological Survey Forest and Rangeland Ecosystem. Laurel, Maryland.

Baumann J. R., and T. J. Kwak. 2011. Trophic relations of introduced Flathead Catfish in an

Atlantic river. Transactions of the American Fisheries Society 140:1120-1134.

Bayless, J. D., and W. B. Smith. 1962. Survey and classification of the Neuse river and

tributaries, North Carolina. NC Wildlife Resources Commission, Raleigh, North Carolina

Bovee, K. D., and R. T. Milhous. 1978. Hydraulic simulation in instream flow studies: theory

and technique. Instream Flow Information Paper 5, U.S. Fish and Wildlife Service 33:130

pp.

Bovee, K. D. 1986. Development and evaluation of habitat suitability criteria for use in the

instream flow incremental methodology. U.S. Fish and Wildlife Service Biological

Report 86(7).

Braswell, A. L, and R. E. Ashton Jr. 1985. Distribution, ecology, and feeding habits of Necturus

lewisi (Brimley). Brimleyana 10:13-35.

Burr, B. M., B. R. Kuhajda, W. W. Dimmick, and J. M. Grady. 1989. Distribution, biology, and

conservation status of the Carolina Madtom, Noturus furiosus, an endemic North

Carolina catfish. Brimleyana 15:57-86.

Burr, B. M., and J. N. Stoeckel. 1999. The natural history of madtoms (genus Noturus), North

America’s diminutive catfishes. Pages 51-101 in E.R. Irwin, W. A. Hubert, C. F. Rabeni,

H. L. Schramm, Jr., and T. Coon, editors. Catfish 2000: proceedings of the international

ictalurid symposium. American Fisheries Society, Symposium 24, Bethesda, Maryland.

Clarkson R. W., P. C. Marsh, S. E. Stefferud, and J. A. Steufford. 2005. Conflicts between native

fish and nonnative sport fish management in the southwestern United States. Fisheries

30:20-27.

Collier, B. A., J. E. Groce, M. L. Morrison, J. C. Newnam, A. J. Campomizzi, S. L. Farrell, H. A.

Mathewson, R. T. Snelgrove, R. J. Carroll, and R. N. Wilkins. 2012. Predicting patch

occupancy in fragmented landscapes at the rangewide scale for an endangered species: an

example of an American warbler. Diversity and Distributions 18:158-167.

Conant, S. 1988. Saving endangered species by translocation: are we tinkering with evolution?

Bioscience 38:254-257.

Cooke, S. J., C. M. Bunt, S. J. Hamilton, C. A. Jennings, M. P. Pearson, M. S. Cooperman, and

D. F. Markle. 2005. Threats, conservation strategies, and prognosis for suckers

(Catostomidae) in North America: insights from regional case studies of a diverse family

of non-game fishes. Biological Conservation 121:317-331.

Dunham, J. B., A. E. Rosenberger, R. F. Thurow, C. A. Dolloff, and P. J. Howell. 2009.

Coldwater fish in wadeable streams. Pages 1-20 in S. A. Bonar, W. A. Hubert, and D. W.

Willis, editors. Standard methods for sampling North American freshwater fishes.

American Fisheries Society. Bethesda, Maryland.

Ensign, W. E., P. L. Angermeier, and C. A. Dolloff. 1995. Use of line transect methods to

estimate abundance of benthic stream fishes. Canadian Journal of Fisheries and Aquatic

Sciences 52:213-222.

Fisher, W. L. 1987. Benthic fish sampler for use in riffle habitats. Transactions of the American

Fisheries Society. 116:768-772.

Hankin D. G., and G. H. Reeves. 1988. Estimating total fish abundance and habitat area in small

streams based on visual estimation methods. Canadian Journal of Fisheries and Aquatic

Sciences 45:834-844.

Jelks, H. L., S. J. Walsh, N. M. Burkhead, S. Contreras-Balderas, E. Diaz-Pardo, D. A.

Hendrickson, J. Lyons, N. E. Mandrak, F. McCormick, J. S. Nelson, S. P. Platania, B. A.

Porter, C. B. Renaud, J. J. Schmitter-Soto, E. B. Taylor, and M. L. Warren. 2008.

Conservation status of imperiled North American freshwater and diadromous fishes.

Fisheries 33:372-407.

Kwak, T. J., W. E. Pine III, and D. S. Waters. 2006. Age, growth, and mortality of introduced

Flathead Catfish in Atlantic rivers and a review of other populations. North American

Journal of Fisheries Management 26:73-87.

Kwak, T. J., M. T. Porath, P. H. Michaletz, and V. H. Travnichek. 2011. Catfish science: status

and trends in the 21st century. Pages 755-780 in P. H. Michaletz and V. H. Travnichek,

editors. Conservation, ecology, and management of catfish: the second international

symposium, American Fisheries Society Symposium 77, Bethesda, Maryland.

LeGrand, H. E., S. E. McRae, S. P. Hall, and J. T. Finnegan. 2008. Natural heritage program list

of the rare animal species of North Carolina. North Carolina Natural Heritage Program,

Division of Natural Resource Planning and Conservation, North Carolina Department of

Environment and Natural Resources. Raleigh, North Carolina.

MacKenzie, D. I., J. D. Nichols, G. B. Lachman, S. Droege, J. A. Royle, and C. A. Langtimm.

2002. Estimating site occupancy rates when detection probabilities are less than one.

Ecology 83:2248-2255.

MacKenzie, D. I., and J. A. Royle. 2005. Designing occupancy studies: general advice and

allocating survey effort. Journal of Applied Ecology 42:1105-1114.

MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 2006.

Occupancy estimation and modeling: inferring patterns and dynamics of species

occurrence. Academic Press. Burlington, Massachusetts.

Magoulick, D. D., and D. T. Lynch. 2015. Occupancy and abundance modeling of the

endangered Yellowcheek Darter in Arkansas. Copeia 103:433-439.

Midway, S. R., T. J. Kwak, and D. D. Aday. 2010a. Habitat suitability of the Carolina Madtom,

an imperiled, endemic stream fish. Transactions of the American Fisheries Society

139:325-338.

Midway, S. R., D. D. Aday, T. J. Kwak, and K. Gross. 2010b. Cover preference of the Carolina

Madtom (Noturus furiosus), an imperiled endemic southeastern stream fish. Journal of

Freshwater Ecology 25:151-154.

Moyer, G. R., and A. Williams. Undated. Warm springs conservation genetics lab standard

operating procedure: collecting tissue samples for genetic analysis. Warm Springs

Conservation Genetics Lab. Warm Springs, Georgia.

NatureServe. 2014. Noturus furiosus. The IUCN Red List of Threatened Species 2014:

e.T14901A19033935. Downloaded on 10 May 2016.

NCDEQ (North Carolina Department of Environmental Quality). 2009. Neuse River basinwide

water quality plan. NC Department of Environmental Quality. Raleigh, North Carolina.

NCDEQ (North Carolina Department of Environmental Quality). 2010. Tar-Pamlico River Basin

NCDWR (North Carolina Division of Water Resources). 2001. State water supply plan. NC

Department of Environment and Natural Resources. Raleigh, North Carolina.

NCDWR (North Carolina Division of Water Resources). 2010. Neuse River basin water

resources plan. NC Department of Environment and Natural Resources. Raleigh, North

Carolina.

NCNHP (North Carolina Natural Heritage Program). 2016. Natural Heritage Program list of rare

animal species of North Carolina 2016. Raleigh, North Carolina

NCNHP (North Carolina Natural Heritage Program). 1997. The upper Tar River basin: Swift

Creek and Fishing Creek subbasins. Biological conservation database. Raleigh, North

Carolina.

NCWRC (North Carolina Wildlife Resources Commission). 2005. North Carolina wildlife action

plan. Raleigh, North Carolina.

Neves, R. J., A. E. Bogan, J. D. Williams, S. A. Ahlstedt, and P. W. Hartfield. 1997. Status of

aquatic mollusks in the southeastern United States: a downward spiral of diversity. Pages

43-85 in G. W. Benz, and D. E. Collins, editors. Aquatic fauna in peril: the southeastern

perspective. Southeast Aquatic Research Institute, Lenz Design and Communications,

Decatur, Georgia.

Orth, D. J., and O. E. Maughan. 1982. Evaluation of the incremental methodology or

recommending instream flows for fishes. Transactions of the American Fisheries Society

111:413-445.

Persinger, J. W., D. J. Orth, and T. J. Newcomb. 2004. A comparison of snorkeling versus

throwable anode electrofishing for evaluating stream fish habitat use. Journal of

Freshwater Ecology 19:547-557.

Peterson, J. T., and C. F. Rabeni. 2001. Evaluating the efficiency of a one-square-meter quadrat

sampler for riffle-dwelling fish. North American Journal of Fisheries Management 21:76-

85.

Piller, K. R., H. L. Bart, Jr., and J. A. Tipton. 2004. Decline of the Frecklebelly Madtom in the

Pearl River based on contemporary and historical surveys. Transactions of the American

Fisheries Society 133:1004-1013.

Pine, W. E, III, T. J. Kwak, D. S. Waters, and J. A. Rice. 2005. Diet selectivity of introduced

Flathead Catfish in coastal rivers. Transactions of the American Fisheries Society 134:

901-909.

Pine, W. E, III, T. J. Kwak, and J. A. Rice. 2007. Modeling management scenarios and the

effects of an introduced apex predator on a coastal riverine fish community. Transactions

of the American Fisheries Society 136:105-120.

Rakes, P. L., J. R. Shute, and P. W. Shute. 1999. Reproductive behavior, captive breeding, and

restoration ecology of endangered fishes. Environmental Biology of Fishes 55:31-42.

Robison, H. W., and G. L. Harp. 1985. Distribution, habitat and food of the Ouachita Madtom,

Noturus lachneri, a Ouachita River drainage endemic. Copeia 1985:216-220.

Ruiz, J. C., and J. T. Peterson. 2007. An evaluation of the relative influence of spatial, statistical,

and biological factors on the accuracy of stream fish species presence models.

Transactions of the American Fisheries Society 136:1640-1653.

Shute, J. R., P. L. Rakes, and P. W. Shute. 2005. Reintroduction of four imperiled fishes in

Abrams Creek, Tennessee. Southeastern Naturalist 4:93-110

Simonson, T. D., and R. J. Neves. 1992. Habitat suitability and reproductive traits of the

Orangefin Madtom Noturus gilberti (Pisces: Ictaluridae). American Midland Naturalist

127:115-124.

Smith, W. B., and J. D. Bayless. 1964. Survey and classification of the Tar River and tributaries,

North Carolina. NC Wildlife Resources Commission. Raleigh, North Carolina.

Taylor, R. T. 1969. A revision of the catfish genus Noturus (Rafinesque), with an analysis of

higher groups in the Ictaluridae. Smithsonian Institution Press, Washington DC.

Tear, T. H., J. M. Scott, P. H. Hayward, and B. Griffith. 1993. Status and prospects for success

of the endangered species act: a look at recovery plans. Science 262:976-977.

Thompson W. L. 2003. Hankin and Reeves’ approach to estimating fish abundance in small

streams: limitations and alternatives. Transactions of the American Fisheries Society

132:69-75.

Throneberry, J. K. Reintroduction success of Smoky Madtom Noturus baileyi and Yellowfin

Madtom Noturus flavipinnis in Abrams Creek, Great Smoky Mountains National Park.

Master’s thesis. Tennessee Tech University. Cookeville, Tennessee.

Thurow, R. F., C. A. Dolloff, and J. E. Marsden. 2012. Visual observation of fishes and aquatic

habitats. Pages 781-817 in A. V. Zale, D. L. Parrish, and T. M. Sutton, editors. Fisheries

techniques, 3rd edition. American Fisheries Society. Bethesda, Maryland.

Tiemann, J. S., and D. P. Gillette. 2004. Effects of lowhead dams on riffle-dwelling fishes and

macroinvertebrates in a midwestern river. Transactions of the American Fisheries Society

133:705-717.

Van Nieuwenhuyse, E. E., and J. D. LaPerriere. 1986. Effects of placer gold mining on primary

production in subarctic streams of Alaska. Water Resources Bulletin 22:91-99.

Warren, M. L., P. L. Angermeier, B. M. Burr, and W. R. Haag. 1997. Decline of a diverse fish

fauna: patterns of imperilment and protection in the southeastern United States. Pages

105-164 in G. W. Benz, and D. E. Collins, editors. Aquatic fauna in peril: the

southeastern perspective. Southeast Aquatic Research Institute, Lenz Design and

Communications, Decatur, Georgia.

Waters, T. F. 1995. Sediment in streams: sources, biological effects, and control. American

Fisheries Society Monograph 7, Bethesda, Maryland.

Weaver, D. M., T. J. Kwak, and K. H. Pollock. 2014. Sampling characteristics and calibration of

snorkel counts to estimate stream fish populations. North American Journal of Fisheries

Management 34:1159-1166.

Wildhaber, M. L., V. M. Tabor, D. W. 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.

Wood, C. J., and R. B. Nichols. 2008. Status assessment of the Carolina Madtom final report.

North Carolina Wildlife Resources Commission. Raleigh, North Carolina.

Wood, C. J., and R. B. Nichols. 2011. Status assessment of the Carolina Madtom: a rare North

Carolina endemic. Pages 295-303 in P.H. Michaletz and V.H. Travnichek, editors.

Conservation, ecology, and management of catfish: the second international symposium,

American Fisheries Society Symposium 77, Bethesda, Maryland.

2016 Lat/Lon 2017 Lat/Lon Basin and Site Waterway Reach Site 1 Site 2 Site 3 Site 4 Tar 1 Conetoe Creek 35.6967 - 77.4884 - - - - Tar 2 Deep Creek 35.9475 - 77.4758 - - - - Tar 3 Town Creek 35.7919 - 77.5583 - - - - Tar 4 Beech Swamp 36.0886 - 77.5202 - - - - Tar 5 Fishing Creek 35.9739 - 77.5404 35.9891 - 77.5447 35.9849 -77.5454 35.9804 - 77.5408 35.9735 - 77.5404 Tar 6 Swift Creek - 35.9659 - 77.5861 35.955 - 77.5914 35.9544 - 77.5895 - Tar 7a Fishing Creek - 36.1282 - 77.6337 36.1137 - 77.6260 36.1148 - 77.6258 - Tar 7b Fishing Creek - 36.1375 - 77.6471 36.1337 - 77.6410 36.1312 - 77.6366 - Tar 8a Swift Creek 36.0742 - 77.8697 36.0734 - 77.8683 36.0682 - 77.8626 36.0674 - 77.8539 36.0680 - 77.8482 Tar 8b Swift Creek - 36.1246 - 77.9525 36.1293 - 77.9491 36.1227 - 77.9404 36.1117 - 77.9203 Tar 9a Little Fishing Creek 36.1861 - 77.8758 36.2030 - 77.8704 36.1869 - 77.8759 36.1826 - 77.8772 36.1777 - 77.8801 Tar 9b Little Fishing Creek - 36.1768 - 77.8871 36.1721 - 77.8912 36.1670 - 77.8862 36.1631 - 77.8837 Tar 9c Little Fishing Creek - 36.2573 - 77.8872 36.2575 - 77.8818 36.2414 - 77.8920 36.2362 - 77.8910 Tar 10 Tar River - 35.9340 - 78.1474 35.9264 - 78.1466 35.9191 - 78.1417 - Tar 11a Fishing Creek 36.1695 - 77.9230 36.1693 - 77.9230 36.1641 - 77.9172 36.1637 - 77.9033 - Tar 11b Fishing Creek - 36.1979 - 77.9971 36.1876 - 77.9936 36.1859 - 77.9904 - Tar 12 Tar River 35.9354 - 78.1477 - - - - Tar 13 Shocco Creek 36.2106 - 78.1051 - - - - Tar 15 Tar River 36.1752 - 78.4964 - - - - Neuse 5 Toisnot Swamp 35.6120 - 77.8048 - - - - Neuse 7 Contentnea Creek 35.6101 - 77.8649 - - - -

Table 1 Continued.

2016 Lat/Lon 2017 Lat/Lon Basin and Site Waterway Reach Site 1 Site 2 Site 3 Site 4 Neuse 9 Contentnea Creek 35.6977 - 78.0611 35.6978 - 78.0610 35.6967 - 78.0598 35.6918 - 78.0503 - Neuse 10 Mill Creek 35.3451 - 78.1841 - - - - Neuse 11 Buffalo Creek 35.5883 - 78.2117 - - - - Neuse 12a Little River 35.6103 - 78.2136 35.6085 - 78.2105 36.6028 - 78.2013 35.6003 - 78.1974 35.5945 - 78.1968 Neuse 12b Little River - 35.5611 - 78.1596 35.5522 - 78.1635 35.5518 - 78.1626 - Neuse 12c Little River - 35.6635 - 78.2559 35.5904 - 78.1788 35.5884 - 78.1765 - Neuse 13 Swift Creek 35.5187 - 78.3817 - - - - Neuse 14 Middle Creek 35.5078 - 78.4017 - - - - Neuse 15 Eno River 36.0721 - 78.9086 - - - -

Table 2. Number of Carolina Madtoms collected during the summer 2016 and 2017. Areas surveyed in 2016 were 150-m reaches, and areas surveyed in 2017 were 30-m sites along 4-km river reaches.

2016 Collections 2017 Collections Location Reach Site 1 Site 2 Site 3 Site 4 Tar 1 0 - - - - Tar 2 0 - - - - Tar 3 0 - - - - Tar 4 0 - - - - Tar 5 0 0 0 0 0 Tar 6 0 0 0 0 - Tar 7a 0 2 0 4 - Tar 7b - 2 3 6 - Tar 8a 15 3 9 2 2 Tar 8b - 1 1 3 4 Tar 9a 0 0 0 0 0 Tar 9b - 0 0 0 0 Tar 9c - 0 0 0 0 Tar 10 0 0 0 0 - Tar 11a 0 0 2 0 - Tar 11b - 0 0 0 - Tar 12 0 - - - - Tar 13 0 - - - - Tar 15 0 - - - - Neuse 5 0 - - - - Neuse 7 0 - - - - Neuse 9 0 0 0 0 - Neuse 10 0 - - - - Neuse 11 0 - - - - Neuse 12a 0 0 0 0 0 Neuse 12b 0 0 0 0 -

Table 2 Continued.

2016 Collections 2017 Collections Location Reach Site 1 Site 2 Site 3 Site 4 Neuse 12c 0 0 0 0 - Neuse 13 0 - - - - Neuse 14 0 - - - - Neuse 15 0 - - - -

2016 2017

Species No. Observed No. of Sites No. Observed No. of Sites Carolina Madtom 15 1 44 14 Margined Madtom 60 9 153 15 Channel Catfish 49 11 94 7 Flathead Catfish 12 1 7 4 Neuse River Waterdog 29 8 61 12

Model AIC ΔAIC AICw Ψ(dominant),p(mean velocity) 34.00 0.00 0.2531 Ψ(dominant),p(depth) 36.34 2.34 0.0786 Ψ(subdominant),p(mean velocity) 36.42 2.42 0.0755 Ψ(dominant),p(distance to cover) 36.75 2.75 0.0640 Ψ(.),p(mean velocity) 37.03 3.03 0.0556 Ψ(depth),p(mean velocity) 37.25 3.25 0.0498 Ψ(dominant),p(.) 37.36 3.36 0.0472 Ψ(large woody),p(mean velocity) 37.39 3.39 0.0465 Ψ(cobble),p(mean velocity) 37.55 3.55 0.0429

Microhabitat Covariate AICw (Ψ) AICw (p) Depth 0.0232 0.0247 Mean Column Velocity 0.0104 0.0777 Dominant Substrate 0.1107 - Subdominant Substrate 0.0338 - Distance to Nearest Cover 0.0118 0.0227 Large Woody Debris 0.0230 - Cobble 0.0186 -

Table 6. Temporal comparison of Carolina Madtom microhabitat use between Midway et al. (2010a) and the current study.

Midway et al. 2010a This Study Variable Mean/Mode 95% Confidence Interval Range Mean/Mode 95% Confidence Interval Range Depth (m) 0.42 0.39 - 0.45 0.01 - 0.92 0.47 0.42 - 0.52 0.18 - 1.00 Bottom Velocity (m/s) 0.14 0.12 - 0.16 0.00 - 0.43 0.10 0.08 - 0.12 -0.03 - 0.39 Mean-Column Velocity (m/s) 0.22 0.20 - 0.24 0.00 - 0.58 0.22 0.18 - 0.26 -0.02 - 0.77 Substrate Sand - - Sand - - Cover Type Cobble - - Cobble - -

Table 7. Comparison of the percentage by area of available microhabitat at surveyed sites, suitable for Carolina Madtom occupancy in the Tar and Neuse river basins.

Tar Neuse Variable Suitable Range % Available % Available Depth 0.30 - 0.49 m 22.22 23.80 Bottom Velocity 0.30 - 0.39 m/s 1.97 2.71 Mean Column Velocity 0.20 - 0.29 m/s 13.25 10.96 0.40 - 0.49 m/s 3.50 2.91

0.70 - 0.79 m/s 0.70 0.42

Dominant Substrate Cobble 3.24 7.91 Subdominant Substrate Gravel 35.43 33.66 Cover Type Artificial 0.39 0.97

Table 8. Top performing Generalized Linear Models showing microhabitat correlation to Carolina Madtom capture locations ranked by AIC values. Individual covariates in bold had significant (p < 0.05) positive or negative correlation with Carolina Madtom capture locations.

Model Parameters AIC Dominant Substrate 148.43 Mean Velocity + Dominant Substrate 148.85 Dominant Substrate + Subdominant Substrate 149.37 Depth + Dominant Substrate 149.91 Mean Velocity + Subdominant Substrate 160.94 Mean Velocity 162.44 Depth + Mean Velocity 163.84 Depth + Subdominant Substrate 165.00 Subdominant Substrate 165.08 Null 165.58 Depth 166.31

Figures

Figure 1. Map of Carolina Madtom (CMT) survey results combined over both summers sampled (2016 and 2017). Anecdotal collections came from the mainstem Tar River and Little River.

0.6 (a) Depth 0.6 (b) Bottom Velocity Use Use 0.5 Availability 0.5 Availability D = 0.270 D = 0.299 0.4 P = < 0.001 0.4 P = < 0.001 0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0 0.10 0.30 0.50 0.70 0.90 1.10 0.00 0.20 0.40 0.60 0.80 Depth (m) Bottom Velocity (m/s)

0.6 (c) Mean-Column Velocity 0.8 (d) Dominant Substrate Use Use 0.5 Availability 0.7 Availability D = 0.342

D = 0.355 0.6 0.4 P = < 0.001 P = < 0.001 0.5 0.3 0.4

0.2 0.3 Proportion 0.2 0.1 0.1 0 0.00 0.20 0.40 0.60 0.80 0.0 Silt Sand Gravel Cobble Boulder Mean-Column Velocity (m/s) Dominant Substrate

0.8 (e) Subdominant Substrate 0.8 (f) Cover Type Use Use 0.7 Availability 0.7 Availability 0.6 D = 0.356 0.6 G = 120.75 P = < 0.001 0.5 P = < 0.001 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 0.0 Artificial Leaf Cobble Small Large None Silt Sand Gravel Cobble Boulder Pack Woody Woody Subdominant Substrate Cover Type

1.0 (a) Depth 1.0 (b) Bottom Velocity

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0 0.10 0.30 0.50 0.70 0.90 1.10 0.00 0.20 0.40 0.60 0.80 Depth (m) Bottom Velocity (m/s)

0.8 0.8

0.6 0.6

0.4 0.4 Suitability 0.2 0.2

0.0 0.0 0.00 0.20 0.40 0.60 0.80 Silt Sand Gravel Cobble Boulder Mean-Column Velocity (m/s) Dominant Substrate

(e) Subdominant Substrate 1.0 1.0 (f) Cover Type

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0 Silt Sand Gravel Cobble Boulder Artificial Leaf Cobble Small Large Subdominant Substrate Pack Woody Woody Cover Type

Figure 3. Carolina Madtom microhabitat suitability for (a) depth, (b) bottom velocity, (c) mean- column velocity, (d) dominant substrate, (e) subdominant substrate, and (f) cover type based on use and availability data collected from the Tar River basin.

0.6 (a) Depth 0.6 (b) Bottom Velocity Tar Availability Tar Availability 0.5 Neuse Availability 0.5 Neuse Availability D = 0.064 D = 0.064 0.4 P = 0.001 0.4 P = 0.001 0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0 0.10 0.30 0.50 0.70 0.90 1.10 0.00 0.20 0.40 0.60 0.80 Depth (m) Bottom Velocity (m/s)

0.6 (c) Mean-Column Velocity 0.8 (d) Dominant Substrate Tar Availability Tar Availability 0.7 0.5 Neuse Availability Neuse Availability D = 0.061 D = 0.058 0.6 0.4 P = 0.003 P = 0.005 0.5 0.3 0.4 0.3 0.2 Proportion 0.2 0.1 0.1 0.0 0.0 0.00 0.20 0.40 0.60 0.80 Silt Sand Gravel Cobble Boulder Mean-Column Velocity (m/s) Dominant Substrate

0.8 (e) Subdominant Substrate 0.8 (f) Cover Type Tar Availability Tar Availability 0.7 0.7 Neuse Availability Neuse Availability G = 18.123 0.6 D = 0.054 0.6 P = 0.002 0.5 P = 0.011 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 0.0 Silt Sand Gravel Cobble Boulder Artificial Leaf Cobble Large Small None Pack Woody Woody Subdominant Substrate Cover Type Figure 4. Frequency distributions of available microhabitat in the Tar and Neuse river basins for (a) depth, (b) bottom velocity, (c) mean-column velocity, (d) dominant substrate, (e) subdominant substrate, and (f) cover type from the Tar River basin.

CHAPTER 2

Evaluation of Artificial Cover Units as a Sampling Technique for Madtoms in Rivers

Abstract

Instream habitat loss and degradation is a major threat to freshwater fishes, and is a problem specifically among nongame species, due to a lack of research and knowledge concerning these species’ habitat requirements. Instream cover is an important component of instream habitat requirements, especially in benthic species that require cover for reproduction and shelter from larger predators. One such benthic species is the Carolina Madtom (Noturus furiosus), a small nongame catfish endemic to the Tar and Neuse river basins of North Carolina.

To augment understanding of instream cover dynamics, we deployed artificial cover units in both basins to determine if artificial cover units could be an effective passive sampling technique for increasing accuracy of detection and abundance of Carolina Madtoms. Artificial cover units were deployed at 8 sites in the Tar and Neuse river basins, and 30 Carolina Madtoms were collected from artificial cover units at 2 sites in the Tar river basin. Occupancy modeling estimated

Carolina Madtom detection probability using artificial cover units at 0.92. Compared to other standardized survey methods, artificial cover units were found to be an efficient, passive sampling technique for detecting Carolina Madtoms. Observations also revealed that artificial cover units were used in reproduction and nesting for Carolina Madtoms. The application of these results allows natural resource managers an improved opportunity to assess and conserve this important, threatened species through an inexpensive, efficient, passive sampling device that can provide valuable spawning habitat, protection from predators, and help mitigate the negative effects of instream habitat degradation.

Introduction

Quality instream habitat is essential for biodiversity and aquatic organism survival in lotic ecosystems. Degradation and loss of stream habitat has become a common stressor affecting biota in regions with high freshwater diversity, such as the southeastern United States

(Angermeier 1995; Jelks et al. 2008). Such habitat loss directly influences declines in freshwater populations and community diversity (Ricciardi and Rasmussen 1999). Instream habitat relations and population declines due to habitat degradation are well documented, yet many river systems lack habitat-based management plans. In systems where stream habitat is managed, conservation efforts are typically allocated to more economically or recreationally important species, leaving nongame species habitat requirements understudied and poorly understood (Cooke et al. 2005).

Furthermore, nongame and sportfish species often have different habitat affinities and requirements, potentially harming nongame stream fish when habitat is strictly managed for sportfish species (Aadland 1993; Kwak and Freeman 2010). Therefore, habitat management for nongame species remains a vital conservation objective to maintain stream system biodiversity.

Instream habitat can be defined as the place where fish can find the physical and chemical features required for life (Maddock 1999; Orth and White 1999). An important feature of stream habitat is available instream cover, which is any physical object or formation that offers concealment or visual isolation (Orth and White 1999). Cover is a critically important element of stream habitat for many species, as instream cover allows fishes a location to spawn and rear eggs, as well as a location to be sheltered from predators (Mann 1996; Mackenzie and

Greenberg 1998). Benthic species, in particular, rely on instream cover, as their association with the substrate and related cover provides them an ecological niche distinct from other midwater species (Rahel and Stein 1988; Magoulick 2004).

Study and application of stream habitat and instream cover relations is important in conservation of nongame and benthic species, as habitat degradation is a primary threat to these populations (Jelks et al. 2008). Angermeier (1995) found that benthic and endemic species were among the most impacted by habitat degradation. For example, over 25% of fishes in the benthic dwelling families Percidae and Ictaluridae are classified as jeopardized (Piller et al. 2004).

Especially important members of the Ictaluridae family are the small, nongame species of madtoms (Noturus spp.). Most members of the genus are both benthic and endemic to restricted distributions, leaving them especially vulnerable to habitat degradation and imperilment.

Robison and Harp (1985) found that anthropogenic habitat degradation had negatively affected populations of the Ouachita Madtoms (Noturus lachneri), while Etnier and Starnes (1991) concluded that benthic species such as madtoms and darters (Ammocrypta, Crystallaria,

Etheostoma, Percina) were the most jeopardized fish families due to loss of available instream cover for nesting and reproduction. One approach to mitigating instream habitat loss to conserve these species is through introduction of artificial cover or augmentation of existing instream cover of natural materials. Introduction of artificial cover has been documented as beneficial to salmonids and other sportfish (Moring and Nicholson 1994; Eklöv and Greenberg 1998;

Heggenes and Traaen 1988), and studies of benthic-dwelling nongame darters have demonstrated that artificial cavities were suitable structures for nesting and spawning activities (Lindquist et al.

1984; Piller and Burr 1999).

The Carolina Madtom (Noturus furiosus) is a small, nongame, cover-associated catfish that is endemic to only two river basins in North Carolina. Due to recent population declines, the

Carolina Madtom is currently listed as State Threatened (LeGrand et al. 2008; NCNHP 2016).

Suitable cover is required for Carolina Madtom spawning and parental care activities, and

53 cobble, leaf packs, woody debris, and native unionid mussel shells are used as cover items (Burr et al. 1989; Midway 2010a). Carolina Madtoms will also occupy artificial habitats for nesting.

Burr et al. (1989) found that Carolina Madtoms inhabited 355-ml aluminum cans, and Wood and

Nichols (2008) collected individuals from aluminum cans and glass beverage containers.

Midway et al. (2010a) conducted an extensive before-after-control-impact study using artificial cover units to determine if they increased abundance of Carolina Madtoms in river reaches and found that abundance was significantly higher in experimental reaches augmented with artificial cover units, compared to control reaches. Midway et al. (2010b) also conducted laboratory cover preference trials in which Carolina Madtoms were placed in aquaria and presented with three natural cover objects and one artificial cover unit. Carolina Madtoms selected the artificial cover unit in 63% of the experimental trials, indicating clear preference of artificial cover over natural cobble, leaf pack, and mussel shell cover.

Carolina Madtoms preferentially occupy artificial cover units in the laboratory, and they provide suitable cover habitat in streams (Midway et al. 2010a, 2010b); thus, they have the potential to serve as sampling technique for the species. The objective of this research was to assess artificial cover units as an effective passive sampling technique for detection of Carolina

Madtoms with a minimal amount of field sampling effort relative to active sampling techniques

(e.g., snorkeling or electrofishing). The Tar and Neuse river basins have both experienced effects of urbanization and sedimentation (NCDEQ 2009, NCDWR 2010), and Carolina Madtoms have undergone drastic population declines (Chapter 1, this thesis). This evaluation may also serve as a comparison of findings by Midway et al. (2010a), who deployed these same cover units during a period (2007-2008) when the Carolina Madtom was more widespread and abundant in the

Neuse and Tar river basins. If Carolina Madtom occupancy of artificial cover units remains high

54 at current reduced population levels, they may provide managers a simple and affordable method of sampling, monitoring, and habitat enhancement to help conserve the species.

Methods

We constructed artificial cover units, following the design of Midway et al. (2010a,

2010b)(Figure 1), deployed them in a standardized protocol at 8 stream sites in the range of the

Carolina Madtom, and then monitored their occupancy by the Carolina Madtom and other aquatic species.

Construction

Artificial cover units were built to specifications designed by Midway et al. (2010a,

2010b). Due to lack of availability of their exact original building materials, artificial cover units were constructed using materials similar to the original design. The bottom 133 mm was removed from a 150-mm diameter terra cotta clay pot using a Dremel Model 4000 rotary cutting tool with a diamond grit cutting wheel to be used as the cover enclosure. An approximately 19 by

32 mm “mouse hole” shaped entryway and three 10 mm vent slots were cut out of the enclosure.

The enclosure was then glued to an inverted 133-mm diameter planter saucer, and landscaping river rocks, approximately 20-30 cm diameter, were glued into the concave portion of the planter saucer to serve as a weight and provide stability for the unit (Figure 1). To complete the unit, a rubber stopper was inserted into the drain hole of the terra cotta enclosure and cut off flush with the top of the unit. The stopper was then glued into place from the outside, as to not introduce any unnecessary chemicals from gluing on the inside of the cover unit.

Deployment

We selected 8 sites based on surveying efforts of Wood and Nichols (2011), Midway et al. (2010), and Chapter 1 (this thesis). Six sites were chosen in the Tar River basin and two sites were chosen in the Neuse River basin (Table 1; Figure 2). All 8 sites had Carolina Madtom occurrences in 2007 confirmed through snorkel surveying efforts (Wood and Nichols 2011).

Sites in the Tar River basin covered most of the extant range of the animal, with multiple sites along Fishing, Swift, and Little Fishing creeks. Sites in the Neuse River basin covered the remaining extant range of the Carolina Madtom in the basin, Contentnea Creek and Little River.

At each site, 150-m reaches were delineated for snorkel surveying efforts. Twelve artificial cover units were deployed along the 150-m reach in areas deemed quality habitat in a cross-sectional grid pattern as to include all types of habitat across the width of the stream in the cover analysis.

Cover units were deployed with vent slots facing upstream and the entryway facing downstream as to not fill the unit with debris or sediment washing downstream. Cover units were deployed from July through September 2016 and June through November 2017 and were checked between

3 to 6 times per season.

Monitoring

Artificial cover units were deployed for a 14-day soak period, then were checked for occupancy by Carolina Madtoms and associated benthic, cover-oriented species. During each observation event, artificial habitats were approached from downstream by a snorkeling observer with a dipnet. The observer then gently grasped the cover unit with a finger over the entryway to confine any occupant, lifted the cover unit out of the water, and emptied the contents of the unit into the dipnet. Area beneath the cover unit was also observed for potential specimens that occurred under the units rather than inside them. The cover units were then re-deployed in the

56 same location and were left for another 14-day soak period. Collected madtoms were weighed to the nearest 0.1 gram, and total length was measured in millimeters. A right-side pelvic fin clip was also taken for genetic analysis. After data collection was completed, each madtom was returned to the stream nearest the cover unit they occupied and released alive. Microhabitat characteristics were measured at all Carolina Madtom points-of-capture. Depth (m), bottom velocity (m/s), mean-column velocity (m/s), dominant substrate composition, and subdominant substrate composition were measured at the point directly beside the occupied cover unit.

Statistical Analyses

Means were calculated for Carolina Madtom fish length and weight and for depth, bottom velocity, mean-column velocity at capture locations, and modes were identified for dominant and subdominant substrate. To determine efficiency of artificial cover units as a sampling technique, we fit a single-season-single-species occupancy model (Mackenzie et al. 2002) to determine occupancy (Ψ) and detection (p) probabilities and compare them to occupancy and detection values from previous Carolina Madtom studies. Detection histories were created using each successive artificial cover unit observation event as a repeat visit to the location. In total, artificial cover unit surveying occurred at 8 sites, and the number of observation events ranged from 3 to 6 per sites, resulting in 8 sampling locations with detection histories ranging from 3 to

6 repeat visits. Occupancy modeling was completed using the software program PRESENCE

(USGS Patuxent Wildlife Research Center, Laurel, MD; https://www.mbr- pwrc.usgs.gov/software/presence.html)

Results

In total, 30 Carolina Madtoms were collected from artificial cover units at 2 of the 8 sites in which they were deployed. These 30 captures were comprised of 21 unique collections and 9 recaptures, as assessed by presence or absence of the right pelvic fin. Collections came from sites

7 and 8 in the Tar River basin on Fishing Creek and Swift Creek (Table 2; Figure 2). No

Carolina Madtoms were collected from cover units deployed at two sites in the Neuse River basin. All sites were also snorkel surveyed in 2016 or 2017. Carolina Madtoms were captured at

3 of the 8 sites through snorkel survey efforts. Carolina Madtoms were detected at two sites following a standardized snorkel survey protocol developed in Chapter 1 of this thesis, and the species was detected at one additional site during supplemental snorkeling, not following the standard protocol. Considering only standardized snorkeling surveys, detection of Carolina

Madtoms was in complete agreement with detections by artificial cover units. Our modeled detection probability sampling with artificial cover units was 0.92, which was similar to that of snorkeling surveys (Table 3). The estimated occupancy probability associated with artificial cover unit sampling was 0.25, similar to that based on snorkel surveys during recent surveys.

Carolina Madtoms occupying cover units ranged from 72 to 121 mm in total length and weighed 4.21 g to 20.54 g. They were found occupying cover units placed in microhabitats with a mean water depth of 0.40 m, bottom velocity of 0.20 m/s, mean-column velocity of 0.35 m/s, and substrates composed of sand and gravel (Table 4). In addition to Carolina Madtoms, other species were found occupying the cover units. Ictalurids, including Margined Madtoms (Noturus insignis), Channel Catfish (Ictalurus punctatus), and Flathead Catfish (Pylodictis olivaris), were observed inside the cover units (Table 5). Other benthic and cover-oriented fishes were also

58 observed occupying the cover units, including darters (Etheostoma sp.), sunfish (Lepomis sp.), and American Eels (Anguilla rostrata).

Additional descriptive observations were made regarding Carolina Madtom occupancy of artificial cover units. Artificial cover units originally deployed by Midway et al. (2010) during

2007-2008 remained and were discovered at site 8 in the Tar River basin (Figure 1). Four units were found, still functional and free of debris after 10 years in the stream. Carolina Madtoms were observed occupying these units during snorkel surveying and artificial cover unit monitoring events. Carolina Madtoms were also observed guarding eggs inside the cover units, indicating nesting and successful spawning function of the artificial habitat. Another common observation at site 8 in the Tar River basin was the tendency of Carolina Madtoms to use the entire cover unit as cover. If space was available between the rock-filled saucer of the unit base and the substrate, Carolina Madtoms could be found occupying this cavity.

Discussion

Our findings indicated that artificial cover units are an efficient passive sampling device for detecting presence or absence of Carolina Madtoms. Compared to our standardized snorkeling effort, cover units were as effective at detecting presence of madtoms at the sites where both methods were employed. Occupancy modeling results using cover units were also similar to results from snorkel surveying. The probability of detecting Carolina Madtoms using artificial cover units was 0.92. Compared to 0.94 estimated by Wood and Nichols (2011) for snorkel surveying, and 0.81 from our snorkel surveying (Chapter 1), artificial habitats are similarly efficient at detecting Carolina Madtoms. These findings are in accord with those of

Midway et al. (2010a) that showed deploying similar cover units increased abundance of

Carolina Madtom observations. Midway et al. (2010b) also found that Carolina Madtoms preferred artificial cover units over natural cover in laboratory choice experiments, and Burr et al. (1989) and Wood and Nichols (2008) also recorded Carolina Madtoms occupying beverage containers and other litter as instream cover. Extensive supplemental snorkeling effort may be more effective for detecting madtoms in areas of low abundance, suggested by our finding at Site

9 in the Neuse River basin, where the species was detected during supplemental snorkeling, but not in artificial cover units. However, multiple intensive snorkeling efforts are time-consuming and may not be cost-effective for routine monitoring. Artificial cover units are inexpensive and easy to construct and can be deployed and observed with much less effort than snorkeling.

Passive minnow trap sampling for presence or absence is currently the standard protocol for population status assessment of the Neuse River Waterdog (Necturus lewisi), another rare endemic of the Tar and Neuse river basins (Braswell and Ashton Jr. 1985), and likewise, widespread deployment of artificial cover units may be the most time-efficient and cost-effective method for monitoring the extant population status of the Carolina Madtom. Furthermore, our collection of four Margined Madtoms, one of the co-occurring madtom species in the sampled basins, and 51 other ictalurids in artificial cover units suggests that this method may be applicable to other species.

Artificial cover units could also be used as long-term habitat enhancement tools. While snorkel surveying Site 8 in the Tar River basin during 2016, we captured Carolina Madtoms inside four cover units originally deployed by Midway et al. (2010a) in 2007. Through 10 years of substrate alterations and flow fluctuations, the units remained operational and provided cover for madtoms. Even after the passing of Hurricane Matthew during the fall of 2016, which was one of the most intense rainfall events in eastern North Carolina and caused drastic flooding and

60 massive current flow through Swift Creek, these cover units were discovered again in 2017, validating their durability and effectiveness as long-term instream cover.

Physical instream habitat is an important factor in successful occupancy of artificial cover units. Cover units were placed in a cross-sectional grid pattern throughout the reaches in varying depths, water velocities, and substrates. However, occupied units were deployed in sand and gravel substrate, in moderate depth, and low to medium velocity areas. These microhabitat characteristics are similar to those found most suitable for Carolina Madtoms by Midway et al.

(2010) and our concurrent microhabitat study (Chapter 1). Cover units deployed in areas with silt substrate or low velocities were frequently found with sediment inside when returning after the

14-day soak period, and units deployed in shallow areas were prone to de-watering as stream levels receded during dry periods. These observations suggest that random grid placement of artificial cover units is not effective, and proper microhabitat analysis should be conducted prior to placing units to maximize occupancy potential.

In addition to serving as an efficient sampling technique, artificial cover units can provide spawning habitat and protection from predators for Carolina Madtoms. Adequate cover is necessary for successful Carolina Madtom nesting and spawning (Burr et al. 1989). Both the Tar and Neuse river basins are degraded and have experienced habitat alteration and loss, and madtoms are especially at risk from habitat degradation as substrate, flow, and instream cover are some of the first factors to become altered in impacted streams (Warren et al. 1997).

Deployment of artificial cover units can compensate for lost habitat that is essential for madtom viability. Male Carolina Madtoms were found guarding eggs inside artificial habitats, further validating their function as spawning habitat. Cover is also important in protecting prey species from predators. The Tar and Neuse river basins both have populations of introduced nonnative

Flathead Catfish, a voracious predator. Flathead Catfish have been known to consume native bullhead catfish (Ameiurus spp.) as well as other Noturus spp. in North Carolina (Pine et al.

2007; Baumann and Kwak 2011), and cover to avoid predation is vital for survival of the species.

Unlike other cover, which may be lifted or displaced during Flathead Catfish foraging, these artificial units provide a single small entryway which may prevent predators access to any fish inside.

The Carolina Madtom is an important component of stream ecosystems. Not only do they provide important functions, they also add to the great fish diversity found in the southeastern

U.S. Given recent declines of the species, proper conservation and management will be vital to retain the species. Habitat loss is the greatest threat to madtom conservation (Etnier and Starnes

1991), and artificial cover units may enhance degraded habitat in impacted basins. In addition to providing critical spawning habitat and protection from predators, our findings demonstrate that artificial cover units can also serve as an efficient sampling device that may enhance scientists’ ability to conserve this important, but imperiled species.

References

Aadland, L. P. 1993. Stream habitat types: their fish assemblages and relationship to flow. North

American Journal of Fisheries Management 13:790-806.

Angermeier, P. L. 1995. Ecological attributes of extinction-prone species: loss of freshwater

fishes of Virginia. Conservation Biology 9:143-158.

Baumann J. R., and T. J. Kwak. 2011. Trophic relations of introduced catfish in an Atlantic river.

Transactions of the American Fisheries Society 140:1120-1134.

Braswell, A. L., and R. E. Ashton Jr. 1985. Distribution, ecology, and feeding habits of Necturus

lewisi (Brimley). Brimleyana 10:13-36.

Burr, B. M., B. R. Kuhajda, W. W. Dimmick, and J. M. Grady. 1989. Distribution, biology, and

conservation status of the Carolina Madtom, Noturus furiosus, an endemic North

Carolina catfish. Brimleyana 15:57-86.

Cooke, S. J., C. M. Bunt, S. J. Hamilton, C. A. Jennings, M. P. Pearson, M. S. Cooperman, and

D. F. Markle. 2005. Threats, conservation strategies, and prognosis for suckers

(Catostomidae) in North America: insights from regional case studies of a diverse family

of non-game fishes. Biological Conservation 121:317-331.

Eklöv, A. G., and L. A. Greenberg. 1998. Effects of artificial instream cover on the density of 0+

Brown Trout. Fisheries Management and Ecology 5:45–53.

Etnier, D. A., and W. C. Starnes. 1991. An analysis of Tennessee’s jeopardized fish taxa. Journal

of the Tennessee Academy of Science 66:129–134.

Heggenes, J., and T. Traaen. 1988. Daylight responses to overhead cover in stream channels for

fry of four salmonid species. Ecography 11:194–201.

Jelks, H. L., S. J. Walsh, N. M. Burkhead, S. Contreras-Balderas, E. Diaz-Pardo, D. A.

Hendrickson, J. Lyons, N. E. Mandrak, F. McCormick, J. S. Nelson, S. P. Platania, B. A.

Porter, C. B. Renaud, J. J. Schmitter-Soto, E. B. Taylor, and M. L. Warren. 2008.

Conservation status of imperiled North American freshwater and diadromous fishes.

Fisheries 33:372-407.

Kwak, T. J., and M. C. Freeman. 2010. Assessment and management of ecological integrity.

Pages 353-394 in W. A. Hubert and M. C. Quist, editors. Inland fisheries management in

North America, third edition. American Fisheries Society, Bethesda, Maryland.

LeGrand, H. E., S. E. McRae, S. P. Hall, and J. T. Finnegan. 2008. Natural heritage program list

of the rare animal species of North Carolina. North Carolina Natural Heritage Program,

Division of Natural Resource Planning and Conservation, North Carolina Department of

Environment and Natural Resources. Raleigh, North Carolina.

Lindquist, D. G., J. R. Shute, P. W. Shute, and L. M. Jones. 1984. Selection of sites for egg

deposition and spawning dynamics in the Waccamaw Darter. Environmental Biology of

Fishes 11:107-112.

Mackenzie, A. R., and L. Greenberg. 1998. The influence of instream cover and predation risk

on microhabitat selection of Stone Loach Barbatula barbatula (L.). Ecology of

Freshwater Fish 7:87-94.

Maddock, I. 1999. The importance of physical habitat assessment for evaluating river health.

Freshwater Biology 41:373-391.

Magoulick, D. D. 2004. Effects of predation risk of habitat selection by water column fish,

benthic fish, and crayfish in stream pools. Hydrobiologia 527:209-221.

Mann, R. H. K. 1996. Environmental requirements of European non-salmonid fish in rivers.

Hydrobiologia 323:223-235.

Midway, S. R., T. J. Kwak, and D. D. Aday. 2010a. Habitat suitability of the Carolina Madtom,

an imperiled, endemic stream fish. Transactions of the American Fisheries Society

139:325-338.

Midway, S. R., D. D. Aday, T. J. Kwak, and K. Gross. 2010b. Cover preference of the Carolina

Madtom (Noturus furiosus), an imperiled endemic southeastern stream fish. Journal of

Freshwater Ecology 25:151-154.

Moring, J. R., and P. H. Nicholson. 1994. Evaluation of three types of artificial habitats for fishes

in a freshwater pond in Maine, USA. Bulletin of Marine Science 55:1149–1159.

NCDEQ (North Carolina Department of Environmental Quality. 2009. Neuse river basinwide

water quality plan. NC Department of Environment and Natural Resources. Raleigh,

North Carolina.

NCDWR (North Carolina Division of Water Resources). 2010. Neuse river basin water resources

plan. NC Department of Environment and Natural Resources. Raleigh, North Carolina.

NCNHP (North Carolina Natural Heritage Program). 2016. Natural heritage program list of rare

animal species of North Carolina 2016. Raleigh, North Carolina.

Orth, D. J., and R. J. White. 1999. Stream habitat management. Pages 249-283 in C. C. Kohler

and W. A. Hubert, editors. Inland Fisheries Management in North America, 2nd edition.

American Fisheries Society, Bethesda, Maryland.

Piller, K. R., and B. M. Burr. 1999. Reproductive Biology and spawning habitat supplementation

of the Relict Darter, Etheostoma chienense, a federally endangered species.

Environmental Biology of Fishes 55:145-155.

Piller, K. R., H. L. Bart, Jr., and J. A. Tipton. 2004. Decline of the Frecklebelly Madtom in the

Pearl River based on contemporary and historical surveys. Transactions of the American

Fisheries Society 133:1004–1013.

Pine, W. E, III, T. J. Kwak, and J. A. Rice. 2007. Modeling management scenarios and the

effects of an introduced apex predator on a coastal riverine fish community. Transactions

of the American Fisheries Society 136:105-120.

Rahel, F. J., and R. A. Stein. 1988. Complex predator-prey interactions and predator intimidation

among crayfish, piscivorous fish, and small benthic fish. Oecologia 75:94-98.

Ricciardi, A., and J. B. Rasmussen. 1999. Extinction rates of North American freshwater fauna.

Conservation Biology 13:1220–1222.

Robison, H. W., and G. L. Harp. 1985. Distribution, habitat and food of the Ouachita Madtom,

Noturus lachneri, a Ouachita River drainage endemic. Copeia 1985:216-220.

Warren, M. L., P. L. Angermeier, B. M. Burr, and W. R. Haag. 1997. Decline of a diverse fish

fauna: patterns of imperilment and protection in the southeastern United States. Pages

105-164 in G. W. Benz, and D. E. Collins, editors. Aquatic fauna in peril: the

southeastern perspective. Southeast Aquatic Research Institute, Lenz Design and

Communications, Decatur, Georgia.

Wood, C. J., and R. B. Nichols. 2008. Status assessment of the Carolina Madtom final report.

North Carolina Wildlife Resources Commission. Raleigh, North Carolina.

Wood, C. J., and R. B. Nichols. 2011. Status assessment of the Carolina Madtom: a rare North

Carolina endemic. Pages 295-303 in P. H. Michaletz and V. H. Travnichek, editors.

Conservation, ecology, and management of catfish: the second international symposium,

American Fisheries Society, Symposium 77, Bethesda, Maryland.

Table 1. Location of artificial cover unit deployment sites in the Tar and Neuse river basins during 2016 and 2017.

Basin and Site Waterway Longitude Latitude Tar 5 Fishing Creek 35.9739 -77.5404 Tar 6 Swift Creek 35.9666 -77.5855 Tar 7 Fishing Creek 36.1137 -77.6280 Tar 8 Swift Creek 36.0742 -77.8697 Tar 9 Little Fishing Creek 36.1861 -77.8758 Tar 11 Fishing Creek 36.1695 -77.9230 Neuse 9 Contentnea Creek 35.6977 -78.0611 Neuse 12 Little River 35.6103 -78.2136

Basin and Site Waterway Snorkeling detection Artificial cover unit detection Tar 5 Fishing Creek No No Tar 6 Swift Creek No No Tar 7 Fishing Creek Yes Yes Tar 8 Swift Creek Yes Yes Tar 9 Little Fishing Creek No No Tar 11 Fishing Creek No No Neuse 9 Contentnea Creek No No Neuse 12 Little River Yes* No

Table 3. Comparison of occupancy modeling results using snorkel survey methods from Wood and Nichols (2008) and Chapter 1 of this thesis and artificial cover units.

Study and Method Occupancy (Ψ) Detection (p) Wood and Nichols (2008) Snorkel 0.75 0.94 Cope (Chapter 1) Snorkel 0.35 0.81 Cope Artificial Cover Units 0.25 0.92

Table 4. Length, weight, and microhabitat characteristics from Carolina Madtoms (N = 21) collected in artificial cover units.

Variable Mean/Mode SD Range Length (mm) 93.52 12.34 72.00 - 121.00 Weight (g) 9.69 3.92 4.21 - 20.54 Depth (m) 0.40 0.19 0.16 - 0.72 Bottom Velocity (m/s) 0.20 0.14 0.02 - 0.54 Mean Column Velocity (m/s) 0.35 0.23 0.07 - 0.77 Dominant Substrate Sand - - Subdominant Substrate Gravel - -

Table 5. Observation of co-occurring ictalurid species and other fishes of interest captured in artificial cover units during summer 2016 and 2017.

Species Abundance Margined Madtom 4 Channel Catfish 49 Flathead Catfish 2 Darters 1 Sunfish 1 American Eel 1

Figures

Figure 1. (Top) Artificial cover unit constructed and deployed in Swift Creek in 2007 by Midway et al. (2010a). (Bottom) Same artificial cover unit recovered in Swift Creek during summer 2016.

Tar River Basin

Neuse River Basin

CHAPTER 3

Genetic Structure and Diversity of the Endemic Carolina Madtom

and Conservation Implications

Abstract

Identification and conservation of genetic diversity within and among freshwater fish populations is important to increase understanding to better manage and conserve imperiled species and preserve aquatic biodiversity. Imperiled species are especially vulnerable to loss of genetic diversity, mainly due to anthropogenic alterations of natural habitat. One such imperiled species is the Carolina Madtom (Noturus furiosus), a small nongame catfish endemic to the Tar and Neuse river basins of North Carolina. Genetic structure has not been studied in the species, and given recent population declines in both basins, identification of remaining genetic diversity within the species is vital for planning conservation efforts. To assess the status and trends of the genetic structure of the Carolina Madtom, we analyzed genetic markers from 173 individuals to

(1) define the genetic population structure of the Carolina Madtom, (2) determine inter- and intra-basin genetic differences of populations in the Tar and Neuse river basins, and (3) present management implications to guide conservation efforts. Using 10 microsatellite primers developed for the related Yellowfin Madtom (Noturus flavipinnis), we successfully identified genetic structure of the Carolina Madtom. Resulting analyses quantified low genetic diversity in the species. Mean (+SD) M-ratios for the Tar (0.414+0.117) and Neuse (0.117+0.102) basin populations indicated that both populations have experienced demographic bottlenecks, with that experienced by the Neuse River basin population more severe. Effective population size (Ne) estimates for the respective populations were small, on the order of tens of individuals, indicating low genetic diversity within populations. However, multilocus population differentiation metrics,

G’ST (0.135+0.031) and DEST (0.125+0.029), were significantly different from zero (p < 0.001), indicating significant genetic differentiation between the Tar and Neuse river basin populations, suggesting development of separate management plans for each basin. Application of our results will further inform managers on the status of the genetic variation in Carolina Madtom and guide conservation toward informed protective listing and management decisions to maintain the viability of this important endemic species with respect to both demographic and genetic variation needs in order to increase population numbers and fitness of the species.

Introduction

1985). Many factors, including urbanization, erosion, sedimentation, and pollution degrade and destroy habitats that freshwater fishes require to survive and reproduce (Jelks et al. 2008).

Perhaps the most detrimental stressor to imperiled fish populations is the construction of dams on freshwater systems (Poff et al. 1997). Damming is a major cause of population fragmentation and altered flow, both of which can negatively affect biodiversity and health. Not only does degraded and fragmented habitat directly affect fish by reducing the amount of available suitable habitat, but it also divides the species into multiple small populations, which may be at greater risk for loss of genetic variation (Vrijenhoek et al. 1985). Fragmented populations may no longer intermix due to impassable dams or large areas of unsuitable habitat between population patches.

Small, isolated populations are at risk of losing genetic diversity through inbreeding or random genetic drift, which can greatly reduce the fitness of such populations (Frankel and Soule 1981).

As such, identifying genetic structure and variation in such small populations is important for conservation efforts, especially in imperiled and endemic species (Ryman 1991).

Loss of biodiversity and genetic variation is of special concern in the southeastern United

States, because many of the problems listed above, including habitat degradation, fragmentation, damming, and flow alteration are common in the region (Jelks et al. 2008). Loss of genetic diversity is also of concern because the southeast contains 62% of the nation’s freshwater fish species; however, it also contains the greatest number of imperiled fishes (Warren et al. 1997;

Jelks et al. 2008). Many of these imperiled fishes are nongame species. Nongame fishes are critical to freshwater ecosystems by providing important ecological functions and contributing to healthy ecosystem biodiversity. However, many nongame fishes are in need of conservation and are understudied and minimally managed compared to more economically or recreationally important species (Cooke et al. 2005; Kwak et al. 2011). As more nongame fishes face imperilment, the need for genetic conservation increases. In recent years, captive breeding and reintroduction efforts have become common among imperiled freshwater species, especially among darters (Ammocrypta, Crystallaria, Etheostoma, Percina) and madtoms (Noturus spp.)

(Rakes et al. 1999; Shute et al. 2005; Throneberry 2009). Thorough research on genetic structure among the extant populations must be first conducted as to not mix and decrease diversity among the few remaining populations. For example, reintroduction efforts for the Yellowfin Madtom

(Noturus flavipinnis) and Smoky Madtom (Noturus baileyi) followed the development of microsatellite genetic markers and assessment of genetic differentiation among the parent and reintroduced populations (Williams and Moyer 2012). Such genetic markers also have been developed to quantify genetic diversity in many sportfishes, as well as some other nongame fishes (Farias et al. 2003; Dutton et al. 2008; Hallerman et al. 2015). However, many species are lacking available research on genetic diversity, hampering conservation and management efforts.

One such understudied nongame species is the Carolina Madtom (Noturus furiosus), a small catfish native to the Tar and Neuse river basins of North Carolina. Due to recent population declines, the Carolina Madtom is now listed as State Threatened (LeGrand 2008;

NCNHP 2016). Although populations are declining and the species is becoming more imperiled, the Carolina Madtom has been sparsely studied, with only three major surveying events describing and assessing status of extant populations (Bayless and Smith 1962; Smith and

Bayless 1964; Burr et al. 1989; Wood and Nichols 2011), and no population genetic research has been conducted. Populations in the Neuse River basin, in particular, have been intensively impacted, with a 92% loss of historical occurrences in the basin over the past 50 years (Wood and Nichols 2011; Chapter 1). Thus, identification of the genetic diversity of the remaining populations in the two basins is critical for planning to conserve the species, as reintroduction or translocation efforts may be implemented.

Our research was designed to provide a timely assessment of the status and trends of the genetic structure of the Carolina Madtom. Our objectives were to (1) define the genetic population structure of the Carolina Madtom, (2) determine intra- and inter-basin genetic differences within and between populations in the Tar and Neuse river basins, and (3) propose management implications to guide conservation efforts of the species.

Study Area

This research was conducted in the Neuse and Tar river basins of the Piedmont and

Bern (NCDWR 2010). The basin covers an area of 10,034 km2 and spans 18 counties.

Currently, 13% of the basin is considered urban, 45% forested, and 29% crop and pasture land

(NCWRC 2005). In 2007, the American Rivers Foundation listed the Neuse River as one of the

10 most endangered rivers in the United States (American Rivers 2007). Non-point source pollution from agriculture and forestry has degraded water quality and habitats throughout the basin. Commercial farming inputs, such as animal waste and fertilizers, contribute 60% of the nitrates and phosphates in the system (NCWRC 2005). Due to the dense human population in the basin, many municipalities have constructed dams and withdraw water from the created impoundments for human use, affecting river flow, and treated sewage effluent is discharged back to the river. Habitat loss is also an issue in the Neuse River basin, as increasing human population results in the loss of natural areas and increases in impervious surfaces (NCDWR

2010).

2001). The Tar-Pamlico basin is more rural and less impacted by human activities than the Neuse

Methods

Collections

Genetic material was collected non-lethally via clipping the right pelvic fin from all captured Carolina Madtoms. Carolina Madtoms were mainly captured via standardized snorkel

79 surveys, in which waterways were snorkeled and observed individuals were collected by dipnet

(Wood and Nichols 2011; Chapter 1). Collections came from North Carolina Museum of Natural

Sciences archived samples, North Carolina Wildlife Resources Commission biologist surveys, and from sampling by personnel associated with this study. In total, 200 Carolina Madtom fin- clips were taken, including 174 from the Tar and 26 from the Neuse river basins (Figure 1). Of these 200 fin-clips, 173 were used for Carolina Madtom genetic structure and diversity analyses

(Table 1). Individuals were captured from all major tributaries of the Tar River basin, including

Fishing Creek, Little Fishing Creek, Swift Creek, and the mainstem Tar River, and individuals from the Neuse River basin were captured in Contentnea Creek and Little River. Fin-clips were taken in the field and placed into vials with 95% ethanol. After returning from the field, samples were placed into a -20℃ freezer for storage until processing. All fin-clips were deposited and documented in the North Carolina Museum of Natural Sciences before being transported to

Virginia Tech University, Blacksburg, VA, for analysis.

DNA Markers

DNA was extracted using the Qiagen Blood and Tissue DNA Extraction Kit. DNA concentration and purity were assessed using a Lite spectrophotometer (Biodrop, Cambridge,

UK). DNA extractions consisted of 173 individuals, 147 from the Tar River basin and 26 from the Neuse River basin. The panel of DNA samples was screened for variation at 11 microsatellite loci developed by Williams and Moyer (2012) for the congeneric Yellowfin Madtom.

Amplification protocols followed the authors’ protocols, modified as appropriate to yield optimized results using BioRad MyCycler thermocyclers. The 15-µL reactions consisted of 50 ng/μl of template DNA, 5X Promega reaction buffer, 2.00 mM MgCl2, 1.25 mM dNTP mix, 0.5

μM of each forward and reverse primer, and 0.08 units of Promega Taq polymerase (Promega

Corporation, Madison, WI). PCR reactions followed the thermal profile of initial denaturation at

94oC (10 min); followed by a touchdown PCR protocol with 35 cycles of denaturation (94oC, 30 sec.), annealing, and extension (74oC, 30 sec.); the initial annealing temperature was 60oC (1 min.) and decreased by 0.2oC/cycle. Final extension was 74oC for 5 min. The presence of amplification products was checked by running an aliquot of reaction products through an ethidium bromide-stained TBE agarose gel and observation of the gel under UV light.

Multiplexed PCR reaction products were subjected to fragment-size analysis at the Cornell

University Core Facility (Ithaca, NY) using an ABI 3730xl automated DNA sequencer.

Data Analysis

MICROCHECKER (van Oosterhout et al. 2004) was used to check the data for evidence of null alleles, large-allele dropout or stuttering. We used program STRUCTURE, version 2.3.4

(Pritchard et al. 2000), which utilizes an individual-based model-based approach to assign individuals to populations, to infer the most likely number of genetically distinct populations.

We utilized the admixture ancestry Bayesian model with correlated allele frequencies (Falush et al 2003). We performed a burn-in of 10,000 runs and conducted 100,000 MCMC repetitions for

K = 1-15 clusters and with 5 independent runs for each value of K. The true number of populations (K) was assessed two different ways. First, we sought the most likely K as indicated in the STRUCTURE output as the highest value of Ln Pr(X/K). Second, we assessed the most likely K by calculating delta K using the Evanno et al. (2005) method. The method compares each K to ΔK (ΔK = mean (|L''(K)|) / SD (L(K))).

Polymorphism in individual populations and for data pooled across populations was quantified as number of alleles per locus (A), and observed (HO) and expected (HE) heterozygosities using Microsatellite Toolkit (Park 2001) and Arlequin version 3.5.1.3 (Excoffier

81 and Lischer 2010). Arlequin was used to assess linkage disequilibrium with 10,000 permutations.

Arlequin was used to apply the Hardy-Weinberg exact test (Haldane 1954) to the null hypothesis of random union of gametes. A Markov chain method with 10,000 dememorization steps and

100,172 iterations was used to estimate without bias the exact p-value for this test (Levene 1949,

Guo and Thompson 1992). All p-values were adjusted for multiple comparisons using the sequential Bonferroni method (Rice 1989).

Analysis of molecular variance (AMOVA) and calculation of F-statistics followed the methods of Weir and Cockerham (1984), Excoffier et al. (1992), and Weir (1996) and was performed using Arlequin using 10,000 permutations. Calculation of F-statistics and RST (Slatkin

1995) was performed using Arlequin. G’ST (Hedrick 2005) and DEST (Jost 2008) were calculated using GenAlEx 6.5 (Peakall and Smouse 2012), and their significance was assessed using 1,000 permutations of the data. Significance tests were performed with 10,000 permutations. The

Garza-Williamson (G-W) index, which indicates a bottleneck at values below 0.70 when sample size is less than 75 (Garza and Williamson 2001), was calculated using Arlequin.

Effective population size was estimated using LDNe, version 1.31 (with the random mating model, lowest allele frequency of 0.02, and parametric confidence intervals; Waples and

Do 2008) and NeEstimator (Do et al. 2014). These software applications utilize the bias- corrected linkage disequilibrium (Waples and Do 2008), molecular coancestry (Nomura 2008), and heterozygosity excess (Zhdanova and Pudovkin 2008) methods.

Results

All 11 microsatellite primer pairs developed for the Yellowfin Madtom amplified

Carolina Madtom DNA. Amplification products of 10 loci were readily interpretable (Appendix

Table 1); however, results for locus NflD139 were not interpretable. Analyzed globally using

MICROCHECKER, genotype frequencies at the ten loci scored were not in Hardy-Weinberg equilibrium (Table 2), suggesting segregation of null alleles or the effects of population genetic processes, such as inbreeding or Wahlund effect (i.e., the mixing of differentiated populations).

Significant linkage disequilibrium (LD) was observed at 79 of 81 loci-by-locus comparisons in the Tar River population and 65 of 81 in the Neuse River population. It is highly unlikely that all the microsatellite loci screened are linked in the sense of being located on the same chromosome.

Rather, we interpret these positive test results as the consequence of violation of a critical assumption underlying tests for LD, that the population tested be large enough that segregation of alleles at all loci occurs at random. In this instance, because the individuals genotyped were drawn from small populations, genotypes at independent loci were correlated by descent from a limited number of breeders.

We performed STRUCTURE analysis to seek clustering in individual-and population- level patterns of genetic differentiation and to inform the approach to subsequent statistical analyses. The lnP(X/K) criterion declined as K increased well beyond the number of collections made (Appendix Table 2). The delta K criterion indicated that K = 2 was the best-supported number of clusters. At K = 2, there was no geographic basis for distribution of genetic variation

(Figure 2), and at higher levels of K, differentiation among individuals, but not geographically defined collections, became more narrowly drawn. This pattern of results is indicative of lack of population-level genetic variation in Carolina Madtom. Recognizing that Carolina Madtom populations in the respective rivers would be demographically independent and hence separate management units, we performed all subsequent analyses for groupings within each river and, as appropriate, overall.

All microsatellite loci were variable for Carolina Madtoms in each drainage (Table 3), with a mean (+SD) of A = 17.7 (+7.7) alleles per locus in the Tar River basin and A = 11.8 (+7.3) in the Neuse River basin. For all loci on both populations, observed heterozygosity HO was less than expected heterozygosity HE, indicating significant (p < 0.001) departures from Hardy-

Weinberg equilibrium expectations. This finding may be attributed to segregation of null alleles, inbreeding, or both.

Analysis of molecular variation was conducted, and F-statistics were calculated to quantify the distribution of genetic variation within and among Carolina Madtoms of the Tar and

Neuse river basins (Table 4). Results of AMOVA showed that most variation (98.04%) was partitioned within the respective populations and rather little (1.96%) between the Tar and Neuse river basin populations. All F-statistics were significantly different from zero. FIT, a metric of total departure of genotype frequencies from Hardy-Weinberg equilibrium expectations, was

0.502, a reasonably high value. FIS, which quantifies within-population departures of genotype frequencies from Hardy-Weinberg expectations, was 0.490, positive indicating a rather large deficit of heterozygotes. FIS was somewhat greater in Neuse than in Tar populations of Carolina

Madtoms. Using this algorithm, FST, a metric of population-level differentiation, was 0.019, which is rather low. RST, a metric of population-level differentiation which assumes stepwise mutation of microsatellite alleles and includes consideration of allele size, was somewhat higher as expected at RST = 0.035, but still rather low. Using metrics developed for quantifying population-level differentiation using microsatellite loci, G’ST was 0.135 (+SE = 0.031) and DEST was 0.125 (+0.029), both higher than FST, as expected, and both significantly different from zero

(p < 0.001).

We reworked the analysis using four subpopulations with adequate population sizes, (i.e. with separate consideration of the Swift Creek, Fishing Creek, and Tar River mainstem within the Tar River basin, and Contentnea Creek within the Neuse River basin). Overall FST was 0.032

(+ SE = 0.004), G’ST = 0.137 (+0.031), and DEST = 0.125 (+0.029). Matrices of subpopulation- by-subpopulation G’ST and DEST values are shown in Table 5.

The Garza-Williamson index, or M-ratio, is a microsatellite marker-based indicator of recent genetic population bottlenecks in a population (Garza and Williamson 2001). The M-ratio for a locus is the number of alleles exhibited by a population at a particular locus as a proportion of alleles possible within the allelic size range at that locus. M-ratios for demographically stable populations averaged 0.87 and for bottlenecked populations 0.64 in the original Garza and

Williamson (2001) report. Mean (+SD) M-ratios (Table 6) for the Tar and Neuse river basins

(0.414+0.117 and 0.341+0.102, respectively) indicate that both populations have experienced demographic bottlenecks, the Neuse River basin experiencing a more severe one.

Using the linkage disequilibrium method, effective population size Ne estimates for the respective populations (Table 7) were small, on the order of tens of individuals. Molecular coancestry-derived estimates of the effective number of breeders, Neb, were 11.0 (jack-knifed

95% CI = 5.6 – 18.2) for the overall system, 10.5 (5.4 – 17.1) for the Tar drainage, and 12.7 (3.5

– 27.8) for the Neuse drainage. Use of the heterozygosity excess method led to unresolved Ne estimates with confidence intervals including zero and infinity.

Discussion

Carolina Madtoms genetic structure was successfully quantified using microsatellite primers. Ten microsatellite primers from the congeneric Yellowfin Madtom successfully cross-

85 amplified in our study of the Carolina Madtom. Using these genetic markers, we found that

Carolina Madtoms had low genetic diversity throughout both the Tar and Neuse river basins, as determined through multiple testing methods. Carolina Madtoms are relatively abundant in the

Tar River basin, and would be expected to show moderate levels of genetic variability. However, populations in the Neuse River basin have declined demographically, and the samples screened would be expected to harbor less genetic variability. This outcome was apparent with individuals from the Neuse River basin displaying reduced numbers of alleles and lower M-ratios than Tar

River basin populations. M-ratio results for both basins revealed recent genetic bottlenecking events, with the Neuse River basin experiencing a slightly more severe one. Both basins also exhibited low effective population sizes, in the order of tens of individuals, explaining the low levels of genetic diversity within populations. However, using measures of genetic variation among individual sub-basin populations, we found considerable genetic differentiation among the Tar River, Fishing Creek, Swift Creek, and Contentnea Creek populations, large enough to warrant separate basin management plans.

A substantial success of the study was the amplification of 11 microsatellite primer pairs developed by Williams and Moyer (2012) for the Yellowfin Madtom. While cross-species primer transfer was shown to be relatively unsuccessful for the related Smoky Madtom, for which only three primer pairs successfully amplified, we successfully amplified and got readable results from 10 of 11 tested primer pairs. Cross-species transferability of microsatellite primers is highly variable, and successful amplification is not common; as such, these results are promising for interspecies analyses (Scribner et al. 1996; Barbara et al. 2007). However, as is the case in most cross-species primer transfers, genetic results were not perfect. Genotype frequencies from our Carolina Madtom samples were not in Hardy-Weinberg equilibrium, with a severe lack of

86 heterozygous genotypes. Ecologically, this can be explained by inbreeding, which favors homozygotes and reduces the number of heterozygotes in a population. However, we cannot unequivocally conclude that these results are due to inbreeding, as the cross-species primer pairs returned high frequencies of null alleles, for which proper annealing of PCR primers does not occur, creating the false appearance of excess homozygotes in the population.

Our study was the first to describe the genetic structure of Carolina Madtoms.

Identification of genetic structure and variability is critical for management and conservation decisions for a species. Identification of such genetic variability is especially important among endangered and endemic species, which are at greatest risk of sudden loss of populations and genetic variability (Ryman 1991). As such, identification of genetic variability in the imperiled, endemic Carolina Madtom may guide informed decisions about possible translocation and population augmentation decisions.

Identification of genetic variability is common among endangered species, and the results from the Carolina Madtom are not unique. Identification of genetic variation in Shortnose

Sturgeon (Acipenser brevirostrum), Yazoo Darter (Etheostoma raneyi), and the Neotropical catfish Steindachneridion parahybae have shown that anthropogenic habitat alteration has negatively affected the genetic diversity of the species (Quattro et al. 2002; Sterling et al. 2012; da Fonseca et al. 2017). Research on the Caddo Madtom (Noturus taylori) has also shown similarly low genetic diversity due to small, fragmented populations (Turner and Robison 2006).

However, much like the Carolina Madtom, Caddo Madtoms exhibited structuring among individual subpopulations, even at a relatively small spatial scale. Comparison with other benthic habitat specialist species, Brindled Madtom (Noturus miurus) and Paleback Darter (Etheostoma pallididorsum), revealed that habitat specialist species such as these are more likely to exhibit

87 genetic variation at smaller spatial scale because of their inability to migrate between populations separated by patches of unsuitable habitat. Such genetic variation was seen in the Carolina

Madtom, as three separate sub-basin populations in the Tar River basin exhibited significant genetic differentiation although they are not greatly separated spatially (both the Fishing Creek and Swift Creek tributaries are within 11 km of each other off the mainstem Tar River).

Interbasin genetic variation is also common in imperiled fishes. An imperiled, endemic cyprinid

(Chondrostoma lusitanicum) exhibited significant genetic variation to warrant separate basins as individual management units (Mesquita et al. 2001). Diadromous species such as smelt

(Retropinna spp.) also have exhibited interbasin genetic variability, even though their marine larval stage allows for connectivity among multiple river basins (Hughes et al. 2014).

Connectivity between the Tar and Neuse rivers likely last occurred during the Pleistocene epoch when sea level was 100 meters lower than currently (Bloom 1983). With the rise of sea level,

Carolina Madtom populations in the two river basins became subject to independent evolutionary histories. Genetic differentiation of Tar and Neuse river populations was considerable, and coupled with population losses, individual basin management plans may be warranted.

Conservation Implications

Although we successfully amplified primers developed for the congeneric Yellowfin

Madtom, species-specific microsatellite primers for the Carolina Madtom would allow for greater confidence in genetic analyses and provide greater accuracy and genetic resolution to guide conservation of the species. The segregation of null alleles was apparent in our data set, which impacted our ability to reach well-supported inferences of population genetic structure and underlying processes in the Carolina Madtom. As such, specific microsatellite loci primer pairs developed specifically for Carolina Madtoms would allow for more accurate analyses of genetic

88 data. Development of primers could be approached in two ways. First, genomic DNA could be sheared and applied to a column bearing beads with affinity oligonucleotides targeting specific microsatellite motifs; eluted DNA would be sequenced. Second, genomic DNA could be shotgun-sequenced and microsatellite arrays sought within the resulting DNA fragments. Using either approach, candidate primer pairs would be designed and tested for amplification, repeatability, and variability, with the resulting markers potentially exhibiting linkage and

Hardy-Weinberg equilibrium.

Given the results from our study, a variety of management options may be available to conserve the remaining genetic variation of the Carolina Madtom. The calculations of G’ST and

DEST revealed that 4 separate sub-basin populations experienced significant genetic differentiation. Thus, the two basins, and possibly all 4 individual sub-basin populations may be treated as separate management units. Separate management of the basins would ensure that genetic diversity is not lost in the event of translocation or population reintroduction events.

Reintroduction efforts are common among federally listed threatened and endangered species as viable population recovery goals (Tear et al. 1993). However, concerns have been raised over reintroduction efforts as species often are reintroduced into habitat that retains problematic factors which caused the original population collapse (Conant 1988). Even with such concerns, multiple madtom species have been successfully reintroduced in the southeastern United States.

The Smoky Madtom and Yellowfin Madtom have both been successfully propagated in captivity, released into stream habitat in their native ranges in Tennessee, and successful reproduction and population increases in the wild post-reintroduction have been documented

(Rakes et al. 1999; Shute et al. 2005; Throneberry 2009). Given these findings, reintroduction is an attainable goal for Carolina Madtoms, but any interbasin genetic transfer should be carefully

89 and cautiously deliberated, presuming a goal to maintain genetic diversity, and any population reintroduction or augmentation into the impacted Neuse River basin would require identification and amelioration of the stressors and impacts that caused the population collapse in that basin prior to activities.

Human-mediated impacts to the Neuse River basin are impediments to conservation of the Carolina Madtom. The sub-basin population from the Neuse River basin at Contentnea Creek has not been detected in recent years despite extensive sampling efforts, leaving all known genetic diversity to come from a small number of individuals occurring in the Little River. The threats facing the basin also make reintroduction or augmentation efforts a challenge in the

Neuse River. Nonnative Flathead Catfish (Pylodictis olivaris) have been introduced into the basin (Kwak et al. 2006). Flathead Catfish are voracious predators that have been shown to suppress native fish populations by up to 50% and to positively select ictalurid prey, including native bullhead catfish (Ameiurus spp.) and madtom species in North Carolina rivers (Pine et al.

2005, 2007; Baumann and Kwak 2011). Snorkel surveying during our study documented co- occurrence of Carolina Madtoms and Flathead Catfish in the Little River in the Neuse River basin, which may be harboring the last individuals in the entire basin. Therefore, reintroduction of captively propagated Carolina Madtoms may not be feasible because any reintroduced individuals would be at risk of immediate Flathead Catfish predation.

With the threat of Flathead Catfish predation in the Neuse River basin, a reasonable approach to population expansion may be the establishment of experimental populations from captively propagated individuals in reaches away from areas where predation is likely. Timely collection of remaining spawning adults may be considered in the Little River to provide brood stock for reintroduction purposes. Using captively propagated individuals from the Neuse River

90 basin may allow introducing Carolina Madtoms into areas with suitable habitat that are currently not occupied by Flathead Catfish to determine if captively propagated individuals can successfully survive and reproduce in the wild. Historical Carolina Madtom occurrences have been recorded in the Eno River of the Neuse River basin above Falls Lake, north of Raleigh,

North Carolina (Wood and Nichols 2011). This location may be suitable for an experimental

Neuse River basin population, as there is recorded suitable habitat for Carolina Madtom occurrence, and the Eno River currently has no record of Flathead Catfish occurrence. Similarly, the adjoining Flat River in the same area is another possible location for an experimental population.

The Carolina Madtom is an important component of the stream ecosystem, functioning both as an important ecological link and as part of the network of endemics that render North

Carolina and the southeastern United States freshwater systems biologically diverse and unique.

Given the species’ recent decline and current small, fragmented populations, conservation is vital not only to maintain a unique, endemic genome, but to also maintain diversity and ecological integrity of these waterways. The application of our results will further inform managers on the status of the genetic variation in Carolina Madtom and guide protective listing, planning, and decisions to maintain the viability of this endemic species with respect to both demographic and genetic variation needs to increase population numbers and the fitness of the species.

References

American Rivers Foundation. 2007. America’s most endangered rivers of 2007. American Rivers

Foundation, Washington, DC.

Barbara, T., C. Palma-Silva, G. M. Paggi, F. Bered. M. F. Fay, and C. Lexer. 2007. Cross-

species transfer of nuclear microsatellite makers: potential and limitations. Molecular

Ecology 16:3759-3767.

Baumann J. R., and T. J. Kwak. 2011. Trophic relations of introduced Flathead Catfish in an

Atlantic river. Transactions of the American Fisheries Society 140:1120-1134.

Bayless, J. D., and W. B. Smith. 1962. Survey and classification of the Neuse River and

tributaries, North Carolina. NC Wildlife Resources Commission, Raleigh, North

Carolina.

Bloom, A. L. 1983. Sea level and coastal changes. Pages 42-51 in H. E. Wright, editor. Late-

Quaternary environments of the United States, volume 2. University of Minnesota Press,

Minneapolis.

Burr, B. M., B. R. Kuhajda, W. W. Dimmick, and J. M. Grady. 1989. Distribution, biology, and

conservation status of the Carolina Madtom, Noturus furiosus, an endemic North

Carolina catfish. Brimleyana 15:57-86.

Conant, S. 1988. Saving endangered species by translocation: are we tinkering with evolution?

BioScience 38:254-257.

Cooke, S. J., C. M. Bunt, S. J. Hamilton, C. A. Jennings, M. P. Pearson, M. S. Cooperman, and

D. F. Markle. 2005. Threats, conservation strategies, and prognosis for suckers

(Catostomidae) in North America: insights from regional case studies of a diverse family

of non-game fishes. Biological Conservation 121:317-331.

92 da Fonseca, F. S., R. R. Domingues, E. M. Hallerman, and A. W. S. Hilsdorf. 2017. Genetic

diversity of an imperiled Neotropical catfish and recommendations for its restoration.

Frontiers in Genetics 8:1-12.

Do, C., R. S. Waples, D. Peel, G. M. Macbeth, B. J. Tillett, and J. R. Ovenden. 2014.

NeEstimator, v2: re-implementation of software for the estimation of contemporary

effective population size (Ne) from genetic data. Molecular Ecology Resources 14:209-

214.

Dutton, D. J., J. H. Roberts, P. L. Angermeier, and E. M. Hallerman. 2008. Microsatellite

markers for the endangered Roanoke Logperch, Percina rex (Percidae) and their potential

utility for other darter species. Molecular Ecology Resources 8:831-834.

Evanno, G., S. Regnaut, and J. Goudet. 2005. Detecting the number of clusters of individuals

using the software STRUCTURE: a simulation study. Molecular Ecology 14:2611-2620.

Excoffier, L., P. E. Smouse, and J. M. Quattro. 1992. Analysis of molecular variance inferred

from metric distances among DNA haplotypes: application to human mitochondrial DNA

restriction data. Genetics 131:479-491.

Excoffier, L., and H. E. Lischer. 2010. Arlequin suite, ver. 3.5: a new series of programs to

perform population genetics analyses under Linux and Windows. Molecular Ecology

Resources 10:564-567.

Falush, D., Stephens, M. and Pritchard, J. K. 2003. Inference of population structure using

multilocus genotype data: linked loci and correlated allele frequencies. Genetics

164:1567-1587.

Farias, I. P., T. Hrbek, H. Brinkmann, I. Sampaio, and A. Meyer. 2003. Characterization and

isolation of DNA microsatellite primers for Arapaima gigas, an economically important

but severely over-exploited fish species of the Amazon basin. Molecular Ecology

Resources 3:128-130.

Frankel, O .H., and M. E. Soule. 1981. Conservation and Evolution. Cambridge University Press.

Cambridge, United Kingdom.

Garza, J. C., and E. G. Williamson. 2001. Detection of reduction in population size using data

from microsatellite loci. Molecular Ecology 10:305-318.

Goudet, J. 2005. FSTAT version 2.9.3.2. Department of Ecology and Evolution, Biology

Building, UNIL, CH-1015 LAUSANNE, Switzerland.

https://www2.unil.ch/popgen/softwares/fstat.htm.

Guo, S. W., and E. A. Thompson. 1992. Performing the exact test of Hardy-Weinberg proportion

for multiple alleles. Biometrics 1:361-372.

Haldane, J. S. 1954. An exact test for randomness of mating. Journal of Genetics 52:631-635.

Hallerman, E. M., N. Johnson, and D. Dutton. 2015. Implications of microsatellite DNA

variation upon management of Virginia Smallmouth Bass populations. Pages 363-378 in

M. D. Tringali, J. M. Long, T. W. Birdsong, and M. S. Allen, editors. Black Bass

diversity: multidisciplinary science for conservation. American Fisheries Society,

Bethesda, Maryland.

Hedrick, P. W. 2005. A standardized genetic differentiation measure. Evolution 59:1633-1638.

Hughes, J. M., D. J. Schmidt, J. I. Macdonald, J. A. Huey, and D. A. Crook. 2014. Low

interbasin connectivity in facultatively diadromous fish: evidence from genetics and

otolith chemistry. Molecular Ecology 23:1000-1013.

Jelks, H. L., S. J. Walsh, N. M. Burkhead, S. Contreras-Balderas, E. Diaz-Pardo, D. A.

Hendrickson, J. Lyons, N. E. Mandrak, F. McCormick, J. S. Nelson, S. P. Platania, B. A.

Porter, C. B. Renaud, J. J. Schmitter-Soto, E. B. Taylor, and M. L. Warren. 2008.

Conservation status of imperiled North American freshwater and diadromous fishes.

Fisheries 33:372-407.

Jost, L. 2008. GST and its relatives do not measure differentiation. Molecular Ecology 17:4015-

4026.

Kwak, T. J., W. E. Pine III, and D. S. Waters. 2006. Age, growth, and mortality of introduced

Flathead Catfish in Atlantic rivers and a review of other populations. North American

Journal of Fisheries Management 26:73-87.

Kwak, T. J., M. T. Porath, P. H. Michaletz, and V. H. Travnichek. 2011. Catfish science: status

and trends in the 21st century. Pages 755-780 in P. H. Michaletz and V. H. Travnichek,

editors. Conservation, ecology, and management of catfish: the second international

symposium, American Fisheries Society Symposium 77, Bethesda, Maryland.

LeGrand, H. E., S. E. McRae, S. P. Hall, and J. T. Finnegan. 2008. Natural heritage program list

of the rare animal species of North Carolina. North Carolina Natural Heritage Program,

Division of Natural Resource Planning and Conservation, North Carolina Department of

Environment and Natural Resources. Raleigh, North Carolina.

Levene H. (1949). On a matching problem arising in genetics. The Annals of Mathematical

Statistics 20:91-94.

Meirmans, P. G., and P. W. Hedrick. 2011. Assessing population structure: FST and related

measures. Molecular Ecology 11:5-18.

Mesquita, N., G. Carvalho, P. Shaw, E. Crespo, and M. M. Coelho. 2001. River basin-related

genetic structuring in an endangered fish species, Chondrostoma lusitanicum, based on

mtDNA sequencing and RFLP analysis. Heredity 86:253-264.

NCDEQ (North Carolina Department of Environmental Quality. 2010. Tar-Pamlico River basin

2010 water quality plan. NC Department of Environmental Quality. Raleigh, North

Carolina.

NCDWR (North Carolina Division of Water Resources). 2001. State water supply plan. NC

Department of Environment and Natural Resources. Raleigh, North Carolina.

NCDWR (North Carolina Division of Water Resources). 2010. Neuse River basin water

resources plan. NC Department of Environment and Natural Resources. Raleigh, North

Carolina.

NCNHP (North Carolina Natural Heritage Program). 1997. The upper Tar River basin: Swift

Creek and Fishing Creek subbasins. Biological conservation database. Raleigh, North

Carolina.

NCNHP (North Carolina Natural Heritage Program). 2016. Natural Heritage Program list of rare

animal species of North Carolina 2016. Raleigh, North Carolina.

NCWRC (North Carolina Wildlife Resources Commission). 2005. North Carolina wildlife action

plan. Raleigh, North Carolina.

Nomura, T. 2008. Estimation of effective number of breeders from molecular coancestry of

single cohort sample. Evolutionary Applications 1:462-474.

Park, S. J. 2001. Microsatellite toolkit for Excel. Smurfit Institute of Genetics, Trinity College,

University of Dublin, Ireland.

Peakall, R. and P. E. Smouse. 2012. GenAlEx 6.5: genetic analysis in Excel. Population genetic

software for teaching and research-an update. Bioinformatics 28:2537-2539.

Pine, W. E, III, T. J. Kwak, D. S. Waters, and J. A. Rice. 2005. Diet selectivity of introduced

Flathead Catfish in coastal rivers. Transactions of the American Fisheries Society

134:901-909.

Pine, W. E, III, T. J. Kwak, and J. A. Rice. 2007. Modeling management scenarios and the

effects of an introduced apex predator on a coastal riverine fish community. Transactions

of the American Fisheries Society 136:105-120.

Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E. Sparks,

and J. C. Stromberg. 1997. The natural flow regime: a paradigm for river conservation

and restoration. BioScience 47:769-784.

Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using

multilocus genotype data. Genetics 155:945-959.

Quattro, J. M., T. W. Greig, D. K. Cokyendall, B. W. Bowen, and J. D. Baldwin. 2002. Genetic

issues in aquatic species management: the Shortnose Sturgeon (Acipenser brevirostrum)

in the southeastern United States. Conservation Genetics 3:155-166.

Rakes, P. L., J. R. Shute, and P. W. Shute. 1999. Reproductive behavior, captive breeding, and

restoration ecology of endangered fishes. Environmental Biology of Fishes 55:31-42.

Raymond, M., and F. Rousset. 1995. GENEPOP, version 1.2: population genetics software for

exact tests and ecumenicism. Journal of Heredity 86:248-249.

Ryman, N. 1991. Conservation genetics considerations in fishery management. Journal of Fish

Biology 39:211-224.

Scribner, K. T., J. R. Gust, and R. L. Fields. 1996. Isolation and characterization of novel salmon

microsatellite loci: cross-species amplification and population genetic applications.

Canadian Journal of Aquatic Sciences 53:833-841.

Shute, J. R., P. L. Rakes, and P. W. Shute. 2005. Reintroduction of four imperiled fishes in

Abrams Creek, Tennessee. Southeastern Naturalist 4:93-110.

Slatkin, M. 1995. A measure of population subdivision based on microsatellite allele frequencies.

Genetics 139:457-462.

Smith, W. B., and J. D. Bayless. 1964. Survey and classification of the Tar River and tributaries,

North Carolina. NC Wildlife Resources Commission. Raleigh, North Carolina.

Sterling, K. A., D. H. Reed, B. P. Noonan, M. L. Warren Jr. 2012. Genetic effects of habitat

fragmentation and population isolation on Etheostoma raneyi (Percidae). Conservation

Genetics 13:859-872.

Tear, T. H., J. M. Scott, P. H. Hayward, and B. Griffith. 1993. Status and prospects for success

of the endangered species act: a look at recovery plans. Science 262:976-977.

Throneberry, J. K. Reintroduction success of Smoky Madtom Noturus baileyi and Yellowfin

Madtom Noturus flavipinnis in Abrams Creek, Great Smoky Mountains National Park.

Master’s thesis. Tennessee Tech University. Cookeville, Tennessee.

Turner, T. F., and H. W. Robison. 2006. Genetic diversity of the Caddo Madtom, Noturus

taylori, with comments on factors that promote genetic divergence in fishes endemic to

the Ouachita highlands. The Southwestern Naturalist 51:338-345.

Van Oosterhout, C., W. F. Hutchinson, D. P. Will, and P. Shipley. 2004. MICRO-CHECKER:

software for identifying and correcting genotyping errors in microsatellite data.

Molecular Ecology Resources 4:535-538.

Vrijenhoek, R. C., M. E. Douglas, and G. K. Meffe. 1985. Conservation genetics of endangered

fish populations in Arizona. Science 229:400-402.

Waples, R. S., and C. H. Do. 2008. LDNe: a program for estimating effective population size

from data on linkage disequilibrium. Molecular Ecology Resources 8:753-756.

Warren, M. L., P. L. Angermeier, B. M. Burr, and W. R. Haag. 1997. Decline of a diverse fish

fauna: patterns of imperilment and protection in the southeastern United States. Pages

105-164 in G. W. Benz, and D. E. Collins, editors. Aquatic Fauna in Peril: The

Southeastern Perspective. Southeast Aquatic Research Institute, Lenz Design and

Communications, Decatur, Georgia.

Weir, B. S. 1996. Genetic Data Analysis II. Sinauer and Associates, Sunderland, Massachusetts.

Weir, B. S., and C. C. Cockerham. 1984. Estimating F-statistics for the analysis of population

structure. Evolution 38:1358-1370.

Williams, A. S., and G. R. Moyer. 2012. Isolation and characterization of 21 microsatellite loci

for the federally endangered Yellowfin Madtom (Noturus flavipinnis) with cross species

amplification in N. baileyi. Conservation Genetics Resources 4:221-223.

Wood, C. J., and R. B. Nichols. 2011. Status assessment of the Carolina Madtom: a rare North

Carolina endemic. Pages 295-303 in P. H. Michaletz and V. H. Travnichek, editors.

Conservation, ecology, and management of catfish: the second international symposium,

American Fisheries Society Symposium 77, Bethesda, Maryland.

Zhdanova, O., and A. I. Pudovkin. 2008. Nb_HetEx: a program to estimate the effective number

of breeders. Journal of Heredity 99:694-695.

Table 1. Collections of Carolina Madtom tissue used for genetic analysis from the North Carolina Museum of Natural Sciences, North Carolina Wildlife Resource Commission, and this study.

Basin Waterbody N Tar Fishing Creek 52 Tar Little Fishing Creek 3 Tar Swift Creek 56 Tar Tar River 33 Tar Town Creek 3 Neuse Contentnea Creek 20 Neuse Little River 6

100

Table 2. Results of MICROCHECKER analysis of genotype frequencies at 10 microsatellite loci in Carolina Madtoms, with estimated frequency of null alleles using the van Oosterhout method.

Locus Null Present Estimated Frequency NflA3 yes 0.208 NflC138 yes 0.250 NflC143 yes 0.473 NflC145 yes 0.472 NflD105 yes 0.380 NflD123 yes 0.095 NflD129 yes 0.243 NflD137 yes 0.263 NflD146 yes 0.198 NflD145 yes 0.120

101

Tar River basin Neuse River basin

Locus A HO HE Locus A HO HE NflA3 12 0.485 0.851 NflA3 8 0.458 0.757 NflC143 10 0.023 0.877 NflC143 9 0.000 0.875 NflC145 8 0.000 0.806 NflC145 6 0.000 0.718 NflD105 21 0.200 0.887 NflD105 8 0.187 0.778 NflD123 34 0.771 0.956 NflD123 27 0.785 0.964 NflD129 15 0.458 0.892 NflD129 11 0.428 0.866 NflD137 13 0.390 0.867 NflD137 9 0.461 0.812 NflD138 14 0.454 0.882 NflD138 9 0.321 0.867 NflD145 23 0.746 0.942 NflD145 15 0.555 0.923 NflD146 20 0.561 0.911 NflD146 16 0.518 0.902 Mean 17 0.409 0.887 Mean 11.8 0.371 0.846 SD 7.7 0.266 0.043 SD 7.3 0.249 0.078

102

A. Analysis of molecular variance.

Degrees of Sum of Variance Percentage of Source of variation freedom squares components variation

Among populations 1 7.164 0.04248 Va 1.96 Among individuals 171 543.159 1.04772 Vb 48.26 within populations

Within individuals 173 187.00 1.08092 Vc 49.79 Total 345 737.324 2.17113 -

B. Fixation indices:

FIT FIS FST 0.502 0.490 0.019

C. Results of significance tests (10,100 permutations):

Variable p(random value > observed value)

Vb and FIS 0.00000 + 0.00000

Va and FST 0.00911 + 0.00097

Vc and FIT 0.00000 + 0.00000

D. Population-specific FIS indices (10,100 permutations):

Population FIS p(random FIS > observed FIS) Tar 0.486 0.000000 Neuse 0.520 0.000000

103

Swift Creek Tar River Fishing Creek Contentnea Creek Swift Creek - 0.148 0.053 0.147 Tar River - 0.101 0.179 Fishing Creek 0.002 0.002 - 0.183 Contentnea Creek 0.002 0.002 0.002 -

Swift Creek Tar River Fishing Creek Contentnea Creek Swift Creek - 0.140 0.050 0.138 Tar River 0.001 - 0.095 0.167 Fishing Creek 0.054 0.022 - 0.172 Contentnea Creek 0.007 0.004 0.004 -

104

Table 6. Locus-specific and overall M-ratios (Garza and Williamson 2001) for Carolina Madtom populations in the Tar and Neuse river basins.

Locus Tar Neuse Mean SD NflA3 0.521 0.421 0.471 0.071 NflC143 0.476 0.360 0.418 0.082 NflC145 0.533 0.400 0.466 0.094 NflD105 0.295 0.242 0.269 0.037 NflD123 0.343 0.317 0.33 0.018 NflD129 0.365 0.407 0.386 0.029 NflD137 0.52 0.333 0.426 0.131 NflD138 0.518 0.333 0.425 0.13 NflD145 0.193 0.118 0.155 0.053 NflD146 0.377 0.484 0.431 0.076 Mean 0.414 0.341 0.378 0.051 SD 0.117 0.102 0.109 0.009

105

Table 7. Estimated effective population size, Ne (mean + 95% CI) for stated groups of Carolina Madtoms using the linkage disequilibrium method.

Lowest allele frequency used Population 0.05 0.02

Overall 42.9 (38.3 – 48.2) 80.8 (73.0 – 89.8)

Tar River basin 39.0 (34.5 – 44.1) 70.7 (63.7 – 78.9)

Neuse River basin 76.9 (39.0 – 493.1) 63.7 (40.7 – 130.4)

106

Figures

Figure 1. Collection sites of Carolina Madtoms for genetic analysis. Collections were from the North Carolina Museum of Natural Sciences, North Carolina Wildlife Resource Commission, and this study.

107

K = 1

K = 2

K = 3

K = 4

108

APPENDIX

109

Appendix A: Supporting Information for Chapter 3

110

Table SI 1. Allele frequencies at 10 microsatellite loci for Carolina Madtoms.

Locus Population NflA3 Tar Neuse 234 0.74 - 236 5.51 - 238 22.06 6.25 240 15.81 41.67 242 20.22 25.00 244 14.34 8.33 246 9.19 8.33 248 6.25 6.25 250 3.31 - 252 1.10 2.08 254 0.74 - 256 0.74 2.08

NflC143 Tar Neuse 353 1.55 - 355 13.57 4.17 357 15.89 8.33 359 17.83 16.67 361 9.69 20.83 363 12.02 16.67 365 12.40 12.50 367 7.75 12.50 369 6.98 - 373 2.33 4.17 379 - 4.17

111

NflC145 Tar Neuse 149 0.70 - 151 16.08 3.70 153 24.48 7.41 155 24.48 44.44 157 16.78 25.93 159 14.69 14.81 161 2.10 - 163 0.70 - 165 - 3.70

NflD105 Tar Neuse 118 1.05 - 122 10.00 - 124 12.11 3.13 126 20.53 37.50 128 13.16 28.13 130 14.21 3.13 132 3.16 - 136 0.53 3.13 138 0.53 - 140 1.05 - 144 0.53 - 146 1.58 - 148 1.58 - 150 0.53 - 152 3.68 6.25 154 8.95 12.50 156 3.16 6.25 158 1.58 - 160 1.05 -

112

164 0.53 - 188 0.53 -

NflD123 Tar Neuse 345 2.50 - 347 2.86 - 349 2.50 7.14 351 2.14 1.79 353 3.21 1.79 355 2.50 - 357 2.86 5.36 359 0.71 - 361 3.57 - 365 0.71 1.79 367 1.43 1.79 369 1.43 3.57 371 6.43 8.93 373 6.07 5.36 375 8.57 10.71 377 5.36 7.14 379 8.21 1.79 381 5.71 5.36 383 7.14 3.57 385 3.93 5.36 387 3.21 1.79 389 2.86 3.57 391 2.14 1.79 393 2.86 1.79 395 2.86 3.57 397 1.07 3.57 399 1.07 -

113

401 0.36 - 403 1.43 1.79 405 0.71 - 407 2.50 - 411 0.36 1.79 413 0.36 - 427 - 1.79 429 - 1.79 431 - 3.57 443 0.36 1.79

NflD129 Tar Neuse 120 1.74 - 122 6.25 - 124 5.56 5.36 126 5.56 8.93 128 15.63 1.79 130 17.71 23.21 132 14.24 5.36 134 10.76 19.64 136 7.99 7.14 138 4.86 7.14 140 5.21 17.86 142 2.78 1.79 144 0.69 - 146 0.69 - 150 - 1.79 160 0.35 -

114

NflD137 Tar Neuse 109 1.88 - 111 8.65 - 113 10.15 9.62 115 10.53 15.38 117 26.32 36.54 119 16.54 13.46 121 5.26 5.77 123 2.63 - 125 4.51 - 127 4.51 11.54 129 3.76 3.85 131 3.01 1.92 133 2.26 - 139 - 1.92

NflD138 Tar Neuse 306 0.35 3.57 308 5.94 - 310 12.59 12.50 312 15.73 - 314 20.63 19.64 316 11.19 19.64 318 7.69 14.29 320 8.04 16.07 322 8.04 3.57 324 3.15 - 326 3.15 - 328 2.10 7.14 330 0.70 - 332 0.70 3.57

115

NflD145 Tar Neuse 251 0.38 - 253 2.31 - 255 4.23 - 257 5.00 5.56 259 4.62 20.37 261 6.92 5.56 263 10.77 7.41 265 8.08 7.41 267 6.92 3.70 269 10.00 7.41 271 5.77 7.41 273 5.38 14.81 275 6.15 5.56 277 4.62 5.56 279 3.08 3.70 281 5.00 - 283 2.31 1.85 285 3.08 - 287 1.92 3.70 289 2.31 - 291 0.77 - 293 0.38 -

NflD146 Tar Neuse 328 1.08 - 330 8.27 - 332 13.67 - 334 12.59 1.85 336 11.87 3.70 338 6.47 24.07

116

340 9.71 5.56 342 11.87 11.11 344 6.47 12.96 346 4.32 3.70 348 2.16 1.85 350 2.88 3.70 352 1.44 3.70 354 1.80 7.41 356 1.44 7.41 358 0.72 7.41 360 1.44 - 362 1.08 1.85 364 - 1.85 366 0.36 1.85 380 0.36 -

117

Table SI 2. Results of STRUCTURE analysis for K = 1-15 clusters of multilocus genotypes. Each value of LnP(X/K) represents the mean of 5 independent runs.

K LnP(X/K) Delta K 1 -7775 N/A 2 -7054 -721 3 -6697 -357 4 -6521 -175 5 -6347 -174 6 -6207 -139 7 -6125 -84 8 -6040 -85 9 -5927 -68 10 -5914 -58 11 -5924 +10 12 -5841 -83 13 -5834 -7 14 -5730 -104 15 -5671 -59

118