Habitat-Specific Production of a Fall Line River Shoal Macroinvertebrate Assemblage

Home , Fall line, Pleuroceridae, Podostemum ceratophyllum

HABITAT-SPECIFIC PRODUCTION OF A FALL LINE RIVER

SHOAL MACROINVERTEBRATE

ASSEMBLAGE

TIMOTHY DAVIS WYNN

ARTHUR C. BENKE, COMMITTEE CHAIR

ALEXANDER D. HURYN AMELIA K. WARD JONATHAN P. BENSTEAD JACK W. FEMINELLA

A DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Biological Sciences in the Graduate School of The University of Alabama

TUSCALOOSA, ALABAMA

2012

ABSTRACT

Fall Line shoals are zones of geomorphic complexity within a river basin, and have been recognized as sites of high aquatic macroinvertebrate diversity and production. The shoals of the free-flowing Cahaba River, in central Alabama, represent some of the most significant remaining examples of this channel feature that was once common throughout many rivers of the southeastern United States prior to widespread river regulation. The goal of this dissertation is to examine how the major habitats of a Cahaba River shoal influence the distribution and secondary production of the macroinvertebrate assemblage. Chapter 2 quantifies the variety of habitat types across the shoal reach and examines the temporal biomass dynamics of the 2 most common in-stream macrophytes, Justicia americana and Podostemum ceratophyllum. Chapter 3 presents a new method for obtaining in situ growth rates of several species of the diverse pleurocerid snail assemblage. The results of this method were later used to estimate production for this family.

Chapter 4 describes the distribution, biomass, and production of the nonnative Asiatic clam,

Corbicula fluminea, across the shoal reach, highlighting its dependence on Justicia habitat.

Finally, Chapter 5 incorporates the preceding chapters into a study of the distribution of macroinvertebrate assemblage production across bare bedrock, Justicia, and Podostemum habitats, as well as the entire shoal reach. Total annual production of all macroinvertebrates was

56.1 g AFDM m-2 y-1 in bedrock, 284.4 in Podostemum, and 177.3 in Justicia habitats. Total habitat-weighted production of the shoals reach was 87.1 g m-2 y-1, with bedrock contributing

24.3%, Podostemum 22.7%, and Justicia 53.0% to this total. This study supports the view that

Fall Line shoals can support high habitat diversity and production, and that the more complex habitats (e.g., those with macrophytes) enhance benthic invertebrate diversity and production.

Also, the influence of a given habitat depends largely on its relative abundance, and this study demonstrated that the Justicia habitat can have a dominant influence on diversity and production of a river reach. This work advances our understanding of the roles of shoal habitats in maintaining the diversity and function of this endangered river channel feature.

iii

DEDICATION

This dissertation is dedicated to my dad, Jeffrey Vernon Wynn, who took the time to show his sons the beauty and mystery of nature.

LIST OF ABBREVIATIONS AND SYMBOLS

Fig. Figure

No. Number ind. Individuals spp. Species, plural sp. Species, singular

% Percent km Kilometers m Meters cm Centimeters mm Millimeters

μm Micrometers

AW Aperture width

L Shell length bs Body size g Instantaneous growth rate g Grams mg Milligrams

AFDM Ash-free dry mass

DM Dry mass

# Pound

SA Surface area

L Liter

T Temperature

C Degrees Celsius

 Degrees

‘ Minutes

“ Seconds t Time in days y Year mo Month d day min Minute s Second

%C Percent cover

Ni No. of habitat-specific observations

Nt No. of total habitat observations pH Negative logarithmic value of Hydrogen ion conentration

NO2-N Nitrite-nitrogen

NO3-N Nitrate-nitrogen

PO4-P Phosphate-phosphorus

CaCO3 Calcium carbonate

SE Standard error n Number of independent observations

F F statistic p Probability of rejecting the null-hypothesis if that hypothesis is true

α Significance level r2 Coefficient of determination

ANOVA Analysis of variance ln Natural logarithm log Logarithm

< Less than

<< Much less than

≤ Less than or equal to

> Greater than

≥ Greater than or equal to

+ Plus

- Minus

± Plus or minus x Multiplication sign x Independent variable

= Equal to

~ Approximately

∆ Change in et al. And others

vii

i.e. That is

e.g. For example

vs. Versus

® Trademark

Inc. Incorporated

Co. County

AL Alabama

MD Maryland

UA The University of Alabama

USFWS United States Fish and Wildlife Service

USGS United States Geological Survey

BR Bare bedrock,

BRD Dry bedrock

POD Podostemum

JUS Justicia

JUSD Dry Justicia

JUStotal Submerged and emergent Justicia tissues

JUSemerge Emergent Justicia tissues

JUSsub Submerged Justicia tissues

HYM Hymenocallis

BO Boulder

BOD Dry boulder

G Gravel

viii

S Sand

M Mud

SG Snags

DP Deep pool

W Ind. mass

W Size class mass

W Mean mass between size classes

N Density

B Biomass

B Mean annual biomass

P Production

P/B Production to biomass ratio

CPI Cohort production interval

X Present

JAN January

FEB February

MAR March

APR April

JUN June

JUL July

AUG August

SEP September

OCT October

NOV November

DEC December

ACKNOWLEDGEMENTS

This dissertation would not have been possible without the generous help and encouragement of many people. First, I would like to express my sincere appreciation to all the members of my committee, Drs. Arthur C. Benke, Jonathan P. Benstead, Alexander D. Huryn,

Amelia K. Ward, and Jack W. Feminella, for contributing their invaluable time and insight toward the planning and execution of this work. I am especially grateful to my advisor, Dr.

Arthur C. Benke, for guiding me through this long and difficult process, and for believing in me even when I doubted myself. Working under his mentorship has been a true honor.

I would also like to express my gratitude to my family for their undying support, guidance, and friendship. Special thanks go to my wonderful wife and best friend, Kimberly, who has stood beside me every step of the way, and has given selflessly to help me achieve this goal. Mark Dedmon assisted me through every sampling trip, and was a cheerful companion throughout long exhausting days in the field. Laura Frost was my tireless lab assistant who helped make an impossible task possible. I would also like to thank fellow students Jeffrey

Pollock, Michael Venarsky, James Ramsey, Stephanie Parker, and Michael Kendrick for assistance in the field and lab, and for many helpful discussions. Thanks as well to Dr. Christina

Staudhammer for helpful advice about statistical methods. Finally, special thanks go to Lori

Tolley-Jordan, who first showed me around the Cahaba shoals, and introduced me to its amazing snail fauna. Financial support was provided by the University of Alabama Department of

Biological Sciences, the UA Graduate School, and the UA National Alumni Association.

CONTENTS

ABSTRACT...... ii

DEDICATION...... iv

LIST OF ABBREVIATIONS AND SYMBOLS ...... v

ACKNOWLEDGEMENTS...... xi

LIST OF TABLES...... xiii

LIST OF FIGURES ...... xiv

LIST OF APPENDICES...... xvi

CHAPTER 1: INTRODUCTION...... 1

CHAPTER 2: QUANTIFICATION OF BENTHIC HABITATS AND VEGETATION DYNAMICS IN A FALL LINE RIVER SHOAL...... 8

CHAPTER 3: IN SITU ESTIMATION OF PLEUROCERID SNAIL GROWTH USING A TETHERING TECHNIQUE IN A FALL LINE RIVER SHOAL...... 29

CHAPTER 4: PRODUCTION OF THE NONNATIVE ASIATIC CLAM, CORBICULA FLUMINEA, IN A SOUTHEASTERN RIVER BEDROCK SHOAL...... 45

CHAPTER 5: HABITAT-SPECIFIC PRODUCTION OF A FALL LINE RIVER SHOAL MACROINVERTEBRATE ASSEMBLAGE...... 63

CHAPTER 6: GENERAL CONCLUSIONS...... 116

LITERATURE CITED ...... 121

xii

LIST OF TABLES

2.1 Mean annual velocity and mean annual depth for both macrophyte habitats, Justicia americana and Podostemum ceratophyllum...... 21

2.2 Conversion of mean annual biomass per unit habitat to mean annual biomass per unit shoal surface area ...... 22

3.1 Length-mass regression relationships for Pleurocera vestita, Elimia clara, and Elimia showalteri...... 36

3.2 Simple regression model parameters and statistics for each species as well as combined genus and family-level models ...... 39

4.1 Size-frequency table illustrating the calculation of secondary production of Corbicula fluminea in the Justicia americana beds of Hargrove Shoals...... 55

5.1 Habitat-specific mean annual biomass, annual production, and annual P/B for the 41 most common benthic invertebrate taxa from the three most abundant habitats of Hargrove Shoals...... 73

5.2 Habitat-weighted mean annual density, biomass, annual production, and annual P/B for the 41 most common benthic invertebrate taxa of Hargrove Shoals ...... 84

xiii

LIST OF FIGURES

1.1 The Cahaba River Basin, central Alabama...... 6

1.2 Hargrove Shoals, Cahaba River, Bibb Co., Alabama...... 7

2.1 Percent cover of the variety of habitat types occurring in Hargrove Shoals at the peak of growing seasons 2007 and 2008 ...... 16

2.2 Biomass of Justicia americana in Hargrove Shoals ...... 18

2.3 Biomass of emergent Justicia americana in Hargrove Shoals ...... 19

2.4 Biomass of Podostemum ceratophyllum in Hargrove Shoals...... 20

3.1 Description of snail tethering technique ...... 31

3.2 Average weekly water temperature of Hargrove Shoals ...... 37

3.3 Regression plots of body size and temperature with growth for Pleurocera vestita, Elimia clara, and Elimia showalteri...... 38

4.1 Size-frequency histograms of Corbicula fluminea in Justicia americana beds of Hargrove Shoals...... 53

4.2 Mean density and biomass of Corbicula fluminea in Justicia americana beds ...... 54

5.1 Annual production of Trichoptera taxa in three main habitats (bedrock, Podostemum ceratophyllum, and Justicia americana) of Hargrove Shoals...... 76

5.2 Annual production of Diptera taxa in three main habitats (bedrock, Podostemum ceratophyllum, and Justicia americana) of Hargrove Shoals...... 78

5.3 Annual production of gastropod taxa and Corbicula fluminea in three main habitats (bedrock, Podostemum ceratophyllum, and Justicia americana) of Hargrove Shoals...... 81

5.4 Annual production of major taxa per square meter of bedrock, Podostemum, and Justicia habitats, respectively, as well as habitat-weighted production per square meter of Hargrove Shoals...... 83

xiv

5.5 Habitat-weighted percent contribution of major taxa to the mean annual biomass and annual production of Hargrove Shoals ...... 94

LIST OF APPENDICES

5.1 List of all benthic invertebrate taxa collected from three main habitats of Hargrove Shoals...... 102

5.2 Annual summary data for the 41 most common taxa found in three main habitats of Hargrove Shoals...... 107

xvi

CHAPTER 1

INTRODUCTION

Habitat structure and distribution have long been recognized as major determinants of ecosystem structure and function (MacArthur 1972, Southwood 1977, 1988, Huryn and Wallace

1987, Pringle et al. 1988). In streams, local geomorphology and habitat features influence the structure and function of benthic invertebrate assemblages (Hynes 1970, Benke et al. 1984,

Huryn and Wallace 1987, Grubaugh and Wallace 1995, Grubaugh et al. 1996, 1997, Hutchens et al. 2004). Regional geology and climate shape the geomorphology of stream catchments, which, in turn, determines the variety and distribution of habitats available for stream biota (Huryn and

Wallace 1987). Habitat stability and complexity influence the spatial distribution of benthic fauna, and it is widely thought that systems containing an abundant and complex array of habitat types will support a diverse assemblage of species (Wallace and Webster 1996, Vinson and

Hawkins 1998).

Habitat features alter the local conditions available to the invertebrate assemblage by influencing substrate surface area, hydraulic conditions, and retention of organic matter (Huryn and Wallace 1987). For example, continuous areas of shallow submerged bedrock can promote rapid current velocity and the entrainment of organic matter, whereas more structurally complex habitats (e.g. emergent macrophytes) tend to reduce local velocity and promote organic matter deposition (Fritz and Feminella 2003, Fritz et al. 2004). Since invertebrates are generally restricted to narrow ranges of habitat, the flow of energy through a system, such as through

plants, animals, and microbes, can be largely driven by the availability and spatial distribution of

habitat patches within the stream mosaic (Hynes 1970, Wallace and Webster 1996).

Habitat-specific secondary production

Habitat-specific studies of invertebrate (or secondary) production have allowed for a better understanding of the relative contributions of different habitat types to the overall energy flow through a stream system. In general, secondary production is the flow of biomass through a heterotrophic population or assemblage through time, and estimates of production provide the most comprehensive view of the roles of various taxa to overall functioning of the ecosystem

(Benke and Huryn 2006). Simple estimates of abundance, density, and biomass only provide instantaneous ‘snapshots’ of the invertebrate assemblage, whereas secondary production incorporates the individual growth, survivorship, and turnover rates of populations into one integrated variable (Benke 1984, 1993).

Many studies have illustrated the value of applying a habitat-specific approach to the study of stream macroinvertebrate assemblages, and in some streams, relatively rare habitats, such as submerged wood (snags), have been shown to account for disproportionate amounts of assemblage production (e.g., Benke et al. 1984). The habitat-specific approach has also proven useful in studying the contributions of different habitats to the production of bedrock shoals,

which are known to be highly heterogeneous in terms of habitat for invertebrates. For example,

Grubaugh et al. (1997) demonstrated that cobbles and bedrock covered with the submersed

macrophyte Podostemum ceratophyllum in Little Tennessee River shoals supported some of the

highest estimates of macroinvertebrate production ever recorded for any habitat. The present

study as well as others preceding it (e.g., Nelson and Scott 1962, Grubaugh and Wallace 1995)

highlighted the potential of bedrock shoal reaches to be zones of high production throughout a

river’s course, particularly in the presence of Podostemum.

Bedrock shoals and the Cahaba River

Bedrock shoals are a distinctive channel form that occurs most prevalently in river

sections as they transition across the Fall Line from upland physiographic provinces (e.g.,

Piedmont, Valley and Ridge) to the lowland Coastal Plain. This Fall Line transition is manifested in river geomorphology as series of waterfalls, rapids, or exposed bedrock. The increased width and shallow depths of mid-order shoal reaches allows for a greater influx of solar energy and potential for increased primary production of algae, moss, and vascular plants compared to lower-order streams (Vannote 1980, Grubaugh and Wallace 1995). Consequently, shoals often possess a variety of aquatic macrophyte species, some of which can cover a substantial portion of available substrate while increasing surface area for microbial, macroinvertebrate, and vertebrate assemblages (Grubaugh and Wallace 1995, Hutchens et al.

2004). Thus, rocky shoals possess several characteristics that enhance their capacity to support

high macroinvertebrate production.

Unfortunately, the number of medium to large (i.e., mean discharge >10 m3/s) free-

flowing rivers possessing significant shoal reaches was dramatically reduced throughout the

contiguous United States during the 20th century (Williams et al. 2008). Most of these rivers

became highly regulated for navigation, flood control, hydropower, or water supply (Stanford

and Ward 1979, Benke 1990, Benke and Cushing 2005). In the southeastern US, dams often

were constructed at the Fall Line of major rivers where rocky shoals were concentrated.As a

result, these rivers lost much (if not all) of these habitats along with many unique aquatic species

(Lydeard and Mayden 1995). Given the increasingly rare nature of shoal reaches, it is prudent to

seek an understanding of the distribution of remaining shoal habitats and examine the

relationships between these habitats and their associated organisms.

The present study focuses on the Cahaba River, an unregulated medium-sized river in

Central Alabama. It possesses some of the most expansive bedrock shoals remaining in

Alabama, and probably the entire Southeast from Virginia to Mississippi. The Cahaba shoals

provide an excellent template from which to study the relationships between habitat

characteristics and benthic assemblage structure and function. Studies on rivers such as the

Cahaba also represent a comparative baseline condition by which to assess the effects of ongoing and increasing disturbance from upstream urban development in the watershed (Shepard et al.

1994, Grubaugh and Wallace 1995). In addition, the macroinvertebrates of the Cahaba River shoals have yet to be intensively studied in relation to its variety of habitat resources, and this dissertation, in part, attempts to fill this gap in our knowledge. This study was conducted on the longest of the Cahaba shoals (Hargrove Shoals), within the Cahaba River National Wildlife

Refuge (Figs 1.1 and 1.2).

The central question of this dissertation is: How do the major habitats of a Cahaba River shoal influence the distribution and secondary production of the macroinvertebrate assemblage?

Accordingly, the following chapters focus on several related aspects of this question. Chapter 2 provides a detailed description and quantification of the major habitats across the reach (bare bedrock, bedrock covered by Podostemum ceratophyllum, and depositional areas dominated by

Justicia americana), and examines the temporal distribution of Podostemum and Justicia

biomass. Chapter 3 describes a new method utilizing a tethering strategy to estimate in situ

growth rates of a diverse pleurocerid snail assemblage, an important component of this shoal

system. Regression models developed with this method are needed to estimate the production of

this important group via the instantaneous growth method. Chapter 4 describes the distribution

of the non-native Asiatic clam (Corbicula fluminea) among 3 primary shoal habitats, and

estimates its density, biomass, and secondary production in sediments associated with Justicia americana, its principal habitat in this reach. Finally, Chapter 5 represents the culmination of the

3 preceding chapters by characterizing the distribution and production of macroinvertebrate

assemblages associated with the 3 major shoal habitats, and estimating reach-scale production

for Hargrove Shoals.

Fig. 1.1. The Cahaba River Basin, central Alabama. The study site, Hargrove Shoals, is outlined by a small rectangle.

Fig. 1.2. Hargrove Shoals, Cahaba River, Bibb Co., Alabama. This photograph depicts a view of the upstream boundary of Hargrove Shoals from its western bank in late spring. The shorter and more widespread vegetation is American waterwillow (Justicia americana), and the flowering plant is the shoals spiderlily (Hymenocallis coronaria), known locally as the “cahaba lily”.

CHAPTER 2

QUANTIFICATION OF BENTHIC HABITATS AND VEGETATION DYNAMICS IN A FALL LINE RIVER SHOAL

Abstract

Most large rivers in the Mobile River Basin have been heavily altered by human

development, which has caused the destruction of many sections of bedrock shoals. The largest

remaining exception to this trend is the Cahaba River whose expansive shoals remain intact.

Given the increasingly rare nature of shoal reaches and their tendency to be centers of diversity

and production, it is important to seek a greater understanding of the distribution of the

remaining shoal habitats and to examine the relationships between these habitats and their

associated organisms. The goals of this study were to (1) characterize the variety and relative proportions of available habitat for vegetation and aquatic macroinvertebrates in a Cahaba River shoal, and (2) to examine the temporal dynamics of the two dominant shoal macrophytes:

American water-willow, Justicia americana, and hornleaf riverweed, Podostemum

ceratophyllum. The four most abundant shallow habitats were: bedrock > Justicia > boulder >

Podostemum, accounting for a combined 79.5% of shoal area. Gravel, sand, mud, snags, pools,

and the shoals spiderlily (Hymenocallis coronaria) together accounted for the remaining ~20%

of shoal area. Vegetated habitats covered 35.1% of shoal area. Mean annual macrophyte

biomass (± 1SE) was: Justicia = 782.15 ± 96.8 g AFDM m-2, and Podostemum = 108.12 ± 15.4.

Justicia showed no significant variation in total biomass over time, but emergent tissues

exhibited a distinctive senescent period during winter months. Podostemum also exhibited a

decreased biomass during winter. This study represents the most intensive survey of the

diversity of habitats found in a Cahaba River shoal. Riverine macrophytes such as those in this study play a large role in creating this habitat diversity, and likely contribute to the maintenance of the corresponding faunal diversity found there.

Introduction

Most medium and large rivers of the coterminous United States became highly regulated during the 20th century, particularly for navigation, flood control, hydropower, and water supply

(Stanford and Ward 1979, Benke 1990, Benke and Cushing 2005). In the southeastern US, dams

often were constructed at the Fall Line of major rivers where they flowed across rocky shoals

from one physiographic province (e.g., Piedmont, Valley and Ridge) to another (Coastal Plain).

As a result, these rivers lost much of their preexisting shoal habitats – shallow river reaches

underlain with exposed bedrock substrate. This loss of habitat unfortunately caused a

corresponding loss of many unique aquatic species that depended upon shoals (Lydeard and

Mayden 1995). Given the increasingly rare nature of shoal reaches, it is important to better understand of the distribution of the remaining shoal habitats and to examine the relationships between these habitats and their associated organisms. This better understanding could lead to more informed management decisions and conservation efforts.

Shoal reaches of southeastern rivers are heterogeneous, with increased width and shallow depths allowing for a greater influx of solar energy and potential for increased primary production of algae, moss, and vascular plants (Vannote 1980, Grubaugh and Wallace 1995).

Consequently, shoals often possess several aquatic macrophyte species, some of which can cover a substantial portion of available substrate while increasing surface area for usage by microbes,

macroinvertebrates, and vertebrate assemblages (Grubaugh and Wallace 1995, Hutchens et al.

2004). Given these characteristics, shoal reaches can represent concentrations of diversity and

production at the scale of the river, while shoal macrophytes can represent concentrations of

diversity and production at the scale of the shoal reach (Grubaugh et al. 1996, 1997).

Previous work has recognized the importance of 2 macrophyte species in particular: the

submersed Podostemum ceratophyllum (Michx.) (hereafter Podostemum or POD) and the

emergent Justicia americana L. (Vahl.) (hereafter Justicia or JUS). Both Podostemum and

Justicia are significant modifiers of shoal habitat, providing stable refugia for invertebrates

during high flow (Minshall 1984, Grubaugh and Wallace 1995, Fritz et al. 2003, 2004).

Grubaugh and Wallace (1995) observed a positive correlation between the biomass of

Podostemum and the abundance and production of the associated macroinvertebrate community in the Middle Oconee River, Georgia. Similarly, Grubaugh et al. (1997) recognized the influence of Podostemum on invertebrate community composition and production in the Little

Tennessee River, North Carolina, one of the highest estimates of secondary production to date.

Hutchens et al. (2004) experimentally determined a strong positive correlation between

Podostemum surface area and invertebrate abundance and biomass in the same river. Fritz et al.

(2004) demonstrated that Justicia increases streambed stability with its root/rhizome system, and

reduces local current velocity with its vertical stems in Halawakee Creek, a third-order tributary of the Chattahoochee River, Alabama. The reduction of velocity caused accumulation of sediments and organic matter, creating favorable habitat for many invertebrates adapted to depositional conditions (Fritz et al. 2004).

This paper focuses on the Cahaba River, the largest unregulated tributary in the Mobile

River Basin. Besides the Mississippi River, the Mobile is the largest system (by discharge)

flowing into the Gulf of Mexico, and is renowned for its high aquatic biodiversity (Ward et al.

1995). Unfortunately, most major tributaries have been altered by navigation structures or hydropower dams, causing the inundation or removal of many shoal reaches (Benke 1990, Ward et al. 1992). The largest exception is the Cahaba River, whose expansive shoals remain intact.

The Cahaba’s unregulated nature and its many shoal reaches are likely a major reason for the persistence of many rare and unique species including the shoals spiderlily, and several federally listed snails, mussels, and fishes (Ward et al. 1992, Lydeard and Mayden 1995). This river thus provides an excellent template from which to study the relationships between physical habitat structure and community structure and function. Furthermore, a quantitative study of a Cahaba shoal habitats represents a relative baseline by which to assess the effects of ongoing and increasing disturbance from upstream economic development in the basin (Shepard et al. 1994,

Grubaugh and Wallace 1995).

The goals of this study were (1) to characterize the relative proportions of available habitat for aquatic plants and animals in a Cahaba River shoal, and (2) to examine the temporal biomass dynamics of the 2 dominant shoal macrophytes: Justicia americana, and Podostemum ceratophyllum.

Materials and Methods

Study site

The Cahaba River is a free-flowing 6th order river, spanning 2 distinct physiographic provinces throughout its 304 km length (Chapter 1, Fig. 1.1). The Upper Cahaba flows through the Valley and Ridge province, composed of predominantly limestone/dolomite geology, whereas the Lower Cahaba flows through the Coastal Plain province, characterized by shifting alluvial sediments (Ward et al. 2005). The transition between the two provinces, known as the

Fall Line, is characterized by a series of exposed bedrock shoals. In this section the river

channel becomes especially wide (up to ~150 m) and shallow, increasing its exposure to the sun

and allowing for the growth of large macrophyte beds. These beds are predominantly composed

of three species: the emergent American water willow (Justicia americana) and shoals spiderlily

(Hymenocallis coronaria), and the submergent hornleaf riverweed (Podostemum ceratophyllum).

The climate of the Cahaba Basin is humid-subtropical (mean annual temperature 16.7

2 Cair, 18.1 Cwater), with average annual rainfall of 138 cm. Basin area is 4730 km . Mean daily

discharge is 80 m3/s (USGS gauge, Centreville, AL), with mean annual high flows reaching 177 m3/s in winter-early spring, and low flows reaching ~25 m3/s in early autumn (Ward et al. 2005).

This study was conducted in a reach of the Cahaba River known as Hargrove Shoals (33

04’ 10” N, 87 04’ 43” W, Bibb Co., AL), the last and amongst the largest of the shoals in the

Fall Line series. This reach lies within the federally owned Cahaba River National Wildlife

Refuge (NWR), and is ~2 km upstream of the confluence with the Little Cahaba River. The

basin area draining into the NWR is approximately 1536 km2, with a mean discharge of ~33 m3/s

(Pierson et al. 1989). Channel dimensions of the shoal are ~550 m in length with a mean width

of 141.5 m. Channel geomorphology consists of multiple bands of shallow (mean depth = 20 cm) bedrock/macrophyte beds separated by several short pools.

Water chemistry near the sampling location is influenced by nutrient runoff from the

greater Birmingham area with NO2-N + NO3-N ranging from 0.25 to 0.30 mg/L, and PO4-P ranging from <0.003 to 0.025 mg/L (Ward et al. 2005).

Habitat quantification

Shoal habitats were quantified in late summer-early autumn over 2007, 2008 when discharge was near baseflow. A multiple point-transect survey technique was used, similar to that described by Bowden et al. (2006). Starting at the upstream boundary of the shoal, transects were established at 10 m intervals. Habitat type (and depth in 2008) was then recorded at 5 m points along each transect. Habitat categories included: bare-bedrock (BR), dry-bedrock (BRD),

Podostemum (POD), Justicia (JUS), dry Justicia (JUSD), Hymenocallis (HYM), boulder (BO),

dry boulder (BOD), gravel (G), sand (S), mud (M), snags (SG), and deep pool (DP). The

number of “hits” representing each habitat type was then converted to estimate percent cover

(C%) by each habitat with the following equation: C%=(Ni/Nt) x 100 where Ni is the number of

“hits” represented by habitat type i, and Nt is the total number of point observations (Bowden et

al. 2006). The resulting habitat cover estimates were used to make the final determination of

habitats to be sampled and, ultimately, to weight vegetation biomass, as well as invertebrate

biomass and secondary production estimates in further studies (Huryn and Wallace 1987, Wohl

et al. 1995).

Adjustments were made to the 2007 survey results to allow for comparison with the 2008 survey. This modification was necessary because of the exclusion of the pool habitat category in

2007. All 2007 habitat percentages were adjusted by multiplying each estimate by: 100 - %

DP2008 based on the assumption that the percentage of pool habitats remained constant between the 2 years. Standard errors were calculated based on the variation of each habitat percentage

across all transects.

Vegetation sampling

A stratified random design was used to quantitatively sample Justicia and Podostemum, the 2 major vegetation habitats, on an approximately monthly basis for 1 y from June 2008 to

June 2009 (Cummins 1962, Wohl et al. 1995). The shoals spiderlily (Hymenocallis coronaria) was not sampled because it was usually embedded within Justicia habitat, it occupied a much smaller area on the shoals than the other 2 species, and sampling would have been too destructive to this “charismatic” species. Five replicate Surber samples (0.09 m2; mesh size 250

μm) were collected from each habitat for each date. Before sampling each habitat patch, velocity and depth were measured with a velocity probe (Marsh-McBirney Inc., Frederick, MD) and meter stick. Velocity and depth was also measured for 5 patches of bare bedrock for comparison with vegetated habitats. Justicia was sampled by placing the Surber frame over a patch of vegetation, clipping all emergent growth, and then extracting the root/rhizome-mass with all associated sediments and invertebrates. Podostemum was sampled by placing the surber frame over the substrate and scraping all vegetation into the net with a putty knife and stiff-bristled brush (Grubaugh et al. 1997). The emergent portion of Justicia samples was packaged in plastic bags and transported on ice, whereas all other material was fixed in the field with 70% ethanol for transport to the laboratory.

Estimation of macrophyte biomass generally followed Grubaugh et al. (1997). After thorough separation of vegetation from animals and sediments, aquatic vegetation was dried to a constant weight at 60 C for 5-7 d, then ashed at 550C for 12 h and reweighed to obtain ash-free dry mass (AFDM). Emergent Justicia vegetation (JUSemerge) was dried and ashed separately from the root mass to obtain estimates of both total plant biomass (JUStotal) as well as root/rhizome biomass (JUSsub).

All biomass, velocity, and depth data were ln(x+1) transformed prior to analysis. One-

way ANOVA was applied to each habitat category separately, and also to JUStotal (JUSsub +

JUSemerge), to examine the potential variation of biomass through time. Mean annual biomass for

each macrophyte was then adjusted based on the relative abundance of each habitat (as estimated

from the habitat quantification surveys) to convert habitat-specific estimates to shoal area estimates.

Results

Habitat quantification

For both survey years, 27 transects were established along the length of the shoals, with

727 habitat points recorded in 2007 and 703 points recorded in 2008. The four most abundant

shallow habitats were: BR>JUS>POD>BO in 2007 and BR>JUS>BO>POD in 2008 (Fig. 2.1).

In 2007, BR accounted for 36.9%, followed by JUS, POD, and BO (31.4, 9.8, and 5.1%,

respectively). In 2008, BR accounted for 37.7%, followed by JUS, BO, and POD (23.6, 10.8,

and 8.0, respectively). Vegetated habitats (JUS+POD+HYM+JUSD) covered 43.2% of

Hargrove shoals in 2007 and 35.1% in 2008, which amounted to 33,620 m2 in 2007 and 27,317

m2 in 2008 of the total shoal surface area of ~77,825 m2. Of the remaining habitats, DP

accounted for 6.4% in 2008, and was assumed to cover the same percent area in 2007 (Fig 2.1).

G, S, M, and SG accounted for a combined 4.0% in 2007 and 5.8% in 2008. Despite its

conspicuous display of spring blooms, HYM covered only 1.7 and 1.1% of shoals area in 2007

and 2008, respectively.

50 2007 40 2008 30

20 % Cover 10

S D D G S D M P BR BO SG D JU O YM US P BR H BOD J

Fig. 2.1. Mean (± 1 SE) percent cover (by area) of the variety of habitat types occurring in Hargrove Shoals in late summer 2007 and 2008. Habitat categories included: bare bedrock (BR), dry bedrock (BRD), Podostemum (POD), Justicia (JUS), dry Justicia (JUSD), Hymenocallis coronaria (HYM), boulder (BO), dry boulder (BOD), gravel (G), sand (S), mud (M), snags (SG), and deep pool (DP). Percent coverage by DP in 2007 was assumed to equal that recorded for 2008 (see text for details).

Vegetation biomass and temporal distribution

-2 Mean annual biomass (g AFDM m ± 1 SE) was: JUStotal = 782.15 ± 96.8, JUSsub =

637.94 ± 49.9, JUSemerge = 144.21 ± 46.8, and POD = 108.12 ± 15.4. Monthly mean biomass (g

-2 AFDM m ) values ranged from: JUStotal = 443.93 to 1217.22, JUSsub = 443.93 to 905.92,

JUSemerge = 0 to 377.61, and POD = 42.41 to 197.85 (Figs. 2.2-2.4).

JUStotal and JUSsub showed no significant variation of biomass with time (One-way

ANOVA, p = 0.1971 and 0.3095, respectively) (Fig 2.2), but JUSemerge biomass varied

significantly with time, exhibiting a distinctive senescent period during the winter months (One-

way ANOVA, p ≤ 0.0001) (Fig 2.3). POD also showed a decreased biomass during winter (One-

way ANOVA, p = 0.0031) (Fig. 2.4), though only November and January were significantly

different from the observed maximum biomass of June 2008 (both p < 0.03, Bonferroni post hoc

tests).

Mean current velocity in POD was significantly higher than in BR or JUS habitats (One-

way ANOVA, p ≤ 0.0001), while velocity in BR was higher than in JUS (p = 0.021). Mean depth was not significantly different between POD and JUS or between BR and JUS, while POD occurred in slightly shallower areas of bedrock (p = 0.041) (Table 2.1). Because of temporal variation in discharge, all habitats showed significant variation in velocity and depth by date

(One-way ANOVA, p ≤ 0.0001).

Mean annual biomass estimates (g AFDM m-2 ± 1 SE) adjusted for relative habitat

proportions for 2007 were (in g AFDM m-2 shoal SA ± 1 SE): JUS = 245.29 ± 53.31, POD =

10.56 ± 3.99. For 2008, adjusted mean annual biomass was JUS = 180.24 ± 52.31, POD = 8.61

± 3.36 (Table 2.2).

) -2 1400 emergent 1200 submergent 1000 800 600 400 200 0

8 08 8 8 09 Mean biomass (g AFDM m (g Mean biomass -08 - 0 0 -09 - l g-08 - v-08 c- un Ju u ct-0 o e ay J A Sep O N D Jan Feb-09 Mar-09 Apr-09 M

Fig. 2.2. Mean biomass (± 1 SE) of Justicia americana in Hargrove Shoals from June 2008 to June 2009.

) -2 450 400 350 300 250 200 150 100 50 0

8 8 8 8 8 9 9 9

Mean biomass (g AFDM m AFDM (g biomass Mean 0 0 0 -0 -0 0 -09 n- p c-0 b r- n- u Jul- e u J Aug-08 S Oct-0 Nov-08 De Jan-09 Fe Ma Apr May-09 J

Fig. 2.3. Mean biomass (± 1 SE) of emergent Justicia americana in Hargrove Shoals from June 2008 to June 2009.

) -2 250 225 200 175 150 125 100 75 50 25 0

8 8 8 8 8 9 9 Mean biomass (g AFDM m AFDM (g biomass Mean -0 -0 0 -0 -09 l p t- v n-09 b y-0 Ju e c o ec-0 a e un-09 Jun-08 Aug-08 S O N D J F Mar-0 Apr-09 Ma J

Fig. 2.4. Mean biomass (± 1 SE) of Podostemum ceratophyllum in Hargrove Shoals from June 2008 to June 2009.

Table 2.1. Mean annual velocity (m/s ± 1 SE) (n = 50) and mean annual depth (cm ± 1 SE) for bedrock and both macrophyte habitats. BR = bedrock, JUS = Justicia americana, POD = Podostemum ceratophyllum.

Habitat Velocity Depth

BR 0.56 ± 0.05 17.07 ± 1.06

JUS 0.39 ± 0.04 14.98 ± 1.01

POD 1.01 ± 0.15 14.07 ± 1.02

Table 2.2. Conversion of mean annual biomass (g AFDM m-2 ± 1 SE) per unit habitat to mean annual biomass (g AFDM m-2 ± 1 SE) per unit shoal surface area. JUS = Justicia americana, POD = Podostemum ceratophyllum, B = biomass, SA = surface area (m2).

Habitat proportion Mean annual B per unit shoal SA

Habitat Mean annual B 2007 2008 2007 2008

JUS 782.15 ± 96.76 0.31 ± 0.03 0.23 ± 0.04 245.29 ± 53.31 180.24 ± 52.31

POD 108.12 ± 15.42 0.10 ± 0.02 0.08 ± 0.02 10.56 ± 3.99 8.61 ± 3.36

Discussion

Habitat quantification

Within the Cahaba River, Hargrove Shoals possesses a variety of habitat types, at least 12

of which were quantified over 2 separate years. Four of these habitats (BR, JUS, POD, BO)

account for ~80 to 83% of total surface area. With the inclusion of dry bedrock, dry Justicia,

and dry boulder, the combined coverage of these 4 main habitat types increases to 87 to 88% of

total surface area. This inclusion is based on the likelihood that these dry habitats are submerged

for parts of the year, and that the surveys were completed during low discharge phases of the hydrograph. Boulder likely is similar to bedrock as a habitat type and Podostemum was sometimes found on boulder surfaces. In these cases, the habitat type was counted as

Podostemum. The 2 main vegetated habitats (Justicia and Podostemum) accounted for ~40% of

the shoal area. This relatively high fraction of vegetated shoals is significant as lotic

macrophytes often provide high primary production and exceptional habitat for

macroinvertebrates. Podostemum in particular, has been shown to support high invertebrate

production in multiple studies (e.g. Nelson and Scott 1962, Grubaugh et al. 1995, 1997), and

both plants may act as refugia during high-flow events, possibly reducing the degree of

disturbance to the benthic community (Sedell et al. 1990).

Gravel, sand, mud, and snag habitats, which are sometimes major habitat types in other

streams and rivers, all were relatively rare in Hargrove Shoals. Even so, they add to the habitat

diversity and possibly to the biological diversity of this reach. Snags are often a hotspot of

diversity in some rivers (e.g., Benke et al. 1984), but probably have a similar species

composition as the bedrock and Podostemum in the shoals. While the spiderlily is conspicuous

during flowering in the spring, it only represents a small fraction of the habitats (~1 to 2%) in

this reach. It is typically surrounded by Justicia and is likely protected by it to some extent.

Survey results generally were consistent between years, although some points of variation

stand out. Justicia decreased from 31.4 in 2007 to 23.0% in 2008, and boulder increased from

5.1 in 2007 to 10.8% in 2008. It is possible that the relative proportions of the vegetated habitats

varied between the years, but it is less likely the proportion of boulder habitats changed. Thus,

the variation between years is most likely attributable to sampling error.

Vegetation biomass and temporal distribution

Emergent Justicia demonstrated the expected seasonal pattern of growth and senescence,

total plant biomass did not. This result is not surprising, given the relatively low percentage of

total plant biomass accounted for by emergent tissues even at the peak of the growing season,

and the persistence of submerged tissues throughout the year. Timing of maximum Justicia biomass occurred in early November, which was later in the year than expected. In a study of

Justicia biomass and production in the New River, Virginia, Hill (1981) recorded maximum

biomass in August, and observed the emergent tissues to be mostly senescent by mid-October.

Given the limited timescale of our study, it is uncertain whether this sampling year was extraordinary in its timing of senescence or whether these patterns repeat with annual regularity.

Also, the relatively colder temperatures of the higher latitude New River may explain this

variation.

-2 -2 Peak biomass of JUStotal reached 1,217.3 g AFDM m , with 839.6 g AFDM m

-2 attributed to JUSsub, and 377.6 g AFDM m attributed to JUSemerge. These estimates are

considerably lower than those recorded from the New River in Virginia. Hill (1981) reported

-2 -2 peak JUStotal biomass of 2,524 g AFDM m , with 2,076.7 g AFDM m from below ground, and

447.8 g AFDM m-2 from above ground portions. It appears that these differences were at least partly due to substratum differences among the rivers. Hill reported that Justicia beds of the

New River occur in sand-gravel substrate. In contrast, the Justicia of Hargrove shoals is established atop and within the shallow cracks of exposed bedrock substrate. This limits the depth of root development, and may reduce the maximum biomass attainable in this area.

Nonetheless, Justicia proves to be an adaptive plant that can thrive across a wide range of conditions.

Peak biomass of Podostemum occurred in late June 2008 with 197.8 ± 36.3 g AFDM m-2, while the lowest biomass occurred in January 2009 with 42.4 g AFDM m-2. Temporal dynamics of Podostemum followed a similar pattern to that of Hill and Webster (1984) in the New River.

Their study reported peak biomass in late August of 244.8 g AFDM m-2 in one site and 193.8 g

AFDM m-2 in another. Grubaugh and Wallace (1995) however, reported peak biomass in

November of 1045 g AFDM m-2 from their estimates in the Middle Oconee River, Georgia, with monthly biomass ranging from 297 to 1045 g AFDM m-2. Interestingly, monthly biomass estimates in the Oconee followed no discernable seasonality, and their observed maximum stands alone as an anomalous peak. Grubaugh et al. (1997) found annual biomass of

Podostemum ranged from 122 to 212 g AFDM m-2 in the 7th order shoals of the Little Tennessee

River, North Carolina.

The higher velocities associated with Podostemum versus Justicia or bedrock further illustrate the depositional character of Justicia habitats, while Podostemum tends to occur in more erosive conditions. While Justicia may have an affinity towards slower currents, it also is known to reduce local velocity within patches (Fritz et al. 2004). This study was not designed to elucidate the cause/effect relationship of this observation.

Adjusting mean annual macrophyte biomass estimates from the habitat-scale to the

shoals-scale based on relative habitat abundance provides a better understanding of the

contributions of each species to the overall shoal reach. For Justicia, adjusted mean annual

biomass ranged from ~180 to 245 g AFDM m-2 shoal SA, whereas Podostemum estimates

ranged from ~9 to 11 g AFDM m-2 shoal SA (Table 2.2). Though Podostemum may be a relatively productive habitat in terms of its associated invertebrate assemblage (Chapter 5), its availability (8 to 10%) is much more limited than Justicia (23 to 31%). Finally, while both

plants greatly increase substrate surface area and complexity (Hutchens et al. 2004), the

expansive nature of Justicia emphasizes its potential influence on the physical and biological

processes along this section of the river.

This study has described the habitat diversity found in a bedrock shoal reach of an

unregulated river, and quantified the biomass dynamics of its major macrophytes. Riverine

macrophytes such as those in this study play a major role in creating the observed habitat

diversity found in this and other shoal reaches. Furthermore, this habitat diversity is at least

partially responsible for the maintenance of the corresponding faunal diversity for the river as a

whole, which underscores the importance of preserving the shoals that remain.

Acknowledgements

I would like to thank Mark Dedmon, Kimberly Voss, Michael Venarsky, and James

Ramsey for field assistance, Laura Frost for lab assistance, Dr. Christina Staudhammer for

statistical advice, and the Cahaba River National Wildlife Refuge for access to Hargrove Shoals.

This work was completed under permit number TE163435-0 from the USFWS granted to Dr.

Arthur C. Benke.

Literature Cited

Benke, A.C. 1990. A perspective on America’s vanishing streams. Journal of the North American Benthological Society 9:77-88.

Benke, A.C., and C.E. Cushing (eds.). 2005. Rivers of North America. Elsevier Academic Press, London. 1144 pp.

Benke, A.C., T.C. Van Arsdall, Jr., and D.M. Gillespie. 1984. Invertebrate productivity in a subtropical blackwater river: the importance of habitat and life history. Ecological Monographs 54:25-63.

Bowden, W.B., J.M. Glime, and T. Riis. 2006. Macrophytes and bryophytes Pp. 381- 406 in: Hauer, F.R., and G.A. Lamberti (eds.), Methods in Stream Ecology, 2nd edition. Elsevier Academic Press, London. 877 pp.

Cummins, K.W. 1962. An evaluation of some techniques for the collection and analysis of benthic samples with special emphasis on lotic waters. American Midland Naturalist 67:477-504.

Fritz, K.M., and J.W. Feminella. 2003. Substratum stability associated with the riverine macrophyte Justicia americana. Freshwater Biology 48:1630-1639.

Fritz, K.M., M.M. Gangloff, and J.W. Feminella. 2004. Habitat modification by the stream macrophyte Justicia americana and its effects on biota. Oecologia 140:388-397.

Grubaugh, J.W., and J.B. Wallace. 1995. Functional structure and production of the benthic community in a Piedmont river: 1956-1957 and 1991-1992. Limnology and Oceanography 40:490-501.

Grubaugh, J.W., J.B. Wallace, and E.S. Houston. 1996. Longitudinal changes of macroinvertebrate communities along an Appalachain stream continuum. Canadian Journal of Fisheries and Aquatic Sciences 53:896-909.

Grubaugh, J.W., J.B. Wallace, and E.S. Houston. 1997. Production of benthic macroinvertebrate communities along a southern Appalachian river continuum. Freshwater Biology 37:581-596.

Hill, B.H. 1981. Distribution and production of Justicia americana in the New River, Virginia. Castanea 46:162-169.

Hill, B.H., and J.R. Webster 1984. Productivity of Podostemum ceratophyllum in the New River, Virginia. American Journal of Botany 71:130-136.

Huryn, A.D., and J.B. Wallace. 1987. Local geomorphology as a determinant of macrofaunal production in a mountain stream. Ecology 68:1932-1942.

Hutchens, J.J.,Jr., J.B. Wallace, and E.D. Romaniszyn. 2004. Role of Podostemum ceratophyllum Michx. in structuring benthic macroinvertebrate assemblages in a southern Appalachian River. Journal of the North American Benthological Society 23:713-727.

Lydeard, C., and R.L. Mayden. 1995. A diverse and endangered aquatic ecosystem of the Southeast United States. Conservation Biology 9:800-805.

Minshall, G.W. 1984. Aquatic insect-substratum relationships, in V.H. Resh, and D.M. Rosenberg (eds.) The ecology of aquatic insects. Praeger Scientific, New York, NY, USA.

Pierson, M.J., W.M. Howell, R.A. Stiles, M.F. Mettee, P.E. O’Neil, R.D. Suttkus, and J.S. Ramsey. 1989. Fishes of the Cahaba River system in Alabama. Bulletin 134. Geological survey of Alabama, Tuscaloosa.

Sedell, J.R., G.H. Reeves, F.R. Hauer, J.A. Stanford, and C.P. Hawkins. 1990. Role of refugia in recovery from disturbances: modern fragmented and disconnected river systems. Environmental Management 14:711-724.

Shepard, T.E., P.E. O’Neil, S.W. McGregor, and S.C. Harris. 1994. Water quality and biomonitoring studies in the upper Cahaba River drainage of Alabama. Bulletin 160. Geological Survey of Alabama, Tuscaloosa.

Stanford, J. A., and J. V. Ward. 1979. Stream regulation in North America. Pages 215-236 in J. V. Ward and J. A. Stanford, editors. The ecology of regulated streams. Plenum Press, New York.

Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137.

Ward, A.K., G.M. Ward, and S.C. Harris. 1992. Water quality and biological communities of the mobile River drainage, eastern Gulf of Mexico region. Pp. 279-304 in: C.D. Becker and D.A. Neitzel (eds.), Water Quality in North American River Systems. Batelle Press, Columbus, Ohio.

Ward, G.M., P.M. Harris, and A.K. Ward. 2005. Gulf coast rivers of the southeastern United States. Pp. 125-178 in: A.C. Benke, and C.E. Cushing (eds.), Rivers of North America. Elsevier Academic Press, London. 1144 pp.

Wohl, D.L., J.B. Wallace, and J.L. Meyer. 1995. Benthic macroinvertebrate community structure, function and production with respect to habitat type, reach and drainage basin in the southern Appalachians (U.S.A.). Freshwater Biology 34:447-464.

CHAPTER 3

IN SITU ESTIMATION OF PLEUROCERID SNAIL GROWTH USING A TETHERING TECHNIQUE IN A FALL LINE RIVER SHOAL

Abstract

Growth models were developed for three species of pleurocerid snail, (Pleurocera vestita, Elimia clara, and E. showalteri) using an in situ tethering technique. Monthly interval

loss of individuals ranged from 3-50%. Observed growth rates ranged from 0 to 0.043 mg mg-1 d-1 for P. vestita, and from 0 to 0.038 for both E. clara and E. showalteri. Growth rates displayed

a significant negative relationship with increasing body size (all p ≤ 0.0001). Mean interval

temperature, however, showed no significant correlation with snail growth for P. vestita or E. clara. Final regression models predict instantaneous growth (mg mg-1 d-1) of individuals based

solely on body size (mg AFDM). Models yielded results similar to those reported by previous

studies, thus validating this technique for predicting instantaneous growth rates necessary for

estimating secondary production.

Introduction

Pleurocerid snails are important components of many southeastern US stream

communities. This diverse group of gastropods has been shown to dominate macroinvertebrate biomass in many streams they inhabit (Rosemond et al. 1993, Huryn et al. 1994). Grazing of periphyton by high densities of pleurocerids can substantially alter the structure, biomass, and production of this basal food resource (Rosemond et al. 1993), which, in turn, influences their

own growth and production (Cross and Benke 2002) as well as other invertebrate taxa (Hawkins

and Furnish 1987, Hill 1991).

Growth rates are needed to estimate the secondary production of this group via the

instantaneous growth method. This method generally provides the most accurate estimates of

production for populations whose cohort structure cannot be determined from field data,

especially when calculated using independently derived growth models such as those developed

in this study (Benke 1993). Often, snail growth rates are measured in the field with enclosed

chambers (e.g., Richardson et al. 1988, Cross and Benke 2002), and occasionally with mark-

recapture techniques (e.g., Huryn et al. 1994). Both methods are well suited to small streams,

but are considerably less applicable to larger (i.e., mean discharge > 10 m3/s) free-flowing rivers,

where high flows could easily destroy in-situ growth chambers. Also, the relatively expansive

channel area of larger rivers makes the recapture of marked snails unlikely.

Here, monofilament tethers were used to track individual growth of 3 common species of

pleurocerid snail (Pleurocera vestita, Elimia clara, and E. showalteri) (Fig. 3.1A) across a gradient of body size and water temperature. This tethering technique is designed to allow for the tracking of individual growth while providing access to natural substrate for grazing. This approach dramatically reduces the number of individuals required compared with a mark- recapture study. Hypotheses were that snail growth is (1) negatively related to body mass, and

(2) positively related to temperature. Regression models incorporating these two variables were used to estimate pleurocerid secondary production when combined with biomass from quantitative field sampling (Chapter 5).

AW P. vestita E. clara E. showalteri

A. B. C.

ID tag

tethered snail D. E.

Fig. 3.1. Description of snail tethering technique. Species examined in this study (A). Gluing monofilament to P. vestita (B). Measuring aperture-width (AW) (C). Configuration of tethering apparatus (D), and placement on submerged bedrock (E).

Materials and Methods

Study site

This study was carried out in a shoal reach of the 6th order Cahaba River in Bibb Co., AL

known as Hargrove Shoals (33 04’ 10” N, 87 04’ 43” W). This reach is found within the

Cahaba River National Wildlife Refuge, established in 2002. Hargrove Shoals is one of the largest in a series of exposed limestone/dolomite bedrock shoal reaches found along the Fall

Line, a transitional zone between the upland Valley and Ridge and the lowland Coastal Plain physiographic provinces (Ward et al. 1992). Mean discharge at the study site is 33 m3/s (Pierson

et al. 1989), and mean water temperature is 18.1 C (Ward et al. 2005). The pH of the Cahaba is

circumneutral at (7.7), and alkalinity is high (84 mg/L as CaCO3) because of the underlying

carbonate geology (Ward et al. 2005).

Hargrove Shoals, like other shoals in the Fall Line series, is characterized by wide and

shallow channel dimensions (mean width = 141.5 m, mean depth = 20 cm), which allow for high

benthic solar exposure and consequently high production of algae and macrophytes. These

qualities provide ideal conditions for invertebrate grazers of periphyton, especially snails from

the family Pleuroceridae. Pleurocerids are known to occur in high diversity and abundance throughout the Cahaba shoals (Stein 1976).

Snail growth study

Thirty individuals of each snail species (Fig. 3.1A), representing a range of body size, were tethered in the field with monofilament line (20 cm, 4# test) by means of cyanoacrylate glue (i.e. ‘super-glue’) (Fig. 3.1B), and then anchored to 0.5 m segments of rebar (3 ind./rebar) with cable ties (Fig. 3.1D). An identification tag made from engraved polyethylene plastic was assigned to each individual and attached to the rebar in the same manner (Fig. 3.1D). The rebar

apparatuses were then placed horizontally on bedrock in shallow water (~10-20 cm) (Fig. 3.1E).

Initial aperture widths (AW) were measured upon capture and then subsequently measured at

~30 d intervals for a period of 1 y (Fig. 3.1C). Dead or missing snails were replaced when necessary. After each interval measurement, snails were replaced onto different patches of bedrock to reduce the possibility of overgrazing and food scarcity. AW measurements were converted to ash-free dry mass (AFDM, mg) using a length-mass regression developed for each species with the methods of Huryn et al. (1994). Daily instantaneous biomass growth rates for an interval were calculated as: g = ln(Wt+∆t/Wt)/Δt where g = daily instantaneous growth (mg

mg-1 d-1), W = individual AFDM (mg), and Δt = time (d). Water temperature data was continuously recorded over the study at 30 min intervals with in-stream data loggers (HOBO®

Onset Computer Corporation) for the duration of the study. Growth values from small Elimia

spp. (<6 mm AW) were pooled for analysis because of uncertainty of species identification.

To develop predictive models of snail growth, instantaneous growth rates were ln(x) transformed and regressed against the geometric mean of ln(x) transformed individual body mass

and mean water temperature during that interval. Regression models were developed based on

significance of variables at α = 0.05, and also to meet the assumption of homogeneous variance.

Initially, multiple regression analyses included both body size and mean water temperature as

predictive variables of instantaneous growth. Significance of each variable was then evaluated at

α = 0.05 and removed from the final model if α > 0.05. Models were developed for each of the

three species separately, along with compiled genus (Elimia spp.) and family-level

(Pleuroceridae) models.

Results

The species-specific length-mass relationships were AFDM = 0.0031AW4.36 for P.

vestita, AFDM = 0.0058AW3.88 for E. clara, and AFDM = 0.0052AW3.97 for E. showalteri, where AFDM = ash-free dry mass of soft tissue (mg), AW = width of shell at the aperture (mm)

(Table 3.1). Monthly interval loss of individuals ranged from 20 to 50% for P. vestita, 3 to 27% for E. clara, and 10 to 33% for E. showalteri. Observed growth rates (as mg mg-1 d-1) ranged

from 0 to 0.043 for P. vestita, and 0 to 0.038 for both E. clara and E. showalteri. Mean interval

temperatures (C) ranged from 9.1 to 29.2 for P. vestita, 9.6 to 28.7 for E. clara, and 9.6 to 25.3

for E. showalteri (Fig. 3.2). Temperature ranges vary by species because interval-growth

measurements were often conducted across a several-day time span.

Initial examination of the residuals from the multiple regression models revealed a

heterogeneous variance caused by high numbers of zero growth values recorded across the entire

body size gradient and temperature conditions. Because of this anomaly, growth models were

reformulated using a subset of the original data that excluded zero growth values yielding the

following models: P. vestita: ln(g) = - 4.126 - 0.519ln(bs) - 0.010T, E. clara: ln(g) = - 5.147 -

0.193ln(bs) - 0.011T, E. showalteri: ln(g) = - 6.570 - 0.260ln(bs) + 0.104T, where g =

instantaneous growth (mg mg-1 d-1), bs = body size (mg), and T = mean interval temperature

(C).

Further analysis of the above multiple regression models revealed a significant negative relationship between body size and growth rates (all p ≤ 0.0001, Table 3.2). Mean interval

temperature however, showed no significant correlation with snail growth for 2 of the 3 species.

Only the growth rates of E. showalteri were positively correlated with mean interval temperature

(r2 = 0.061, p = 0.0089). Simple regression plots of instantaneous growth with body size and

temperature illustrate these relationships (Fig. 3.3). To maintain consistency between models,

and because temperature only explained 6.1% of model variance for E. showalteri, mean interval

temperature was excluded from all final regression models. Final regression models predict

instantaneous growth (mg mg-1 d-1) of individuals based solely on body size (mg AFDM) (Table

3.2). These reformulated (i.e., without zero growth values) simple regression models predict

ranges of growth (mg mg-1 d-1) from 0.001 to 0.047 for P. vestita, from 0.002 to 0.007 for E. clara, and from 0.001 to 0.010 for E. showalteri (Table 3.2).

Table 3.1. Length-mass regression relationships for Pleurocera vestita, Elimia clara, and Elimia showalteri. Equations are based on a power function AFDM = a AWb, where AFDM = ash-free dry mass of soft tissue (mg), AW = width of shell at the aperture (mm), n = number of snails in model, Range = range of body sizes (AW) in model (mm), a and b are constants (± 1 SE), % Ash = ash content (± 1 SE) of soft tissue.

Snail Taxon n Range a b r2 % Ash

P. vestita 21 3.8-11.1 0.0031 4.36 ± 0.19 0.96 13.92 ± 1.48 E. clara 25 3.9-11.4 0.0058 3.88 ± 0.15 0.97 10.67 ± 0.86 E. showalteri 33 3.6-13.1 0.0052 3.97 ± 0.14 0.96 10.39 ± 0.68

15 Degrees Celsius 10

0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun 2008 2009

Fig. 3.2. Average weekly water temperature of Hargrove Shoals, Cahaba River, Bibb Co., AL.

-2

-4

-6

-8 P. vestita -10

-2

-4 ) g -6 ln(

-8 E. clara -10 -2

-4

-6

-8 E. showalteri -10 -4 -2 0 2 4 6 5 1015202530

ln(bs) T

Fig. 3.3. Regression plots of body size and temperature with growth for Pleurocera vestita, Elimia clara, and Elimia showalteri, where bs = body size (mg AFDM), T = mean interval temperature (C), and g = daily instantaneous growth (mg mg-1 d-1). Plots exclude data points where g = 0. Lines through data points represent the least squares regression line, and equations for body size-growth regressions are reported in Table 3.2.

Table 3.2. Simple regression model parameters and statistics of growth vs. body size for each species as well as combined genus and family-level models. Regressions predict instantaneous growth (mg mg-1 d-1) based on body size (mg AFDM) and are of the form ln(g) = a + bln(bs), where g = instantaneous growth (mg mg-1 d-1), bs = body size (mg), and a and b are fitted constants. Range of body size in model = size range (mg AFDM) of individuals used to develop each model, Range of predicted g = growth rates (%d-1) predicted by regression for the range of body sizes used in model. Asterisks denote models where 0 growth values were excluded. For models including 0 growth values, g was ln(x+1) transformed, and body size was ln(x) transformed, whereas both variables were ln(x) transformed for models excluding 0 growth values. All models were significant at p ≤ 0.0001.

Range of body Range of Snail Taxon n a b size in model predicted g F r2 P. vestita 109 0.013 -0.002 0.08-133.43 0.32-0.80 39.6 0.27 P. vestita* 81 -4.346 -0.516 0.08-133.43 0.10-4.71 55.2 0.41 E. clara 180 0.006 -0.001 0.12-255.73 0.05-0.39 33.3 0.16 E. clara* 124 -5.325 -0.201 0.15-242.47 0.16-0.71 20.2 0.14 E. showalteri 114 0.007 -0.001 0.12-208.33 0.17-0.49 22.7 0.17 E. showalteri* 77 -5.183 -0.296 0.15-198.80 0.12-0.98 20.5 0.21 E. spp. 242 0.006 -0.001 0.12-255.73 0.05-0.39 51.7 0.18 E. spp.* 162 -5.338 -0.205 0.15-242.47 0.16-0.71 25.8 0.14 Pleuroceridae 351 0.008 -0.001 0.08-255.73 0.25-0.55 78.7 0.18 Pleuroceridae* 243 -5.074 -0.296 0.08-242.47 0.12-1.31 65.4 0.21

Discussion

The b values calculated from length-mass regression models ranged from 3.88 to 4.36

(Table 3.1), and were larger than most reported estimates (e.g., Huryn et al. 1995, Benke et al.

1999, Tolley-Jordan 2008). In general, b values tend to be ~3 across a wide range of taxa, and

most previous estimates for pleurocerids range from 2.42 to 3.32 (Benke et al. 1999), although

one study (Rosemond et al. 1993) has reported a value similar those presented here for E. clavaeformis (b = 3.98). These high b values imply that the increase in soft tissue biomass of

these snails increases at a greater than cubic relationship to aperture width. Given that standard

procedure was strictly followed (Huryn et al. 1994), and that high r2 values (≥0.96) demonstrate

the good fit of these models to the data, we considered them to be valid.

Although temperature was predicted to have an influence on the growth of snails, the

relationship was not significant, and its potential effect was likely overwhelmed by the influence

of body size. Previous studies of snail growth have also used body size as the sole predictor. In

a study of pleurocerid growth in the headwaters of an adjacent basin, Huryn et al. (1995)

developed predictive growth regression models similar to ours based on body size alone.

Additionally, while Hall et al. (2006) found temperature to be a significant factor predicting

growth of the invasive hydrobiid snail, Potamopyrgus antipodarum, they noted that body size

accounted for most of the variation observed in their data.

Percent loss between intervals, due largely to mortality, was occasionally very high (up to

50%), as was the frequency of zero growth observations across all body sizes. Regardless of

size, individuals can become entangled, effectively reducing the available radius of movement

and possibly access to food, which can lead to slower growth. However, while is it reasonable to

expect mature snails to exhibit negligible or zero growth, this is unlikely to be the case for

juveniles. Small individuals are more likely to experience reduced freedom of movement simply

as a result of the presence of the tether, which suggests the possibility of a ‘tether effect’ that

may influence the growth of small snails disproportionately. This reasoning justified the

reformulation of the growth models excluding all observed zero growth observations.

Reformulated models undoubtedly excluded a number of valid zero growth estimates, but this

was the only objective strategy by which to exclude data points. Residuals of the resulting

alternate models exhibited greatly improved homoscedasticity, indicating a more valid

application of regression analysis, and these models likely represent a more accurate reflection of

natural conditions.

The tethering technique developed here provided reasonable estimates of growth (ranging

from 0.16 to 0.71 %/d for E. clara, 0.12 to 0.98 %/d for E. showalteri, and 0.10 to 4.71 %/d for

P. vestita, Table 3.2) when compared with previous estimates for pleurocerid snails. In two field enclosure studies conducted in smaller tributaries of this basin, Richardson et al. (1988) reported a lower range of growth (0.01 to 0.16 %/d) for both E. clara and E. cahawbensis, while Tolley-

Jordan (2008) recorded maximum growth rates of small E. cahawbensis (4.75 %/d) similar to the

highest values observed in our study. Using a mark-recapture strategy, Huryn et al. (1994, 1995)

developed similar regression models of growth for two species of Elimia with estimates ranging

from ~0.12 to 0.41 %/d.

Several studies have demonstrated a wide range of possible growth rates among

pleurocerids depending upon several factors, such as individual size/age, and relative food

abundance. In a study conducted in Walker Branch, Tennessee, Rosemond et al. (1993) reported

growth of E. clavaeformis to be ~0.38 %/d under ambient nutrient conditions while increasing to

~0.72 %/d under conditions of elevated Nitrogen and Phosphorus. In a study of pleurocerid

competition, Cross and Benke (2002) reported growth rates for E. cahawbensis and E. carinifera

ranging from ~0 %/d for larger individuals up to 2.50 %/d for small snails under enriched

periphyton conditions. These studies show that pleurocerid growth rates can vary considerably

depending on the individual age/size and food availability. Based on these wide ranging results, it appears that the ranges of predicted growth rates from our models encompasses these values, and are reasonable.

These comparisons of our model predictions to previous studies provide confidence that our methods produced realistic results. The fact that small snails in Hargrove Shoals exhibit relatively fast growth rates (especially P. vestita) suggests an abundance of periphyton and lack

of food limitation commonly observed in lower-order woodland streams of the southeast US

(e.g. Hill 1992, Rosemond et al. 1993, Cross and Benke 2002). This is not surprising given the wide unshaded channel and relative abundance of nutrients from upstream inputs (i.e.,

Birmingham).

Despite the lack of an expected temperature-growth relationship, and potential tether effects on small snails, these models produce results well within the realm of previous observations for this group. Ultimately, these models will prove useful for estimating the secondary production of this important component of the Cahaba shoals macroinvertebrate

community (Chapter 5).

Acknowledgements

I would like to thank Dr. Alexander Huryn, who originated the idea for this technique,

Kimberly Voss for field assistance, and the Cahaba River National Wildlife Refuge for access to

Hargrove Shoals. This work was completed under permit number TE163435-0 from the USFWS

granted to Dr. Arthur C. Benke.

Literature Cited

Benke, A. C. 1993 Baldi Memorial Lecture: Concepts and patterns of invertebrate production in running waters. Verh. Internat. Verein. Limnol. 25:15-38.

Benke, A.C., A.D. Huryn, L.A. Smock, and J.B. Wallace. 1999. Length-mass relationships for freshwater macroinvertebrates in North America with a particular reference to the southeastern United States. Journal of the North American Benthological Society 18:308-343.

Cross, W.F., and A.C. Benke. 2002. Intra- and interspecific competition among coexisting lotic snails. Oikos 96:251-264

Hall, R.O., M.F. Dybdahl, and M.C. Vanderloop. 2006. Extremely high secondary production of introduced snails in rivers. Ecological Applications 16:1121-1131.

Hawkins, C.P., and J.K. Furnish. 1987. Are snails important competitors in stream ecosystems? Oikos 49:209-220.

Hill, W.R. 1991. Food limitation and interspecific competition in snail-dominated streams. Canadian Journal of Fisheries and Aquatic Sciences 49:1257-1267.

Huryn, A.D., A.C. Benke, and G.M. Ward. 1995. Direct and indirect effects of geology on the distribution, biomass, and production of the freshwater snail Elimia. Journal of the North American Benthological Society 14:519-534.

Huryn, A.D., J.W. Koebel, and A.C. Benke. 1994. Life history and longevity of the pleurocerid snail Elimia: a comparative study of eight populations. Journal of the North American Benthological Society 13:540-556.

Richardson, T.D., J.F. Scheiring, and K.M. Brown. 1988. Secondary production of two lotic snails (Pleuroceridae: Elimia). Journal of the North American Benthological Society 7:234-245.

Rosemond, A.D., P.J. Mullholland, and J.W. Elwood. 1993. Top-down and bottom-up control of stream periphyton: effects of nutrients and herbivores. Ecology 74:1264-1280.

Stein, C.B. 1976. Gastropods. In Endangered and threatened species of Alabama. H. Boschung, ed. Bulletin of the Alabama Museum of Natural History 2:21-41.

Tolley-Jordan, L.R. 2008. The biology of Pleuroceridae (Gastropoda: Caenogastropoda) in the Cahaba River basin, Alabama, USA. Ph.D. dissertation, The University of Alabama.

CHAPTER 4

PRODUCTION OF THE NONNATIVE ASIATIC CLAM, CORBICULA FLUMINEA, IN A SOUTHEASTERN RIVER BEDROCK SHOAL

Abstract

The Asiatic clam, Corbicula fluminea, is probably the most widespread and abundant

nonnative invertebrate species in streams and rivers of the United States. Evidence of potential

negative effects on native bivalve diversity and abundance, however, remains equivocal. As part

of an invertebrate production study in a shoals reach of an unregulated southeastern US river,

Corbicula was a major inhabitant warranting individual attention. The goals of this study were to assess the relative distribution of Corbicula among 3 primary shoal habitats, and to estimate

density, biomass, and annual secondary production in sediments associated with the emergent

macrophyte Justicia americana, its primary shoal habitat. Corbicula was concentrated in

Justicia sediments, but had extremely low densities in patches of Podostemum and bedrock.

Size-frequency histograms indicate a 3-y lifespan with highest levels of reproduction in summer.

Mean monthly density (± 1 SE) in Justicia beds ranged from 846 (± 160) to 13,638 (± 6,154)

with a mean annual density of 3,293 (± 1,372)/m2. Mean annual biomass, production, and

annual P/B were 8.4 ± 1 g DM/m2, 14.9 g DM m-2 y-1, and 1.8, respectively. After adjusting for

proportion of shoal area covered by Justicia (~26%), density, biomass, and production were 857

± 357/m2, 2.2 ± 0.3 g DM m-2, and 3.9 g DM m-2 y-1, respectively of total shoal surface area.

This study represents the first production estimate of Corbicula in bedrock shoal habitats and

illustrates the dramatic influence of macrophytes in modifying otherwise inhospitable benthic habitat characteristics for Corbicula.

Introduction

The Asiatic clam, Corbicula fluminea, was first introduced to the Northwestern United

States in the 1920s, and quickly spread across the country to become the most widespread nonnative species found in US streams and rivers (Strayer 1999). It was first encountered in

Alabama in the Mobile River in 1962 (Counts 1986). In 1964, it was found in the Alabama

River, and then in the unregulated Cahaba River the following year (Counts 1986). With

Corbicula having been established in the free-flowing Cahaba River for >40 y, the general goal

of this paper is to examine Corbicula production and distribution within one of the river’s major

shoal habitats.

Corbicula possesses several qualities that make it a successful nonnative species. Its self

fertilizing hermaphroditic life history allows it to efficiently produce large numbers of offspring

(103-105 adult-1 y-1) (Strayer 1999). Growth rates can be rapid, with time to reproductive

maturity being reached in 3 to 6 mo, and spawning occurring 1 or 2 times per year (McMahon

and Bogan 2001). It is a generalist filter feeder, and is also known to feed with its pedal organ

on organic matter within the underlying sediments (McMahon 1983). Corbicula can

successfully colonize a wide variety of lotic and lentic habitats, as long as sufficiently

oxygenated depositional habitats are available (McMahon 1983, Payne et al. 1989). Limits to its

distribution appear to be excessively cold water temperatures (< 2C, French and Schloesser

1991), low oxygen concentration, and exposure to desiccation (Strayer 1999).

Much concern has been expressed regarding the possible negative effects of Corbicula

invasions on native bivalve diversity. Its ubiquitous distribution and potential for high density,

biomass, and production (e.g., Sousa et al. 2008), makes it reasonable to expect a significant

influence on the natural communities it invades, although supporting evidence remains equivocal

(Gardner et al. 1976, Strayer 1999, Dillon 2000). Corbicula can filter substantial amounts of

seston from the water column, exploiting the same food resource used by unionid mussels and

other native bivalves (Cohen et al. 1984, Phelps 1994). Densities reaching the tens of thousands

per m2 can dominate available benthic habitat (Gardner et al. 1976, Sousa et al. 2008) and

potentially impact native mussels. Gardner et al. (1976) reported a decline of unionids in the

Altamaha River during a Corbicula population explosion, and Vaughn and Spooner (2006) observed a negative relationship between Corbicula and unionid density on a small (0.25 m2) scale, but detected no similar relationship at a larger reach-scale. This lack of definitive evidence has caused many to conclude that Corbicula does not pose the same threat to native

fauna as the notorious Dreissena polymorpha (Dillon 2000). However, Corbicula has not been

shown to competitively exclude unionids from undisturbed streams, but it does seem appear

highly adapted to exploit heavily managed waterways where native bivalve densities have

already been reduced (Kraemer 1979, Strayer 1999, Vaughn and Spooner 2006).

The Cahaba River is well known for its free-flowing (unregulated) hydrology, its

extensive shoal habitat, and its high biodiversity (Lydeard and Mayden 1995). As part of a

habitat-specific invertebrate production analysis on a reach of the Cahaba shoals, it became

apparent that Corbicula was a major component of the benthic fauna. Given its abundance, and

reports that diversity of native bivalves in this region have been in decline for some time

(Lydeard and Mayden 1995, Neves et al. 1997), a detailed description of Corbicula production

dynamics in this shoal reach was warranted. The specific goals of this study were to describe the relative distribution of Corbicula among 3 primary shoal habitats, and to estimate its density,

biomass, and secondary production in sediments associated with the emergent macrophyte,

Justicia americana, its primary habitat within the reach. Production was selected for analysis

because it is particularly useful in assessing the relative success of species within and among habitats and ecosystems (Benke 1993, Benke and Huryn 2010). In contrast to most other

Corbicula production studies conducted in lakes or Coastal Plain rivers characterized by

unconsolidated benthic sediments (e.g., Aldridge and McMahon 1978, Sickel 1976, Gardner et

al. 1976, Stites et al. 1995), this study examines facilitation by a widespread shoal macrophyte

on the distribution and production of Corbicula within an extensive bedrock shoals.

Materials and Methods

Study site

The study was conducted along a large shoal reach of the middle Cahaba River known as

Hargrove Shoals (33 04’ 10” N, 87 04’ 43” W), which exists at the Fall Line transition from

the Valley and Ridge to the Coastal Plain physiographic provinces of central Alabama.

Hargrove Shoals lies within the Cahaba River National Wildlife Refuge established by the US

Fish and Wildlife Service in 2002 (USFWS 2002). This shoal reach is the largest among the series of exposed limestone-dolomite bedrock shoals remaining in the state, and is known to harbor several species of endemic aquatic snails (Lioplax cyclostomaformis, Lepyrium

showalteri, Leptoxis ampla) and one rare shoal macrophyte (Hymenocallis coronaria). Channel

dimensions of this reach are ~550 m in length with a mean width of 141.5 m and a mean depth of

20 cm (Chapter 2). Mean discharge at the study site is 33 m3/s (Pierson et al. 1989), and mean

water temperature is 18.1 C (Ward et al. 2005). The pH of the Cahaba is circumneutral at 7.7,

and alkalinity is high at 84 mg/L as CaCO3 due to the underlying carbonate geology (Ward et al.

2005). Although unregulated hydrologically, the Cahaba is not isolated from the influences of

human development. The shoals receive nutrient rich runoff from the upstream urban center of

Birmingham with NO2-N + NO3-N ranging from 0.25 to 0.30 mg/L, and PO4-P ranging from

<0.003 to 0.025 mg/L (Ward et al. 2005).

Benthic habitat of the shoals is dominated by large patches of exposed bedrock

intermixed with expansive beds of the emergent macrophyte Justicia americana. Bedrock and

Justicia habitats alone occupy ~41% and 26% of shoal surface area, respectively (Chapter 2).

Two other macrophytes are present in lower abundance. The submerged Podostemum

ceratophyllum (~8% cover), while the rare emergent shoals spiderlily, Hymenocallis coronaria, is found in sparse (~1% cover) clumps amongst Justicia patches. The remaining shoal area

consists of boulders (11%), gravel (3%), sand (3%), mud (<1%), woody debris (<1%), and deep

(>1 m) pools (6%) (Chapter 2). Mean base-flow current velocity (m/s) is 0.56 in bedrock, 1.01

in Podostemum, and 0.39 in Justicia habitats. Mean base-flow depth (cm) is 17 in bedrock, 14 in

Podostemum, and 15 in Justicia habitats (Chapter 2, Table 2.1). Justicia has been shown to

significantly modify the benthos by creating large zones of reduced current velocity and

increased sediment deposition (Fritz and Feminella 2003, Fritz et al. 2004). Deposited sediments

are stabilized by perennial root and rhizome structures, creating suitable habitat for Corbicula.

Field sampling

Sampling occurred approximately monthly for 1 y from June 2008-June 2009. A

stratified random design was used to sample Justicia, Podostemum, and bedrock habitats

(Cummins 1962, Wohl et al. 1995). Five replicate Surber samples (0.09 m2; mesh size 250 μm)

were collected of each habitat for each date. Justicia habitats were sampled by placing the

Surber frame over a patch of vegetation, removing all emergent growth, and then extracting the

root/rhizome mass with all associated sediments and invertebrates. Podostemum and bedrock

habitats were sampled by placing the surber frame over the substrate and scraping all benthic

matter into the net with a putty knife and stiff-bristled brush (Grubaugh et al. 1997). Samples

were then elutriated and sorted in the field to remove and replace any federally protected snails;

all other benthic material was fixed in the field with 70% ethanol for transport to the laboratory.

In the laboratory, macrophyte vegetation was separated from the samples and thoroughly

washed to remove and retain any trapped material. Samples were then washed through 1 mm

and 250 µm nested sieves. All animals retained on the 1 mm sieve were sorted at 15x

magnification. Those retained by the 250 µm sieve were subsampled with a sample splitter

(Waters 1969) to a fraction of 1/2, 1/8, or 1/32 the original sample, and sorted under 30x

magnification (Grubaugh et al. 1996).

Production analyses

Corbicula lengths were measured to the nearest 1 mm, except for individuals < 0.75 mm

which were classified as 0.5 mm. Biomass was calculated by converting size-specific (1 mm

size classes) densities (No. individuals/m2) into grams dry mass (g/m2) using the length-mass regression equation from Stites et al. (1995): log(DM) = - 1.69 + 2.45log(L), where DM = dry

mass (mg), and L = shell length (mm).

Annual secondary production was estimated with the size-frequency method for individuals found in the Justicia habitat (Benke 1984). Corbicula individuals were grouped into

2 mm size classes for the initial production calculation. Size-frequency histograms, also using 2

mm size classes, were produced to determine life span, or the cohort production interval (CPI).

It is necessary to adjust the initial size-frequency production calculation with a CPI correction

factor; i.e., production = unadjusted values x 12/CPI (in months) (Benke 1979). The resulting

habitat specific estimate was then further adjusted based on relative proportion of the Justicia habitat (Chapter 2) to estimate Corbicula production per m2 total shoal surface area.

Results

Corbicula was primarily concentrated in the sediments associated with Justicia beds, and

was found in extremely low densities in patches of Podostemum or bedrock. Mean annual

density (No./m2 ± 1 SE) in Justicia was 3,293 ± 1372, compared to Podostemum (80 ± 24), and

bare bedrock (55 ± 32). In Podostemum, 96.5% of individuals were ≤ 3 mm, and the largest

individual collected was 8 mm. In bedrock, 99.6% of individuals were ≤ 2 mm, and the largest

individual collected was 5 mm. Life history, biomass, and production analyses are thus based

only on samples of Justicia habitat. Size-frequency histograms from 5 of the 10 months sampled exhibit a tri-modal distribution of size groups (Fig. 4.1), suggesting 3 simultaneous cohorts and a maximum lifespan (or CPI) of 3 y.

Mean (± 1 SE) monthly density of Corbicula in Justicia beds ranged from 13,638 ±

6,154/m2 in June 2008 to 846 ± 160 in Feb 2009 (Fig. 4.2A). Mean (± 1 SE) monthly biomass

did not track mean density, and ranged from 14.8 ± 3.7 (g/m2) in March 2009 to 3.3 ± 1.3 in

Sept. 2008 (Fig. 4.2B). High density observations of June-September 2008 and were driven

largely by the 0.5 and 1 mm size classes. For these months, the 2 smallest size classes accounted

for 93, 87, and 78% of all collected individuals, respectively. The remaining months exhibited a

more even size class distribution. Biomass fluctuated more erratically, but was driven by

relatively few larger individuals.

Mean annual biomass was 8.4 g DM/m2, and was dominated by individuals ≥13 mm in

length (Table 4.1). Annual production was 14.9 g DM m-2 y-1, based on a 3-y CPI, with an

annual P/B value of 1.8 (Table 4.1). Adjusting the mean annual density, biomass, and annual

production estimates for the proportion of shoal area covered by Justicia population centers

(~26%, Chapter 2) yields values of 857/m2, 2.2 g DM m-2, and 3.9 g DM m-2 y-1 of total shoal surface area, respectively.

25-26 21-22 17-18 13-14 9-10 5-6

Shell Length (mm) (mm) Length Length Shell Shell <3 26 28 8 6 6 22 16 24 28 2 JUN JUL SEP OCT NOV JAN FEB MAR APR JUN 2008 2009

Fig. 4.1. Size-frequency histograms of Corbicula fluminea in Justicia americana beds of Hargrove Shoals, Cahaba River, Bibb Co., AL. Horizontal bars represent relative densities (No./m2) of each size class for each date. Widths of some bars were compressed for clarity, while not altering locations of peaks or valleys. Note 3 peaks in JUN 2008, OCT, NOV, FEB, and JUN 2009 indicating three simultaneous cohorts and a 3-y lifespan.

20000

17500 A

15000 ) 2 12500

10000

7500 Density (No./m 5000

2500

20 18 B 16

) 14 2 12

Biomass (g/m Biomass 6

0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul

2008 2009

Fig. 4.2. Mean (± 1 SE) density (A) and biomass (B) of Corbicula fluminea in Justicia americana beds from 26 June, 2008 through 2 June, 2009.

Table 4.1. Size-frequency table illustrating the estimation of secondary production of Corbicula fluminea in the Justicia americana beds of Hargrove Shoals, Cahaba River, Bibb Co., AL. Annual production was adjusted based on an estimated CPI of 36 mo (Benke 1979).

Shell Length N W B ∆N W W ∆N x 13 (mm) (no./m2) (mg) (g/m2) (no./m2) (mg) (g/m2) (g/m2)

<3 2451 0.08 0.20 2311 0.26 0.60 7.80 3-4 140 0.44 0.06 29 0.88 0.03 0.34 5-6 111 1.33 0.15 15 2.09 0.03 0.41 7-8 96 2.84 0.27 12 3.96 0.05 0.64 9-10 84 5.07 0.42 7 6.58 0.05 0.62 11-12 76 8.10 0.62 -19 10.05 -0.19 -2.49 13-14 96 11.99 1.15 -4 14.41 -0.05 -0.69 15-16 99 16.82 1.67 30 19.74 0.60 7.77 17-18 69 22.65 1.56 17 26.09 0.44 5.73 19-20 52 29.53 1.54 41 33.52 1.39 18.05 21-22 11 37.51 0.40 6 42.07 0.25 3.24 23-24 5 46.64 0.22 3 51.80 0.15 2.01 25-26 2 56.97 0.10 2 56.97 0.10 1.33

B = 8.37 Puncorrected = 44.74 Cohort P/B= 44.74/8.37 = 5.35 Annual P/B= 14.91/8.37 = 1.78 Annual P = 44.74 x 12/36 = 14.91 Negative values were included in summation.

Discussion

This study represents the first production estimate of Corbicula in bedrock shoal habitats,

and illustrates the dramatic influence of shoal macrophytes in modifying otherwise inhospitable

local benthic habitat characteristics for this species. The fact that no individuals >8 mm were

found in Podostemum or bare bedrock habitats indicates that their presence in these habitats is

transient. These clams would almost certainly be washed downstream and into more

depositional habitats such as Justicia or deep pools as they grow larger. Justicia beds, however,

provide both a favorable depositional substrate for Corbicula colonization, and shelter from the high flows that frequent this unregulated river. Its stem and leaf tissues slow local current velocity, which increases deposition of mineral (sand and gravel) and organic particles, while its

perennial root and rhizome tissues trap and stabilize these particles (Fritz and Feminella 2003,

Fritz et al. 2004). The presence of such stable unconsolidated substrate may be the most

important determinant of Corbicula abundance, as was found to be the case by Payne et al.

(1989) for several Mississippi streams

Our estimation of 3 simultaneous cohorts and a 3-y lifespan is consistent with Stites et al.

(1995), though our data were less clear in this regard. Other studies have estimated life spans

ranging from 1 y (Aldridge and McMahon 1978) in Lake Arlington, Texas, to 5 y (Mouthon

2001) in the Saone River, France. Analysis of size-frequency histograms does not clearly reveal

discrete periods of recruitment in this population as large numbers of small individuals were

present almost every month sampled, though much higher densities of juveniles were found in

June and July 2008 (Fig. 4.2A). A more highly replicated sample regime may be required to

provide more clarity of recruitment patterns as the variation between replicates can be large. For

instance, in the detailed life history analyses of Stites et al. (1995), twelve replicate samples were collected from one population center for 2 y to clearly delineate size-frequency distributions.

Extremely high densities of small individuals (≤1 mm) observed in Justicia from June-

September 2008 suggest a pulse of recruitment during summer, but the subsequent disappearance of such high numbers in the following months is evidence of high juvenile mortality or emigration. Mean density (1,000/m2) for most other months was still quite high in this habitat, when compared to previous studies that sampled population center habitats (e.g., Payne et al.

1989, Stites et al. 1995). Maximum densities of similar magnitude to this study have been observed by Gardner et al. (1976) in the Altamaha River, Georgia. In their study documenting the invasion of this river in the early 1970s, they observed maximum summer densities of immature individuals of almost 10,000/m2, but this peak was of short duration and surrounded by monthly densities of ~100/m2. After adjusting the Justicia habitat densities for proportion of shoal area covered (26%), maximum density drops from >13,000/m2 to 3,550.1/m2 of shoal surface area, and for most other months (Oct. 2008-June 2009) densities averaged ~279/m2.

These average values are more in line with most previous channel-scale observations in other rivers (Gardner et al. 1976, Payne et al. 1989, Hornbach 1992, Stites et al. 1995, Sousa et al.

2008).

The range of Corbicula biomass observed in Justicia (~3 to 15 g DM m-2) lies within the lower end of previously reported values for southeastern rivers. Stites et al. (1995) reported a range of 1 to 5 g DM m-2 within the main channel of the Ogeechee River, whereas Sickel (1979) reported 4.4 to 26.2 g DM m-2 in the Altamaha River. Monthly values as high as 425 g AFDM m-2 have been reported for the River Minho estuary in Portugal (Sousa et al. 2008). When habitat-specific values are adjusted for shoal area, monthly biomass estimates range from only

0.86 to 3.85 g DM m-2 of shoal surface, which suggests that this shoal reach as a whole is not an extraordinary center of Corbicula biomass.

Despite the large volume of research on Corbicula, relatively few estimates of secondary

production have been reported for this species. Annual secondary production of Corbicula in

Justicia habitats (~15 g DM m-2 y-1) is comparable to estimates reported by Sickel (1976) in the

Altamaha River (15-16 g DM m-2 y-1), Marsh (1985) in an Arizona drainage canal (26 g DM m-2

y-1), and Stites et al. (1995) in Ogeechee River population centers (10-23 g DM m-2 y-1). These

estimates, however, are far from approaching the upper limit of Corbicula production. Given

optimal conditions, which appears to be the case in the River Minho estuary, this species can

produce as much as ~487 g DM m-2 y-1 (Sousa et al. 2008).

Once again, when Justicia population center estimates are adjusted to the scale of the

shoal reach, production estimates become more modest in magnitude (~4 g DM m-2 y-1). This

value is more consistent with Stites et al. (1995), in their estimate of channel wide production of

the Ogeechee River (1-2 g DM m-2 y-1). The lower estimate of production at the reach-scale emphasizes the fact that Justicia beds represent important islands of productive habitat within a

generally suboptimal bedrock shoal reach.

While it is clear that Justicia is a center of Corbicula production amongst the shoal

habitats within Hargrove Shoals, one can only speculate whether or not Justicia is a uniquely

productive habitat at larger spatial scales within the Cahaba River. Since the presence of well-

oxygenated sediments appears to be the most important condition for Corbicula success

(McMahon 1983, Payne et al. 1989), it would not be surprising to discover similar (or even higher) levels of density, biomass, and production in adjacent deeper pool habitats within the

shoals or in downstream Coastal Plain sediments. Only further, more extensive sampling will answer this question.

Acknowledgements

I would like to thank Mark Dedmon for field assistance, Laura Frost and Michael

Kendrick for lab assistance, and the Cahaba River National Wildlife Refuge for access to

Hargrove Shoals. This work was completed under permit number TE163435-0 from the USFWS granted to Dr. Arthur C. Benke.

Literature Cited

Aldridge, D.W., and R.F. McMahon. 1978. Growth, fecundity and bioenergetics in a natural population of the Asiatic freshwater clam Corbicula manilensis Phillipi from north central Texas. Journal of Molluscan Studies 44:49-70.

Benke, A.C. 1984. Secondary production of aquatic insects. Pages 289-322 In V.H. Resh and D.M. Rosenberg (eds.). The ecology of aquatic insects. Praeger, New York.

Benke, A.C. 1979. A modification of the Hynes method for estimating secondary production with particular significance for multivoltine populations. Limnology and Oceanography 24:168-174.

Benke, A. C. 1993 Baldi Memorial Lecture: Concepts and patterns of invertebrate production in running waters. Verh. Internat. Verein. Limnol. 25:15-38.

Benke, A. C. and A. D. Huryn. 2010. Benthic invertebrate production–facilitating answers to ecological riddles in freshwater ecosystems. Journal of the North American Benthological Society 29:264-285.

Cohen, R.R.H., P.V. Dresler, E.J.P. Phillips, and R.L. Cory. 1984. The effect of the Asiatic clam Corbicula fluminea on phytoplankton of the Potomac River, Maryland. Limnology and Oceanography. 29:170-180.

Counts, C.L. 1986. The zoogeography and history of the invasion of the United States by Corbicula fluminea (Bivalvia: Corbiculidae). American Malacological Bulletin Special Edition No. 2:7-39.

Cummins, K.W. 1962. An evaluation of some techniques for the collection and analysis of benthic samples with special emphasis on lotic waters. The American Midland Naturalist 67:477-504.

Dillon, R.T. 2000. The ecology of freshwater molluscs. Cambridge University Press.

French, J.R.P., and D.W. Schloesser. 1991. Growth and overwinter survival of the Asiatic clam, Corbicula fluminea, in the St. Clair River, Michigan. Hydrobiologia 219:165-170.

Fritz, K.M., and J.W. Feminella. 2003. Substratum stability associated with the riverine macrophyte Justicia americana. Freshwater Biology 48:1630-1639.

Fritz, K.M., M.M. Gangloff, and J.W. Feminella. 2004. Habitat modification by the stream macrophyte Justicia americana and its effects on biota. Oecologia 140:388-397.

Gardner, J.A., W.R. Woodall, A.A. Staats, and J.F. Napoli. 1976. The invasion of the Asiatic clam (Corbicula manilensis Philippi) in the Altamaha River, Georgia. Nautilus 90:117- 125.

Grubaugh, J.W., J.B. Wallace, and E.S. Houston. 1997. Production of benthic macroinvertebrate communities along a southern Appalachian river continuum. Freshwater Biology 37:581-596.

Grubaugh, J.W., J.B. Wallace, and E.S. Houston. 1996. Longitudinal changes of macroinvertebrate communities along an Appalachian stream continuum. Canadian Journal of Fisheries and Aquatic Science 53:896-909.

Hornbach, D.J. 1992. Life history traits of a riverine population of the Asian clam Corbicula fluminea. American Midland Naturalist 127:248-257.

Kraemer, L.R. 1979. Corbicula (Bivalvia: Sphaeriacea) vs. indigenous mussels (Bivalvia: Unionacea) in U.S. rivers: a hard case for interspecific competition? American Zoologist 19:1085-1096.

Lydeard, C., and R.L. Mayden. 1995. A diverse and endangered aquatic ecosystem of the southeast United States. Conservation Biology 9:800-805.

Marsh, P.C. 1985. Secondary production of introduced Asiatic clam Corbicula fluminea in a central Arizona canal. Hydobiologia 24:103-110.

McMahon, R.F. 1983. Ecology of an invasive pest bivalve, Corbicula. Pages 505-561. In W.D. Russell-Hunter (ed.). The Mollusca. Vol. 6. Ecology. Academic Press, New York.

McMahon, R.F., and A.E. Bogan. 2001. Mollusca: Bivalvia. In J.H. Thorp, and A.P. Covich (eds.) Ecology and classification of North American freshwater invertebrates. Academic Press, San Diego, CA.

Mouthon, J. 2001. Life cycle and population dynamics of the Asian clam Corbicula fluminea (Bivalvia: Corbiculidae) in the Rhone River at Creys-Malville (France). Archiv fur Hydrobiologie 151:571-589.

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 43-86 In G.W. Benz and D.E. Collins (eds.) Aquatic fauna in peril: the southeastern perspective. Special Publication 1, Southeast Aquatic Research Institute, Lenz Design & Communications, Decatur, Georgia.

Payne, B.S., A.C. Miller, P.D. Hartfield, and R.F. McMahon. 1989. Variation in size demography of lotic populations of Corbicula fluminea (Muller). Nautilus 103:78-82.

Phelps, H.L. 1994. The Asiatic clam (Corbicula fluminea) invasion and stream-level ecological change in the Potomac River estuary neat Washington, D.C. Estuaries 17:614-621.

Sickel, J.B. 1976. An ecological study of the Asiatic clam, Corbicula manilensis (Phillipi, 1841), in the Altamaha River, Georgia, with emphasis on population dynamics, productivity and control methods. Ph.D. dissertation, Emory University, Atlanta, Georgia.

Sousa, R., A.J.A., Nogueira, M.B. Gaspar, C. Antunes, L. Guilhermino. 2008. Growth and extremely high production of the non-indigenous invasive species Corbicula fluminea (Muller, 1774): possible implications for ecosystem functioning. Estuarine, Coastal, and Shelf Science 80:289-295.

Stites, D.L, A.C. Benke, and D.M. Gillespie. 1995. Population dynamics, growth, and production of the Asiatic clam, Corbicula fluminea, in a blackwater river. Canadian Journal of Fisheries and Aquatic Science 52:425-437.

Strayer, D.L. 1999. Effects of alien species on freshwater mollusks in North America. Journal of the North American Benthological Society 18:74-98.

U.S. Fish and Wildlife Service. 2002. Cahaba River national wildlife refuge. Federal Register Document 02-32210. 67:78238.

Vaughn, C.C., and D.E. Spooner. 2006. Scale-dependent associations between native freshwater mussels and invasive Corbicula. Hydrobiologia 568:331-339.

Waters, T.F. 1969. Subsampler for dividing large samples of invertebrate drift. Limnology and Oceanography 14:813-815

CHAPTER 5

HABITAT-SPECIFIC PRODUCTION OF A FALL LINE RIVER SHOAL MACROINVERTEBRATE ASSEMBLAGE

Abstract

Habitat-specific sampling of benthos allows for a better understanding of the

contributions of various habitats to the overall functional character of a stream reach. The goals of this study were (1) to characterize the distribution and production of the macroinvertebrate assemblages associated with 3 major shoal habitats: bare bedrock, Podostemum ceratophyllum, and Justicia americana; and (2) to estimate the reach-scale production of the macroinvertebrate

assemblage of a free-flowing southeastern river shoal. A total of 90 taxa were identified (most to

genus) in this reach; 49 were found in bedrock, 57 in Podostemum, and 80 in Justicia.

Secondary production was estimated for the 41 most abundant taxa. Total annual production (in

g AFDM m-2 y-1) of all macroinvertebrates was 56.1 in bedrock, 284.4 in Podostemum, and

177.3 in Justicia habitats, with hydropsychid caddisflies and chironomid dipterans being the

most productive groups. Total annual production of the shoals reach was 87.1 g m-2 y-1 of channel surface area, with bedrock contributing 24.3%, Podostemum 22.7%, and Justicia 53.0%

to this total. This study supports the view that Fall Line shoals posses high habitat diversity and

production, and that the more complex habitats (e.g., those with macrophytes) enhance benthic

invertebrate diversity and production. Lastly, the influence of a given habitat depends largely on

its relative abundance, and the Justicia habitat was shown to be especially influential on the diversity and production in this reach.

Introduction

Efforts to understand functional aspects of the ecology of streams and rivers often have focused on production, including the secondary production of the macroinvertebrate assemblage.

Macroinvertebrate assemblages are a conspicuous and diverse component of stream/river communities and estimates of their production (a portion of secondary production) are widespread for small streams (Benke 1993, Benke and Huryn 2010). In larger streams and rivers

(mean annual discharge >10 m3/s), however, production estimates for macroinvertebrate assemblages are relatively scarce (Benke 1993). Large river production studies have occurred in polluted rivers (Berrie 1972, Flossner 1976, 1982), Coastal Plain rivers (Benke et al. 1984), heavily regulated rivers (Voshell 1985, Edwards et al. 1989, Cross et al. 2011), and reaches of rivers with exposed bedrock (Nelson and Scott 1962, Grubaugh and Wallace 1995, Grubaugh et al. 1997).

To date, only 3 previous studies have described production of macroinvertebrate assemblages of expansive bedrock shoals, 2 of which occurred at the same location. First,

Nelson and Scott (1962) performed a production analysis of bedrock outcrop shoals in the

Oconee River, Georgia, and investigated the relationships between organic matter and invertebrate production. Several decades later, Grubaugh and Wallace (1995) revisited this site, repeating the production analyses to evaluate the effects of changing land-use patterns over this time period. They found high levels of production (182 g m-2 y-1) associated with Podostemum covered bedrock, about twice that found in the earlier study. The 3rd study took place along the

Little Tennessee River continuum, which documented extremely high estimates of production

(up to 364 g m-2 y-1 in Podostemum-covered cobble habitats) from two 7th order shoal reaches

(Grubaugh et al. 1997). Interestingly, habitat characteristics of both sites were heavily

influenced by the submergent hydrophyte Podostemum ceratophyllum (Michaux).

Podostemum-covered cobbles or bedrock have been shown to provide important habitat

for macroinvertebrate production (Nelson & Scott 1962, Voshell 1985, Grubaugh and Wallace

1995, Grubaugh et al. 1997), especially filtering-collectors such as hydropsychid caddisflies, and black flies. The dense carpet-like mats of Podostemum, which adhere to bedrock and cobble

substrata, create stable attachment sites for these insects while positioning them within the

current that delivers their sestonic food resources. An experimental study by Hutchens et al.

(2004) demonstrated a strong positive relationship between biomass/surface area of Podostemum and abundance and biomass of associated macroinvertebrates.

Another stream macrophyte sometimes associated with bedrock in streams that has received less attention from benthic invertebrate ecologists is the emergent plant Justicia americana (Vahl). This plant exhibits a more cosmopolitan distribution in both lotic and lentic

systems, and can establish high biomass under certain conditions (e.g., perennially shallow water

depths, slow current velocity) (Chapter 2). Beds of Justicia can significantly alter local physical

conditions of streams they occupy by stabilizing bottom sediments and reducing local current velocity (Fritz et al. 2004). Little is known about the influence of Justicia on the structure and

production of its associated macroinvertebrate assemblage, and how this influences overall

production of the stream reach.

The present study focuses on a heterogeneous Fall Line shoal reach of an unregulated

river in central Alabama. Its channel form is dominated by extensive slabs of exposed bedrock,

including some portions covered by Podostemum ceratophyllum. It also possesses an abundance

of Justicia throughout its channel. This feature sets this shoal apart from those of previous

studies, and this study represents the first examination of this macrophyte’s role as an important

habitat for invertebrate production in a bedrock shoal reach.

The goals of this study were (1) to characterize the distribution and production of the

macroinvertebrate assemblages associated with 3 major shoal habitats: bare bedrock,

Podostemum, and Justicia; and (2) to estimate the reach-scale production of the

macroinvertebrate assemblage of a Cahaba River shoals in central Alabama. It is expected that

habitats within the shoals will support contrasting degrees of taxonomic diversity and

assemblage production, with the more complex habitats (i.e., macrophytes) supporting greater

levels of each. It is predicted that (1) Podostemum will support the highest level of habitat- specific production and be dominated by filter-feeders, (2) Justicia will support the next highest

production, but be dominated by deposition-zone taxa (e.g., shredders, gathering-collectors, and

bivalves), (3) bare bedrock will have the lowest production, and be dominated by grazers feeding

on what is likely to be high benthic primary production, and (4) Justicia will make the highest

reach-level contribution to secondary production, given its expansive distribution.

Materials and Methods

Study site

This study was conducted in a large shoal reach of the 6th order free-flowing Cahaba

River in Bibb County, central Alabama. The study site (Hargrove Shoals, 33 04’ 10” N, 87 04’

43” W) is located within the Cahaba River National Wildlife Refuge established by the US Fish

and Wildlife Service in 2002 (USFWS 2002). Hargrove Shoals exists near the downstream

boundary of the Fall Line transition zone connecting the Valley and Ridge upland and Coastal

Plain lowland physiographic provinces, and is the largest in a series of shallow

limestone/dolomite bedrock shoals that characterize this section of the river. Channel

dimensions of this reach are ~550 m in length, and average 141.5 m wide and 20 cm deep at base

flow (Chapter 2). This shallow depth allows for complete access of the channel by foot during

periods of average to low flow. Mean annual discharge at the study site is 33 m3/s (Pierson et al.

1989), and mean water temperature is 18.1 C (Ward et al. 2005). Streamwater pH is 7.7, and the

underlying carbonate geology yields a high alkalinity of 84 mg/L as CaCO3 (Ward et al. 2005).

Nutrient concentrations are influenced by enriched runoff from the greater Birmingham area

resulting in NO2-N + NO3-N and PO4-P levels that reach 0.30 mg/L and 0.025 mg/L,

respectively (Ward et al. 2005).

The geomorphology of Hargrove Shoals as well as other Cahaba River shoals results in a

high degree of benthic habitat variability. The underlying bedrock provides the stable foundation

for many habitat types (Chapter 2). Exposed bare bedrock encompasses ~41% of available

surface area. Expansive beds of the emergent macrophyte Justicia americana, established atop

and anchored within bedrock fissures, accounts for another 26% of shoal surface area (Chapter

2). Two other macrophytes are also present in lower abundance at this site. The submerged

Podostemum ceratophyllum covers ~8%, and creates carpet-like mats of vegetation that firmly

adhere to bedrock and boulders, whereas the rare emergent shoals spiderlily, Hymenocallis

coronaria, is found in sparse (~1%) clumps of bulbs anchored within bedrock fissures among the

patches of Justicia. The remaining shoal area consists of boulders (11%), gravel (3%), sand

(3%), mud (<1%), woody debris (<1%), and deep (>1 m) pools (6%) (Chapter 2). Knowledge of

relative habitat distributions allow for the weighting of habitat specific production estimates to the reach-scale (Huryn and Wallace 1987, Wohl et al. 1995).

Mean base-flow current velocity (m/s) is 0.56 in bedrock, 1.01 in Podostemum, and 0.39

in Justicia habitats (Chapter 2, Table 2.1). Justicia beds have been shown to create zones of

reduced current velocity and increased sediment deposition (Fritz and Feminella 2003, Fritz et al.

2004). Deposited sediments are stabilized by perennial roots and rhizomes, creating ideal habitat

for deposition-zone adapted invertebrates. Podostemum provides ideal habitat for invertebrate

taxa requiring a stable substratum within zones of higher current velocity (e.g., hydropsychid

caddisflies, Hutchens et al. 2004). This variety of habitats provides the potential for the presence

of a diverse and productive benthic macroinvertebrate assemblage.

Field sampling

A stratified random design was used to quantitatively sample 3 dominant habitats: bare

bedrock, Justicia, and Podostemum on roughly monthly intervals for 1 y (June 2008-June 2009).

Five replicate Surber samples (0.09 m2; mesh size 250 μm) were collected from 5 discrete

patches of each habitat for each date. Justicia habitats were sampled by placing the Surber frame

over a patch of vegetation, removing all emergent growth first, and then extracting the

root/rhizome mass with all associated sediments and invertebrates. Podostemum and bedrock habitats were sampled by placing the Surber frame over the substratum and scraping all benthic matter into the net with a putty knife and stiff-bristled brush (Grubaugh et al. 1997). Samples were then elutriated and sorted in the field to remove, enumerate, record aperture widths, and replace any federally protected snails (e.g., Leptoxis ampla, Lepyrium showalteri, and Lioplax cyclostomaformis). All other benthic material was fixed in the field with 70% ethanol for

transport to the laboratory.

In the laboratory, macrophyte vegetation was separated from the samples and thoroughly

washed to remove and retain any trapped material (detailed methods and analyses found in

Chapter 2). Samples were then washed through 1 mm and 250 µm nested sieves. All animals retained on the 1 mm sieve were sorted at 15x magnification. Those retained by the 250 µm sieve were subsampled with a sample splitter (Waters 1969) to a fraction of 1/2, 1/8, or 1/32 the original sample, and sorted under 30x magnification (Grubaugh et al. 1996). Molluscs were removed, identified and measured from all 5 sample replicates, whereas all other invertebrates were analyzed from only 3 replicates.

Production analyses

Macroinvertebrates were identified to the lowest taxonomic level practicable. Most

insects were identified to genus, with except for the Chironomidae, which were identified to sub-

family or tribe (Merritt and Cummins 1996). Crustacea also were identified to genus, except for

the Ostracoda (Pennak 2001, Thorp and Covich 2010). Mollusca, with the exceptions of

Ancylidae, Planorbidae, Somatogyrus spp., and Campeloma sp. were identified to species

(Tolley-Jordan 2008). Annelids were identified as either oligochaetes or leeches (Hirudinea).

Lengths of all individuals (aperture widths for snails) were measured to the nearest 1 mm, except

for individuals < 0.75 mm, which were classified as 0.5 mm.

Abundance data were then screened to determine which taxa were represented in

numbers sufficient to warrant further analyses. Only those taxa with mean annual densities

>1/m2 in at least 1 habitat were included in biomass and production analyses; the remaining taxa

were considered rare. Size-specific densities (No. ind./m2) of most taxa were converted to

biomass (mg DM/m2, or mg AFDM/m2) with published length-mass regressions (Benke et al.

1999), whereas laboratory-derived equations were used for the pleurocerid gastropods (Chapter

3). For taxa lacking corresponding regression models, equations for the closest possible related

taxon were used. For example, a family-level pleurocerid snail model (Chapter 3) was applied to

Leptoxis ampla, Campeloma sp., and Lioplax cyclostomaformis. Biomass of Somatogyrus spp.,

Lepyrium showalteri, and Ancylidae was estimated with a published equation for Potamopyrgus

antipodarum (Hall et al. 2006). Finally, an equation for Hirudinea was derived from average published a and b values from the length-mass regressions for 3 species (Edwards et al. 2009),

whereas an equation for Oligochaeta was obtained from A.D. Huryn (unpublished data). If

length-mass equations predicted dry mass, these values were converted to AFDM by applying

appropriate % ash correction factors from Benke et al. (1999), Méthot et al. (2012), and Williams

and McMahon (1989).

Annual habitat-specific secondary production was estimated by 4 different methods: the

instantaneous growth, size-frequency, empirical model, and P/B (rough estimate based on mean

B and published P/B values). The choice of method was based on the availability of

independently derived growth models and the quality of average size-frequency distributions.

Instantaneous growth models from Benke and Jacobi (1994) were applied to baetid, caenid,

heptageniid, and tricorythodid mayflies. Chironomid and simuliid production was estimated

using growth models from Benke (1998) and Hauer and Benke (1987), respectively. Growth

models for mayflies, chironomids, and simuliids were developed from a river at a similar

latitude, similar temperature pattern, and similar size (Ogeechee River, Georgia). Previously

derived growth models for the pleurocerid snails were applied to this group (Chapter 3). The

size-frequency method was applied to all other taxa for which sufficient data were present.

Cohort production interval (CPI) correction factors (Benke 1979) were derived from size-

frequency distributions when possible (e.g., Corbicula fluminea, see Chapter 4 for methods) and

from the literature when necessary (e.g., Hydropsyche spp.). Many taxa were sufficiently

abundant for size-frequency production estimation in only 1 or 2 of the sampled habitats. In

such instances, P/B values obtained with the size-frequency method were multiplied by mean annual biomass of a taxon from the habitat(s) for which data were sparse to obtain a rough

estimate of production.

For the Oligochaeta, whose tendency to fragment prohibited accurate estimation of

density, body fragment biomass was estimated from the length of segments with a length-mass

regression (A.D. Huryn, unpublished data) for each sampling date and averaged to calculate mean annual biomass for each habitat. This method likely results in conservative estimates of

biomass. Production was then estimated for each habitat by multiplying mean annual biomass

by an assumed annual P/B value of 10 (Smock et al. 1985). Some taxa were found in insufficient

numbers to apply the size-frequency method in any habitat. When possible in these instances,

empirical models were use to estimate annual P/Bs (e.g., Corydalus sp. and Nyctiophylax sp.)

(Benke 1993). If application of empirical models was impractical, annual P/B was assumed to

be 5 for univoltine taxa (e.g., Lepyrium showalteri) and 10 for bivoltine taxa (Waters 1977).

Production of all taxa in each habitat was summed to estimate the annual assemblage production

at the habitat-scale. Finally, habitat-specific production estimates were adjusted based on the

relative proportions of the 3 sampled habitats (Chapter 2) to estimate reach-scale production.

Results

Taxa richness and distribution

A total of 90 were identified (most to genus) from the sampling of Hargrove Shoals.

Forty-nine taxa were found in bedrock, 57 in Podostemum, and 80 in Justicia (Appendix 5.1).

Among the aquatic insects, 19 families were represented in bedrock, 24 in Podostemum, and 35 in Justicia habitats. Many taxa were found in all habitats sampled, although several showed a

more limited distribution, especially with regard to Justicia. For example, Orconectes spp., the

only crayfish collected was only found in this habitat although in insufficient numbers for

production estimation. Other taxa only found in Justicia include: Tipula spp., ephemerid

mayflies, all odonates, leuctrid and nemourid stoneflies, and the unionid mussel Obliquaria sp.

Habitat-specific production

Secondary production was estimated for the 41 most abundant taxa. Total annual

production (in g AFDM m-2 y-1) of all macroinvertebrates was 56.1 in bedrock, 248.4 in

Podostemum, and 177.3 in Justicia habitats (Table 5.1). Each of the most productive taxonomic

groups will be described according to their distribution among the 3 habitats.

Trichoptera.-- Caddisfly production was high in all habitats, with 17.6 g m-2 y-1 in

bedrock, 111.3 g m-2 y-1 in Podostemum, and 44.6 g m-2 y-1 in Justicia. This production was

primarily from 2 genera within the family Hydropsychidae (Table 5.1, Fig. 5.1). Hydropsyche

spp. accounted for 17.3 g m-2 y-1 in bedrock, 83.2 g m-2 y-1 in Podostemum, and 16.6 g m-2 y-1 in

Justicia. Cheumatopsyche spp. production was similarly high in Podostemum (27.6 g m-2 y-1)

and Justicia (26.2 g m-2 y-1) but very low on bedrock (0.2 g m-2 y-1). Hydroptila spp., Chimarra sp., and Nyctiophylax sp. were very low in all habitats.

Table 5.1. Habitat-specific mean annual biomass (B, mg ash-free dry mass [AFDM]/m2 ± 1 SE), annual production (P, mg AFDM m-2 y-1), and annual P/B (y-1) for the 41 most common benthic invertebrate taxa from the three most abundant habitats of Hargrove Shoals. Habitats include: BR = bedrock, POD = Podostemum ceratophyllum, and JUS = Justicia americana. B and P values are reported per unit habitat surface area. Totals for B and P are reported in g AFDM/m2 and g AFDM m-2 y-1, respectively.

Habitat BR POD JUS Taxon B P P/B B P P/B B P P/B Tricladida Dugesiidae 0.0 0.1 5.6 35.1 208.1 7.6 19.5 57.5 3.5 Oligochaeta 1.2 11.5 10.0 7.3 72.8 10.0 438.4 4384.2 10.0 Hirudinea 19.6 75.9 3.9 1.5 5.9 3.9 2283.5 8765.4 3.8 Isopoda Lirceus spp. 0.7 6.5 8.9 11.8 105.8 8.9 1133.4 10143.2 8.9 Amphipoda Hyalella spp. 0.3 3.5 11.8 0.6 6.8 11.8 55.0 648.3 11.8 Ephemeroptera Baetis spp. 9.5 780.6 81.9 182.1 14050.1 77.1 32.9 2628.5 79.8 Heterocloeon spp. 11.0 1012.7 92.3 73.2 7008.5 95.8 7.2 650.1 90.3 Caenis sp. 0.3 27.4 87.5 6.0 315.0 52.6 0.1 8.1 58.3 Serratella sp. 0.1 1.2 9.5 34.6 328.8 9.5 39.7 377.0 9.5 Stenonema spp. 3.7 288.7 78.4 25.7 1852.3 72.0 88.2 7108.3 80.6 Isonychia sp. 0.0 0.1 9.2 6.0 55.7 9.2 46.0 425.0 9.2 Tricorythodes spp. 2.4 136.3 56.1 23.3 1369.5 58.7 25.7 1601.1 62.3

Table 5.1. Continued. Habitat BR POD JUS Taxon B P P/B B P P/B B P P/B Odonata Haeterina sp. 0.0 0.0 -- 0.8 3.4 4.4 57.1 255.0 4.4 Argia spp. 0.0 0.1 4.8 1.6 7.9 4.8 21.9 71.6 4.8 Megaloptera Corydalus sp. 0.0 0.0 -- 0.3 2.4 7.5 5.2 39.2 7.5 Lepidoptera Petrophila sp. 6.4 30.9 4.8 13.9 82.0 5.9 4.4 16.2 3.7 Trichoptera Cheumatopsyche spp. 11.1 182.8 16.5 2119.4 27627.8 13.0 1838.0 26189.3 14.3 Hydropsyche spp. 845.4 17259.5 20.2 6483.2 83213.9 11.8 1250.6 16622.8 12.5 Hydroptila spp. 11.4 156.7 13.7 19.3 250.0 13.0 7.3 80.5 11.1 Chimarra sp. 0.1 1.4 9.8 16.6 163.1 9.8 162.6 1599.9 9.8 Nyctiophylax sp. 0.0 0.0 -- 5.7 48.5 8.5 7.0 59.0 8.5 Coleoptera Microcylloepus spp. 9.2 31.9 3.5 183.7 649.8 3.5 347.6 1230.5 3.5 Stenelmis spp. 6.4 39.6 6.2 724.8 3824.3 5.3 380.3 1653.6 4.4 Diptera Chironomini 33.4 7783.4 233.4 118.0 30673.3 260.0 42.6 11104.5 260.9 Orthocladiinae 117.2 18253.1 155.8 271.0 40450.4 149.3 149.7 23094.5 154.2 Tanytarsini 2.1 440.7 211.2 35.5 6534.4 183.8 15.4 2945.1 191.2 Tanypodinae 0.4 93.3 257.7 33.0 3571.4 108.2 1.6 394.7 251.1 Simulium spp. 11.53 1011.26 87.71 215.7 15716.3 72.86 129.2 9497.2 73.49 Tipula spp. 0.0 0.0 -- 0.0 0.0 -- 1010.4 13054.6 13.1

Table 5.1. Continued. Habitat BR POD JUS Taxon B P P/B B P P/B B P P/B Gastropoda Ancylidae 0.1 0.4 5.3 0.2 1.0 5.3 3.0 15.7 5.3 Lepyrium showalteri 2.5 12.7 5.0 2.2 10.8 5.0 9.6 47.8 5.0 Somatogyrus spp. 6.4 30.1 4.7 215.8 1098.1 4.7 72.1 277.5 4.1 Elimia ampla 707.1 594.9 0.8 1028.6 876.6 0.9 860.6 624.7 0.7 Elimia cahawbensis 86.0 98.1 1.1 1087.8 1129.1 1.0 1484.6 1495.2 1.0 Elimia clara 625.7 666.6 1.1 1365.1 1453.2 1.1 2278.3 2235.2 1.0 Elimia showalteri 1335.1 953.3 0.7 3488.6 2592.4 0.7 4361.5 3042.1 0.7 Pleurocera vestita 79.3 64.8 0.8 350.5 211.5 0.6 2149.5 1519.7 0.7 Leptoxis ampla 6923.2 6082.3 0.9 2892.0 2650.5 0.9 1463.6 1191.3 0.8 Campeloma sp. 0.0 0.0 -- 0.0 0.0 -- 834.7 1391.2 1.7 Lioplax cyclostomaformis 0.0 0.0 -- 90.3 150.5 1.7 3945.3 6575.5 1.7 Bivalvia Corbicula fluminea 0.8 1.5 1.8 6.1 10.9 1.8 7949.7 14164.5 1.8

Total 10.87 56.13 5.4 21.18 248.38 11.9 35.01 177.29 5.2

3030 BedrockBedrock 2525

2020 1515

1010

5 5

0 0

8595 PodostemumPodostemum 8090 ) ) -1 -1 3030 y y -2 -2 2525 2020 1515 1010 Production m (g AFDM Production (g AFDM m 5 5 0 0

3030 JusticiaJusticia

2525

2020

1515

1010 5 5

0 0 . . . . p...... pp. p pp p p sp.p spp s sppp s sppp spp. s xs x sp p s . las . rra sp.rra sp.. a a . hes he . he he p ti p aa a ylx yl e p e p la tila p rr sp a sp h sp ch s ti s a yl c y e p a ra h ax y he s h ro til im r p l ps c p yc d p h a io hy o y to s y ro C im ct p dr ps a p H d h y io y o m to y C N ct H dr u a H y y he m N H C u he C

Fig. 5.1. Annual production of Trichoptera taxa in three main habitats (bedrock, Podostemum ceratophyllum, and Justicia americana) of Hargrove Shoals. Production is reported per unit habitat surface area. AFDM = ash free dry mass.

Diptera.-- Dipteran production also was high in all habitats, with 27.6 g m-2 y-1 in

bedrock, 97.0 g m-2 y-1 in Podostemum, and 60.1 g m-2 y-1 in Justicia. High production was

primarily due to Chironomidae with their extremely high annual P/B values (Table 5.1, Fig. 5.2).

Within the Chironomidae, the Orthocladiinae had an annual production of 18.3 g m-2 y-1 in

bedrock, 40.5 g m-2 y-1 in Podostemum, and 23.1 g m-2 y-1in Justicia. Chironomini accounted for

7.8 g m-2 y-1 in bedrock, 30.7 g m-2 y-1 in Podostemum, and 11.1 g m-2 y-1in Justicia. Tanytarsini

produced 0.4 g m-2 y-1 in bedrock, 6.5 g m-2 y-1 in Podostemum, and 2.9 g m-2 y-1 in Justicia.

Tanypodinae production was relatively low in all habitats. Simuliidae (i.e., Simulium spp.) and

Tipulidae (i.e., Tipula spp.) accounted for considerably less production than chironomids,

although each had higher production than many other insect taxa. Simulium production was 1.0

g m-2 y-1 in bedrock, 15.7 g m-2 y-1 in Podostemum, and 9.5 g m-2 y-1 in Justicia. Tipula spp. was

only present in Justicia, where its production was 13.1 g m-2 y-1.

Ephemeroptera.-- Mayfly production was not nearly as high as caddisflies and dipterans

in general, but was the third highest among the insects, and was concentrated in Podostemum

(25.0 g m-2 y-1) and Justicia (13.0 g m-2 y-1) (Table 5.1). Like the chironomids and simuliids, the

relatively high production of some mayflies was due to high P/B estimates. The Baetidae

composed most of this production, with Baetis spp. reaching 14.1 g m-2 y-1 in Podostemum, and

2.6 g m-2 y-1 in Justicia. Heterocloeon spp. was 7.0 g m-2 y-1 in Podostemum, and 0.7 g m-2 y-1 in

Justicia. Stenonema spp. (Heptageniidae) production was 1.9 g m-2 y-1 in Podostemum, and 7.1 g

m-2 y-1 in Justicia, whereas Tricorythodes spp. (Tricorythidae) was 1.4 g m-2 y-1 in Podostemum, and 1.6 g m-2 y-1 in Justicia (Table 5.1). The remaining mayflies all were <1.0 g m-2 y-1 in all habitats.

45 Bedrock 40 35 30 25 20 20 15

45 Podostemum 40

) ) 35 -1 -1 y y 30 -2 -2 25

Production (g AFDM m Production (g AFDM m 10 5 0

45 Justicia 40 40 35

5 0 e i . e .. eae i i pp.p. . ii aaee p aiinaeiina minimininin ssppp inn a spp.p indn o spp arsinis i idindinaenin s ilaii nnomm s ytrrs dd la ad ladd noo mumm tata oopo llaulaau l la oon iuu yy ypp uu oc c irir l li nn nny iipp h ho h h uu aa aa TT rt rt CC imim TT TT OO SS

Fig. 5.2. Annual production of Diptera taxa in three main habitats (bedrock, Podostemum ceratophyllum, and Justicia americana) of Hargrove Shoals. Production is reported per unit habitat surface area. AFDM = ash free dry mass.

Coleoptera.-- Beetle larva production was restricted to 2 genera within the Elmidae with highest production found in Podostemum (4.5 g m-2 y-1) and Justicia (2.9 g m-2 y-1). Stenelmis spp. reached 3.8 g m-2 y-1 in Podostemum and 1.7 g m-2 y-1 in Justicia (Table 5.1). The less

abundant Microcylloepus spp. only exceeded 1.0 (1.2 g m-2 y-1) in Justicia.

Other Insects.-- Production of other insect taxa was quite low, with only 2 genera of

Odonata (Haeterina sp., and Argia spp.) and 1 genus of Megaloptera (Corydalus sp.). Both

odonate taxa and Corydalus sp. were concentrated in Justicia beds, but total production of all 3

taxa was <<1.0 g m-2 y-1 (Table 5.1). Petrophila sp. (Lepidoptera) was most productive in

Podostemum at 0.08 g m-2 y-1 (Table 5.1).

Gastropoda.-- Snail production was relatively high in all habitats (but not as high as

many insects), primarily because of their high biomass (Table 5.1, Fig. 5.3). Production was 8.5

g m-2 y-1 in bedrock, 10.2 g m-2 y-1 in Podostemum, and 18.4 g m-2 y-1 in Justicia. Snail

production was concentrated primarily with the Pleuroceridae, and somewhat less so with the

Viviparidae and Hydrobiidae. The federally endangered viviparid Lioplax cyclostomaformis had

relatively high annual production (6.6 g m-2 y-1) compared with most other snails in any given

habitat, though it was almost entirely restricted to the sediments found within Justicia beds. The

federally threatened (but locally abundant) pleurocerid Leptoxis ampla possessed similarly high

production (6.1 g m-2 y-1) in bedrock, while being somewhat lower in Podostemum (2.7 g/m2) and Justicia (1.2 g m-2 y-1). Production of the common pleurocerids was more evenly distributed

among the 3 habitats. However, Pleurocera vestita exhibited somewhat higher production in

Justicia (1.5 g m-2 y-1), compared with bedrock and Podostemum, whereas Elimia showalteri, E.

clara, and E. cahawbensis occurred mostly in either vegetated habitat vs. bare bedrock. Annual

production of Elimia showalteri was 1.0 g m-2 y-1 in bedrock, 2.6 g m-2 y-1 in Podostemum, and

3.0 g m-2 y-1 in Justicia; E. clara was 0.7 g m-2 y-1 in bedrock, 1.5 g m-2 y-1 in Podostemum, and

2.2 g m-2 y-1 in Justicia; E. cahawbensis was 0.1 g m-2 y-1 in bedrock, 1.1 g m-2 y-1 in

Podostemum, and 1.5 g m-2 y-1 in Justicia. Somatogyrus spp. seemed to prefer Podostemum (1.1 g m-2 y-1) over bedrock or Justicia (Table 5.1).

Bivalvia.-- Estimates of bivalve production were limited to the non-native Corbicula

fluminea, which has been previously discussed (Chapter 4). The distribution of this species was

almost exclusively restricted to the depositional Justicia beds, with annual production reaching

14.2 g m-2 y-1 in this habitat (Fig. 5.3, Table 5.1). Production of Corbicula in bedrock and

Podostemum was negligible.

Other invertebrates.-- Crustacean annual production was dominated by the isopod

Lirceus spp., which was most abundant within Justicia beds. This genus produced 10.1 g m-2 y-1

in this habitat, while only producing 0.01 g m-2 y-1 in bedrock and 0.1 g m-2 y-1 in Podostemum

(Table 5.1). The amphipod Hyallela spp. was also concentrated within Justicia beds (0.6 g m-2 y-

1), having only 3.5 mg m-2 y-1 and 6.8 mg m-2 y-1 in bedrock and Podostemum habitats,

respectively (Table 5.1). Annelid annual production was also concentrated in Justicia,

amounting to a combined 13.1 g m-2 y-1. Oligochaetes produced 4.4 g m-2 y-1, while the leeches

(Hirudinea) produced 8.8 g m-2 y-1 in this habitat.

7 Bedrock 6 5 4

2 1 0

7 Podostemum 6

) -1 5 y -2 4

2 Production m (g AFDM 1 0

14 Justicia 7 6 5 4 3 2 1 0 . . la ri ra a s p ri e p s a p te a pl i ita p e a s i e l l ns t s lt id a m in m a m e es s a yl or a w iac b u w c om f um is o iaa w yr o n l a l x sh lim a rav g h A e m to E lim h e o s p to laf p ia E a c at m m s u e m c ro u a lo ic L li ia u om ri C c rb E m le S py cy o li P e x C E L la op Li

Fig. 5.3. Annual production of gastropod taxa and Corbicula fluminea in three main habitats (bedrock, Podostemum ceratophyllum, and Justicia americana) of Hargrove Shoals. Production is reported per unit habitat surface area. AFDM = ash free dry mass.

Habitat comparison among major groups and Habitat-weighted production

An overview of production for the 3 major habitats revealed certain similarities and some striking differences. Production of all habitats was dominated by caddisflies and dipterans (Fig.

5.4, upper 3 graphs). Snails were the third most productive group in all habitats except

Podostemum where mayflies came in third. Justicia clearly had the broadest diversity of production, with significant portions from crustaceans, annelids, and bivalves, although the latter was a single nonnative species.

Total annual production of all major macroinvertebrates after habitat-weighting of

Hargrove Shoals was 87.1 g m-2 y-1 of channel surface area, with bedrock contributing 24.3%,

Podostemum 22.7%, and Justicia 53.0% to this total (Table 5.2). Main taxonomic contributors

to this total remained the Diptera and Trichoptera, which accounted for 33.8 and 27.1 g m-2 y-1, respectively (Table 5.2, Fig. 5.4, bottom graph). The Mollusca also contributed substantially to reach production (12.5 g m-2 y-1) (Fig. 5.4). Snails accounted for 8.8 g m-2 y-1, and Corbicula

fluminea, for 3.7 g m-2 y-1. Other invertebrates such as the mayflies, annelids (Oligochaeta and

Hirudinea) and crustaceans accounted for another 6.2, 3.5, and 2.8 g m-2 y-1, respectively (Table

5.2, Fig. 5.4).

70 Bedrock 60 50 40 30 20 10 0 120 Podostemum 100 80 60 )

-1 40 y -2 20 0 70 Justicia

Production (g AFDM m 60 50 40 30 20 10 0 70 Hargrove Shoals 60 50 40 30 20 10 0 ra ra ra ea da ia da te te te c li lv o ip p p ta e va p D o ro s nn i ro ch e ru A B st ri m C a T e G ph E Fig. 5.4. Annual production of major taxa per square meter of bedrock, Podostemum, and Justicia habitats, respectively, as well as habitat-weighted production per square meter of Hargrove Shoals. AFDM = ash free dry mass. Note different scale for Podostemum.

Table 5.2. Habitat-weighted mean annual density (N, no./m2), biomass (B, mg ash-free dry mass [AFDM]/m2 ± 1 SE), annual production (P, mg AFDM m-2 y-1), and annual P/B (y-1) for the 41 most common benthic invertebrate taxa of Hargrove Shoals. Values of N, B, and P are reported per unit shoal surface area. Totals for B and P are reported in g AFDM/m2 and g AFDM m-2 y-1, respectively.

Taxon N B P P/B Tricladida Dugesiidae 96.0 ± 56.7 7.9 ± 4.7 31.6 4.0 Oligochaeta N/A 150.7 ± 115.1 1151.3 84.7 Hirudinea 31.8 ± 8.9 601.9 ± 243.4 2310.7 3.8 Isopoda Lirceus spp. 1920.0 ± 1770.8 296.3 ± 258.8 2651.2 8.9 Amphipoda Hyalella spp. 53.3 ± 33.7 14.5 ± 7.6 170.6 11.8 Ephemeroptera Baetis spp. 276.9 ± 139.3 26.7 ± 11.9 2098.2 78.7 Heterocloeon spp. 126.7 ± 82.4 11.8 ± 8.6 1109.5 93.7 Caenis sp. 2.4 ± 2.4 0.6 ± 0.6 37.5 59.4 Serratella sp. 21.0 ± 14.7 13.1 ± 9.4 124.8 9.5 Stenonema spp. 75.4 ± 29.9 26.4 ± 9.8 2106.7 79.8 Isonychia sp. 11.1 ± 5.0 12.5 ± 5.9 115.1 9.2 Tricorythodes spp. 84.2 ± 39.6 9.5 ± 4.0 577.3 61.0 Odonata Haeterina sp. 9.6 ± 5.4 14.9 ± 6.5 66.6 4.5 Argia spp. 6.3 ± 3.5 5.8 ± 2.6 19.3 3.3 Megaloptera Corydalus sp. 3.9 ± 3.9 1.4 ± 1.4 10.4 7.5 Lepidoptera Petrophila sp. 14.7 ± 10.5 4.7 ± 3.1 22.4 4.8 Trichoptera Cheumatopsyche spp. 1125.4 ± 394.7 651.5 ± 186.1 9087.9 13.9 Hydropsyche spp. 2362.6 ± 1230.9 1160.9 ± 330.7 17464.2 15.0 Hydroptila spp. 160.5 ± 85.0 7.7 ± 3.6 99.9 12.9 Chimarra sp. 142.0 ± 79.7 43.7 ± 13.2 430.0 9.8 Nyctiophylax sp. 2.5 ± 1.9 2.3 ± 1.9 19.2 8.5 Coleoptera Microcylloepus spp. 371.2 ± 143.6 108.6 ± 44.4 384.1 3.5 Stenelmis spp. 581.7 ± 189.7 159.2 ± 49.4 750.1 4.7

Table 5.2. continued

Taxon N B P P/B Diptera Chironomini 1172.9 ± 322.3 33.1 ± 7.4 8268.7 250.2 Orthocladiinae 3264.6 ± 1379.2 104.7 ± 51.3 16115.0 153.9 Tanytarsini 672.8 ± 214.4 7.6 ± 3.1 1453.5 190.5 Tanypodinae 11.1 ± 7.3 3.2 ± 2.9 422.5 133.1 Simulium spp. 899.9 ± 455.7 46.6 ± 23.6 4105.9 88.0 Tipula spp. 18.8 ± 11.7 263.0 ± 137.8 3398.1 12.9 Gastropoda Ancylidae 23.8 ± 15.6 0.8 ± 0.6 4.3 5.2 Lepyrium showalteri 7.0 ± 4.2 3.6 ± 2.5 18.1 5.0 Somatogyrus spp. 487.7 ± 241.2 38.4 ± 16.4 171.1 4.5 Elimia ampla 45.8 ± 28.2 572.5 ± 283.1 456.7 0.8 Elimia cahawbensis 59.9 ± 15.8 505.6 ± 160.5 516.2 1.0 Elimia clara 131.9 ± 66.8 937.7 ± 298.1 948.9 1.0 Elimia showalteri 154.6 ± 61.9 1916.6 ± 734.6 1357.8 0.7 Pleurocera vestita 27.1 ± 10.8 617.3 ± 262.1 436.8 0.7 Leptoxis ampla 143.7 ± 39.9 3220.8 ± 984.1 2813.8 0.9 Campeloma sp. 1.1 ± 0.6 217.3 ± 137.5 362.1 1.7 Lioplax cyclostomaformis 14.5 ± 4.1 1034.2 ± 358.9 1723.6 1.7 Bivalvia Corbicula fluminea 884.2 ± 371.2 2070.1 ± 276.4 3688.4 1.8

Total 15,500.6 ± 7583.2 14.9 ± 5.0 87.1 6.0

Discussion

Habitat heterogeneity is known to enhance species diversity (Downes et al. 1998, Brown

2003), temporal stability (Lancaster 2000, Brown 2003), abundance and biomass (Hutchens et al.

2004). While this study was not explicitly designed to test any specific mechanism regarding the

influence of habitat heterogeneity, the three habitats sampled in this reach represent, to some

degree, a gradient of habitat heterogeneity. Bedrock is the least physically complex of the

habitats, whereas the differences in structural complexity between Podostemum and Justicia are somewhat less apparent. The differences between these two vegetated habitats may not simply lie in the possible differences in surface area per unit stream area, but may be more related to the variety and range of different microhabitats created by these plants throughout their reach-scale distributions. Nevertheless, the results presented here support the general notion that complex macrohabitats, in this case a shoal reach, have the capacity to support high levels of taxonomic diversity and production due, at least in part, to the variable contributions of a variety of differing habitats (e.g., bedrock, Podostemum, and Justicia).

Taxa richness and distribution

Estimates of taxa richness (90 taxa) in Hargrove Shoals are undoubtedly conservative in

terms of actual species diversity given the varied and often coarse degree of taxonomic resolution applied in this study. Certainly, limiting the resolution of the Chironomidae alone to sub-family or tribe causes a potentially dramatic underestimate in richness. In their study of the

Little Tennessee River continuum, Grubaugh et al. (1997) collected a total of 86 taxa in one 7th order reach, and 63 taxa in a reach nearly 30 km downstream. These values include chironomids identified to the genus level, which alone accounted for 20 taxa in each site. Therefore,

assuming a similar degree of chironomid diversity for Hargrove Shoals one may expect to find at

least 110 macroinvertebrate genera in this reach.

The taxa richness estimated for Hargrove Shoals surpasses that of Grubaugh and Wallace

(1995), who found 42 taxa in an Oconee River bedrock outcrop shoal reach, though direct

comparison is somewhat difficult due to minor differences in taxonomic resolution.

Nevertheless, their richness values are comparable to those found in Podostemum patches of

Hargrove Shoals (57 taxa, Appendix 5.1), which is not surprising given that the Oconee shoals are dominated by Podostemum covered habitats.

The most obvious differences in habitat structure between the above studies and

Hargrove Shoals are the presence of expansive Justicia beds in the latter, combined with the

comparatively low abundance of Podostemum habitats. The presence of Justicia across the reach creates a wide variety of habitat conditions not found in shoals lacking this plant, which may

help explain the greater diversity observed here. Justicia probably provides a greater level of

microhabitat variety when compared to Podostemum, which tends to be more abundant in zones

of higher water velocity (Chapter 2, Table 2.1). The expansive distribution of Justicia across the

entire river channel spans the spectrum of available physical conditions, from nearly lentic eddy

zones to patches directly adjacent to the swift channel thalweg. Some Justicia patches collect substantial amounts of sediment and benthic organic matter (e.g., >10 cm deep), while other patches create only a thin layer of rhizomes clinging to the surface of otherwise bare bedrock

(personal observation). This variety of microhabitat conditions fulfills the requirements of a number of wetland-plant/deposition-zone taxa (e.g., Ephydridae, Ephemeridae, Tipula spp.,

Corbicula fluminea), while still providing suitable conditions for several productive rheophilic

taxa (e.g., net-spinning caddisflies, simuliids), and supporting a more diverse invertebrate

assemblage overall.

Habitat-specific production

Annual production of bare bedrock (56.1 g m-2 y-1) was considerably lower than either

Podostemum (248.4 g m-2 y-1) or Justicia (177.3 g m-2 y-1). Even so, the level of production

observed in bedrock would be considered relatively high when compared to most other habitat-

specific or reach-level estimates for macroinvertebrate communities (Benke 1993). Several

factors likely contribute to the high production of Hargrove Shoals habitats in general. Shoals are complex in terms of overall reach geomorphology, and variety of stable meso-scale habitat structure. The extremely wide and shallow channel allows for full sun exposure and presumably high autochthonous primary production, which provides an abundance of food resources for primary consumers. Also, the carbonate geology of this basin yields a highly alkaline water chemistry that has been associated with higher primary and secondary production (Krueger and

Waters 1983). Finally, anthropogenic nutrient inputs from the city of Birmingham may enhance algal, microbial, and higher plant production (Ward et al. 2005).

Production in each habitat was dominated by two insect families: the hydropsychid caddisflies, and the chironomid dipterans, whose combined contributions accounted for 78% of production in bedrock, 77% in Podostemum, and 45% in Justicia. Podostemum has once again

proven to be an important site of insect production, with estimates of 3 major orders (i.e.,

Trichoptera, Diptera, and Ephemeroptera) exceeding that found in either Justicia or especially bedrock (Fig. 5.4). The bulk of production in Podostemum, however, came from a more limited

group of taxa than found in Justicia. While nearly half of Justicia production was associated

with hydropsychids and chironomids, a greater diversity of taxa contributed substantially to the

overall total for the habitat (Fig. 5.4). For instance, deposition-zone taxa such as tipulid dipterans (Tipula spp.), annelids (oligochaetes and leeches), crustaceans (especially Lirceus spp.), bivalves (Corbicula fluminea), and some gastropods (e.g., Lioplax cyclostomaformis) account for the bulk of remaining production in Justicia, and were rarely (if ever) found in the

other habitats (Fig 5.4).

Differences of production between habitats are most likely related to varying degrees of

structural complexity at both the microhabitat and mesohabitat scales. Low structural complexity of bedrock, relative to vegetated habitats, provides less surface area for invertebrate

colonization and is likely to provide little refuge from hydraulic stress (Lancaster and Belyea

1997, Taniguchi and Tokeshi 2004). Both vegetated habitats, however, offer complex 3-

dimensional environments that greatly multiply the space available for colonization of

periphyton and invertebrates, and likely create refuge from periodic flow disturbance (Gregg and

Rose 1982, Fritz et al. 2004, Hutchens et al. 2004). Hutchens et al. (2004) developed a

regression relating Podostemum biomass to surface area and found a 3 to 4 fold increase over

that of bare bedrock. This surface area was then found to be positively correlated to abundance

and biomass of the invertebrate assemblage. Unfortunately, no such relationship has been

developed for Justicia, but there is obviously some degree of increase in available surface area

created by this plant. Considering that mean monthly submergent biomass (i.e., roots, rhizomes,

and submerged stems) ranged from 444 to 906 g AFDM m-2 (Chapter 2), it is clear that the

surface area provided by this plant for invertebrate colonization is substantial. For comparison,

mean monthly biomass of Podostemum throughout the sampling period ranged from 42 to 198 g

AFDM m-2 (Chapter 2).

Habitat-weighted production

Habitat-weighted annual production estimates adjust habitat-specific values (per m2 of a given habitat SA) to the scale of the average m2 surface area of the entire shoal. These adjusted

values allow for a better understanding of how habitats and taxa contribute to invertebrate

production at the reach-scale. This is important because some habitats can be quite productive,

yet relatively rare. For example, though Podostemum has proven to be an extremely productive

habitat, its relatively sparse distribution (8%) in this shoal dampens its influence on overall reach production. In contrast, the more modest production of the bedrock habitat is amplified by its abundant distribution (38%) to such an extent that the two habitats have similar degrees of influence on reach-scale production (BR = 24%, POD = 23%). Finally, the moderately high production within Justicia beds combined with its expansive distribution (26%) leads to its great

influence on reach-scale production (JUS = 53%).

Comparisons of total assemblage production with other studies

All 3 habitats examined in this study proved to be quite productive when compared to

habitats in many other rivers. High levels of habitat-specific production, especially for habitats associated with Podostemum, have been documented for other bedrock-dominated shoal reaches

in the southeastern United States, and are among the highest ever estimated (Benke 1993, Benke

and Huryn 2010). Grubaugh and Wallace (1995) estimated production of a Podostemum covered bedrock outcrop of the Middle Oconee River to be 182 g m-2 y-1. In addition, they revised the

historical estimate of Nelson and Scott (1962) for the same reach to be 90 g m-2 y-1. In a 7th order section of the Little Tennessee River, Grubaugh et al. (1997) found extremely high production in both Podostemum-covered cobble (364 g m-2 y-1), and bedrock outcrop (143 to 168

g m-2 y-1). They suggested that Podostemum-covered cobble was a more productive habitat

compared to bedrock due to its substratum complexity as well as its connection to the hyporheic

zone.

One noteworthy difference distinguishes the bedrock production estimates reported here from those of both Grubaugh and Wallace (1995) and Grubaugh et al. (1997). Neither of the previously published studies discriminated between what has been termed here as ‘bare bedrock’

(i.e., lacking all vegetation) and Podostemum-covered bedrock which are treated as separate

habitats in this study. In fact, both previous studies demonstrated that the vast majority of

available bedrock outcrop in both the Oconee and Little Tennessee Rivers was actually covered

in this macrophyte, which helps explain their comparatively high bedrock production estimates.

Podostemum is much more limited in its distribution throughout Hargrove Shoals, and only covers 8% of the channel area (Chapter 2). Therefore, estimates of ‘bedrock outcrop’ production from these systems are more appropriately comparable to the Podostemum habitat in Hargrove

Shoals.

Reach-scale production of Hargrove Shoals (87.1 g m-2 y-1) was also at the high end of

reported estimates for stream reaches (Benke 1993), while being similar to previous estimates of

mid-order shoals of this region. Estimates from the Oconee shoals were no different from

habitat-specific estimates due to the dominance of Podostemum-covered bedrock habitat in this

reach, and ranged from 90 g m-2 y-1 (Grubaugh and Wallace 1995, from the data of Nelson and

Scott 1962) to 182 g m-2 y-1 (Grubaugh and Wallace 1995). Reach-scale production of the Little

Tennessee shoals ranged from 121 to 154 g m-2 y-1 at two 7th order sites (Grubaugh et al. 1997).

Comparisons of snail production with other studies

Another aspect of Hargrove Shoals that sets it apart from other rivers is the presence of

such a diverse (6 families and at least 13 species, Appendix 5.1) snail assemblage. No previous

study has provided such detail regarding the production of such a diverse snail assemblage.

While only a minor contributor to habitat (4 to 15%) or reach (10%, Fig. 5.5) production, the gastropods of Hargrove Shoals are noteworthy in terms of diversity and reach biomass (9.1 g m-2 or 61%, Fig. 5.5). Few (if any) streams or rivers compare with the Cahaba in this regard, though several may have prior to human alteration (e.g., Coosa, Tallapoosa, Tennessee).

Few estimates of lotic snail assemblage production exist in the literature, and even fewer studies report estimates comparable to those found in Hargrove Shoals (bedrock = 8.5 g m-2 y-1,

Podostemum = 10.2 g m-2 y-1, Justicia = 18.4 g m-2 y-1, shoal reach = 8.8 g m-2 y-1) (Fig. 5.4).

For instance, snails (e.g., Elimia spp.) only accounted for <1% (i.e., <1.5 g m-2 y-1) of habitat

weighted production in 6th and 7th order shoal reaches of the Little Tennessee (Grubaugh et al.

1997). In the Oconee shoals, total snail production was 28 g m-2 y-1, while biomass was a

considerably lower 7.3 g m-2 (Grubaugh and Wallace 1995). The majority (87%) of this high

production, however, was accounted for by the single hydrobiid genus Somatogyrus, which thrives in the Podostemum habitat abundant in this reach. Elimia spp. accounted for the

remainder (3.6 g m-2 y-1). Neither study attempted to identify the snail assemblage to species as

was done here (for most taxa) in the Cahaba.

One study, which took place in the Mobile River Basin, estimated production for the

Elimia spp. assemblages of six smaller streams from two distinct geologies (Huryn et al. 1995).

The three Piedmont streams supported a mean Elimia spp. production of 1.6 g m-2 y-1, while the

three Valley and Ridge streams (two of which drain into the Cahaba River) supported a

production of 2.5 g m-2 y-1 (Huryn et al. 1995). For comparison, reach-scale production of

Elimia spp. in Hargrove Shoals was 3.3 g m-2 y-1. This relatively small increase in production

from the small streams to Cahaba River may be somewhat surprising except for the likelihood

that Elimia spp. represents the near totality of snail production in the smaller streams, whereas

Elimia spp. only represents 37% of snail production in the Hargrove Shoals reach.

While the Hargrove Shoals snail assemblage appears to be relatively productive for this

group, one study of the macroinvertebrate assemblage of Polecat Creek, Wyoming, reported the

production of the nonnative New Zealand mudsnail, Potamopyrgus antipodarum, of 194 g m-2 y-

1 (Hall et al. 2006). This estimate is among highest values ever recorded for any invertebrate taxon, and higher than most assemblage level production estimates. P. antipodarum dominated

the invertebrate assemblage of this stream, constituting 92% of total production, and exhibited

what must be near the upper limit of possibility for snail production. Clearly, this is the result of

a species relieved of natural limitations to its production, as these values are not realized within

its native range (Hall et al. 2006).

60 Biomass

60 Production % Contribution 50

0 ra ra ra ea da ia da te te te c li lv o ip p p ta e va p D o ro s nn i ro ch e ru A B st ri m C a T e G ph E

Fig. 5.5. Habitat-weighted percent contribution of major taxa to the mean annual biomass and annual production of Hargrove Shoals.

Production vs. Biomass estimates

Several major taxonomic groups examined in this study illustrate the importance of using

secondary production as opposed to biomass when attempting to understand the contributions of individual taxa or functional groups to the function of an ecosystem. For example the biomass of

the chironomid and simuliid Diptera was universally low at both spatial scales compared to trichopterans or mollusks (Table 5.1, Table 5.2, Fig. 5.5). Despite this, the dipteran contribution

to assemblage production was substantial, owing to their high rates of individual growth and

consequently high rates of annual biomass turnover (annual P/Bs ranging from 72.9 to 260 y-1)

(Table 5.1, Fig. 5.5). Similarly, the baetid, caenid, heptageniid, and tricorythodid mayflies accounted for very little biomass compared to their contribution to assemblage production.

Annual P/Bs for these taxa ranged from 52.6 to 95.8 y-1 (Table 5.1). These extremely high

growth rates and annual P/Bs were first documented in studies of similar populations in the

Ogeechee River, Georgia (Benke 1998, Hauer and Benke 1987, Benke and Jacobi 1994), but

have been demonstrated in several studies since then.

In contrast to the distribution of production among the major taxonomic groups, the distribution of biomass among groups weighs heavily toward the mollusks, particularly the snails

(Fig. 5.5). Snails are the most visually conspicuous invertebrates on the shoals, but their

dominance of biomass (Fig. 5.5, upper graph) does not result in dominance of production (Fig.

5.5, lower graph). This is because low daily growth rates (Chapter 3) and long life spans (also

see Huryn et al. 1994) lead to extremely low annual P/Bs (ranging from 0.7 to 1.7 y-1 for

Pleuroceridae), and thus relatively low annual production (Table 5.1). Clearly, one would come

to dramatically different conclusions about the relative contributions of these major taxonomic

groups to ecosystem function if biomass was the sole metric.

Conclusions

Streams are often described as mosaics of habitats at multiple spatial and temporal scales

(Hynes 1970, Pringle et al. 1988, Palmer and Poff 1997, Ward et al. 2002), and many studies have demonstrated the importance of considering the relative contributions of specific habitat types to the overall diversity and functional structure of a stream reach (Benke et al. 1984, Huryn and Wallace 1987, Grubaugh et al. 1997). Still, relatively few attempts have been made at describing the habitat-specific production of entire (or nearly entire) macroinvertebrate assemblages of unregulated rivers. This study contributes to this effort by attempting to characterize the distribution of production across a complex Fall Line shoal reach of the Cahaba

River, the longest unregulated river in the state of Alabama. This study also represents the first major effort to describe the relative contributions of various habitats, particularly that of Justicia americana, to reach diversity and production. Finally, no other study has documented the

production of such a diverse snail assemblage. Hopefully this work will provide a useful

foundation from which to build future research in this area, and provide a basis for conservation

of shoal habitats and species.

As demonstrated here and elsewhere, shoals are important sites of habitat diversity and

production. In Hargrove Shoals, both Podostemum and Justicia proved to be highly productive

at the habitat-scale, while bedrock revealed its functional importance more so at the reach-scale,

due to its spatial dominance. One habitat, Justicia, has proven to be especially influential at both

scales. At the habitat-scale, it harbors a higher diversity than bare bedrock or Podostemum, while supporting similarly high levels of production as Podostemum. At the reach-scale, this

high production is made even more significant (relative to other habitats) by its expansive

distribution across the shoal reach.

Acknowledgements

I would like to thank Mark Dedmon for field assistance, Laura Frost for lab assistance,

Lori Tolley-Jordan for her expertise on Cahaba snails, and the Cahaba River National Wildlife

Refuge for access to Hargrove Shoals. This work was completed under permit number

TE163435-0 from the USFWS granted to Dr. Arthur C. Benke.

Literature Cited

Benke, A.C. 1979. A modification of the Hynes method for estimating secondary production with particular significance for multivoltine populations. Limnology and Oceanography 24:168-174.

Benke, A. C. 1993 Baldi Memorial Lecture: Concepts and patterns of invertebrate production in running waters. Verh. Internat. Verein. Limnol. 25:15-38.

Benke, A.C. 1998. Production dynamics of riverine chironomids: extremely high biomass turnover rates of primary consumers. Ecology 79:899-910.

Benke, A.C. and D.I. Jacobi. 1994. Production dynamics and resource utilization of snag- dwelling mayflies in a blackwater river. Ecology 75:1219-1232.

Benke, A.C., T.C. Van Arsdall, Jr., and D.M. Gillespie. 1984. Invertebrate productivity in a subtropical blackwater river: the importance of habitat and life history. Ecological Monographs 54:25-63.

Berrie, A.D. 1972. Productivity of the River Thames at Reading. Symp. Zool. Soc. Lond. 29:69-86.

Brown, B.L. 2003. Spatial heterogeneity reduces temporal variability in stream insect communities. Ecology Letters 6:316-325.

Cross, W.F., C.V. Baxter, K.C. Donner, E.J. Rosi-Marshall, T.A. Kennedy, R.O. Hall, H.A. Wellard Kelly, and R.S. Rogers. 2011. Ecosystem ecology meets adaptive management: food web response to a controlled flood on the Colorado River, Glen Canyon. Ecological Applications 21:2016–2033.

Downes, B.J., P.S. Lake, E.S.G. Schreiber, and A. Glaister. 1998. Habitat structure and regulation of local species diversity in a stony, upland stream. Ecological Monographs 68:237-257.

Edwards, C.J., P.L. Hudson, W.G. Duffy, S.J. Nepszy, C.D. McNabb, R.C. Haas, C.R. Liston, B. Manny, W.N. Busch. 1989. Hydrological, morphometrical, and biological characteristics of the connecting rivers of the International Great Lakes: a review. Pages 240-264 in D.P. Dodge (editor). Proceedings of the International Large River Symposium. Can. Spec. Publ. Fish. Aquat. Sci. 106.

Edwards, F.K., R.B. Lauridsen, L. Armand, H.M. Vincent, and J.I. Jones. 2009. The relationship between length, mass and preservation time for three species of freshwater leeches (Hirudinea). Fundamental and Applies Limnology Archiv. Fur Hydrobiologie 173:321-327.

Flossner, D. 1976. Biomasse und production des lithalen makrobenthos der mittleren Saale. Limnologica 10:123-153.

Flossner, D. 1982. Untersuchungen zur biomasse und production des makrozoobenthos der Ilm und mittleren Saale. Limnologica 14:297-327.

Fritz, K.M., and J.W. Feminella. 2003. Substratum stability associated with the riverine macrophyte Justicia americana. Freshwater Biology 48:1630-1639.

Fritz, K.M., M.M. Gangloff, and J.W. Feminella. 2004. Habitat modification by the stream macrophyte Justicia americana and its effects on biota. Oecologia 140:388-397.

Gregg, W.W., and F.L. Rose. 1982. The effects of aquatic macrophytes of the stream microenvironment. Aquatic Botany 14:309-324.

Grubaugh, J.W., and J.B. Wallace. 1995. Functional structure and production of the benthic community in a Piedmont river: 1956-1957 and 1991-1992. Limnology and Oceanography 40:490-501.

Grubaugh, J.W., J.B. Wallace, and E.S. Houston. 1997. Production of benthic macroinvertebrate communities along a southern Appalachian river continuum. Freshwater Biology 37:581-596.

Hall, R.O., M.F. Dybdahl, and M.C. Vanderloop. 2006. Extremely high secondary production of introduced snails in rivers. Ecological Applications 16:1121-1131.

Hauer, F.R., and A.C. Benke. 1987. Influence of temperature and river hydrograph on black fly growth rates in a subtropical blackwater river. Journal of the North American Benthological Society 6:251-261.

Huryn, A.D., and J.B. Wallace. 1987. Local geomorphology as a determinant of macrofaunal production in a mountain stream. Ecology 68:1932-1942.

Hynes, H.B.N. 1970. The ecology of running waters. University of Toronto Press, Toronto.

Krueger, C.C., and T.F. Waters. 1983. Annual production of macro-invertebrates in three streams of different water quality. Ecology 64:840-850.

Lancaster, J. 2000. Geometric scaling of microhabitat patches and their efficacy as refugia during disturbance. Journal of Animal Ecology 69:442-457.

Lancaster, J., and L.R. Belyea. 1997. Nested hierarchies and scale-dependence of mechanisms of flow refugium use. Journal of the North American Benthological Society 16:221-238.

Merritt, R.W., and K.W. Cummins. 1996. An Introduction to the Aquatic Insects, 3rd edition. Kendall/Hunt, Dubuque, Iowa.

Méthot, G., C. Hudon, P. Gagnon, B. Pinel-Alloul, A. Armellin, and A. Tourville Poirier. 2012. Macroinvertebrate size–mass relationships: how specific should they be? Freshwater Science 31:750-764.

Nelson, D.J., and D.C. Scott. 1962. Role of detritus in the productivity of a rock-outcrop community in a Piedmont stream. Limnology and Oceanography 7:396-413.

Palmer, M.A., and N.L. Poff. 1997. Hetergeneity in streams: the influence of environmental heterogeneity on patterns and processes in streams. Journal of the North American Benthological Society 16:169-173.

Pringle, C.M., R.J. Naiman, G. Bretscho, J.R. Karr, M.W. Oswood, J.R. Webster, R.L. Welcomme, and M.J. Winterbourn. 1988. Patch dynamics in lotic systems: the stream as a mosaic. Journal of the North American Benthological Society 7:503-524.

Smith, D.G. 2001. Pennak's Freshwater Invertebrates of the United States: Porifera to Crustacea, 4th edition. John Wiley and Sons, Inc.

Smock, L.A., E. Gilinsky, and D.L. Stoneburner. 1985. Macroinvertebrate production in a southeastern United States stream. Ecology 66:1491-1503.

Taniguchi, H., and M. Tokeshi. 2004. Effects of habitat complexity on benthic assemblages in a variable environment. Freshwater Biology 49:1164-1178.

Thorp, J.H., and A.P. Covich (eds.) 2010. Ecology and classification of North American freshwater invertebrates. Academic Press (Elsevier).

Tolley-Jordan, L.R. 2008. The biology of Pleuroceridae (Gastropoda: Caenogastropoda) in the Cahaba River basin, Alabama, USA. Ph.D. dissertation, The University of Alabama, Tuscaloosa, Alabama.

U.S. Fish and Wildlife Service. 2002. Cahaba River national wildlife refuge. Federal Register Document 02-32210. 67:78238.

Voshell, J.R. 1985. Trophic basis of production for macroinvertebrates in the New River below Bluestone Dam. Department of Entomology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia.

Ward, J.V., K. Tockner, D.B. Arscott, and C. Claret. 2002. Riverine landscape diversity. Freshwater Biology 47:517-539.

100

Waters, T.F. 1969. Subsampler for dividing large samples of invertebrate drift. Limnology and Oceanography 14:813-815

Waters, T.F. 1977. Secondary production in inland waters. Advances in Ecological Research 10:91-164.

Williams, C.J., and R.F. McMahon. 1989. Annual variation of tissue biomass and carbon and nitrogen content in the freshwater bivalve Corbicula fluminea relative to downstream dispersal. Canadian Journal of Zoology 67:82-90.

101

Appendix 5.1. List of all benthic invertebrate taxa collected from three main habitats of Hargrove Shoals, where BR = bedrock, POD = Podostemum ceratophyllum, and JUS = Justicia americana habitats. Presence in each habitat is indicated by an X. Totals at bottom indicate number of taxa found across all habitats (in parenthesis), and in each habitat sampled.

Habitat Taxon BR POD JUS Annelida Hirudinea X X X Oligochaeta X X X Crustacea Decapoda Cambaridae Orconectes X Amphipoda Hyalellidae Hyalella X X X Copepoda Harpacticoida X X X Isopoda Asellidae Lirceus X X X Ostracoda X X X Coleoptera Elmidae Ancyronyx X Dubiraphia X X Machronychus X X Microcylloepus X X X Neoelmis X X Stenelmis X X X Hydrophilidae Berosus X X Lampyridae X Psephenidae Ectopria X Psephenus X

102

Appendix 5.1. Continued. Habitat Taxon BR POD JUS Diptera Ceratopogonidae Atrichopogon X X X Bezzia X Culicoides X Chironomidae Chironomini X X X Orthocladiinae X X X Tanypodinae X X X Tanytarsini X X X Empididae Hemerodromia X X X Ephydridae Hydrellia X Notophila X Simuliidae Simulium X X X Tipulidae Helius X Tipula X Ephemeroptera Baetidae Baetis X X X Heterocloeon X X X Caenidae Caenis X X X Ephemerellidae Eurylophella X Serratella X X X Ephemeridae Ephemera X Hexagenia X Heptageniidae Stenacron X X Stenonema X X X

103

Appendix 5.1. Continued. Habitat Taxon BR POD JUS Isonychiidae Isonychia X X X Tricorythidae Tricorythodes X X X Hymenoptera Pteromalidae X Lepidoptera Pyralidae Crambus X Petrophila X X X Noctuidae Simyra X X Megaloptera Corydalidae Corydalus X X Odonata Aeshnidae Boyeria X Calopterygidae Haeterina X X Coenagrionidae Amphiagrion X X Argia X X X Enallagma X Gomphidae Dromogomphus X Gomphus X Plecoptera Leuctridae Leuctra X Nemouridae Amphinemoura X Perlidae Neoperla X Perlesta X

104

Appendix 5.1. Continued. Habitat Taxon BR POD JUS Taeniopterygidae Taeniopteryx X X Psocoptera X X Trichoptera Brachycentridae Brachycentrus X Micrasema X Glossosomatidae Glossosoma X Hydropsychidae Cheumatopsyche X X X Hydropsyche X X X Hydroptilidae Hydroptila X X X Ochrotrichia X Limnephilidae Pycnopsyche X Philopotamidae Chimarra X X X Polycentropodidae Nyctiophylax X X Polycentropus X Arachnida Acari X X X Diplopoda Polydesmida X Bivalvia Corbiculidae Corbicula fluminea X X X Unionidae Obliquaria X Gastropoda Ancylidae X X X Hydrobiidae Lepyrium showalteri X X X Somatogyrus X X X

105

Appendix 5.1. Continued. Habitat Taxon BR POD JUS Physidae Physa X X Planorbidae X X Pleuroceridae Elimia ampla X X X Elimia cahawbensis X X X Elimia clara X X X Elimia showalteri X X X Leptoxis ampla X X X Pleurocera vestita X X X Viviparidae Campeloma X Lioplax cyclostomaformis X X Nematoda X X Hydrozoa Hydridae Hydra X Turbellaria Tricladida Dugesiidae X X X total (90) 49 57 80

106

Appendix 5.2. Annual summary data for the 41 most common taxa found in three main habitats of Hargrove Shoals. Habitats include: BR = bedrock, POD = Podostemum ceratophyllum, and JUS = Justicia americana. N = mean annual density (no./m2 ± 1 SE), B = mean annual biomass (mg ash-free dry mass [AFDM]/m2 ± 1 SE), P = annual production (mg AFDM m-2 y-1), P/B = annual P/B (y-1), CPI = cohort production interval (mo). Method = technique used to estimate P, where IG = instantaneous growth, SF = size- frequency, and P/B = P/B method. Ref. = source of IG model, or CPI data. Under Ref., a = Smock et al. 1985, b = Martien & Benke 1977, c = Pickard & Benke 1996, d = Benke & Jacobi 1994, e = Benke 1993, f = Hall et al. 2011, g = Benke & Wallace 1997, h = Benke 2002, i = Benke 1998, j = Hauer & Benke 1987, k = SF data, l = Assumed, m = Chapter 3, n = Assumed from Huryn et al. 1994, o = Stites et al. 1995. Totals for B and P are reported in g AFDM/m2 and g AFDM m-2 y-1, respectively.

Habitat Taxon N B P P/B CPI Method Ref. BR Tricladida Dugesiidae 7.3 ± 7.0 0.0 ± 0.0 0.1 5.6 12 P/B a Oligochaeta -- 1.5 ± 1.2 11.5 10 12 P/B a Hirudinea 6.5 ± 5.7 19.6 ± 19.3 75.9 3.9 12 P/B a Isopoda Lirceus spp. 1.8 ± 1.4 0.7 ± 0.5 6.5 8.9 12 P/B b Amphipoda Hyalella spp. 0.7 ± 0.7 0.3 ± 0.3 3.5 11.86 P/B c Ephemeroptera Baetis spp. 176.8 ± 98.9 9.5 ± 3.7 780.6 81.9-- IG d Heterocleon spp. 67.8 ± 56.1 11.0 ± 9.7 1012.7 92.3-- IG d Caenis sp. 1.1 ± 1.1 0.3 ± 0.3 27.4 87.5 -- IG d Serratella sp. 0.6 ± 0.4 0.1 ± 0.1 1.2 9.56 P/B d Stenonema spp. 4.3 ± 2.3 3.7 ± 2.5 288.7 78.4-- IG d Isonychia sp. 3.2 ± 2.9 0.0 ± 0.0 0.1 9.26 P/B d Tricorythodes spp. 48.2 ± 46.1 2.4 ± 2.2 136.3 56.1-- IG d Odonata Haeterina sp. 0 0 0 -- 12 -- a Argia spp. 0.4 ± 0.4 0.0 ± 0.0 0.1 4.8 12 P/B a

107

Appendix 5.2. Continued Habitat Taxon N B P P/B CPI Method Ref. BR Megaloptera Corydalus sp. 0 0 0 -- 12 -- e Lepidoptera Petrophila sp. 24.3 ± 15.5 6.4 ± 3.9 30.9 6.4 11 SF f Trichoptera Cheumatopsyche spp. 33.5 ± 15.7 11.1 ± 4.8 182.8 11.16 SF g Hydropsyche spp. 2387.5 ± 1598.0 845.4 ± 401.4 17,259.5 845.46 SF g Hydroptila spp. 281.2 ± 161.9 11.4 ± 4.8 156.7 11.46 SF g Chimarra sp. 0.7 ± 0.5 0.1 ± 0.1 1.4 0.16 P/B g Nyctiophylax sp. 0 0 0 -- 6 -- e Coleoptera Microcylloepus spp. 40.1 ± 18.6 9.2 ± 4.3 31.9 3.5 12 SF h Stenelmis spp. 55.7 ± 23.3 6.4 ± 2.7 39.6 6.2 12 SF h Diptera Chironomini 1265.1 ± 399.8 33.4 ± 7.1 7783.4 233.4-- IG i Orthocladiinae 3413.9 ± 1619.9 117.2 ± 60.4 18253.1 155.8-- IG i Tanytarsini 247.4 ± 76.7 2.1 ± 0.6 440.7 211.2-- IG i Tanypodinae 9.2 ± 6.2 0.4 ± 0.2 93.3 257.7-- IG i Simulium spp. 274.6 ± 174.4 11.5 ± 6.9 1011.3 87.71-- IG i Tipula spp. 0 0 0 -- 4 SF j

108

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. BR Gastropoda Ancylidae 4.3 ± 3.1 0.1 ± 0.0 0.4 5.3 12 P/B k Lepyrium showalteri 5.5 ± 2.8 2.5 ± 1.4 12.7 5.0 12 P/B k Somatogyrus spp. 85.1 ± 46.9 6.4 ± 3.1 30.1 4.7 18 SF j Elimia ampla 35.7 ± 18.8 707.1 ± 383.5 594.9 0.8-- IG l Elimia cahawbensis 6.9 ± 4.0 86.0 ± 53.2 98.1 1.1-- IG l Elimia clara 57.8 ± 26.7 625.7 ± 274.4 666.6 1.1-- IG l Elimia showalteri 73.6 ± 35.9 1335.1 ± 573.7 953.3 0.7-- IG l Pleurocera vestita 2.7 ± 1.4 79.3 ± 36.0 64.8 0.8-- IG l Leptoxis ampla 294.3 ± 84.0 6923.2 ± 2147.2 6082.3 0.9-- IG l Campeloma sp. 0 0 0 -- 36 P/B m Lioplax cyclostomaformis 0 0 0 -- 36 P/B m Bivalvia Corbicula fluminea 55.1 ± 32.0 0.8 ± 0.4 1.5 1.8 36 P/B n

Total 8972.9 ± 4588.8 10.9 ± 4.0 56.1 5.4

109

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. POD Tricladida Dugesiidae 776.8 ± 509.7 35.1 ± 21.2 208.1 7.6 12 SF a Oligochaeta -- 9.5 ± 7.3 72.8 10 12 P/B a Hirudinea 0.4 ± 0.4 1.5 ± 1.5 5.9 3.9 12 P/B a Isopoda Lirceus spp. 116.4 ± 111.2 11.8 ± 11.1 105.8 8.9 12 P/B b Amphipoda Hyalella spp. 6.3 ± 5.0 0.6 ± 0.4 6.8 11.86 P/B c Ephemeroptera Baetis spp. 1628.3 ± 708.8 182.1 ± 86.7 14,050.1 77.1-- IG d Heterocloeon spp. 1100.0 ± 598.8 73.2 ± 38.4 7008.5 95.8-- IG d Caenis sp. 6.5 ± 5.7 6.0 ± 5.3 315.0 52.6-- IG d Serratella sp. 76.6 ± 45.1 34.6 ± 22.9 328.8 9.56 P/B d Stenonema spp. 51.9 ± 25.2 25.7 ± 17.4 1852.3 72.0-- IG d Isonychia sp. 8.2 ± 5.7 6.0 ± 4.1 55.7 9.26 P/B d Tricorythodes spp. 174.9 ± 69.2 23.3 ± 13.3 1369.5 58.7-- IG d Odonata Haeterina sp. 7.2 ± 6.9 0.8 ± 0.7 3.4 4.4 12 P/B a Argia spp. 1.1 ± 0.5 1.6 ± 1.1 7.9 4.8 12 P/B a

110

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. POD Megaloptera Corydalus sp. 2.9 ± 2.9 0.3 ± 0.3 2.4 7.5 12 P/B e Lepidoptera Petrophila sp. 18.3 ± 11.2 13.9 ± 9.0 82.0 5.9 11 SF f Trichoptera Cheumatopsyche spp. 3555.2 ± 921.1 2119.4 ± 510.2 27,627.8 13.06 SF g Hydropsyche spp. 10,283.4 ± 4125.7 6483.2 ± 1052.3 83,213.9 11.86 SF g Hydroptila spp. 366.6 ± 173.0 19.3 ± 9.8 250.0 13.06 SF g Chimarra sp. 515.0 ± 330.7 16.6 ± 12.6 163.1 9.86 P/B g Nyctiophylax sp. 18.3 ± 14.6 5.7 ± 4.1 48.5 8.56 P/B e Coleoptera Microcylloepus spp. 1017.7 ± 519.0 183.7 ± 99.9 649.8 3.5 12 SF h Stenelmis spp. 3979.9 ± 1141.7 724.8 ± 231.5 3824.3 5.3 12 SF h Diptera Chironomini 4726.0 ± 653.6 118.0 ± 20.8 30,673.3 260.0-- IG i Orthocladiinae 9998.1 ± 3746.7 271.0 ± 122.4 40,450.4 149.3-- IG i Tanytarsini 3991.1 ± 1383.3 35.5 ± 16.6 6534.4 183.8-- IG i Tanypodinae 34.4 ± 34.4 33.0 ± 33.0 3571.4 108.2-- IG i Simulium spp. 3841.4 ± 1883.4 215.7 ± 108.7 15,716.3 72.86-- IG i Tipula spp. 0 0 0 -- 4 P/B j

111

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. POD Gastropoda Ancylidae 11.5 ± 11.5 0.2 ± 0.2 1.0 5.3 12 P/B k Lepyrium showalteri 9.2 ± 5.6 2.2 ± 1.3 10.8 5.0 12 P/B k Somatogyrus spp. 2574.8 ± 1268.1 215.8 ± 71.3 1098.1 4.7 18 SF j Elimia ampla 64.3 ± 24.9 1028.6 ± 305.0 876.6 0.9-- IG l Elimia cahawbensis 136.9 ± 71.9 1087.8 ± 661.9 1129.1 1.0-- IG l Elimia clara 144.1 ± 27.3 1365.1 ± 368.8 1453.2 1.1-- IG l Elimia showalteri 284.3 ± 74.9 3488.6 ± 1239.2 2592.4 0.7-- IG l Pleurocera vestita 9.8 ± 5.9 350.5 ± 290.4 211.5 0.6-- IG l Leptoxis ampla 199.7 ± 52.4 2892.0 ± 811.4 2650.5 0.9-- IG l Campeloma sp. 0 0 0 -- 36 P/B m Lioplax cyclostomaformis 0.9 ± 0.7 90.3 ± 73.4 150.5 1.7 36 P/B m Bivalvia Corbicula fluminea 79.7 ± 25.3 6.1 ± 2.7 10.9 1.8 36 P/B n

Total 49,818.1 ± 18,602.0 21.2 ± 6.3 248.4 11.9

112

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. JUS Tricladida Dugesiidae 120.6 ± 51.7 19.5 ± 11.5 57.5 3.5 12 SF a Oligochaeta -- 573.8 ± 438.4 4384.2 10 12 P/B a Hirudinea 112.7 ± 26.0 2283.5 ± 906.6 8765.4 3.8 12 SF a Isopoda Lirceus spp. 7337.8 ± 6766.8 1133.4 ± 990.2 10,143.2 8.9 12 SF b Amphipoda Hyalella spp. 202.0 ± 126.9 55.0 ± 28.5 648.3 11.86 SF c Ephemeroptera Baetis spp. 309.2 ± 175.0 32.9 ± 13.7 2628.5 79.8-- IG d Heterocloeon spp. 51.9 ± 51.9 7.2 ± 7.2 650.1 90.3-- IG d Caenis sp. 5.7 ± 5.7 0.1 ± 0.1 8.1 58.3-- IG d Serratella sp. 56.3 ± 41.9 39.7 ± 28.8 377.0 9.56 SF d Stenonema spp. 267.7 ± 104.0 88.2 ± 28.8 7108.3 80.6-- IG d Isonychia sp. 35.5 ± 13.5 46.0 ± 21.3 425.0 9.26 SF d Tricorythodes spp. 200.3 ± 64.1 25.7 ± 8.1 1601.1 62.3-- IG d Odonata Haeterina sp. 34.7 ± 18.7 57.1 ± 24.8 255.0 4.4 12 SF a Argia spp. 23.2 ± 12.8 21.9 ± 9.5 71.6 4.8 12 SF a

113

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. JUS Megaloptera Corydalus sp. 14.0 ± 14.0 5.2 ± 5.2 39.2 7.5 12 P/B e Lepidoptera Petrophila sp. 15.5 ± 14.3 4.4 ± 3.7 16.2 3.7 11 SF f Trichoptera Cheumatopsyche spp. 3186.5 ± 1211.4 1838.0 ± 551.9 26189.3 14.36 SF g Hydropsyche spp. 2470.7 ± 1151.9 1250.6 ± 367.1 16622.8 12.56 SF g Hydroptila spp. 97.2 ± 39.0 7.3 ± 3.9 80.5 11.16 SF g Chimarra sp. 386.8 ± 204.4 162.6 ± 46.7 1599.9 9.86 SF g Nyctiophylax sp. 3.9 ± 2.9 7.0 ± 6.2 59.0 8.56 P/B e Coleoptera Microcylloepus spp. 1056.2 ± 365.9 347.6 ± 133.8 1230.5 3.5 12 P/B h Stenelmis spp. 935.6 ± 345.6 380.3 ± 114.7 1653.6 4.4 12 P/B h Diptera Chironomini 1227.0 ± 459.0 42.6 ± 12.0 11104.5 260.9-- IG i Orthocladiinae 4537.4 ± 1805.8 149.7 ± 72.2 23094.5 154.2-- IG i Tanytarsini 1004.4 ± 289.1 15.4 ± 5.8 2945.1 191.2-- IG i Tanypodinae 18.9 ± 8.4 1.6 ± 0.8 394.7 251.1-- IG i Simulium spp. 1883.3 ± 921.4 129.2 ± 58.4 9497.2 73.49-- IG i Tipula spp. 72.1 ± 44.8 1010.4 ± 529.5 13054.6 13.14 P/B j

114

Appendix 5.2. Continued. Habitat Taxon N B P P/B CPI Method Ref. JUS Gastropoda Ancylidae 81.8 ± 52.0 3.0 ± 2.1 15.7 5.3 12 SF k Lepyrium showalteri 16.1 ± 10.4 9.6 ± 7.1 47.8 5.0 12 P/B k Somatogyrus spp. 962.0 ± 470.6 72.1 ± 36.8 277.5 4.1 18 SF j Elimia ampla 104.4 ± 73.5 860.6 ± 439.1 624.7 0.7-- IG l Elimia cahawbensis 178.1 ± 33.0 1484.6 ± 337.1 1495.2 1.0-- IG l Elimia clara 378.8 ± 209.7 2278.3 ± 634.9 2235.2 1.0-- IG l Elimia showalteri 400.3 ± 162.9 4361.5 ± 1611.9 3042.1 0.7-- IG l Pleurocera vestita 97.1 ± 37.8 2149.5 ± 865.8 1519.7 0.7-- IG l Leptoxis ampla 65.0 ± 15.4 1463.6 ± 423.1 1191.3 0.8-- IG l Campeloma sp. 4.3 ± 2.4 834.7 ± 528.3 1391.2 1.7 36 P/B m Lioplax cyclostomaformis 55.6 ± 15.7 3945.3 ± 1356.5 6575.5 1.7 36 P/B m Bivalvia Corbicula fluminea 3292.7 ± 1371.9 7949.7 ± 1060.4 14164.5 1.8 36 SF n

Total 31,303.1 ± 16,792.5 35.0 ± 11.6 177.3 5.2

115

CHAPTER 6

GENERAL CONCLUSIONS

Fall Line shoals represent extraordinary zones of geomorphic variability and complexity

within a river basin. The dominance of erosion resistant bedrock substrata leads to extreme

channel widths (up to 150 m) and shallow average depths (20 cm), allowing for full exposure to

the sun and the potential for high levels of autochthonous primary production of algae, moss, and

vascular macrophytes (Vannote et al. 1980). Plants successfully establish themselves within the

fissures and voids in the bedrock substrata (e.g., Justicia americana, Hymenocallis coronaria), or by adhesion to rock surfaces with holdfast anchors (e.g., Podostemum ceratophyllum). This unique combination of channel features and in-stream macrophytes work in concert to create ideal habitat conditions for a diverse and productive assemblage of benthic macroinvertebrate taxa. The habitat heterogeneity found in Fall Line shoals makes habitat-specific secondary production analysis an optimal approach to address questions regarding how habitat distribution influences ecosystem function.

The central question of this dissertation was: How do the major habitats of a Cahaba

River shoal influence the distribution and secondary production of the macroinvertebrate assemblage? In order to provide information about the physical habitats available to the invertebrate assemblage, Chapter 2 provided a detailed description and quantification of the major habitats across the reach. The habitat survey helped focus invertebrate sampling efforts toward the 3 most abundant habitats: bare bedrock, bedrock covered by Justicia americana, and

116

bedrock covered by Podostemum ceratophyllum, which covered 38%, 26%, and 8%,

respectively, of reach surface area in 2008. This relative distribution data was later used to

estimate reach-scale assemblage production of the non-native Asiatic Clam (Corbicula fluminea) in Chapter 4, and of the macroinvertebrate assemblage in Chapter 5. Emergent biomass of

Justicia was comparable to that of the New River, Virginia, while submersed biomass was considerably lower due likely to the bedrock substratum, which may limit the biomass attainable in this reach (Hill 1981). Podostemum biomass was similar to that found in other southeastern

shoals (Hill and Webster 1984, Grubaugh and Wallace 1995, Grubaugh et al. 1997). An

examination of the temporal distribution of Justicia revealed consistently high levels of

submersed biomass, even though emergent biomass senesced during the winter. This

persistently high biomass, combined with its expansive distribution, suggests an important role

for this plant in modifying benthic habitat conditions for macroinvertebrates. The seasonally

senescent emergent biomass also provides a potentially important input of organic matter for

detritivore taxa (Hill 1981).

Chapter 3 described a new method for obtaining estimates of in situ growth rates of a

diverse pleurocerid snail assemblage. Growth rates observed with this tethering method

displayed a significant negative relationship with increasing body size. However, because mean

interval temperature showed no significant correlation with snail growth for two of the three

species, final growth models were based only on snail body size. Estimates generated with these

models were relatively low, similar to those reported in the literature, which validated this technique. Regression models developed in this chapter were later used in Chapter 5 to estimate the production of this important group via the instantaneous growth method.

117

During sampling efforts, it became apparent that the nonnative Asiatic Clam, Corbicula fluminea, was quite abundant within the sediments associated with Justicia beds. This

observation provided the rationale for a more detailed examination of this species, which is the

most abundant and widespread nonnative invertebrate in the region. Therefore, Chapter 4

described the distribution of Corbicula among 3 primary shoal habitats, and provided an estimate

of its density, biomass, and secondary production in the Justicia habitat. Annual secondary

production of Corbicula in Justicia habitats was comparable to estimates reported by several

previous studies (Sickel 1976, Marsh 1985, and Stites et al. 1995). This study represents the first

production estimate of Corbicula in bedrock shoal habitats and illustrates the dramatic influence

of a shoal macrophyte in modifying otherwise inhospitable benthic habitat characteristics for

Corbicula.

Finally, Chapter 5 represented the culmination of the 3 preceding chapters by

characterizing the distribution and production of macroinvertebrate assemblages associated with

the 3 major shoal habitats, and estimating reach-scale production for the entire Cahaba River

shoal. Hargrove Shoals supports an impressive level of invertebrate diversity, higher than that found in similar shoal reaches of Georgia or North Carolina (Grubaugh and Wallace 1995,

Grubaugh et al. 1997). Taxa richness was greatest in Justicia, followed by Podostemum, and

then bedrock. While production in all habitats was high compared to most published assemblage

estimates, bedrock was the least productive of the habitats. Both vegetated habitats supported

high levels of invertebrate production, similar to previously published estimates for Podostemum

(Grubaugh and Wallace 1995, Grubaugh et al. 1997). Production in each habitat was dominated by two insect families: the hydropsychid caddisflies (due to high biomass), and the chironomid dipterans (due to extremely high biomass turnover rates). Podostemum was once again proven to

118

be an important site for insect production, with estimates of 3 major orders (i.e., Trichoptera,

Diptera, and Ephemeroptera) exceeding that found in either Justicia or especially bedrock.

Production in Justicia, however, was composed of a greater diversity of taxa, including several adapted to deposition zones. Justicia also contributed most to habitat-weighted reach production, due largely to its expansive distribution. Bedrock and Podostemum contributed similarly to reach production, due to the relatively limited distribution of Podostemum.

As this dissertation demonstrates, Fall Line shoals can possess a great variety of habitats, which can support a diverse and productive macroinvertebrate assemblage. While many studies have demonstrated the importance of considering the relative contributions of specific habitat types to the overall diversity and functional structure of a stream reach (Benke et al. 1984, Huryn and Wallace 1987, Grubaugh et al. 1997), few have estimated the contributions of major bedrock shoal habitats to the macroinvertebrate production of large free-flowing rivers. This dissertation represents the first effort to describe the production of the macroinvertebrate assemblage of such habitats throughout one of the largest shoal reaches of the Cahaba River, the longest unregulated river in the state of Alabama. The application of habitat-specific sampling, combined with knowledge of relative habitat distributions has allowed for a simultaneous view of the roles of several major habitats to both patch-scale as well as overall reach function.

As this and other studies have illustrated, Podostemum is an important site for the production of several major macroinvertebrate taxa. This plant is more abundant in some river shoal reaches (Grubaugh and Wallace 1995, Grubaugh et al. 1997), and completely dominates the functional character of these sites. While this was not the case in Hargrove Shoals, its influence on the production of the macroinvertebrate assemblage was much greater than its limited distribution would suggest. It has been suggested that Podostemum could be a useful

119

indicator species for healthy streams due to its relative intolerance to siltation and low dissolved oxygen (Meijer 1976, Hutchens et al. 2004). This being the case, periodic surveys of

Podostemum distribution in shoals where it is present may assist in revealing long term trends in overall ecosystem health.

Justicia has been studied with regards to its role as a physical habitat modifier in small

streams (Fritz and Feminella 2003, Fritz et al. 2004), but this is the first and most detailed

examination of its role as an important habitat for invertebrate production. Upon casual

observation, it is apparent that Justicia has a dominant presence on these shoals. This dissertation illustrated how not only the vast distribution of this plant influenced the overall reach, but also how the varied and unique microhabitats created by Justicia beds supported high

habitat-scale production as well as a diverse distribution of this production. Additionally,

Justicia provided important depositional habitat for many taxa, and what appeared to be critical

habitat for the federally endangered cylindrical lioplax snail (Lioplax cyclostomaformis). Efforts

in monitoring this imperiled species should certainly be concentrated in this habitat, where it was

often abundant.

The shoals of the Cahaba River represent some of the most significant remaining

examples of a channel feature that was once common throughout many rivers of the southeastern

United States. While we have some idea of the immensity of snail and unionid mussel

extinctions in river shoals caused by widespread river regulation (Lydeard et al. 1995, Williams et al. 2008), we will never fully understand the greater impacts to biotic diversity and ecosystem function without quantitative studies such as described in this dissertation. This dissertation

therefore has increased our ecological understanding of an endangered shoal habitat and will

hopefully contribute to a stronger rationale for its conservation.

120

LITERATURE CITED

Benke, A.C. 1984. Secondary production of aquatic insects. Pages 289-322 In V.H. Resh and D.M. Rosenberg (eds.). The Ecology of Aquatic Insects. Praeger, New York.

Benke, A.C. 1990. A perspective on America’s vanishing streams. Journal of the North American Benthological Society 9:77-88.

Benke, A.C., and C.E. Cushing (eds.). 2005. Rivers of North America. Elsevier Academic Press, London. 1144 pp.

Benke, A. C., and A. D. Huryn. 2006. Secondary production of macroinvertebrates. Pages 691- 710 In Hauer, F. R., and G. A Lamberti (eds.). Methods in Stream Ecology, 2nd ed. Academic Press, Burlington.

Benke, A.C., T.C. Van Arsdall, Jr., and D.M. Gillespie. 1984. Invertebrate productivity in a subtropical blackwater river: the importance of habitat and life history. Ecological Monographs 54(1):25-63.

Fritz, K.M., and J.W. Feminella. 2003. Substratum stability associated with the riverine macrophyte Justicia americana. Freshwater Biology 48:1630-1639.

Fritz, K.M., M.M. Gangloff, and J.W. Feminella. 2004. Habitat modification by the stream macrophyte Justicia americana and its effects on biota. Oecologia 140:388-397.

Grubaugh, J.W., and J.B. Wallace. 1995. Functional structure and production of the benthic community in a Piedmont river: 1956-1957 and 1991-1992. Limnology and Oceanography 40:490-501.

Grubaugh, J.W., J.B. Wallace, and E.S. Houston. 1997. Production of benthic macroinvertebrate communities along a southern Appalachian river continuum. Freshwater Biology 37:581-596.

Hill, B.H. 1981. Distribution and production of Justicia americana in the New River, Virginia. Castanea 46:162-169.

121

Hill, B.H., and J.R. Webster 1984. Productivity of Podostemum ceratophyllum in the New River, Virginia. American Journal of Botany 71:130-136.

Huryn, A.D., and J.B. Wallace. 1987. Local geomorphology as a determinant of macrofaunal production in a mountain stream. Ecology 68:1932-1942.

Hutchens, J.J.,Jr., J.B. Wallace, and E.D. Romaniszyn. 2004. Role of Podostemum ceratophyllum Michx. In structuring benthic macroinvertebrate assemblages in a southern Appalachian River. Journal of the North American Benthological Society 23:713-727.

Hynes, H.B.N. 1970. The ecology of sream insects. Annual Reviews of Entomology 15:25-42.

Lydeard, C., and R.L. Mayden. 1995. A diverse and endangered aquatic ecosystem of the Southeast United States. Conservation Biology 9:800-805.

MacArthur, R.H. 1972. Geographical Ecology: Patterns in the Distribution of Species. Princeton University Press, Oxford. 269 pp.

Marsh, P.C. 1985. Secondary production of introduced Asiatic clam Corbicula fluminea in a central Arizona canal. Hydobiologia 24:103-110.

Meijer, W. 1976. A note on Podostemum ceratophyllum Michx., as an indicator of clean streams in and around the Appalachian Mountains. Castanea 41:319-324.

Nelson, D.J., and D.C. Scott. 1962. Role of detritus in the productivity of a rock-outcrop community in a Piedmont stream. Limnology and Oceanography 7:396-413.

Southwood, T.R.E. 1977. Habitat, the templet for ecological strategies? Journal of Animal Ecology 46:336-365.

122

Southwood, T.R.E. 1988. Tactics, strategies and templets. Oikos 52:8-13.

Stanford, J. A., and J. V. Ward. 1979. Stream regulation in North America. Pages 215-236 in J. V. Ward and J. A. Stanford, editors. The ecology of regulated streams. Plenum Press, New York.

Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The river continuum concept. The Canadian Journal of Fisheries and Aquatic Science 37:130-137.

Vinson, M.R., and C.P. Hawkins. 1998. Biodiversity of stream insects: variation at local, basin, and regional scales. Annual Reviews of Entomology 43:271-293.

Wallace, J.B., and J.R. Webster. 1996. The role of macroinvertebrates in stream ecosystem function. Annual Review of Entomology 41:115-139.

Williams, J.D., A.E. Bogan, and J.T. Garner. 2008. Freshwater Mussels of Alabama and the Mobile Basin in Georgia, and Tennessee. University of Alabama Press, Tuscaloosa. 960 pp.

123