Abstract Poteat, Monica Deshay

Home , Ambassis jacksoniensis, Aquatic insect, Drunella cornutella, Drunella flavilinea, Drunella grandis, Drunella walkeri

ABSTRACT

POTEAT, MONICA DESHAY. Comparative Trace Metal Physiology in Aquatic Insects. (Under the direction of Dr. David B. Buchwalter).

Despite their dominance in freshwater systems and use in biomonitoring and bioassessment programs worldwide, little is known about the ion/metal physiology of aquatic insects. Even less is known about the variability of trace metal physiologies across aquatic insect species. Here, we measured dissolved metal bioaccumulation dynamics using radiotracers in order to 1) gain an understanding of the uptake and interactions of Ca, Cd and

Zn at the apical surface of aquatic insects and 2) comparatively analyze metal bioaccumulation dynamics in closely-related aquatic insect species. Dissolved metal uptake and efflux rate constants were calculated for 19 species. We utilized species from families

Hydropsychidae (order Trichoptera) and Ephemerellidae (order Ephemeroptera) because they are particularly species-rich and because they are differentially sensitive to metals in the field

– Hydropsychidae are relatively tolerant and Ephemerellidae are relatively sensitive. In uptake experiments with Hydropsyche sparna (Hydropsychidae), we found evidence of two shared transport systems for Cd and Zn – a low capacity-high affinity transporter below 0.8

µM, and a second high capacity-low affinity transporter operating at higher concentrations.

Cd outcompeted Zn at concentrations above 0.6 µM, suggesting a higher affinity of Cd for a shared transporter at those concentrations. While Cd and Zn uptake strongly co-varied across

12 species (r = 0.96, p < 0.0001), neither Cd nor Zn uptake significantly co-varied with Ca uptake in these species. Further, Ca only modestly inhibited Cd and Zn uptake, while neither

Cd nor Zn inhibited Ca uptake at concentrations up to concentrations of 89 nM Cd and 1.53

µM Zn. Ca, Cd and Zn bioaccumulation parameters all varied across orders of magnitude within the two families examined. Familial differences were striking across uptake rate constants and bioconcentration factors, however not across efflux rate constants. Body size was an important driver of uptake rate constants (and consequentially, bioconcentration factors). While the variation in Cd uptake and efflux rate constants was previously shown to be heavily influenced by phylogenetic position in species from orders Ephemeroptera,

Plecoptera and Trichoptera, the tremendous variability displayed by these two families effectively erased the phylogentic signal. In analyses of all available Cd uptake and efflux rate constants for aquatic insects from this study and from the literature, we discovered clade

(as genus) to explain significant amounts of variation across metal bioaccumulation parameters. Analyses across more taxonomically-divergent aquatic organisms identified phylogenetic signal in efflux rate constants across four aquatic phyla. Further, the strong co- variation of Zn and Cd efflux rate constants in hydropsychids and ephemerellids led to successful predictions of Zn uptake rate constants from known Cd values in five of six aquatic insect species. This work has led to two overall conclusions which should be taken into consideration when using aquatic insect data in regulatory toxicology. First, aquatic insects appear to have trace metal physiologies which differ from other aquatic organisms

(e.g., fish, Daphnia) regarding the trafficking of dissolved metal. Hypocalcemia does not appear to be the mechanism of toxicity in dissolved Cd/Zn as evidenced by the lack on interactions between Ca and Cd/Zn. This potentially contributes to their acute metal tolerance in the laboratory (as opposed to their observed metal sensitivity in the field). Second, aquatic insects, particularly within families Hydropsychidae and Ephemerellidae, have highly variable physiologies. This has important implications in that we need to better understand how well “surrogate species” represent their fellow congeners and account for the variation in aquatic risk assessments aimed at protecting insect diversity.

by Monica Deshay Poteat

A dissertation submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Toxicology

Raleigh, North Carolina

2014

APPROVED BY:

______Dr. David B. Buchwalter Dr. Gerald A. LeBlanc Committee Chair

______Dr. Tom Augspurger Dr. Eric A. Stone

BIOGRAPHY

Monica Deshay Poteat was born in Kannapolis, NC, on July 6, 1988, to parents Pat and Judy Poteat. After graduating from A.L. Brown High School in 2006, she stayed in

North Carolina to pursue a biology/pre-med degree at Elon University (because what else can you do with a biology degree other than go to medical school?). While at Elon, Monica quickly became interested in environmental research. She was able to conduct independent research projects on tree allelopathy in Great Smoky Mountains National Park and nitrogen cycling at the Virginia Institute for Marine Sciences. It was during these research experiences that Monica decided that the laboratory was where she would rather base her career.

During her last year and a half at Elon, Monica began researching graduate programs in the biological sciences. It was during this time that she stumbled upon the field of

Environmental Toxicology. She began emailing faculty in programs of interest to her, and became especially interested in the comparative, interdisciplinary work in Dr. David

Buchwalter’s laboratory (despite never having even thought about aquatic insects before).

Monica joined the Buchwalter lab in June of 2010, 10 days after graduating from Elon

University with a B.S. in Biology. She hit the ground running and hasn’t looked back since.

ACKNOWLEDGMENTS

First and foremost, I’d like to thank my advisor, Dr. David Buchwalter. He has been an exemplary major (and life) advisor. Under his guidance, I have become a better scientist, better writer, better presenter, and I’ve learned to better balance work and play. Through his confidence in me, I gained confidence in myself. I would also like to thank my committee members Dr. Gerald LeBlanc, Dr. Tom Augspurger and Dr. Eric Stone. I greatly appreciate their input and assistance throughout my time at NC State.

I am indebted to the members of the Buchwalter laboratory, past and present, including Dr. Justin Conley, Allison Camp, Jeanne Burr, Shane Scheibener, Beth Dittman,

Dr. Kyoung Sun Kim, and Dr. Lingtian Xie, for their assistance in the field and laboratory.

Other Toxicology folks were also extremely helpful throughout my time at NC State. Janet

Roe kept me on track with what forms to fill out and when to do it. Jackie Broughton was especially helpful with funding questions and the many travel authorizations that took me from Raleigh to the Smokies.

Other outside collaborators were invaluable to this project. Dr. Ted Garland, Jr. provided assistance with learning phylogenetic statistics and the software programs used to calculate the statistics. Dr. Luke Jacobus provided invaluable aid in insect identification throughout my research and drove many (many many) miles to aid in insect collection in the

Smokies. Luke also was also always willing to lend an open ear or a helping hand on anything I asked. Eric Fleek provided many spur-of-the-moment insect identifications that were greatly appreciated.

iii

I thank my family and friends for their support and never-ending encouragement throughout this journey through graduate school. I thank my parents for always encouraging me to follow my dreams, for cheering me on throughout my academic career, and for always being supportive of any decisions, good or bad, along the way. I especially thank my husband, Graham Medlin, for his love and patience throughout my graduate career. He made many trips to the Smokies with me as an extra pair of hands, and he was always willing to keep me company during the many late nights spent in lab running long time course experiments (and was always up for a Cook-Out run after a late night in the lab).

Last, but certainly not least, this work would not have been possible without my funding sources. The National Science Foundation, the SETAC/ICA Chris Lee Award for

Metals Research, and North Carolina State University generously provided funding which made the completion of this project possible.

TABLE OF CONTENTS

LIST OF TABLES ...... vii LIST OF FIGURES ...... ix

General Introduction ...... 1 Overview on aquatic insects ...... 1 Aquatic insects in ecological assessments ...... 2 Trace metals and their effects on aquatic insects in the laboratory and field ...... 3 Modeling of trace metal bioaccumulation ...... 5 Areas of concern in regulatory toxicology ...... 8 Comparative phylogenetic methods ...... 10 Research objectives and approach ...... 12 References ...... 16

CHAPTER 1: Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna...... 25 Abstract ...... 26 Introduction ...... 27 Materials and Methods ...... 31 Results ...... 37 Discussion ...... 42 Acknowledgements ...... 48 References ...... 49 Figures...... 55

CHAPTER 2: Calcium uptake in aquatic insects: Influences of phylogeny and metals (Cd and Zn) ...... 61 Abstract ...... 62 Introduction ...... 63 Materials and Methods ...... 65 Results ...... 72 Discussion ...... 76 Acknowledgements ...... 81 References ...... 82 Figures...... 87

CHAPTER 3: Phylogeny and size differentially influence dissolved Cd and Zn bioaccumulation parameters among closely related aquatic insects ...... 93 Abstract ...... 94 Introduction ...... 95 Materials and Methods ...... 97

Results ...... 103 Discussion ...... 107 Acknowledgements ...... 111 References ...... 112 Figures...... 117 Tables ...... 122 Supplemental Information ...... 125

CHAPTER 4: Four reasons why traditional metal toxicity testing with aquatic insects is irrelevant ...... 126 Text ...... 127 References ...... 131 Figure ...... 132

CHAPTER 5: The importance of retaining a phylogenetic perspective in traits-based analyses ...... 133 Abstract ...... 134 Introduction ...... 135 Materials and Methods ...... 139 Results ...... 141 Discussion ...... 143 Acknowledgements ...... 148 References ...... 149 Tables ...... 154 Figures...... 157

CHAPTER 6: Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms ...... 159 Abstract ...... 160 Introduction ...... 161 Materials and Methods ...... 164 Results ...... 169 Discussion ...... 174 Acknowledgements ...... 179 References ...... 180 Figures...... 184 Supplemental Information ...... 187

Summary and Conclusions ...... 199

LIST OF TABLES

CHAPTER 3: Phylogeny and size differentially influence dissolved Cd and Zn bioaccumulation parameters among closely related aquatic insects

Table 1. Cd and Zn uptake rate constants (ku) for 19 aquatic insect species ...... 122

Table 2. Cd and Zn efflux rate constants (ke) for 18 aquatic insect species ...... 123

Table 3. Cd and Zn bioconcentration factors (BCF) for 18 aquatic insect species ...124

CHAPTER 5: The importance of retaining a phylogenetic perspective in traits-based community analyses

Table 1. Traits examined for 42 species of aquatic insects ...... 154

Table 2. Simple and multiple linear regression results for the effects of taxonomic- and trait- based parameters on Cd bioaccumulation parameters ...... 156

CHAPTER 6: Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms

Table S1. Cd and Zn efflux rate constants (kes) for 47 species of aquatic insects ....189

Table S2. Cd and Zn efflux rate constants (kes) for 30 aquatic species ...... 191

Table S3. Analysis of phylogenetic signal for Cd and Zn kes, including comparisons of the ln likelihoods for each set of arbitrary branch lengths compared with that of a star phylogeny (indicating no hierarchal structure) ...... 193

vii

LIST OF FIGURES

CHAPTER 1: Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna

Figure 1. Uptake rates of Zn and Cd at concentrations spanning 15.3 nM to 15.3 µM Zn and 26.7 nM to 8.9 µM Cd ...... 55

Figure 2. Competition (or lack thereof) of Zn and Cd at two sets of concentrations ...... 56

Figure 3. Effects of increasing concentrations of Mn and Ca on the accumulation (uptake and adsorption) and uptake of Zn and Cd ...... 57

Figure 4. Effects of increased Ca concentrations on the accumulation and adsorption of Mn ...... 58

Figure 5. Ca absorption in H. sparna in response to increasing concentrations of ruthenium red ...... 59

Figure 6. Zn and Cd uptake in H. sparna in response to increasing concentrations of ruthenium red ...... 60

CHAPTER 2: Calcium uptake in aquatic insects: Influences of phylogeny and metals (Cd and Zn)

Figure 1. Time course of 45Ca uptake in two aquatic insects, E. invaria and H. alhedra ...... 87

Figure 2. Comparisons of newly acquired 45Ca in 12 species of aquatic insects, both absorbed and adsorbed ...... 88

Figure 3. Ca uptake rates of 12 aquatic insect species versus the average body weight of individuals ...... 89

Figure 4. Relationship between Zn and Cd kus for 12 aquatic insect species ...... 90

Figure 5. Relationship between Ca uptake rates and Zn or Cd kus for 12 aquatic insect species ...... 91

Figure 6. The effect of either dissolved Cd or Zn on Ca uptake across 4 species, 2 ephemerellids and 2 hydropsychids ...... 92

viii

CHAPTER 3: Phylogeny and size differentially affect dissolved Cd and Zn bioaccumulation among closely related aquatic insects

Figure 1. Relationship between Zn and Cd kus in 19 aquatic insect species and Zn and Cd kes in 18 species ...... 117

Figure 2. Relationship between Zn and Cd BCFs in 18 aquatic insect species ...... 118

Figure 3. Log-transformed Zn and Cd kus of 18 species versus the average log- transformed body weight of individuals ...... 119

Figure 5. Log-transformed Zn and Cd BCFs of 18 species versus the average log-transformed body weight of individuals ...... 120

Figure 6. Zn BCFs for 18 species plotted onto their phylogeny ...... 121

Supplemental Figure 1. Log-transformed Zn and Cd kes of 18 species versus the average log- transformed body weight of individuals ...... 125

CHAPTER 4: Four reasons why traditional metal toxicity testing with aquatic insects is irrelevant

Figure 1. Time it takes 34 aquatic insect species to reach a steady state tissue concentration of Cd ...... 132

CHAPTER 5: The importance of retaining a phylogenetic perspective in traits-based community analyses

Figure 1. Conceptual model of how trait values can vary across clades ...... 157

Figure 2. Cd ku, ke and BCF values for 42 species ...... 158

CHAPTER 6: Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms

Figure 1. Predicted and measured values of Zn kes from 6 EPT species ...... 184

Figure 2. Zn and Cd kes of 13 closely related aquatic insect species mapped onto their phylogeny ...... 185

Figure 3. All compiled Zn and Cd ke values from the literature ...... 186

GENERAL INTRODUCTION

Overview on aquatic insects

Aquatic insects are an invertebrate faunal group ubiquitous in freshwater systems worldwide, and they dominate freshwater systems in terms of both numbers and diversity.

Often accounting for between 75-90% of the invertebrate species pool,1,2 the global tally of aquatic insect species tops 100,000 species described to date.3 In North America alone, there are around 6,500 species.4 The abundance and diversity of aquatic insects make them integral to the structure and function of freshwater systems.5

The tremendous biodiversity of aquatic insects is likely traced back to their unique evolutionary history. Whereas freshwater fish, crustaceans, annelids and mollusks all have marine origins,6–8 aquatic insects are directly descended from terrestrial ancestors.9

Interestingly, all aquatic insects are not descended from the same terrestrial ancestor – there were multiple freshwater invasions of insects over evolutionary time.9,10 These different invasions have led to modern aquatic insect communities comprised of aquatic insect groups which have been aquatic for varying lengths of time and which have overcome the obstacles of freshwater life in different ways.

With each successful invasion of freshwater, insects had to overcome many substantial obstacles associated with life in freshwater. One such obstacle was the need to sequester ions from hypo-osmotic surroundings.11 Terrestrial insects perform ion regulation in the centralized locations of the gastrointestinal tract,12 Malpighian tubules13 and rectum,14 and it is assumed that the terrestrial ancestors of today’s aquatic insects also performed ionoregulatory duties solely in the gut. With each invasion to freshwater, aquatic insects

relocated a portion of this ionoregulatory burden to the apical body surface in the form of different structures such as chloride cells on the gills of mayflies,15 ventral chloride epithelia of caddisflies,15 and anal papillae of mosquito larvae.16,17

Aquatic insects in ecological assessments

Biological communities, specifically aquatic insect communities, are a fundamental focus of environmental assessments because they integrate stressors and pollutants in a way that other measures (e.g., pollutant concentrations in the water column) cannot.18,19 Aquatic insects are central to many environmental assessment programs worldwide because they are ubiquitous in freshwater systems, possess a great diversity of species,20 and are easily sampled in freshwater habitats.21 Further, their sedentary nature reflects conditions in their surrounding environments, integrating effects of stressors over a period of time22 and giving a more complete picture of a contaminants effects on freshwater communities than would a water sample at a single time point. Some species (particularly mayflies) are particularly sensitive to environmental contaminants,23–25 whereas others (e.g., chironomids), are extremely tolerant of environmental contaminants.26

Traditionally, environmental bioassessment and biomonitoring programs are based on taxonomic information, often at the family27–29 or genus19 levels of identification. Further, in ecological bioassessments of community composition, aquatic insect species are often lumped together into larger species groups in accordance with each state’s or jurisdiction’s protocols.30 For instance, a metric of bioassessment is often the number of EPT (orders

Ephemeroptera, Plecoptera, Trichoptera) taxa in a given space. Unfortunately, even species

within the same genera can vary greatly in their pollutant sensitivities. Water quality tolerance values (tolerant, facultative, or intolerant) were different for species in the same genus in almost 70% of aquatic insect genera examined which had values for more than one species.30 Thus, there is a continued debate on the appropriate level of taxonomic identification in community bioassessment and biomonitoring programs.30,31

Trace metals and their effects on aquatic insects in the lab and field

Trace metals, both essential and nonessential, naturally occur throughout the earth’s crust. Natural processes such as volcanic eruptions, weathering, and forest fires release metals from the earth, mobilizing and sending metals into the air and water.32 Because metals are mobilized through natural processes, metals occur naturally in freshwater systems.

However, the amount of metals in freshwater systems has seen an increase due to anthropogenic (e.g., mining, smelting) practices.33 High concentrations of metals are often found downstream of mines, industries and smelters because when mobilized, metals which were previously locked in the earth are released.32,34

Within freshwater systems, metals can take on many forms. Freshwater systems naturally differ in pH, dissolved organic material, and conductivity, all parameters which can alter the speciation of metal ions. The total dissolved metal concentration in freshwater includes metals as free ions, as well as metals bound to organic (e.g., dissolved organic carbon) and inorganic ligands (e.g., bicarbonate, chloride).35 Free metal ions in the water column are of toxicological interest, as free metal ions are bioavailable to freshwater organisms via dissolved routes of exposure.36 Thus, while two systems may have comparable

total metal concentrations, the biologically-active metal concentrations can be radically different. Because of this, models to predict the bioaccumulation (and toxicity) of metals in freshwater organisms typically account for the free metal ion concentrations in media rather than total metal concentrations.

Aquatic insects are relatively sensitive to metals in the field in comparison to most other freshwater taxa.24,37,38 Aquatic insects are often some of the first species to disappear when metal contamination occurs. Metal contamination shapes the structure of aquatic insect communities,24 where the loss of certain taxa correlates with higher metal exposures. These differential responses of aquatic insects to metals has led to the use of certain taxa (e.g., the

Hydropsyche (order Trichoptera) genus) as biomonitors, where the metal concentrations in the body tissues of individual Hydropsyche insects has been correlated with the disappearance of certain, more sensitive taxa (e.g., Heptageniidae (order Ephemeroptera) species).37,39

Contrary to their responses to metals in the field, aquatic insects are quite tolerant of dissolved metals in acute, water-only exposures in the laboratory.40–43 In fact, aquatic insects are often the most tolerant species in acute dissolved metal toxicity tests, with metal LC50 values 4 orders of magnitude higher than metal concentrations found in nature.40 Further, Cd

LC50s for aquatic insects in the Environmental Protection Agency’s ambient water quality criteria documents44 are orders of magnitude higher than concentrations known to adversely affect benthic macroinvertebrates in nature.45,46

In the following studies, we concentrate specifically on trace metals zinc (Zn) and cadmium (Cd). Zn and Cd are two borderline transition elements which co-occur in metal

ores and surface waters throughout the world.47 While natural, their concentrations in the environment are ever increasing due to industrial activities throughout the world (especially in the United States). Zn and Cd share similar biochemical properties in aquatic environments. They share the same divalent charge as free ions, the same number of valence electrons, similar electron configuration,48 and a similar electronegativity values.49 Further,

Zn and Cd share comparable affinities for sulfur, oxygen and nitrogen ligands.47

Modeling of trace metal bioaccumulation

Bioaccumulation is an integrative indicator of metal exposures to aquatic organisms in polluted environments because unlike other contaminants, metals are not metabolized.50

Metal bioaccumulation can be quite complex, with many factors contributing to the bioaccumulation of metals in aquatic environments. Four factors in particular introduce uncertainty into metal bioaccumulation in aquatic organisms: metal specificity, exposure route, environmental influences, and species-specific characteristics.51 These factors make it particularly difficult to model metal bioaccumulation in aquatic organisms, making it exceptionally difficult to link chemical/metal exposure in the environment with body burdens.

One of the first methods to attempt the modeling of metal bioaccumulation in aquatic organisms was the free ion activity model (FIAM).52 This model identified water chemistry as an important parameter in metal bioaccumulation from dissolved exposures. Because total metal concentrations in the environment were poor predictors of metal bioaccumulation, this

model allowed for only the free metal ion – the most biologically active form of the metal – to be accounted for after taking into account water chemistry parameters.53

A model currently in use to determine metal bioavailability, the Biotic Ligand Model

(BLM),36 is largely based on FIAM, with a heavy emphasis on understanding the free metal ion concentration present to interact with the biotic ligand (the physiologically active site).

The inherent assumption of the BLM model is that an organism experiences toxicity from dissolved metals as a result of metal binding to the biotic ligand on the surface of gills.54,55 In addition to the complexation of the metal to the biotic ligand, the model incorporates the complexation of the metal ion to other environmental ligands (e.g., dissolved organic matter), the concentration of the free metal ion in solution, and the competition of the free metal ion with other cations capable of binding the biotic ligand.

One of the main benefits of the BLM is the ability to incorporate site-specific water chemistry values into the analysis.55 Thus, it allows the determination of the toxicity of a metal to an organism in a specific freshwater system. The BLM has been used to successfully predict the toxicity of different metals in a variety of organisms – namely fish and Daphnia spp.56,57 In fact, the BLM is so successful that the Environmental Protection Agency now uses it to derive water quality criteria for select metals (e.g., copper58).

While the BLM has many advantages, there are also drawbacks to using the BLM for modeling metal bioaccumulation.59 The BLM relies heavily on the assumption that properties of the gill/metal interaction are fixed, however these (and the toxicity values associated with them) have been shown to be dynamic.54 One of the main criticisms of the BLM is the disregard of factors other than water chemistry and dissolved metal uptake at the biotic

ligand on the body surface. As mentioned previously, aquatic organisms can take up metals via their surrounding media as well as diet,60 and the BLM ignores any metal accumulation or toxicity which may occur due to the dietary intake of metals.

An alternative metal bioaccumulation model, the dynamic multi-pathway bioaccumulation model (DYN-BAM) is a biologically-based conceptual model first described by Luoma and Rainbow.51 This model describes species-specific physiological characteristics of organisms which drive metal bioaccumulation.61 It has become a powerful tool to predict trace metal bioaccumulation among species. This model works under the assumption that metal bioaccumulation is a result of three physiological processes: uptake of dissolved metals, uptake of dietary metals, and elimination of metals. Further, this model further allows for site-specific concentrations and conditions to be added.51 The model is as follows:

, where

-1 = the steady-state metal concentration in the organisms (µg g ), = uptake rate

-1 -1 -1 constant from dissolved metals (L g d ), = concentration in water (µg L ), =

-1 -1 assimilation efficiency (%), = ingestion rate (g g d ), = metal concentration in food

-1 -1 -1 (µg g ), = proportional daily loss (d ), = growth rate (d ).

The ku parameter indicates how fast an organism will accumulate dissolved free metal ions. Together with metal concentrations in the solution, this determines how much metal an organism will accumulate from its surrounding media. Because this value involves measuring

uptake from solution, water chemistry can affect values acquired. Metal uptake from food is described by assimilation efficiency, ingestion rate and the concentration of metal in food.

Assimilation efficiency indicates the fraction of ingested metal which is retained in biological

62 tissue, and ingestion rate indicates the speed at which an organism eats. Lastly, the ke is an important driver of metal bioaccumulation among co-occurring species.51 The sizeable differences in ke values across species can be striking, with Cd ke values in aquatic insects

-1 37,51 ranging orders of magnitude, from 0.003 – 0.21 d . Further, ke is considered a key determinant of the metal loads that a species will need to detoxify or store in order to carry on in a metal-contaminated environment.

Areas of concern in regulatory toxicology

Acute, water-only toxicity tests with dissolved metals are a component of the methods used to develop regulations for metal concentrations in the environment. Currently, regulatory agencies work under the assumption that dissolved metals drive acute toxicity in aquatic organisms.61 Therefore, these acute toxicity tests rely on the assumption that all metals are toxic to freshwater species via aqueous exposure. A physiological mechanism for acute metal (Cd, Zn) toxicity is hypocalcemia and has been described as such in organisms such as fish63 and crustaceans.64 However despite their dominance in freshwater ecosystems, the trace metal physiology of aquatic insects in relation to dissolved metal exposures remains poorly understood.

Dissolved Zn and Cd are hypothesized to traverse Ca transporters in freshwater taxa.

The similar biochemical properties shared by Zn and Cd are also shared by the

macronutrient, calcium (Ca). Ions of the same charge and biochemical characteristics are known to utilize (compete for) the same transporters65,66 given the lack of specificity of transport systems.67 Studies with fish,68–71 crustaceans,72,73 and mollusks74 have all given evidence for the shared transport systems of Ca and trace metals Cd and Zn. It is assumed that this is also the case for aquatic insects.

In regulatory toxicology, there is a widespread reliance on toxicity values derived from surrogate laboratory species and their application to other taxa. For example, the common use of rainbow trout as a surrogate for salmonids (including endangered Pacific salmon species) occurs without a robust understanding of whether the rainbow trout’s sensitivity to a given pesticide is comparable to its relatives. Similarly, Daphnia magna is often the “go to” surrogate species meant to represent planktonic crustaceans – a hugely diverse group. Unfortunately, we do not know how well traditional laboratory organisms represent species within their species groups in toxicity tests, and we have a limited understanding of the toxicant sensitivity variability which exists in species groups (but see refs. 97-100).

Interspecies extrapolations are especially relied upon in the derivation of water quality criteria (WQC) in the United States. WQC provide concentrations of contaminants that should not be exceeded for set frequencies or durations in order to avoid both acute and chronic toxicity to aquatic life.75 The determination of these criteria involves compiling all available toxicity data, screening those data to ensure they meet data quality objectives, and synthesizing results in order to build species’ sensitivity distributions. These distributions then indicate the acute and chronic maximum concentration limit which will protect 95% of

tested species; the extent to which tested species are a robust and random sample of aquatic life in general then determines the extent to which the derived criteria is protective of other, untested species. Only 8 LC50s are required for the development of acute WQC, with each

LC50 calculated for a species from one of 8 specified groups. The inherent assumption here is that the LC50 of a representative species adequately represents the toxicity of the compound for an entire species group, and that this value will lead to WQC which will be protective of this group. However, we have no idea the variance which occurs across aquatic insect species in particular, and thus have little idea of how well surrogate species represent their larger groups.

Comparative phylogenetic methods

Because of the heavy reliance on aquatic insect species to monitor metal contaminated freshwater systems, it is important to understand not only the metal physiologies of these insects, but also how their physiologies vary across species. A tool to further quantify the variance of physiological values across species is phylogenetic statistics.

Whereas conventional statistics treat species as independent entities,76–78 phylogenetic statistics allow us to take hierarchal relationships based on evolutionary relatedness into account when analyzing comparative data.79 Some degree of relatedness exists between all species whether it be large or small, and close relatives tend to resemble each other both physically and genetically more so that distant relatives. It is now accepted by comparative biologists that multispecies data sets should be analyzed with phylogenetic statistics in lieu of conventional methods.80

The tendency for closely related species to be more similar is termed “phylogenetic signal” and can be is quantified by the K-statistic.81 The main function of the K-statistic is to ask whether a given phylogenetic tree (branch lengths and topology) fits the data (in this case, toxicity values), better than if the data is randomly permuted across the tips of the tree.81 The K-statistic quantifies phylogenetic signal across traits (e.g., toxicity values) using statistical programs which analyze the data while taking into account species’ positions on a phylogenetic tree. Specifically, the K-statistic indicates the amount of phylogenetic signal present in a comparative (multi-species) dataset relative to what should be expected for a trait given that the trait evolved by Brownian motion along a specific tree topology. Generally values for phylogenetic signal range from 0-1, with “0” indicating no phylogenetic signal and values closer to “1” indicating that there is a strong phylogenetic signal (e.g., what would be expected under Brownian motion along a specified tree topology).

In large datasets, the presence of phylogenetic signal is common. For example, in their analysis of 53 traits (morphological, behavioral, life history, physiological) where comparative datasets included more than 20 species, Garland et al. demonstrated that more than 92% of these datasets exhibit significant phylogenetic signal.79 Yet, different types of traits have more concordance with phylogeny than others. For example, behavioral traits, which are considered to be more evolutionarily malleable and adaptive, exhibited phylogenetic signal, albeit a weaker signal than other physiological, morphological, or life history traits.81 Physiological traits, the type of trait which this dissertation focused on primarily, exhibit a range of k-statistics – with most k-statistics falling between 0.25 and 1.79

Therefore, phylogenetic signal can range from relatively low (~0.25) to high (>1).

While phylogenetic statistics have been in use since the early 1980s, these methods are just beginning to gain traction in toxicology (see refs 82–85). To date, phylogenetic methods have been used to examine the influences of evolutionary history on pesticide sensitivity in amphibians,83 predict species tolerances to pesticides in aquatic species,85 and have been proposed to help develop methods for biomonitoring based on the strength or weakness of phylogentic signal.84 Phylogenetic signal in physiological traits across closely related species (e.g., within the same family or genus) has yet to be explored.

Previously, our laboratory quantified phylogenetic signal in Cd uptake and efflux rate constants in 21 species spanning orders Ephemeroptera (mayflies), Plecoptera (stoneflies),

82 and Trichoptera (caddisflies). Across orders, both ku values and ke values exhibited significant moderate phylogenetic signal. However, some families (notably Ephemerellidae and Hydropsychidae) exhibited a large variation within the small sample size of species examined (both at n = 2). This observation led to one of the large, overarching questions explored by this dissertation – How different are closely related species (within the same family or genus), and are these differences driven by evolutionary histories?

Research objectives and approach

Acute laboratory toxicity tests suggest that aquatic insects are tolerant of dissolved metals in the laboratory, however aquatic insects are often some of the first species to disappear in metal contaminated sites.24 Further, the regulatory community is only required to use a single aquatic insect to derive metal regulatory values without knowing how variable the physiologies of these evolutionarily diverse species affect their metal toxicity. Here, our

objective was to gain an understanding of not only dissolved metal (Ca, Cd, Mn, Zn) bioaccumulation in aquatic insect species, but also of the variability of metal bioaccumulation traits across aquatic insects.

For this work, species from two ecologically-important and species rich families were used: Ephemerellidae (Order: Ephemeroptera (mayflies)) and Hydropsychidae (Order:

Trichoptera (caddisflies)). Previously, these families were shown to be especially variable in a larger scale analyses of metal bioaccumulation in aquatic insects.82 In North America,

Ephemerellidae has an estimated 75 species,86 and Hydropsychidae has an estimated 184 species.87 All species for this project were collected from Great Smoky Mountains National

Park in North Carolina and Tennessee.

We focused specifically on metals Cd and Zn for reasons mentioned above. They co- occur in metal ores and are found in freshwater systems together. Further, they are expected to induce toxic effects via uptake through Ca transport systems. Manganese (Mn) was also analyzed for two reasons – the Mn (II) form has the ability to compete for uptake with Ca,

Cd and Zn,88–90 and in the oxide form, Mn can be a sink for other metals91 on the integument of the insect.88

All metal bioaccumulation work in this project utilized radiotracers. Gamma-emitting radioisotopes Zn65, Cd109 and Mn54 allowed for the observation of metal uptake and depuration over time in the same individual insects. Further, we were able to count two

(109Cd and 65Zn) and three (109Cd, 54Mn and 65Zn) isotopes simultaneously using spillover corrections verified across single- and multi- isotope standards. Beta-emitting isotope Ca45 allowed for the measurement of Ca uptake in aquatic insects.

In our first study, we used one hydropsychid species, Hydropsyche sparna, to study the uptake and interactions of Zn, Cd Ca and Mn (a Ca analog88,92). We analyzed the aqueous accumulation kinetics of Cd and Zn across 4 orders of magnitude. We quantified interactions between all metals to determine if competition happened between metals. Different rinsing approaches were used to determine differences between the total accumulated metal

(absorption + adsorption) and total internalized metal (absorption only). Lastly, different pharmacological blockers were used in order to attempt to identify potential uptake pathways for Ca, Cd and Zn.

We further studied Ca uptake across 12 species within families Hydropsychidae

(order Trichoptera) and Ephemerellidae (order Ephemeroptera). Only two studies previously analyzed Ca uptake in aquatic insects (see refs 93, 94), and the field remained relatively untouched for aquatic insects. Ca uptake and adsorption was characterized across 12 species, and the effects of Cd and Zn on Ca uptake were quantified in four species. Uptake rates (Ca) and uptake rate constants (Cd, Zn) were assessed for evidence of shared transport pathways.

We quantified Cd and Zn uptake and efflux rate constants in 19 species within families Hydropsychidae and Ephemerellidae in order to understand differences in metal bioaccumulation across these variable families at fine levels of taxonomic resolution. We calculated bioconcentration factors in a concentration- and time- independent manner as another endpoint to assess. We quantified the co-variation of Zn and Cd bioaccumulation traits, and further analyzed the effects of body weight and phylogeny on these traits in the hopes of elucidating patterns for predictions.

From our metal bioaccumulation studies, we applied our results to explain the large disconnect between laboratory and field studies of metal toxicity in aquatic insects. We analyzed our metal uptake and interactions results along with past studies with aquatic insects in order to define four key reasons why traditional dissolved acute metal toxicity tests fail to capture the sensitivity of aquatic insects to metals.

In our next study, we took all of the dissolved Cd bioaccumulation data generated in from the above studies and combined it with all other available uptake and efflux rate constants in aquatic insects.82,95 We calculated bioconcentration factors for all aquatic insect species examined. The effects of clade (as genus, family, and order), as well as the effects of body weight and feeding guild membership96 were quantified and models were compared to determine the predictor with the most influence over Cd bioaccumulation traits. We also determined the presence and strength of phylogenetic signal across the physiological traits examined.

Lastly, after obtaining Zn and Cd efflux rate constants for Ephemerellidae and

Hydropsychidae species in the laboratory, we then combined this data with all Zn and Cd ke data from the literature. We compared values across four aquatic phyla. We then used phylogenetic statistics to find evolutionary patterns across aquatic insects as well as all aquatic organisms.

References

(1) Arscott, D. B.; Jackson, J. K.; Kratzer, E. B. Role of rarity and taxonomic resolution in a regional and spatial analysis of stream macroinvertebrates. J. North Am. Benthol. Soc. 2006, 25, 977–997.

(2) Minshall, G. W. Community structure in natural stream systems. In The Stream Ecosystem; 1969.

(3) Dijkstra, K.-D. B.; Monaghan, M. T.; Pauls, S. U. Freshwater biodiversity and aquatic insect diversification. Annu. Rev. Entomol. 2014, 59, 143–163.

(4) Merritt, R. W.; Cummins, K. W.; Berg, M. B. An Introduction to the Aquatic Insects of North America; Merritt, R. W.; Cummins, K. W.; Berg, M. B., Eds.; Kendall-Hunt: Dubuque, IA, 2008.

(5) Wallace, J. B.; Webster, J. R. The role of macroinvertebrates in stream ecosystem function. Annu. Rev. Entomol. 1996, 41, 115–139.

(6) Anger, K. The conquest of freshwater and land by marine crabs: Adaptations in life- history patterns and larval bioenergetics. J. Exp. Mar. Biol. Ecol. 1995, 193, 119–145.

(7) Lee, C. E.; Bell, M. A. Causes and consequences of recent freshwater invasions by saltwater animals. Trends Ecol. Evol. 1999, 14, 284–288.

(8) Lee, C. E. Rapid and repeated invasions of fresh water by the copepod Eurytemora affinis. Evolution (N. Y). 1999, 53, 1423–1434.

(9) Kristensen, N. P. Phylogeny of insect orders. Annu. Rev. Entomol. 1981, 26, 135–157.

(10) Bradley, T. J.; Briscoe, A. D.; Brady, S. G.; Contreras, H. L.; Danforth, B. N.; Dudley, R.; Grimaldi, D.; Harrison, J. F.; Kaiser, J. A.; Merlin, C.; et al. Episodes in insect evolution. Integ. Comp. Biol. 2009, 49, 590–606.

(11) Taylor, C. W. Calcium regulation in insects. Adv. In Insect Phys. 1986, 19, 155–186.

(12) Dow, J. A. T. Insect midgut function. Adv. Insect Physiol. 1986, 19, 187–328.

(13) Maddrell, S. H. P. Characteristics of epithelial transport in insect Malpighian tubules. Curr. Top. Membr. Trans. 1980, 14, 427–463.

(14) Stobbart, R. H.; Shaw, J. Salt and water balance: Excretion. In The Physiology of Insects; Rothstein, M., Ed.; Academic Press: New York, 1974.

(15) Komnick, H. Chloride cells and chloride epithelia of aquatic insects. Int. Rev. Cytol. 1977, 49, 285–329.

(16) Komnick, H.; Schmidtz, M. Cutaneous chloride uptake from hypo-osmitoc concentration by chloride cells of Corixa punctata. J. Insect Physiol. 1977, 23, 165– 173.

(17) Donini, A.; O’Donnell, M. J. Analysis of Na+, Cl-, K+, H+ and NH4+ concentration gradients adjacent to the surface of anal papillae of the mosquito Aedes aegypti: application of self-referencing ion-selective microelectrodes. J. Exp. Biol. 2005, 208, 603–610.

(18) Rosenberg, D. M.; Resh, V. H. Introduction to freshwater biomonitoring and benthic macroinvertebrates. In Freshwater Biomonitoring and Benthic Macroinvertebrates; Rosenberg, D. M.; Resh, V. H., Eds.; Chapman & Hall: New York, 1993.

(19) Barbour, M. T.; Yoder, C. O. The multimetric approach to bioassessment, as used in the United States of America. In Assessing the Bioogical Quality of Freshwater RIVPACs and Other Techniques; Wright, J. F.; Sutcliffe, D. W.; Furse, M. T., Eds.; Freshwater Biological Association, 2000; pp. 281–292.

(20) Resh, V. H.; Norris, R. H.; Barbur, M. T. Design and implementation of rapid assessment approaches for water resource monitoring using benthic macroinvertebrates. Aust. J. Ecol. 1995, 20, 108–121.

(21) Rosenberg, D. M.; Danks, H. V.; Lehmkuhl, D. M. Importance of insects in environmental impact assessment. Environ. Manag. 1986, 10, 773–783.

(22) Bonada, N.; Prat, N.; Resh, V. H.; Statzner, B. Developments in aquatic insect biomonitoring: a comparative analysis of recent approaches. Annu. Rev. Entomol. 2006, 51, 495–523.

(23) Clements, W. H.; Cadmus, P.; Brinkman, S. F. Responses of aquatic insects to Cu and Zn in stream microcosms: Understanding differences between single species tests and field responses. Env. Sci. Technol. 2013, 47, 7506–7513.

(24) Clements, W. H.; Carlisle, D. M.; Lazorchak, J. M.; Johnson, P. C. Heavy metals structure benthic communities in Colorado mountain streams. Ecol. Appl. 2000, 10, 626–638.

(25) Schmidt, T. S.; Clements, W. H.; Zuellig, R. E.; Mitchell, K. A; Church, S. E.; Wanty, R. B.; San Juan, C. A.; Adams, M.; Lamothe, P. J. Critical tissue residue approach

linking accumulated metals in aquatic insects to population and community-level effects. Env. Sci. Technol. 2011, 45, 7004–7010.

(26) Winner, R.; Boesel, M.; Farrell, M. Insect community structure as an index of heavy- metal pollution in lotic ecosystems. Can. J. Fish. Aquat. Sci. 1980, 37, 647-655.

(27) Hilsenhoff, W. L. Rapid field assessment of organic pollution with a family-level biotic index. J. N. Am. Benthol. Soc. 1988, 7, 65–68.

(28) Chessman, B. C. Rapid assessment of rivers using macroinvertebrates: A procedure based on habitat-specific sampling, family level identification and a biotic index. Aust. J. Ecol. 1995, 20, 122–129.

(29) Simpson, J. C.; Norris, R. H. Bioassessment of river quality: development of AUSRIVAS model and outputs. In Assessing the Bioogical Quality of Freshwater RIVPACs and Other Techniques; Wright, J. F.; Sutcliff, D. W.; Furse, M. T., Eds.; Freshwater Biological Association: Ambleside, UK, 2000; pp. 125–142.

(30) Resh, V. H.; Unzicker, J. D. Water quality monitoring and aquatic organisms: The importance of species identification. Water Environ. Fed. 1975, 47, 9–19.

(31) Waite, I. R.; Herlihy, A. T.; Larsen, D. P.; Urquhart, N. S.; Lemm, D. J. The effects of macroinvertebrate taxonomic resolution in large landscape bioassessments: an example from Mid-Atlantic Highlands, U.S.A. Freshw. Biol. 2004, 49, 474–489.

(32) Callender, E. Heavy metals in the environment- Historical trends. In Treatise on Geochemistry; Sherwood Lollar, B., Ed.; Elsevier, 2003; Vol. 5, pp. 67–103.

(33) Adriano, D. C. Trace elements in the Terrestrial Environment; Springer: NY, 1986.

(34) Nriagu, J. O. Global metal pollution: poisoning the biosphere? Environ. Sci. Policy Sustain. Dev. 1990, 32, 7–33.

(35) Campbell, P. G. C. Interactions between trace metals and aquatic organisms: A critique of the free-ion activity model. In Metal Speciation and Bioavailability in Aquatic Organisms; Tessier, A.; Turner, D., Eds.; John Wiley: New York, 1995; pp. 45–102.

(36) Di Toro, D. M.; Allen, H. E.; Bergman, H. L.; Meyer, J. S.; Paquin, P. R.; Santore, R. C. Biotic ligand model of the acute toxicity of metals. 1. Technical basis. Env. Toxicol. Chem. 2001, 20, 2383–2396.

(37) Cain, D. J.; Luoma, S. N.; Wallace, W. G. Linking metal bioaccumulation of aquatic insects to their distribution patterns in a mining-impacted river. Environ. Toxicol. Chem. 2004, 23, 1463–1473.

(38) Rainbow, P. S. Trace metal concentrations in aquatic invertebrates: Why and so what? Env. Pollut. 2002, 120, 497–507.

(39) Luoma, S. N.; Cain, D. J.; Rainbow, P. S. Calibrating biomonitors to ecological disturbance: A new technique for explaining metal effects in natural waters. Integ. Env. Assess. Manag. 2010, 6, 199–209.

(40) Brix, K. V.; DeForest, D. K.; Burger, M.; Adams, W. J. Assessing the relative sensitivity of aquatic organisms to divalent metals and their representation in toxicity datasets compared to natural aquatic communities. Hum. Ecol. Risk Assess. An Int. J. 2005, 11, 1139–1156.

(41) Brix, K. V; DeForest, D. K.; Adams, W. J. The sensitivity of aquatic insects to divalent metals: A comparative analysis of laboratory and field data. Sci. Total Env. 2011, 409, 4187–4197.

(42) Buchwalter, D. B.; Cain, D. J.; Clements, W. H.; Luoma, S. N. Using biodynamic models to reconcile differences between laboratory toxicity tests and field biomonitoring with aquatic insects. Env. Sci. Technol. 2007, 41, 4821–4828.

(43) Brix, K. V.; DeForest, D. K.; Adams, W. J. Assessing acute and chronic copper risks to freshwater aquatic life using species sensitivity distributions for different taxonomic groups. Environ. Toxicol. Chem. 2001, 20, 1846–1856.

(44) United States Environmental Protection Agency. 2001. Update of Ambient Water Quality Criteria for Cadmium. Accessed from: http://water.epa.gov/scitech/swguidance/standards/upload/2001_04_13_criteria_cadmi um_cad2001upd.pdf

(45) Buchwalter, D. B.; Luoma, S. N. Differences in dissolved cadmium and zinc uptake among stream insects: Mechanistic explanations. Env. Sci. Technol. 2005, 39, 498– 504.

(46) Clements, W. H. Small-scale experiments support causal relationships between metal contamination and macroinvertebrate community responses. Ecol. Appl. 2004, 14, 954–967.

(47) Nieboer, E.; Richardson, D. H. S. The replacement of the non-descript term “heavy metals” by a biologically and chemically significant classification of metal ions. Env. Pollut. B 1980, 1, 3–26.

(48) Newman, M. C.; McCloskey, J. T. Predicting relative toxicity and interactions of divalent metal ions: Microtox(R) bioluminescence assay. Env. Toxicol. Chem. 1996, 15, 275–281.

(49) Allred, A. L. Electronegativity values from thermochemical data. J. Inorg. Nucl. Chem. 1961, 17, 215–221.

(50) Phillips, D. J. H.; Rainbow, P. S. Biomonitoring of Trace Aquatic Contaminants; Alden Press Ltd: Oxford, 1994.

(51) Luoma, S. N.; Rainbow, P. S. Why is metal bioaccumulation so variable? Biodynamics as a unifying concept. Env. Sci. Technol. 2005, 39, 1921–1931.

(52) Morel, F. M. Principles of Aquatic Chemistry; John Wiley & Sons Ltd: Toronto, ON, Canada, 1983.

(53) Campbell, P. G. C. Interactions between trace metals and aquatic organisms: A critique of the free-ion activity model. In Metal Speciation and Bioavailability in Aquatic Organisms; Tessier, A.; Turner, D., Eds.; John Wiley: New York, 1995; pp. 45–102.

(54) Niyogi, S.; Wood, C. M. Effects of chronic waterborne and dietary metal exposures on gill metal-binding: Implications for the biotic ligand model. Hum. Ecol. Risk Assess. 2003, 9, 813–846.

(55) Niyogi, S.; Wood, C. M. Biotic ligand model, a flexible tool for developing site- specific water quality guidelines for metals. Env. Sci. Technol. 2004, 38, 6177–6192.

(56) De Schamphelaere, K. A. C.; Janssen, C. R. A biotic ligand model predicting acute copper toxicity for Daphnia magna: The effects of calcium, magnesium, sodium, potassium, and pH. Env. Sci. Technol. 2002, 36, 48–54.

(57) Heijerick, D. G.; De Schamphelaere, K. A. C.; Janssen, C. R. Predicting acute zinc toxicity for Daphnia magna as a function of key water chemistry characteristics: Development and validation of a biotic ligand model. Env. Toxicol. Chem. 2002, 21, 1309–1315.

(58) United States Environmental Protection Agency. 2007. Aquatic life ambient freshwater quality criteria- copper. Accessed from: http://www.itrcweb.org/contseds- bioavailability/References/2009_04_27_criteria_copper_2007_criteria-full.pdf

(59) Erickson, R. J. The biotic ligand model approach for addressing effects of exposure water chemistry on aquatic toxicity of metals: Genesis and challenges. Env. Toxicol. Chem. 2013, 32, 1212–1214.

(60) Hare, L. Aquatic insects and trace metals: Bioavailability, bioaccumulation, and toxicity. Crit. Rev. Toxicol. 1992, 22, 327–369.

(61) Schlekat, C. E.; Lee, B.-G.; Luoma, S. N. Dietary Metals Exposure and Toxicity to Aquatic Organisms: Implications for Ecological Risk Assessment. In Coastal and Estuarine Risk Assessment; CRC Press, LLC, 2002.

(62) Wang, W.-X.; Fisher, N. S. Assimilation efficiencies of chemical contaminants in aquatic invertebrates: A synthesis. Env. Toxicol. Chem. 1999, 18, 2034–2045.

(63) Spry, D. J.; Wood, C. M. Ion flux rates, acid-base status, and blood gases in rainbow trout, Salmo gairdneri, exposed to toxic zinc in natural soft water. Can. J. Fish. Aquat. Sci. 1985, 42, 1332–1341.

(64) Felten, V.; Charmantier, G.; Mons, R.; Geffard, A.; Rousselle, P.; Coquery, M.; Garric, J.; Geffard, O. Physiological and behavioural responses of Gammarus pulex (Crustacea: Amphipoda) exposed to cadmium. Aquat. Toxicol. 2008, 86, 413–425.

(65) Hinkle, P. M.; Kinsella, P. A.; Osterhoudt, K. C. Cadmium uptake and toxicity via voltage-sensitive calcium channels. J. Biol. Chem. 1987, 262, 16333–16337.

(66) Simkiss, K.; Taylor, M. G. Transport of Metals Across Membranes. In Metal Speciation and Bioavailability in Aquatic Systems; Tessier, A.; Turner, D. R., Eds.; John Wiley & Sons Ltd: West Sussex, England, 1995; pp. 1–45.

(67) Williams, R. J. P.; da Silva, J. J. R. F. The distribution of elements in cells. Coord. Chem. Rev. 2000, 200, 247–348.

(68) Hogstrand, C.; Wilson, R. W.; Polgar, D.; Wood, C. M. Effects of zinc on the kinetics of branchial calcium uptake in freshwater rainbow trout during adaptation to waterborne zinc. J. Exp. Biol. 1994, 186, 55–73.

(69) McCormick, S. D.; Hasegawa, S.; Hirano, T. Calcium uptake in the skin of a freshwater teleost. P. Natl. Acad .Sci. USA 1992, 89, 3635–3638.

(70) Chang, M.-H.; Lin, H.-C.; Hwang, P. P. Effects of cadmium on the kinetics of calcium uptake in developing tilapia larvae, Oreochromis mossambicus. Fish Physiol. Biochem. 1997, 16, 459–470.

(71) Wicklund, A.; Runn, P. Calcium effects on cadmium uptake, redistribution, and elimination in minnows, Phoxinus phoxinus, acclimated to different calcium concentrations. Aquat. Toxicol. 1988, 13, 109–121.

(72) Tan, Q. G.; Wang, W. X. Acute toxicity of cadmium in Daphnia magna under different calcium and pH Conditions: Importance of influx rate. Env. Sci. Technol. 2011, 45, 1970–1976.

(73) Wright, D. A. Effect of salinity on cadmium uptake by tissues of shore crab Carcinus maenas. J. Exp. Biol. 1977, 67, 137–146.

(74) Bjerregaard, P.; Depledge, M. H. Cadmium accumulation in Littorina littorea, Mytilus edulis and Carcinus maenas: The influence of salinity and calcium ion concentrations. Mar. Biol. 1994, 119, 385–395.

(75) Stephan, C. E.; Mount, D. I.; Hansen, D. J.; Gentile, J. R.; Chapman, G. A.; Brungs, W. A. Guidelines for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. United States Environmental Protection Agency, 1985.

(76) Felsenstein, J. Phylogenies and the comparative method. Am. Nat. 1985, 125, 1–15.

(77) Grafen, A. The phylogenetic regression. Phil. Trans. R. Soc. Lond. B. 1989, 326, 119– 156.

(78) Pagel, M. D.; Harvey, P. H. Comparative methods for examining adaptation depend on evolutionary models. Folia Primatol. 1989, 53, 203–220.

(79) Garland, T.; Bennett, A. F.; Rezende, E. L. Phylogenetic approaches in comparative physiology. J. Exp. Biol. 2005, 208, 3015–3035.

(80) Blomberg, S. P.; Garland, Jr., T.; Ives, A. R. Tempo and mode in evolution: Phylogenetic inertia, adaptation and comparative methods. J. Evol. Biol. 2002, 15, 899–910.

(81) Blomberg, S. P.; Garland, Jr., T.; Ives, A. R. Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution. 2003, 57, 717–745.

(82) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.; Garland, Jr., T. Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. P. Natl. Acad. Sci. USA 2008, 105, 8321–8326.

(83) Hammond, J. I.; Jones, D. K.; Stephens, P. R.; Relyea, R. A. Phylogeny meets ecotoxicology: Evolutionary patterns of sensitivity to a common insecticide. Evol. Appl. 2012, 5, 593–606.

(84) Carew, M. E.; Miller, A. D.; Hoffmann, A. A. Phylogenetic signals and ecotoxicological responses: Potential implications for aquatic biomonitoring. Ecotoxicology 2011, 20, 595–606.

(85) Guenard, G.; von der Ohe, P. C.; de Zwart, D.; Legendre, P.; Lek, S. Using phylogenetic information to predict species tolerances to toxic chemicals. Ecol. Appl. 2011, 21, 3178–3190.

(86) Mayfly Central. Website, http://www.entm.purdue.edu/mayfly/.

(87) Morse, J. C. A checklist of the Trichoptera of North America , including Greenland and Mexico. Trans. Am. Entomol. Soc. 1993, 119, 47–93.

(88) Dittman, E. K.; Buchwalter, D. B. Manganese bioconcentration in aquatic insects: Mn oxide coatings, molting loss, and Mn (II) thiol scavenging. Env. Sci. Technol. 2010, 44, 9182–9188.

(89) Lasier, P. J.; Winger, P. V.; Bogenrieder, K. J. Toxicity of manganese to Ceriodaphnia dubia and Hyalella azteca. Arch. Environ. Contam. Toxicol. 2000, 38, 298–304.

(90) Stubblefield, W. A.; Brinkman, S. E.; Davies, P. H.; Garrison, T. D. Effects of water hardness on the toxicity of manganese to developing brown trout (Salmo trutta). Env. Toxicol. Chem. 1997, 16, 2082–2089.

(91) Trivedi, P.; Axe, L. Modeling Cd and Zn sorption to hydrous metal oxides. Env. Sci. Technol. 2000, 34, 2215–2223.

(92) Markich, S. J.; Jeffree, R. A. Absorption of divalent trace metals as analogues of calcium by Australian freshwater bivalves: An explanation of how water hardness reduces metal toxicity. Aquat. Toxicol. 1994, 29, 257–290.

(93) Barkai, A. I.; Williams, R. W. The exchange of calcium in larvae of the mosquito Aedes aegypti. J. Exp. Biol. 1983, 104, 139–148.

(94) Gillis, P. L.; Wood, C. M. The effect of extreme waterborne cadmium exposure on the internal concentrations of cadmium, calcium, and sodium in Chironomus riparius larvae. Ecotoxicol. Env. Saf. 2008, 71, 56–64.

(95) Cain, D.; Croteau, M.-N.; Luoma, S. N. Bioaccumulation dynamics and exposure routes of Cd and Cu among species of aquatic mayflies. Env. Toxicol. Chem. 2011, 30, 2532–2541.

(96) Poff, N. L.; Olden, J. D.; Vieira, N. K. M.; Finn, D. S.; Simmons, M. P.; Kondratieff, B. C. Functional trait niches of North American lotic insects: Traits-based ecological applications in light of phylogenetic relationships. J. N. Am. Benthol. Soc. 2006, 25, 730–755.

(97) Mayer, F. L.; Ellersieck, M. R. Manual of acute toxicity: Interpretations of database for 410 chemicals and 66 species of freshwater animals. 1986, Resource Publication 160. U.S. Fish and Wildlife Service, Washington, D.C.

(98) Raimondo, S.; Jackson, C. R.; Barron, M. G. Influence of taxonomic relatedness and mode of action in acute interspecies estimation models for aquatic species. Environ. Sci. Technol. 2010, 44, 7711-7716.

(99) Dyer, S. D.; Versteeg, D. J.; Belanger, S. E.; Chaney, J. G.; Raimondo, S.; Barron, M. G. Comparison of species sensitivity distributions derived from interspecies correlation models to distributions used to derive water quality criteria. Environ. Sci. Technol. 2008, 42, 3076-3083.

(100) Raimondo, S.; Mineau, P.; Barron, M. G. Estimation of chemical toxicity to wildlife species using interspecies correlation models. Environ. Sci. Technol. 2007, 41, 5888- 5894.

CHAPTER 1

Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna

Monica D. Poteat1, Mauricio Díaz-Jaramillo2, and David B. Buchwalter1

1Department of Environmental and Molecular Toxicology, North Carolina State University, Raleigh, NC 27695, United States

2Aquatic Research Unit, EULA-Chile Environmental Center, Universidad de Concepción, Concepción, Chile

Published In: Journal of Experimental Biology. 2012; Volume 215; pages 1575-1583.

Abstract

Despite their ecological importance and prevalent use as ecological indicators, the trace element physiology of aquatic insects remains poorly studied. Understanding divalent metal transport processes at the water-insect interface is important because these metals may be essential (e.g., Ca), essential and potentially toxic (e.g., Zn) or non-essential and toxic

(e.g., Cd). We measured accumulation kinetics of Zn and Cd across dissolved concentrations ranging 4 orders of magnitude and examined interactions with Ca and Mn in the caddisfly

Hydropsyche sparna. Here we provide evidence for at least two transport systems for both

Zn and Cd, the first of which operating at concentrations below 0.8 µM (and is fully saturable for Zn). We observed no signs of saturation of a second lower-affinity transport system at concentrations up to 8.9 μM Cd and 15.3 μM Zn. In competition studies at 0.6 μM

Zn and Cd, the presence of Cd slowed Zn accumulation by 35% while Cd was unaffected by

Zn. At extreme concentrations (listed above), Cd accumulation was unaffected by the presence of Zn whereas Zn accumulation rates were reduced by 58%. Increasing Ca from

31.1 µM to 1.35 mM resulted in only modest decreases in Cd and Zn uptake. Manganese decreased adsorption of Cd and Zn to the integument but not internalization. L-type Ca channel blockers verapamil and nifedipine and plasma membrane Ca-ATPase inhibitor carboxyeosin had no influence on Ca, Cd or Zn accumulation rates while ruthenium red, a

Ca-ATPase inhibitor, significantly decreased accumulation of each in a concentration- dependent manner.

Introduction

Current understanding of ion trafficking in aquatic insects is not commensurate with their ecological importance and use as environmental monitors. There are over 6,500 species of aquatic insects described in North America alone1 and this faunal group typically comprises 70-95% of the invertebrate species pool in freshwater ecosystems2,3. This ecological dominance, coupled with the responsiveness of insect communities to environmental change, has led to the extensive use of aquatic insects as ecological indicators.

In anthropogenically-altered ecosystems, insect communities have been used as biomonitors of ecological conditions (e.g., ref 4), and individual taxa have been used as biomonitors of changes in metal tissue concentrations over time and space (e.g., refs 5 and 6).

Species within the genus Hydropsyche (Order: Trichoptera) have been used as biomonitors of tissue metal concentrations because they readily accumulate trace metals, yet are able to tolerate high levels of metal exposure.7,8 The recent calibration of community composition data with tissue metal concentrations in Hydropsyche spp. along a metal contamination gradient further extends the use of this genus in biomonitoring,9 providing a connection between these two distinct types of biomonitoring. However, to more fully utilize

Hydropsyche as a biomonitor, it is critical to better understand trace metal bioaccumulation dynamics. This paper focuses on interactions among metals in relation to uptake across the apical membrane of an epithelial cell, specifically at the organism-water interface, and adsorption (the degree to which metals adhere to the integument).

In Hydropsyche, the uptake of trace metals occurs at anal papillae,10 specialized organs with ionoregulatory activity.11 Papillae take up essential metals and inevitably

nonessential metals, either of which are potentially harmful.12 Yet, metals such as Cd can also adsorb to the body surfaces of aquatic organisms, acting as surface active toxicants on fish gills13 and insect papillae,14 or bind to physiologically inert surfaces such as chitin.6

Little is known about the transport systems (e.g., channels, pumps) responsible for apical metal entry in Hydropsyche nor the degree to which metals interact and/or compete at these sites. Further, the influences of competing cations on adsorption processes are also poorly understood.

Zinc and Cd are borderline transition metals (sensu ref 15) that co-occur in ores in the earth’s crust and in surface waters, with Zn typically much more abundant. In aquatic environments, Zn and Cd ions exhibit similar physical and biochemical properties. Both ions have the same charge, same number of outer shell electrons, and a similar electron configuration.16 They also have comparable electronegativity values17 and share similar affinities for sulfur, oxygen, and nitrogen ligands.15 Ions of the same charge and relative size

(such as Ca and Cd) frequently compete for the same channels or transporters18,19 because transport systems often cannot make a distinction between similar ions.20 A strong interspecific covariance of Zn and Cd uptake rates has been observed under similar water chemistries in aquatic insects,21 mussels22 and crustaceans.23 Intra-specifically, Zn and Cd transport rates are reported to be similarly affected by pharmacological agents targeting calcium transport systems in aquatic insects21 and mussels.22,24 Covariance in uptake of the two trace metals is suggestive of a shared transport system (potentially Ca), though the essentiality of Zn in a wide variety of physiological processes would indicate that Zn specific transport systems are also potentially important to Zn (and possibly Cd) uptake.

Calcium is a Class A metal (sensu ref. 15) and macronutrient that has a similar effective radius (0.94 Å) to Cd (0.92 Å), though less similar to Zn (0.70 Å).25 If Cd and Zn share transporters with Ca in these aquatic species, it is possible for competition to occur.

Zinc and Cd have been shown to inhibit Ca influx in fish and crustaceans.26 Conversely, Ca has been shown to be strongly protective against Zn and Cd uptake in secretory cell line

GH4C1 containing well-characterized Ca channels18 and Zn uptake in rainbow trout

(Oncorhynchus mykiss).27 This competition has been ascribed to a limited number of binding sites on the gills of fish;28 however details on competition in aquatic insects remain vague.

Relative to Ca, Mn is less frequently studied and considerably more complex. In the environment, Mn most commonly occurs in (II) and oxide forms.29 Mn (II) has the capacity to interact and/or compete with other ions such as Ca30–32 and has been described as a Ca analog.33 In oxide forms, Mn can be highly reactive with other elements and may indeed be a sink for metals in sediment systems.34 Recent work has also demonstrated that Mn oxides form on the integument of insects including Hydropsyche,30 providing yet another opportunity to alter trace element uptake and adsorption.

Here we examined the accumulation kinetics of Zn and Cd both individually and jointly to explore potential interactions between them at both environmentally relevant and extreme concentrations. We further tested the hypothesis that Mn could interact with Cd and

Zn via competition (Mn (II)) and complexation (Mn oxides). Experiments conducted under different Ca concentrations allowed us to examine Ca – Cd/Zn interactions as well as the influence of Mn on these processes. We used different rinsing techniques to explore both total metal accumulation and internalization. Throughout this paper, we use the term

accumulation to refer to total metal accumulation (adsorbed and absorbed metal), and the term uptake to refer only to internalized metals. Lastly, we used pharmacological blockers

(verapamil, nifedipine, carboxyeosin, and ruthenium red) to specifically target different Ca transport systems in an attempt to identify possible uptake pathways for Ca, Cd, and Zn in

Hydropsyche sparna.

Materials and Methods

Insect collection and handling

Hydropsyche sparna larvae were field collected from the Eno River in North Carolina from September 2010 to October 2011 using a D-frame kicknet. Larvae were transported to the laboratory in a cooler with stream water, cobble substrate, and aeration. Acclimation occurred for a minimum of 48 hours in a walk-in cold room (12.7°C; 12:12 light:dark photoperiod) in aerated American Society for Testing and Materials (ASTM) artificial very soft water (VSW) (145 µM NaHCO3; 43.6 µM CaSO4•2H2O; 62.3 µM MgSO4; 6.71 µM

KCl). Insects were not fed during this period. Voucher specimens were preserved in 75% ethanol for each experiment and verified by an independent taxonomist. Only larvae that appeared healthy were used for experimentation.

Radioactivity measurement

The gamma-emitting radioisotopes 65Zn, 109Cd, and 54Mn were measured using a

Perkin-Elmer Wallac Wizard 1480 Automatic Gamma Counter (Shelton, CT, USA). Each isotope was acquired as a chloride salt in HCl (65Zn and 54Mn from Perkin-Elmer (Billerica,

MA, USA) and 109Cd from Los Alamos National Laboratory (Los Alamos, NM, USA)).

Working stock solutions were made by diluting each isotope in 0.1 N HNO3. Protocols for counting two (109Cd and 65Zn) and three (109Cd, 54Mn and 65Zn) isotopes simultaneously were established with spillover corrections and verified against single and mixed isotope standards as appropriate. All radioactivity measurements (solutions and larvae (in vivo)) were

conducted for 3 minutes and sufficient radioactivity of working solutions ensured that counting errors were small (mean ± s.d.; 1.76 ± 1.3%).

45 Calcium (as CaCl2 in H2O) was obtained from Perkin-Elmer (Billerica, MA, USA) and diluted in 0.1 N HNO3. Water samples (1 mL) were counted using 20 mL scintillation vials with 16 mL Scintisafe® liquid scintillation cocktail (Perkin-Elmer; Waltham, MA,

USA). Larvae were digested in 1 mL Soluene® (Perkin-Elmer; Waltham, MA, USA) at 60

°C for 24 hours prior to addition of scintillation cocktail. All 45Ca samples (water and larvae) were counted for 3 minutes using a Beckman LS6500 Multipurpose Scintillation Counter.

Counting errors remained small (mean ± s.d.; 3.21 ± 1.6%) and Lumex values were <5%.

Zinc and cadmium: Accumulation kinetics and competition

Accumulation rates of Cd and Zn were each measured individually in larvae in initial experiments at a range of concentrations spanning 4 orders of magnitude to represent environmentally relevant and extreme dissolved concentrations (Zn: 0.0153, 0.153, 1.53, and

15.3 µM; Cd: 0.0089, 0.089, 0.89, and 8.9 µM). When results suggested the presence of a saturable transport system at environmentally relevant concentrations, two subsequent experiments were performed to better resolve these kinetics: (Experiment 1- Zn: 0.0275,

0.050, 0.10, 0.20, 0.40 µM and Cd: 0.0125, 0.025, 0.050, 0.10, 0.20 µM; Experiment 2- Zn:

0.2, 0.4, 0.8, 1.2 µM and Cd: 0.1, 0.2, 0.4, 0.8 µM). Solutions were prepared in bulk (700 mL solutions) to ensure identical treatment among replicates for a given element. For Zn treatments, all bulk solutions were spiked with 65Zn tracer to achieve exposure activities of

-1 102 kBq L with stable Zn (as ZnCl2) comprising the majority of Zn in solution. Similarly,

Cd solutions were spiked with 109Cd tracer to achieve exposure activities of 29.5 kBq L-1 with stable Cd (as CdCl2) comprising the majority of Cd in solution. The pH of each bulk solution was adjusted to 7.20 ± 0.02 with the addition of 0.1 N NaOH. For each replicate

(n=8), 70 mL of bulk solution were distributed into individual 100 mL aerated high-density polyethylene (HDPE) cups containing a small square of teflon mesh as substrate. Each replicate consisted of a single larva, and parafilm was used to reduce evaporative losses.

Insects were exposed to dissolved concentrations for a total of 9 hours. At 3, 6, and 9 hours, larvae were removed, rinsed with VSW, assayed in vivo for radioactivity (see above), and returned to exposures. After the last time point larvae were blotted dry and wet weights were recorded.

Single and dual-labeled treatments at environmentally low (46 nM Zn and 2.7 nM

Cd), high (0.6 µM Zn and 0.6 µM Cd and extreme (15.3 µM Zn and 8.9 µM Cd) concentrations were completed in order to explore potential competition between these elements. Bulk solutions and replicates (n=10 for environmentally relevant and high, n=8 for extreme) were prepared with radiotracer as above. Accumulation rates of the dual-labeled treatment were compared to the single metal exposures of the same concentrations. Single metal exposures were included in the analysis of accumulation kinetics.

Assessing the influence of potentially competing cations: Calcium and manganese.

To assess the ability of Ca and Mn to compete with Cd and Zn accumulation, we used a 4x2 factorial design containing four Mn concentrations (0 µM, 0.24 µM, 1.95 µM and 19.8

µM) with 54Mn as a radiotracer (specific activity: 48.5 kBq L-1) and two Ca concentrations

(31.1 ± 2.68 µM and 1.35 ± 0.44 mM). Cadmium (19.99 ± 2.86 nM; specific activity: 29.5 kBq L-1) and Zn (462.03 ± 73.7 nM; specific activity: 102 kBq L-1) concentrations were fixed in all treatment groups. The low Ca solution contained the standard VSW Ca concentration, and the high Ca solution consisted of VSW spiked with additional CaSO4 to achieve the Ca concentration of very hard water (ASTM standard).

Replicates (n=10) were prepared as above and assayed for radioactivity at 3, 6, 9 and

24 hour time points. Wet weights were obtained after the 24 hour exposure. Five insects in each treatment were then rinsed with 0.05 M EDTA alone for 30 seconds to chelate and remove adsorbed metal, and the remaining five were rinsed with 0.1 M ascorbate for 30 seconds to reduce metal oxides followed by a 30 second rinse in 0.05 M EDTA. Following rinses, larvae were re-assayed for radioactivity. We interpret metals removed by EDTA to be adsorbed metals likely in the (II) oxidation state. We interpret the remaining metals associated with larvae after both ascorbate and EDTA rinses to be absorbed (internalized) metals. Finally, the difference between the EDTA alone and ascorbate/EDTA rinses are interpreted as metals associated with oxide phases.

Calcium transport system blockers

The accumulation of Ca, Cd, and Zn were examined in the presence of four Ca blockers: L-type Ca channel blockers verapamil and nifedipine, plasma membrane Ca-

ATPase inhibitor carboxyeosin, and ruthenium red, a Ca-ATPase inhibitor. Ruthenium red and nifedipine were obtained from Sigma-Aldrich (St. Louis, MO, USA), verapamil HCl

from MP Biomedicals (Solon, OH, USA), and carboxyeosin from MGT, Inc (Eugene OR,

USA).

To facilitate the measurement of 45Ca influx in larvae, we prepared VSW with ½ the

45 -1 Ca content (21.8 µM CaSO4•2H2O) and radiotracer Ca (specific activity: 131 kBq L ). At least two concentrations of each Ca blocker were tested. Larvae were exposed to verapamil and nifedipine concentrations of 0, 1, 10, and 100 µM. Nifedipine required DMSO as a carrier, and appropriate DMSO controls were used. Exposure concentrations for ruthenium red and carboxyeosin were 0, 10, and 100 µM. Each treatment had 8 replicates- each with one animal contained in a Parafilm-covered 50 mL HDPE cup with 30 mL solution, aeration, and Teflon mesh substrate. After a 6 hour exposure, larvae were removed, weighed, rinsed with 0.05 M EDTA for 30 seconds to remove adsorbed Ca, and digested in 1 mL Soluene®.

Replicates were assayed for radioactivity individually (see above).

Zinc and Cd total accumulation and uptake rates in the presence of Ca blockers were measured using dual-labeled exposures of 306 nM Zn and 17.8 nM Cd with radiotracers used as above in VSW. Insects were exposed to treatment concentrations of Ca blockers identical to those used for Ca experiments. For each replicate (n=8), an insect was exposed to 80 mL solution in a 100 mL Parafilm-covered HDPE cup with aeration and Teflon mesh. Insects were assayed in vivo for radioactivity after 3, 6, and 9 hours in exposure. After 9 hours, wet weights were obtained and insects were rinsed with 0.05 M EDTA for 30 seconds before re- assaying for radioactivity to determine absorbed-only metal.

Data analysis

Data analysis was performed using GraphPad Prism (v5.04). Total accumulation rates of Zn, Cd, and Mn were determined by linear regression, and uptake rate constants were derived in the traditional fashion as the slope of accumulation vs. concentration plots.

Michaelis-Menten parameters were obtained through non-linear regression. Student t-tests and one-way analysis of variance (ANOVA) with Tukey’s post-hoc test were used to determine significant differences between treatments. All values are given as value ± standard error of the mean (s.e.m.). Values were considered significantly different at p ≤

0.05.

Results

Zinc and cadmium accumulation kinetics

The accumulation kinetics of Zn at low concentrations ranging from 0.0153 to 0.153

µM and Cd at concentrations of 0.0027 to 0.2 µM were well described by linear regression models, and corresponding uptake rate constants were generated for both Zn (0.53 ± 0.06 L g-1 d-1, r2 = 0.94) and Cd (0.26 ± 0.03 L g-1 d-1, r2 = 0.90) on a wet weight basis. (Fig. 1A). At

~0.2 µm, Zn accumulation rates begin to slow and are briefly saturated at concentrations between 0.4 µM and 0.8 µM. These data fit Michaelis-Menten type kinetics better than linear

-1 -1 2 models (Vmax = 0.157 ± 0.02 µmol g d ; Km = 0.256 ± 0.09 µM, r = 0.64). It is apparent that Zn is transported by a second system at concentrations greater than 0.8 µM, as accumulation again becomes rapid and shows no signs of saturation up to 15.3 uM. The evidence for saturable transport of Cd within the concentrations tested (0.0027 µM to 8.9

µM) is weaker (Fig. 1B). There is an apparent slight slowing of accumulation within the same range of concentrations where Zn transport is clearly saturated (0.4 µM to 0.8 µM), however linear models fit these data better than Michaelis-Menten kinetics. We do note that the uptake rate constant for concentrations below 0.8 µM was 0.227 ± 0.006 L g-1 d-1 (r2 =

0.993), 46% lower than the uptake rate constant for the whole range of concentrations (0.42 ±

0.01 L g-1 d-1 (r2 = 0.99). While Cd saturation was not apparent, this vast difference in uptake rate constants suggests the presence of at least two transport systems.

Competition experiments were performed with single and dual isotope exposures at three concentrations. At environmentally low concentrations (46 nM Zn and 2.7 nM Cd), accumulation rates of each element were unchanged by the presence of the other (data not

shown). At higher concentrations of 0.6 µM Zn and 0.6 µM Cd, Zn accumulation was slowed

35% in the presence of Cd while Cd accumulation remained unaffected by Zn (Fig 2A). At extreme concentrations (15.3 µM Zn and 8.9 µM Cd), accumulation rates of Zn were 58% slower in the presence of Cd (Fig 2B). These results suggest that Cd generally out-competes

Zn for apical entry in H. sparna.

Manganese interactions with cadmium and zinc

To assess the possible protective effect of Mn on Cd and Zn total accumulation and uptake, we held Cd (20 nM) and Zn (462 nM) constant while varying Mn concentrations from 0-19.8 µM. This experiment was conducted at 2 levels of Ca, 31.1 µM and 1.35 mM.

Manganese decreased total Cd and Zn accumulation in a concentration dependent manner under both Ca conditions (Fig. 3A-D). Total Zn accumulation was reduced by 61% in insects exposed to 19.8 µM Mn relative to Mn-free water contained in low Ca exposures and by 53% in high Ca exposures. Correspondingly, total Cd accumulation in larvae was reduced by 47% and 48% at 19.8 µM Mn in low and high Ca exposures, respectively. The effect of Mn on Zn and Cd accumulation seems to be primarily due to differences in adsorption, not uptake.

Analysis of post ascorbate/EDTA rinsed larvae suggests that Mn had no significant effects on

Zn or Cd uptake (internalized concentrations) at any Mn concentration at either high or low ambient Ca concentrations (Fig. 3A-D).

Different rinses with EDTA and ascorbate were used to examine interactions of Zn and Cd with oxide phases on the surface of the animals. We observed differences in the magnitude of metals removed by EDTA alone vs. ascorbate/EDTA rinse, suggesting that

some Cd and Zn was associated with an oxide phase, but not enough to reduce internalization of these metals. At 31 µM Ca (across Mn treatments), an average of 36% of adsorbed Zn and

31% of adsorbed Cd were associated with an oxide phase. At 1.35 mM Ca, an average of

36% of adsorbed Zn and 28% of adsorbed Cd were associated with an oxide phase. This suggests that substantial amounts of Cd and Zn were associated with an oxide phase, but that neither Mn competition for uptake or sequestration by an oxide phase limited Cd and Zn uptake in these experiments.

Calcium interactions with cadmium and zinc

To examine the potential for Ca to protect against the total accumulation and uptake of Cd and Zn, we examined accumulation kinetics at Ca concentrations differing 43-fold

(Fig. 3A-D). In the absence of Mn, neither total accumulation nor uptake of Cd and Zn differed with Ca treatment. The influence of Ca on total Cd and Zn accumulation and uptake was subtle at different Mn concentrations and was never statistically significant, partially due to small sample sizes. Because Mn exerted no effect on internalization (the uptake of Zn and

Cd were statistically similar across Mn treatments within each Ca treatment), data for Zn and

Cd uptake were pooled (Fig. 3E-F). A 43-fold increase in Ca decreased Zn uptake by 31%, (p

= 0.04), however the 22% decrease in Cd uptake was not significant. It is apparent that Ca does not exert a major influence on Cd or Zn total accumulation or uptake in this species at the tested concentrations.

Manganese accumulation dynamics

In the presence of Cd and Zn, total Mn accumulation increases rapidly with increasing Mn concentration (Fig 4A-C). Interestingly, the profile of Mn accumulation is different than that of Cd and Zn, particularly with respect to adsorption. Much of the Mn adsorption was in the form of oxides. In four of six treatment combinations, oxide phase Mn comprised the majority of the total adsorbed Mn, averaging 57%, and was never less than

40%. By contrast, Cd and Zn association with oxide phases was lower, averaging 32% and

33%, respectively. Calcium also had a much stronger effect on Mn accumulation. Increasing

Ca concentrations from 31.1 µM to 1.35 mM significantly reduced total Mn accumulation in an idiosyncratic manner (56%, 15%, and 21% at Mn concentrations of 0.24, 1.95, and 19.8

µM, respectively). Similarly, increasing Ca concentrations from 31.1 µM to 1.35 mM significantly reduced Mn uptake in all treatments (reductions of 75%, 58%, and 57% in 0.24,

1.95, and 19.8 µM Mn, respectively).

Calcium transport system blockers

Different pharmacological agents were used to target Ca transport systems to elucidate possible uptake pathways for Ca, Cd, and Zn. Verapamil, nifedipine, and carboxyeosin were all found to be ineffective at concentrations up to 100 µM against Ca, Cd, and Zn uptake at the concentrations tested (data not shown). However, the Ca-ATPase inhibitor ruthenium red decreased the uptake rate of Ca and the accumulation rates of Zn and

Cd in a concentration-dependent manner. Relative to controls, 10 and 100 µM ruthenium red reduced the uptake rate of Ca 53% and 93.4%, respectively (Fig. 5), thus demonstrating a

concentration-dependent inhibition of Ca influx. At 10 µM ruthenium red, Zn accumulation rates were reduced by 60% and Cd accumulation rates were slowed by 67% in comparison to controls (Fig. 6A, B). At 100 µM ruthenium red, Zn accumulation rates were reduced by

89% and Cd accumulation rates were reduced by 87%.At a concentration of 10 µM, ruthenium red significantly reduced Zn uptake by 28% but failed to significantly reduce Cd uptake (Fig. 6C, D). At 100 µM, ruthenium red significantly reduced Zn uptake by 80% and

Cd uptake by 71%.

Discussion

The extensive use of aquatic insects as biomonitors requires that we better understand their fundamental physiological processes. In the case of trace metal contamination in streams, the genus Hydropsyche has received considerable attention from numerous authors7–

9,35,36 and prompts us to better understand both metal accumulation mechanisms and interactions among metals. Here we focused on metal interactions in relation to uptake from water, but also examined adsorption in light of the view that metals are often viewed as surface active toxicants. Moreover, insects occupy important trophic positions as primary food sources for fish and birds with both absorbed and adsorbed metals potentially available via diet.

Zinc and cadmium accumulation kinetics

Environmental background concentrations of Zn rarely exceed 0.612 µM (40 µg L-

1),5,37,38 however higher concentrations are present in anthropogenically-affected areas such as mining sites. In our experiments, Zn transport briefly saturates within the range of environmental background concentrations, with saturation occurring between 0.4 µM and 0.8

µM Zn. Environmental background concentrations of Cd are much lower, ranging from 89 pM to 4.4 nM (10 to 500 ng L-1),39,40 however like Zn, higher concentrations can be present in contaminated areas. In our experiments, Cd transport never reaches saturation at environmental background concentrations, nor does it saturate at concentrations up to 0.8

µM. However, the Cd accumulation rate does increases at concentrations greater than 0.89

µM Cd. This suggests that at least 2 transport systems are involved in the accumulation of Zn

and Cd from water into H. sparna tissues. There is an array of routes available for trace metal uptake including transport systems for both trace essential metals (such as Zn) and macronutrients (such as Ca). While there may be multiple pathways for Zn and Cd entry, saturation did not occur and accumulation rates were still increasing in a linear fashion at the highest concentrations tested (15.3 µM Zn and 8.9 µM Cd). Competition experiments between Cd and Zn suggest that these two metals may be transported, at least partially, by similar transport systems. At very low concentrations of each metal (46 nM Zn and 2.7 nM

Cd) no competition was observed, possibly because different transporters are at work, or a shared transport system has ample capacity to transport both without evidence of competition. At higher concentrations (0.6 µM, where Zn transport is briefly saturated, and

Cd transport rates appear to slow slightly), Cd clearly out-competes Zn. Finally at the extreme concentrations (15.3 µM Zn and 8.9 µM Cd) where presumably a different transport system predominates, Cd again out-competes Zn. Because Cd out-competes Zn at concentrations falling within ranges for both observed transport systems, it can be inferred that Cd and Zn are entering through the same two (at least) transport systems. This also suggests a higher affinity of Cd for both shared transport systems observed in H. sparna. The ionic radius of Cd is more similar to Ca than Zn;25 therefore, if one or both of these transport systems that Zn and Cd traverse in H. sparna are Ca transport systems, Cd may have an advantage in entry. Alternatively, Zn transport systems may be responsible for observed Cd and Zn transport.

Studies on the competition of Zn and Cd remain scarce in aquatic insects. Previous work on Hydropsyche californica observed Zn out-competing Cd at extremely low

concentrations of 0.08 nM Zn and 0.17 nM Cd.21 The closest concentration tested in our experiment had concentrations of 46 nM Zn and 2.7 nM Cd, at which we found no interactions between metals in accumulation. Other competition studies between Zn and Cd in different aquatic taxa yield varying results. For example, a large excess of Zn was needed to interact with Cd in bivalve species,24,41 and Cd was found to have a higher metal-gill binding affinity than Zn in juvenile rainbow trout,13 supporting our hypothesis of shared transport systems with higher affinities for Cd. Other studies show no interactions between

Cd and Zn. Rainbow et al.42 found no interaction between Zn and Cd in crustaceans Carcinus maenas, Pachygrapsus marmoratus, and Orchestia gammarellus while Wang and Fisher22 found no interaction in the mussel Mytilus edulis.

Manganese interactions with cadmium and zinc

We hypothesized that Mn could significantly alter Zn and Cd accumulation in H. sparna either through competition with Mn (II) or by complexation/association with oxides.

29 Manganese oxides (MnOx) are highly reactive phases with high sorptive capacities. They are reported to react with the reduced forms of other metals in soils and sediments including

Cd43–45 and Zn,46,47 thereby changing their bioavailability. As MnO was previously observed forming on the integument of aquatic insects,30 we hypothesized a high potential for interaction with Cd and Zn in these studies. Results herein showed that Mn reduced adsorption but not uptake of Cd and Zn, despite the apparent formation of significant oxide phases on the exoskeleton. We note that Cd and Zn were found in oxide phases of even the 0

μM Mn treatment groups, suggesting that these field collected larvae came into the lab with some form of oxide phase already present on the integument.

Calcium interactions with cadmium and zinc

In our studies, a 43-fold increase in Ca concentration resulted in only modest decreases in Zn and Cd adsorption and uptake. The literature is mixed with respect to the protective effects of Ca against Cd and Zn. In toxicity assays where organisms are often exposed to ecologically irrelevant metal concentrations, Ca often provides significant protection against Cd48,49 and Zn50 toxicity, and is a major player in Biotic Ligand

Modeling.13 Under environmentally relevant exposure regimes, the literature provides conflicting information regarding the interaction of Ca with Cd and Zn uptake. Calcium has been found to be protective against the uptake of Cd and Zn in aquatic insects H. californica and Drunella flavilinea (Buchwalter and Luoma, 2005), juvenile rainbow trout (O. mykiss),51 water fleas (Daphnia magna),52 mollusks (Littorina littorea),53 and crabs (C. maenas).54

However, other studies with mussels (M. edulis) and clams (Macoma balthica) have shown

Ca to be ineffective against the accumulation of Cd or Zn.22,53

Calcium transport system blockers

While Ca was weakly inhibitory toward Cd and Zn uptake in the present study, Ca channel blockers verapamil and nifedipine were unsuccessful in blocking Ca, Cd, and Zn at concentrations of 21.8 µM Ca, 306 nM Zn and 17.8 nM Cd. Therefore, an L-type voltage gated Ca channel is likely not responsible for the influx of these ions into H. sparna at these

tested concentrations (within the range of the higher-affinity, saturable transport system). It is possible that verapamil and nifedipine might inhibit Cd and Zn uptake at concentrations of

Cd and Zn above 0.8 µM, within the range of a second higher capacity transport system.

Experimental findings in regards to these Ca pharmacological blockers in aquatic insects are mixed. Verapamil failed to inhibit the influx of Zn and/or Cd in aquatic insects

Chironomus staegeri55 and H. californica21 while successfully inhibiting metal flux in D. flavilinea.21 In rainbow trout, verapamil and nifedipine were ineffective in blocking Zn and

Cd influx.56 However, several studies report significant inhibition of Zn and Cd by verapamil and nifedipine in mollusks Crassostrea virginica,57 M. edulis,22,24 and M. balthica,22 D. magna,58 and rainbow trout59 suggesting a shared Ca channel as means of apical entrance for these species.

Ruthenium red, a specific Ca-ATPase inhibitor,60 significantly inhibited the uptake of

Zn, Cd, and Ca in a concentration-dependent manner. The effects of ruthenium red on ion transport in aquatic species remains less well-known than those of other more well-studied

Ca blockers such as verapamil. However, ruthenium red has been shown to inhibit Ca uptake in the mosquito larvae Aedes aegypti.61 Our work suggests that a Ca-ATPase transporter is responsible for the aqueous influx of Ca, Cd, and Zn into H. sparna at the low, environmentally-relevant concentrations tested whether directly, indirectly by regulating electrochemical gradients, or otherwise. It is unknown whether a Ca-ATPase inhibitor would inhibit Cd or Zn at concentrations above 0.8 µM (within the range of the second lower- affinity transport system found).

A host of anthropogenic activities (e.g., mining, natural gas extraction, urbanization) change the ionic and metal composition of surface waters. Yet our understanding of fundamental ion transport physiology remains remarkably poor, especially in aquatic insects.

We suggest that improving our physiological understanding of these important organisms will lead to improvements in their use as biomonitors.

Acknowledgements

The authors thank W. Crouch (NCDENR) for taxonomic assistance. G.L. LeBlanc

(NCSU), J.M. Conley (NCSU), D.J. Cain (USGS) and anonymous reviewers provided valuable editorial comments. M.D.-J. was supported by CONICYT Chile fellowship and

M.D.P. was supported by NSF (IOS 0919614).

References

(1) Merritt, R. W.; Cummins, K. W.; Berg, M. B., Eds. An Introduction to the Aquatic Insects of North America, 4th, ed.; Kendall-Hunt: Dubuque, IA, 2008.

(2) Arscott, D. B.; Jackson, J. K.; Kratzer, E. B. Role of rarity and taxonomic resolution in a regional and spatial analysis of stream macroinvertebrates. J. N. Am. Benthol. Soc. 2006, 25, 977–997.

(3) Minshall, G. W. Community structure in natural stream systems. In The Stream Ecosystem. Boston: AAAS Symposium; 1969.

(4) Clements, W. H.; Carlisle, D. M.; Lazorchak, J. M.; Johnson, P. C. Heavy metals structure benthic communities in Colorado mountain streams. Ecol. Appl. 2000, 10, 626–638.

(5) Cain, D. J.; Luoma, S. N.; Wallace, W. G. Linking metal bioaccumulation of aquatic insects to their distribution patterns in a mining-impacted river. Environ. Toxicol. Chem. 2004, 23, 1463–1473.

(6) Hare, L. Aquatic insects and trace metals: Bioavailability, bioaccumulation, and toxicity. Crit. Rev. Toxicol. 1992, 22, 327–369.

(7) Cain, D. J.; Buchwalter, D. B.; Luoma, S. N. Influence of metal exposure history on the bioaccumulation and subcellular distribution of aqueous cadmium in the insect Hydropsyche californica. Env. Toxicol. Chem. 2006, 25, 1042–1049.

(8) Luoma, S. N.; Rainbow, P. S. Metal Contamination in Aquatic Environments; Cambridge University Press: New York, 2008.

(9) Luoma, S. N.; Cain, D. J.; Rainbow, P. S. Calibrating biomonitors to ecological disturbance: A new technique for explaining metal effects in natural waters. Integ. Env. Assess. Manag. 2010, 6, 199–209.

(10) Komnick, H. Chloride cells and chloride epithelia of aquatic insects. Int. Rev. Cytol. 1977, 49, 285–329.

(11) Koch, H. J. The absorption of chloride ions by the anal papillae of Diptera larvae. J. Exp. Biol. 1938, 15, 152–160.

(12) Tessier, L.; Boisvert, J. L.; Vought, L. B. M.; Lacoursiere, J. O. Anomalies on capture nets of Hydropsyche slossonae larvae (Trichoptera; Hydropsychidae) following a sublethal chronic exposure to cadmium. Env. Pollut. 2000, 108, 425–438.

(13) Niyogi, S.; Wood, C. M. Biotic ligand model, a flexible tool for developing site- specific water quality guidelines for metals. Env. Sci. Technol. 2004, 38, 6177–6192.

(14) Vuori, K. M. Rapid behavioral and morphological responses of hydropsychid larvae (Trichoptera, Hydropsychidae) to sublethal cadmium exposure. Environ. Pollut. 1994, 84, 291–299.

(15) Nieboer, E.; Richardson, D. H. S. The replacement of the non-descript term “heavy metals” by a biologically and chemically significant classification of metal ions. Env. Pollut. B. 1980, 1, 3–26.

(16) Newman, M. C.; McCloskey, J. T. Predicting relative toxicity and interactions of divalent metal ions: Microtox(R) bioluminescence assay. Env. Toxicol. Chem. 1996, 15, 275–281.

(17) Allred, A. L. Electronegativity values from thermochemical data. J. Inorg. Nucl. Chem. 1961, 17, 215–221.

(18) Hinkle, P. M.; Kinsella, P. A.; Osterhoudt, K. C. Cadmium uptake and toxicity via voltage-sensitive calcium channels. J. Biol. Chem. 1987, 262, 16333–16337.

(19) Simkiss, K.; Taylor, M. G. Transport of Metals Across Membranes. In Metal Speciation and Bioavailability in Aquatic Systems; Tessier, A.; Turner, D. R., Eds.; John Wiley & Sons Ltd: West Sussex, England, 1995; pp. 1–45.

(20) Williams, R. J. P.; da Silva, J. J. R. F. The distribution of elements in cells. Coord. Chem. Rev. 2000, 200, 247–348.

(21) Buchwalter, D. B.; Luoma, S. N. Differences in dissolved cadmium and zinc uptake among stream insects: Mechanistic explanations. Env. Sci. Technol. 2005, 39, 498– 504.

(22) Wang, W. X.; Fisher, N. S. Effects of calcium and metabolic inhibitors on trace element uptake in two marine bivalves. J. Exp. Mar. Biol. Ecol. 1999, 236, 149–164.

(23) Rainbow, P. S. Physiology, Physicochemistry and Metal Uptake - A Crustacean Perspective. Mar. Pollut. Bull. 1995, 31, 55–59.

(24) Vercauteren, K.; Blust, R. Uptake of cadmium and zinc by the mussel Mytilus edulis and inhibition by calcium channel and metabolic blockers. Mar. Biol. 1999, 135, 615– 626.

(25) Williams, R. J. P.; Frausto da Silva, J. R. The Natural Selection of the Chemical Elements; Clarendon Press: Oxford, 1996.

(26) Wright, D. A. Trace metal and major ion interactions in aquatic animals. Mari. Pollut. Bull. 1995, 31, 8–18.

(27) Barron, M. G.; Albeke, S. Calcium control of zinc uptake in rainbow trout. Aquat. Toxicol. 2000, 50, 257–264.

(28) Playle, R. C.; Gensemer, R. W.; Dixon, D. G. Copper accumulation on gills of fathead minnows - Influence of water hardness, complexation and pH of the gill microenvironment. Environ. Toxicol. Chem. 1992, 11, 381–391.

(29) Tebo, B. M.; Bargar, J. R.; Clement, B. G.; Dick, G. J.; Murray, K. J.; Parker, D.; Verity, R.; Webb, S. M. Biogenic manganese oxides: Properties and mechanisms of formation. Annu. Rev. Earth Planet. Sci. 2004, 32, 287–328.

(30) Dittman, E. K.; Buchwalter, D. B. Manganese bioconcentration in aquatic insects: Mn oxide coatings, molting loss, and Mn (II) thiol scavenging. Env. Sci. Technol. 2010, 44, 9182–9188.

(31) Stubblefield, W. A.; Brinkman, S. E.; Davies, P. H.; Garrison, T. D. Effects of water hardness on the toxicity of manganese to developing brown trout (Salmo trutta). Env. Toxicol. Chem. 1997, 16, 2082–2089.

(32) Lasier, P. J.; Winger, P. V.; Bogenrieder, K. J. Toxicity of manganese to Ceriodaphnia dubia and Hyalella azteca. Arch. Environ. Contam. Toxicol. 2000, 38, 298–304.

(33) Markich, S. J.; Jeffree, R. A. Absorption of divalent trace metals as analogues of calcium by Australian freshwater bivalves: An explanation of how water hardness reduces metal toxicity. Aquat. Toxicol. 1994, 29, 257–290.

(34) Trivedi, P.; Axe, L. Modeling Cd and Zn sorption to hydrous metal oxides. Env. Sci. Technol. 2000, 34, 2215–2223.

(35) Barata, C.; Lekumberri, I.; Vila-Escale, M.; Prat, N.; Porte, C. Trace metal concentration, antioxidant enzyme activities and susceptibility to oxidative stress in the tricoptera larvae Hydropsyche exocellata from the Llobregat river basin (NE Spain). Aquat. Toxicol. 2005, 74, 3–19.

(36) Evans, R. D.; Balch, G. C.; Evans, H. E.; Welbourn, P. M. Simultaneous measurement of uptake and elimination of cadmium by caddisfly (Trichoptera: Hydropsychidae) larvae using stable isotope tracers. Env. Toxicol. Chem. 2002, 21, 1032–1039.

(37) Eisler, R. Zinc hazards to fish, wildlife, and invertebrates: A synoptic review. Biological Report 10, 1993.

(38) Van Genderen, E.; Adams, A.; Cardwell, R.; Volosin, J.; Santore, R.; Rodriguez, P. An evaluation of the bioavailability and aquatic toxicity attributed to ambient zinc concentrations in fresh surface waters from several parts of the world. Integr. Environ. Assess. Manag. 2009, 5, 426–434.

(39) Jensen, A.; Bro-Rasmussen, F. Environmental cadmium in Europe. Rev. Environ. Contam. Toxicol. 1992, 125, 101–181.

(40) Xue, H. B.; Sigg, L. Cadmium speciation and complexation by natural organic ligands in fresh water. Anal. Chim. Acta. 1998, 363, 249–259.

(41) Jackim, E.; Morrison, G.; Steele, R. Effects of environmental factors on radiocadmium uptake by four species of marine bivalves. Mar. Biol. 1977, 40, 303–308.

(42) Rainbow, P. S.; Amiard-Triquet, C.; Amiard, J. C.; Smith, B. D.; Langston, W. J. Observations on the interaction of zinc and cadmium uptake rates in crustaceans (amphipods and crabs) from coastal sites in UK and France differentially enriched with trace metals. Aquat. Toxicol. 2000, 50, 189–204.

(43) Chen, Z. S.; Lee, G. J.; Liu, J. C. The effects of chemical remediation treatments on the extractability and speciation of cadmium and lead in contaminated soils. Chemosphere. 2000, 41, 235–242.

(44) Dong, D. M.; Nelson, Y. M.; Lion, L. W.; Shuler, M. L.; Ghiorse, W. C. Adsorption of Pb and Cd onto metal oxides and organic material in natural surface coatings as determined by selective extractions: New evidence for the importance of Mn and Fe oxides. Water. Res. 2000, 34, 427–436.

(45) Mench, M. J.; Didier, V. L.; Loffler, M.; Gomez, A.; Masson, P. A mimicked in-situ remediation study of metal-contaminated soils with emphasis on cadmium and lead. J. Env. Qual. 1994, 23, 58–63.

(46) Li, X. D.; Shen, Z. G.; Wai, O. W. H.; Li, Y. S. Chemical forms of Pb, Zn and Cu in the sediment profiles of the Pearl River Estuary. Mari. Pollut. Bull. 2001, 42, 215– 223.

(47) Singh, A. K.; Hasnain, S. I.; Banerjee, D. K. Grain size and geochemical partitioning of heavy metals in sediments of the Damodar River - a tributary of the lower Ganga, India. Env. Geol. 1999, 39, 90–98.

(48) Carroll, J. J.; Ellis, S. J.; Oliver, W. S. Influences of hardness constituents on the acute toxicity of cadmium to brook trout (Salvelinus fontinalis). Bull. Env. Contam. Toxicol. 1979, 22, 575–581.

(49) Meinelt, T.; Playle, R. C.; Pietrock, M.; Burnison, B. K.; Wienke, A.; Steinberg, C. E. W. Interaction of cadmium toxicity in embryos and larvae of zebrafish (Danio rerio) with calcium and humic substances. Aquat. Toxicol. 2001, 54, 205–215.

(50) Heijerick, D. G.; De Schamphelaere, K. A. C.; Janssen, C. R. Predicting acute zinc toxicity for Daphnia magna as a function of key water chemistry characteristics: Development and validation of a biotic ligand model. Env. Toxicol. Chem. 2002, 21, 1309–1315.

(51) Hollis, L.; McGeer, J. C.; McDonald, D. G.; Wood, C. M. Protective effects of calcium against chronic waterborne cadmium exposure to juvenile rainbow trout. Env. Toxicol. Chem. 2000, 19, 2725–2734.

(52) Tan, Q.-G.; Wang, W.-X. The influence of ambient and body calcium on cadmium and zinc accumulation in Daphnia magna. Env. Toxicol Chem 2008, 27, 1605–1613.

(53) Bjerregaard, P.; Depledge, M. H. Cadmium accumulation in Littorina littorea, Mytilus edulis and Carcinus maenas: The influence of salinity and calcium ion concentrations. Mar. Biol. 1994, 119, 385–395.

(54) Wright, D. A. Effect of salinity on cadmium uptake by tissues of shore crab Carcinus maenas. J. Exp. Biol. 1977, 67, 137–146.

(55) Craig, A.; Hare, L.; Tessier, A. Experimental evidence for cadmium uptake via calcium channels in the aquatic insect Chironomus staegeri. Aquat. Toxicol. 1999, 44, 255–262.

(56) Rogers, J. T.; Wood, C. M. Characterization of branchial lead-calcium interaction in the freshwater rainbow trout Oncorhynchus mykiss. J. Exp. Biol. 2004, 207, 813–825.

(57) Roesijadi, G.; Unger, M. E. Cadmium uptake in gills of the mollusc Crassostrea virginica and inhibition by calcium channel blockers. Aquat. Toxicol. 1993, 24, 195– 206.

(58) Tan, Q. G.; Wang, W. X. Acute toxicity of cadmium in Daphnia magna under different calcium and pH Conditions: Importance of influx rate. Env. Sci. Technol. 2011, 45, 1970–1976.

(59) Li, Z.-H.; Li, P.; Randak, T. Protective roles of calcium channel blocker against cadmium-induced physiological stress in freshwater teleost Oncorhynchus mykiss. Water Air Soil Pollut. 2011, 220, 293–299.

(60) Watson, E. L.; Vincenzi, F. F.; Davis, P. W. Ca2+-activated membrane ATPase - selective inhibition by ruthenium red. Biochim. Biophys. Acta. 1971, 249, 606–610.

(61) Barkai, A. I.; Williams, R. W. The exchange of calcium in larvae of the mosquito Aedes aegypti. J. Exp. Biol. 1983, 104, 139–148.

Figures

10 5.0 B

5 A )

) 2.5

-1 -1

0.25 0.25

day

day -1 0.20 -1 0.20 0.15 0.15

0.10 0.10

Uptake Rates Uptake

Uptake Uptake Rates

mol Zn g Zn mol

mol Cd g Cd mol



 ( 0.05 ( 0.05 0.00 0.00 0.0 0.2 0.4 0.6 0.8 5 10 15 0.0 0.2 0.4 0.6 0.8 2 4 6 8 Zn Concentration ( M) Cd Concentration (M)

Figure 1. A). At concentrations ranging from 15.3 nM to 15.3 µM, at least two components of Zn accumulation are apparent. At concentrations below 0.8 µM, -1 -1 evidence for saturation is present (Vmax = 0.157 ± 0.023 µmol g d ; Km = 0.256 ± 0.089 µM, r2 = 0.64). B). At concentrations ranging from 26.7 nM to 8.9 µM, Cd accumulation fit a linear model (r2 = 0.993). At concentrations below 0.8 µM Cd, there is no evidence for saturation. However, this range of concentrations has an uptake rate constant 46% lower than that for the whole range of concentrations (0.23 ± 0.01 L g-1 d-1 vs. 0.42 ± 0.01 L g-1 d-1). Data points represent mean + s.e.m.

0.20 A: 0.6 M Zn; 0.6 M Cd 10 B: 15.3 M Zn; 8.9 M Cd

) -1 8

0.15 day

-1 * 6 0.10 4 *

Uptake Rates Uptake 0.05

mol metal g metal mol 2

 ( 0.00 0

Zn Only Zn + Cd Cd Only Cd + Zn Zn Only Zn + Cd Cd Only Cd + Zn

Figure 2. A). At concentrations of 0.6 µM Cd and 0.6 µM Zn, the presence of Cd inhibited Zn accumulation by 35%, but Cd is unaffected by the presence of Zn. B.) At extreme concentrations of 15.3 µM Zn and 8.9 µM Cd, the uptake of Zn is reduced 58% in the presence of Cd, but Cd is unaffected by the presence of Zn. Bars represent uptake rates (mean + s.e.m., n= 8) of Zn (open bars) and Cd (closed bars). (* = p < 0.05).

Zn Uptake Cd Uptake

@ 31.1 M Ca @ 31.1 M Ca A 1.0 B 12 Total accumulation Total accumulation Post-EDTA Rinse Post EDTA Rinse Absorbed (internalized) Absorbed (internalized)

wet wt) wet

wet wt) wet -1 -1 0.5

(nmol Cd g Cd (nmol

(nmol Zn g Zn (nmol Cd Tissue Concentration Tissue Cd Zn Tissue Concentration Tissue Zn 0 0.0 0.00 0.24 1.95 19.80 0.00 0.24 1.95 19.80 Mn (M) Mn (M) C 12 @ 1.35 mM Ca 1.0 D @ 1.35 mM Ca

wet wt)

-1 -1 0.5

(nmol Zn g Zn (nmol

(nmol Cd g Cd (nmol Zn Tissue Concentration Tissue Zn 0 Concentration Tissue Cd 0.0 0.00 0.24 1.95 19.80 0.00 0.24 1.95 19.80 Mn (M) Mn (M)

E 0.20 F 2.5

2.0 0.15

1.5 wet wt.)wet

wet wt.) wet 0.10

* -1 -1 1.0

0.05

(nmol g (nmol

(nmol g (nmol Total absorbed Zn absorbed Total 0.5 Cd absorbed Total

0.0 0.00

M Ca M Ca  

@ 1.35 mM Ca @ 31.1 @ 31.1 @ 1.35 mM Ca

Figure 3. A-D). The influence of Mn and Ca on Cd and Zn uptake in H. sparna. Bars represent the acquisition (24 hour mean ± s.e.m.) of radiotracer (newly acquired metal). Open bars represent total metal accumulation (adsorbed and absorbed) (n=10). Five individuals per treatment group were rinsed with 0.05 M EDTA and re-assayed (grey bars). Five individuals were rinsed with a reducing agent (0.1 M ascorbate) followed by 0.05M EDTA and re-assayed (black bars). We interpret the black bars as representing absorbed metal. The difference between grey and black bars represents metals associated with oxide phases on the body surface. E). Mean absorbed Zn (± s.e.m.) in response to low and high Ca concentrations after pooling data from all Mn treatments. A 43-fold increase in Ca decreased Zn uptake by 31% . F.) Mean absorbed Cd (± s.e.m.) in response to low and high Ca concentrations after pooling data from all Mn treatments. Cd uptake was not decreased significantly. (* = p < 0.05)

0.24 M Mn A 100 a Total accumulation b Post-EDTA Rinse d Post-EDTA+Ascorbate Rinse e

wet wt.) 10 -1 c

(nmol g (nmol f

Mn Tissue Concentration Tissue Mn 1 31.1 M Ca 1.35 mM Ca

B 1.95 M Mn 1000 a d b e

wet wt.) 100 c -1

f (nmol g (nmol

Mn Tissue Concentration Tissue Mn 10 31.1 M Ca 1.35 mM Ca

C 19.8 M Mn 10000 a c bd d b

wet wt.) 1000 -1

e (nmol g (nmol

Mn Tissue Concentration Tissue Mn 100 31.1 M Ca 1.35 mM Ca

Figure 4. A-C). The influence of Ca on Mn uptake at three Mn concentrations in H. sparna. Bars represent the acquisition (24 hour mean ± s.e.m.) of radiotracer (newly acquired metal). Open bars represent total metal accumulation (adsorbed and absorbed) (n=10). Five individuals per treatment group were rinsed with 0.05 M EDTA and re-assayed (grey bars). Five individuals were rinsed with a reducing agent (0.1 M ascorbate) followed by 0.05M EDTA and re-assayed (black bars). We interpret the black bars as representing absorbed metal. The difference between grey and black bars represents metals associated with oxide phases on the body surface. Different letters signify significant differences (p < 0.05).

) 40

-1 hr

-1 30

20 b

(nmol Ca g Ca (nmol 10 Ca Absorption Rate Absorption Ca c 0

M RR M RR Control   10 100

Figure 5. The influence of Ca-ATPase inhibitor ruthenium red (RR) on Ca influx rate in H. sparna (n=8). At 10 µM RR, Ca influx was reduced by 53%, and 100 µM RR reduced Ca influx by 93.4%. Bars represent mean + s.e.m. Different letters signify significant differences (p < 0.05).

80 A 4 B Control Control 60 10 M RR 3 10 M RR

wet wt.) 100 M RR 100 M RR

wet wt.) -1 40 -1 2

20 1

(nmol Zn g Zn (nmol (nmol Cd g (nmol

Total Tissue Concentration Tissue Total 0 Total Tissue Concentration Tissue Total 0 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 Time (days) Time (days)

30 C 1.5 D

20 1.0 wet wt.)

wet wt.) wet *

-1 -1

10 0.5 * Total Zn Uptake Zn Total

* Uptake Cd Total

(nmol Zn g Zn (nmol (nmol Cd g Cd (nmol 0 0.0

M RR M RR M RR    M RR Control Control  10 10 100 100

Figure 6. A-B) Total accumulation rates (adsorbed and absorbed) of Zn (panel A) and Cd (panel B) are reduced by ruthenium red (RR) in a concentration dependent fashion (mean ± s.e.m.; n = 8). C) Ruthenium red decreased Zn uptake by 28% at 10 µM and 80% at 100 µM. D) Ruthenium red failed to decrease Cd uptake significantly at 10 µM but decreased uptake by 71% at 100 µM. Bars represent mean + s.e.m. An asterisk (*) represents a significant difference from the control treatment (p<0.05).

CHAPTER 2

Calcium uptake in aquatic insects: Influences of phylogeny and metals (Cd and Zn)

Monica D. Poteat, and David B. Buchwalter

Environmental and Molecular Toxicology, Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695, United States

Published In: Journal of Experimental Biology. 2014; Volume 217; pages 1180-1186.

Abstract

Calcium sequestration in the hypo-osmotic freshwater environment is imperative in maintaining calcium homeostasis in freshwater aquatic organisms. This uptake process is reported to have the unintended consequence of potentially toxic heavy metal (Cd, Zn) uptake in a variety of aquatic species. However, calcium uptake remains poorly understood in aquatic insects, the dominant invertebrate faunal group on most freshwater ecosystems.

Here we examined Ca uptake and interactions with heavy metals (Cd, Zn) at low ambient Ca levels (12.5 µmol l-1) in 12 aquatic insect species within Ephemerellidae and

Hydropsychidae, two families differentially responsive to trace metal pollution. We found Ca uptake varied 70-fold across the 12 species studied. Body weight and clade (family) were found to significantly influence both Ca uptake and adsorption (p < 0.05). Zn and Cd uptake rate constants (kus) exhibited a strong correlation (r = 0.96, p < 0.0001), suggesting a shared transport system. Ca uptake failed to significantly correlate with either Zn or Cd kus. Further, neither Zn nor Cd exhibited inhibitory effects toward Ca uptake. In fact we saw evidence of modest stimulation of Ca uptake rates in some metal treatments. This work suggests that insects generally differ from other freshwater taxa in that aqueous Ca uptake does not appear to be compromised by Cd or Zn exposure. It is important to understand the trace metal and major ion physiology of aquatic insects due to their ecological importance and widespread use as ecological indicators.

Introduction

Calcium homeostasis in living organisms plays a crucial role in several fundamental physiological processes including the maintenance of appropriate cellular and tissue permeability, stability of intracellular matrices, and neurological and muscular activity.1 A major component of Ca homeostasis involves acquiring adequate Ca from the environment.

For freshwater organisms, the acquisition of Ca from the external environment often involves the movement of ionic Ca against concentration gradients by specialized ionocytes containing transport systems that are better characterized in some groups (e.g. fish)2–4 than others (e.g. aquatic insects).5,6

Aquatic insects dominate the invertebrate species pool in freshwater ecosystems and largely differ from their crustacean, annelidan, molluscan and piscean co-habitants in that they are direct descendents of terrestrial ancestors7 rather than of more proximate marine origin.8–10 Insect invasions of freshwater habitats are hypothesized to have occurred several times over evolutionary history,7 resulting in modern aquatic insect communities composed of groups that have been aquatic for varying lengths of time and that may have solved the problem of successful ionoregulation in freshwater habitats in different ways.11 We work under the assumption that the ancestral strategy of aquatic insect progenitors involved exclusively dietary Ca acquisition, with centralized osmoregulatory functions in the gut, malpighian tubules and rectum.12 We hypothesize that multiple successful freshwater invasions have resulted in groups that have differentially relocated portions of their Ca trafficking function to the outer body surface. Such differences are manifested as diverse

rates of Ca influx from the surrounding water via specialized structures such as chloride cells

(Ephemeroptera), chloride epithelia, or anal papillae (Trichoptera).13

One unintended consequence of apical Ca sequestration directly from the surrounding water is that some of these transport systems appear to also transport other ions, including potentially toxic metals. Studies with fish,14–17 crustaceans18,19 and mollusks20 show evidence of shared uptake pathways between Ca and heavy metals Cd and Zn, with the potentially toxic metals competing with Ca for uptake. Data for aquatic insects is less clear however.

Here we assess the variability of Ca uptake among several members of two common aquatic insect families described as responding differently to trace metal in field studies.21,22

We look for evidence of shared transport pathways between Ca, Cd and Zn by correlating their uptake rates. We further assess whether environmentally relevant concentrations of either dissolved Cd or Zn show evidence of Ca uptake inhibition (as seen in other aquatic species). Throughout this paper, the term “uptake” refers only to absorbed (internalized) metals whereas “accumulation” refers to both absorbed and adsorbed (on body surface) metal in sum.

Materials and Methods

Aquatic insect collection and acclimation

All aquatic insect larvae were collected using a D-frame kicknet from Ca poor (~25

μmol l-1 Ca) streams in Great Smoky Mountains National Park within North Carolina and

Tennessee. Collecting focused specifically on two species-rich families, Ephemerellidae

(Order: Ephemeroptera) and Hydropsychidae (Order: Trichoptera) in order to explore Ca, Zn and Cd uptake and interactions among close relatives. Larvae were transported back to the laboratory in coolers with aerated stream water and cobble substrate. Acclimation occurred for a minimum of 48 hours in a walk-in cold room where all experimentation occurred

(12.7°C, 12h:12h light:dark photoperiod). If kept in the lab for extended periods of time (> 1 week), insects were fed natural periphyton biofilms (Stroud Water Research Center,

Avondale, PA) and fasted for at least 24 hours before experimentation.

Larvae were acclimated to one of two experimental waters depending on the experiment for which larvae were utilized. Waters consisted of American Society for Testing

-1 and Materials (ASTM) very soft water (VSW) (µmol l : 145 NaHCO3; 62.3 MgSO4; 6.71

-1 KCl) with either VSW Ca concentrations (43.6 µmol l CaSO4•2H2O) or low Ca

-1 concentrations (12.5 µmol l CaSO4•2H2O, “extra soft water, ESW”). VSW was used when measuring 65Zn and 109Cd uptake whereas ESW was used in experiments measuring 45Ca uptake. Voucher specimens for each species were verified by an independent taxonomist.

Only larvae that appeared healthy were used in experimentation.

Radioactivity measurement

45 45 The β-emitting isotope Ca was obtained as CaCl2 in H2O, (Perkin-Elmer,

45 Billerica, MA, USA) and diluted in 0.1 N HNO3 to make a working stock solution. Ca was used to examine Ca uptake and Ca adsorption to body surfaces. Exposure solutions ranged from 122 to 174 Bq l-1 in all experiments. All samples (water, EDTA and larval digests) were counted in 20 mL scintillation vials containing 16 mL Scintisafe® liquid scintillation cocktail

(LSC) on a Beckman LS6500 Multipurpose Scintillation Counter. Water samples and EDTA rinsate sample volumes were 1 mL and 2 mL, respectively. Larvae were digested individually in 20 mL scintillation vials with 1.5 mL Soluene® for 2-3 days at room temperature before the addition of LSC and subsequent radioactivity counting. Each individual 45Ca sample was counted for 10 minutes to ensure error and lumex values were <

5%.

The uptake of Zn and Cd were measured in larvae using the γ-emitting isotopes 65Zn

109 65 65 and Cd. Zn (as ZnCl2 in HCl) was obtained from Perkin-Elmer (Billerica, MA, USA) and 109Cd was obtained from Los Alamos National Laboratories (Los Alamos, NM, USA).

Both γ-isotopes were diluted in 0.1 N HNO3 to make working stock solutions. Protocols for counting both 65Zn and 109Cd simultaneously were established with spillover corrections and verified against single isotope samples. All samples (water and larvae) were counted in 20 mL scintillation vials using a Perkin-Elmer Wallac Wizard 1480 Automatic Gamma Counter.

Water samples (1 mL) were counted to verify radiotracer concentrations. Because in vivo γ- counting is non-destructive, individual larvae were counted multiple times throughout an

experiment with 15 mL VSW in the scintillation vial. All samples were counted for 3 minutes to ensure counting errors of < 5%.

Calcium uptake experiments

First, Ca time course experiments were conducted in two species, Ephemerella invaria and Hydropsyche alhedra, to establish a suitable time point for use in further comparative studies. The goal was to identify a time point within the initial linear phase of the uptake curve while achieving sufficient activity to minimize counting error. Bulk solutions consisted of ESW (Ca concentration: 12.5 µmol l-1) and 45Ca tracer, with the pH of each solution adjusted to 7.20 ± 0.02 using 0.1 N NaOH. Because β-emitting isotope analysis is destructive to the sample, individual larvae were designated for analysis at each of 5-6 time points spanning 48 hours (E.invaria: n=10; H. alhedra: n=7). Each replicate consisted of a single larva with 40 mL of solution in an aerated high-density polyethylene (HDPE) cup containing a small square of Teflon mesh as substrate and Parafilm® to reduce evaporative loss.

At each time point (E. invaria: 3, 6, 9, 12, 24, 47 hours; H. alhedra: 6, 9, 12.5, 24, 48 hours) larvae were removed from solution and rinsed with ESW for at least 3 minutes. In order to chelate and remove Ca adsorbed to the surface of the animals, each larva was bathed for 30 seconds in individual scintillation vials containing 2 mL 0.05 mol l-1 EDTA (see

Poteat et al., 2012). Larvae then were blotted dry, weighed and digested in 1.5 mL Soluene® before adding LSC and counting.

Determination of Ca uptake and adsorption

Ca uptake and body surface adsorption was measured in 12 species of aquatic insects, 7 species from family Ephemerellidae (Teleganopsis deficiens, Ephemerella invaria,

Ephemerella catawba, Ephemerella hispida, Drunella walkeri, Drunella cornutella,

Eurylophella verisimilis) and 5 species from family Hydropsychidae (Parapsyche cardis,

Arctopsyche irrorata, Diplectrona modesta, Hydropsyche sparna, Hydropsyche alhedra).

Bulk solutions consisted of ESW with 45Ca (122 to 174 Bq l-1), and the pH of each solution was adjusted to 7.20 ± 0.02 with the addition of 0.1 N NaOH. Ca concentrations before the addition of radiotracer were determined via ICP-MS at the Environmental and Agricultural

Testing Services Laboratory (Department of Soil Sciences, North Carolina State University,

Raleigh, NC, USA) and were within 2.5% of the target concentration of 12.5 µmol Ca. For each species, 5-10 replicates were used, each consisting of a single larva in an aerated HDPE beaker with 40 mL solution, Teflon mesh as substrate and Parafilm® to minimize evaporative loss. Following 6 hours of exposure, larvae were rinsed with ESW for at least 3 minutes and transferred to individual 20 mL scintillation vials containing 2 mL 0.05 mol l-1 EDTA for an additional 30 seconds to remove superficially adsorbed Ca. Finally insects were removed, rinsed with ESW, blotted dry, weighed fresh and digested as described above. EDTA rinsate sample counts were interpreted as adsorbed Ca whereas insect digest sample counts were interpreted as internalized (absorbed) Ca.

Determination of Zn and Cd uptake rate constants

23 Zn and Cd uptake rate constants (kus) (see ref. ) were determined for the same 12 species as above from time course experiments using γ-emitting isotopes 65Zn and 109Cd jointly in dual label exposures. Previous work testing single vs. dual metal exposures showed no interactions in uptake between Zn or Cd in three species (H. sparna, D. cornutella,

D.tuberculata) within the concentration ranges used here (Poteat, unpublished data). Bulk solutions containing environmentally-relevant concentrations of 45.9 nmol l-1 Zn and 2.67 nmol l-1 Cd, 138 nmol l-1 Zn and 8.01 nmol l-1 Cd, and 413 nmol l-1 Zn and 24.0 nmol l-1 Cd were prepared to ensure identical treatments of replicates within each species. 65Zn activities ranged from 104 - 207 Bq l-1 and 109Cd activities ranged from 30 - 59 Bq l-1 in all exposure solutions with stable Zn (as ZnCl2) and Cd (as CdCl2) comprising the majority of metals in solution. The pH of each solution was adjusted to 7.20 ± 0.02 with 0.1 N NaOH. Each replicate consisted of a single larva within a HDPE beaker with 80 mL solution, Teflon mesh as substrate and Parafilm® to reduce evaporative loss. For each of the three dual metal solutions of Zn and Cd, there were between 5 and 10 replicates for each of the 12 species.

Larvae were exposed to dissolved concentrations for a total of 9 hours. At 3, 6 and 9 hours, each larva was removed from solution, rinsed with VSW, assayed in vivo for radioactivity and returned to the exposure. After larvae were assayed at the last time point, each larva was blotted dry and wet weights were obtained. Within each species, uptake rates were determined at each concentration measured for Zn and Cd, and the slope of the line of uptakes rates versus concentration at which the uptake rate was derived was taken as the uptake rate constant (ku).

Assessing the effects of Zn and Cd on Ca uptake

To assess the effects that dissolved Zn or Cd had on Ca uptake, we measured 45Ca uptake in the presence of Cd (0.89, 8.90, 89.0 nmol Cd l-1) or Zn (0.0153, 0.153, 1.53 µmol

Zn l-1) (individually) in 4 species (2 ephemerellids and 2 hydropsychids). Results were compared to control 45Ca uptake rates. In these experiments, Cd and Zn concentrations were chosen to test a range from environmentally relevant concentrations up to extreme concentrations. For each species, bulk solutions were made for each treatment containing

45 ESW, Ca tracer and either stable Zn (as ZnCl2) or Cd (as CdCl2) as needed. The highest concentrations of Cd and Zn (89.0 nmol l-1 and 1.53 µmol l-1, respectively) were verified by

ICP-MS analyses; however the lower concentrations were below the limit of detection and could not be verified. Within each species, each treatment had 6-10 replicates, each consisting of a single larva in an aerated HDPE cup with 40 mL solution, Teflon mesh as substrate, and Parafilm® to reduce evaporative loss.

Larvae were exposed to treatment solutions for 6 hours. When removed, larvae were rinsed with ESW for at least 3 minutes, rinsed with 0.05 mol l-1 EDTA for 30 seconds in individual 20 mL scintillation vials, blotted dry, weighed and digested in 1.5 mL Soluene® before adding LSC and counting for radioactivity.

Data analysis

Data analysis was performed using GraphPad Prism (v5.04) and SigmaPlot.

Allometry and multiple linear regression analyses were performed on log-transformed data; other analyses were performed on raw data. Multiple linear regression was used to determine

the effects of body weight and clade (family) on Ca uptake rates across species. Student’s t- tests were performed to determine the effects of dissolved Zn and Cd on Ca uptake relative to controls. Results were considered significant when p < 0.05.

Results

Calcium uptake and adsorption across aquatic insect species

Under identical water chemistry conditions, E. invaria and H. alhedra vary considerably in their uptake of Ca (Fig. 1). Slopes of the initial uptake rates (time 0 through

12 hours) were 10.9 fold faster in E. invaria (9.07 ± 0.08 nmol Ca g-1 hr-1) than in H. alhedra

(0.83 ± 0.16 nmol Ca g-1 hr-1) (p < 0.0001). Based on this experiment, a 6 hour time point was chosen for subsequent comparative 45Ca experiments in other taxa because it fell within the initial linear phase of the uptake curve while ensuring sufficient activity to minimize counting error.

Across 12 species, 6-hour Ca uptake (absorption) varied over 70-fold (from 3.6 to

253.2 nmol Ca g-1 tissue in T. deficiens and P. cardis, respectively) (Fig. 2A). Ca uptake varied 18.9 fold among ephemerellids (13.4 to 253.2 nmol Ca g-1), and 6.1-fold among hydropsychids (3.6 to 21.9 nmol Ca g-1). On average, Ca absorption was 9.4-fold faster in ephemerellids (80.9 ± 29.6 nmol Ca g-1) than in hydropsychids (8.59 ± 7.7 nmol Ca g-1).

The analysis of EDTA rinsates revealed that in each species tested, more Ca was found adsorbed to the external surfaces of the larvae than was internalized (Fig. 2B).

Adsorbed Ca ranged from 65 % (P. cardis) to 91% (D. modesta) of the total Ca load acquired. When considered on an adsorbed Ca per mass basis, ephemerellids adsorbed 9.2- fold more Ca than hydropsychids (426.6 vs 46.3 nmol Ca g-1, respectively). However, on a percent of total Ca adsorbed basis, ephemerellids only adsorbed 1.2-fold more Ca than hydropsychids.

Effects of allometry and clade

Both body weight and family significantly affected Ca uptake rates in the aquatic insects we tested. Analysis of log-transformed Ca uptake rates against log-transformed weights across species revealed a strong correlation between body weight and Ca uptake (r =

-0.89, p < 0.0001) (Fig. 3). Further, multiple linear regression analysis revealed that both weight (p < 0.0001) and family (p < 0.0001) significantly affected Ca uptake rates in these aquatic insect families (adjusted r2 = 0.942) (Eq. 1).

, (1)

-1 -1 where y is the log-transformed Ca uptake rate (nmol g hr ), is the log-transformed wet weight, and is family, where “0” represents ephemerellids and “1” represents hydropsychids.

Body weight (p = 0.002) and family (p = 0.006) also significantly affected Ca adsorption rates (adjusted r2 = 0.86) (Eq. 2).

, (2)

-1 where y is the amount of Ca adsorbed after a 6 hour exposure (nmol g ), is the log- transformed wet weight, and is family, where “0” represents ephemerellids and “1” represents hydropsychids. Overall, both Ca uptake and adsorption were heavily influenced by body weight and family.

Comparisons of Ca, Zn and Cd uptake

Cd and Zn kus were strongly correlated across all 12 species of aquatic insects tested

(r = 0.96, p < 0.0001) (Fig. 4). Zn kus ranged 91-fold and Cd kus ranged 108-fold across all species. The linear regression of Zn kus against Cd kus was described as:

, (3) with the slope suggesting that the uptake of Cd is faster than the uptake of Zn across species.

Both Cd (r = -0.79, p = 0.002) and Zn (r = -0.77, p = 0.004) kus were significantly correlated with species’ body weight (data not shown).

Ca uptake rates failed to correlate significantly with either Zn (r = 0.49, p = 0.10;

Fig. 5A) or Cd (r = 0.54, p = 0.072; Fig. 5B) kus across all 12 species. To assess the potentially confounding influence of body weight on this correlation, we removed 6 species from the analysis that had statistically different body weights in Ca experiments as compared to Cd/Zn experiments. Again, the uptake rates of Ca failed to correlate significantly with either Zn (r = 0.57, p = 0.23) or Cd (r = 0.57, p = 0.24) kus across the remaining 6 species with statistically identical weights between experiments (Fig 5A,B). Across these same 6 species with identical weights across Ca and Cd/Zn experiments, Cd and Zn kus were still strongly correlated (r = 0.93, p = 0.0081). In sum, Cd and Zn uptake is tightly correlated across species whereas Ca uptake fails to correlate significantly with either Cd or Zn uptake.

Effects of Zn and Cd on Ca uptake

Across 4 aquatic insect species, neither Cd nor Zn inhibited Ca uptake over a 6 hour period (Fig. 6A-D). Conversely, in E. invaria, small amounts of Cd or Zn actually stimulated

Ca uptake (Fig. 6A). The lowest Cd exposure concentration (0.89 nmol Cd l-1) elicited an increase of 61% in newly acquired Ca, while 0.0153 and 0.153 μmol Zn l-1 elicited increases of 75% and 60% in newly acquired Ca, respectively. Neither Zn nor Cd had any other significant effects on the uptake of newly acquired Ca in other species, however both E. catawba (Fig. 6B) and D. modesta (Fig. 6D) exhibited trends similar to those seem in E. invaria. Taken together, these results suggest that neither Cd nor Zn have significant affinity for the calcium transport systems used by these insects under low Ca conditions.

Discussion

As a diverse faunal group of over 6500 species,24 aquatic insects typically account for

70-95% of the invertebrate species in freshwater ecosystems.24,25 The dominance of aquatic insects in these ecosystems and species’ differential responsiveness to environmental stressors has led to their extensive use as ecological indicators worldwide.26,27 Here, we compared Ca, Cd and Zn uptake (and their interactions) in Hydropsychidae and

Ephemerellidae— two common families in freshwaters worldwide thought to differ in their metal sensitivity. This work represents the first attempt to comparatively study Ca uptake and interactions with Zn and Cd in aquatic insects.

Analyzing the Ca physiologies of species within the families Ephemerellidae and

Hydropsychidae allowed us to examine the extent to which variability exists in ionoregulatory processes within close relatives. In the comparative examination of Ca uptake among these 12 species, we found Ca uptake to vary across 2 orders of magnitude, ranging from 0.6 to 42.2 nmol Ca g-1 hr-1 at very low ambient Ca concentration (12.5 µmol l-1 Ca).

Body size strongly influenced Ca uptake rates, with faster Ca uptake rates observed for smaller organisms. However, even after accounting for body weight, there were clear differences between families, with ephemerellids exhibiting faster Ca uptake than hydropsychids.

Few Ca uptake studies using aquatic insects exist with which to compare our results.

We were only able to find Ca influx data for 2 dipteran species (order: Diptera). Mosquito larvae (Aedes aegypti) were reported to have Ca uptake rates of 0.0335 nmol Ca hr-1 larva-1

(body weights were not reported) at a concentration of 0.1 mM Ca28 and Chironomus riparus

has a maximum Ca uptake rate of 0.388 µmol Ca g-1 hr-1 at concentrations up to 2.58 mmol l-

1.29 It is difficult to quantitatively compare these studies to our work because the influence of adsorbed Ca was likely unaccounted for in previous studies. In our study, adsorption proved to be an important consideration in Ca accumulation for all species examined within this study, accounting for between 65 and 91% of all Ca accumulated through a 6 hour time course. Rinsing larvae with water for > 3 minutes did not remove adsorbed 45Ca in our study, but EDTA rinses removed significant 45Ca. It is clear that Ca uptake in all of the species we examined is considerably faster than reported values for dipterans.

In our study, body weight significantly contributed to differences both in Ca uptake rates and uptake rate constants for Cd and Zn. This finding is consistent with other studies of

Ca4,15,30 and metal11,31,32 uptake. In insects, species within a single family utilize the same specialized structures for ion transport (e.g., chloride cells, chloride epithelia, anal papillae), therefore the increased surface area: mass ratios present in smaller species would produce the higher Ca uptake rates observed.

We also show that in addition to body size, clade (specifically family) was a significant factor in determining Ca uptake rates within aquatic insects. This finding is consistent with other studies11,33 that demonstrate a strong phylogenetic basis for differences across species in ion (metal) transport processes. Our data also suggest that the

Ephemerellidae is a more physiologically variable group than is the Hydropsychidae. The use of phylogenetic frameworks for better understanding and predicting species performance and responses to environmental stressors is still in its infancy34–36 and represents a potentially powerful approach for overcoming the inherent limitations that experimentalists face when

attempting to apply findings from a limited number of species to the tremendous biodiversity that exists in nature.

The strong correlation of Cd and Zn uptake seen in our study is suggestive of a shared transport system for these metals. Across the literature, the correlation between Zn and Cd uptake is seen in other aquatic organisms, specifically in mussels37 and crustaceans.38

Previously work also showed Cd and Zn uptake in H. sparna to mirror each other across a wide range of concentrations (Zn: 0.0153 to 15.2 µmol l-1; Cd: 0.0089 to 8.9 µmol l-1), with

Cd apparently outcompeting Zn for uptake when concentrations of each were sufficiently

5 high (>0.6 uM). While Zn and Cd kus have been shown to correlate within species, the strength of the correlations across aquatic insect species, specifically closely-related species, is novel.

In contrast, the lack of significant correlation of Ca uptake rates with Cd or Zn kus was somewhat surprising given the robust literature that suggests shared transport in other taxa. Further, the lack of competition between either Cd or Zn with Ca uptake was also surprising in light of other studies in fish,39–45 daphnids,46 mollusks20 and crabs19 where these metals are described as competing with Ca for transport. Conversely, there is evidence of a lack of competition between Ca and Cd/Zn also seen in mussels37,47 and crabs.48

Within aquatic insects, evidence for shared transport sites for Ca and Cd/Zn is equivocal with only a few studies available in the literature. We previously reported a modest protective (inhibitory) effect of elevated Ca (1.35 mmol l-1) on Cd and Zn uptake.5 Similarly, increasing Ca content from 178 to 712 µmol l-1 inhibited Cd uptake by only 13% in Drunella flavilinea, and increasing Ca content from 178 µmol l-1 to 1.42 mmol l-1 inhibited (albeit

modestly) both Zn and Cd uptake in Hydropsyche californica by 36% and 34%, respectively.49 Craig et al. found that 1 mmol l-1 Ca (10x the control concentration) inhibited

Cd uptake in Chironomus staegeri by 46%.50 However, in Chironomus riparius, Gillis and

Wood found that Cd uptake was not significantly inhibited by Ca until the Ca concentration was increased to 2.41 mmol l-1,29 and they were unable to determine if this was a result of competitive or non-competitive interactions.

Results from pharmacological channel blockers and inhibitors are also equivocal. We previously demonstrated similarly decreasing trends in Ca, Cd and Zn uptake by H. sparna when exposed to the Ca-ATPase inhibitor Ruthenium Red. Presumably, this compound either blocked a specific transporter used by the ions or altered the transmembrane potential of the water-organism interface.5 Verapamil, nifedipine (both L-Type Ca channel blockers) and carboxyeosin (plasma membrane Ca-ATPase) were all unsuccessful in blocking Ca, Cd or Zn uptake from solution. The influx of Cd and/or Zn was also unaffected by the Ca channel blocker verapamil in the aquatic insect H. californica,49 however verapamil blocked metal uptake in Chironomus staegeri50 and D. flavilinea.49

To our knowledge, only this study (Ephemeroptera and Trichoptera) and the Gillis and Wood study (Diptera)29 examined the effects of Cd and/or Zn on Ca influx in aquatic insects. This limited dataset suggests that aquatic insects differ from other freshwater taxa in that neither Cd nor Zn appears to out-compete Ca for transport. Yet, all three ions are readily taken up. One potential explanation for this finding is that Ca transporters in insects are more selective than those in other taxa. A second explanation is that different transport systems are used for Ca than those used by Cd and Zn (with the latter using a specialized Zn transport

systems perhaps). A third possibility is that because Ca homeostasis is so important, complimentary (or redundant) Ca transport systems exist to ensure that if one system is compromised (for example, via metal exposure) another can compensate. It is possible that the apparent stimulation of Ca uptake in the presence of Cd or Zn that we observed represents such a compensatory response, though much more research would be required to validate such a proposition.

Any of the above explanations raise the possibility that ionoregulatory disturbance may not be the proximal mechanism of Cd and/or Zn toxicity in this important faunal group.

We note that aquatic insects generally do not exhibit toxic responses to acute Cd or Zn dissolved exposures at environmentally relevant concentrations (but see ref. 51) whereas fish and daphnids are responsive via ionoregulatory disturbance. Gillis and Wood ascribed the Cd tolerance of C. riparius to the lack of Ca influx inhibition.29 Interestingly, we see the same lack of metal induced Ca influx inhibition in both hydropsychids (generally described as metal tolerant) and ephemerellids, which are among the first groups to disappear in metal contaminated systems.22

With their dominance of freshwater ecosystems and widespread use of aquatic insects as indicators of stream health, it is imperative to understand the fundamental physiological differences between species in traits which account for sensitivity. By examining the physiologies of close relatives, it may become possible to predict the physiological performance of closely-related (or distantly related) species by taking into account fundamental characteristics such as phylogeny and body weight.

Acknowledgements

We would like to thank Eric Fleek (North Carolina Department of Environmental and

Natural Resources) and Luke Jacobus (Indiana University – Purdue University Columbus) for their taxonomic expertise. Gerald LeBlanc (North Carolina State University), Tom

Augspurger (US Fish and Wildlife Service), Justin Conley (North Carolina State University) and anonymous reviewers provided valuable editorial comments. This research was supported by the National Science Foundation (NSF) [IOS 0919614].

References

(1) Clark, E. W. A review of literature on calcium and magnesium in insects. Ann. Entomol. Soc. Am. 1958, 51, 142–154.

(2) Flik, G.; Van Der Velden, J. A.; Dechering, K. J.; Verbost, P. M.; Schoenmakers, T. J. M.; Kolar, Z. I.; E, W. B. S. Ca2+ and Mg2+ transport in gills and gut of tilapia, Oreochromis mossambicus: A review. J. Exp. Zool. 1993, 265, 356-365.

(3) Perry, S. F.; Flik, G. Characterization of branchial transepithelial calcium fluxes in freshwater trout, Salmo gairdneri. Am. J. Physiol. 1988, 254, R491–R498.

(4) Perry, S. F.; Wood, C. M. Kinetics of branchial calcium uptake in the rainbow trout: Effects of acclimation to various external calcium levels. J. Exp. Biol. 1985, 116, 411– 433.

(5) Poteat, M. D.; Díaz-Jaramillo, M.; Buchwalter, D. B. Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna. J. Exp. Biol. 2012, 215, 1575–1583.

(6) Taylor, C. W. Calcium regulation in insects. Adv. In Insect Phys. 1986, 19, 155–186.

(7) Kristensen, N. P. Phylogeny of insect orders. Annu. Rev. Entomol. 1981, 26, 135–157.

(8) Anger, K. The conquest of freshwater and land by marine crabs: Adaptations in life- history patterns and larval bioenergetics. J. Exp. Mar. Biol. Ecol. 1995, 193, 119–145.

(9) Lee, C. E.; Bell, M. A. Causes and consequences of recent freshwater invasions by saltwater animals. Trends Ecol. Evol. 1999, 14, 284–288.

(10) Lee, C. E. Rapid and repeated invasions of fresh water by the copepod Eurytemora affinis. Evolution. 1999, 53, 1423–1434.

(11) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.; Garland, Jr., T. Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. P. Natl. Acad. Sci. USA. 2008, 105, 8321–8326.

(12) Stobbart, R. H.; Shaw, J. Salt and water balance: Excretion. In The Physiology of Insects; Rothstein, M., Ed.; Academic Press: New York, 1974.

(13) Komnick, H. Chloride cells and chloride epithelia of aquatic insects. Int. Rev. Cytol. 1977, 49, 285–329.

(14) Chang, M.-H.; Lin, H.-C.; Hwang, P. P. Effects of cadmium on the kinetics of calcium uptake in developing tilapia larvae, Oreochromis mossambicus. Fish Physiol. Biochem. 1997, 16, 459–470.

(15) Hogstrand, C.; Wilson, R. W.; Polgar, D.; Wood, C. M. Effects of zinc on the kinetics of branchial calcium uptake in freshwater rainbow trout during adaptation to waterborne zinc. J. Exp. Biol. 1994, 186, 55–73.

(16) McCormick, S. D.; Hasegawa, S.; Hirano, T. Calcium uptake in the skin of a freshwater teleost. P. Natl. Aca . Sci. USA. 1992, 89, 3635–3638.

(17) Wicklund, A.; Runn, P. Calcium effects on cadmium uptake, redistribution, and elimination in minnows, Phoxinus phoxinus, acclimated to different calcium concentrations. Aquat. Toxicol. 1988, 13, 109–121.

(18) Tan, Q. G.; Wang, W. X. Acute toxicity of cadmium in Daphnia magna under different calcium and pH Conditions: Importance of influx rate. Env. Sci. Technol. 2011, 45, 1970–1976.

(19) Wright, D. A. The effect of calcium on cadmium uptake by the shore crab Carcinus maenus. J. Exp. Biol. 1977, 67, 163–174.

(20) Bjerregaard, P.; Depledge, M. H. Cadmium accumulation in Littorina littorea, Mytilus edulis and Carcinus maenas: The influence of salinity and calcium ion concentrations. Mar. Biol. 1994, 119, 385–395.

(21) Cain, D. J.; Luoma, S. N.; Wallace, W. G. Linking metal bioaccumulation of aquatic insects to their distribution patterns in a mining-impacted river. Environ. Toxicol. Chem. 2004, 23, 1463–1473.

(22) Clements, W. H.; Carlisle, D. M.; Lazorchak, J. M.; Johnson, P. C. Heavy metals structure benthic communities in Colorado mountain streams. Ecol. Appl. 2000, 10, 626–638.

(23) Buchwalter, D. B.; Cain, D. J.; Clements, W. H.; Luoma, S. N. Using biodynamic models to reconcile differences between laboratory toxicity tests and field biomonitoring with aquatic insects. Env. Sci. Technol. 2007, 41, 4821–4828.

(24) Merritt, R. W.; Cummins, K. W.; Berg, M. B., Eds. An Introduction to the Aquatic Insects of North America, 4th, ed.; Kendall-Hunt: Dubuque, IA, 2008.

(25) Arscott, D. B.; Jackson, J. K.; Kratzer, E. B. Role of rarity and taxonomic resolution in a regional and spatial analysis of stream macroinvertebrates. J. N. Am. Benthol. Soc. 2006, 25, 977–997.

(26) Hawkins, C. P.; Norris, R. H.; Hogue, J. N.; Feminella, J. W. Development and evaluation of predictive models for measuring the biological integrity of streams. Ecol. Appl. 2000, 10, 1456–1477.

(27) Wallace, J. B.; Webster, J. R. The role of macroinvertebrates in stream ecosystem function. Annu. Rev. Entomol. 1996, 41, 115–139.

(28) Barkai, A. I.; Williams, R. W. The exchange of calcium in larvae of the mosquito Aedes aegypti. J. Exp. Biol. 1983, 104, 139–148.

(29) Gillis, P. L.; Wood, C. M. The effect of extreme waterborne cadmium exposure on the internal concentrations of cadmium, calcium, and sodium in Chironomus riparius larvae. Ecotoxicol. Env. Saf. 2008, 71, 56–64.

(30) Hogstrand, C.; Webb, N.; Wood, C. M. Covariation in regulation of affinity for branchial zinc and calcium uptake in freshwater rainbow trout. J. Exp. Biol. 1998, 201, 1809–1815.

(31) Wang, W.-X.; Fisher, N. S. Modeling the influence of body size on trace element accumulation in the mussel Mytilus edulis. Mar. Ecol.- Prog. Ser. 1997, 161, 103–115.

(32) Zhang, L.; Wang, W.-X. Size-dependence of the potential for metal biomagnification in early life stages of marine fish. Env. Toxicol. Chem. 2007, 26, 787–794.

(33) Poteat, M. D.; Garland, Jr., T.; Fisher, N. S.; Wang, W.-X.; Buchwalter, D. B. Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms. Env. Sci. Technol. 2013, 47, 7989–7995.

(34) Carew, M. E.; Miller, A. D.; Hoffmann, A. A. Phylogenetic signals and ecotoxicological responses: Potential implications for aquatic biomonitoring. Ecotoxicology. 2011, 20, 595–606.

(35) Guenard, G.; von der Ohe, P. C.; de Zwart, D.; Legendre, P.; Lek, S. Using phylogenetic information to predict species tolerances to toxic chemicals. Ecol. Appl. 2011, 21, 3178–3190.

(36) Hammond, J. I.; Jones, D. K.; Stephens, P. R.; Relyea, R. A. Phylogeny meets ecotoxicology: Evolutionary patterns of sensitivity to a common insecticide. Evol. Appl. 2012, 5, 593–606.

(37) Wang, W.-X.; Fisher, N. S. Effects of calcium and metabolic inhibitors on trace element uptake in two marine bivalves. J. Exp. Mar. Biol. Ecol. 1999, 236, 149–164.

(38) Rainbow, P. S. Physiology, Physicochemistry and metal uptake - A crustacean perspective. Mar. Pollut. Bull. 1995, 31, 55–59.

(39) Barron, M. G.; Albeke, S. Calcium control of zinc uptake in rainbow trout. Aquat. Toxicol. 2000, 50, 257–264.

(40) Hogstrand, C.; Verbost, P. M.; Bonga, S. E. W.; Wood, C. M. Mechanisms of zinc uptake in gills of freshwater rainbow trout: Interplay with calcium transport. Am. J. Physiol.- Reg. I. 1996, 39, R1141–R1147.

(41) Hollis, L.; McGeer, J. C.; McDonald, D. G.; Wood, C. M. Protective effects of calcium against chronic waterborne cadmium exposure to juvenile rainbow trout. Env. Toxicol. Chem. 2000, 19, 2725–2734.

(42) Ojo, A. A.; Wood, C. M. In vitro characterization of cadmium and zinc uptake via the gastro-intestinal tract of the rainbow trout (Oncorhynchus mykiss): Interactive effects and the influence of calcium. Aquat. Toxicol. 2008, 89, 55–64.

(43) Spry, D. J.; Wood, C. M. Ion flux rates, acid-base status, and blood gases in rainbow trout, Salmo gairdneri, exposed to toxic zinc in natural soft water. Can. J. Fish. Aquat. Sci. 1985, 42, 1332–1341.

(44) Spry, D. J.; Wood, C. M. A kinetic method for the measurement of zinc influx in vivo in the rainbow trout, and the effects of waterborne calcium on flux rates. J. Exp. Biol. 1989, 142, 425–446.

(45) Wright, D. A. Trace metal and major ion interactions in aquatic animals. Mar. Pollut. Bull. 1995, 31, 8–18.

(46) Tan, Q.-G.; Wang, W.-X. The influence of ambient and body calcium on cadmium and zinc accumulation in Daphnia magna. Env. Toxicol. Chem. 2008, 27, 1605–1613.

(47) Vercauteren, K.; Blust, R. Uptake of cadmium and zinc by the mussel Mytilus edulis and inhibition by calcium channel and metabolic blockers. Mar. Biol. 1999, 135, 615– 626.

(48) Rainbow, P. S.; Black, W. H. Physicochemistry or physiology: Cadmium uptake and effects of salinity and osmolality in three crabs of different ecologies. Mar. Ecol.- Prog. Ser. 2005, 286, 217–229.

(49) Buchwalter, D. B.; Luoma, S. N. Differences in dissolved cadmium and zinc uptake among stream insects: Mechanistic explanations. Env. Sci. Technol. 2005, 39, 498– 504.

(50) Craig, A.; Hare, L.; Tessier, A. Experimental evidence for cadmium uptake via calcium channels in the aquatic insect Chironomus staegeri. Aquat. Toxicol. 1999, 44, 255–262.

(51) Xie, L.; Buchwalter, D. B. Cadmium exposure route affects antioxidant responses in the mayfly Centroptilum triangulifer. Aquat. Toxicol. 2011, 105, 199–205.

Figures

4 0 0

E p h e m e r e lla in v a r ia

)

a t

C H y d ro p sy c h e a lh e d ra

4 3 0 0

u g

2 0 0

l o

w 1 0 0

( N

0 0 1 0 2 0 3 0 4 0 5 0 T im e (H o u r s )

Figure 1) Time course of 45Ca uptake in two aquatic insects, E. invaria and H. alhedra. Each time point consisted of either 7 (H. alhedra) or 10 (E. invaria) individuals. Symbols are mean ± s.d.

1 0 0 0 A ) A b s o r b e d

t a

h A d s o r b e d

) e

(

t 1 0 0

e d

i 1

g q

c a

a 1 0

l l

w m

( 1

1 2 3 4 5 6 7 8 9 0 1 2 d B 1 1 1

e 1 0 0 B

b r

o s

d E p h e m e r e llid a e H y d r o p sy c h id a e A

s v

d 5 0

e b

s b

a C

0 % s a a a i a s s a a a a n i d r ll li i t t n r r b i e i d a s r d ie a w te r r e v p lk m a o a e ic a s u i r d p h f in t i a n s c r o l a h r i . i s a e . c . w r . . d E . o e P m . . . E c v ) A . H H ) E ) D . ) ) T b ) . (h i D k l) ) ( ) d e D E ( ) ( ( c ( ( (j (a ( f) ) ( (g

Figure 2) Comparisons of newly acquired 45Ca in 12 species of aquatic insects, both absorbed and adsorbed. (A) Bars represent mean absorption and adsorption after a 6 hour exposure. (B) Average percentages of absorption and adsorption that accounted for total Ca accumulation.

)

1 -

r 6 0 h E p h e m e r e llid a e

1 4 0 a -

g H y d r o p s y c h id a e

a 2 0

o 1 5

g m

n (

c 1 0 f b

e e

k a t 5

p k j d

U l i h

a 0

C 0 1 0 2 0 3 0 4 0 W e t W e ig h t (m g )

Figure 3) Ca uptake rates of 12 aquatic insect species versus the average body weight of individuals. Letters represent the names of species defined in Fig. 2. Symbols are mean ± s.d., (n = 5-10).

E p h e m e r e llid a e

1 .0 H y d r o p s y c h id a e

) 1 -

d b

- g

c L

( 0 .5 g f

u k

n d

Z e r = 0 .9 6 l i j h k p < 0 .0 0 0 1 0 .0 0 .0 0 .5 1 .0

- 1 - 1 C d k u (L g d )

Figure 4) Relationship between Zn and Cd kus for 12 aquatic insect species. Letters are the names of species defined in Figure 2. Symbols are mean ± s.d.

)

1 -

r 6 0 6 0 E p h e m e r e llid a e

h A a B a

1 4 0 4 0 H y d r o p sy c h id a e -

a 2 0

l 2 0

o 1 5 g m

n g

( c b

1 0 f c b

e e f 1 0 e

k a

t 5 k p h d

i j d h l k U

l i j

a 0 0 C 0 .0 0 .2 0 .4 0 .6 0 .8 1 .0 0 .0 0 .5 1 .0 - 1 - 1 Z n k (L g d ) - 1 - 1 u C d k u (L g d )

Figure 5) Relationship between Ca uptake rates and (A) Zn or (B) Cd kus for 12 aquatic insect species. Black symbols represent species for which there was no significant difference in weight between the Ca and metal uptake experiments, gray symbols represent species for which there was a significant difference in weight between experiments. Letters are the names of species defined in Fig. 2. Symbols are mean ± s.d.

)

s s 1 5 0 A ) E . in v a r ia 1 0 0 C ) H . s p a r n a a

m *

t 8 0

e *

1 0 0 w

6 0

- g

4 0 l 5 0 o

2 0 m

( 0

0 a

C 1 5 0 1 5

) B ) E . c a ta w b a D ) D . m o d e s ta

h 6

(

d 1 0 0 1 0

e b

r o

s 5 0 5 b

y l

0 0 w

e l d d d n n n l d d d n n n

o C C C 1 Z 1 Z 1 Z o C C C 1 Z 1 Z 1 Z tr -1 -1 -1 - - - tr -1 -1 -1 - - -

N n l l l l l l n l l l l l l o l l l l l l o l l l l l l C o o o o o o C o o o o o o m m m m m m m m m m m m n n n    n n n    9 9 9 3 3 3 9 9 9 3 3 3 8 . 8 5 5 5 8 . 8 5 5 5 . 8 1 1 . . 8 1 1 . 0 . 1 0 . 1 .0 0 .0 0 0 0

Figure 6) The effect of either dissolved Cd or Zn on Ca uptake across 4 species, 2 ephemerellids (A,B) and 2 hydropsychids (C,D). Bars represent mean ± s.d. Each treatment within each species had n = 8-10.

CHAPTER 3

Phylogeny and size differentially influence dissolved Cd and Zn bioaccumulation parameters among closely related aquatic insects

Monica D. Poteat and David B. Buchwalter

Environmental and Molecular Toxicology Program, Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695, United States

Published in: Environmental Science and Technology. 2014; Volume 48; pages 5274-5281.

Abstract

Evolutionarily distinct lineages can vary markedly in their accumulation of, and sensitivity to, contaminants. However, less is known about variability among closely related species. Here, we compared dissolved Cd and Zn bioaccumulation in 19 species spanning 2 species-rich aquatic insect families - Ephemerellidae (order Ephemeroptera (mayflies)), generalized to be metal sensitive, and Hydropsychidae (order Trichoptera (caddisflies)), generalized to be metal tolerant. Across all species, Zn and Cd uptake rate constants (kus), efflux rate constants (kes) and bioconcentration factors (BCFs) strongly co-varied, suggesting that these metals share transport pathways in these distinct lineages. Kus and BCFs were substantially larger in Ephemerellidae than in Hydropsychidae, whereas kes did not dramatically differ between the two families. Body size played an important role in driving ku differences among species, but had no influence on kes. While familial differences in metal bioconcentration were striking, each family exhibited tremendous variability in all bioaccumulation parameters. At finer levels of taxonomic resolution (within families), phylogeny did not account for differences in metal bioaccumulation. These findings suggest that intra-family variability can be profound and have important practical implications in that we need to better understand how well “surrogate species” represent their fellow congeners and family members.

Introduction

Aquatic insects are the most dominant invertebrate faunal group in freshwater ecosystems, often accounting for between 75-90% of the invertebrate species pool.1,2 They are also an inherently diverse faunal group, with global biodiversity estimates of 100,000 species3 stemming from multiple freshwater invasions by terrestrial ancestors throughout evolutionary history.4 The sheer dominance of aquatic insects in freshwater systems coupled with their differential responses to pollutants has led to their widespread use in biomonitoring and bioassessment programs worldwide.

While biomonitoring and bioassessment programs take advantage of the variable pollution responsiveness of these species, our understanding of the physiological drivers of these sensitivity differences remains poorly understood. Further, our understanding of physiological variability among species (and species groups) remains surprisingly limited.

Similarly, regulatory toxicology programs are reliant on the widespread use of surrogate species to represent larger groups of species without a clear understanding of how much variability might occur at broader levels of biological organization. For example, the requirement to have only a single aquatic insect represented in toxicity datasets used to establish water quality criteria in the United States5 functionally means that a single species often represents the entire class (~6,500 North American species6).

We can begin to understand variability across species by examining physiological/toxicological traits in a comparative context.7–10 Previous work analyzed the uptake and efflux of Cd across 21 aquatic insect species representing orders Ephemeroptera,

Plecoptera and Trichoptera (EPT) and found metal fluxes to vary drastically across the three

orders.7 Importantly, variability across species was largely driven by phylogeny, with insects of the same taxonomic lineage generally having more similar metal flux parameters than insects from different taxonomic lineages.7 Thus, while highly variable, metal bioaccumulation parameters in these aquatic insects were heavily influenced by evolutionary history and generally followed phylogeny.

Here we made an attempt to understand the variability of aquatic insect responses to trace metal pollution by examining dissolved Cd and Zn fluxes in closely related aquatic insect species. We examined Cd and Zn bioaccumulation specifically because they co-occur in ores in the earth’s crust and therefore commonly co-occur in metal contaminated environments. Both are borderline transition metals,11 though Zn is an essential element and typically more abundant than Cd, a nonessential element. Further, these ions have similar biochemical and physical properties.11–13

We measured Cd and Zn uptake and efflux rate constants (ku and ke, respectively) in

19 and 18 species, respectively, with species representing two common aquatic insect families previously suggested to be especially variable in Cd fluxes.7 The mayfly family

Ephemerellidae (order Ephemeroptera) is generalized to be metal sensitive whereas the caddisfly family Hydropsychidae (order Trichoptera) is generalized to be metal tolerant.14

We explored whether the trafficking patterns of Zn are similar to those of Cd across species tested, and assessed the influence of body weight on Cd and Zn flux parameters. Lastly, by examining the fluxes of these two metals among several closely related species, we asked if phylogeny could continue to explain interspecies variability within families.

Materials and Methods

Insect collection and handling

Larvae of 19 aquatic insect species spanning families Ephemerellidae and

Hydropsychidae were collected from Great Smoky Mountains National Park during the time period of July 2010 to August 2013. All larvae were collected using a D-frame kicknet from cool, cobble-bottomed streams and transported back to the laboratory at North Carolina State

University as described previously.15 Upon arrival, larvae were held under laboratory conditions for a minimum of 48 hours before experimentation.

Experiments were performed in a walk-in cold room with a controlled climate

(12.7°C, 12h:12h light:dark photoperiod). All experiments utilized American Society for

-1 Testing Materials (ASTM) very soft water (VSW) (mg L : 12 NaHCO3, 7.5 CaSO4·2H2O,

7.5 MgSO4, 0.5 KCl) because all insects were collected from streams with low ambient calcium (~1 mg L-1). Only larvae that appeared healthy were used for experimentation. All ephemerellid larval identifications were determined using morphological characters, and some identifications were confirmed using DNA barcoding when data were available.16

Hydropsychid larval identifications were performed using morphological characters.17

Radioactivity measurement

The γ-emitting isotopes 65Zn and 109Cd were used to measure metal fluxes across

65 65 109 109 insect species. Isotopes Zn (as ZnCl2 in HCl) and Cd (as CdCl2 in HCl) were diluted

109 in 0.1 N HNO3 for working stock solutions. Methods for the simultaneous counting of Cd and 65Zn were established and verified against single and dual standards. Protocols for

counting dual labeled exposures were utilized because previous work in ephemerellids

(Poteat, unpublished) and Hydropsyche sparna 15 showed no difference in the metal uptake or efflux of insects exposed to single or dual metal exposures at environmentally relevant concentrations. Water and larvae samples were counted in 20 mL scintillation vials using a

Perkin-Elmer Wallac Wizard 1480 Automatic Gamma Counter. The non-destructive nature of in vivo γ-counting allowed us to count individual larvae multiple times throughout experimental time periods in 15 mL VSW within scintillation vials. All samples were counted for 3 minutes to ensure low counting errors (< 5%).

Determination of uptake rate constants

Zn and Cd uptake rate constants (ku) were determined for 19 species of aquatic insects – 6 species of hydropsychids (Arctopsyche irrorata, Diplectrona modesta,

Hydropsyche alhedra, Hydropsyche slossonae, Hydropsyche sparna, Parapsyche cardis) and

13 species of ephemerellids (Dannella sp., Drunella cornutella, Drunella longicornis,

Drunella tuberculata, Drunella walkeri, Ephemerella catawba, Ephemerella crenula,

Ephemerella hispida, Ephemerella invaria, Ephemerella rossi, Ephemerella subvaria,

Eurylophella verisimilis, Teleganopsis deficiens) – in methods described previously.18

Briefly, three dual-labeled bulk solutions of the following environmentally-relevant Zn and

Cd concentrations were prepared for each species: 3 µg L-1 Zn and 0.3 µg L-1 Cd, 9 µg L-1 Zn and 0.9 µg L-1 Cd, and 27 µg L-1 Zn and 2.7 µg L-1 Cd. Solutions all contained between 104 -

207 Bq L-1 Zn and 30 - 59 Bq L-1 Cd, with the remaining dissolved metal being stable Zn (as

ZnCl2) and stable Cd (as CdCl2). Water samples (1 mL) were counted to verify radiotracer

concentrations. The pH of all solutions was adjusted to 7.2 ± 0.02 using 0.1 N NaOH. A total of 5-10 replicates were utilized per concentration level within each species, with each replicate consisting of a single larva within a high-density polyethylene (HDPE) cup containing 80 mL solution, Teflon mesh as substrate and Parafilm® to reduce evaporative loss. Therefore, 15 – 30 individual insect larvae were used to determine kus for each species.

Larvae were exposed to solutions for a total of 9 hours. After 3 and 6 hours of exposure, larvae were removed from the exposure, rinsed with VSW, assayed in vivo for radioactivity and returned to the exposure solution. After 9 hours of exposure, insects were removed from the exposure, rinsed with VSW, assayed for radioactivity, blotted dry and weighed to obtain the wet weight of individuals. Uptake rates over time were determined at each concentration measured for Zn and Cd. The slope of the line of uptakes rates versus the concentration at which the uptake rate was derived was taken as the uptake rate constant (ku) for each species.

Determination of efflux rate constants

Efflux rate constants (ke) were determined for all insects used in the above uptake experiments (except for T. deficiens, due to a lack of sufficient quantities collected) using methods described previously.7,19-21 Briefly, insect larvae from each species were exposed to a dual labeled solution containing 3 µg L-1 Zn and 0.3 µg L-1 Cd for a total of 4-5 days to ensure adequate uptake of radiolabel. The solutions contained 104 - 207 Bq L-1 Zn and 30 -

-1 59 Bq L Cd, and the remaining dissolved metal was stable Zn (as ZnCl2) and Cd (as CdCl2).

The pH of all solutions was adjusted with 0.1 N NaOH to 7.20 ± 0.02. Insects were exposed

individually in 80 mL aerated solution in HDPE beakers containing Teflon mesh as substrate and Parafilm® to reduce evaporative loss.

After 4-5 days, insects were removed from the exposure solution, rinsed with VSW and assayed to ensure the adequate uptake of radiolabel. Each larva (5-10 replicates per species) was then placed in an individual 1 L HDPE container with 500 mL aerated VSW,

Teflon mesh as substrate and Parafilm® to reduce evaporative loss. Larvae were assayed in vivo daily for 10 days, and solutions were checked for radioactivity to ensure that there was no appreciable metal available for re-uptake. After larvae were assayed on day 10, wet weights were obtained.

Efflux rate constants were determined as the slope of the natural log of the proportion of metal retained in the body tissue and the time of depuration22 after excluding days 0 and 1 to minimize the confounding influence of desorbed metals as follows:

Where = tissue concentration at time (day) t

= initial tissue concentration

-1 = efflux rate constant (d )

t = time (days).

Determination of bioconcentration factors

Bioconcentration factors were determined for 18 aquatic insect species for which we were able to obtain both kus and kes as follows:

100

Where BCF = bioconcentration factor

-1 -1 = uptake rate constant (L g d )

-1 = efflux rate constant (d ).

Mean metal BCFs were determined using mean kus and kes for each species. The minimum BCF estimate (see Table 1) was determined using the mean metal ku - 1 s.d. and the mean metal ke + 1 s.d. The maximum BCF estimate was determined using the mean metal ku + 1 s.d. and the mean metal ke - 1 s.d.

Data analysis

All conventional statistical analyses were performed using GraphPad Prism (v6.02) and SigmaPlot. Student t-tests were used to determine differences in traits between families.

Multiple linear regressions (performed on log-transformed data) were used to determine the effects of body weight and clade (family) on metal bioaccumulation parameters. Results were considered significant when p < 0.05.

Phylogenetic statistical methods were used to determine whether metal bioaccumulation parameters measured in this study were influenced by the evolutionary history of the aquatic insects. Relevant phylogenies were constructed using the available literature for ephemerellids23 and hydropsychids.24,25 The K-statistic was used to quantify phylogenetic signal (the tendency for more closely related species to resemble each other26), and the randomization test based on the mean-squared error was used to test for the

101

significance of the phylogenetic signal. Log-transformed traits were corrected allometrically before phylogenetic analyses.

102

Results

Metal bioaccumulation parameters across metals and families

Cd and Zn values for each trait (kus, kes, BCFs) strongly co-varied across all species in this study. Zn and Cd kus strongly co-varied across 19 species (r = 0.95, p < 0.0001) (Fig.

1a). Zn and Cd kes strongly co-varied across 18 species (r = 0.89, p < 0.0001) (Fig. 1b). Zn and Cd BCFs also strongly co-varied across the 18 species (r = 0.86, p < 0.0001) (Fig. 2).

Zn and Cd kus showed remarkable variation in the aquatic insect species analyzed

-1 -1 (Table 1). Across all 19 species in this study, Zn kus ranged 0.01 to 1.58 L g d and Cd kus

-1 -1 ranged 0.01 to 1.96 L g d . Of 19 insect species, 8 had comparable Zn and Cd kus (within a standard deviation of each other’s values). Ten species had higher Cd kus than Zn kus, and only one species had higher Zn kus than Cd kus. A regression of Zn vs. Cd kus (Fig. 1a) had a slope of 0.84 ± 0.06 (r2 = 0.91, p < 0.0001).

Zn and Cd kes were also highly variable across the species examined (Table 1). Zn kes

-1 -1 ranged from 0.013 to 0.18 d and Cd kes ranged 0.010 to 0.17 d . Of 18 insect species, 10 had comparable Zn and Cd kes (within a standard deviation of each other’s values). Three species had higher Cd kes than Zn kes, and 5 had higher Zn kes than Cd kes. A regression of

2 Zn vs. Cd kes (Fig. 1b) had a slope of 0.71 ± 0.09 (r = 0.80, p < 0.0001).

Mean Zn and Cd BCFs for 18 species within families Hydropsychidae and

Ephemerellidae ranged over 3 orders of magnitude for each metal (Table 1). Mean Zn BCFs ranged 265 – 56,089, and mean Cd BCFs ranged 235 – 88,423. Of 18 insect species, 13 had comparable Zn and Cd BCFs (within each other’s minimum – maximum BCF range seen in

Table 1). Five species had higher Cd BCFs than Zn BCFs, and no species had higher Zn

103

BCFs than Cd BCFs. A regression of Zn vs. Cd BCFs (Fig. 2) had a slope of 0.50 ± 0.07 (r2

= 0.74, p < 0.0001).

Variability of metal bioaccumulation in Ephemerellidae and Hydropsychidae

Zn and Cd kus were noticeably different between the two families observed here

-1 -1 (Table 1). Hydropsychid species (n = 6) had Zn kus ranging from 0.010 – 0.17 L g d and

-1 -1 Cd kus ranging from 0.01 – 0.37 L g d . Ephemerellid species (n = 13) had Zn kus ranging

-1 -1 -1 -1 from 0.15 – 1.58 L g d and Cd kus ranging from 0.12 – 1.96 L g d . Overall, hydropsychid species had significantly lower Zn (p = 0.003) and Cd (p = 0.01) kus than did ephemerellid species.

Contrary to the pattern present in kus, there was no clear familial grouping in the ranges of ke values present (Table 1). Zn and Cd ke values for hydropsychids were nested within the range of values for ephemerellids. Hydropsychid values ranged 0.028 – 0.13 d-1

-1 across Zn kes and 0.042 – 0.17 d across Cd kes. Ephemerellid species had Zn kes ranging

-1 -1 0.013 – 0.18 d and Cd kes ranging 0.010 – 0.19 d .

The range of BCF values for both metals overlapped, with hydropsychids occupying a lower range for Cd (235 – 4,297) and Zn (265 – 2,375). Ranges in ephemerellid BCFs were substantially larger for Cd (1,053 – 88,423) and Zn (1,071 – 56,089). Overlapping ranges in the two families rendered mean family differences marginally significant for Cd (p = 0.08) and Zn (p = 0.06).

104

Effects of allometry and family on metal bioaccumulation

Log-transformed Zn kus were negatively correlated with log-transformed body weight

(r = -0.75, p = 0.0002) (Fig. 3a). Further, multiple linear regression analysis found both body weight (p = 0.05) and family (p = 0.002) to be significant predictors of Zn kus. Log- transformed Cd kus were negatively correlated with log-transformed body weight (r = -0.71, p = 0.0007) (Fig. 3b). Multiple linear regression analysis found that both body weight (p =

0.06) and family (p = 0.07) could be biologically important determinants of Cd kus. Unlike kus, neither body weight nor family significantly affected either Zn or Cd kes (p > 0.05) (Fig.

S1).

Log-transformed Zn BCFs were negatively correlated with log-transformed body weight (r = -0.65, p = 0.0035) (Fig. 4a). Multiple linear regression analysis found family to be an important predictor of Zn BCF values (p = 0.03), but body weight was not a significant predictor of Zn BCF values (p = 0.21). Log-transformed Cd BCFs were negatively correlated with log-transformed body weight (r = -0.65, p = 0.0037) (Fig. 4b). Multiple linear regression revealed neither body weight (p = 0.12) nor family (p = 0.16) were significant predictors of

Cd BCFs.

Phylogenetic analyses of metal bioaccumulation parameters

While familial differences in Zn and Cd kus and BCFs were apparent, the variance seen within families was not explained by evolutionary history. Across Zn and Cd kus, kes, and BCFs, no K-statistic was found to be statistically significant. This signified that all traits

105

analyzed failed to exhibit phylogenetic signal after accounting for the effects of body weight

(p > 0.05) within the species tested in this study (for example, Fig. 5).

This lack of phylogenetic influence over metal bioaccumulation parameters can be partially explained by the very large variation seen in metal bioaccumulation within genera, specifically in the family Ephemerellidae. Species within the ephemerellid genus Drunella (n

-1 -1 -1 -1 = 4) had Zn kus ranging 0.21 – 1.58 L g d and Cd kus ranging 0.28 – 1.96 L g d , values which cover almost the entire range of kus for ephemerellid species. Species within the ephemerellid genus Ephemerella (n = 6) also had highly variable kus, with Zn kus ranging

-1 -1 -1 -1 0.26 – 0.91 L g d and Cd kus ranging 0.27 – 0.98 L g d (Table 1).

Zn and Cd kes were also especially variable in ephemerellid genera. Across the genus

-1 -1 Drunella, Zn kes ranged 0.05 – 0.11 d and Cd kes ranged 0.01 – 0.19 d . Interestingly, the

Drunella Cd kes spanned the entire range of values for ephemerellids. The genus

-1 Ephemerella also had a wide range of ke values, with Zn kes ranging 0.031 – 0.18 d and Cd

-1 kes ranging 0.019 – 0.18 d (Table 1).

Zn and Cd BCFs were variable across all genera for which we were able to measure multiple species. Across the hydropsychid genus Hydropsyche (n = 3), Cd BCFs ranged 256

– 4,297 and Zn BCFs ranged 265 – 1,139. The genus Drunella had BCFs spanning almost the entire ephemerellid range, with Cd BCFs spanning 1,486 – 88,423 and Zn BCFs spanning

1,966 – 56,089. Across the genus Ephemerella, BCFs were also quite variable, with Cd BCFs ranging 1,053 – 41,764 and Zn BCFs ranging 1,524 – 19,140 (Table 1).

106

Discussion

One objective of this study was to compare Cd and Zn bioaccumulation parameters in two species-rich aquatic insect families known to be especially variable in metal fluxes.7 We focused on the dissolved route of exposure to expediently make comparisons across several taxa, though we acknowledge that dietary routes of exposure often drive bioaccumulation differences among species.20,27 Evidence for shared dissolved uptake pathways of Cd and Zn is found in a range of freshwater fauna including aquatic insects,15,18,28 mussels 29 and crustaceans,30 though other studies have shown a lack of evidence for shared uptake pathways across aquatic taxa.31,32 The strong co-variation observed in Zn and Cd fluxes is suggestive of shared transport systems in aquatic insects, however at environmentally relevant concentrations, we have no evidence that they compete. Strong co-variances in Zn and Cd efflux have been observed across aquatic organisms spanning mollusks to fish to

21 arthropods. In fact, the co-variation between Cd and Zn kes in aquatic insects was so strong that Cd kes from 11 ephemerellid and hydropsychid species successfully predicted Zn kes

21 from known Cd kes in 5 of 6 EPT taxa using a simple linear regression.

This is the first study to our knowledge to examine Cd and Zn BCFs across aquatic insect species. Here we used a time independent (steady state) estimate of BCF based on rate constants of uptake and elimination to avoid common pitfalls associated with the use of this term. This BCF estimate is useful to compare the relative tendencies of these species to accumulate Cd and Zn exclusively from solution. While we found Zn and Cd BCFs to co- vary, we found evidence to suggest that Cd generally bioaccumulates more readily than Zn.

Previous work in aquatic insects has shown Cd to have a higher affinity for shared co-

107

15 transporters of Zn. This could be a reason we generally observed higher Cd kus than Zn kus

(and higher Cd BCFs than Zn BCFs).

A second objective of this study was to compare metal bioaccumulation parameters in two families described as differentially responsive to metal pollution. Ephemerellid mayfly species are commonly observed to be more sensitive to metal contamination in the field, as they account for some of the first species to disappear at contaminated sites.14 Conversely, hydropsychid caddisfly species are able to persist in metal contaminated environments,14 withstanding higher concentrations of dissolved metals.33 Here we see that ephemerellids had elevated uptake of dissolved metals as compared to hydropsychids, leading to higher BCFs.

Ephemerellid species are especially known for their comparatively high uptake of metals in relation to other aquatic insects in the laboratory.27,28 We suspect that the ephemerellid tendency to strongly bioconcentrate metals contributes to their observed metal sensitivity in the field.34 However other important considerations in metal sensitivities such as diet27,35 and the ability to detoxify metals7 were not taken into account in this study.

Body weight appeared to influence Zn and Cd kus, and to a lesser extent BCFs, in

Ephemerellidae and Hydropsychidae. Allometric scaling is common in physiological traits across aquatic species, including the dissolved metal uptake of freshwater organisms.7,36,37 In this study, ephemerellids were generally smaller larvae and tended to bioaccumulate more metal from dissolved exposures than the larger caddisflies.

A third objective of this study was to examine the physiological variation occurring among closely related species. While familial differences in some metal bioaccumulation parameters were apparent within our dataset, phylogenetic patterns previously observed

108

across EPT orders7 appeared to break down at the genus/species level of taxonomic identification. Evolutionary patterns in comparative datasets are well documented across a range of physiological traits38 including metal and ion transport traits.7,21 However, most if not all of these datasets include species from a wide range of families, orders and even phyla.

In toxicology, it is often assumed that closely related species, particularly ones within the same genus or family, are physiologically similar. Aquatic insect families and genera known to be tolerant of metal pollution have recently come into use as a way to monitor bioavailable metals as well as predict community effects in polluted freshwater systems.39–42

In particular, total metal body burdens of species of the hydropsychid genus Hydropsyche are often used in this regard due to the taxon’s persistence and ability to reflect metal bioavailabilities in contaminated environments.40,41 However, the inherently variable nature of some species groups (including Hydropsyche) can potentially influence the results of studies which utilize coarser family- or genus- level identifications of test species and should be considered.

In general, measuring Cd and Zn bioaccumulation parameters in both ephemerellid mayflies and hydropsychid caddisflies gave us a unique opportunity to observe patterns in the physiological variability of two species-rich families with radically different evolutionary histories. The mayfly family Ephemerellidae has an estimated 75 species in North America43 and is descended from one of the oldest aquatic insect lineages which has been aquatic for up to an estimated 400 million years.44 The caddisfly family Hydropsychidae has an estimated

184 species in North America45 and is descended from a lineage which has been aquatic for up to 234 million years.46 These two families are descendants of different terrestrial

109

ancestors4 which used unique physiological methods to overcome the challenges to living in freshwater, thus enabling each lineage to develop differing physiologies which potentially lead to varying metal bioaccumulation parameters.

To our knowledge, this study is the first attempt to comparatively examine the physiologies of closely related aquatic insect species of the same families and genera.

Whether variation in metal bioaccumulation across species is caused by underlying evolutionary patterns or differences in body size, it is important to acknowledge the extreme physiological variances seen across even closely related species. Biomonitoring and bioassessment programs often use measures of biodiversity as important endpoints upon which regulatory decisions are made, and such biodiversity measures are highly dependent on taxonomic resolution (e.g., species-, genus-, family-level identifications). It is important to recognize that not all species within a given family (or even genus) will physiologically resemble each other, and even species within the same genera can vary drastically in their metal flux physiologies.

110

Acknowledgements

We thank Luke Jacobus (Indiana University – Purdue University Columbus) and Eric

Fleek (North Carolina Department of Environmental and Natural Resources) for their aid in insect species identification. Allison Camp (NCSU), Justin Conley (NCSU), Gerald LeBlanc

(NCSU), and anonymous reviewers provided valuable editorial comments. This work was supported by NSF (IOS 0919614), the ICA Chris Lee Award for Metals Research and the

Society of Environmental Toxicology and Chemistry (SETAC). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the ICA or

SETAC.

111

References

(2) Minshall, G. W. Community structure in natural stream systems. In The Stream Ecosystem. Boston: AAAS Symposium; 1969.

(3) Dijkstra, K.-D. B.; Monaghan, M. T.; Pauls, S. U. Freshwater biodiversity and aquatic insect diversification. Annu. Rev. Entomol. 2014, 59, 143–163.

(4) Bradley, T. J.; Briscoe, A. D.; Brady, S. G.; Contreras, H. L.; Danforth, B. N.; Dudley, R.; Grimaldi, D.; Harrison, J. F.; Kaiser, J. A.; Merlin, C.; et al. Episodes in insect evolution. Integ. Comp. Biol. 2009, 49, 590–606.

(5) Stephan, C. E.; Mount, D. I.; Hansen, D. J.; Gentile, J. R.; Chapman, G. A.; Brungs, W. A. Guidelines for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. United States Environmental Protection Agency: Washington, D.C., 1985.

(6) Merritt, R. W.; Cummins, K. W.; Berg, M. B., Eds. An Introduction to the Aquatic Insects of North America, 4th, ed.; Kendall-Hunt: Dubuque, IA, 2008.

(7) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.; Garland, Jr., T. Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. P. Natl. Acad. Sci. USA 2008, 105, 8321–8326.

(8) Hammond, J. I.; Jones, D. K.; Stephens, P. R.; Relyea, R. A. Phylogeny meets ecotoxicology: Evolutionary patterns of sensitivity to a common insecticide. Evol. Appl. 2012, 5, 593–606.

(9) Guenard, G.; von der Ohe, P. C.; de Zwart, D.; Legendre, P.; Lek, S. Using phylogenetic information to predict species tolerances to toxic chemicals. Ecol. Appl. 2011, 21, 3178–3190.

(10) Carew, M. E.; Miller, A. D.; Hoffmann, A. A. Phylogenetic signals and ecotoxicological responses: Potential implications for aquatic biomonitoring. Ecotoxicology 2011, 20, 595–606.

(11) Nieboer, E.; Richardson, D. H. S. The replacement of the non-descript term “heavy metals” by a biologically and chemically significant classification of metal ions. Env. Pollut. B. 1980, 1, 3–26.

112

(12) Newman, M. C.; McCloskey, J. T. Predicting relative toxicity and interactions of divalent metal ions: Microtox(R) bioluminescence assay. Env. Toxicol. Chem. 1996, 15, 275–281.

(13) Allred, A. L. Electronegativity values from thermochemical data. J. Inorg. Nucl. Chem. 1961, 17, 215–221.

(14) Clements, W. H.; Carlisle, D. M.; Lazorchak, J. M.; Johnson, P. C. Heavy metals structure benthic communities in Colorado mountain streams. Ecol. Appl. 2000, 10, 626–638.

(15) Poteat, M. D.; Díaz-Jaramillo, M.; Buchwalter, D. B. Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna. J. Exp. Biol. 2012, 215, 1575–1583.

(16) Webb, J. M.,; Jacobus, L. M.; Funk, D. H.; Zhou, X.; Kondratieff, B.; Geraci, C. J.; DeWalt, R. E.; Baird, D.; Richard, B.; Phillips, I.; Herbert, P. D. N. A DNA barcode library for North American Ephemeroptera: Progress and prospects. PLoS One 2012. 7 (5), e38063.

(17) Beatty, S. R. Taxonomy document with standard taxonomic efforts levels for Trichoptera. Biological Assessment Unit, Division of Water Quality, NC Department of Environmental and Natural Resources.

(18) Poteat, M.; Buchwalter, D. Calcium uptake in aquatic insects: Influences of phylogeny and metals (Cd and Zn). J. Exp. Biol. 2014, 217, 1180-1186.

(19) Buchwalter, D. B.; Cain, D. J.; Clements, W. H.; Luoma, S. N. Using biodynamic models to reconcile differences between laboratory toxicity tests and field biomonitoring with aquatic insects. Env. Sci. Technol. 2007, 41, 4821–4828.

(20) Martin, C. A.; Cain, D. J.; Luoma, S. N.; Buchwalter, D. B. Cadmium ecophysiology in seven stonefly (Plecoptera) species: Delineating sources and estimating susceptibility. Env. Sci. Technol. 2007, 41, 7171–7177.

(21) Poteat, M. D.; Garland, Jr., T.; Fisher, N. S.; Wang, W.-X.; Buchwalter, D. B. Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms. Env. Sci Technol. 2013, 47, 7989–7995.

(22) Blackmore, G.; Wang, W.-X. Uptake and efflux of Cd and Zn by the green mussel Perna viridis after metal preexposure. Env. Sci. Technol. 2002, 36, 989–995.

113

(23) Ogden, T. H.; Osborne, J. T.; Jacobus, L. M.; Whiting, M. F. Combined molecular and morphological phylogeny of Ephemerellinae (Ephemerellidae: Ephemeroptera), with remarks about classification. Zootaxa 2009, 1991: 28-42.

(24) Geraci, C. J.; Al-Saffar, M. A.; Zhou, X. DNA barcoding facilitates description of unknown faunas: a case study on Trichoptera in the headwaters of the Tigris River, Iraq. J. N. Am. Benthol. Soc. 2011, 30 (1), 163-173.

(25) Zhou, X. Robinson, J. L.; Geraci, C. J.; Parker, C. R.; Flint Jr., O. S.; Etnier, D. A.; Ruiter, D.; DeWalt, R. E.; Jacobus, L. M.; Herbert, P. D. N. Accelerated reconstruction of a regional DNA-barcode reference library: caddisflies (Trichoptera) in the Great Smoky Mountains National Park. BioOne 2011, 30 (1), 131-162.

(26) Blomberg, S. P.; Garland, Jr., T.; Ives, A. R. Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution 2003, 57, 717–745.

(27) Cain, D.; Croteau, M.-N.; Luoma, S. N. Bioaccumulation dynamics and exposure routes of Cd and Cu among species of aquatic mayflies. Env. Toxicol. Chem. 2011, 30, 2532–2541.

(28) Buchwalter, D. B.; Luoma, S. N. Differences in dissolved cadmium and zinc uptake among stream insects: Mechanistic explanations. Env. Sci. Technol. 2005, 39, 498– 504.

(29) Wang, W. X.; Fisher, N. S. Effects of calcium and metabolic inhibitors on trace element uptake in two marine bivalves. J. Exp. Mar. Biol. Ecol. 1999, 236, 149–164.

(30) Rainbow, P. S. Physiology, physicochemistry and metal uptake - A crustacean perspective. Mar. Pollut. Bull. 1995, 31, 55–59.

(31) Rainbow, P. S.; Amiard-Triquet, C.; Amiard, J. C.; Smith, B. D.; Langston, W. J. Observations on the interaction of zinc and cadmium uptake rates in crustaceans (amphipods and crabs) from coastal sites in UK and France differentially enriched with trace metals. Aquat. Toxicol. 2000, 50, 189–204.

(32) Wang, W.-X.; Fisher, N. S. Effects of calcium and metabolic inhibitors on trace element uptake in two marine bivalves. J. Exp. Mar. Biol. Ecol. 1999, 236, 149–164.

(33) Clements, W. H.; Cadmus, P.; Brinkman, S. F. Responses of aquatic insects to Cu and Zn in stream microcosms: Understanding differences between single species tests and field responses. Env. Sci. Technol. 2013, 47, 7506–7513.

114

(34) Poteat, M. D.; Buchwalter, D. B. Four reasons why traditional metal toxicity testing with aquatic insects is irrelevant. Env. Sci. Technol. 2014, 48, 877–888.

(35) Xie, L.; Buchwalter, D. B. Cadmium exposure route affects antioxidant responses in the mayfly Centroptilum triangulifer. Aquat. Toxicol. 2011, 105, 199–205.

(36) Wang, W.-X.; Fisher, N. S. Modeling the influence of body size on trace element accumulation in the mussel Mytilus edulis. Mar. Ecol.- Prog. Ser. 1997, 161, 103–115.

(37) Zhang, L.; Wang, W.-X. Size-dependence of the potential for metal biomagnification in early life stages of marine fish. Env. Toxicol. Chem. 2007, 26, 787–794.

(38) Garland, T.; Bennett, A. F.; Rezende, E. L. Phylogenetic approaches in comparative physiology. J. Exp. Biol. 2005, 208, 3015–3035.

(39) De Jonge, M.; Tipping, E.; Lofts, S.; Bervoets, L.; Blust, R. The use of invertebrate body burdens to predict ecological effects of metal mixtures in mining-impacted waters. Aquat. Toxicol. 2013, 142-143, 294-302.

(40) Cain, D. J.; Luoma, S. N.; Wallace, W. G. Linking metal bioaccumulation of aquatic insects to their distribution patterns in a mining-impacted river. Env. Sci. Tecnnol. 2004, 23, 1463–1473.

(41) Luoma, S. N.; Cain, D. J.; Rainbow, P. S. Calibrating biomonitors to ecological disturbance: A new technique for explaining metal effects in natural waters. Integ. Env. Assess. Manag. 2010, 6, 199–209.

(42) Schmidt, T. S.; Clements, W. H.; Zuellig, R. E.; Mitchell, K. a; Church, S. E.; Wanty, R. B.; San Juan, C. A.; Adams, M.; Lamothe, P. J. Critical tissue residue approach linking accumulated metals in aquatic insects to population and community-level effects. Env. Sci. Technol. 2011, 45, 7004–7010.

(43) Mayfly Central Website; http://www.entm.purdue.edu/mayfly/.

(44) Whitfield, J. B.; Kjer, K. M. Ancient rapid radiations of insects: Challenges for phylogenetic analysis. Annu. Rev. Entomol. 2008, 53, 449–472.

(45) Morse, J. C. A checklist of the Trichoptera of North America, including Greenland and Mexico. Trans. Am. Entomol. Soc. 1993, 119, 47–93.

(46) Malm, T.; Johanson, K. A.; Wahlberg, N. The evolutionary history of Trichoptera (Insecta): A case of successful adaptation to life in freshwater. Syst. Entomol. 2013, 38, 459–473.

115

(47) Pagel, M. D. A method for the analysis of comparative data. J. Theor. Biol. 1992, 156, 431–442.

(48) Huson, D. H.; Scornavacca, C. Dendroscope 3: An interactive viewer for rooted phylogenetic trees and networks. Syst. Biol. 2012, 61, 1061–1067.

116

Figures

2 0 .3

a j b

) 1

- 1 .0

i o

)

1 0 .2 - 1 q m

- n

g ( l

e g

L m l (

r k

u 0 .5 p h n d k q e

0 .1 a s Z c k n n j

Z o g h b c b k r = 0 .9 5 p r = 0 .8 9 a i f e f p < 0 .0 0 0 1 r p < 0 .0 0 0 1 0 .0 d 0 .0 0 .0 0 .5 1 .0 2 0 .0 0 0 .0 5 0 .1 0 0 .1 5 0 .2 0 0 .2 5 - 1 - 1 - 1 C d k u (L g d ) C d k e (d )

Figure 1) Relationship between Zn and Cd kus in 19 aquatic insect species (a) and Zn and Cd kes in 18 species (b). Symbols are mean ± s. d. Letters represent the names of species identified in Table 1.

117

1 0 0 0 0 0 r = 0 .8 6 i p < 0 .0 0 0 1 r j p o 1 0 0 0 0 h F q C l B k m

n b n Z 1 0 0 0 g e a f E p h e m e r e llid a e d c H y d r o p s y c h id a e 1 0 0 1 0 0 1 0 0 0 1 0 0 0 0 1 0 0 0 0 0 C d B C F

Figure 2) Relationship between Zn and Cd BCFs in 18 aquatic insect species. Letters represent the names of species identified in Table 1. The dotted line is a 1:1 line.

118

E p h e m e r e llid a e H y d r o p s y c h id a e 0 .5 a j i

0 .0 ) d q o

1 l m

e -

s p d m r h

-0 .5 r

1 n -

o k f g b

s -1 .0 L

n e (

a u r a

t d k -1 .5

n o

Z L -2 .0 r = - 0 .7 5 f p = 0 .0 0 0 2 -2 .5 0 .0 0 .5 1 .0 1 .5 2 .0 0 .5 b j o i m

0 .0

) d

1 h e - q s

l r

d m

-0 .5 p b r 1 n

- e

o k

s g d

L -1 .0

n (

a a u

r c

-1 .5

o C L -2 .0 r = - 0 .7 1 f p = 0 .0 0 0 7 -2 .5 0 .0 0 .5 1 .0 1 .5 2 .0 L o g -tr a n sfo r m e d m a ss (m g )

Figure 3) Log-transformed Zn (a) and Cd (b) kus of 19 species versus the average log- transformed body weight of individuals. Letters represent the names of species identified in Table 1.

119

5 F a i

C j

o r B

4 q h n

Z l m b d k n e 3 e g d a m f

r c o

f 2

s n

a r

t 1 E p h e m e r e llid a e - r = -0 .6 5 g H y d r o p s y c h id a e o p = 0 .0 0 3 5 L 0 0 .0 0 .5 1 .0 1 .5 2 .0

F 6

C b

B j i

d o r

C h

p 4 q e

d m

e l b n

m k g a

r d o

f c f

s 2

r t

- r = -0 .6 5

o p = 0 .0 0 3 7

L 0 0 .0 0 .5 1 .0 1 .5 2 .0 L o g -tr a n sfo r m e d m a ss (m g )

Figure 4) Log-transformed Zn (a) and Cd (b) BCFs of 18 species versus the average log- transformed body weight of individuals. Letters represent the names of species identified in Table 1.

120

Figure 5) Zn BCFs for 18 species plotted onto their phylogeny. Pagel’s arbitrary branch lengths47 are depicted. The phylogeny was constructed using Dendroscope.48

121

Table 1. Cd and Zn uptake rate constants (ku) for 19 aquatic insect species.

Mean body Genus species wt. (mg) Cd ku ± s.d. Zn ku ± s.d. Arctopsyche irrorata (a)* 60.7 0.070 ± 0.00087 0.046 ± 0.0018 Diplectrona modesta (b) 10.4 0.27 ± 0.019 0.17 0.018 Hydropsyche alhedra (c) 15.7 0.044 ± 0.015 0.029 ± 0.0043 Hydropsyche slossonae (d) 15.2 0.12 ± 0.015 0.042 ± 0.014 Hydropsyche sparna (e) 5.8 0.37 ± 0.026 0.095 ± 0.0031 Parapsyche cardis (f) 27.1 0.0098 ± 0.0040 0.010 ± 0.0040 Dannella sp. (g) 4.5 0.12 ± 0.030 0.15 ± 0.0060 Drunella cornutella (h) 8.0 0.74 ± 0.015 0.50 ± 0.022 Drunella longicornis (i) 6.4 0.78 ± 0.13 0.92 ± 0.12 Drunella tuberculata (j) 3.1 1.96 ± 0.068 1.58 ± 0.017 Drunella walkeri (k) 7.2 0.28 ± 0.029 0.21 ± 0.014 Ephemerella catawba (l) 5.0 0.77 ± 0.24 0.58 ± 0.18 Ephemerella crenula (m) 7.1 0.73 ± 0.048 0.58 ± 0.047 Ephemerella hispida (n) 13.9 0.27 ± 0.037 0.26 ± 0.0080 Ephemerella invaria (o) 3.5 0.98 ± 0.15 0.91 ± 0.037 Ephemerella rossi (p) 3.9 0.30 ± 0.16 0.44 ± 0.13 Ephemerella subvaria (q) 1.1 0.66 ± 0.19 0.80 ± 0.080 Eurylophella verisimilis (r) 6.2 0.54 ± 0.045 0.50 ± 0.044 Teleganopsis deficiens (s) 2.5 0.65 ± 0.030 0.47 ± 0.026 *Letters are symbols representing hydropsychid (a-f) and ephemerellid (g-s) species in other graphics throughout this paper.

122

Table 2. Cd and Zn efflux rate constants (ke) for 18 aquatic insect species.

Mean body Genus species wt. (mg) Cd ke ± s.d. Zn ke ± s.d. Arctopsyche irrorata (a)* 60.7 0.075 ± 0.046 0.079 ± 0.035 Diplectrona modesta (b) 10.4 0.075 ± 0.017 0.071 ± 0.024 Hydropsyche alhedra (c) 15.7 0.17 ± 0.010 0.11 ± 0.014 Hydropsyche slossonae (d) 15.2 0.14 ± 0.022 0.13 ± 0.018 Hydropsyche sparna (e) 5.8 0.085 ± 0.0092 0.083 ± 0.014 Parapsyche cardis (f) 27.1 0.042 ± 0.027 0.028 ± 0.018 Dannella sp. (g) 4.5 0.11 ± 0.077 0.14 ± 0.090 Drunella cornutella (h) 8.0 0.035 ± 0.011 0.050 ± 0.017 Drunella longicornis (i) 6.4 0.010 ± 0.0014 0.016 ± 0.0033 Drunella tuberculata (j) 3.1 0.022 ± 0.0054 0.057 ± 0.0090 Drunella walkeri (k) 7.2 0.19 ± 0.047 0.11 ± 0.029 Ephemerella catawba (l) 5.0 0.16 ± 0.030 0.15 ± 0.030 Ephemerella crenula (m) 7.1 0.17 ± 0.058 0.18 ± 0.060 Ephemerella hispida (n) 13.9 0.18 ± 0.020 0.17 ± 0.010 Ephemerella invaria (o) 3.5 0.023 ± 0.0035 0.047 ± 0.011 Ephemerella rossi (p) 3.9 0.019 ± 0.0073 0.031 ± 0.011 Ephemerella subvaria (q) 1.1 0.053 ± 0.019 0.086 ± 0.022 Eurylophella verisimilis (r) 6.2 0.017 ± 0.011 0.013 ± 0.0084 *Letters are symbols representing hydropsychid (a-f) and ephemerellid (g-s) species in other graphics throughout this paper.

123

Table 3. Cd and Zn bioaccumulation factors (BCF) for 18 aquatic insect species. Mean body Genus species wt. (mg) Mean Cd BCF Mean Zn BCF (min – max) (min – max) Arctopsyche irrorata (a)* 60.7 931 584 (572 – 2,393) (390 – 1,085) Diplectrona modesta (b) 10.4 3,609 2,375 (2,747 – 4,962) (1,587 – 3,952) Hydropsyche alhedra (c) 15.7 256 265 (157 – 367) (200 – 348) Hydropsyche slossonae (d) 15.2 824 325 (618 – 1,105) (188 – 508) Hydropsyche sparna (e) 5.8 4,297 1,139 (3,598 – 5,166) (940 – 1,421) Parapsyche cardis (f) 27.1 235 372 (84 – 939) (140 – 1,433) Dannella sp. (g) 4.5 1,053 1,071 (471 – 4,054) (626 – 3,120) Drunella cornutella (h) 8.0 21,441 10,040 (16,036 – 31, 685) (7,143 – 15,972) Drunella longicornis (i) 6.4 78,204 56,089 (56,714 – 106,953) (40,433 – 79,656) Drunella tuberculata (j) 3.1 88,423 27,477 (68,727 – 120,689) (23,510 – 32,916) Drunella walkeri (k) 7.2 1,486 1,966 (1,067 – 2,180) (1,438 – 2,895) Ephemerella catawba (l) 5.0 4,806 3,791 (2,791 – 7,752) (2,186 – 6,179) Ephemerella crenula (m) 7.1 4,208 3,225 (2,944 – 6,746) (2,218 – 5,247) Ephemerella hispida (n) 13.9 1,053 1,524 (1,171 – 1,917) (1,393 – 1,671) Ephemerella invaria (o) 3.5 41,764 19,140 (30,587 – 56,980) (14,886 – 25,986) Ephemerella rossi (p) 3.9 16,033 14,472 (5,397 – 40,296) (7,546 – 28,654) Ephemerella subvaria (q) 1.1 12,434 9,279 (6,486 – 25,029) (6,650 – 13,716) Eurylophella verisimilis (r) 6.2 31,877 38,485 (17,529 – 103,464) (21,355 – 117,382) *Letters are symbols representing hydropsychid (a-f) and ephemerellid (g-s) species in other graphics throughout this paper.

124

Supporting Information

E p h e m e r e llid a e

) 1

- 0 .0 H y d r o p s y c h id a e d

(

e k -0 .5

n Z g m n

d l

d e -1 .0 k c

m q e a

r j b o

f o h s

n -1 .5

p f

a r

- i

g r

o -2 .0

L 0 .0 0 .5 1 .0 1 .5 2 .0

) 1

- 0 .0

d (

k -0 .5

k n d l c

C m g d d -1 .0 e e b a

m q

r f h o -1 .5

f o

s j

p r n

a -2 .0 i

t -

g o -2 .5

L 0 .0 0 .5 1 .0 1 .5 2 .0 L o g -tr a n sfo r m e d m a ss (m g )

Figure S1. Log-transformed Zn (a) and Cd (b) kes of 18 species versus the average log- transformed body weight of individuals. Letters represent the names of species identified in Table 1.

125

CHAPTER 4

Four reasons why traditional metal toxicity testing with aquatic insects is irrelevant

Monica D. Poteat and David B. Buchwalter

Environmental and Molecular Toxicology, Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695, United States

Published In: Environmental Science and Technology. 2014; Volume 48; pages 887-888.

126

Trace metal contamination of freshwater ecosystems is a problem worldwide, and insects are typically the predominant invertebrate faunal group in these systems. Metals can shape community structure, as evidenced by reduced biodiversity in affected areas. Aquatic insects are often some of the first species to disappear from metal-contaminated sites, despite the fact that laboratory toxicity tests would suggest that aquatic insects are insensitive to metals. In fact, typical laboratory results would indicate that insects only respond to dissolved metals at concentrations orders of magnitude larger than those found in the most insect-depleted contaminated sites. Even with mounting evidence highlighting the obvious disconnect between laboratory toxicity tests and field observations regarding metal toxicity to aquatic insects, water quality criteria for metals continues to rely primarily on toxicity values derived from short term dissolved-only exposures. Below we discuss four key reasons as to why such tests don’t provide relevant data for this important faunal group, focusing upon recent advances in our understanding of bioaccumulation and mechanisms of toxicity.

1) Reaching steady state tissue concentrations takes time, lots of time.

Here, we used previously published1–3 cadmium uptake and efflux rate constants to model the time it would take 34 species of EPT taxa (Ephemeroptera, Plecoptera,

Trichoptera) to reach steady state tissue concentrations. Steady state times were determined by the following equations,

where = steady state concentration, ku = uptake rate constant, ke = efflux rate constant, and Cw = dissolved metal concentration, and

127

where = time to steady state concentration, = steady state concentration, = efflux rate constant, and = time.

Median times to steady state concentration from dissolved Cd exposures were 405,

70, and 50 days, respectively, for EPT taxa (Fig. 1). These times to steady state tissue concentration far surpass the time allotted for most acute toxicity tests (96 hours). Exposure duration in traditional toxicity testing is clearly insufficient from a bioaccumulation perspective.

2) Traditional understanding of the dissolved acute toxicity mechanisms don’t seem to apply to aquatic insects.

Some might argue that the surface action of metals (rather than bioaccumulation) better predicts acute toxicity. In rainbow trout for example, acute metal toxicity is well predicted by the surface action of metals on gills, and associated osmoregulatory disturbance.

In particular, metal exposures (e.g., Cd, Cu) have been shown to reduce the influx rates of major ions Ca and Na, respectively. Recent work with aquatic insects has demonstrated a lack of interactions between Ca and heavy metals Zn and Cd at the apical surface of aquatic insects in dissolved exposures.2,3 Increased Ca concentrations only moderately protects against Cd and Zn uptake,2 and the uptake of Cd and Zn occurs without impeding Ca uptake.3

In insects, the prevailing paradigm that Cd and Zn will out-compete Ca for apical entry and result in osmoregulatory disturbance does not appear to be correct. Similar work focusing on metal (e.g., Cu, Ag) effects on Na transport has yet to be performed.

128

3) Tissue burdens of metals acquired from diet far surpass those obtained from dissolved exposures in aquatic insects.

Current acute toxicity methodologies rely solely on dissolved metal exposures to derive water quality criteria. However, in every comparison of dietary vs. dissolved acquisition of metals that we are aware, diet is the predominant route of exposure for aquatic insects. For example, Cain et al.4 compared site specific water chemistry and field measures of tissue Cu and Cd concentrations with lab based bioaccumulation studies to show that dissolved exposures could only account for <5% of actual observed tissue burdens of insects in the field. From this study and others, we can conclude that metals acquired through diet more heavily drive metal bioaccumulation in aquatic insects, thus potentially influencing metal toxicity more so than metal acquired through dissolved exposures alone.

4) Metals obtained from dietary sources may be more physiologically active than metals derived from dissolved exposures.

The inherent assumption that metals derived from solution are more physiologically active than metals derived from diet is questionable. Using the lab-reared mayfly

Centroptilum triangulifer, Xie and Buchwalter5 achieved comparable tissue concentration of

Cd from either dissolved or dietary exposures. They found that dietary Cd exposures resulted in a significant depression of antioxidant enzymes (catalase, superoxide dismutase and glutathione-S-transferase) whereas dissolved exposures of Cd had no effect on these antioxidant enzymes. This suggests that diet-derived metals can be more physiologically

129

active (and potentially toxic) than water derived metals, and helps explain why insects tend to be so unresponsive to traditional toxicity tests that only consider dissolved routes of exposure.

What are we trying to protect?

The goal of water quality criteria in the US is to protect aquatic communities from impairment resulting from contaminant exposures. Insects are the critical faunal group in freshwaters from both structural and functional perspectives, which is why they are the focal group of monitoring programs. Yet data requirements to generate criteria (8 families, including only 1 insect) remain unchanged since 1985. (The suitability of only requiring a single insect species to represent ~7,000 species can be debated elsewhere). Here we provide evidence for why the test methodologies themselves are inadequate and in need of modernization so that water quality criteria are scientifically relevant. We suggest that appropriate toxicity approaches with insects which take into account their unique physiologies can reduce the uncertainties that currently exist in the protectiveness of water quality criteria for metals. We further advocate that the development of relevant insect toxicity models will help bridge the gap that currently exists between ecological monitoring programs (that typically focus upon insect community structure) and toxicity based approaches for setting environmental standards.

130

References

(1) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.; Garland, Jr., T. Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. P. Natl. Acad. Sci. USA. 2008, 105, 8321–8326.

(2) Poteat, M. D.; Díaz-Jaramillo, M.; Buchwalter, D. B. Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna. J. Exp. Biol. 2012, 215, 1575–1583.

(3) Poteat, M.; Buchwalter, D. Calcium uptake in aquatic insects: Influences of phylogeny and metals (Cd and Zn). J. Exp. Biol. 2013.

(4) Cain, D.; Croteau, M.-N.; Luoma, S. N. Bioaccumulation dynamics and exposure routes of Cd and Cu among species of aquatic mayflies. Env. Toxicol. Chem. 2011, 30, 2532–2541.

(5) Xie, L.; Lambert, D.; Martin, C. A.; Cain, D. J.; Luoma, S. N.; Buchwalter, D. B. Cadmium biodynamics in the oligochaete Lumbriculus variegatus and its implications for trophic transfer. Aquat. Toxicol. 2008, 86, 265–271.

131

Figures

) s

y 5 5 0 0 a

D 1 0 0 0

(

]

e 7 5 0

t a

t 5 0 0

y 2 5 0 d

a 1 0 0

S [

5 0

e m

i 0 T a a a r r r te te te p p p o o o r c e h le ic m P r e T h p E

Figure 1) Modeled time (in days) required to attain steady state tissue concentrations of Cd in 34 EPT species. Each symbol represents a different species, and each line represents the median time.

132

CHAPTER 5

The importance of retaining a phylogenetic perspective in traits-based analyses

Monica D. Poteat1, Luke M. Jacobus2, and David B. Buchwalter1

1Department of Environmental and Molecular Toxicology, North Carolina State University, Raleigh, NC 27695, United States

2 Division of Science, Indiana University Purdue University Columbus, Columbus, Indiana, 47203, United States

Formatted for submission to Freshwater Biology.

133

Abstract

1) Here, we analyze the influence of phylogeny on 3 physiological traits linked to the accumulation of the toxic metal Cd from solution across 42 aquatic insect species representing orders Ephemeroptera (mayfly), Plecoptera (stonefly), and Trichoptera

(caddisfly). These traits included the propensity to take up Cd from water (ku), the ability to excrete Cd (ke), and the net result of these two processes (bioconcentration factor- BCF).

2) Ranges in traits related to dissolved Cd bioaccumulation varied in magnitude across lineages (some lineages had a greater tendency to bioaccumulate Cd than others). Overlap in the ranges of trait values in different lineages was common and creates a situation where species from different lineages could share a comparable trait state representing the high end for one lineage and the low end of another. Lineage provides context for traits.

3) Variance around the mean trait state for a given clade was highly variable, suggesting that some groups (e.g., Ephemerellidae) are inherently more variable than others (e.g., Perlidae).

Thus, trait variability across species is at least partially a function of lineage.

4) Akaike information criterion (AIC) comparisons of statistical models were driven by clade and generally improved with increasing taxonomic resolution.

5) Taken together, these results suggest that phylogeny (clade membership) provides context for the evaluation of trait states in community ecology. Failure to consider the evolutionary framework of species traits may hinder the progress of trait based approaches for understanding how communities are structured and respond to environmental change.

134

Introduction

The study of species traits (any detectable phenotypic property of an organism) is emerging in stream ecology as a way to better understand and interpret community composition (Bonada et al., 2006; Menezes, Baird & Soares, 2010). Historically, the study of species traits has developed from two distinct traditions – evolutionary biology and ecology.

Evolutionary biologists are generally concerned with patterns and processes of evolutionary change in the context of lineage and descent (Garland, Bennett & Rezende, 2005), while ecologists often frame traits as adaptations to their surrounding environments (Southwood,

1977) with an apparent greater emphasis on convergent evolution (Poff et al., 2006). Both perspectives have provided valuable insights into biological form and function despite their different philosophical approaches.

To date, the study of traits in benthic ecology has been largely dominated by the functional guild approach of ecologists, essentially disregarding lineage as a driver of traits.

However, important traits for the survival of organisms are often conserved in lineages

(Webb et al., 2002), as constraints imposed on an organism (e.g., physiology, behavior, morphology, etc) are directed by an organism’s phylogenetic history (Southwood, 1988).

Further, a given trait may be differentially linked to other traits as a function of lineage and evolutionary history (Tokeshi, 1999). Poff et al. (2006) referred to this phenomenon as “trait syndromes” and noted the association of these syndromes with different lineages.

Consequently, there are risks assumed when excluding a descent-based perspective in analyzing traits across species. Trait function is reliant on the remainder of an individual’s or a species’ biology (Verberk, van Noordwijk & Hildrew, 2013). For example, a given trait

135

state “x” superimposed on a mayfly background may mean something different than that same trait state superimposed onto a dipteran background, with “x” differentially linked to other traits as a function of lineage and/or representing a low value in one lineage and a high value for the other. The mayfly background may impose a tighter phylogenetic constraint on the evolutionary responsiveness of a trait, whereas a dipteran background may impose less constraint (sensu Blomberg, Garland, Jr. & Ives, 2002). Thus, phylogeny gives traits context, and it has been recognized that ecological studies would benefit from including phylogenetic information in analyses (Webb et al., 2002; Verberk, van Noordwijk & Hildrew, 2013).

The tendency for evolutionarily related species to resemble each other is termed

“phylogenetic signal,” and there is extensive support for the phylogenetic parsing of traits across species (Blomberg, Garland, Jr. & Ives, 2003). Analyses of 53 (morphological, behavioral, life history, or physiological) traits where datasets included more than 20 species showed that 92% of traits exhibited phylogenetic signal (Blomberg et al., 2003). Blomberg et al. show that different types of traits have more concordance with phylogeny than others.

Behavioral traits, which are considered to be more evolutionarily malleable and adaptive, exhibited weaker phylogenetic signal, than physiological, morphological, or life history traits

(Blomberg et al., 2003). While the physiological traits vary in strength of phylogenetic signal, we are particularly interested in them as key mechanistic drivers of species performance.

From another perspective, clades can have a great variation in the ranges of trait values possible (though even highly variable traits can have phylogenetic signal across species). For instance, some traits may show significant variance in the magnitude of values

136

between species groups (Fig. 1a). Here, variance around mean trait values is similar, but lineages occupy different space with respect to magnitude. Additionally, overlap in trait states can create a circumstance where species sharing a comparable trait state could represent the higher end of one lineage’s physiological space, but the lower end of another’s.

Conversely, some lineages may share a comparable mean trait state, but potentially vastly different variance around the mean (Fig. 1b). Here, trait lability is a function of lineage.

Thus, including phylogenetic information can puts a trait in perspective for a given clade.

The failure to consider traits in the context of lineage may introduce more noise than clarity in the analysis of communities.

The uncertainty around current practices which involve the generalization of traits across species (e.g., use of surrogate species in ecotoxicology, coarse taxonomic ranks used in biomonitoring) is thus tied to the extent to which species vary within taxonomic ranks.

Variance among closely related species is not novel- Resh and Unzicker (1975) found that of

89 aquatic insect genera for which water quality tolerances (tolerant, facultative, intolerant) had been ascribed to more than one species, at least 2 tolerance categories were represented in 61 of those genera. Moreover, in the analysis of traits, continuous traits (e.g., body weight) are often reduced to discrete categories (Orlofske & Baird, 2014), ignoring variation which can occur among closely related species.

Here, we analyzed physiological traits directly related to dissolved Cd bioaccumulation in a comparative context to determine the influence of phylogeny on these traits. We compiled Cd uptake rate constants (ku), efflux rate constants (ke), and bioaccumulation factors (BCF) for 42 species within Ephemeroptera (mayfly), Plecoptera

137

(stonefly), and Trichoptera (caddisfly) – three orders which are used often in biomonitoring and bioasssessment programs. We analyzed the evolutionary patterns of variance in Cd bioaccumulation parameter across species within different taxonomic ranks (order, family genus). We further quantified the influence of phylogeny (via phylogenetic signal) on Cd bioaccumulation parameters. Lastly, we compared statistical models including body weight, lineage and feeding guild to examine how ecological traits and evolutionary history might influence the metal bioaccumulation traits examined.

138

Materials and Methods

Data collection

Three studies contributed dissolved Cd ku and ke values for a total of 42 aquatic insect species (Buchwalter et al., 2008; Cain, Croteau & Luoma, 2011; Poteat & Buchwalter,

2014). Bioconcentration factors from Cd ku and ke values were calculated for all species for further analyses (see Poteat & Buchwalter, 2014). These three parameters are physiological traits directly related to dissolved Cd bioaccumulation (Buchwalter et al., 2008). To use data from Cain et al. (2011), we converted dry body weights to wet weights under the assumption that water accounted for 77% of body weight. Cd ku values were also converted accordingly.

Data analysis

All Cd bioaccumulation parameters were log transformed prior to analyses which resulted in normally-distributed data for all three parameters. One-way ANOVA analyses in

GraphPad Prism (v6.02) determined significant differences in Cd bioaccumulation parameters between families. Simple and multiple linear regressions performed in the R 3.0.3 program (Team, 2013) determined the effects of clade (order, family, or genus), body weight, and feeding guild (per Poff et al., 2006) on each Cd bioaccumulation parameter. Clade was analyzed because taxonomic information is the basis for most ecotoxicological and ecological bioassessment studies, weight was included because allometric effects are often seen in physiological traits of aquatic organisms (Wang & Fisher, 1997; Zhang & Wang,

2007; Buchwalter et al., 2008), and feeding guild was chosen to explore possible

139

physiological differences associated with life history. Akaike Information Criterion (AIC) determined the best fit model for each dissolved Cd bioaccumulation parameter. The smaller- is-better formulation was used [AIC = (-2 x ln ML) + 2 x no. of parameters], and the model with the lowest AIC was taken as the best fit model. Simpler models whose AIC was ≤ 2 units larger were considered to have substantial support (Lavin et al., 2008) and preferred over more complex models. Figures were prepared in GraphPad Prism (v6.02).

Phylogenetic methods (Blomberg et al., 2003) were used to determine the extent that

Cd bioaccumulation parameters followed phylogeny across EPT taxa. A phylogeny containing all 42 species of Ephemeroptera (Wang & McCafferty, 2004; Jacobus &

McCafferty, 2008; Ogden et al., 2009a, Ogden et al., 2009b), Plecoptera (Federhen, 2012), and Trichoptera (Geraci, Al-Saffar & Zhou, 2011; Zhou et al., 2011) was constructed from the available literature. Three arbitrary branch lengths (all branches set to equal 1, Pagel’s arbitrary (Pagel, 1992), and Grafen’s arbitrary (Grafen, 1989)) were employed to determine the branch lengths which resulted in the highest ln likelihood (the branch lengths which fit the data best) (see Poteat et al., 2013). Blomberg’s K-statistic (Blomberg et al., 2003) was used to quantify phylogenetic signal (the tendency for closely related species to resemble each other) for each Cd bioaccumulation parameter examined. The randomization test based on the mean-squared error was used to determine significance of the phylogenetic signal.

140

Results

Evolutionary patterns of physiological traits across species

Here, we analyzed a dataset of physiological traits compiled from the literature in order to explore evolutionary patterns across Cd bioaccumulation parameters (Table 1).

Though not all lineages in our dataset have adequate sample size for comparisons, we see dissolved Cd bioaccumulation parameters within taxa which follow patterns described in our conceptual Figure 1. For example, Cd ku, ke and BCF values varied in magnitude across orders (Fig. 2). Mayfly families generally had higher Cd ku and BCF values than stonefly and caddisfly families. Ephemerellidae in particular had significantly higher Cd ku values than

Perlidae and Hydropsychidae species (p < 0.05) (Fig. 2a). There was also substantial overlap in each parameter. The lower range of mayfly ku and BCF values overlapped with the higher ranges of stonefly and caddisfly values. There were no apparent differences in the magnitudes of ke values across orders, however there was substantial overlap in values across all families.

The variance around the mean trait state of clades could also be highly variable, following the trend described in conceptual Figure 1b. For example, families Ephemerellidae and Perlidae had statistically equivalent Cd ke mean values (Fig. 2b), however the ranges of

-1 -1 Cd ke values for Ephemerellidae (0.01 – 0.20 d ) and Perlidae (0.081-0.087 d ) were radically different. Another example is seen in Cd BCFs (Fig. 2c), where Perlidae and

Hydropsychidae have statistically equivalent BCF means, however Hydropsychidae has a range of possible values almost twice that of Perlidae.

141

Across all 42 species in orders Ephemeroptera, Plecoptera, and Trichoptera, moderate phylogenetic signal was present in Cd ku values, signifying that evolutionary history is a significant driver of Cd uptake (K = 0.52, p < 0.002, branch lengths of 1). Phylogenetic signal was not significant across ke values (K = 0.32, p = 0.32, branch lengths of 1). A weaker phylogenetic signal was marginally significant in Cd BCFs (K = 0.35, p = 0.06, branch lengths of 1).

Linkages of traits across species

We also examined different combinations of predictive parameters (traits) and their effects on Cd ku values, ke values, and BCFs (Table 2). More complex comparisons than those seen in Table 2 were not possible due to multicollinearity of some parameters

(particularly, clade and feeding guild membership). Single and multiple linear regression analyses revealed that each parameter and combination of parameters (body weight, feeding guild, and clade as genus, family or order) described some percentage of variation across Cd ku values and Cd BCF values. The linear model which best described Cd ku values across species via AIC analysis included only genus as a predictive parameter (r2 = 0.69, p =

0.0008). The linear model which best described Cd BCF values across species included both genus and body weight as predictive parameters (r2 = 0.58, p = 0.008). No predictive parameter had any significant effects on Cd ke values across species.

142

Discussion

Traits-based analyses have traditionally focused on linking species’ traits within a community to the community’s structure as well as to it’s vulnerability to stressors (Poff et al., 2006). To date, many analyses have focused on demographic and recolonization traits linked with population sustainability (sensu Van Straalen, 1994) largely because these traits are relatively easy to obtain. Physiological traits, which are mechanistically linked to intrinsic sensitivity to a given environmental stressor (sensu Van Straalen, 1994), have been largely ignored in traits analyses with aquatic insects, because few research groups focus on aquatic insect physiology (but see Liess & Von Der Ohe, 2005). Because most water quality problems affect species via physiological processes, physiological traits mechanistically link species’ performance to their environment. We suggest it would be beneficial to better understand how these physiological traits vary across species.

Here, we comparatively analyze three physiological traits directly linked with dissolved Cd bioaccumulation across three orders of aquatic insects – Ephemeroptera,

Plecoptera, and Trichoptera. While we focused solely on the dissolved route of exposure, we acknowledge that diet (Martin et al., 2007; Cain et al., 2011) and detoxification capabilities

(Buchwalter et al., 2008) are also important drivers of metal bioaccumulation and toxicity in aquatic insects. We found that the magnitudes of dissolved bioaccumulation trait values differed across orders and families, and that these values generally support coarse field generalizations. Mayflies are particularly sensitive to Cd in the field, whereas caddisflies are generally thought of as more metal tolerant (Winner, Boesel & Farrell, 1980; Clements,

Cherry & Van Hassel, 1992; Clements et al., 2000; Clements, Cadmus & Brinkman, 2013). It

143

is possible that higher rates of ion/metal uptake, and subsequent higher metal BCFs, contribute to the sensitivity of mayfly species to trace metals in the field.

Aquatic insects are known to be especially physiologically diverse (Merritt, Cummins

& Berg, 2008), with contemporary species stemming from different, unique evolutionary histories (Bradley et al., 2009). Many physiological and life history traits of aquatic insects have been described as variable both intra- and inter-specifically – number of larval instars

(Esperk, Tammaru & Nylin, 2007), desiccation survival of eggs (Sota & Mogi, 1992), body size (Peckarsky, Taylor & McIntosh, 2001), and diet (Hawkins, 1985) to name a few.

However despite their great diversity, the challenging nature of species-level identifications in aquatic insects has led to a reliance on coarser levels of identification (Carew, Miller &

Hoffmann, 2011). In biomonitoring and bioassessments protocols, aquatic insect identifications are often left at the family (Hilsenhoff, 1988; Chessman, 1995; Simpson &

Norris, 2000) or genus (Barbour & Yoder, 2000; Bady et al., 2005) level of taxonomic identification. In applied ecotoxicology, single surrogate laboratory species are often used to represent entire taxonomic groups (Stephan et al., 1985). Additionally, trait analyses often occur at the genus- or family- level of taxonomic identification (Usseglio-Polatera et al.,

2000; Poff et al., 2006; Dolédec, Phillips & Townsend, 2011).

Rarely do we pay attention to the range of trait values within clades. In this study, it was apparent that some clades occupy a large physiological space (e.g., Ephemerellidae), while other clades (e.g., Perlidae) occupy a much smaller physiological space. Variation among taxonomic groups can limit the efficacy of generalizations across species, particularly at the family-level of identification (Lenat & Resh, 2001). Consequently, across disciplines

144

there continues to be a fundamental lack of appreciation for how variable traits can be within either taxonomic or functional groups. Because species within even the same families and genera can vary widely (as seen in this study), it is important to consider this variation when choosing representatives in the laboratory as well as the appropriate taxonomic ranks to take identifications down to in the field. It is also an important consideration in applying laboratory and field data to risk assessments.

Evolutionary history largely explained the variation of dissolved Cd bioaccumulation parameters across species. Here, we show that while vast differences occur in the traits of closely-related clades, evolutionary history still explained part of the variation in ku values and BCFs across species. Phylogenetic patterns are well documented across species traits

(Garland et al., 2005) in aquatic and terrestrial taxa, and the phylogenetic signal observed here underlines the importance of evolutionary history in species traits. Including phylogeny in traits analyses could allow for the linkage of species to their environment by way of their trait combinations (Verberk, van Noordwijk & Hildrew, 2013). In understanding phylogentic patterns of traits across species, it may become possible to predict trait values across untested species (Rubach et al., 2011).

The ecological traits that we examined (body weight, feeding guild) also explained variation within Cd bioaccumulation traits ku and BCF, although to a lesser extent than evolutionary history. We were unable to analyze many multiple linear regression models due to multicolinnearity. The muticollinearity of clade (e.g., genus and family) and feeding guild specifically underlines the importance of evolutionary history across species’ traits, as well as the importance of retaining a phylogenetic perspective in the analysis of ecological (and

145

other) traits. It has been noted that in keeping this phylogenetic perspective outlook, the confounding nature of phylogeny could become an ally in traits analyses (Verberk, van

Noordwijk & Hildrew, 2013).

Allometric effects are found in traits across both terrestrial and aquatic organisms, and have been observed in trace metal bioaccumulation in freshwater organisms (Wang &

Fisher, 1997; Zhang & Wang, 2007; Buchwalter et al., 2008). Body size can vary across species, even within the same genus, due to a variety of natural conditions (e.g., temperature of environment, food availability) and lab-derived artifacts (e.g., instar of a species collected). In traits analyses, it is particularly difficult to assign discrete categories representative of body size for an entire clade due to variations of body size found in nature.

Feeding guild, as other species traits, can have inter- and intra-specific variation.

Aquatic insects are often considered as trophic generalists (Cummins, 1973). Mayflies (and ephemerellids, in particular) are exceptionally facultative in their feeding, with species even within the same genus exhibiting remarkably different feeding habits. While most mayflies are herbivores of some sort, some species can exhibit predatory tendencies (e.g., Drunella spp. (Hawkins, 1985)). Therefore, it is exceptionally hard to place species within a specific guild when their feeding capabilities are fluid.

The variation of efflux rate constants (ke) across species was not explained by evolutionary history, nor was it explained by body weight or feeding guild. Previous work has found that Cd ke values exhibit phylogenetic signal across 4 aquatic phyla (Poteat et al.,

2013) and even within many of the same EPT genera examined here (Buchwalter et al.,

2008). However, the variation within families, particularly within Ephemerellidae and

146

Hydropsychidae, does not parse with phylogeny (Poteat & Buchwalter, 2014). While phylogeny was not a driver of ke values here, it was clear that families (and orders) have differing capacities (ranges) in removing Cd from their tissues.

Overall, our results underline the importance of retaining an evolutionary perspective in the analysis of traits in community ecology. The linkage of traits to the evolutionary history of each organism demands a phylogenetic framework to adequately analyze traits among closely (and distantly) related organisms. Failure to consider the evolutionary framework of species traits may hinder the progress of trait based approaches for understanding how communities are structured and respond to environmental change.

147

Acknowledgements

We thank Dan Cain for supplying additional data for these analyses. Gerald LeBlanc

(NCSU), Justin Conley (NCSU), Allison Camp (NCSU) and Tom Augspurger (USFWS) provided valuable editorial comments. Eric Stone (NCSU) provided statistical support. This work was supported by the National Science Foundation (IOS 0919614), the ICA Chris Lee

Award for Metals Research and the Society of Environmental Toxicology and Chemistry

(SETAC). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the ICA or SETAC.

148

References

Bady P., Doledec S., Fesl C., Gayraud S., Bacchi M. & Scholl F. (2005) Use of invertebrate traits for the biomonitoring of European large rivers: the effects of sampling effort on genus richness and functional diversity. Freshwater Biology, 50, 159–173.

Barbour M.T. & Yoder C.O. (2000) The multimetric approach to bioassessment, as used in the United States of America. In: Assessing the Bioogical Quality of Freshwater RIVPACs and Other Techniques. (Eds J.F. Wright, D.W. Sutcliffe & M.T. Furse), pp. 281–292. Freshwater Biological Association.

Blomberg S.P., Garland, Jr. T. & Ives A.R. (2002) Tempo and mode in evolution: Phylogenetic inertia, adaptation and comparative methods. Journal of Evolutionary Biology, 15, 899–910.

Blomberg S.P., Garland, Jr. T. & Ives A.R. (2003) Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution, 57, 717–745.

Bonada N., Prat N., Resh V.H. & Statzner B. (2006) Developments in aquatic insect biomonitoring: a comparative analysis of recent approaches. Annual Review of Entomology, 51, 495–523.

Bradley T.J., Briscoe A.D., Brady S.G., Contreras H.L., Danforth B.N., Dudley R., et al. (2009) Episodes in insect evolution. Integrative and Comparative Biology, 49, 590–606.

Buchwalter D.B., Cain D.J., Martin C.A., Xie L., Luoma S.N. & Garland, Jr. T. (2008) Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. Proceedings of the National Academy of Sciences of the United States of America, 105, 8321–8326.

Cain D., Croteau M.-N. & Luoma S.N. (2011) Bioaccumulation dynamics and exposure routes of Cd and Cu among species of aquatic mayflies. Environmental Toxicology and Chemistry, 30, 2532–2541.

Carew M.E., Miller A.D. & Hoffmann A.A. (2011) Phylogenetic signals and ecotoxicological responses: Potential implications for aquatic biomonitoring. Ecotoxicology, 20, 595–606.

Chessman B.C. (1995) Rapid assessment of rivers using macroinvertebrates: A procedure based on habitat-specific sampling, family level identification and a biotic index. Australian Journal of Ecology, 20, 122–129.

149

Clements W.H., Cadmus P. & Brinkman S.F. (2013) Responses of aquatic insects to Cu and Zn in stream microcosms: Understanding differences between single species tests and field responses. Environmental Science & Technology, 47, 7506–7513.

Clements W.H., Carlisle D.M., Lazorchak J.M. & Johnson P.C. (2000) Heavy metals structure benthic communities in Colorado mountain streams. Ecological Applications, 10, 626–638.

Clements W.H., Cherry D.S. & Van Hassel J.H. (1992) Assessment of the sensitivity of heavy metals on benthic communities at the Clinch River (Virginia): Evaluation of an index of community sensitivity. Canadian Journal of Fisheries and Aquatic Sciences, 49, 1686–1694.

Cummins K.W. (1973) Trophic relations of aquatic insects. Annual Review of Entomology, 18, 183-206.

Dolédec S., Phillips N. & Townsend C. (2011) Invertebrate community responses to land use at a broad spatial scale: trait and taxonomic measures compared in New Zealand rivers. Freshwater Biology, 56, 1670–1688.

Esperk T., Tammaru T. & Nylin S. (2007) Intraspecific Variability in Number of Larval Instars in Insects. Journal of Economic Entomology, 100, 627–645.

Federhen S. (2012) The NCBI Taxonomy database. Nucleic Acids Research 40, D136–D143.

Garland T., Bennett A.F. & Rezende E.L. (2005) Phylogenetic approaches in comparative physiology. Journal of Experimental Biology, 208, 3015–3035.

Geraci C.J., Al-Saffar M.A. & Zhou X. (2011) DNA barcoding facilitates description of unknown faunas: a case study on Trichoptera in the headwaters of the Tigris River, Iraq. Journal of the North American Benthological Society, 30, 163–173.

Grafen A. (1989) The phylogenetic regression. Philosophical Transactions of the Royal Society B, 326, 119–156.

Hawkins C.P. (1985) Food habits of species of Ephemerellid mayflies (Ephemeroptera: Insecta) in streams of Oregon. The American Midland Naturalist, 18, 183–206.

Hilsenhoff W.L. (1988) Rapid field assessment of organic pollution with a family-level biotic index. Journal of the North American Benthological Society, 7, 65–68.

150

Jacobus L.M. & McCafferty W.P. (2008) Revision of Ephemerellidae Genera (Ephemeroptera). Transactions of the American Entomological Society (1890-), 134, 185-274.

Lavin S.R., Karasov W.H., Ives A.R., Middleton K.M. & Garland T. (2008) Morphometrics of the avian small intestine compared with that of nonflying mammals: A phylogenetic approach. Physiological and Biochemical Zoology, 81, 526–550.

Lenat D.R. & Resh V.H. (2001) Taxonomy and stream ecology — The benefits of genus- and species-level identifications. Journal of the North American Benthological Society, 20, 287–298.

Liess M. & Von Der Ohe P.C. (2005) Analyzing effects of pesticides on invertebrate communities in streams. Environmental Toxicology and Chemistry, 24, 954–965.

Martin C.A., Cain D.J., Luoma S.N. & Buchwalter D.B. (2007) Cadmium ecophysiology in seven stonefly (Plecoptera) species: Delineating sources and estimating susceptibility. Environmental Science & Technology, 41, 7171–7177.

Menezes S., Baird D.J. & Soares A.M.V.M. (2010) Beyond taxonomy: a review of macroinvertebrate trait-based community descriptors as tools for freshwater biomonitoring. Journal of Applied Ecology, 47, 711–719.

Merritt R.W., Cummins K.W. & Berg M.B. (2008) An Introduction to the Aquatic Insects of North America. (Eds R.W. Merritt, K.W. Cummins & M.B. Berg), Kendall-Hunt, Dubuque, IA.

Ogden T.H., Osborne J.T., Jacobus L.M. & Whiting M.F. (2009) Combined molecular and morphological phylogeny of Ephemerellinae (Ephemerellidae:Ephemeroptera), with remarks about classification. Zootaxa, 1991, 28–42.

Ogden T.H., Gattolliat J.L., Sartori M., Staniczek A.H., Soldan T. & Whiting M.F. (2009) Toward a new paradigm in mayfly phylogeny (Ephemeroptera): Combined analysis of morphological and molecular data. Systematic Entomology, 34, 616-634.

Orlofske J.M. & Baird D.J. (2014) Incorporating continuous trait variation into biomonitoring assessments by measuring and assigning trait values to individuals or taxa. Freshwater Biology, 59, 477–490.

Pagel M.D. (1992) A method for the analysis of comparative data. Journal of Theoretical Biology, 156, 431–442.

151

Peckarsky B.L., Taylor B.W. & McIntosh A.R. (2001) Variation in mayfly size at metamorphosis as a developmental response to risk of predation. Ecology, 82, 740–757.

Poff N.L., Olden J.D., Vieira N.K.M., Finn D.S., Simmons M.P. & Kondratieff B.C. (2006) Functional trait niches of North American lotic insects: Traits-based ecological applications in light of phylogenetic relationships. Journal of the North American Benthological Society, 25, 730–755.

Poteat M.D. & Buchwalter D.B. (2014) Phylogeny and size differentially influence dissolved Cd and Zn bioaccumulation parameters among closely related aquatic insects. Environmental Science & Technology, 48, 5274–5281.

Poteat M.D., Garland, Jr. T., Fisher N.S., Wang W.-X. & Buchwalter D.B. (2013) Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms. Environmental Science & Technology, 47, 7989–7995.

Resh V.H. & Unzicker J.D. (1975) Water quality monitoring and aquatic organisms: The importance of species identification. Water Environment Federation, 47, 9–19.

Rubach M.N., Ashauer R., Buchwalter D.B., De Lange H.J., Hamer M., Preuss T.G., et al. (2011) Framework for traits-based assessment in ecotoxicology. Integrated Environmental Assessment and Management, 7, 172–186.

Simpson J.C. & Norris R.H. (2000) Bioassessment of river quality: development of AUSRIVAS model and outputs. In: Assessing the Bioogical Quality of Freshwater RIVPACs and Other Techniques. (Eds J.F. Wright, D.W. Sutcliff & M.T. Furse), pp. 125–142. Freshwater Biological Association, Ambleside, UK.

Sota T. & Mogi M. (1992) Interspecific variation in desiccation survival time of Aedes (Stegomyia) mosquito eggs is correlated with habitat and egg size. Oecologia, 90, 353– 358.

Southwood T.R.E. (1977) Habitat, the templet for ecological strategies? Journal of Animal Ecology, 46, 337–365.

Southwood T.R.E. (1988) Tactics, strategies and templets. Oikos, 52, 3–18.

Stephan C.E., Mount D.I., Hansen D.J., Gentile J.R., Chapman G.A. & Brungs W.A. (1985) Guidelines for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. United States Environmental Protection Agency.

Van Straalen N.M. (1994) Biodiversity of ecotoxicological responses in animals. Netherlands Journal of Zoology, 441/442, 112–129.

152

Team R.C. (2013) R: A language and environment for statistical computing. http://www.R– project.org.

Tokeshi M. (1999) Species Coexistance. Blackwell Science, Ltd., Oxford.

Usseglio-Polatera P., Bournaud M., Richoux P. & Tachet H. (2000) Biological and ecological traits of benthic freshwater macroinvertebrates: relationships and definition of groups with similar traits. Freshwater Biology, 43, 175–205.

Verberk W.C.E.P., van Noordwijk C.G.E. & Hildrew A.G. (2013) Delivering on a promise: Integrating species traits to transform descriptive community ecology into a predictive science. Freshwater Science, 32, 531-547.

Wang T.-Q. & McCafferty W.P. (2004) Heptageniidae (Ephemeroptera) of the world. Part I: Phylogenetic higher classification. Transactions of the American Entomological Society, 130, 11–45.

Wang W.-X. & Fisher N.S. (1997) Modeling the influence of body size on trace element accumulation in the mussel Mytilus edulis. Marine Ecology Progress Series, 161, 103– 115.

Webb C.O., Ackerly D.D., Mcpeek M.A., Donoghue M.J. & Webb C. (2002) Phylogenies and community ecology. Annual Review of Ecology, Evolution, and Systematics, 33, 475–505.

Winner R., Boesel M. & Farrell M. (1980) Insect community structure as an index of heavy- metal pollution in lotic ecosystems. Canadian Journal of Fisheries and Aquatic Sciences, 37.

Zhang L. & Wang W.-X. (2007) Size-dependence of the potential for metal biomagnification in early life stages of marine fish. Environmental Toxicology and Chemistry, 26, 787– 794.

Zhou X., Robinson J.L., Geraci C.J., Parker C.R., Flint O.S., Etnier D. a., et al. (2011) Accelerated construction of a regional DNA-barcode reference library: Caddisflies (Trichoptera) in the Great Smoky Mountains National Park. Journal of the North American Benthological Society, 30, 131–162.

153

Table 1. Traits examined in this study for 42 species of aquatic insects.

Genus species Wet weight Cd ku Cd ke Cd BCF Reference (mg) (L g-1 d-1) (d-1) Isonychia sp. 48.2 0.028 0.019 1491 1 Isonychia tusculanensis 51.5 0.021 0.013 1658 1 Nixe sp. 9.8 0.25 0.070 3614 2 Maccaffertium ithaca 77.0 0.085 0.0078 10,910 1 Rhithrogena morrisoni 27.4 0.054 0.0010 54,000 1 Epeorus albertae 11.0 0.076 0.12 633 2 Dannella sp. 4.5 0.12 0.11 1091 3 Eurylophella verisimilis 6.2 0.54 0.017 31,765 3 Drunella grandis 157.4 0.32 0.12 2660 1 Drunella flavilinea 27.6 2.05 0.1 20,500 2 Drunella longicornis 6.4 0.78 0.01 78,000 3 Drunella cornutella 8 0.74 0.035 21,143 3 Drunella walkeri 7.2 0.28 0.19 1474 3 Drunella tuberculata 3.1 1.96 0.022 89,091 3 Ephemerella tibialis 6.9 2.60 0.20 13,000 2 Ephemerella rossi 3.9 0.30 0.019 15,789 3 Ephemerella excruciens 6.6 0.55 0.014 39,593 1 Ephemerella crenula 7.1 0.73 0.17 4294 3 Ephemerella subvaria 1.1 0.66 0.053 12,453 3 Ephemerella invaria 3.5 0.98 0.023 42,609 3 Ephemerella hispida 13.9 0.27 0.18 1500 3 Ephemerella catawba 5.0 0.77 0.16 4813 3 Pteronarcys dorsata 139.0 0.13 0.086 1462 1 Claassinea sabulosa 96.7 0.10 0.082 1214 1 Paragnetina sp. 92.9 0.16 0.085 1935 1 Doroneuria baumanni 335.6 0.030 0.084 354 1 Calineuria californica 105.1 0.10 0.081 1259 1 Acroneuria abnormis 195 0.21 0.087 2425 1 Hesperoperla pacifica 144.3 0.059 0.082 727 1 Skwala sp. 78.8 0.049 0.079 618 1 Isogenoides hansonni 108.5 0.027 0.11 243 1 Baumanella alameda 34.9 0.0085 0.051 168 1 Diplectrona modesta 10.4 0.27 0.075 3600 3 Parapsyche cardis 27.1 0.0098 0.042 233 3 Arctopsyche irrorata 60.7 0.07 0.075 933 3 Cheumatopsyche sp. 18.9 0.043 0.027 1579 1 Hydropsyche californica 24.4 0.42 0.20 2107 1 Hydropsyche sparna 5.8 0.37 0.085 4353 3 Hydropsyche alhedra 15.7 0.044 0.17 259 3 Hydropsyche slossonae 15.2 0.12 0.14 857 3 Rhyacophila sp. 34.2 0.042 0.14 309 1 Rhyacophila fuscula 69.7 0.011 0.12 92 1 References: 1 – (Buchwalter et al., 2008); 2 – (Cain et al., 2011); 3 – (Poteat & Buchwalter, 2014)

154

Table 1 Continued Genus species Order Family Feeding Reference guild Isonychia sp. Ephemeroptera Isonychiidae CF 1 Isonychia tusculanensis Ephemeroptera Isonychiidae CF 1 Nixe sp. Ephemeroptera Heptageniidae H 2 Maccaffertium ithaca Ephemeroptera Heptageniidae H 1 Rhithrogena morrisoni Ephemeroptera Heptageniidae CG 1 Epeorus albertae Ephemeroptera Heptageniidae CG 2 Dannella sp. Ephemeroptera Ephemerellidae CG 3 Eurylophella verisimilis Ephemeroptera Ephemerellidae CG 3 Drunella grandis Ephemeroptera Ephemerellidae H 1 Drunella flavilinea Ephemeroptera Ephemerellidae H 2 Drunella longicornis Ephemeroptera Ephemerellidae H 3 Drunella cornutella Ephemeroptera Ephemerellidae H 3 Drunella walkeri Ephemeroptera Ephemerellidae H 3 Drunella tuberculata Ephemeroptera Ephemerellidae H 3 Ephemerella tibialis Ephemeroptera Ephemerellidae CG 2 Ephemerella rossi Ephemeroptera Ephemerellidae CG 3 Ephemerella excruciens Ephemeroptera Ephemerellidae CG 1 Ephemerella crenula Ephemeroptera Ephemerellidae CG 3 Ephemerella subvaria Ephemeroptera Ephemerellidae CG 3 Ephemerella invaria Ephemeroptera Ephemerellidae CG 3 Ephemerella hispida Ephemeroptera Ephemerellidae CG 3 Ephemerella catawba Ephemeroptera Ephemerellidae CG 3 Pteronarcys dorsata Plecoptera Pteronarcyidae S 1 Claassinea sabulosa Plecoptera Perlidae P 1 Paragnetina sp. Plecoptera Perlidae P 1 Doroneuria baumanni Plecoptera Perlidae P 1 Calineuria californica Plecoptera Perlidae P 1 Acroneuria abnormis Plecoptera Perlidae P 1 Hesperoperla pacifica Plecoptera Perlidae P 1 Skwala sp. Plecoptera Perlodidae P 1 Isogenoides hansonni Plecoptera Perlodidae P 1 Baumanella alameda Plecoptera Perlodidae P 1 Diplectrona modesta Trichoptera Hydropsychidae CF 3 Parapsyche cardis Trichoptera Hydropsychidae CF 3 Arctopsyche irrorata Trichoptera Hydropsychidae CF 3 Cheumatopsyche sp. Trichoptera Hydropsychidae CF 1 Hydropsyche californica Trichoptera Hydropsychidae CF 1 Hydropsyche sparna Trichoptera Hydropsychidae CF 3 Hydropsyche alhedra Trichoptera Hydropsychidae CF 3 Hydropsyche slossonae Trichoptera Hydropsychidae CF 3 Rhyacophila sp. Trichoptera Rhyacophilidae P 1 Rhyacophila fuscula Trichoptera Rhyacophilidae P 1 CF = collector/filterer, H = herbivore, CG = collector/gatherer, S = shredder, P = predator References: 1 – (Buchwalter et al., 2008); 2 – (Cain et al., 2011); 3 – (Poteat & Buchwalter, 2014)

155

Table 2. Simple and multiple linear regression results for the effects of taxonomic- and trait- based parameters on Cd bioaccumulation parameters.

Cd ku Cd BCF Predictive Parameters adj. r2 p-value AIC adj. r2 p-value AIC Weight 0.33 <0.0001 69.27 0.29 0.0001 87.38 Genus 0.69 0.0008 46.69* 0.47 0.03 85.33 Genus + weight 0.70 0.001 45.63 0.58 0.008 74.30* Family 0.62 <0.0001 50.53 0.50 <0.0001 77.48 Family + weight 0.63 <0.0001 50.51 0.51 <0.0001 77.71 Order 0.26 0.0009 73.96 0.43 <0.0001 79.03 Order + weight 0.40 <0.0001 66.79 0.47 <0.0001 76.67 Feeding guild 0.41 <0.0001 66.49 0.48 <0.0001 77.12 Feeding guild + weight 0.46 <0.0001 63.54 0.48 <0.0001 78.04 Feeding guild + weight + 0.49 <0.0001 63.02 0.49 <0.0001 78.19 order

156

Figures

Figure 1. Conceptual model of how trait values can vary within clades (species groups). Circles represent an average trait value for a given species within a clade. Dotted lines are global average values for each clade. A) Clades 1 and 2 have trait value ranges which vary in magnitude. Overlap in the ranges of trait values in clades 1 and 2 creates a situation where species sharing a comparable trait state could represent the high end for one lineage, and the low end of another. B) Clades 3 and 4 have similar average trait values, but the variance around the mean trait is highly variable.

157

a b c Is o n y c h iid a e H e p ta g e n iid a e E p h e m e re llid a e P te ro n a rc id a e P e rlid a e P e rlo d id a e H y d ro p s y c h id a e R h y a c o p h ilid a e

0 1 2 3 4 5 1 2 3 0 5 0 5 0 5 0 0 0 0 0 0 0 0 ...... 0 0 1 1 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 2 4 0 0 0 0 0 2 4 6 8 0 1 -1 -1 -1 C d k u (L g d ) C d k e (d ) C d B C F

Figure 2. Dissolved Cd (a) ku (b) ke and (c) BCF values for 42 species of aquatic insects. Species are divided into families, and each symbol is a different species. Lines are mean familial values.

158

CHAPTER 6

Evolutionary patterns in trace metal (Cd and Zn) efflux capacity in aquatic organisms

Monica D. Poteat1, Theodore Garland, Jr.2, Nicholas S. Fisher,3 Wen-Xiong Wang,4 and David B. Buchwalter1

1Department of Environmental and Molecular Toxicology, North Carolina State University, Raleigh, NC 27695, United States

2Department of Biology, University of California, Riverside, California 92521, United States

3School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York, 11794, United States

4Division of Life Science, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong

Published In: Environmental Science and Technology. 2012; Volume 215; pages 1575- 1583.

159

Abstract

The ability to eliminate (efflux) metals is a physiological trait that acts as a major driver of bioaccumulation differences among species. This species-specific trait plays a large role in determining the metal loads that species will need to detoxify in order to persist in chronically contaminated environments, and therefore contributes significantly to differences in environmental sensitivity among species. To develop a better understanding of how efflux varies within and among taxonomic groupings, we compared Cd and Zn efflux rate constants

(kes) among members of two species-rich aquatic insect families, Ephemerellidae and

Hydropsychidae, and discovered that kes strongly covaried across species. This relationship allowed us to successfully predict Zn efflux from Cd data gathered from aquatic species belonging to other insect orders and families. We then performed a broader, comparative analysis of Cd and Zn kes from existing data for arthropods, mollusks, annelids and chordates

(77 species total) and found significant phylogenetic patterns. Taxonomic groups exhibited marked variability in ke magnitudes and ranges, suggesting that some groups are more constrained than others in their abilities to eliminate metals. Understanding broader patterns of variability can lead to more rational extrapolations across species and improved protectiveness in water quality criteria and ecological assessment.

160

Introduction

The related fields of applied ecotoxicology and ecological bioassessment are challenged by the tremendous biodiversity that exists in nature. The root of this challenge is a fundamental lack of understanding of how patterns of sensitivity to environmental stressors are distributed within and among taxonomic groupings and phylogenetic lineages. In the case of applied ecotoxicology, the widespread use of surrogate species to represent entire taxonomic groups1 largely stems from the practical reality that a relatively small set of available and convenient test species with standardized test methods have been developed for laboratory use. Regulatory entities are largely forced to assume that each test species represents its particular group, even when there is evidence to the contrary.2 The regulatory community generally accepts that when data from these species are combined, a distribution of sensitivities can be used to inform decisions about how best to protect real communities,3–6 but the need for improvements has been noted.7

In the ecological bioassessment of real communities, the lumping of species, whether based on taxonomy (phylogenetic relationships) or function, is borne out of practical challenges, including both taxonomic uncertainties and a conceptual vagueness regarding how the presence of an organism in a sample represents larger groups. Our lack of clarity on the general issue of representativeness introduces major uncertainty in the final products of both applied ecotoxicology (pollution limits and standards) and ecological bioassessment as mandated by the Clean Water Act (US) and related statutes.

One way to explore or refine our concept of representativeness is by explicitly examining inter-specific variability in a trait of interest (e.g. a complex trait such as

161

sensitivity to a contaminant, or subordinate traits that contribute to sensitivity) in defined species groupings.8 By doing so, we can begin to ask questions relevant to environmental protection and/or assessment: How much does a trait of interest vary across members of a taxonomic grouping? How variable are the magnitudes and ranges of values for a particular trait across different groups? What is the likelihood that a given species represents the central tendencies or "outliers" of a given taxonomic grouping? Does variance parse phylogenetically? Can we identify emergent patterns that might eventually lead to rational extrapolation and prediction?

Here, we take a modest first step in attempting to answer these questions by examining variability within and among taxonomic groups in two physiological traits relevant to trace metal pollution. Specifically, we examine the ability of aquatic species to

9 eliminate Cd and Zn from tissues as measured by the efflux rate constant (ke). Although not a proxy for metal sensitivity per se, this loss trait (ke) has emerged from the biodynamic modeling literature as a major driver of metal bioaccumulation differences among co-

10 10,11 occurring species. These differences can be profound, and as such, the ke is a primary determinant of the metal loads that a given species will need to detoxify or store in order to persist in a given exposure scenario.

In this study, we explore patterns in Zn and Cd loss (efflux) across aquatic organisms.

Zn and Cd were utilized for this study due to their co-occurrence in nature, covariance in bioaccumulation parameters,12–15 and abundance of values in the literature. We examined patterns of measured kes for Cd and Zn among several members of two common and species- rich aquatic insect families (Ephemerellidae and Hydropsychidae). We compared the ranges

162

of Cd and Zn kes within and between families and asked if Cd ke values could be used to predict Zn ke values in other aquatic insect species. We also searched for patterns in Zn and

Cd kes across and within other taxonomic groupings with an exhaustive compilation of values from the literature.

163

Materials and Methods

Insect collection and handling

Insect larvae were field collected from streams in Great Smoky Mountains National

Park (GRSM) and Basin Creek (BC) (all from North Carolina, USA) from July 2010 to

August 2012 using a D-frame kicknet. All larvae were collected from cool, cobble-bottomed, riffle-pool type streams. Larvae were transported to the laboratory and acclimated in a manner described previously.12 Only larvae that appeared healthy were used for experimentation. All species identifications were verified by taxonomic experts.

Collecting efforts focused upon the species-rich families Hydropsychidae (Order:

Trichoptera) and Ephemerellidae (Order: Ephemeroptera) to enable comparisons among close relatives. Hydropsychid species are net-spinning collectors and had average sizes ranging from 10.5 to 67.3 mg. Ephemerellid species are collector-gatherers and ranged in average size from 3.4 to 7.4 mg.

Determination of aquatic insect efflux rate constants

Efflux rate constants (kes) for all aquatic insects were determined using gamma- emitting radioisotopes 65Zn and 109Cd as described elsewhere.8,16,17 In vivo gamma counting is non-destructive, enabling the repeated measurements of the same individual organisms. All experimental procedures utilized American Society for Testing Materials artificial very soft

-1 water (ASTM VSW) (mg L : 12 NaHCO3, 7.5 CaSO4·2H2O, 7.5 MgSO4, 0.5 KCl) because the native streams where insects were collected have very low Ca content. All experiments were performed at 12.7 ± 0.5ºC. Most experiments utilized dual-metal exposures because

164

previous work with Hydropsyche sparna showed no difference in single or dual metal exposures with the concentrations utilized. Larvae (5-10 individuals per species) were exposed to 102 kBq L-1 Zn and 29.5 kBq L-1 Cd with the remainder of the metal being stable

Zn in the form of ZnCl2 and stable Cd in the form of CdCl2. Total metal concentrations

(stable metal plus tracer) were 3 µg L-1 Zn and 0.3 µg L-1 Cd. Environmentally, background concentrations rarely exceed 40 µg L-1 for Zn11,18,19 and 0.5 µg L-1 for Cd,20,21 however these concentrations can be exceeded at anthropomorphically effected sites. Therefore, our exposure concentrations were environmentally relevant. Exposures lasted for 4-5 days to ensure that adequate radiolabel would be retained in tissues following 10 days of efflux and that accumulated tissue concentrations would not cause overt toxicity. The pH of each bulk solution was adjusted to 7.20 ± 0.02 using 0.1 N NaOH. Exposures occurred in aerated solutions in high-density polyethylene (HDPE) cups containing Teflon® mesh as substrate.

Parafilm® covers minimized evaporative loss.

After exposure to the dissolved solutions, we ensured that larvae had acquired adequate signal from each isotope by rinsing them with VSW and assaying in vivo for radioactivity in scintillation vials containing 15 mL VSW. Larvae were then placed in individual aerated 1 L HDPE containers containing 500 mL VSW, Teflon® mesh as substrate, and Parafilm® to reduce evaporative loss. Each replicate consisted of a single larva, with 5-10 replicates per species. Larvae were assayed for radioactivity daily for 10 days and returned to their individual VSW containers. Spot checking of the solutions for radioactivity ensured that re-uptake of released metals did not confound the interpretation of results. After insects were assayed on the 10th day, larvae were blotted dry and wet weights were

165

determined. The ke was determined as the slope of the natural log of the proportion of metal retained in the body tissue and the time of depuration22 after excluding days 0 and 1 as follows:

Where = tissue concentration at time (day) t

= initial tissue concentration

t = time (days)

Predictions of kes in aquatic insects

To test whether the ke of one metal was predictable based on the ke of the other, we constructed a regression and 95% confidence interval of Zn ke vs Cd ke from the first 11 insect species tested (all from GRSM). We then obtained Cd ke values for six additional species. Cd kes for three of these species (Isonychia tuscalanensis (Ephemeroptera:

Isonychidae), Acroneuria abnormis (Plecoptera: Perlidae), and Isogenoides hansoni

(Plecoptera: Perlodidae)) were previously published,8 and the other three (Alloperla sp.

(Plecoptera: Peltoperlidae), Hydropsyche sparna (Trichoptera: Hydropsychidae), and

Hydropsyche alhedra (Trichoptera: Hydropsychidae)) were generated for this study. For experimentation, H. alhedra and H. sparna were collected from GRSM while the remainder were collected from BC. Based on these Cd ke values, we predicted Zn kes based on the regression line of the first 11 species for which we obtained Zn and Cd kes. These predictions

65 were then tested by measuring Zn kes using Zn radiotracer in the same manner described above (but independent of Cd).

166

Data compilation for other taxa

Zn and Cd kes were compiled from the literature for several aquatic species representing chordates, mollusks, arthropods, and annelids. Because much more data were available for kes utilizing dissolved exposures, we preferred those values over dietary values when both were available. However, we did include dietary data when no dissolved data were available, provided that the dietary exposures did not result from single-pulse feedings

(in order to analyze only true physiological loss). When multiple kes were reported for a given species, we used a mean value.

Data analysis

All conventional statistical analyses were performed using GraphPad Prism (v5.04).

Results were considered statistically significant with a p-value < 0.05. Regression analyses of kes from insects were performed using untransformed data, while data were log-transformed for comparisons between taxonomic groups.

Phylogenetic methods were employed to analyze log-transformed ke values across the

13 aquatic insect species gathered for this study as well as across all aquatic organisms.

Whereas conventional statistics treat species as independent and assume complete unrelatedness, phylogenetic methods allow for the comparison of species while taking into account their evolutionary relatedness.23 Blomberg et al.s K statistic24 was used to quantify phylogenetic signal (the tendency for related species to resemble each other), and their randomization test based on the mean squared error was used to test for the significance of phylogenetic signal. Phylogenetically independent contrasts were used to detect covariation

167

25,26 of Zn and Cd kes. Because body weight had no significant effect on the analyses, log- transformed traits were not corrected allometrically. See Supplemental Material for more information on utilized phylogenetic methods.

168

Results

Patterns of Zn and Cd kes across aquatic insects

Zn and Cd kes across 13 aquatic insect species from families Ephemerellidae and

Hydropsychidae ranged over an order of magnitude for both Zn (0.013-0.15 d-1) and Cd

-1 (0.010-0.17 d ) (Fig. 1). Across all species, Zn and Cd kes were strongly correlated using conventional statistics (r = 0.93, p<0.0001). Body weight was not a significant driver in ke differences among species. Whether data were log transformed or analyzed raw, correlations of Cd or Zn and body weight were not statistically significant and ranged from -0.36 – 0.15 in ephemerellids, -0.05 – 0.47 in hydropsychids, and 0.24-0.40 in all taxa. This finding suggests that species-specific physiological differences dwarfed any possible influence of body weight.

Differences in ke values among species did not correlate with body weight within ephemerellids, hydropsychids or across all taxa, suggesting that species-specific physiological differences dwarfed any possible influence of body weight.

Across 13 species of aquatic insects, Cd ke values exhibited significant phylogenetic signal (the tendency for related species to resemble each other) (K = 0.86, p = 0.013) (Fig. 2

& Table S3). Although phylogenetic signal did not reach statistical significance for Zn ke values (K =0.67, p = 0.078), this lack of significance is most likely an artifact of the small

24 sample size. Phylogenetically, Zn and Cd ke values were strongly correlated (rIC = 0.90, p <

0.0001).

-1 -1 Among 7 ephemerellids, most species’ kes fell between 0.013 d and 0.057 d for Zn, and 0.010 d-1 and 0.035 d-1 for Cd. Ephemerella catawba was a clear outlier, with Zn and Cd

169

-1 -1 kes of 0.15 d and 0.16 d , respectively. Excluding E. catawba, the mean (± s.d.) kes for Zn and Cd were 0.036 ± 0.019 d-1 and 0.021 ± 0.008 d-1 respectively. Among 6 hydropsychids,

-1 -1 -1 Zn ke values ranged from 0.028 d to 0.13 d , while Cd ke values ranged from 0.042 d to

-1 0.14 d . The mean (± s.d.) ke values across hydropsychids for Zn and Cd were 0.083 ± 0.034 d-1 and 0.098 ± 0.048 d-1, respectively. Excluding E. catawba from the analysis, it is clear that on average ephemerellids eliminate Zn (p = 0.014) and Cd (p = 0.003) from tissues much more slowly than hydropsychids, despite the huge variances around the means for these groups.

Predictions of Zn kes in aquatic insects

A regression of Zn ke on Cd ke based on the first 11 species we tested (Fig. 1) resulted in a line with the form:

2 Zn(ke) = 0.84[Cd(ke)] + 0.014 (r = 0.92, p < 0.0001).

The strength of this regression led us to test whether Zn ke values could be predicted based on Cd ke values from a suite of other aquatic insect species. Predictions of Zn ke values using the linear regression of Zn ke vs Cd ke fell within the 95% confidence interval of the linear regression in 5 of the 6 EPT (Orders: Ephemeroptera, Plecoptera, Trichoptera) species tested

(Fig. 1). Of the 5 species that fell within the 95% confidence interval, the experimental measurements deviated between 2.9-25% from the predicted Zn ke values (I. hansoni, 2.9%;

H. sparna, 3.2%; A. abnormis, 4.1%; I. tuscalanensis, 23%; Alloperla sp., 25%). Although experimental values for I. tuscalanensis and Alloperla sp. differed the most from their predicted Zn ke values, the difference between their predicted and experimental Zn ke values

170

was less than 1% per day in both cases (only 0.007 d-1 and 0.009 d-1, respectively). H. alhedra did not fall within the 95% confidence interval, but it also outside of the range of Cd ke values used to generate the regression. H. sparna ke values for Zn and Cd were identical for dual labeled (Cd+ Zn) and individually labeled metal experiments (p > 0.05).

Based on the success of this prediction exercise, we predicted Zn ke values for an additional 22 species of EPT taxa based on published ke values for Cd (Table S1). Data for other aquatic insect orders (Diptera and Megaloptera) were available; however, we only carried out predictions for EPT taxa due to our success in predicting Zn ke values for EPT taxa previously.

Taxonomic group comparisons

To explore broad patterns in ke values for other taxonomic groupings, we compiled

Zn and/or Cd ke values for all aquatic taxa available in the literature. Including the data generated as part of this study for insects, we compiled data from 77 species representing four phyla (Zn ke values for 45 species, Cd ke values for 75 species, and ke values for both metals for 43 species (Tables S1 and S2)).

Aquatic insects had the most data available, allowing us to analyze the variance among species within taxonomic rankings ranging from genus to order. Mayflies (Order:

Ephemeroptera) had the most variance of the three orders analyzed, and the variance remained high across all levels of taxonomic rank. Within 2 genera, Ephemerella (Family:

Ephemerellidae) and Drunella (Family: Ephemerellidae) each showed high variances of Cd ke values with coefficients of variation (CVs) of 100% (n=5) and 86% (n=5), respectively.

171

Within the family Ephemerellidae, CVs for Cd and Zn ke values were 96% (n=12) and 89%

(n=7), respectively. Across Ephemeroptera, Cd and Zn ke values had CVs of 99% (n=19) and

100% (n=9), respectively.

Caddisflies (Order: Trichoptera) showed less variance in efflux than mayflies. Within genus Hydropsyche (Family: Hydropsychidae), Cd ke values had a CV of 31% (n=5). Within

Hydropsychidae, Cd and Zn ke values had CVs were 53% (n=9) and 42% (n=6), respectively.

Finally, Cd and Zn ke values for Trichoptera had CVs of 53% (n=11) and 42% (n=6), respectively.

Stoneflies (Order: Plecoptera) showed the least variance overall across taxonomic ranks that we were able to compare. In the family Perlidae, Cd ke values exhibited a CV of only 3% (n=6). Overall, Plecoptera had Cd ke values that exhibited a CV of 28% (n=11), again, much lower than other order-level variances among aquatic insects.

At the phylum level of classification, 53 Arthropoda species (10 orders, 18 families) exhibited the most variability of Zn and Cd ke values with CVs of 82% and 80%, respectively. Zn ke values for arthropods ranged 100-fold (Fig. 3A), and Cd ke values ranged

297-fold (Fig. 3B). Arthropods also exhibited the largest mean kes. Arthropoda had a mean

-1 -1 Zn ke value of 0.072 d and a mean Cd ke of 0.080 d .

Across 13 species (5 orders, 10 families), Mollusca had the lowest CVs for both Zn and Cd ke values at 65% and 55%, respectively. Zn ke values ranged 16-fold (Fig. 3A), and

Cd ke values ranged 25.4-fold (Fig. 3B). Mollusca had the lowest average Zn and Cd ke values of 0.022 d-1 and 0.015 d-1, respectively.

172

Chordata, across 8 species (4 orders, 8 families), had CVs for Zn and Cd ke values of

75% and 70%, respectively. Zn ke values ranged 8.4-fold (Fig. 3A), and Cd ke values ranged

-1 -1 29.6-fold (Fig. 3B). Chordates had a mean Zn ke of 0.024 d and mean Cd ke of 0.038 d .

Across all four phyla, Arthropoda had significantly higher Zn ke values than both

Mollusca and Chordata (p=0.003), as well as significantly higher Cd ke values than Mollusca

(p=0.0002). Annelida did not have adequate representation to analyze trends in Zn and Cd ke values, with only 2 and 3 species’ values available, respectively (Fig. 3A and 3B).

Phylogenetic signal was detected for Zn ke values across 45 species (K =0.47, p=0.014) and

Cd ke values across 75 species (K=0.59, p<0.0001), signifying that both Zn and Cd ke values agree with phylogeny (Table S3).

Correlations between Cd and Zn kes

Efflux rate constants (log transformed) for Zn and Cd in 17 species of aquatic insects

(all available insect values) exhibited a strong covariance (r = 0.92, p < 0.0001). The inclusion of four non-insect arthropods (Subphylum: Crustacea) in the analysis still showed significant correlation (r = 0.81, p < 0.0001) as did the inclusion of 12 molluscs, 8 chordates and 2 annelids (43 species total) (r=0.73, p<0.0001). A phylogenetic correlation of Zn and

Cd ke values across all 43 taxa was similar (rIC=0.61, p<0.0001). Efflux rate constants within

12 species of Mollusca and within 8 species of Chordata failed to correlate significantly, but an analysis of all 26 non-insect species showed a weaker correlation (r=0.48, p = 0.01).

173

Discussion

Here, we used metal efflux as a starting point to examine patterns of interspecific variability in a physiological trait that pertains to metal accumulation. Proximal sensitivity to metals is a complex trait, encompassing bioaccumulation (uptake and efflux), detoxification capacity 27,28 and target site sensitivity (which remains poorly studied).28 Other demographic traits related to recovery and dispersion also contribute to population persistence.29,30

However, the ability to eliminate (efflux) acts as a major driver of bioaccumulation differences among species and plays a large role in determining the metal loads that species will need to detoxify in contaminated environments.

One goal of this study was to generate a dataset of closely related species within 2 species-rich insect families to determine whether phylogenetic signal previously observed at more coarse taxonomic levels8 is retained at finer taxonomic levels. Despite finding a surprisingly wide range of ke values within the Ephemerellidae and Hydropsychidae, we found evidence of statistically significant phylogenetic signal within the ke values. Blomberg et al. found that the significance of phylogenetic signal was strongly related to the number of species used in the study.24 For instance, 92% of studies with > 20 species showed significant phylogenetic signal, while only 41% of studies with < 20 studies showed significant phylogenetic signal.24 The analyses performed here only had values for 13 species, thus decreasing the power of the statistical tests. However, these preliminary tests show that phylogenetic signal does exist at the genus/species levels of taxonomic rank in metal bioaccumulation parameters.

174

Using conventional statistics, we found these values covaried strongly enough to allow us to accurately predict Zn ke values from existing Cd ke values for 5 out of 6 additional aquatic insect species including large (average 142 mg) predatory and small

(average 5 mg) shredder stoneflies as well as other mayfly and caddisfly species. This strong pattern of covariation among insect Zn and Cd ke values suggests similar trafficking and binding within subcellular pools of the animals. This is the first time to our knowledge that a correlation between the ke values of 2 metals has been used for the prediction of an unknown metal bioaccumulation value.

A second goal of this study was to explore patterns of previously published Cd and

Zn ke values across broader taxonomic groups. We limited our analysis to efflux values in this study for several reasons. Ke values generally are more robust to exposure route differences,9,31,32 exposure concentration,33,34 and water chemistry10,35 than are other metal bioaccumulation parameters. Robust comparisons of other bioaccumulation parameters (e.g. ku) would require a more standardized exposure regime than was present in the available literature (but see 8). Several patterns emerged from our analysis. The ability of organisms to efflux metal differed within taxonomic groupings, and the ranges of efflux values within and across taxonomic groups also varied. Despite this, we still observed significant phylogenetic signal in both Cd (K = 0.59) and Zn (K = 0.47) ke values. We also found a strong covariation using both conventional (r = 0.73) and phylogenetic (r = 0.61) statistics. The phylogenetic signal present in Zn and Cd ke values indicates that phylogenetic statistics should be used to analyze the data further,23,26 therefore, further analyses or extrapolations should rely on phylogenetic statistics.

175

We note that inter-laboratory differences in experimental procedures can influence physiological parameters and could have introduced variance into our compiled dataset. In particular differences in laboratory temperatures likely influenced metabolic rate, which in turn has been shown to affect physiological traits dictating metal accumulation,36,37 including

38 efflux. For ke values obtained in this study for aquatic insects, all insects were acclimated to the same experimental temperatures (12.7°C). However, amongst all values obtained in the literature, temperature was variable, ranging from 10-25°C. This range of experimental temperatures likely contributed to wider physiological variation within the compiled dataset.

Overall, it is clear that taxonomic groups have different ranges (capacities) with respect to their ability to eliminate Cd and Zn from tissues. For example, the fastest ke values in mollusks would be considerably below the average rate for aquatic insects. Species only tolerate metal exposures through a combination of effluxing and detoxifying or storing the metal.28 Previous studies in insects revealed an inverse relationship between the ability to eliminate and detoxify Cd8 (e.g., species that stored metal effectively did not efflux quickly).

It remains unclear whether slow metal efflux is the result of metal being tightly bound

(detoxified) in physiologically inert structures or if stronger detoxification capacity arises when organisms are poor effluxers. At a subcellular level, caddisfly species store significantly less metal (Cd) in metallothionein-like proteins (MTLPs) than mayfly species17 while maintaining higher ke values (as seen in this study). Mollusks are known to accumulate extremely high concentrations of metals, likely a result of their overall low rates of efflux and abilities to sequester Cd and Zn in metal rich granules.22,39,40

176

We also observed that variances do not scale similarly within a given taxonomic rank.

For instance, there is more variation among ke values in some individual insect genera than we observed within the whole mollusk phylum. Numerous potential sources of variability occur among taxonomic groupings (e.g., phyla) that remain poorly explored. It may be the case that physiological variance increases with the species richness (evolutionary radiation).

Arthropods are by far the most species-rich phylum, with a species richness estimated at between 5-10 million species (including both aquatic and terrestrial species).41 Mollusca has a total estimated species richness of 50,000 to 200,000.42 However, within insects we observe that mayflies vary more widely than do stoneflies despite both groups being comparable with respect to lineage age and biodiversity. It appears as though age of a given evolutionary lineage (clade) is not a good predictor of physiological variability. Mollusks (545 mya43), mayflies (290 mya44) and stoneflies (360 mya45) are each ancient lineages, but only mayflies showed widely varying kes – corresponding to a wider array of morphological bauplans

(body plans). We have no explanation for the presence of outliers (e.g. E. catawba) based on ecology, morphology or life history.

If variation in other more complex traits (e.g., sensitivity to a particular contaminant) is similarly distributed, then it calls to question whether taxonomic rank is a reasonable criterion on which to base practical decisions (e.g., use of single representatives of 8 families1 for the development of water quality criteria and the use of family-level taxonomic work in bioassessments). The ability of a single species to adequately represent a larger taxonomic group is inversely related to the amount of variance that occurs in that taxonomic group, unless it is known a priori that the particular representative is at the sensitive

177

(protective) end of the range. It is more likely, for example, that a randomly selected mollusk is reasonably close to the mollusk mean than a randomly selected insect would be to the insect mean for these ke measures. It is clear that species-rich groups (such as insects) cannot be adequately represented by single test species.

By exploring variation within and across taxonomic groups we can begin to develop a better understanding of the issue of representativeness which is so deeply embedded in practical environmental management and assessment. Examining the variance that occurs in this one subordinate trait begins to give us an idea of just how much variation exists in the physiological traits that influence sensitivity of aquatic organisms. Although it is impossible to test toxicological characteristics of every species for every contaminant, there are likely larger patterns (and hypotheses) that can be gleaned from existing data. Ultimately, incorporation of phylogenetically based methodologies to explore broader patterns among species has the potential to improve our ability to make rational extrapolations across species and taxonomic groups and lead to more robust environmental standards and ecological assessments.

178

Acknowledgements

The authors would like to thank Luke Jacobus (IUPUC) and Eric Fleek (NC DENR) for their taxonomic expertise. Gerald LeBlanc (NCSU), Tom Augspurger (USGS), Eric

Stone (NCSU) and anonymous reviewers provided valuable editorial comments. This work was supported by NSF (IOS 0919614).

179

References

(1) Stephan, C. E.; Mount, D. I.; Hansen, D. J.; Gentile, J. R.; Chapman, G. A.; Brungs, W. A. Guidelines for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. United States Environmental Protection Agency. 1985. http://water.epa.gov/scitech/swguidance/standards/criteria/current/upload/2009_01_13 _criteria_85guidelines.pdf.

(2) Koivisto, S. Is Daphnia magna an ecologically representative zooplankton species in toxicity tests? Env. Pollut. 1995, 90, 263–267.

(3) Solomon, K. R.; Baker, D. B.; Richards, R. P.; Dixon, K. R.; Klaine, S. J.; La Point, T. W.; Kendall, R. J.; Weisskopf, C. P.; Giddings, J. M.; Giesy, J. P.; et al. Ecological risk assessment of atrazine in North American surface waters. Env. Toxicol. Chem. 1996, 15, 31–76.

(4) Hall, Jr., L. W.; Scott, M. C.; Killen, W. D. Ecological risk assessment of copper and cadmium in surface waters of Chesapeake Bay watershed. Env. Toxicol. Chem. 1998, 17, 1172–1189.

(5) Hose, G. C.; Van den Brink, P. J. Confirming the species-sensitivity distribution concept for endosulfan using laboratory, mesocosm, and field data. Arch. Env. Con. Tox. 2004, 47, 511–520.

(6) Maltby, L.; Blake, N.; Brock, T. C. M.; van den Brink, P. J. Insecticide species sensitivity distributions: Importance of test species selection and relevance to aquatic ecosystems. Env. Toxicol. Chem. 2005, 24, 379–388.

(7) Forbes, V. E.; Calow, P. Species sensitivity distributions revisited: A critical appraisal. Hum. Ecol. Risk. Assess. 2002, 8, 473–492.

(8) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.; Garland, Jr., T. Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. P. Natl. Acad. Sci. USA. 2008, 105, 8321–8326.

(9) Wang, W.-X.; Fisher, N. S.; Luoma, S. N. Kinetic determinations of trace element bioaccumulation in the mussel Mytilus edulis. Mar. Ecol.- Prog. Ser. 1996, 140, 91– 113.

(10) Luoma, S. N.; Rainbow, P. S. Why is metal bioaccumulation so variable? Biodynamics as a unifying concept. Env. Sci. Technol. 2005, 39, 1921–1931.

180

(11) Cain, D. J.; Luoma, S. N.; Wallace, W. G. Linking metal bioaccumulation of aquatic insects to their distribution patterns in a mining-impacted river. Environ. Toxicol. Chem. 2004, 23, 1463–1473.

(12) Poteat, M. D.; Díaz-Jaramillo, M.; Buchwalter, D. B. Divalent metal (Ca, Cd, Mn, Zn) uptake and interactions in the aquatic insect Hydropsyche sparna. J. Exp. Biol. 2012, 215, 1575–1583.

(13) Buchwalter, D. B.; Luoma, S. N. Differences in dissolved cadmium and zinc uptake among stream insects: Mechanistic explanations. Env. Sci. Technol. 2005, 39, 498– 504.

(14) Wang, W.-X.; Fisher, N. S. Effects of calcium and metabolic inhibitors on trace element uptake in two marine bivalves. J. Exp. Mar. Biol. Ecol. 1999, 236, 149–164.

(15) Rainbow, P. S. Physiology, physicochemistry and metal uptake - A crustacean perspective. Mar. Pollut. Bull. 1995, 31, 55–59.

(16) Martin, C. A.; Cain, D. J.; Luoma, S. N.; Buchwalter, D. B. Cadmium ecophysiology in seven stonefly (Plecoptera) species: Delineating sources and estimating susceptibility. Env. Sci. Technol. 2007, 41, 7171–7177.

(17) Buchwalter, D. B.; Cain, D. J.; Clements, W. H.; Luoma, S. N. Using biodynamic models to reconcile differences between laboratory toxicity tests and field biomonitoring with aquatic insects. Env. Sci. Technol. 2007, 41, 4821–4828.

(18) Eisler, R. Zinc hazards to fish, wildlife, and invertebrates: A synoptic review. In Biological Report 10. 1993. Washington, DC: US Department of Interior, Fish and Wildlife Service.

(19) Van Genderen, E.; Adams, A.; Cardwell, R.; Volosin, J.; Santore, R.; Rodriguez, P. An evaluation of the bioavailability and aquatic toxicity attributed to ambient zinc concentrations in fresh surface waters from several parts of the world. Integr. Environ. Assess. Manag. 2009, 5, 426–434.

(20) Jensen, A.; Bro-Rasmussen, F. Environmental cadmium in Europe. Rev. Environ. Contam. Toxicol. 1992, 125, 101–181.

(21) Xue, H.; Sigg, L. Cadmium speciation and complexation by natural organic ligands in fresh water. Anal. Chim. Acta. 1998, 363, 249–259.

(22) Blackmore, G.; Wang, W.-X. Uptake and efflux of Cd and Zn by the green mussel Perna viridis after metal preexposure. Env. Sci. Technol. 2002, 36, 989–995.

181

(23) Garland, T.; Bennett, A. F.; Rezende, E. L. Phylogenetic approaches in comparative physiology. J. Exp. Biol. 2005, 208, 3015–3035.

(24) Blomberg, S. P.; Garland, Jr., T.; Ives, A. R. Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution. 2003, 57, 717–745.

(25) Garland, Jr., T.; Midford, P. E.; Ives, A. R. An introduction to phylogenetically based statistical methods, with a new method for confidence intervals on ancestral states. Am. Zool. 1999, 39, 374–388.

(26) Garland Jr., T.; Ives, A. R. Using the past to predict the present: Confidence intervals for regression equations in phylogenetic comparative methods. Am. Nat. 2000, 155, 346–364.

(27) Rainbow, P. S. Trace metal bioaccumulation: Models, metabolic availability and toxicity. Env. Int. 2007, 33, 576–582.

(28) Rainbow, P. S. Trace metal concentrations in aquatic invertebrates: Why and so what? Env. Pollut. 2002, 120, 497–507.

(29) Rubach, M. N.; Ashauer, R.; Buchwalter, D. B.; De Lange, H. J.; Hamer, M.; Preuss, T. G.; Topke, K.; Maund, S. J. Framework for traits-based assessment in ecotoxicology. Integ. Env. Assess. Manag. 2011, 7, 172–186.

(30) Van Straalen, N. M. Biodiversity of ecotoxicological responses in animals. Neth. J. Zool. 1994, 441/442, 112–129.

(31) Fisher, N. S.; Teyssié, J.-L.; Fowler, S. W.; Wang, W.-X. Accumulation and retention of metals in mussels from food and water: A comparison under field and laboratory conditions. Env. Sci. Technol. 1996, 30, 3232–3242.

(32) Lee, B.-G.; Wallace, W. G.; Luoma, S. N. Uptake and loss kinetics of Cd, Cr and Zn in the bivalves Potamocorbula amurensis and Macoma balthica: Effects of size and salinity. Mar. Ecol.- Prog. Ser. 1998, 175, 177–189.

(33) Yu, R.-Q.; Wang, W.-X. Trace metal assimilation and release budget in Daphnia magna. Limnol. Ocean. 2002, 47, 495–504.

(34) Guan, R.; Wang, W. Dietary assimilation and elimination of Cd, Se, and Zn by Daphnia magna at different metal concentrations. Env. Toxicol. Chem. 2004, 23, 2689–2698.

182

(35) Ni, I.-H.; Chan, S. M.; Wang, W.-X. Influences of salinity on the biokinetics of Cd, Se, and Zn in the intertidal mudskipper Periophthalmus cantonensis. Chemosphere. 2005, 61, 1607–1617.

(36) Baines, S. B.; Fisher, N. S.; Kinney, E. L. Influence of temperature on dietary metal uptake in Arctic and temperate mussels. Mar. Ecol.- Prog. Ser. 2005, 289, 201–213.

(37) Van Campenhout, K.; Bervoets, L.; Blust, R. Assimilation efficiencies of Cd and Zn in the common carp (Cyprinus carpio): Effects of metal concentration, temperature and prey type. Env. Pollut. 2007, 145, 905–914.

(38) Baines, S. B.; Fisher, N. S. Modeling the effect of temperature on bioaccumulation of metals by a marine bioindicator organism, Mytilus edulis. Env. Sci. Tecnnol. 2008, 42, 3277–3282.

(39) Pan, K.; Wang, W.-X. Reconstructing the biokinetic processes of oysters to counteract the metal challenges: Physiological acclimation. Env. Sci. Technol. 2012, 46, 10765– 10771.

(40) Bonneris, E.; Giguere, A.; Perceval, O.; Buronfosse, T.; Masson, S.; Hare, L.; Campbell, P. G. C. Sub-cellular partitioning of metals (Cd, Cu, Zn) in the gills of a freshwater bivalve, Pyganodon grandis: Role of calcium concretions in metal sequestration. Aquat. Toxicol. 2005, 71, 319–334.

(41) Ødegaard, F. How many species of arthropods? Erwin’s estimate revised. Biol. J. Linn. Soc. 2000, 71, 583–597.

(42) Lydeard, C.; Cowie, R. H.; Ponder, W. F.; Bogan, A. E.; Bouchet, P.; Clark, S. A.; Cummings, K. S.; Frest, T. J.; Gargominy, O.; Herbert, D. G.; et al. The global decline of nonmarine mollusks. Bioscience. 2010, 54, 321–330.

(43) United States Geologic Survey. Mollusks. U.S. Dep. Inter. 1999.

(44) Brittain, J. E.; Sartori, M. Ephemeroptera (Mayflies). In Encyclopedia of Insects; Resh, W. H.; Carde, R. T., Eds.; Academic Press: San Diego, 2003; pp. 373–380.

(45) Hagner-Holler, S.; Pick, C.; Girgenrath, S.; Marden, J. H.; Burmester, T. Diversity of stonefly hexamerins and implications of insect storage proteins. Insect Biochem. Mol. Biol. 2007, 37, 1064–1074.

(46) Huson, D. H.; Scornavacca, C. Dendroscope 3: An interactive viewer for rooted phylogenetic trees and networks. Syst. Biol. 2012, 61, 1061–1067.

183

Figures

0 .1 8

Is o n y c h ia s p . (E )

0 .1 5

A c r o n e u r ia a b n o r m is ( P ) )

1 A llo p e r la s p . ( P )

- 0 .1 2

( Iso g e n o id e s s p . ( P )

e 0 .0 9 H y d r o p sy c h e a lh e d r a ( T )

n 0 .0 6 H y d ro p s y c h e sp a rn a ( T ) Z 0 .0 3

0 .0 0 0 .0 0 0 .0 5 0 .1 0 0 .1 5 0 .2 0 -1 C d k ( d ) e

Figure 1) Predicted and measured values of Zn kes from 6 EPT species. Closed black circles represent Zn and Cd kes of 11 species within aquatic insect families Ephemerellidae and Hydropsychidae used to construct the regression line and associated 95% confidence interval. Five of the 6 species tested fell within the 95% confidence interval. Symbols represent mean ± std. error of measured Zn and Cd kes. E=Ephemeroptera, P=Plecoptera, T=Trichoptera.

184

Z n k e 's

C d k e 's

0 .0 0 0 .0 5 0 .1 0 0 .1 5 0 .2 0

-1 E fflu x R a te C o n s ta n ts (d )

Figure 2) Zn and Cd kes for 13 aquatic insect species representing the families Ephemerellidae and Hydropsychidae. Pagels arbitrary branch lengths are depicted. The phylogenetic tree was constructed using Dendroscope.46

185

A Mollusca

Chordata

Annelida

Arthropoda

0.0 0.1 0.2 0.3 -1 Zn ke (d )

Mollusca B

Chordata

Annelida

Arthropoda

0.0 0.1 0.2 0.3 -1 Cd ke (d )

Figure 3. All compiled Zn (Panel A) and Cd (Panel B) ke values from the literature. Values include both dietary and dissolved kes separately. In Arthropoda, closed circles = aquatic insects, open circles = other species.

186

Supporting Information

Phylogenetic Methods

Phylogenetic analyses were conducted on all species with available Zn (45 species) and Cd (75 species) kes, species with both kes available (43 species), as well Zn and Cd kes for aquatic insects within families Ephemerellidae and Hydropsychidae collected specifically for this study (13 species). The phylogenetic topologies were determined using the National

Center for Biotechnology Information (NCBI) taxonomy database of nuclueotide sequences curated in GenBank.1 The topology of Ephemerellidae species was further determined using the literature.2–5 Topologies were entered in Newick format into Mesquite6 in order to create trees, and branch lengths were arbitrarily determined (see below). All analyses were completed with log transformed kes.

The MatLab program Physig_LL.m was used to compute the K statistic.7 Three arbitrary branch lengths (all branches set to equal 1, Pagel’s arbitrary,8 and Grafen’s arbitrary9) were employed with the Zn-only and Cd-only topologies in order to determine branch lengths which resulted in the highest ln likelihood. The K statistic from the branch lengths that resulted in the highest likelihood was used.

10 6 The PDTREE module in Mesquite was used to detect covariation of Zn and Cd kes using phylogenetically independent contrasts.11,12 For the tree with all (43) species, branch lengths of 1 were utilized, and for the insect only (13 species) tree, Pagel’s arbitrary branch lengths were used. No significant correlation was detected with the absolute values of the standardized phylogenetically independent contrasts and their standard deviations.13,14 This

187

assured that the branch lengths adequately fit the tip data (assuming a Brownian-motion like model of trait evolution).

188

Table S1) Cd and Zn efflux rate constants (kes) for 47 species of aquatic insects. Cd ke Zn ke Order Genus species (d-1) (d-1) Ref. E Centroptilum triangulifer a 0.004F 15 E Drunella cornutella 0.035D 0.05D n/a E Drunella flavilinea 0.10F 0.098* 16 E Drunella grandis 0.12D 0.11* 17 E Drunella longicornis 0.010D 0.016D n/a E Drunella tuberculata 0.022D 0.057D n/a E Ephemerella catawba 0.16D 0.15D n/a E Ephemerella excrucians 0.014D 0.026* 17 E Ephemerella invaria 0.023D 0.047D n/a E Ephemerella rossi 0.019D 0.031D n/a D 17 E Ephemerella subvaria 0.095 0.094*

E Epeorus albertae 0.12F 0.11* 16

E Eurylophella verisimilis 0.017 0.013D n/a

E Hexagenia limbata 0.086F 0.086* 18

E Isonychia sp. 0.019D 0.030* 17 D D 17 E Isonychia tuscalanensis 0.013 0.023 E Maccaffertium ithaca 0.0078D 0.021* 17 F 16 E Nixe sp. 0.07 0.073* E Rhithrogena morrisoni 0.001D a 17

E Serratella tibialis 0.20F 0.18* 16

P Acroneuria abnormis 0.076F 0.078* 19

P Acroneuria abnormis 0.087D 0.084D 17

P Alloperla sp. 0.026D 0.027D n/a

P Baumanella alameda 0.051D 0.057* 17

P Baumanella alameda 0.065F 0.069* 19

P Calineuria californica 0.049F 0.055* 19

E = Ephemeroptera; P = Plecoptera; T = Trichoptera; D = Diptera; M = Megaloptera; a = value unable to be predicted from linear regression based on the range of Cd; * = predicted value, - = value not predicted because not EPT species, F fed exposure, D dissolved exposure

189

Table S1 Continued

-1 -1 Cd ke (d ) Zn ke (d ) Order Genus species Ref. P Calineuria californica 0.081D 0.082* 17 P Claassenia sabulosa 0.091F 0.090* 19 P Claassenia sabulosa 0.082D 0.083* 17 P Doroneuria baumani 0.084D 0.084* 17 P Hesperoperla pacifica 0.078F 0.080* 19 P Hesperoperla pacifica 0.082D 0.083* 17 P Isogenoides hansoni 0.11D 0.11D 17 P Paragnetina sp. 0.118F 0.11* 19 P Paragnetina sp. 0.085D 0.085* 17 P Pteronarcys dorsata 0.086D 0.086* 17 P Skwala sp. 0.079D 0.081* 17 T Arctopsyche irrorata 0.075D 0.079D n/a T Chemautopsyche sp. 0.027D 0.037* 17 D D T Diplectrona modesta 0.075 0.071 n/a

T Hydropsyche alhedra 0.17D 0.11D n/a

T Hydropsyche betteni 0.21D a 20

T Hydropsyche californica 0.20D a 17 D D T Hydropsyche slossonae 0.14 0.13 n/a T Hydropsyche sparna 0.085D 0.083D n/a D D T Parapsyche cardis 0.042 0.028 n/a T Rhyacophila fuscula 0.12D 0.11* 17 D 17 T Rhyacophila sp. 0.14 0.13* F 21 D Chaoborus albatus 0.0253 - F 21 D Chaoborus americanus 0.0368 - F 21 D Chaoborus flavicans 0.0038 - D Chaoborus punctipennis 0.0029F - 21 M Sialis velata 0.029F - 22

190

Table S2) Cd and Zn efflux rate constants (kes) for 30 aquatic species. -1 -1 Cd ke (d ) Zn ke (d ) Phylu Genus species Ref. m Ar Balanus amphitrite 0.0071F 0.0039F 23,24 Ar Chaetogammarus marinus 0.29D - 25 Ar Daphnia magna 0.047D 0.19D 26 Ar Daphnia magna 0.031F 0.22F 26–28 Ar Elminius modestus 0.0181F 0.0022F 29 Ar Penaeus indicus - 0.05D 30 Ar Temora longicornis 0.297F 0.079F 31 Ar Temora longicornis 0.108D 0.108D 31 C Acanthopagrus schlegeli 0.089F 0.016F 32 C Ambassis jacksoniensis 0.013D 0.017D 33 D D 34 C Lutjanus argentimaculatus 0.025 0.015

C Lutjanus argentimaculatus 0.047F 0.015F 34

C Menidia menidia 0.064D 0.059D 35

C Periophthalmus cantonensis 0.022D 0.019D 36

C Periophthalmus cantonensis 0.044F 0.014F 36 D D 37 C Psetta maxima 0.020 0.007 C Scyliorhinus canicula 0.008D 0.010D 37 D D 38 C Sparus auratus 0.065 0.037 Ar = Arthropoda, C = Chordata, M = Mollusca, An = Annelida, F fed exposure, D dissolved exposure, - = value not available

191

Table S2 continued -1 Cd ke (d ) Zn ke Phylum Genus species (d-1) Ref. M Babylonia formosae habei 0.0055F 0.0057F 39 M Chlamys nobilis 0.0100F 0.032F 40 M Crassostrea rivularis 0.014F 0.014F 41 M Dreissena polymorpha 0.012F - 42 M Dreissena polymorpha 0.011D - 42 M Macoma balthica 0.018F 0.012F 43 M Mactra veneriformis 0.025F 0.048F 44 M Mytilus edulis 0.028F 0.014F 45 M Mytilus edulis 0.023D 0.044D 46 M Mytilus galloprovincialis 0.014D 0.0088D 47 M Nassarius teretiusculus 0.0011F 0.0138F 39 M Perna viridis 0.012D 0.041D 48 M Perna viridis 0.0200F 0.029F 49 M Potamocorbula amurensis 0.011F 0.027F 43 M Ruditapes philippinarum 0.026F 0.025F 44,49 M Saccostrea glomerata 0.004F 0.003F 41 An Arenicola marina 0.003D 0.037D 50 An Lumbriculus variegatus 0.002D - 51 An Nereis diversicolor 0.027D 0.035D 52 Ar = Arthropoda, C = Chordata, M = Mollusca, An = Annelida, F fed exposure, D dissolved exposure, - = value not available

192

Table S3. Analyses of phylogenetic signal for Cd and Zn kes, including comparisons of the ln likelihoods for each set of arbitrary branch lengths compared with that of a star phylogeny (indicating no hierarchal structure to the data). Species Metal Branch K- p-value ln tested transformation statistic likelihood All available Zn Branch lengths = 1 0.41 0.023 -30.5 species Pagel’s arbitrary 0.47 0.014 -30.12* Grafen’s arbitrary 0.12 0.037 -38.61 Star -28.41 All available Cd Branch lengths = 1 0.59 < 0.0001 -56.58 species Pagel’s arbitrary 0.39 < 0.0001 -61.7 Grafen’s arbitrary 0.11 < 0.0001 -66.42 Star -62.16 Aquatic Cd Branch lengths = 1 0.78 0.025 -5.35 insects (This Pagel’s arbitrary 0.86 0.013 -4.9 study, Grafen’s arbitrary 0.47 0.048 -6.95 13 species) Star -6.25 Aquatic Zn Branch lengths = 1 0.62 0.22 -4.75 insects (This Pagel’s arbitrary 0.67 0.078 -4.07* study, Grafen’s arbitrary 0.35 0.21 -6.34 13 species) Star -3.57 *This set of arbitrary branch lengths gave a higher likelihood than other branch length, but it did not surpass the star phylogeny (no hierarchal structure). Because the ln likelihood comparison is not equivalent to testing the null hypothesis of no phylogenetic signal and because the phylogenetic signal is moderate,53 it is possible for phylogenetic signal to exist even though the ln likelihood of the tree is less than that of the star phylogeny.

193

References

(1) Federhen, S. The NCBI Taxonomy database. Nucl. Acid. Res. 2012, 40, D136–D143.

(2) Eubank, L. S.; Telarroja, S.; Ogden, T. H. Spiny crawlers on trees: Phylogeny of Ephemerellidae (Ephemeroptera). Unpubl. poster 2011.

(3) Jacobus, L. M.; McCafferty, W. P. Revision of Ephemerellidae Genera (Ephemeroptera). T. Am. Entomol. Soc. 2008, 134, 185–274.

(4) Ogden, T. H.; Osborne, J. T.; Jacobus, L. M.; Whiting, M. F. Combined molecular and morphological phylogeny of Ephemerellinae (Ephemerellidae: Ephemeroptera), with remarks about classification. Zootaxa. 2009, 1991, 28–42.

(5) Ogden, T. H.; Gattolliat, J. L.; Sartori, M.; Staniczek, A. H.; Soldan, T.; Whiting, M. F. Toward a new paradigm in mayfly phylogeny (Ephemeroptera): Combined analysis of morphological and molecular data. Syst. Entomol. 2009, 34, 616–634.

(6) Maddison, W. P.; Maddison, D. R. Mesquite: A modular system for evolutionary analysis. Version 2.0. 2007.

(7) Blomberg, S. P.; Garland, Jr., T.; Ives, A. R. Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution. 2003, 57, 717–745.

(8) Pagel, M. D. A method for the analysis of comparative data. J. Theor. Biol. 1992, 156, 431–442.

(9) Grafen, A. The phylogenetic regression. Phil. Trans. R. Soc. Lond. B. 1989, 326, 119– 156.

(10) Midford, P. E.; Garland Jr., T.; Maddison, W. P. . PDAP Package of Mesquite. Version 1.07. 2005.

(11) Garland, Jr., T.; Midford, P. E.; Ives, A. R. An introduction to phylogenetically based statistical methods, with a new method for confidence intervals on ancestral states. Am. Zool. 1999, 39, 374–388.

(12) Garland Jr., T.; Ives, A. R. Using the past to predict the present: Confidence intervals for regression equations in phylogenetic comparative methods. Am. Nat. 2000, 155, 346–364.

(13) Garland Jr., T.; Huey, R. B.; Bennett, A. F. Phylogeny and thermal physiology in lizards: A reanalysis. Evolution. 1991, 45, 1969–1975.

194

(14) Garland Jr., T.; Harvey, P. H.; Ives, A. R. Procedures for the analysis of comparative data using phylogenetically independent contrasts. Syst. Biol. 1992, 41, 18–32.

(15) Kim, K. S.; Funk, D. H.; Buchwalter, D. B. Dietary (periphyton) and aqueous Zn bioaccumulation dynamics in the mayfly Centroptilum triangulifer. Ecotoxicology. 2012, 21, 2288–2296.

(16) Cain, D.; Croteau, M.-N.; Luoma, S. N. Bioaccumulation dynamics and exposure routes of Cd and Cu among species of aquatic mayflies. Env. Toxicol. Chem. 2011, 30, 2532–2541.

(17) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.; Garland, Jr., T. Aquatic insect ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. P. Natl. Acad. Sci. USA. 2008, 105, 8321–8326.

(18) Michaud, A. L.; Hare, L.; Campbell, P. G. C. Exchange rates of cadmium between a burrowing mayfly and its surroundings in nature. Limnol. Ocean. 2005, 50, 1707– 1717.

(19) Martin, C. A.; Cain, D. J.; Luoma, S. N.; Buchwalter, D. B. Cadmium ecophysiology in seven stonefly (Plecoptera) species: Delineating sources and estimating susceptibility. Env. Sci. Technol. 2007, 41, 7171–7177.

(20) Evans, R. D.; Balch, G. C.; Evans, H. E.; Welbourn, P. M. Simultaneous measurement of uptake and elimination of cadmium by caddisfly (Trichoptera: Hydropsychidae) larvae using stable isotope tracers. Env. Toxicol. Chem. 2002, 21, 1032–1039.

(21) Croteau, M.-N.; Hare, L.; Tessier, A. Differences in Cd accumulation among species of the lake-dwelling biomonitor Chaoborus. Can. J. Fish. Aquat. Sci. 2001, 58, 1737– 1746.

(22) Croisetière, L.; Hare, L.; Tessier, A. A field experiment to determine the relative importance of prey and water as sources of As, Cd, Co, Cu, Pb, and Zn for the aquatic invertebrate Sialis velata. Env. Sci. Technol. 2006, 40, 873–879.

(23) Rainbow, P. S.; Blackmore, G.; Wang, W. X. Effects of previous field-exposure history on the uptake of trace metals from water and food by the barnacle Balanus amphitrite. Mar. Ecol.- Prog. Ser. 2003, 259, 201–213.

(24) Wang, W.-X.; Qui, J.-W.; Qian, P.-Y. The trophic transfer of Cd, Cr. and Se in the barnacle Balanus amphitrite from planktonic food. Mar. Ecol.- Prog. Ser. 1999, 187, 191–201.

195

(25) Clason, B.; Langston, W. J.; Zauke, G.-P. Bioaccumulation of trace metals in the amphipod Chaetogammarus marinus (Leach, 1815) from the Avon and Tamar estuaries (UK): Comparison of two-compartment and hyperbolic toxicokinetic models. Mar. Env. Res. 2004, 57, 171–195.

(26) Yu, R.-Q.; Wang, W.-X. Trace metal assimilation and release budget in Daphnia magna. Limnol. Ocean. 2002, 47, 495–504.

(27) Yu, R.-Q.; Wang, W.-X. Biokinetics of cadmium, selenium, and zinc in freshwater alga Scenedesmus obliquus under different phosphorus and nitrogen conditions and metal transfer to Daphnia magna. Env. Pollut. 2003, 129, 443–456.

(28) Guan, R.; Wang, W. Dietary assimilation and elimination of Cd, Se, and Zn by Daphnia magna at different metal concentrations. Env. Toxicol. Chem. 2004, 23, 2689–2698.

(29) Rainbow, P.; Wang, W. Comparative assimilation of Cd, Cr, Se, and Zn by the barnacle Elminius modestus from phytoplankton and zooplankton diets. Mar. Ecol.- Prog. Ser. 2001, 218, 239–248.

(30) Nunez-Nogueira, G.; Rainbow, P. S. Kinetics of zinc uptake from solution, accumulation and excretion by the decapod crustacean Penaeus indicus. Mar. Biol. 2005, 147, 93–103.

(31) Wang, W.-X.; Fisher, N. S. Accumulation of trace elements in a marine copepod. Limnol. Ocean. 1998, 43, 273–283.

(32) Zhang, L.; Wang, W.-X. Size-dependence of the potential for metal biomagnification in early life stages of marine fish. Env. Toxicol. Chem. 2007, 26, 787–794.

(33) Creighton, N.; Twining, J. Bioaccumulation from food and water of cadmium, selenium and zinc in an estuarine fish, Ambassis jacksoniensis. Mar. Pollut. Bull. 2010, 60, 1815–1821.

(34) Xu, Y.; Wang, W.-X. Exposure and potential food chain transfer factor of Cd, Se and Zn in marine fish Lutjanus argentimaculatus. Mar. Ecol.- Prog. Ser. 2002, 238, 173– 186.

(35) Dutton, J.; Fisher, N. S. Intraspecific comparisons of metal bioaccumulation in the juvenile Atlantic silverside Menidia menidia. Aquat. Biol. 2010, 10, 211–226.

196

(36) Ni, I.-H.; Chan, S. M.; Wang, W.-X. Influences of salinity on the biokinetics of Cd, Se, and Zn in the intertidal mudskipper Periophthalmus cantonensis. Chemosphere. 2005, 61, 1607–1617.

(37) Mathews, T.; Fisher, N. S. Dominance of dietary intake of metals in marine elasmobranch and teleost fish. Sci. Total Env. 2009, 407, 5156–5161.

(38) Mathews, T.; Fisher, N. S.; Jeffree, R. A.; Teyssié, J.-L. Assimilation and retention of metals in teleost and elasmobranch fishes following dietary exposure. Mar. Ecol.- Prog. Ser. 2008, 360, 1–12.

(39) Wang, W.-X.; Ke, C. Dominance of dietary intake of cadmium and zinc by two marine predatory gastropods. Aquat. Toxicol. 2002, 56, 153–165.

(40) Pan, K.; Wang, W.-X. The subcellular fate of cadmium and zinc in the scallop Chlamys nobilis during waterborne and dietary metal exposure. Aquat. Toxicol. 2008, 90, 253–260.

(41) Ke, C.; Wang, W.-X. Bioaccumulation of Cd, Se, and Zn in an estuarine oyster (Crassostrea rivularis) and a coastal oyster (Saccostrea glomerata). Aquat. Toxicol. 2001, 56, 33–51.

(42) Roditi, H. A.; Fisher, N. S. Rates and routes of trace element uptake in zebra mussels. Limnol. Ocean. 1999, 44, 1730–1749.

(43) Lee, B.-G.; Wallace, W. G.; Luoma, S. N. Uptake and loss kinetics of Cd, Cr and Zn in the bivalves Potamocorbula amurensis and Macoma balthica: Effects of size and salinity. Mar. Ecol.- Prog. Ser. 1998, 175, 177–189.

(44) Shi, D.; Wang, W.-X. Understanding the differences in Cd and Zn bioaccumulation and subcellular storage among different populations of marine clams. Env. Sci. Technol. 2004, 38, 449–456.

(45) Wang, W.-X.; Fisher, N. S. Modeling the influence of body size on trace element accumulation in the mussel Mytilus edulis. Mar. Ecol.- Prog. Ser. 1997, 161, 103–115.

(46) Baines, S. B.; Fisher, N. S. Modeling the effect of temperature on bioaccumulation of metals by a marine bioindicator organism, Mytilus edulis. Env. Sci. Tecnnol. 2008, 42, 3277–3282.

(47) Fisher, N. S.; Teyssié, J.-L.; Fowler, S. W.; Wang, W.-X. Accumulation and retention of metals in mussels from food and water: A comparison under field and laboratory conditions. Env. Sci. Technol. 1996, 30, 3232–3242.

197

(48) Blackmore, G.; Wang, W.-X. Uptake and efflux of Cd and Zn by the green mussel Perna viridis after metal preexposure. Env. Sci. Technol. 2002, 36, 989–995.

(49) Chong, K.; Wang, W.-X. Comparative studies on the biokinetics of Cd, Cr, and Zn in the green mussel Perna viridis and the Manila clam Ruditapes philippinarum. Env. Pollut. 2001, 115, 107–121.

(50) Casado-Martinez, M. C.; Smith, B. D.; Delvalls, T. A.; Rainbow, P. S. Pathways of trace metal uptake in the lugworm Arenicola marina. Aquat. Toxicol. 2009, 92, 9–17.

(51) Xie, L.; Lambert, D.; Martin, C. A.; Cain, D. J.; Luoma, S. N.; Buchwalter, D. B. Cadmium biodynamics in the oligochaete Lumbriculus variegatus and its implications for trophic transfer. Aquat. Toxicol. 2008, 86, 265–271.

(52) Rainbow, P. S.; Smith, B. D.; Luoma, S. N. Biodynamic modelling and the prediction of Ag, Cd and Zn accumulation from solution and sediment by the polychaete Nereis diversicolor. Mar. Ecol.- Prog. Ser. 2009, 390, 145–155.

(53) Fusco, G.; Garland, Jr., T.; Hunt, G.; Hughes, N. C. Developmental trait evolution in trilobites. Evolution. 2011, 66, 314–329.

198

SUMMARY AND CONCLUSIONS

Aquatic insects dominate freshwater systems in terms of numbers and diversity.

While we utilize aquatic insects in order to understand how contaminants (e.g., metals) impact benthic communities, we know little about their ion/metal transport physiology. We know even less about the physiological variance present across aquatic insect clades (e.g., families, genera). There were two main objectives of the studies encompassed in this dissertation. First, we wanted to better understand how aquatic insects traffic dissolved metals. We also wanted to determine how variable closely related aquatic insect species are in their dissolved metal bioaccumulation dynamics, and whether we could find patterns in physiological values across species which might lead to predictions. Overall, we discovered that aquatic insects have unique metal transport physiologies from other aquatic organisms.

Comparative studies revealed that closely related aquatic insects are exceptionally variable in their ion and metal physiologies, and predictions are possible using patterns present in the data across species.

Throughout our studies, there were multiple pieces of evidence for shared apical transport systems for Cd and Zn in aquatic insects. This was not unexpected due to their similar biochemical properties and propensity for using the same transporters in other aquatic organisms. Zn and Cd had similar patterns of uptake kinetics across concentrations spanning four orders of magnitude in the hydropsychid Hydropsyche sparna (Chapter 1), and there was evidence that multiple transport systems are used at varying concentration ranges.

Further experiments in H. sparna revealed that Cd outcompeted Zn at the higher two concentrations examined of three total concentrations. Lastly, the strong co-variation of Zn

199

and Cd bioaccumulation traits across species also suggested that Zn and Cd share the same transport system(s) (Chapter 3). Taken in sum, it appears that Cd and Zn share a transport system (or systems), and both are taken up without inhibition until concentrations of ~0.6

µM. At higher concentrations, Cd has a higher affinity for uptake across a shared transport system(s).

Contrary to what we expected, our results show that Cd and Zn could be potentially using different transporters than Ca. Increasing Ca concentrations by orders of magnitude only modestly inhibited Cd and Zn uptake in H. sparna. Neither Cd nor Zn inhibited Ca uptake in four species of aquatic insects up to concentrations of 89 nM and 1.53µM, respectively. There was even evidence of modest stimulation of Ca uptake in low levels of metal exposure in one ephemerellid species. Lastly, while Cd and Zn uptake rate constants strongly co-varied across species (again suggesting shared transport), neither Cd nor Zn uptake significantly co-varied with Ca uptake.

As a result of our experiments and analyses in Chapters 1-3, we suggest that aquatic insects have physiologies which differ from other freshwater organisms, potentially due to the unique evolutionary history of aquatic insects. We hypothesize that this difference could be one reason for the large disconnect between aquatic insects responses to metals in the field and in the laboratory (during acute dissolved-only exposures). Other reasons outlined in

Chapter 4 for this disconnect include the long time period required to reach steady state metal tissue concentrations of metals, the large percentage of metals acquired from dietary sources in the field, and the potential for dietary metal to be more physiologically active than dissolved metal. From these studies, we suggest that the current methodologies for assessing

200

metal toxicity are inadequate for assessing the sensitivity of aquatic insects to metals and that more relevant methodologies (such as including dietary exposures) could better link the laboratory performance of aquatic insects with their performance in the field.

Another significant mechanistic discovery of this work was that adsorption of metals can play an important role in their “uptake” values. Adsorption accounted for a substantial amount of the total accumulation of both Ca and Mn (Chapters 1 and 2). In these studies, we were able to remove adsorbed Ca and Mn using EDTA, and adsorbed Mn oxides using a combination of an EDTA rinse and an ascorbate rinse. The presence of more Mn oxides on the body surface decreased the adsorption of Cd and Zn onto the body surface in a concentration dependent manner, but did nothing to the internalization of Cd and Zn. We conclude that failing to account for adsorption in ion uptake studies, particularly studies which only utilize one time point, can lead to artificially-inflated uptake values for ions.

Across all metal bioaccumulation parameters examined in this study, aquatic insect species from families Hydropsychidae and Ephemerellidae were especially variable.

However, we were able to identify factors which influenced dissolved metal bioaccumulation. Ca uptake varied 70-fold across 12 aquatic insect species examined in this study (Chapter 2). Clade membership (family) exerted significant influence on Ca uptake rates, and Ca uptake was at least partially described by differences in body weight – smaller animals had higher Ca uptake rates while larger animals had slower uptake rates.

Zn and Cd ku, ke, and BCF values were highly variable across species tested, spanning orders of magnitude (Chapter 3). Body weight was a significant driver of Zn and

Cd ku values, and consequently, BCF values. Familial differences spanned orders of

201

magnitude in both ku and BCF values, and phylogenetic signal broke down at these finer levels of taxonomic resolution due to the high variability observed across closely related species. Neither body size nor phylogenetic history significantly influenced ke values across species within these two families. While it is commonplace in toxicology as well as and other disciplines to generalize across taxonomic groupings (e.g., genera, families, orders), our work shows that species, even within the same genera, can have extremely different physiologies. Thus, when generalizing across taxonomic groups, it’s important to understand the variance which can exist across as well as within groups.

We expanded our analyses of dissolved Cd bioaccumulation parameters to include other aquatic insect taxa (Chapter 5). We found that in 42 species spanning three orders-

Ephemeroptera, Plecoptera, and Trichoptera- patterns emerged in dissolved bioaccumulation parameters across species. Some families, particularly mayfly families, had significantly higher bioaccumulation parameters than others, and some families were much more variable than others. Phylogenetic signal was present in ku and BCF values, but not ke values. Further, models including genera (and body weight) as a predictor of bioaccumulation parameters explained more variation in parameters than models including family, order, body weight alone, or the ecological trait of feeding guild membership. In this work, we demonstrated the importance of considering phylogeny in traits analyses.

We further expanded our scope in Chapter 6 to include all aquatic species for which

Zn and/or Cd ke values were available. Patterns in the ranges of ke values across phyla were apparent- arthropods had particularly variable efflux values whereas mollusks were much less variable across species. Evolutionary history was a significant driver of ke values across

202

all taxa. In aquatic insects examined, we were able to use the strong co-variation of Zn and

Cd ke values to predict Zn ke values for untested species from known Cd ke values. This was the first global analysis of the ability to eliminate metals across different taxonomic groups.

This work again demonstrated the importance of considering evolutionary history in the analysis of traits across species.

In summary, we find that aquatic insects likely have different physiologies than other common test organisms (e.g., Daphnia magna, fish) regarding the trafficking of dissolved metal which leads to their metal tolerance in the laboratory (as opposed to their observed metal sensitivity in the field). Hypocalcemia does not appear to be the mechanism of toxicity in dissolved Cd/Zn exposures as evidenced by the lack on interactions between Ca and

Cd/Zn. We also discovered that closely related aquatic insect species, even species within the same genus, can have very different dissolved metal bioaccumulation parameters. While phylogenetic signal breaks down at this fine level of taxonomic identification, phylogeny still plays an important role in the interspecies variability of dissolved metal bioaccumulation parameters across different clades (e.g., families). Further, other patterns exist throughout metal bioaccumulation parameters (e.g., co-variation of traits) which made predictions of metal bioaccumulation parameters possible.

203