Aquatic-To-Terrestrial Contaminant Flux in the Scioto River Basin, Ohio, USA

Aquatic-to-terrestrial contaminant flux in the Scioto River basin, Ohio, USA

THESIS

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Jeremy M. Alberts

Graduate Program in Environment and Natural Resources

The Ohio State University

2012

Committee:

Dr. Mažeika S.P. Sullivan, Advisor

Dr. Amanda D. Rodewald

Dr. Stanley D. Gehrt

Copyrighted by

Jeremy M. Alberts

2012

Abstract

Aquatic emergent insects provide important prey subsidies to riparian consumers. These aquatic-to-terrestrial feeding relationships provide a pathway through which aquatic contaminants are “reterrestrialized” into riparian food webs. However, influences of land use and land cover (LULC) on the magnitude of aquatic-to-terrestrial contaminant transfers remain largely unexplored.

To that end, I investigated aquatic-to-terrestrial contaminant fluxes at 11 study reaches in the Olentangy and Scioto Rivers (OH, USA), representing urban, agricultural, and mixed land uses. At nine study reaches, I collected benthic sediment, aquatic emergent insects, ants (Formica subsericea), spiders of the family Tetragnathidae, riparian vegetation, and periphyton. At eight of these reaches, as well as additional four reaches where I erected nest-boxes, I sampled riparian swallows including: bank (Riparia riparia), northern rough-winged (Stelgidopteryx serripennis), tree (Tachycineta bicolor), and cliff (Petrochelidon pyrrhonota) swallows. Subsequently, all biological samples were analyzed for δ13C and δ15N. Sediment, ants, spiders, and swallows were tested for a suite of toxic elements including arsenic (As), selenium (Se), lead (Pb), and mercury

(Hg).

Two-source (δ13C and δ15N) mixing models indicated that Tetragnathidae were highly reliant on aquatic insects ( ̅= 76.9%, SD = 8.9%), whereas ant dependence was

ii less but with greater variability ( ̅ = 27.8%, SD = 25.1%). Characteristics of shoreline habitat including standing dead trees and % overhanging vegetation explained 70 and

42% of the variation in the contribution of aquatic prey to F. subsericea and

Tetragnathidae, respectively.

Spider density was positively related to land-cover characteristics associated with urbanization (% impervious surfaces, % invasive shrubs, population density) and nearshore habitat. Shoreline habitat also was strongly related to the overall flux (i.e., contaminant load assimilated into consumer tissue) of Se (R2 = 0.58) and As (R2 = 0.51) to the tetragnathid spider assemblage, and Pb flux to spiders was higher in urban and agricultural reaches than in mixed reaches (F = 6.10, P = 0.025). F. subsericea density exhibited a positive relationship with urbanization (R2 = 0.83). As and Se flux to F. subsericea assemblages was positively related to urbanization (R2 = 0.70) as well as shoreline habitat, and Pb flux was higher in urban reaches than other land use types (F =

8.68, P = 0.017).

For swallows, Hg concentrations were significantly higher at rural reaches than at urban reaches (t = -2.96, P = 0.003, df = 24), and Hg concentrations in swallows were positively related to Hg concentrations in sediment (R2 = 0.23, P = 0.030), though no relationships were evident for Se in swallows. We found that swallow Hg concentrations were significantly higher in rural than urban reaches (t = -2.96, P = 0.003, df = 24), and marginally so for Se (t = -1.54, P = 0.068, df = 24). To an extent, these relationships iii appear to be mediated by swallow reliance on aquatic emergent insect prey. For example, swallows that exhibited a higher proportion of aquatic prey in their diet and fed at a higher trophic level also exhibited elevated Se levels. I also found that both Se and

Hg concentrations in adult swallows were significantly higher than those observed in juveniles (Se: t = -3.47, P = 0.013, df = 4; Hg: t = -4.35, P = 0.006, df = 4).

Collectively, my results indicate that LULC mediates aquatic contaminant flux to terrestrial consumers via regulation of aquatic resource utilization. For riparian arthropods, differences in density associated with landscape variability can result in a significant discrepancy between the magnitude of contaminant export by aquatic emergent insects and the realized contaminant flux to riparian food webs. At a broader spatial scale, riparian swallows may represent a useful assessment tool for contaminant exposure in linked aquatic-terrestrial systems.

Dedication

I dedicate this thesis to Jennifer. Without her endless patience and support, this never would have been possible. For everything you’ve done for me, and for all of the time spent away with which you’ve had to deal, I thank you from the bottom of my heart. I couldn’t have done it without you. I would also like to thank my parents, who have been more encouraging and supportive than any person has the right to expect. Thank you both very much, for everything.

Acknowledgments

I would like to thank the funding sources for this project, The Ohio State University

School of Environment and Natural Resources, The Ohio Agricultural Research and

Development Center SEEDS program, and the United States Forest Service. I would also like to thank my committee, Dr. Stan Gehrt and Dr. Amanda Rodewald, for their valued assistance and direction throughout the course of this project. I also extend my gratitude to the array of undergraduates, lab mates, faculty, and staff that helped me in various ways. There is not enough space here to thank every individual, but know that your assistance was greatly appreciated. Finally, I would like to give special thanks to my advisor, Dr. Mažeika Sullivan, for his patience and guidance with all aspects of this research project. This would have been a nearly impossible task without his assistance and support.

Vita

June 1997 ...... Wheeling Park High School

2001...... B.S. Biology, Muskingum College

2010 to present ...... Graduate Teaching Associate, School of

Environment and Natural Resources,

The Ohio State University

Field of Study

Major Field: Environment and Natural Resources

vii

Table of Contents

Abstract ...... ii

Acknowledgments...... vii

Vita ...... vii

Table of Contents ...... viii

List of Tables ...... ixx

List of Figures………………………………………………………………………….…xi

Chapter 1: Background and Literature Review ...... 1

Chapter 2: Land use mediates aquatic-to-terrestrial contaminant flux ...... 31

Chapter 3: Riparian swallows as integrators of landscape change in a multiuse river system: implications for aquatic-to-terrestrial transfers of contaminants ...... 76

References ...... 113

Appendix A: Study Reach Description ...... 134

Appendix B: Arthropod Contaminant Concentrations ...... 135

Appendix C: Arthropod Density ...... 137

Appendix D: Swallow Contaminant Concentrations ...... 138

viii

List of Tables

Table 1.1. Common aquatic contaminants, their impacts on humans and wildlife, and major pathways through which they enter aquatic systems...... 3

Table 1.2. Estimated mass of contaminants exported from aquatic to terrestrial systems by aquatic emergent insects...... 12

Table 2.1. Biomass and contaminant concentrations of aquatic emergent insects of the families Chironomidae and Hydropsychidae and contaminant export estimates from nine study reaches in the Scioto River basin, Ohio, USA. Note: Hydropsychidae was only collected from one urban reach...... 62

Table 2.2. Contribution of aquatic prey to riparian arthropod consumer diet for study reaches in the Scioto River basin, Ohio (calculated using IsoError v 1.04; Phillips and

Greg 2001)...... 63

Table 2.3. Adjacent (500 m on each side of river) land-use and land-cover (LULC) principal component analysis: eigenvalues and the percent variance captured by the principal components (eigenvalues > 1), along with each principal component’s loadings and the proportion of the variance (r2) each variable shared with the PCA axes. Bold type indicates those axes used as successful predictor variables in regression models...... 64

Table 2.4. Estimated flux (aquatic contaminants uptaken by riparian consumers) of contaminants to Tetragnathidae and F. subsericea from aquatic prey sources for (a) all study reaches and (b) urban, agricultural, and mixed reaches ...... 65

Table 2.5. Explanatory variables and their coefficients in the significant regression models of contaminant flux to F. subsericea ...... 66

Table 3.1. Concentrations of Se and Hg in blood of riparian swallows and sediment from study reaches in the Scioto River basin, Ohio, USA. Sites are presented upstream to downstream. Note: CLSW = cliff swallow, BANS = bank swallow, TRSW = tree swallow...... 104

Table 3.2. δ13C and δ15N values for riparian swallows, periphyton, terrestrial vegetation, emergent aquatic insects, and terrestrial invertebrates for all study reaches in the Scioto

River basin, Ohio, USA. Sites are presented upstream to downstream. Note: CLSW = cliff swallow, BANS = bank swallow, TRSW = tree swallow, NRWS = northern rough- winged swallow...... 105

Table 3.3. Adjacent (500 m on each side of river) land-use and land-cover (LULC) principal component analysis: eigenvalues and the percent variance captured by the principal components (eigenvalues > 1), along with each principal component’s loadings and the proportion of the variance ( ) each variable shared with the PCA axes. Bold type indicates the axis used as successful predictor variable in regression models...... 106

List of Figures

Figure 1.1. Representation of reciprocal food-web linkages (e.g., energy flows as represented by arrows) in a stream-riparian ecosystem (from Sullivan and Rodewald

2012) ...... 5

Figure 1.2. Tetragnatha sp. at Scioto River, Chillicothe, OH, USA. Tetragnathid spiders are nocturnal riparian predators that seek refuge during the day in trees and shrubs. They emerge at sunset to build webs on riparian vegetation or other structures...... 7

Figure 1.3. Fledgling northern rough-winged swallows (Stelgidopteryx serripennis) at the Scioto River, Columbus, OH, USA...... 10

Figure 1.4. Likelihood of dieldrin in fish exceeding levels of concern for fish-eating wildlife (from USGS 2012)...... 18

Figure 2.1. Nine 1-km study reaches of the Scioto River basin, Ohio, USA...... 67

Figure 2.2 Relationship between downstream distance from an area of extensive point source inputs into the Scioto River and (a) Se concentrations of Chironomidae (solid line) and Hydropsychidae (dashed line) and (b) Hg concentrations of Chironomidae (solid line) and Hydropsychidae (dashed line) across all study reaches. Slopes for the linear regression models were highly significant for Hydropsychidae Se (y = 4829 – 24.3x, R2 =

0.86, P < 0.001) and Hg (y = 87.5 - 0.23x, R2 = 0.66, P = 0.012), as well as

Chironomidae Se (y = 3730 – 9.3x, R2 = 0.73, P = 0.007) and Hg (y = 88 – 0.18x, R2 = xi

0.59, P = 0.025). Note: Values from study reaches located upstream of the point source were excluded from regression analysis...... 68

Figure 2.3. Relationship between downstream distance from an area of extensive point source inputs into the mainstem of the Scioto River and (a) As (diamond, solid line), Hg

(circle, dashed line), and Pb (x, dotted line) concentrations of Tetragnathidae and (b) Se concentrations of Tetragnathidae across all study reaches. Slopes for the linear regression models were highly significant for As (y = 764 – 2.6x, R2 = 0.68, P = 0.011),

Hg (y = 331 – 1.4x, R2 = 0.72, P = 0.007), Pb (y = 526 – 2.9x, R2 = 0.93, P < 0.001) and

Se (y = 8.8 – 0.004x, R2 = 0.63, P = 0.019). Note: Values from study reaches located upstream of the point source were excluded from regression analysis...... 69

Figure 2.4. Relationship between Urbanization Axis (i.e., PC1) and F. subsericea density. The slope for the linear regression model was highly significant (y = 3.8 +

0.54x, R2 = 0.83, P < 0.001). Inset: Comparison of F. subsericea density across reaches grouped by urban, agricultural, and mixed land uses (F = 8.81, P = 0.010, N = 9; Tukey’s

HSD, P < 0.05). Different letters signify land uses for which F. subsericea density was significantly different. All values are shown as mean ± 1SE...... 70

xii

(y = 75.5 + 4.9x, R2 = 0.42, P = 0.060) and highly significant for F. subsericea (y = 22.6

+ 18.0x, R2 = 0.70, P = 0.005)...... 71

Figure 2.6. Relationship between Shoreline Habitat Axis (i.e., PC3) and (a) As and (b)

Se flux; (c) relationship between Mature Trees Axis (i.e., PC2) and Pb flux to

Tetragnathidae. Slopes for the regression models were nearly significant for As (y =

0.008 + 0.003x, R2 = 0.51, P = 0.031), and significant for Se (y = 0.079 + 0.024x, R2 =

0.58, P = 0.016) and for Pb (y = 0.005 – 0.001x, R2 = 0.52, P = 0.012) ...... 73

Figure 2.7. (a) Pb export by emergent aquatic insects across reaches grouped by urban, agricultural, and mixed land-use classes (F = 1.40, P = 0.316, N = 9; Tukey’s HSD, P <

0.05), (b) Pb flux to Tetragnathidae across reaches grouped by urban, agricultural, and mixed land-use classes (F = 6.10, P = 0.025, N = 9; Tukey’s HSD, P < 0.05), and (c) Pb flux to F. subsericea across reaches grouped by urban, agricultural, and mixed land-use classes (F = 8.68, P = 0.017, N = 9; Tukey’s HSD, P < 0.05). Different letters signify land-use types for which results were significantly different. All values are shown as mean ± 1SE...... 74

Figure 2.8. Relationship between Urbanization Axis (i.e., PC1) and Pb flux to F. subsericea. The slope for the linear regression model was highly significant (y = 0.005 +

0.001x, R2 = 0.71, P = 0.004)...... 75

Figure 3.1. Eleven study reaches of the Scioto River basin, Ohio, USA. Popouts indicate riverine study reaches and nest box locations, and are not to scale...... 107 xiii

Figure 3.2. Relationship between Hg concentrations (log10) in riparian swallows and sediment (ppb). The slope for the regression model was highly significant (y = 2.95 +

0.46x, R2 = 0.23, P = 0.030) ...... 108

Figure 3.3. Dual isotope plots of δ13C and δ15N values (mean ± 1SE) for the Scioto River basin (Ohio, USA) aquatic and terrestrial food-web components of the Scioto River basin

(Ohio, USA) in (a) rural and (b) urban reaches...... 109

Figure 3.4. Relationship between riparian swallow and periphyton δ13C at urban (solid line) and rural (dotted line) reaches. The slopes for the regression models were highly significant for urban reaches (y = -29.91 – 0.27x, R2 = 0.69, P < 0.001), but not significant for rural reaches (y = -25.6 - 0.08x, R2 = 0.11, P = 0.106)...... 110

Figure 3.5. Contribution, by percentage of biomass, of aquatic emergent insects to swallow boluses from nest box reaches...... 111

15 Figure 3.6. Relationship between swallow Se concentration (log10) and (a) swallow δ N and (b) swallow δ13C. The slopes for the regression models were significant for both

δ15N (y = 3.26 + 0.55x, R2 = 0.76, P < 0.001) and δ13C (y = 1.90 - 0.30x, R2 = 0.27, P =

0.032) ...... 112

Figure 3.7. Relationship between Urbanization Axis (i.e., PC1) and (a) Hg concentration

(log10) and (b) Se concentration (log10) in riparian swallows. Degree of urbanization increases from -3 to 4. The slope for the regression models were significant for both Hg

xiv

(y = 4.91-0.11x, R2 = 0.38, P = 0.004) and Se (y = 9.12 – 0.17x, R2 = 0.63, P < 0.001)…

...... 113

Figure 3.8. Relationship between Urbanization Axis (i.e., PC1) and 15N in riparian swallows. Shoreline habitat increases from -1.5 to 0.5. The slope for the regression model was highly significant (y = 10.9 - 0.23x, R2 = 0.38, P = 0.031)...... 114

Chapter 1: Background and Literature Review Introduction

Whereas material and energy exchanges in stream ecosystems have traditionally been viewed as dominated by inputs from land to water (Power 2004), recent research has highlighted the importance of reciprocal flows of energy. In particular, the critical subsidies provided by aquatic emergent insect prey to riparian predators are becoming increasingly recognized (Baxter, Fausch & Saunders 2005). In addition to subsidizing riparian and terrestrial food webs, these food-web linkages serve as an aquatic-to- terrestrial vector of contaminants (Walters et al. 2010b; Raikow et al. 2011; Sullivan &

Rodewald 2012). As such, movement of contaminants from aquatic to terrestrial systems through complex trophic linkages, including the human food web, and the risks of ecological exposure represent an area of growing interest (Sullivan & Rodewald 2012).

Contamination of aquatic systems is widespread (Fitzgerald et al. 1998; Lemly

2004; Wiener & Sandheinrich 2010). For example, nearly half of all US river sediment sampling stations were found to have exhibit chemical contamination sufficiently high to adversely affect aquatic life or human health (USEPA 2004). In spite of the scope and scale of the problem, aquatic-to-terrestrial transfers of contaminants is poorly resolved.

In particular, the role of adjacent landscape characteristics in regulating recipient communities (i.e., terrestrial predators of aquatic emergent insects) has not been explored, yet represents a fundamental component of the “flux equation”. Because land-

1 use and land-cover (LULC) alterations are known to affect both community structure and function of a suite of riparian organisms (Allan et al. 2003; Laeser, Baxter & Fausch

2005; Ives et al. 2011), the potential influence of LULC on riparian consumer density, distribution, and aquatic resource utilization represents a critical knowledge gap.

Contaminants in Aquatic Ecosystems

Sediments have been highlighted as the leading source of aquatic contamination in the United States, with nearly 10% of all surface waters containing sediment of sufficient contaminant concentration to pose health concerns for humans and wildlife

(USEPA 2004). Approximately 50% of sediment samples collected by the US

Geological Survey (USGS) National Water Quality Assessment (NAWQA) Program from 1992 to 2001 were found to contain organochlorine (OC) pesticide, whereas pesticide compounds were detected in 97% of streams draining agricultural or urban areas, and in 94% of streams in catchments characterized by mixed land use (i.e., a combination of urban, agricultural, and undeveloped area) (Gilliom et al. 2006).

Table 1.1. Common aquatic contaminants, their impacts on humans and wildlife, and major pathways through which they enter aquatic systems.

Contaminant Human Impacts Wildlife Impacts Sources Atmospheric deposition Mortality, reproductive failure Mortality, reproductive failure (Fitzgerald et al. 1998, Wiener Mercury (Scheuhammer et al. 2007, (Scheuhammer et al. 2007, et al. 2002, Mason 2005, Mergler et al. 2007) Mergler et al. 2007) Walters et al. 2010), coal-fired power plants (U.S. EPA 2011) Banned in 1979, but still found Disruption of immune system Hormone imbalance of thyroid in older products such as function and growth PCB and gonad (Dorea 2008, plastics, motor oil, hydraulic (McFarland and Clarke 1989, Abdelouahab et al. 2008) oil, electrical equipment (U.S. Duffy 2002) EPA 2011)

Banned in 1970's but persists Carcinogen (U.S. EPA 1975, Carcinogen (U.S. EPA 1975, in environment from use as Dieldrin U.S. EPA 1990, Nowell et al. U.S. EPA 1990, Nowell et al. agricultural presticide (Nowell 2009) 2009) et al. 2009)

Birth defects (Winchester et al. Endocrine disruption (Hayes et Most common herbicide in Atrazine 2009), low birth weight al. 2002) United States (USDA 1994) (Ochoa-Acuna 2009)

Pathways of aquatic contamination are linked to both chemical properties of the pollutant as well as physical characteristics of the aquatic system, and may be quite complex. As a result, few if any aquatic ecosystems worldwide have escaped some level and type of contamination, as even remote freshwater systems receive atmospherically transported contaminants (Fitzgerald et al. 1998; Blais 2005).

The effects of contaminants in aquatic and terrestrial systems have been well- documented (Kidd et al. 1998; Borga et al. 2004; Runck 2007; Walters et al. 2010a;

Table 1.1) particularly relative to bioaccumulation of contaminants in aquatic food webs

(Maul et al. 2006; Cristol et al. 2008). Mercury (Hg), for example, is known to cause mortality, reproductive failure, and other detrimental health effects in both humans and wildlife (Mergler et al. 2007). For example, Hg was correlated with altered plasma corticosterone and thyroid hormone concentrations in tree swallow (Tachycineta bicolor) nestlings in Virginia (Wada et al. 2009; Wada et al. 2010) and Forster’s tern (Sterna forsteri) nestlings in California (Herring, Ackerman & Herzog 2012). Polychlorinated biphenyls (PCBs) have been implicated in immunosuppression, hepatic metabolic disorders, and altered neuroendocrine function in Ohio River paddlefish (Polyodon spathula) (Gundersen et al. 2000). Organocholorine pesticides, such as dieldrin, have been linked with reproductive impairment (Garcia-Reyero et al. 2006; Martyniuk et al.

2010) in fish. Heavy metal bioaccumulation (including arsenic (As), cadmium (Cd), copper (Cu), nickel (Ni), and lead (Pb) was found to suppress the immune system of a species of wood ant (Formica aquilonia) in Finland (Sorvari et al. 2007).

Aquatic-Terrestrial Trophic Linkages

Rivers and their adjacent riparian zones are tightly-linked through energy exchanges, and reciprocal transfers of energy through these linkages are essential for functional, healthy ecosystems (Figure 1.1). Transfers of energy, particularly in terms of carbon flow, between terrestrial and freshwater ecosystems are often seen as unidirectional pathways in which terrestrially-derived organic matter, nutrients, and biota provide energy to aquatic consumers (Covich, Palmer & Crowl 1999; Power et al. 2004). In fish, for example, terrestrial insects have been shown to affect individual diet and condition

(Nakano & Murakami 2001), community composition and dynamics (Kawaguchi,

Taniguchi & Nakano 2003), and behavior and population demographics (Baxter, et al

2005; Nakano, et al 1999).

Figure 1.1. Representation of reciprocal food-web linkages (e.g., energy flows as represented by arrows) in a stream-riparian ecosystem (from Sullivan and Rodewald 2012)

However, recent research has highlighted the role of reverse flows of energy exchanges from aquatic to riparian zones in providing an important trophic subsidy to riparian and terrestrial food webs (Baxter, Fausch & Saunders 2005). In particular, research has shown that aquatic prey subsidies can constitute an important component of the energy budgets of riparian consumers. Jackson & Fisher (1986), for example, determined that only ~3% of aquatic emergent insect biomass returned to the stream, and suggested that most aquatic emergent insects were preyed upon by terrestrial consumers. 5

Specifically, aquatic emergent insects provide prey subsidies to multiple taxa: birds

(Blancher & McNicol 1991; Mengelkoch, Niemi & Regal 2004), beetles (Hering &

Plachter 1997), salamanders (Burton 1976), spiders (Williams, Ambrose & Browning

1995; Akamatsu, Toda & Okino 2004), bats (Rainey 2000), and adult odonates

(Sukhacheva 1996), among others.

Web-weaving spiders of the family Tetragnathidae (Figure 1.2) are known to rely heavily on emergent aquatic insect prey (Akamatsu, Toda & Okino 2004; Kato, Iwata &

Wada 2004). For example, Akamatsu, Toda, & Okino (2004) estimated the contribution of aquatic emergent insects to riparian web-building spiders in Japan at 84%, and found that the proportion of aquatic diet in web-building spiders, including Tetragnathidae, was higher in those with webs nearest the river than in those residing further inland.

Similarly, Burdon & Harding (2008) observed that riparian spider biomass and web density were closely related to the biomass of aquatic emergent insects in New Zealand.

Kato et al. (2003) found that tetragnathid spider density declined in response to an experimental reduction in aquatic insect prey during months when aquatic insect emergence would normally peak in Japan, suggesting that the spiders and aquatic prey items are tightly linked.

Riparian ants can also take advantage of aquatic nutrient subsidies. For example,

Paetzold et al. (2006) observed an increase in abundance of riparian Manica rubida when stream-derived invertebrate prey was experimentally added to the stream. Though stable isotope analysis indicated that aquatic insects were not a major dietary component of this species, the ants did respond strongly to an increase in readily available prey items, similar to what might occur during a flood pulse or a large emergence event.

Guzman & Castano-Meneses (2007) found that Camponotus rubrithorax, a common ant species in the Tehuacán Valley region of Mexico, tended to travel fairly

7 short distances in search of food, and selected food resources based on their abundance.

Previous work also indicates that foraging ants might selectively feed on nitrogen-rich resources, like invertebrates (Kay 2002; Bihn, Verhaagh & Brand 2008; Ness, Morris &

Bronstein 2009; Pearce-Duvet & Feener 2010). Pearce-Duvet & Feener (2010) suggested that the ability of ants to detect resources is an important component of ant foraging ecology, especially for those ants inhabiting an area with a varied and unpredictable resource base, whereas Schatz et al. (1997) found that detection ranges for generalist ants may be rather small, in the range of 1-2 cm. These studies indicate that ants inhabiting areas within close proximity of aquatic resources may detect temporally patchy aquatic prey sources.

Aquatic emergent insects are also an important food source for many riparian bird species (Keast 1990; Iwata, Nakano & Murakami 2003). Birds in broad-leaved deciduous forests, in which terrestrial arthropod prey was only available for the 4-5 month leafing period per year, have been shown to be dependent on emergent aquatic insects during the leafless period, with up to 25.6% of the annual total energy budget of the riparian bird community coming from aquatic emergent insects (Jackson & Fisher

1986; Henschel, Mahsberg & Stumpf 2001; Iwata, Nakano & Murakami 2003). Iwata,

Nakano & Murakami (2003) reported that flycatchers and gleaners foraged intensively on aquatic emergent insects in a pattern that mirrored aquatic insect distribution, whereas

Sweeney & Vannote (1982) found that swarming activity by mayflies attracted nighthawks (Chordeiles minor) to areas of emergence.

Riparian swallow species [e.g., bank (Riparia riparia), northern rough-winged

(Stelgidopteryx serripennis), tree (Tachycineta bicolor), and cliff (Petrochelidon pyrrhonota) swallows] (Figure 1.3) are also highly dependent on aquatic emergent insects

(McCarty 1997). Supporting evidence comes from Blancher & McNicol (1991), who found that diet of nestling tree swallows in Ontario was comprised of over 50% aquatic insects. Similarly, Mengelkoch (2005) observed that tree swallows relied on aquatic insects for up to half of their energy budget, though females might rely much more heavily on aquatic insects during egg-laying. Riparian swallows’ dependence on riverine nesting locations combined with their dietary intake of aquatic emergent insects makes them excellent study organisms for investigations of contaminant bioaccumulation

(Bishop et al. 1999; Harris & Elliott 2000; Dods et al. 2005).

Figure 1.3. Fledgling northern rough-winged swallows (Stelgidopteryx serripennis) at the Scioto River, Columbus, OH, USA.

Aquatic-to-Terrestrial Contaminant Flux

The contribution of aquatic fluxes of energy (i.e., nutrients, organisms, and organic material) to riparian and terrestrial food webs suggests that rivers may also function as lateral exporters of contaminants, with a significant risk of “reterrestrialization” of contaminants contained within the aquatic system. In light of flows of carbon and energy via aquatic emergent insects, the potential transfer of contaminants from aquatic to riparian and terrestrial ecosystems is likely significant. However, to date, there are comparatively few studies identifying the potential for flux of contamination from aquatic to terrestrial systems (but see Cristol et al. (2008); Walters, Fritz & Otter (2008);

Walters et al. (2010b)). Additionally, the processes mediating contaminant flux remain largely unexplored.

Relationships between trophic linkages and aquatic contaminants have been widely-investigated (Clements & Newman 2002; Ward et al. 2010). In the Chilliwack

River of southern British Columbia, for example, resident American Dippers (Cinclus mexicanus) (i.e., those birds who remain in the mainstem Chilliwack River year-round and have high proportions of contaminated salmon fry in their diet) were found to have higher PCB and Hg concentrations than their migratory counterparts (i.e., those birds who are found on the Chilliwack mainstem only during the winter and who have a significantly smaller proportion of fish fry in their diet; (Morrissey, Bendell-Young &

Elliott 2004). Blais et al. (2007) describes “biovector transport” as the process by which contaminants that have accumulated in aquatic environments are transported by organisms through movement or foodweb linkages into associated systems. Anadromous fish, for example, are known biovectors of contaminants from marine or lake to stream ecosystems (Merna 1986; Zhang et al. 2001; Sarica et al. 2004). For instance,

Christensen et al. (2005) estimated that Pacific salmon (Oncorynchus spp.) returns in

Canada delivered organochlorine pesticides and PCBs to salmon-feeding grizzly bears

(Ursus arctos horribilis). Salmon carcasses are also known to be consumed by aquatic invertebrates, which can lead to bioaccumulation of aquatic contaminants and facilitate their movement into terrestrial foodwebs (Sarica et al. 2004).

Aquatic emergent insects are a critical vector in the transport of contaminants from aquatic-to-terrestrial ecosystems (Table 2). Raikow et al. (2011) found that araneid

11 spiders at the shoreline were tightly linked to aquatic insect prey, with the latter making up between 60 and 100% of their diet, while PCB concentrations reflected this linkage.

In turn, riparian spiders can serve as vectors of contamination to terrestrial predators feeding at higher trophic levels, including birds (Cristol et al. 2008; Walters et al. 2010b).

Many investigators (e.g., (Echols et al. 2004; Maul et al. 2006; Walters, Fritz & Otter

2008)) have documented the biomagnification of PCBs in tree swallows from aquatic insect prey. Walters et al. (2010b) found that spiders, an important food source for a variety of riparian birds, exhibited increased levels of PCBs that were correlated to higher volumes of aquatic insect prey, whereas Cristol et al. (2008) demonstrated that aquatic

Hg can bioaccumulate in terrestrial bird species living near contaminated water sources.

Table 1.2. Estimated mass of contaminants exported from aquatic to terrestrial systems by aquatic emergent insects.

Contaminant Export Estimate Reach Length Location Reference Hg 20 µg · m-2 · yr-1 N/A Sweden Larsson (1984) PCB 4.1 g · yr-1 2.1 km Tennessee Runck (2007) PCB 6.13g · yr-1 25 km South Carolina Walters et al. (2008)

Contamination of riparian bird species has also been shown to operate across broad temporal and spatial scales. For instance, Berglund et al. (2009) observed that pied flycatchers (Ficedula hypoleuca) exhibited elevated levels of Pb after local emissions had ceased, suggesting that insectivorous birds may remain contaminated for long periods of time after removal of the contaminant source. In tree swallows along a river in Virginia,

Jackson et al. (2011) found Hg concentration levels remained high over 100-km from the

12 point source, with no evidence of a decline in concentrations as distance from the point source increased.

Differential biomagnification patterns and influences have been also been found between adult and nestling birds. A study of Glaucus gulls (Larus hyperboreus) in the

Arctic noted that there was a positive correlation between age and bioaccumulation of contaminants (Malinga, Szefer & Gabrielsen 2010). Likewise, Evers et al. (2005) found that concentrations of methylmercury (MeHg) in adult tree swallows were higher than in juveniles, and Akearok et al. (2010) reported that the earliest hatchlings in clutches of arctic terns (Sterna paradisea), common eiders (Somateria mollissima), and long-tailed ducks (Clangula hyemalis) exhibited significantly higher contaminant concentrations than those hatching later. Hofer et al. (2010) observed similar intra-clutch variability among passerine birds. Tsipoura et al. (2008) suggested, however, that nestlings might efficiently remove mercury from their blood due to rapid growth and deposition of the contaminant in their feathers. Along the South River in Virginia, Bouland et al. (2012) found that aquatically-derived Hg shifted nestling sex ratios towards females in three bird species, including tree swallows.

Adjacent Land use and Land cover

Land use and land cover (LULC) have been shown to be key factors that affect many aspects of watershed structure and function (Allan, Erickson & Fay 1997; Cuffney et al. 2000). Land use within a watershed can interact with a suite of local-scale riparian conditions that may influence aquatic ecosystem condition, including changes in water

13 quality (Meador & Goldstein 2003). Land-use practices resulting in the loading of sediment, nutrients, and contaminants can also alter the metabolism of lakes and streams

(Debenest et al. 2010).

Landscape changes, and in particular alterations to riparian and nearshore habitat, might be expected to significantly alter species-habitat relationships, community dynamics, and network interactions, subsequently influencing both the pathways and magnitude of aquatic-to-terrestrial contaminant transport (Sullivan & Rodewald 2012).

Studies that have explored linked aquatic-terrestrial food webs offer supporting evidence.

For example, the combined effects of loss of riparian vegetation and stream channelization were found to suppress tetragnathid spider density by up to 70% in streams of northern Japan, which the authors partly attributed to a reduction in web- building structure via loss of vegetation structure (Laeser, Baxter & Fausch 2005). Ives et al. (2011) found that streamside vegetation community type was a significant predictor of riparian ant assemblages in Australia.

Changes in community structure and function can also be influenced by riparian

LULC (Polis, Power & Huxel 2004). For example, changes in the structure and composition of riparian vegetation have been shown to strongly influence a suite of riparian arthropod consumers. Kavazos & Wallman (2012), for instance, observed that urbanization was associated with concomitant increases in the exotic proportion of the invertebrate community and in the persistence of non-native species. Ives et al. (2011b) found that generalist species of ant, such as F. subsericea, were more abundant in more highly fragmented habitats like those found within urban environments.

Tetragnathidae are highly-dependent on riparian structure for web placement and foraging success (Riechart & Gillespie 1986). Williams et al. (1995) observed higher densities of Tetragnathidae on river banks with overhanging vegetation than on those without overhanging vegetation. Similarly, Chan et al. (2009) found that almost half of all riparian spider orb-webs were built on either overhanging vegetation or boulders in a stream in China. Williams et al. (1995) found that tetragnathid spiders caught the greatest amount of aquatic prey in webs placed within 2-4m of the stream and in bushes and shrubs overhanging the stream at a height of up to 2m. Other studies have shown that aquatic insect abundance decreases exponentially with distance from the stream channel (Petersen et al. 1999; Iwata, Nakano & Murakami 2003).

Changes in LULC can also influence avian community dynamics in riparian landscapes (Rottenborn 1999; Miller, Wiens & Hobbs 2000; Lussier et al. 2006).

Rottenborn (1999) observed that riparian bird species richness and density decreased in urban areas adjacent to bridges and with lower native vegetation coverage, and found positive relationships between species richness and the width of riparian habitat and distance from the nearest human structure. In Columbus, OH urban riparian systems,

Bakermans & Rodewald (2006) found that acadian flycatchers (Empidonax virescens) tended to avoid dense understory vegetation that may be associated with increasing cover of invasive plant species like amur honeysuckle (Lonicera maackii), while Ausprey &

Rodewald (2011) found that survivorship of northern cardinals (Cardinalis cardinalis) and acadian flycatchers was highest in microhabitats that were the most structurally complex. Oneal and Rotenberry (2009) observed that riparian and insectivorous birds

15 responded most strongly to local-scale habitat variables, suggesting that riparian species may persist even in areas with a high degree of urban development if riparian vegetation remains intact.

In forested and agricultural landscapes, strong associations have also emerged.

Ghilain & Belisle (2008) observed that tree swallow nest-box occupancy was negatively related to agricultural land cover, specifically maize and soybean crops. Likewise,

Brown et al. (2002) observed that colony size in cliff swallows was related to increased habitat heterogeneity near the colony site, and that the largest colonies tended to have little agricultural land cover within their foraging range. Sullivan et al. (2007) found that insectivorous birds in Vermont preferred meandering, low gradient streams, and observed the highest abundance of insectivores in reaches with a riparian zone consisting of a mixture of open-canopied meadow and closed-canopied hardwood forest.

Study System: Scioto River Basin

The Scioto River drains approximately 16,882 km2 in central and southern Ohio, and supports a population density of approximately108 people/km2 (White 2005;

Blocksom 2011). As the second longest river in the state, the Scioto is a valuable resource to the residents of central and southern Ohio, providing critical ecosystem services for industry, agriculture, and municipalities, as well as a source of income and recreation for citizens. The Ohio Department of Natural Resources lists the Scioto River among the best places in Ohio to catch a variety of gamefish including smallmouth bass

(Micropterus dolomieu), spotted bass (Micropterus punctulatus), white crappie (Pomoxis

16 anularis), and flathead catfish (Pylodictis olivaris) (Ohio DNR 2000), and multiple canoe liveries allow visitors the opportunity recreate on the river. A joint project by the Ohio

River Valley Water Sanitation Commission (ORSANCO) and Midwest Biodiversity

Institute recorded over 70 different species of fish present in the Scioto River (Tewes et al. 2007), including a number of rare fishes such as the blue sucker (Cycleptus elongatus), the Tippecanoe darter (Etheostoma tippecanoe), and the shovelnose sturgeon

(Scaphirhynchus platorynchus).

The Scioto flows for 372 km through a complex landscape matrix from its headwaters in Auglaize County, OH to its confluence with the Ohio River in Portsmouth,

OH. Total catchment land-use by area is 69% agricultural, 21% forested, and 9% urbanized (White 2005; Blocksom 2011). In Columbus, the Olentangy River drains into the Scioto River. Twenty-two miles of the Olentangy upstream of Columbus are designated as a State Scenic River, yet it flows through such intense urban development areas that it is within a 30 minute drive of over 1.5 million people (Ohio DNR 2011).

Downstream of Columbus, the Scioto is free-flowing with riparian zones characterized by row crops (corn and soybean), deciduous forest, or urban landscapes. The river mainstem is generally turbid with steep, muddy banks, and exhibits relatively high levels of conductivity and nutrient loads (White 2005; Blocksom 2011).

Sediment contamination within the Scioto River has been documented by both federal and state agencies. According to the US EPA National Sediment Quality Survey

(USEPA 2004), multiple Scioto River sampling sites were classified as Tier 1, indicating that they exhibit “probable adverse effects on aquatic life” due to sediment

17 contamination. The Ohio EPA Division of Surface Water routinely monitors water quality in Ohio’s rivers and lakes through the use of biological surveys, and impairment sources include industrial and municipal point source contamination (Ohio EPA 2012).

A recent fish consumption advisory recommends limited ingestion of fish from the Scioto due to Hg and PCB concentrations found in fish tissue (Ohio EPA 2011). The USGS

Pesticide National Synthesis Project noted that pesticides are often found in waterways associated with agricultural or urban areas (USGS 2012), and the Scioto River basin is located within a region where both land-use types are common (Figure 4).

Figure 1.4. Likelihood of dieldrin in fish exceeding levels of concern for fish-eating wildlife (from USGS 2012).

Objectives and Summary

Fluvial systems have historically been viewed from a unidirectional perspective, with the majority of research addressing the influences of the surrounding terrestrial landscape on 18 streams and rivers. Only recently have reciprocal, aquatic-to-terrestrial transfers of material and energy between streams and their adjacent riparian zones received significant attention. Within the context of an increasing body of knowledge relative to these complex, bidirectional trophic linkages between aquatic and terrestrial ecosystems, and the ubiquitous presence of contaminants in aquatic systems, the overarching objectives of this research were to: (1) examine the potential influence of adjacent LULC

(i.e., riparian and nearshore) on linked aquatic-terrestrial invertebrate trophic dynamics and aquatic-to-terrestrial contaminant flux (Chapter 2), and (2) explore the relationships between adjacent patterns in LULC and the magnitude of contaminant concentration in riparian swallows as mediated by aquatic-to-terrestrial resource utilization. My intention is that this body of work will contribute to a growing understanding of aquatic-to- terrestrial contaminant transfer by addressing how adjacent landscape characteristics may regulate riparian consumer dynamics, and therefore, the uptake of aquatic contaminants.

It is my hope that this work will lead to improved conservation and management of river- riparian ecosystems and encourage a more holistic approach to monitoring and regulation. I conclude with a suite of appendices that provide additional data and information relative to the information presented within Chapters 2 and 3.

References

Abdelouahab, N., Mergler, D., Takser, L., Vanier, C., St-Jean, M., Baldwin, M., Spear, P.A. & Chan, H.M. (2008) Gender differences in the effects of organochlorines, mercury, and lead on thyroid hormone levels in lakeside communities of Quebec (Canada). Environmental Research, 107, 380-392.

Akamatsu, F., Toda, H. & Okino, T. (2004) Food source of riparian spiders analyzed by using stable isotope ratios. Ecological Research, 19, 655-662.

Akearok, J.A., Hebert, C.E., Braune, B.M. & Mallory, M.L. (2010) Inter- and intraclutch variation in egg mercury levels in marine bird species from the Canadian Arctic. Science of the Total Environment, 408, 836-840.

Allan, J.D., Erickson, D.L. & Fay, J. (1997) The influence of catchment land use on stream integrity across multiple spatial scales. Freshwater Biology, 37, 149-161.

Allan, J.D., Wipfli, M.S., Caouette, J.P., Prussian, A. & Rodgers, J. (2003) Influence of streamside vegetation on inputs of terrestrial invertebrates to salmonid food webs. Canadian Journal of Fisheries and Aquatic Sciences, 60, 309-320.

Ausprey, I.J. & Rodewald, A.D. (2011) Postfledgling survivorship and habitat selection across a rural-to-urban landscape gradient. Auk, 128, 293-302.

Bakermans, M.H. & Rodewald, A.D. (2006) Scale-dependent habitat use of Acadian Flycatcher (Empidonax virescens) in central Ohio. Auk, 123, 368-382.

Baxter, C.V., Fausch, K.D. & Saunders, W.C. (2005) Tangled webs: reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biology, 50, 201- 220.

Berglund, A.M.M., Klaminder, J. & Nyholm, N.E.I. (2009) Effects of Reduced Lead Deposition on Pied Flycatcher (Ficedula hypoleuca) Nestlings: Tracing Exposure Routes Using Stable Lead Isotopes. Environmental Science & Technology, 43, 208-213.

Bihn, J.H., Verhaagh, M. & Brand, R. (2008) Ecological stoichiometry along a gradient of forest succession: Bait preferences of litter ants. Biotropica, 40, 597-599.

Bishop, C.A., Mahony, N.A., Trudeau, S. & Pettit, K.E. (1999) Reproductive success and biochemical effects in tree swallows (Tachycineta bicolor) exposed to chlorinated hydrocarbon contaminants in wetlands of the Great Lakes and St. Lawrence River basin, USA and Canada. Environmental Toxicology and Chemistry, 18, 263-271.

Blais, J.M. (2005) Biogeochemistry of persistent bioaccumulative toxicants: processes affecting the transport of contaminants to remote areas. Canadian Journal of Fisheries and Aquatic Sciences, 62, 236-243.

Blancher, P.J. & McNicol, D.K. (1991) Tree swallow diet in relation to wetland acidity. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 69, 2629-2637.

Blocksom, K.A. (2011) Development of a regional macroinvertebrate index for large river bioassessment (vol 9, pg 313, 2009). Ecological Indicators, 11, 757-757.

Borga, K., Fisk, A.T., Hoekstra, P.F. & Muir, D.C.G. (2004) Biological and chemical factors of importance in the bioaccumulation and trophic transfer of persistent organochlorine contaminants in arctic marine food webs. Environmental Toxicology and Chemistry, 23, 2367-2385.

Bouland, A.J., White, A.E., Lonabaugh, K.P., Varian-Ramos, C.W. & Cristol, D.A. (2012) Female-biased offspring sex ratios in birds at a mercury-contaminated river. Journal of Avian Biology, 43, 244-251.

Brown, C.R., Sas, C.M. & Brown, M.B. (2002) Colony choice in Cliff swallows: Effects of heterogeneity in foraging habitat. Auk, 119, 446-460.

Burdon, F.J. & Harding, J.S. (2008) The linkage between riparian predators and aquatic insects across a stream-resource spectrum. Freshwater Biology, 53, 330-346.

Burton, T.M. (1976) An analysis of the feeding ecology of the salamanders (Amphibia urodela) of the Hubbard Brook Experimental Forest New Hampshire, USA. Journal of Herpetology, 10, 187-204.

Christensen, J.R., Macduffee, M., Macdonald, R.W., Whiticar, M. & Ross, P.S. (2005) Persistent organic pollutants in British Columbia grizzly bears: Consequence of divergent diets. Environmental Science & Technology, 39, 6952-6960.

Clements, W.H. & Newman, M.C. (2002) Community Ecotoxicology. John Wiley & Sons, Chichester, UK.

Covich, A.P., Palmer, M.A. & Crowl, T.A. (1999) The role of benthic invertebrate species in freshwater ecosystems. BioScience, 49, 119-127.

Cristol, D.A., Brasso, R.L., Condon, A.M., Fovargue, R.E., Friedman, S.L., Hallinger, K.K., Monroe, A.P. & White, A.E. (2008) The movement of aquatic Mercury through terrestrial food webs. Science, 320, 335-335.

Cuffney, T.F., Meador, M.R., Porter, S.D. & Gurtz, M.E. (2000) Responses of physical, chemical, and biological indicators of water quality to a gradient of agricultural land use in the Yakima River Basin,Washington. Environmental Monitoring and Assessment, 64, 259-270.

Debenest, T., Silvestre, J., Coste, M. & Pinelli, E. (2010) Effects of Pesticides on Freshwater Diatoms. Reviews of Environmental Contamination and Toxicology, Vol 203 (ed. D.M. Whitcare), pp. 87-103. Springer, New York.

Dods, P.L., Birmingham, E.M., Williams, T.D., Ikonomou, M.G., Bennie, D.T. & Elliott, J.E. (2005) Reproductive success and contaminants in tree swallows (Tachycineta bicolor) breeding at a wastewater treatment plant. Environmental Toxicology and Chemistry, 24, 3106-3112.

Dorea, J.G. (2008) Elevated PCB levels in anglers and unsuspected transport of pollutants from aquatic food webs into human foods. Environmental Research, 108, 268-268.

Duffy, J.E., Carlson, E., Li, Y., Prophete, C. & Zelikoff, J.T. (2002) Impact of polychlorinated biphenyls (PCBs) on the immune function of fish: age as a variable in determining adverse outcome. Marine Environmental Research, 54, 559-563.

Echols, K.R., Tillitt, D.E., Nichols, J.W., Secord, A.L. & McCarty, J.P. (2004) Accumulation of PCB congeners in nestling tree swallows (Tachycineta bicolor) on the Hudson River, New York. Environmental Science & Technology, 38, 6240-6246.

Evers, D.C., Burgess, N.M., Champoux, L., Hoskins, B., Major, A., Goodale, W.M., Taylor, R.J., Poppenga, R. & Daigle, T. (2005) Patterns and interpretation of mercury exposure in freshwater avian communities in northeastern North America. Ecotoxicology, 14, 193-221.

Fitzgerald, W.F., Engstrom, D.R., Mason, R.P. & Nater, E.A. (1998) The case for atmospheric mercury contamination in remote areas. Environmental Science & Technology, 32, 1-7.

Garcia-Reyero, N., Barber, D.S., Gross, T.S., Johnson, K.G., Sepulveda, M.S., Szabo, N.J. & Denslow, N.D. (2006) Dietary exposure of largemouth bass to OCPs changes expression of genes important for reproduction. Aquatic Toxicology, 78, 358-369. 22

Ghilain, A. & Belisle, M. (2008) Breeding success of tree swallows along a gradient of agricultural intensification. Ecological Applications, 18, 1140-1154.

Gilliom, R.J., Barbash, J.E., Crawford, C.G., Hamilton, P.A., Martin, J.D., Nakagaki, N., Nowell, L.H., Scott, J.C., Stackelburg, P.E., Thelin, G.P., and D.M. Wolock (2006) The quality of our nation's waters - Pesticides in the nation's streams and ground water, 1992-2001: U.S. Geological Survey Circular 1291, 172p

Gundersen, D.T., Miller, R., Mischler, A., Elpers, K., Mims, S.D., Millar, J.G. & Blazer, V. (2000) Biomarker response and health of polychlorinated biphenyl- and chlordane-contaminated paddlefish from the Ohio River basin, USA. Environmental Toxicology and Chemistry, 19, 2275-2285.

Guzman Mendoza, R. & Castano-Meneses, G. (2007) Selected foraging activity of Camponotus rubrithorax (Hymenoptera : Formicidae) in the Zapotitlan Salinas Valley, Puebla, Mexico. Sociobiology, 50, 435-448.

Harris, M.L. & Elliott, J.E. (2000) Reproductive success and chlorinated hydrocarbon contamination in tree swallows (Tachycineta bicolor) nesting along rivers receiving pulp and paper mill effluent discharges. Environmental Pollution, 110, 307-320.

Hayes, T.B., Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart, A.A. & Vonk, A. (2002) Hermaphroditic, demasculinized frogs after exposure to the herbicide atrazine at low ecologically relevant doses. Proceedings of the National Academy of Sciences of the United States of America, 99, 5476-5480.

Henschel, J.R., Mahsberg, D. & Stumpf, H. (2001) Allochthonous aquatic insects increase predation and decrease herbivory in river shore food webs. Oikos, 93, 429-438.

Hering, D. & Plachter, H. (1997) Riparian ground beetles (Coeloptera, Carabidae) preying on aquatic invertebrates: A feeding strategy in alpine floodplains. Oecologia, 111, 261-270.

Herring, G., Ackerman, J.T. & Herzog, M.P. (2012) Mercury Exposure May Suppress Baseline Corticosterone Levels in Juvenile Birds. Environmental Science & Technology, 46, 6339-6346.

Hofer, C., Gallagher, F.J. & Holzapfel, C. (2010) Metal accumulation and performance of nestlings of passerine bird species at an urban brownfield site. Environmental Pollution, 158, 1207-1213.

Ives, C.D., Hose, G.C., Nipperess, D.A. & Taylor, M.P. (2011a) Environmental and landscape factors influencing ant and plant diversity in suburban riparian corridors. Landscape and Urban Planning, 103, 372-382.

Iwata, T., Nakano, S. & Murakami, M. (2003) Stream meanders increase insectivorous bird abundance in riparian deciduous forests. Ecography, 26, 325-337.

Jackson, A.K., Evers, D.C., Folsom, S.B., Condon, A.M., Diener, J., Goodrick, L.F., McGann, A.J., Schmerfeld, J. & Cristol, D.A. (2011) Mercury exposure in terrestrial birds far downstream of an historical point source. Environmental Pollution, 159, 3302-3308.

Jackson, J.K. & Fisher, S.G. (1986) Secondary production, emergence, and export of aquatic insects of a Sonoran Desert stream. Ecology, 67, 629-638.

Kato, C., Iwata, T., Nakano, S. & Kishi, D. (2003) Dynamics of aquatic insect flux affects distribution of riparian web-building spiders. Oikos, 103, 113-120.

Kato, C., Iwata, T. & Wada, E. (2004) Prey use by web-building spiders: stable isotope analyses of trophic flow at a forest-stream ecotone. Ecological Research, 19, 633- 643.

Kavazos, C.R.J. & Wallman, J.F. (2012) Community composition of carrion-breeding blowflies (Diptera: Calliphoridae) along an urban gradient in south-eastern Australia. Landscape and Urban Planning, 106, 183-190.

Kay, A. (2002) Applying optimal foraging theory to assess nutrient availability ratios for ants. Ecology, 83, 1935-1944.

Keast, A. (1990) The annual cycle and activity on the breeding grounds in a Canadian broad-leafed deciduous forest bird bommunity - relationship to the prey resource base. Keast, a. (Ed.). Biogeography and Ecology of Forest Bird Communities. V+410p. Spb Academic Publishing Bv: the Hague, Netherlands. Illus. Maps, 197- 214.

Kidd, K.A., Hesslein, R.H., Ross, B.J., Koczanski, K., Stephens, G.R. & Muir, D.C.G. (1998) Bioaccumulation of organochlorines through a remote freshwater food web in the Canadian Arctic. Environmental Pollution, 102, 91-103.

Laeser, S.R., Baxter, C.V. & Fausch, K.D. (2005) Riparian vegetation loss, stream channelization, and web-weaving spiders in northern Japan. Ecological Research, 20, 646-651.

Lemly, A.D. (2004) Aquatic selenium pollution is a global environmental safety issue. Ecotoxicology and Environmental Safety, 59, 44-56.

Lussier, S.M., Enser, R.W., Dasilva, S.N. & Charpentier, M. (2006) Effects of habitat disturbance from residential development on breeding bird communities in riparian corridors. Environmental Management, 38, 504-521.

Malinga, M., Szefer, P. & Gabrielsen, G.W. (2010) Age, sex and spatial dependent variations in heavy metals levels in the Glaucous Gulls (Larus hyperboreus) from the Bjornoya and Jan Mayen, Arctic. Environmental Monitoring and Assessment, 169, 407-416.

Martyniuk, C.J., Feswick, A., Spade, D.J., Kroll, K.J., Barber, D.S. & Denslow, N.D. (2010) Effects of acute dieldrin exposure on neurotransmitters and global gene transcription in largemouth bass (Micropterus salmoides) hypothalamus. Neurotoxicology, 31, 356-366.

Maul, J.D., Belden, J.B., Schwab, B.A., Whiles, M.R., Spears, B., Farris, J.L. & Lydy, M.J. (2006) Bioaccumulation and trophic transfer of polychlorinated biphenyls by aquatic and terrestrial insects to tree swallows (Tachycineta bicolor). Environmental Toxicology and Chemistry, 25, 1017-1025.

McFarland, V.A. & Clarke, J.U. (1989) Environmental occurrence, abundance, and potential toxicity of polychlorinated biphenyl congeners - considerations for a congener-specific analysis. Environmental Health Perspectives, 81, 225-239.

Meador, M.R. & Goldstein, R.M. (2003) Assessing water quality at large geographic scales: relations among land use, water physicochemistry, riparian condition, and fish community structure. Environmental Management, 31, 504-517.

Mengelkoch, J.M., Niemi, G.J. & Regal, R.R. (2004) Diet of the nestling Tree Swallow. Condor, 106, 423-429.

Mergler, D., Anderson, H.A., Chan, L.H.M., Mahaffey, K.R., Murray, M., Sakamoto, M. & Stern, A.H. (2007) Methylmercury exposure and health effects in humans: A worldwide concern. Ambio, 36, 3-11.

Merna, J.W. (1986) Contamination of stream fishes with chlorinated hydrocarbons from eggs of Great Lakes salmon. Transactions of the American Fisheries Society, 115, 69-74.

Miller, J.R., Wiens, J.A. & Hobbs, N.T. (2000) How does urbanization affect bird communities in riparian habitats? An approach and preliminary assessment. Nature Conservation. Conservation in production environments: Managing the matrix (ed. J.L.S.D.M.N. Craig), pp. 427-439.

Morrissey, C.A., Bendell-Young, L.I. & Elliott, J.E. (2004) Linking contaminant profiles to the diet and breeding location of American dippers using stable isotopes. Journal of Applied Ecology, 41, 502-512.

Ness, J.H., Morris, W.F. & Bronstein, J.L. (2009) For ant-protected plants, the best defense is a hungry offense. Ecology, 90, 2823-2831.

Ochoa-Acuna, H., Frankenberger, J., Hahn, L. & Carbajo, C. (2009) Drinking-Water Herbicide Exposure in Indiana and Prevalence of Small-for-Gestational-Age and Preterm Delivery. Environmental Health Perspectives, 117, 1619-1624.

Ohio Department of Natural Resources (2000) A Guide to Ohio's Streams. Ohio Chapter American Fisheries Society, Columbus, OH.

Ohio Department of Natural Resources (2011). Olentangy State Scenic River. http://ohiodnr.com/watercraft/sr/tabid/2565/Default.aspx. Accessed December 18, 2011.

Ohio Environmental Protection Agency (2011) http://www.epa.ohio.gov/dsw/fishadvisory/limitmeals.aspx. Accessed February 11, 2012.

Ohio Environmental Protection Agency (2012)http://www.epa.ohio.gov/dsw/tmdl/monitoring_LowerSciotoRiver.aspx. Accessed February 11, 2012.

Oneal, A.S. & Rotenberry, J.T. (2009) Scale-dependent habitat relations of birds in riparian corridors in an urbanizing landscape. Landscape and Urban Planning, 92, 264-275.

Paetzold, A., Bernet, J.F. & Tockner, K. (2006) Consumer-specific responses to riverine subsidy pulses in a riparian arthropod assemblage. Freshwater Biology, 51, 1103- 1115.

Pearce-Duvet, J.C. & Feener, D.H. (2010) Resource discovery in ant communities: do food type and quantity matter? Ecological Entomology, 35, 549-556.

Petersen, I., Winterbottom, J.H., Orton, S., Friberg, N., Hildrew, A.G., Spiers, D.C. & Gurney, W.S.C. (1999) Emergence and lateral dispersal of adult Plecoptera and Trichoptera from Broadstone Stream, UK. Freshwater Biology, 42, 401-416.

Polis, G.A., Power, M.E. & Huxel, G.R. (2004) Food webs at the landscape level. University of Chicago Press, Chicago, Illinois.

Power, M.E., Rainey, W.E., Parker, M.S., Sabo, J.L., Smyth, A., Khandwala, S., Finlay, J.C., McNeely, F.C., Marsee, K. & Anderson, C. (2004) River to watershed subsidies in an old-growth conifer forest. Food webs at the landscape level (eds G.A. Polis, M.E. Power & G.R. Huxel), pp. 217-240. The University of Chicago Press, Chicago, IL, USA.

Power, M.E., W. E. Rainey, M.S. Parker, J. L. Sabo, A. Smyth, S. Khandwala, J.C. Finlay, F.C. McNeely, K. Marsee, and C. Anderson (2004) River to watershed subsidies in an old-growth conifer forest. Foodwebs at the Landscape Level (ed. M.E.P. G.A. Polis, and G. Huxel). University of Chicago Press, Chicago, IL.

Raikow, D.F., Walters, D.M., Fritz, K.M. & Mills, M.A. (2011) The distance that contaminated aquatic subsidies extend into lake riparian zones. Ecological Applications, 21, 983-990.

Rainey, P.M.E.a.W.E. (2000) Food webs and resource sheds: Towards spatially delimiting trophic interactions. Ecological Consequences of Habitat Heterogeneity (ed. J.E.A. Hutchings M.J., and A.J.A. Stewart). Blackwell Scientific, Oxford, U.K.

Rottenborn, S.C. (1999) Predicting the impacts of urbanization on riparian bird communities. Biological Conservation, 88, 289-299.

Runck, C. (2007) Macroinvertebrate production and food web energetics in an industrially contaminated stream. Ecological Applications, 17, 740-753.

Sarica, J., Amyot, M., Hare, L., Doyon, M.R. & Stanfield, L.W. (2004) Salmon-derived mercury and nutrients in a Lake Ontario spawning stream. Limnology and Oceanography, 49, 891-899.

Schatz, B., Lachaud, J.P. & Beugnon, G. (1997) Graded recruitment and hunting strategies linked to prey weight and size in the ponerine ant Ectatomma ruidum. Behavioral Ecology and Sociobiology, 40, 337-349.

Sorvari, J., Rantala, L.M., Rantala, M.J., Hakkarainen, H. & Eeva, T. (2007) Heavy metal pollution disturbs immune response in wild ant populations. Environmental Pollution, 145, 324-328.

Sukhacheva, G.A. (1996) Study of the natural diet of adult dragonflies using an immunological method. Odonatologica, 25, 397-403.

Sullivan, S.M.P. & Rodewald, A.D. (2012) In a state of flux: The energetic pathways that move contaminants from aquatic to terrestrial environments. Environmental toxicology and chemistry / SETAC, 31, 1175-1183.

Sullivan, S.M.P., Watzin, M.C. & Keeton, W.S. (2007) A riverscape perspective on habitat associations among riverine bird assemblages in the Lake Champlain Basin, USA. Landscape Ecology, 22, 1169-1186.

Survey, U.S.G. (2009) Pesticide National Synthesis Project.

Sweeney, B.W. & Vannote, R.L. (1982) Population synchrony in mayflies - a predator satiation hypthesis. Evolution, 36, 810-821.

Tewes R.A., E.E.B., Thomas J., & Hershla L.E. (2007) Evaluation and Development of Large River Biological Assessment Methods and Standardized Protocols for Region V - Boat Electrofishing Methods Comparison Study. Midwest Biodiversity Institute, Columbus, OH.

Tsipoura, N., Burger, J., Feltes, R., Yacabucci, J., Mizrahi, D., Jeitner, C. & Gochfeld, M. (2008) Metal concentrations in three species of passerine birds breeding in the Hackensack Meadowlands of New Jersey. Environmental Research, 107, 218- 228.

United States Geological Survey (2012) http://water.usgs.gov/nawqa/pnsp/. Accessed February 11, 2012

Wada, H., Cristol, D.A., McNabb, F.M.A. & Hopkins, W.A. (2009) Suppressed Adrenocortical Responses and Thyroid Hormone Levels in Birds near a Mercury- Contaminated River. Environmental Science & Technology, 43, 6031-6038.

Wada, H., Yates, D.E., Evers, D.C., Taylor, R.J. & Hopkins, W.A. (2010) Tissue mercury concentrations and adrenocortical responses of female big brown bats (Eptesicus fuscus) near a contaminated river. Ecotoxicology, 19, 1277-1284.

Walters, D.M., Blocksom, K.A., Lazorchak, J.M., Jicha, T., Angradi, T.R. & Bolgrien, D.W. (2010a) Mercury Contamination in Fish in Midcontinent Great Rivers of the United States: Importance of Species Traits and Environmental Factors. Environmental Science & Technology, 44, 2947-2953.

Walters, D.M., Fritz, K.M. & Otter, R.R. (2008) The dark side of subsidies: adult stream insects export organic contaminants to riparian predators. Ecological Applications, 18, 1835-1841.

Walters, D.M., Mills, M.A., Fritz, K.M. & Raikow, D.F. (2010b) Spider-Mediated Flux of PCBs from Contaminated Sediments to Terrestrial Ecosystems and Potential Risks to Arachnivorous Birds. Environmental Science & Technology, 44, 2849- 2856.

Ward, M.P., Jablonski, C., Semel, B. & Soucek, D. (2010) The biological pathway and effect of PCBs on common terns in Lake Michigan. Ecotoxicology, 19, 1513- 1522.

White D., and Miller M. (2005) Ohio River Basin. Rivers of North America (ed. C.E. Cushing), pp. 375-424. Elsevier Academic Press, Burlington, MA.

Wiener, J.G. & Sandheinrich, M.B. (2010) Contaminants in the Upper Mississippi River: historic trends, responses to regulatory controls, and emerging concerns. Hydrobiologia, 640, 49-70.

Williams, D.D., Ambrose, L.G. & Browning, L.N. (1995) Trophic dynamics of 2 sympatric species of riparian spider (Araneae, Tetragnathidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie, 73, 1545-1553.

Winchester, P.D., Huskins, J. & Ying, J. (2009) Agrichemicals in surface water and birth defects in the United States. Acta Paediatrica, 98, 664-669.

Zhang, X.M., Naidu, A.S., Kelley, J.J., Jewett, S.C., Dasher, D. & Duffy, L.K. (2001) Baseline concentrations of total mercury and methylmercury in salmon returning via the Bering Sea (1999-2000). Marine Pollution Bulletin, 42, 993-997.

Chapter 2: Land use mediates aquatic-to-terrestrial contaminant flux via

aquatic emergent insects

Jeremy M. Alberts1,2 and S. Mažeika P. Sullivan1

1School of Environment & Natural Resources, The Ohio State University, 2021 Coffey

Road, Columbus, OH 43210

2Corresponding author: email: [email protected]; Fax: (01)614-292-7342

Abstract

Aquatic emergent insects are important vectors of contaminant transfer from riverine sediment to terrestrial food webs. However, the extent to which land use mediates the magnitude of aquatic-to-terrestrial contaminant transfers remains largely unexplored. To this end, I investigated aquatic-to-terrestrial contaminant fluxes at nine reaches of the Olentangy and Scioto Rivers (OH, USA), representing urban, agricultural, and mixed land uses. We collected emerging adult aquatic insects, ants (Formica subsericea), and spiders of the family Tetragnathidae from study reaches, analyzed them for δ13C stable isotopes, and subsequently tested them for a suite of toxic elements including Arsenic (As), Selenium (Se), Lead (Pb), and Mercury (Hg). Two-source mixing models indicated that Tetragnathidae were highly reliant on aquatic insects ( ̅ = 76.9%,

SD = 8.9%), whereas ant dependence was less but with greater variability ( ̅ = 27.8%, SD

= 25.1%). Characteristics of shoreline habitat including standing dead trees and % overhanging vegetation explained 70 and 42% of the variation in the contribution of aquatic prey to F. Subsericea and Tetragnathidae, respectively. Spider density was closely related to land-use and land-cover characteristics associated with urbanization (% impervious surfaces, % invasive shrubs, population density) and nearshore habitat.

Shoreline habitat also was strongly related to the overall flux (i.e., contaminant load assimilated into consumer tissue) of Se (R2 = 0.58) and As (R2 = 0.51) to the tetragnathid

32 spider assemblage, and Pb flux to spiders was higher in predominantly urban and agricultural reaches than in mixed reaches (F = 6.10, P = 0.025). Density of F. subsericea increased with urbanization (R2 = 0.83). Flux of As and Se to F. subsericea assemblages was positively related to urbanization (R2 = 0.70) as well as shoreline habitat, and Pb flux was higher in predominantly urban reaches than other land use types

(F = 8.68, P = 0.017). Collectively, our results indicate that landscape-mediated differences in aquatic resource utilization and density of riparian invertebrate consumers lead to a significant discrepancy between contaminant export by aquatic emergent insects and realized flux to riparian food webs. Improved understanding of the pathways and influences that regulate these aquatic-to-terrestrial contaminant fluxes are critical for effective risk management and remediation.

Introduction

Contamination of aquatic ecosystems is extremely widespread. For example, nearly half of almost 700,000 river miles assessed in the US have been designated as impaired or threatened with impairment (U.S. EPA 2000). The multiple pathways through which contaminants enter aquatic ecosystems are well described. Many contaminants, such as pesticides, enter aquatic ecosystems via surface runoff from tributaries (Blocksom et al. 2010). Atmospheric deposition is recognized as a major source of many pollutants, including polychlorinated biphenyls (PCBs) and mercury. As a result, few if any aquatic ecosystems worldwide have escaped some level and type of contamination, as even remote freshwater systems receive atmospherically transported contaminants (Fitzgerald et al. 1998; Blais 2005). Pathways are thusly quite variable, and linked to both chemical properties of the pollutant as well as physical characteristics of the aquatic system.

Once in aquatic environments, linkages between feeding relationships and contaminants have been widely-investigated (Clements and Newman 2002; Ward et al.

2010). Numerous examples have been cited relative to biomagnification of contaminants in piscivorous vertebrates (see Evers et al. 2005; Hinck et al. 2009). In cases of

“biovector transport”, contaminants that have accumulated in aquatic environments are transported to receptor sites by organisms through direct movement and/or food webs

(Blais et al. 2007). Anadromous fish, for example, are effective biovectors of contaminants from marine or lake to stream ecosystems (Merna 1986; Zhang et al. 2001;

Sarica et al. 2004), where contaminants can then be moved into terrestrial food webs. A

34 classic example comes from the Pacific Northwest, where Christensen et al. (2005) estimated that late-summer Pacific salmon (Oncorynchus spp.) returns in British

Columbia delivered 70% of volatile organochlorine (OC) pesticides, up to 85% of brominated PBDE congeners, and 90% of PCBs to salmon-feeding grizzly bears (Ursus arctos horribilis). Salmon carcasses are also consumed by aquatic invertebrates, which accumulate contaminants and move them through terrestrial food webs (Sarica et al.

2004).

Within this context, there is growing interest in the potential fluxes of contaminants through the energetic pathways that link aquatic and terrestrial ecosystems

(Sullivan and Rodewald 2012). In particular, aquatic insects that emerge from the stream as adults (hereafter, ‘aquatic emergent insects’) can be a critical food source for riparian and terrestrial consumers including arthropods, birds, mammals, and reptiles (reviewed in

Baxter et al. 2005). For example, Jackson and Fisher (1986) found that only 3% of aquatic emergent insect biomass returned to the stream after hatching, suggesting that most were consumed by terrestrial insectivores such as birds and spiders. Walters et al.

(2008b) estimated that approximately 6.13g · yr-1 of PCB was exported from the stream via riparian spider predation on aquatic emergent insects along a 25-km riparian corridor of a contaminated stream. In turn, riparian spiders can serve as vectors of contamination to terrestrial predators feeding at higher trophic levels (Cristol et al. 2008; Walters et al.

2010).

Landscape changes, and in particular alterations to riparian and nearshore habitat, are expected to significantly alter species-habitat relationships, community dynamics, and

35 network interactions, subsequently influencing both the pathways and magnitude of aquatic-to-terrestrial contaminant transport (Sullivan and Rodewald 2012). Linked stream-riparian food-web studies offer supporting evidence. Paetzold et al. ( 2011) found that chronic stream pollution decreased aquatic insect density resulting in a decrease in the density of riparian spiders of the families Araneidae, Linyphiidae, and

Tetragnathidae. Likewise, the combined effect of riparian vegetation loss and stream channelization was found to suppress Tetragnathidae density up to 70% in streams of northern Japan, partly implicating a reduction in web-building substrate via loss of vegetation structure (Laeser, Baxter and Fausch 2005).

Within this framework, we investigated the potential influence of adjacent land use and land cover (i.e., riparian and near shore) on linked aquatic-terrestrial invertebrate trophic dynamics and aquatic-to-terrestrial contaminant flux along a multiuse river system in Ohio, USA. We hypothesized that (1) riparian predators (spiders and ants) would be contaminated with toxins via consumption of aquatic emergent insects, (2) tetragnathid spiders would show a greater reliance than riparian ants (Formica subsericea) on aquatic food resources (as inferred from stable carbon [δ13C] and nitrogen

[δ15N] isotopes) and would have correspondingly higher concentrations of toxins, and (3) adjacent land use and land cover (LULC) would mediate aquatic-to-terrestrial flux of contaminants by influencing the both the relative aquatic resource utilization and density of riparian spiders and ants.

Materials and methods

STUDY SYSTEM AND EXPERIMENTAL DESIGN

The Scioto River is a multiuse river that drains approximately 16,882 km2 in central and southern Ohio and supports a population density of approximately 108 people

· km-2 (White et al. 2005; Blocksom & Johnson 2009). The Scioto flows for 372 km through a complex landscape matrix from its headwaters in Auglaize County, OH to its confluence with the Ohio River in Portsmouth, OH. Total catchment land-use by area is

69% agricultural, 21% forested, and 9% urbanized (White et al. 2005; Blocksom &

Johnson 2009). In Columbus, the Olentangy River joins the Scioto River. Twenty-two miles of the Olentangy River upstream of Columbus are designated as a State Scenic

River, yet the Olentangy flows through downtown Columbus – the 15th largest city in the

US (pop. 787,033 residents; US Census Bureau 2010). Downstream of Columbus, the

Scioto is free-flowing with riparian zones primarily characterized by deciduous forest, row crops (corn and soybean), and smaller urban centers. The river mainstem is generally turbid with steep, muddy banks, and exhibits relatively high levels of conductivity and nutrient loads (White et al. 2005; Blocksom & Johnson 2009).

Sediment contamination in the Scioto River is recognized by both federal and state agencies. According to the U.S. EPA National Sediment Quality Survey (2004), four Scioto sites were classified as Tier 1 sampling stations, indicating that they exhibited

“probable adverse effects on aquatic life” due to sediment contamination. A recent fish- consumption advisory from the Ohio EPA (2011) recommended limited ingestion of fish from the Scioto due to mercury and PCB concentrations found in fish tissue.

We conducted our study at nine river reaches (defined as a river segment based on similar valley features and channel geomorphology; ~1 km for this study) of the

Scioto River catchment, representing a mixture of urban, mixed (mainly forested and exurban), and agricultural land uses (Figure 1). Although study reaches were distributed along the length of the river, LULC patterns in the watershed and limited access to some stretches precluded equidistant sampling. However, reaches were separated by an average distance of 16.6 river kilometers, although there was significant variability (SD =

16.5km). We used ArcGIS 10.1 (ESRI, Redlands, CA, USA) to calculate percent cover by vegetation, land-use class, and impermeable surfaces within 500 m from the channel

(Zelt and Johnson 2005). All study reaches were subsequently ground-truthed for accuracy. Further surveys were completed at each site to document bridge crossings, vegetation structure, and vegetation composition.

SEDIMENT AND VEGETATION

Following Walters et al. (2010), we collected five sediment cores (~10-cm in depth) from each reach, representing approximately equidistant locations spanning from the top to the bottom of the reach. We collected sediment cores by push-coring approximately 2 cm with a polycarbonate coring device. We sealed sediment composite samples in plastic sleeves, placed them on ice, and froze them at -20 °C until analysis

(Lutz 2008).

We surveyed riparian vegetation along three equidistant 100-m transects along each bank (perpendicular to the stream, n = 15/reach) at each reach following methods

38 outlined in Bartel et al. (2010) and Pennington and Blair (2011). Using 5 x 5 m quadrats, we categorized vegetation structure by height (i.e., < 3 m, 3-5 m, and > 5 m). We visually estimated the % coverage of overhanging vegetation along both banks, and counted standing dead trees with diameter at breast height (DBH) of greater than 10 cm.

We identified all samples using Braun (1989) as a guide.

INVERTEBRATES

We deployed six floating, pyramid-style emergent insect traps at each reach

(Aubin et al. 1973). We distributed traps in close proximity to both banks along the length of the reach, representing available flow habitats as best possible. After a 10-day collection period, we collected all aquatic emergent insects, placed them on ice for transport, and froze them at -20° C until analysis (Walters et al. 2010). In the laboratory, we identified all individuals to family using Johnson and Triplehorn (2004) as a guide.

We sampled F. subsericea using quadrats (3 m2) placed at three intervals (stream bank, 1st riparian terrace, 2nd terrace) along the same five transects at which we sampled vegetation. Quadrat sampling has been shown to yield the most reliable estimate of density and diversity of ground-dwelling arthropods (Andersen 1995). F. subsericea were counted and collected in each quadrat for a period of ten minutes. We surveyed web-weaving riparian spiders of the family Tetragnathidae along both banks of each study reach during July when spiders were at peak abundance (Williams et al. 1995). We estimated spider density by counting the number of orb webs visible within five representative 30-m bank segments, allotting fifteen minutes per 30-m segment. We

39 collected spiders (> 15 per reach) by walking along the transects at night (22:00-00:00) when spiders tend to be most active and visible. We sealed all ants and spiders in plastic sleeves, placed them on ice for transport, and froze them at -20°C until analysis (Walters et al. 2010). In the laboratory, we identified ants to species using Cover and Fisher

(2007) as a guide. We dried and weighed aquatic emergent invertebrates (by family), F. subsericea, and Tetragnathidae (total dry mass per reach [mg · m-2 d-1]; Iwata, Nakano and Murakami 2003).

CONTAMINANT AND STABLE ISOTOPE ANALYSIS

We composited sediment samples by reach. We also combined arthropod samples into one composite per taxon per reach before homogenization by mortar and pestle. Subsequently, composite samples were analyzed for lead (Pb), arsenic (As), mercury, and selenium (Se) at the Diagnostic Center for Population and Animal Health at

Michigan State University.

Aquatic emergent insects, tetragnathid spiders, F. subsericea, and a suite of terrestrial insects (e.g., Cicadellidae, Dolichopodidae) collected as part of a companion study at the same study reaches were analyzed for 13C and 15N stable isotopes. We combined samples into one composite per family per reach, and dried them at 60°C for

48 hours before homogenization and packing in tin capsules. All samples were analyzed for C and N using continuous flow isotope ratio mass spectrometry (EA- IRMS) at the

University of Washington Stable Isotope Core (Pullman, WA, USA). The isotopic composition of samples was expressed using the 13C notation defined in Equation 1:

13 13 12  C = [(Rsample / Rstandard) – 1] * 1000, where R is C/ C [eqn 1], where parts per thousand (‰) are the atomic ratios of the number of atoms in the sample or standard.

Typical analytical precision was 0.2‰ for 13C determination.

NUMERICAL AND STATISTICAL ANALYSIS

For each reach, we calculated the overall density and biomass of the dominant aquatic emergent insects (i.e., Hydropsychidae and Chironomidae) using the mean of the values from the six emergence traps from each study reach (m2). We used these estimates to calculate contaminant export via these aquatic emergent insect families.

Estimates were initially derived for the 10-day sampling period and then extrapolated over a 90-day period to approximate export during a summer season (i.e., June, July,

August). Equation 2 represents the estimated total export: Export = (Emergent Biomass mg∙ m-2 x 10 days) x (Contaminant Concentration ppb) x (Area of Reach m-2) x 9 [eqn

2]. Similar equations have been used by others to estimate export (e.g., Walters et al.

2008). We used the mean density of the sampling area and extrapolated over the total length of the reach to calculate ant density. For spiders, we used the mean density of the sampling area and extrapolated over an area encompassing a zone spanning 5 m from the water’s edge along the length of the reach. Reach biomass estimates were calculated by multiplying the density of each consumer by the mean biomass of individuals over all reaches.

We used isotopic mixing models with one isotope (δ13C) and two sources (aquatic and terrestrial) (IsoError version 1.04 software, Phillips and Gregg 2001) to estimate the

41 feasible contributions of potential aquatic and terrestrial food sources to (1)

Tetragnathidae and (2) F. subsericea at each study reach. The model considers isotopic variances of both prey items and consumers, resulting in more reliable estimates of aquatic or terrestrial dietary proportions and their variances (Phillips and Gregg 2001).

Because of the array of prey available to our riparian consumers, we constrained the models using composite samples of prey individuals of the aquatic families

Chironomidae, Hydropsychidae, Heptageniidae and terrestrial families Dolichopodidae,

Cicadellidae, and Curculionidae. These were the numerically dominant families captured in our sampling periods and best captured the range of variability in δ13C values for our insect samples. Further, these families were often observed in the webs of tetragnathid spiders at our study reaches.

We used spider and ant density and biomass values for each reach, along with contaminant concentrations and contribution of aquatic resources to diet to estimate total realized contaminant flux (i.e., contaminant load assimilated into consumer tissue) to tetragnathid spiders and F. subsericea. Equation 3 represents the transport of aquatic contaminants uptaken by terrestrial consumers: Flux = (Consumer Biomass mg · m-2) x

(Contaminant Concentration ppb) x (% Aquatic Contribution) x (Area of Reach m-2) [eqn

3].

We performed principal component analysis (PCA) on riparian LULC variables that we had selected a priori as candidate predictors of riparian arthropod density. PCA axes with eigenvalues > 1 were retained for use in subsequent linear regression models

(Rencher 1995; Sullivan and Watzin 2008). We then used simple and mixed stepwise

42 linear regression models to select the best predictors for spider and ant density, contribution of aquatic prey to consumer diet, and contaminant flux (i.e., contaminant load assimilated into consumer tissue). Given multiple regression tests, the Bonferroni adjustment for α was α/k = 0.05/3 = 0.0167, where k is the number of tests (Wright 1992).

We used analysis of variance (ANOVA) to test for potential differences in density, contribution of aquatic resources, and contaminant flux of/to spiders and ants among land-use categories. Where significant main effects were detected, we conducted mean comparison tests using Tukey-Kramer HSD.

Results

Contaminant concentration (ppb) in sediment was highly variable for Se ( ̅ = 466,

SD = 143), Pb ( ̅ = 14650, SD = 7338) and Hg ( ̅ = 42.8, SD = 19.2), whereas concentrations of As were below detectable limits (i.e., 12.5 ppm). Biomass, density, and contaminant concentrations of aquatic emergent insects varied widely among study reaches (Table 2.1). Se concentrations in Hydropsychidae and Chironomidae exhibited a strong spatial pattern associated with an area of extensive industrial and wastewater point sources at the southern end of Columbus (Figure 2.2a). Hg exhibited a similar pattern for

Hydropsychidae and Chironomidae (Figure 2.2b). Export (the magnitude of contaminants leaving the stream via aquatic emergent insects) varied greatly among contaminants, but was highest for As, Se, Pb and Hg at mixed reaches (Table 2.1).

Spider density ranged from 0.32 to 0.76 individuals · m-2 among reaches ( ̅ = 0.50, SD =

0.14), whereas ant density ranged from 2.2 to 6.7 individuals · m-2 ( ̅ = 3.83, SD =

1.46).

TROPHIC LINKAGES

Two source mixing model results indicated that spiders relied heavily on aquatic emergent insect prey, ranging from 64.5 to 87% across the study reaches (Table 2.2).

Ant reliance on aquatic insect prey was much less substantial and displayed greater variability, ranging from 0.3% to 87%, with both minimum and maximum values occurring within reaches of predominantly mixed LULC (Table 2.2). Concentrations of

As, Pb, Hg, and Se in spider tissue were strongly related to distance from aforementioned point source (Figure 2.3a and 2.3b), supporting mixing model results.

ADJACENT LAND-USE and LAND-COVER INFLUENCES

Density and Diet

Principal component analysis of the eleven LULC measures identified three axes with eigenvalues > 1 (Table 2.3). PC1 explained about 54% of the variance. Because this axis was predominantly driven by factors related to urban landscapes (e.g., population density [r2 = 0.91, + correlation], % impervious surface coverage [r2 = 0.86,

+] % agriculture [r2 = 0.86, -), we labeled this PC ‘Urbanization Axis’. The second principal component (hereafter ‘Mature Trees Axis’) was driven almost solely by % tall tree cover (r2 = 0.86) and captured about 19% of the remaining variance. ‘Shoreline

Habitat’, the third PC, explained about 14% of the variance and was highly influenced by

44 standing dead trees (r2 = 0.69) and % overhanging vegetation (r2 = 0.48); all other adjacent LULC metrics had r2 < 0.22.

Urbanization (partial R2 = 0.33) and Shoreline Habitat (partial R2 = 0.33) were significant predictors of tetragnathid spider density (F = 7.75, P = 0.013). We also observed a strong, positive relationship between Urbanization and F. subsericea density

(Figure 2.4). F. subsericea density was significantly higher in urban reaches than in agricultural or mixed (F = 8.81, P = 0.010, N = 9; Tukey’s HSD, P < 0.05; Figure 2.4 inset), whereas no relationships were evident between tetragnathid density and land-use classes.

Shoreline Habitat was only weakly related to the contribution of aquatic prey to tetragnathid spider diet but was strongly related to the contribution of aquatic prey to ants

(Figure 2.5). We found no significant relationships between Urbanization or Mature

Trees and either spider or ant density or diet. We also found no significant differences among aquatic prey contribution to either ants or spiders among land-use classes.

Aquatic-to-terrestrial contaminant flux

Flux (i.e., aquatic contaminants uptaken by riparian spiders and ants) estimates are presented in Table 2.4. We found a positive relationship between Shoreline Habitat and contaminant flux to tetragnathid spiders for both As (Figure 2.6a) and Se (Figure

2.6b), a negative relationship between Mature Trees and Pb flux (Figure 2.6c), and no relationship between any PC axis and Hg. While Pb export was not significantly different between land use classes(Figure 2.7a), Pb flux to spiders was significantly

45 higher at urban and agricultural reaches than at mixed reaches (Figure 2.7b).

Urbanization and Shoreline Habitat were significant predictors of contaminant flux to ants for As and Se, although urbanization explained the greatest percent of the variation in both cases (0.51 and 0.50, respectively; Table 2.5). Pb flux was strongly associated with Urbanization (Figure 2.8). Pb flux to ants was also significantly greater at urban reaches than at mixed, and slightly greater than at agricultural reaches (Figure 2.7c).

Although we observed that Pb flux was highest in predominantly urban reaches for ants, and urban and agricultural reaches for spiders, we observed a divergent pattern in relation to aquatic emergent insect Pb export and land-use classes (Figure 2.7).

Discussion

Aquatic emergent insects are an important energy source for terrestrial consumers

(Baxter et al. 2005; Burdon and Harding 2008; Dreyer et al. 2012) and thus serve as critical aquatic-to-terrestrial exposure pathways (Custer et al. 2003; Walters et al. 2008a;

Walters et al. 2008b). However, the mechanisms of trophic contaminant transfer between aquatic insects and terrestrial invertebrate predators are still not fully understood. Our results indicate that adjacent LULC influences the density of riparian arthropod consumers as well as the extent to which they rely on aquatic resources, and in turn, the magnitude of aquatic-to-terrestrial contaminant flux. Given the ubiquity, productivity, and proclivity of aquatic insects to accumulate contaminants (Menzie 1980; Kay et al.

2005; Daley et al. 2011), understanding the factors that regulate these cross-boundary

46 contaminant fluxes will be critical for risk management related to the contamination of aquatic ecosystems.

RIPARIAN CONSUMER DENSITY

Land cover and land use have been shown to be key factors that affect many aspects of watershed structure and function (Allan et al. 1997; Cuffney et al. 2000). Land use within a watershed can interact with local riparian conditions to affect aquatic ecosystem condition, including changes in water quality (Meador and Goldstein 2003) and the metabolism of lakes and streams through the loading of sediments, nutrients, and contaminants (Debenest et al. 2010). Changes in riparian land cover can also significantly change community structure and function (Polis et al. 2004). Riparian vegetation, in particular, has been shown to strongly influence a suite of riparian arthropod consumers. Laeser, et al. (2005), for example, found that the density of

Tetragnathidae decreased by 72% along natural streams where vegetation had been removed, while other web-building families (Theridiidae, Linyphiidae and Araneidae) decreased in density by 87%. Ives et al. (2011a) found that streamside vegetation community type was a significant predictor of riparian ant assemblages. Urbanization has been correlated with decreases in both species richness and density of ground beetles

(Carabidae, Coleoptera) in eight cities in the boreal and temperate zones of North

America, Europe and Asia (reviewed by Niemela & Kotze (2009). Kavazos and

Wallman (2012) observed that increasing urbanization was related to invertebrate communities characterized by a greater relative abundance of non-native species.

In our study, tetragnathid spider density was most closely related to a combination of land-cover characteristics associated with both urbanization and shoreline habitat (% impervious surface coverage, population density, % overhanging vegetation, nearshore standing dead trees). For passive predators, habitat selection is critical, and web placement is thought to be non-random (Riechart and Gillespie1986). In a stream in

China, Chan et al. (2009) found that between 50 and 80% of all riparian spider orb-webs were less than 0.5 m above the water surface, and 45-51% of them were supported by either overhanging vegetation or boulders. These same investigators found that both the abundance and density of orb-webs increased significantly with the addition of artificial substrate along the stream. Williams et al. (1995) found that Tetragnathidae density varied widely between opposite banks of the same stream reach, and implicated greater coverage of overhanging vegetation for web-building habitat. Urban reaches in our study system are characterized by a high density of non-native Lonicera maackii (Amur honeysuckle), an invasive woody shrub in the region (Ohio DNR 2012). Dense patches of this plant overhang the littoral zones of the Scioto River, providing abundant web- building habitat for tetragnathid spiders.

Ant density, F. subsericea were also found in highest densities at more urbanized reaches (Figure 2.5). Krombein (1979) determined that F. subsericea exhibit a preference for deciduous forest habitat, while Ouellette et al. (2010) found that this species was often found near the shoreline, and recorded most ants in association with rocks and logs that provide nest-building structures. Leal et al. (2012) noted that generalist ants, like F. subsericea, are often less affected by environment disturbance

48 than more specialized species and Ives et al. (2011b) found that generalist species of ant were more abundant in highly fragmented habitats (e.g., urban environments).

AQUATIC RESOURCE UTILIZATION

Consistent with previous studies (Akamatsu, et al. 2004; Chan et al. 2009), we observed that tetragnathid spiders predominantly consumed aquatic emergent insects

(Table 2.2). For example, Akamatsu et al. (2004) estimated that the contribution of aquatic insects to riparian web-building spiders was 84% in the Chikuma River, Japan.

In our study, the mean aquatic proportion of tetragnathid spider diet was nearly 77%. In contrast, the contribution of aquatic emergent insects to F. subsericea was approximately

28%, with substantial variability (0.3 - 84.6%). F. subsericea are known generalist predators (Cover and Fisher 2007), and they likely utilize a wide variety of food resources as they become available. Paetzold et al. (2006), for instance, observed an increase in the abundance of riparian ant species Manica rubida when stream-derived invertebrate prey was experimentally added to the stream. Though stable isotope analysis indicated that aquatic insects were not a major dietary component of this species, they did respond strongly to an increase in readily available prey items, similar to what might occur during a flood pulse or a large emergence event. Given plasticity in ant diet, the strength of trophic linkages – and thus contaminant transfer – between aquatic insects and

F. subsericea appears to be less predictable than for tetragnathid spiders.

We also observed that shoreline habitat, principally the % vegetation over the water and the number of standing dead trees (mainly distributed along the shoreline),

49 explained variation in both ant and spider (although weakly) reliance on aquatic insects

(Figure 2.5). Williams et al. (1995) found that tetragnathid spiders caught the greatest amount of aquatic prey in webs placed 2-4 m from the stream at a height of up to 2 m, and in shrubs and tree branches that overhung the water at a height of up to 2 m. Briers et al. (2005) found that over 40% of lycosid spider diet was made up of aquatic insects adjacent to the stream, but was less than 1% at 20 m from the stream. In Japan,

Akamatsu et al. (2004) observed that the proportion of aquatic diet in web-building spiders, including Tetragnathidae, was higher in those with webs nearest the river than those residing further inland. Consistent with our results, these findings suggest that tetragnathid spiders with webs constructed in overhanging vegetation and dead trees at the aquatic-terrestrial interface will capture more aquatic emergent insects than those with webs further into the riparian zone.

F. subsericea, on the other hand, are opportunistic feeders, although they might be expected to selectively feed on nitrogen-rich resources like invertebrates (Kay 2002; Bihn et al. 2008; Ness et al. 2009; Pearce-Duvet and Feener 2010). Pearce-Duvet and Feener

(2010) suggested that the ability of ants to detect resources is an important component of ant foraging ecology, especially for those ants inhabiting areas with a varied and unpredictable resource base such as may be the case in riparian zones. Therefore, greater proximity to and detection frequency of high-quality aquatic food resources made possible by increased habitat availability either near or overhanging the water might be expected to be beneficial to F. subsericea.

AQUATIC-TO-TERRESTRIAL CONTAMINANT FLUX

Whereas multiple factors have been shown to be critical to the accumulation of contaminants in aquatic sediments (Stone et al. 2007; Iannuzzi et al. 2011), our results indicate that that adjacent LULC characteristics may strongly mediate the

“reterrestrialization” of these transfers through influences on consumer density and reliance on aquatic prey items. We observed that these mediating influences are evident at both fine and coarse scales of resolution. For example, Shoreline Habitat explained

>50% of the variation observed in tetragnathid flux estimates for As and Se (Figure 2.6), relating to fine-scale influences of near-shore structural characteristics on spider densities and aquatic prey consumption. However, patterns at a coarser resolution (i.e., by land- use class) were also evident (Figures 2.7b and 2.7c).

The discrepancy between contaminant export via aquatic emergent insects and contaminant flux to riparian ants and spiders provides a synthesis of the evidence supporting the mediating role of adjacent LULC. Export estimates from aquatic emergent insects were greatest for all contaminants at mixed reaches (Table 2.1).

However, Pb flux to Tetragnathidae was greater at both urban and agricultural reaches than at mixed reaches (Figure 2.7b). Similarly, Pb flux to F. subsericea was greater at urban than at mixed reaches (Figure 2.7c). The lack of alignment between export and flux estimates by land-use class highlights the important regulatory role of LULC dynamics on the magnitude of aquatic-to-terrestrial contaminant transfer. The implications of these results become more evident given that our study was conducted in a mixed-use river system with contaminant levels representative of many others river

51 systems found in developed regions (Blocksom and Johnson 2009; Haring et al. 2011).

The mediating effects of adjacent LULC might be expected to be even more pronounced at highly-contaminated locations, such as Superfund sites investigated in other studies related to aquatic-to-terrestrial contaminant transport (e.g. Walters et al. 2008a; Rashleigh et al. 2009; Walters et al. 2010). For example, we estimated that the mean Hg export per

1-km reach was 3.01 mg · 90 days -1, while Runck (2007) found that 4.01 g · yr -1 of Hg was exported over a 2.1-km reach from a highly contaminated stream in Tennessee, indicating a much greater level of Hg available in the system. Additionally, our estimates were based on 1-km river reaches, thereby representing an infinitesimal estimate of total flux coming from the entire river system.

Although we have identified LULC variables that can mediate aquatic-to- terrestrial contaminant flux, there are important caveats to our estimates. It is possible that physiological uptake of contaminants may differ between arthropod consumers, as

Nath et al. (2011) found when comparing heavy metal bioaccumulation in four acridid grasshopper species. Also, we relied on web counts in order to estimate tetragnathid spider density, though adult males generally feed by stealing food from other webs and do not construct their own (Gillespie & Caraco 1987; Williams 1995). As a result, our density estimates, and therefore our contaminant flux estimates, might be overly conservative. However, we are confident that any discrepancy is likely to remain constant across all reaches, and would not compromise our estimates of the relative magnitude of aquatic-to-terrestrial contaminant flux.

CONCLUSION

Overall, our results (1) support recent research examining the role of emergent aquatic insects as vectors of contaminant transport to terrestrial foodwebs (Currie et al.

1997; Walters et al. 2008b; Daley et al. 2011) and (2) provide new insights into the influence of adjacent landscape characteristics on the magnitude of these contaminant fluxes. Specifically, urbanization and nearshore habitat appear to be of particular importance in regulating both the density and diet of riparian consumers, and in turn, aquatic-to-terrestrial contaminant flux estimates. Building on work by others (e.g.,

Walters et al. 2008, Raikow et al. 2011) that have posited that aquatic insect emergence is a limiting step in aquatic-to-terrestrial contaminant transfer, we argue that recipient community dynamics (i.e., terrestrial consumers of aquatic emergent insects) represent a critical mechanism regulating the magnitude and distribution of contaminant assimilation across the landscape.

As management agencies work to mediate risk of exposure to contaminated aquatic sediments, the findings within this study provide important insights for decision makers. Given the pervasive nature of aquatic contamination, aquatic-to-terrestrial contaminant flux is likely to have far-reaching consequences impacting both wildlife and humans (Cristol et al., 2008, Sullivan and Rodewald 2012). Improved understanding of mediating influences on the flux of contaminants from aquatic to terrestrial systems will be critical for effective risk management and regulation. The proliferation of invasive honeysuckle, a facilitator of stronger trophic linkages between the aquatic and terrestrial systems, for example, is a factor that could strongly influence contaminant transfer. In

53 particular, we advocate for additional research that quantifies the magnitude of exposure to terrestrial food webs, including investigations relative to the drivers of riparian consumer population dynamics and foraging behavior in the face of environmental disturbance.

Acknowledgements

Research support was provided by state and federal funds appropriated to The Ohio State

University, Ohio Agricultural Research and Development Center. We would like to thank Adam Kautza and Paradzayi Tagwireyi for their help in the field and laboratory.

References

Akamatsu, F., H. Toda & T. Okino (2004) Food source of riparian spiders analyzed by using stable isotope ratios. Ecological Research, 19, 655-662.

Allan, J. D., D. L. Erickson & J. Fay (1997) The influence of catchment land use on stream integrity across multiple spatial scales. Freshwater Biology, 37, 149-161.

Andersen, J. (1995) A comparison of pitfall trapping and quadrat sampling of Carabidae (Coleoptera) on river banks. Entomologica Fennica, 6, 65-77.

Aubin, A., J. P. Bourassa & Pellissi.M (1973) Effective emergence trap for capture of mosquitos. Mosquito News, 33, 251-252.

Bartel, R. A., N. M. Haddad & J. P. Wright (2010) Ecosystem engineers maintain a rare species of butterfly and increase plant diversity. Oikos, 119, 883-890.

Baxter, C. V., K. D. Fausch & W. C. Saunders (2005) Tangled webs: reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biology, 50, 201- 220.

Bihn, J. H., M. Verhaagh & R. Brand (2008) Ecological stoichiometry along a gradient of forest succession: Bait preferences of litter ants. Biotropica, 40, 597-599.

Blais, J. M. (2005) Biogeochemistry of persistent bioaccumulative toxicants: processes affecting the transport of contaminants to remote areas. Canadian Journal of Fisheries and Aquatic Sciences, 62, 236-243.

Blais, J. M., R. W. Macdonald, D. Mackey, E. Webster, C. Harvey & J. P. Smol (2007) Biologically mediated transport of contaminants to aquatic systems. Environmental Science & Technology, 41, 1075-1084.

Blocksom K.A. & Johnson B.R. (2009) Development of a regional macroinvertebrate index for large river bioassessment.Ecological Indicators, 9, 313-328.

Blocksom, K. A., D. M. Walters, T. M. Jicha, J. M. Lazorchak, T. R. Angradi & D. W. Bolgrien (2010) Persistent organic pollutants in fish tissue in the mid-continental great rivers of the United States. Science of the Total Environment, 408, 1180- 1189. 55

Braun, E. L. 1989. Woody Plants of Ohio: Trees, Shrubs, and Woody Climbers, Native, Naturalized, and Escaped. Columbus, OH: The Ohio State University Press.

Briers, R. A., H. M. Cariss, R. Geoghegan & J. H. R. Gee (2005) The lateral extent of the subsidy from an upland stream to riparian lycosid spiders. Ecography, 28, 165- 170.

Burdon, F. J. & J. S. Harding (2008) The linkage between riparian predators and aquatic insects across a stream-resource spectrum. Freshwater Biology, 53, 330-346.

Chan, E. K. W., Y. Zhang & D. Dudgeon (2009) Substrate Availability May Be More Important than Aquatic Insect Abundance in the Distribution of Riparian Orb-web Spiders in the Tropics. Biotropica, 41, 196-201.

Christensen, J. R., M. Macduffee, R. W. Macdonald, M. Whiticar & P. S. Ross (2005) Persistent organic pollutants in British Columbia grizzly bears: Consequence of divergent diets. Environmental Science & Technology, 39, 6952-6960.

Clements, W. H. & M. C. Newman. 2002. Community Ecotoxicology. Wiley.

Cover, S. P. & B. L. Fisher. 2007. Ants of North America. University of California Press.

Cristol, D. A., R. L. Brasso, A. M. Condon, R. E. Fovargue, S. L. Friedman, K. K. Hallinger, A. P. Monroe & A. E. White (2008) The movement of aquatic Mercury through terrestrial food webs. Science, 320, 335-335.

Cuffney, T. F., M. R. Meador, S. D. Porter & M. E. Gurtz (2000) Responses of physical, chemical, and biological indicators of water quality to a gradient of agricultural land use in the Yakima River Basin,Washington. Environmental Monitoring and Assessment, 64, 259-270.

Currie, R. S., W. L. Fairchild & D. C. G. Muir (1997) Remobilization and export of cadmium from lake sediments by emerging insects. Environmental Toxicology and Chemistry, 16, 2333-2338.

Custer, C.M., Custer, T.W., Dummer, P.M., & K.L. Munney (2003) Exposure and effects of chemical contaminants on tree swallows nesting along the Housatonic River, Berkshire County, Massachusetts, USA, 1998-2000. Environmental Toxicology andChemistry, 22, 1605-1621.

Daley, J. M., L. D. Corkum & K. G. Drouillard (2011) Aquatic to terrestrial transfer of sediment associated persistent organic pollutants is enhanced by bioamplification processes. Environmental Toxicology and Chemistry, 30, 2167-2174.

Debenest, T., J. Silvestre, M. Coste & E. Pinelli. 2010. Effects of Pesticides on Freshwater Diatoms. In Reviews of Environmental Contamination and Toxicology, Vol 203, ed. D. M. Whitcare, 87-103. New York: Springer.

Dreyer, J., D. Hoekman & C. Gratton (2012) Lake-derived midges increase abundance of shoreline terrestrial arthropods via multiple trophic pathways. Oikos, 121, 252- 258.

Evers, D. C., N. M. Burgess, L. Champoux, B. Hoskins, A. Major, W. M. Goodale, R. J. Taylor, R. Poppenga & T. Daigle (2005) Patterns and interpretation of mercury exposure in freshwater avian communities in northeastern North America. Ecotoxicology, 14, 193-221.

Fitzgerald, W. F., D. R. Engstrom, R. P. Mason & E. A. Nater (1998) The case for atmospheric mercury contamination in remote areas. Environmental Science & Technology, 32, 1-7.

Gillespie, R.G. & T. Caraco (1987) Risk-sensitive foraging strategies of two spider populations. Ecology, 68, 887-899.

Haring, J.H., Blocksom, K.A., Smith, M.E., Agnradi, T., Wratschko, M.C., Armstrong, B., Bolgrien, D., & J.M. Lazorchak (2011) Sediment toxicity in mid-continent great rivers (USA). Archives of Environmental Contamination and Toxicology, 60, 57-67.

Hinck, J. E., C. J. Schmitt, K. A. Chojnacki & D. E. Tillitt (2009) Environmental contaminants in freshwater fish and their risk to piscivorous wildlife based on a national monitoring program. Environmental Monitoring and Assessment, 152, 469-494.

Iannuzzi, J., Butcher, M., & T Iannuzi (2011) Evaluation of potential relationships between chemical contaminants and aquatic organisms from the lower Passaic River, New Jersey, USA. Environmental Toxicology and Chemistry, 30, 1721- 1728.

Ives, C.D., Hose, G.C., Nipperess, D.A., & M.P. Taylor (2011a) Environmental and landscape factors influencing ant and plant diversity in suburban riparian corridors. Landscape and Urban Planning, 103, 372-382.

Ives, C. D., G. C. Hose, D. A. Nipperess & M. P. Taylor (2011b) The influence of riparian corridor width on ant and plant assemblages in northern Sydney, Australia. Urban Ecosystems, 14, 1-16.

Iwata, T., S. Nakano & M. Murakami (2003) Stream meanders increase insectivorous bird abundance in riparian deciduous forests. Ecography, 26, 325-337. 57

Jackson , J.K. & S.G. Fisher (1986) Secondary production, emergence, and export of aquatic insects in a Sonoran Desert stream. Ecology, 67, 629-638.

Johnson, N. F. & C. A. Triplehorn. 2004. Borror and Delong's Introduction to the Study of Insects. Thomson Brooks/Cole.

Kavazos, C. R. J. & J. F. Wallman (2012) Community composition of carrion-breeding blowflies (Diptera: Calliphoridae) along an urban gradient in south-eastern Australia. Landscape and Urban Planning, 106, 183-190.

Kay, A. (2002) Applying optimal foraging theory to assess nutrient availability ratios for ants. Ecology, 83, 1935-1944.

Kay, D. P., A. L. Blankenship, K. K. Coady, A. M. Neigh, M. J. Zwiernik, S. D. Millsap, K. Strause, C. Park, P. Bradley, J. L. Newsted, P. D. Jones & J. P. Giesy (2005) Differential accumulation of polychlorinated biphenyl congeners in the aquatic food web at the Kalamazoo River superfund site, Michigan. Environmental Science & Technology, 39, 5964-5974.

Krombein K.V., H. P. D., Smith D.R., and B.D. Burks. 1979. Catalog of Hymenoptera in America North of Mexico. Washington D.C.: Smithsonian Institution Press.

Laeser, S. R., C. V. Baxter & K. D. Fausch (2005) Riparian vegetation loss, stream channelization, and web-weaving spiders in northern Japan. Ecological Research, 20, 646-651.

Leal, I. R., B. K. C. Filgueiras, J. P. Gomes, L. Iannuzzi & A. N. Andersen (2012) Effects of habitat fragmentation on ant richness and functional composition in Brazilian Atlantic forest. Biodiversity and Conservation, 21, 1687-1701.

Lutz M.A., B. M. E. M.-D. M. 2008. Procedures for collecting and processing streambed sediment and pore water for analysis of mercury as part of the National Water- Quality Assessment Program. Open File Report 2008-1279.

Meador, M. R. & R. M. Goldstein (2003) Assessing water quality at large geographic scales: relations among land use, water physicochemistry, riparian condition, and fish community structure. Environmental Management, 31, 504-517.

Menzie, C. A. (1980) Potential significance of insects in the removal of contaminants from aquatic systems. Water Air and Soil Pollution, 13, 473-479.

Merna, J. W. (1986) Contaminatino of stream fishes with chlorinated hydrocarbons from eggs of Great Lakes salmon. Transactions of the American Fisheries Society, 115, 69-74.

Nath, S., H. Anand & P. Haldar (2011) Study of short horned grasshoppers in relation to heavy metals accumulation. Journal of Entomological Research (New Delhi), 35, 291-293.

Ness, J. H., W. F. Morris & J. L. Bronstein (2009) For ant-protected plants, the best defense is a hungry offense. Ecology, 90, 2823-2831.

Niemela, J. & D. J. Kotze (2009) Carabid beetle assemblages along urban to rural gradients: A review. Landscape and Urban Planning, 92, 65-71.

Ohio Department of Natural Resources, Division of Forestry (2012) http://www.dnr.state.oh.us/Home/trees/honey_am/tabid/5376/Default.aspx

Ohio Environmental Protection Agency (2011) http://www.epa.ohio.gov/dsw/fishadvisory/limitmeals.aspx. Accessed February 11, 2012.

Ouellette, G. D., F. A. Drummond, B. Choate & E. Groden (2010) Ant Diversity and Distribution in Acadia National Park, Maine. Environmental Entomology, 39, 1447-1456.

Paetzold, A., J. F. Bernet & K. Tockner (2006) Consumer-specific responses to riverine subsidy pulses in a riparian arthropod assemblage. Freshwater Biology, 51, 1103- 1115.

Paetzold, A., M. Smith, P. H. Warren & L. Maltby (2011) Environmental impact propagated by cross-system subsidy: Chronic stream pollution controls riparian spider populations. Ecology, 92, 1711-1716.

Pearce-Duvet, J. C. & D. H. Feener (2010) Resource discovery in ant communities: do food type and quantity matter? Ecological Entomology, 35, 549-556.

Pennington, D. N. & R. B. Blair (2011) Habitat selection of breeding riparian birds in an urban environment: untangling the relative importance of biophysical elements and spatial scale. Diversity and Distributions, 17, 506-518.

Phillips, D. L. & J. W. Gregg (2001) Uncertainty in source partitioning using stable isotopes. Oecologia, 127, 171-179.

Polis, G. A., M. E. Power & G. R. Huxel. 2004. Food webs at the landscape level. Chicago, Illinois: University of Chicago Press.

Rashleigh, B., M. C. Barber & D. M. Walters (2009) Foodweb modeling for polychlorinated biphenyls (PCBs) in the Twelvemile Creek Arm of Lake Hartwell, South Carolina, USA. Ecological Modelling, 220, 254-264.

Rencher, A. C. 1995. Methods of Multivariate Analysis. New York, NY, U.S.A.: Johny Wiley and Sons, Inc.

Riechart, S.E. & R.G. Gillespie. 1986. Habitat choice and utilisation in web-building spiders. In: Spiders: webs, behaviour, and evolution. Ed. A. Shear. Stanford, CA., Stanford University Press.

Runck, C. (2007) Macroinvertebrate production and food web enegetics in an industrially contaminated stream. Ecological Applications, 17, 740-753.

Sarica, J., M. Amyot, L. Hare, M. R. Doyon & L. W. Stanfield (2004) Salmon-derived mercury and nutrients in a Lake Ontario spawning stream. Limnology and Oceanography, 49, 891-899.

Stone, J. J., C. M. McCutcheon, L. D. Stetler & S. R. Chipps (2011) Interrelationships between Fish Tissue Mercury Concentrations and Water Quality for South Dakota Natural Lakes and Impoundments. Water Air and Soil Pollution, 222, 337-349.

Stone, W. W., C. G. Crawford & R. J. Gilliom (2007) AGRO 140-Estimating pesticide concentrations in US streams from watershed characteristics and pesticide properties. Abstracts of Papers of the American Chemical Society, 233, 219-219.

Sullivan, S. M. P. & A. D. Rodewald (2012) In a state of flux: The energetic pathways that move contaminants from aquatic to terrestrial environments. Environmental toxicology and chemistry / SETAC, 31, 1175-83.

Sullivan, S. M. P. & M. C. Watzin (2008) Relating stream physical habitat condition and concordance of biotic productivity across multiple taxa. Canadian Journal of Fisheries and Aquatic Sciences, 65, 2667-2677.

United States Census Bureau (2010). Interactive Population Map: Most Populous Places. http://2010.census.gov/2010census/popmap/. Accessed June 4, 2012.

United States Environmental Protection Agency (2004) Draft Aquatic Life Water Quality Criteria for Selenium. EPA-822-D-04-001.

Walters, D. M., K. M. Fritz, B. R. Johnson, J. M. Lazorchak & F. H. McCormico (2008a) Influence of trophic position and spatial location on polychlorinated biphenyl (PCB) bioaccumulation in a stream food web. Environmental Science & Technology, 42, 2316-2322.

Walters, D. M., K. M. Fritz & R. R. Otter (2008b) The dark side of subsidies: adult stream insects export organic contaminants to riparian predators.. Ecological Applications, 18, 1835-1841.

Walters, D. M., M. A. Mills, K. M. Fritz & D. F. Raikow (2010) Spider-Mediated Flux of PCBs from Contaminated Sediments to Terrestrial Ecosystems and Potential Risks to Arachnivorous Birds. Environmental Science & Technology, 44, 2849- 2856.

Ward, M. P., C. Jablonski, B. Semel & D. Soucek (2010) The biological pathway and effect of PCBs on common terns in Lake Michigan. Ecotoxicology, 19, 1513- 1522.

White D., Johnston K. & Miller M. (2005) Ohio River basin. In: Rivers of North America (Eds A.C. Benke& C.E. Cushing), pp. 375-424. Elsevier Academic Press, Burlington, Massachusetts.

Williams, D. D., L. G. Ambrose & L. N. Browning (1995) Trophic dynamics of 2 sympatric species of riaprian spider (araneae, tetragnathidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie, 73, 1545-1553.

Wright, S.P. (1992) Adjusted p-values for simultaneous inference. Biometrics, 45, 1005- 1013.

Zelt, R.B., & M.R. Johnson (2005) Protocols for Mapping and Characterizing Land Use/Land Cover in Riparian Zones. U.S. Geological Survey Open-File Report 2005-1302.

Zhang, X. M., A. S. Naidu, J. J. Kelley, S. C. Jewett, D. Dasher & L. K. Duffy (2001) Baseline concentrations of total mercury and methylmercury in salmon returning via the Bering Sea (1999-2000). Marine Pollution Bulletin, 42, 993-997.

Table 2.13) Biomass and contaminant concentrations of aquatic emergent insects of the families Chironomidae and Hydropsychidae and contaminant export estimates from 9 study reaches in the Scioto River basin, Ohio, USA. Note: Hydropsychidae was only collected from one urban reach.

Taxonomic Group Contaminant Concentration mean (SD) Biomass mean (SD) (ppb) (mg · m-2) As Se Pb Hg Chironomidae All reaches 339 (340) 2997 (611) 623 (355) 72 (11) 70.9 (35.9) Agricultural 203 (151) 3576 (485) 775 (435) 88 (1.4) 211 (24.0) Mixed 240 (242) 2614 (388) 419 (249) 64 (7.0) 549 (156) Urban 722 (539) 3378 (672) 982 (192) 74 (2.8) 331 (296) Hydropsychidae All reaches 534 (325) 2637 (1329) 1365 (257) 67 (14) 12.2 (19.4) Agricultural 400 (113) 4722 (76) 1527 (25) 84 (5.7) 6.4 (0.7) Mixed 617 (397) 1870 (389) 1274 (296) 58 (8.9) 119 (133) Urban 380 2303 1496 76 4.3 (6.1)

Contaminant Export mean (SD) (mg · reach-1 · 90 days-1) As Se Pb Hg Emergent Insects 20.4 All reaches (13.0) 115.0 (34.1) 42.6 (20.7) 3.01 (1.29) 5.86 Agricultural (1.44) 84.8 (21.2) 22.9 (0.21) 1.76 (0.43) 26.1 Mixed (13.7) 131.0 (31.7) 50.6 (16.4) 3.64 (1.09) Urban 20.6 (2.8) 107.0 (39.4) 42.4 (35.5) 2.70 (1.74)

Table 2.2. Contribution of aquatic prey to riparian arthropod consumer diet for study reaches in the Scioto River basin, Ohio (calculated using IsoError v 1.04; Phillips and Greg 2001).

Land-use Class Consumer Tetragnathidae F. subsericea ̅ (SE), % ̅ (SE), % All Reaches 76.9 (8.9) 27.8 (25.1) Urban (n =2) 82.5 (6.4) 30.4 (16.1) Agriculture (n =2) 72.3 (10.0) 27.6 (17.7) Mixed (n =5) 76.4 (9.8) 26.8 (33.4)

Table 2.3. Adjacent (500 m on each side of river) land-use and land-cover (LULC) principal component analysis: eigenvalues and the percent variance captured by the principal components (eigenvalues > 1), along with each principal component’s loadings and the proportion of the variance ( ) each variable shared with the PCA axes. Bold type indicates those axes used as successful predictor variables in regression models.

Urbanization Mature Trees Shoreline Habitat Adjacent LULC Metrics Loading r 2 Loading r 2 Loading r 2 % Agriculture -0.38 0.86 0.08 0.01 0.02 0.00 % Shrub Cover (1-3m) 0.34 0.69 0.02 0.00 -0.38 0.22 % Small Tree Cover (3-6m) 0.30 0.54 0.31 0.20 0.05 0.00 % Tall Tree Cover (>6m) -0.01 0.00 0.64 0.86 0.17 0.04 # Standing Dead Trees 0.00 0.00 -0.34 0.24 0.67 0.69 Riparian Forest Width -0.30 0.54 -0.10 0.02 -0.08 0.01 # Bridge Crossings 0.30 0.54 -0.39 0.32 -0.17 0.04 % Impervious Surface 0.38 0.86 -0.23 0.11 -0.04 0.00 # Canopy Layers 0.32 0.61 0.36 0.27 -0.09 0.01 Population Density 0.39 0.91 -0.10 0.02 0.13 0.03 % Overhanging Vegetation 0.26 0.40 0.13 0.04 0.56 0.48

Eigenvalue 5.98 2.10 1.53 % variance 54.4 19.1 14.0

Contaminant Flux mean (SD) (mg · reach-1) Tetragnathidae F. subsericea As Se Pb Hg As Se Pb Hg All reaches 0.009 ( 0.004) 0.086 (0.037) 0.004 (0.003) 0.004 (0.002) 0.004 (0.003) 0.004 (0.004) 0.004 (0.003) 0.000 (0.000) Agricultural 0.010 (0.002) 0.098 (0.007) 0.007 (0.001) 0.004 (0.001) 0.003 (0.001) 0.003 (0.001) 0.003 (0.000) 0.000 (0.000) Mixed 0.008 (0.005) 0.075 (0.045) 0.003 (0.001) 0.003 (0.002) 0.003 (0.003) 0.004 (0.002) 0.002 (0.002) 0.001 (0.001) Urban 0.012 (0.005) 0.102 (0.038) 0.006 (0.003) 0.005 (0.003) 0.006 (0.004) 0.007 (0.006) 0.008 (0.003) 0.000 (0.000)

Table 2.5. Explanatory variables and their coefficients in the significant regression models of contaminant flux to F. subsericea

Contaminant Variable Coefficient SE R 2 F statistic As (n = 9) Intercept 0.0037 0.0010 (P = 0.028) Urbanization 0.0006 0.0006 0.51 0.88 Shoreline Habitat 0.0017 0.0009 0.19 3.68

Se (n = 9) Intercept 0.0041 0.0012 (P = 0.036) Urbanization 0.0007 0.0007 0.50 0.93 Shoreline Habitat 0.0018 0.0011 0.17 2.99

Figure 2.1. Nine 1-km study reaches of the Scioto River basin, Ohio, USA.

Figure 2.2. Relationship between downstream distance from an area of extensive point source inputs into the Scioto River and (a) Se concentrations of Chironomidae (solid line) and Hydropsychidae (dashed line) and (b) Hg concentrations of Chironomidae (solid line) and Hydropsychidae (dashed line) across all study reaches. Slopes for the linear regression models were highly significant for Hydropsychidae Se (y = 4829 – 24.3x, R2 = 0.86, P < 0.001) and Hg (y = 87.5 - 0.23x, R2 = 0.66, P = 0.012), as well as Chironomidae Se (y = 3730 – 9.3x, R2 = 0.73, P = 0.007) and Hg (y = 88 – 0.18x, R2 = 0.59, P = 0.025). Note: Values from study reaches located upstream of the point source were excluded from regression analysis.

Figure 2.3. Relationship between downstream distance from an area of extensive point source inputs into the mainstem of the Scioto River and (a) As (diamond, solid line), Hg (circle, dashed line), and Pb (x, dotted line) concentrations of Tetragnathidae and (b) Se concentrations of Tetragnathidae across all study reaches. Slopes for the linear regression models were highly significant for As (y = 764 – 2.6x, R2 = 0.68, P = 0.011), Hg (y = 331 – 1.4x, R2 = 0.72, P = 0.007), Pb (y = 526 – 2.9x, R2 = 0.93, P < 0.001) and Se (y = 8.8 – 0.004x, R2 = 0.63, P = 0.019). Note: Values from study reaches located upstream of the point source were excluded from regression analysis.

Figure 2.4. Relationship between Urbanization Axis (i.e., PC1) and F. subsericea density. The slope for the linear regression model was highly significant (y = 3.8 + 0.54x, R2 = 0.83, P < 0.001). Inset: Comparison of F. subsericea density across reaches grouped by urban, agricultural, and mixed land uses (F = 8.81, P = 0.010, N = 9; Tukey’s HSD, P < 0.05). Different letters signify land uses for which F. subsericea density was significantly different. All values are shown as mean ± 1SE.

Figure 2.5. Relationship between Shoreline Habitat Axis (i.e., PC3) and contribution of aquatic prey (%) to Tetragnathidae (circle, dashed line) and F. subsericea (diamond, solid line). Slopes for the linear regression models were nearly significant for Tetragnathidae (y = 75.5 + 4.9x, R2 = 0.42, P = 0.060) and highly significant for F. subsericea (y = 22.6 + 18.0x, R2 = 0.70, P = 0.005).

Figure 2.6. Relationship between Shoreline Habitat Axis (i.e., PC3) and (a) As and (b) Se flux; (c) relationship between Mature Trees Axis (i.e., PC2) and Pb flux to Tetragnathidae. Slopes for the regression models were nearly significant for As (y = 0.008 + 0.003x, R2 = 0.51, P = 0.031), and significant for Se (y = 0.079 + 0.024x, R2 = 0.58, P = 0.016) and for Pb (y = 0.005 – 0.001x, R2 = 0.52, P = 0.012).

Figure 2.7. (a) Pb export by emergent aquatic insects across reaches grouped by urban, agricultural, and mixed land-use classes (F = 1.40, P = 0.316, N = 9; Tukey’s HSD, P < 0.05), (b) Pb flux to Tetragnathidae across reaches grouped by urban, agricultural, and mixed land-use classes (F = 6.10, P = 0.025, N = 9; Tukey’s HSD, P < 0.05), and (c) Pb flux to F. subsericea across reaches grouped by urban, agricultural, and mixed land-use classes (F = 8.68, P = 0.017, N = 9; Tukey’s HSD, P < 0.05). Different letters signify land-use types for which results were significantly different. All values are shown as mean ± 1SE.

Figure 2.8. Relationship between Urbanization Axis (i.e., PC1) and Pb flux to F. subsericea. The slope for the linear regression model was highly significant (y = 0.005 + 0.001x, R2 = 0.71, P = 0.004).

Chapter 3: Riparian swallows as integrators of landscape change in a

multiuse river system: implications for aquatic-to-terrestrial transfers of

contaminants

Jeremy M. Alberts1,2 and S. Mažeika P. Sullivan1

1School of Environment & Natural Resources, The Ohio State University, 2021 Coffey

Rd., Columbus, OH 43210

2Corresponding author: email: [email protected]; Fax: (01)614-292-7342

Abstract

Recent research has highlighted the transfer of contaminants from aquatic to terrestrial ecosystems via predation of aquatic emergent insects by riparian consumers.

The influence of adjacent land use and land cover (LULC) on aquatic-to-terrestrial contaminant transfer, however, has received little attention. To that end, at 11 river reaches in the Scioto River basin (OH, USA), we investigated the relationships between

LULC and selenium (Se) and mercury (Hg) concentrations in riparian swallows. Hg concentrations in swallows were significantly higher at rural reaches than urban (t = -

3.58, P < 0.001, df = 30), while Se concentrations were positively associated with adjacent land cover characterized by mature tree cover (R2 = 0.49, P = 0.006). To an extent, these relationships appear to be mediated by swallow reliance on aquatic emergent insects. For example, swallows at urban reaches exhibited a higher proportion of aquatic prey in their diet, fed at a higher trophic level, and exhibited elevated Se levels.

We also found that both Se and Hg concentrations in adult swallows were significantly higher than those observed in juveniles (Se: t = -2.83, P = 0.033, df = 3; Hg: t = -3.22, P

= 0.024, df = 3). Collectively, our results indicate that riparian swallows integrate contaminant exposure in linked aquatic-terrestrial systems and that LULC may strongly regulate aquatic contaminant flux to terrestrial consumers.

Introduction

Movement of materials across the landscape has been central to the study of food webs (Polis 1997; Polis et al. 2004). At the aquatic-terrestrial interface, ecological subsidies, or the flux of material from one habitat or ecosystem to another (Nakano and

Murakami 2001) has emerged as a central theme of cross-boundary linkages (Akamatsu et al. 2004; Vander Zanden and Sanzone 2004; Briers et al. 2005; Ballinger and Lake

2006). Whereas the contribution of terrestrial systems to aquatic food webs via inputs of nutrients and organic matter is well recognized, recent investigations have focused on the importance of aquatic prey subsidies to the energy budgets of riparian consumers

(reviewed in Baxter et al. 2005). A classic example is the reliance of grizzly bears (Ursus arctos horribilis) on fish as an important nutrient source (Mattson and Reinhart 1995;

Varley and Schullery 1996). Others have shown that aquatic prey represent important food sources to riparian bats (Fukui et al. 2006; Hagen and Sabo 2012), raccoons (Bigler et al. 1975; Lord et al. 2002), piscivorous birds (Vessel 1978; Hamas 1994), lizards (Sabo and Power 2002), and terrestrial arthropods (Cloe and Garman 1996; Henschel et al.

2001; Paetzold et al. 2005; Dreyer et al. 2012) .

In particular, aquatic insects that emerge from the stream as adults (hereafter,

‘aquatic emergent insects’) can be a critical food source for insectivorous riparian bird

78 species (Keast 1990; Iwata et al. 2003). For example, Sweeney and Vannote (1982) found that swarming activity by mayflies attracted nighthawks (Chordeiles minor) to areas of emergence, and Iwata et al. (2003) reported that flycatchers and gleaners foraged intensively on aquatic emergent insects in a pattern that mirrored aquatic insect distribution, suggesting that the dynamics of insect emergence can influence avian foraging behavior. Likewise, McCarty (1997) showed that the feeding ecology of tree swallows (Tachycineta bicolor) was also influenced by aquatic insect population dynamics.

Within the context of the energetic pathways that link aquatic and terrestrial ecosystems, there is growing interest in food-web studies at the land-water ecotone.

Many investigators have found that the high reliance of riparian bird species on aquatic prey resources can expose them to contaminants found in aquatic systems (see Sullivan and Rodewald 2012). Cristol et al. (2008) demonstrated that mercury (Hg) can bioaccumulate in terrestrial bird species living near contaminated water sources, while

Wada et al. (2009) found that adrenocortical responses and thyroid hormones were suppressed in tree swallows living near a river contaminated with mercury. In the

Chilliwack River of southern British Columbia, resident American Dippers (Cinclus mexicanus) (i.e., those birds who remain in the mainstem Chilliwack River year-round and have high proportions of contaminated salmon fry in their diet) were found to have higher polychlorinated biphenyl (PCB) and Hg concentrations than their migratory counterparts (i.e., those birds who are found on the Chilliwack mainstem only during the winter and who have a significantly smaller proportion of fish fry in their diet; Morrissey

79 et al. 2004). Several studies have described the bioaccumulation of PCBs in tree swallows from aquatic insect prey (Echols et al. 2004; Maul et al. 2006; Walters et al.

2008). Echols et al. (2004), for instance, found that PCB concentrations in tree swallow nestlings closely tracked concentrations in aquatic insects. Maul et al. (2006) implicated aquatic insects of the family Chironomidae as the key vector of PCBs to nestling tree swallows from an Illinois lake.

Differences in age-related contaminant exposure patterns among birds have also been documented. For example, Evers et al. (2005) found that concentrations of methylmercury (MeHg) in adult tree swallows were higher than in juveniles. Likewise,

Malinga et al. (2010) noted a positive correlation between age and biomagnification of contaminants in Glaucus gulls (Larus hyperboreus) in the Arctic.

Given the widespread contamination of aquatic systems (Fitzgerald et al. 1998;

Hammerschmidt and Fitzgerald 2005; Blocksom et al. 2010), the relative scarcity of information available on the aquatic-to-terrestrial contaminant flux does not currently reflect the potential scale of the problem. In particular, although the impacts of land-use and land-cover (LULC) on riparian birds has been widely investigated (Ormerod & Tyler

1991a; Sullivan et al. 2007; Luther et al. 2008; Pennington et al. 2008; Oneal &

Rotenberry 2009), the relationships between LULC and contaminant exposure to riparian birds from aquatic systems is poorly understood. For example, riparian swallow species

[e.g., bank (Riparia riparia), northern rough-winged (Stelgidopteryx serripennis), tree, and cliff (Petrochelidon pyrrhonota) swallows] are susceptible to bioaccumulation of contaminants because of their dependence on a mixture of stream, riparian, and terrestrial

80 food sources and habitat (Bishop et al. 1999; Harris and Elliot 2000; Dods et al. 2005;

Custer 2011). Further, riparian swallows tend to forage mainly within several hundred meters of their nesting reaches (Mengelkoch et al. 2005). Land-management practices that alter aquatic resource utilization by swallows might be expected to strongly influence their exposure to aquatic contaminants. Roux and Marra (2007), for example, found that

Pb concentrations in urban passerine birds were significantly higher than in their rural counterparts.

Our overarching objective was to explore the relationships between adjacent patterns in LULC and the magnitude of contaminant concentrations in riparian swallows as mediated by aquatic-to-terrestrial contaminant fluxes. Because of historic and contemporary differences in land-management and chemical use, we anticipated that concentrations of contaminants in river sediments, aquatic emergent insects, and riparian swallows would be different between rural and urban sites. Consequently, we anticipated that contaminant exposure to swallows would also differ by LULC class, but would be mediated by the degree of aquatic resource utilization such that swallows feeding at rural reaches (i.e., comprised of a mixture of forest, grassland, and agriculture) would exhibit greater reliance on aquatic emergent insects than swallows feeding at urban reaches.

Finally, we expected that adult swallows would exhibit greater biomagnification of contaminants than nestlings, and that female swallows would exhibit greater biomagnification of contaminants than males.

Materials and Methods

STUDY SYSTEM AND EXPERIMENTAL DESIGN

The Scioto River is a multiuse river that drains approximately 16,882 km2 in central and southern Ohio, flowing for 372 km through agricultural (69%), forested

(21%), and urban (9%) landscapes (White et al. 2005; Blocksom and Johnson 2009). The

Olentangy River meets the Scioto in Columbus, the 15th largest city in the US by population (1,178,899 people; USCB 2010). Once downstream of the city, the Scioto is free-flowing and exhibits riparian zones dominated by deciduous forest, row crops (corn and soybean), and smaller urban centers. The river mainstem is characterized by steep, muddy banks, and turbid waters (White et al. 2005; Blocksom and Johnson 2009).

We sampled adult riparian swallows at twelve river reaches (defined as a river segment based on similar valley features and channel geomorphology, (~1 km for this study) of the Scioto River watershed, representing urban and rural (i.e., mixture of forest, grassland, and agriculture) landscapes. Study reaches were distributed along the length of the river and were separated by an average distance of 16.6 river kilometers, although there was significant variability (SD = 16.5km). We used ArcGIS 10.1 (ESRI,

Redlands, CA, USA), to calculate percent cover by vegetation class within 500 m of the channel following Zelt and Johnson (2005), and performed ground-based surveys to document bridge crossings, vegetation structure, and vegetation composition.

FIELD AND LABORATORY METHODS

Periphyton and Vegetation

Following methods used in Bartel et al. (2010) and Pennington and Blair (2011), we surveyed riparian vegetation along three 100-m transects (upstream, middle, downstream) along both banks (perpendicular to the stream, n = 15/reach) at each reach.

Using 5 m x 5 m quadrats, we estimated vegetation cover by height (< 3 m, 3-5 m, and >

5 m). We visually estimated the % coverage of vegetation overhanging the water’s edge along both banks and counted standing dead trees of diameter at breast height > 10 cm.

We identified all samples using Braun (Braun 1989) as a guide.

Periphyton was collected by scrubbing a 25-cm2 section of the dominant hard substrate (cobble) at each of 5 transects within each reach following Reavie et al.

(2010). Samples were composited by reach into a single polyethylene jar and preserved in 70% ethanol solution until analysis. We collected five terrestrial vegetation samples per transect on both sides of the river (along the survey transects): three samples from the most dominant trees species and two samples from the most dominant species among the shrubs, grasses, herbs and sedges. We stored leaf samples in paper bags until further processing in the laboratory.

Riparian Swallows

At seven of the eleven research sites, we captured swallows using mist nets set over the river or in the riparian zone. We erected six nest boxes at each of four additional sites, following protocols set forth by Golondrinas de las Américas (2010). We equipped each box with an external trapping mechanism (i.e., wig-wag), which allowed the capture of adult tree swallows while they were inside the nest box. We banded all birds with

83 aluminum USFWS size 0 bands for tree (T. bicolor), bank (R. riparia), and northern rough-winged (S. serripennis) swallows, and size 1 bands for cliff (P. pyrrhonota) swallows, at first capture. We drew blood from the jugular vein to be used for stable isotope and contaminant analyses (Ardia 2005; Sullivan and Vierling 2012). We collected blood from a minimum of three adults and four nestlings at each of the experimental nest box reaches, where at least two nests were targeted per reach. At all other reaches we collected blood from a total of four swallows, except at Mosquito, where we were only able to capture three. Once drawn, we immediately transferred blood samples to centrifuge tubes (in 70% ethanol) or Dried Blood Spot (PerkinElmer,

Waltham, MA, USA) cards for contaminant analysis. All swallow collections were performed in compliance with Animal Care and Use Committee protocols

(2011A00000049 and 2009A0215); a valid Ohio Wildlife Collection Permit; and a US

Department of the Interior (USDI), US Geological Survey (USGS) Federal Banding

Permit.

Invertebrates

We deployed six floating emergent insect traps (i.e., pyramid traps) per reach as described by Aubin et al. (1973) along both banks of the reach to capture aquatic emergent insects, distributing them so that they represented dominant flow habitats (i.e., riffle, pool/run) to the degree possible. For terrestrial insects, we suspended half-sized

Malaise traps (1-m high, 1-m long, 0.6-m wide, 0.5-mm mesh, MegaView Science Co.,

Taichung, Taiwan; Townes 1972) from trees at 10 m and 50 m from the active river

84 channel on both banks. We collected all invertebrates at five and ten days. Adult tree swallows typically return to the nest with a bolus of invertebrate prey carried in their beaks, which we removed from their mouths before the parent fed the nestlings following

(see McCarty and Winkler 1999; Echols et al. 2004) at nest-box sites. We preserved these boluses in 70% ethanol.

Upon collection, we placed all invertebrates in jars, put them on ice for transport, and froze them at -20° C until analysis (Lutz 2008). In the laboratory, we identified all individuals to family using Johnson and Triplehorn (2004). For bolus invertebrates, we also classified all individuals by their origin (i.e.,, aquatic or terrestrial according to their larval habitat). Following identification, we dried and weighed each invertebrate by family (total dry mass per reach (mg m-2 d-1 ; Iwata et al. 2003).

Contaminant and Stable Isotope Analysis

All sediment, bolus, and swallow blood samples were analyzed for Hg and Se concentrations at the Diagnostic Center for Population and Animal Health at Michigan

State University.

Terrestrial vegetation, terrestrial insects, periphyton, aquatic emergent insects, and bird blood samples were analyzed for 13C and 15N stable isotopes using continuous flow isotope ratio mass spectrometry (EA- IRMS) at the University of Washington Stable

Isotope Core (Pullman, WA, USA). The isotopic composition of samples was expressed using the 13C notation defined as follows:

13 13 12  C = [(Rsample / Rstandard) – 1] * 1000, where R is C/ C (1)

85 where parts per thousand (‰) are the atomic ratios of the number of atoms in the sample or standard. Typical analytical precision was + 0.2‰ for 13C determination.

Numerical and Statistical Analyses

Because we observed no differences in 13C and 15N values between adult and juvenile swallows, we grouped them by species among study reaches. We then analyzed the contributions of aquatic and terrestrial food sources to riparian swallows at each study reach using (1) dual isotope plots of 13C and 15N, (2) linear regression between 13C of swallows and periphyton, and (3) invertebrates from bolus samples. For dual-isotope plots, spatial correspondence between 13C and 15N values in swallows and aquatic prey indicates greater use of that prey source, as DeNiro & Epstein (1978; 1981) showed that consumers retain the isotopic signatures of that which they consume. We created one plot for each land-use class (i.e., urban and rural) and plotted 13C and 15N values for swallows by species, given variability in isotopic signatures among species. Because of an unbalanced number of swallows at each site, we used weighted linear regression to determine the degree to which swallow 13C tracked periphyton 13C across all study reaches. We used invertebrate composition from bolus samples to further inform aquatic vs. terrestrial prey use by swallows.

We performed principal component analysis (PCA) on riparian LULC variables that we had selected a priori as candidate predictors of riparian swallow prey utilization.

PCA axes with eigenvalues > 1 were retained for use in subsequent linear regression models (Rencher 1995; Sullivan and Watzin 2008). We used weighted regression to 86 examine the best LULC predictors of contaminant concentrations in swallows. We used analysis of variance (ANOVA) to test for potential differences in contribution of aquatic resources and contaminant concentrations in riparian swallows between LULC types.

Given multiple regression tests, the Bonferroni adjustment for α was α/k = 0.05/5 =

0.010, where k is the number of tests (Wright 1992). Concentrations of Se and Hg were log10 transformed before statistical analyses.

Results

Sediment concentrations of Se ranged from 367 to 2440 ppb, while sediment Hg ranged from 31 to 137 ppb. Se in sediment was significantly higher at urban than rural reaches (t = 2.09, P = 0.033, df = 9). Concentrations of Se and Hg in riparian swallows varied widely across all reaches (Table 3.1). Hg concentrations in swallows were positively related to Hg concentrations in sediment (Figure 3.2), though no relationship was evident between Se in swallows and sediment. Bolus results indicated that Se concentrations were greater in aquatic ( ̅ = 3956, SD = 2039) than terrestrial ( ̅ = 1132,

SD = 1089) prey (t = -4.11, P = 0.027, df = 2), but no such relationship was found for Hg

(t = -2.17, P = 0.081, df = 2).

CONTRIBUTION OF AQUATIC RESOURCES TO SWALLOWS

Isotopic values (Table 3.2) for riparian swallows ranged from -26.77 to -20.29‰ for δ13C and from 6.63 to 12.74‰ for δ15N. Signatures for aquatic emergent insect δ13C ranged from -28.27 to -27.04‰, while δ15N ranged from 10.04 to 12.31‰. Terrestrial

87 insect prey δ13C ranged from -25.84 to -22.47‰ whereas δ15N of terrestrial insects ranged from 5.10 to 8.74‰. Terrestrial vegetation δ13C ranged from -31.53 to -28.00‰, and δ15N from 2.72‰ to 4.68‰. Periphyton exhibited the greatest variation, ranging from -21.46 to -11.26‰ for δ13C, and from 6.14 to 12.39‰ for δ15N. Periphyton δ13C was notably enriched at Berliner (-11.26‰) and Gravel Bar (-11.68 ‰), and δ15N was highly enriched at Gravel Bar (12.39‰). In spite of this variation in basal resources, we found no significant differences for either δ13C (t = -0.12, P = 0.455, df = 3) or δ15N (t =

1.33, P = 0.137, df = 3) between adult and juvenile swallows.

Swallows at both rural and urban reaches plotted within the parameters of the stream-riparian food web (Figure 3.3a & 3.3b). For rural (Figure 3.3a) and urban (Figure

3.3b) reaches, isotope plots indicated that swallows were reliant on a mixture of aquatic and terrestrial food resources, although swallows were presumably more δ15N enriched at urban sites (based on the lower δ15N enrichment of primary producers at urban reaches along with the invariant swallow δ15N signatures for swallows at both urban and rural reaches). Swallow δ13C negatively tracked periphyton δ13C at urban, but not rural, reaches (Figure 3.4), indicating that swallows were not heavily reliant on stream primary productivity. However, this relationship does not preclude swallow reliance on aquatic emergent insects, which can be highly reliant on terrestrial inputs of vegetation, as in this study. The variability explained by this regression serves as our minimum contribution of aquatic emergent insects to swallow diet (Figure 3.3).

Bolus results from the four nest-box reaches indicated that swallows at urban locations relied more heavily on aquatic insects than at rural locations (Figure 3.5). By

88 weight, aquatic emergent insects comprised 84.0% of swallow diet (vs. 16.0% from terrestrial invertebrates) at urban locations, whereas aquatic emergent insects comprised only 46.4% of swallow diet (vs. 53.6% from terrestrial invertebrates) at rural nest-box locations. Individuals of the family Coenegrionidae (narrow-winged damselflies) made up nearly 30% of all aquatic individuals collected, whereas Aeshnidae (hawkers or darners) represented the largest-bodied prey item.

Concentrations of Se in riparian swallows were linked with both δ15N and δ13C isotopic values (Figure 3.6a and 3.6b). Swallow Se concentrations increased with δ15N enrichment and δ13C depletion, indicating a shift toward aquatic prey resources.

Concentrations of Hg were not closely related to either swallow δ15N (R2 = 0.00, P =

0.971) or δ13C (R2 = 0.13, P = 0.152) signatures.

INFLUENCE OF LULC

Principal component analysis of eleven LULC measures identified three axes with eigenvalues > 1 (Table 3.3). PC1 captured about 54% of the variance, and was mainly driven by factors related to urbanized landscapes (i.e.,, population density [r2 = 0.91, + correlation], % impervious surface coverage [r2 = 0.86, +], % agriculture [r2 = 0.86, -]).

Accordingly, this PC was named ‘Urbanization Axis’. The second principal component explained about 19% of the variation that remained and was driven almost solely by % tall tree cover (r2 = 0.86, +), and is hereafter referred to as “Mature Trees Axis”. No other metric within this axis had a correlation > 0.32. The third PC, “Shoreline Habitat

Axis”, explained about 14% of the variance and was strongly influenced by standing dead trees (r2 = 0.69) and % overhanging vegetation (r2 = 0.48).

At a coarse landscape resolution, we found that swallow Hg concentrations were significantly higher in rural than urban reaches (t = -2.96, P = 0.003, df = 24), and marginally so for Se (t = -1.54, P = 0.068, df = 24). A combination of urban characteristics and minimal agricultural activity in the riparian zone appeared to be in part driving this association, as we found that the Urbanization axis was negatively related to both Hg (Figure 3.7a) and Se concentration in swallows (Figure 3.7b).

Urbanization was also negatively related to δ15N enrichment (Figure 3.8), which in our system indicates a decreasing reliance on aquatic prey along this axis.

We found that Hg concentrations were significantly higher in adult swallows

( ̅ 163.1 ppb) than juveniles ( ̅ 49.5 ppb) (t = -4.35, P = 0.006, df = 4). Se concentrations in adults ( ̅ 8088 ppb) were also significantly higher than in juveniles

( ̅ 1953 ppb) (t = -3.47, P = 0.013, df = 4). We found no significant differences in Hg or Se concentrations between male and female swallows at either urban or rural reaches.

Discussion

Aquatic emergent insects serve as important prey items for myriad riparian consumers (Akamatsu et al. 2004; Ballinger and Lake 2006). As a result of these strong, cross-system linkages, terrestrial consumers are exposed to aquatic contaminants

(Walters et al. 2008; Akamatsu and Toda 2011). Walters et al. (2008), for example, showed how aquatic emergent insects can transport PCBs from contaminated lake

90 sediments into adjacent riparian and terrestrial food webs. Our understanding of the mechanisms that influence and mediate these exposure pathways remains limited, but given the geographical range and abundance of aquatic emergent insects and riparian consumers, and the global proliferation of aquatic contamination (Menzie 1980; Walters et al. 2008), an improved grasp of these cross-boundary flows will be crucial for managing terrestrial exposure risk from aquatic contaminants.

Many investigators have illustrated the variability in swallow dietary composition.

For example, Blancher and McNicol (1991) determined that the diet of non-egg-laying adults was comprised of 54.9% aquatic insects by biomass. Similarly, Wayland et al.

(1998) found that insects of aquatic origin made up 50-60% of tree swallow diet in western Canada. Collectively, our results indicate that swallows rely on a mixture of terrestrial and aquatic insects, but that reliance on aquatic insects is likely greater at urban vs. rural reaches (Figures 3.2a & 3.2b,3.4).

As we hypothesized, sediment contaminant concentrations differed between land-use classes, as concentrations of Se were higher in urban than rural reaches.

However, Hg and Se concentrations in riparian swallows were lower at urban reaches than rural, despite the fact that urban nest-box swallows depredated aquatic insects more commonly than terrestrial insects, and that bolus samples indicate that aquatic insects exhibited higher concentrations of Se and Hg than terrestrial insects. Concomitantly, concentrations of Hg in swallow blood and river sediment were positively related (Figure

3.2), which suggests that riparian swallows are exposed to aquatic contaminants through aquatic prey, Likewise, Se concentrations in swallows were positively linked to both

δ15N enrichment and δ13C depletion (Figure 3.6), indicators of reliance on aquatic emergent insects in our study system. Swallows also fed at a higher trophic level at our urban reaches (Figure 3.2), indicating a potential increase in aquatic prey utilization. One possible explanation for this apparent discrepancy between contribution of aquatic prey and contamination patterns is that swallows were actively selecting prey that were not dominant within the system, and therefore not included in our isotopic or contaminant analyses. Blancher and McNicol, (1991) for example, found that mayflies were an important component of tree swallow diet despite composing a small fraction of the insects caught in emergent traps. It is also possible that varying bioavailability amongst metal speciations found within the sediments leads to unequal transfer to swallows through predation of aquatic emergent insects. For instance, Langner et al. (2012) observed that blood Hg concentrations in osprey (Pandion haliaetus) chicks did not reflect Hg concentrations in river sediments, which they suggested was due to decreased methylation, and thus bioavailability, of Hg.

Influences of land cover on avian foraging behavior likely explain patterns we observed in swallow contaminant concentrations. For example, Ormerod and Tyler

(1991b) found that grey wagtails (Motacilla cinerea) changed their diet to incorporate more lepidopteran larvae in broadleaved forests than in moorlands or coniferous forests in Wales. Such dietary shifts in response to land-cover variation could have a profound effect on contaminant uptake. Biomagnification of aquatic contaminants in tree swallows, for instance, is known to be linked with their reliance on aquatic emergent insects (Brasso and Cristol 2008; Hallinger et al. 2011; Jackson et al. 2011), which is

92 consistent with our results for Se (Figure 3.6). Many riparian swallows tend to forage in open areas (Brown et al. 2002; Ghilain and Belisle 2008), indicating a likely preference for feeding over open water vs. heavily canopied riparian areas. Our urban reaches were characterized by densely forested riparian areas, leaving little access to open foraging areas other than directly over water, which might explain why urban swallows appeared to feed more heavily on aquatic emergent insects than their rural counterparts.

Contrary to our expectations, we found no differences in the aquatic contribution to the diet of male and female, or adult and juvenile swallows. Blancher and McNicol

(1991) reported that diets of egg-laying females were comprised of approximately 71% aquatic insects, compared to just less than 55% for other non-egg-laying adults. These same authors found that nestling diet was almost 65% aquatic, higher than our bolus findings from rural reaches, and lower than bolus results at urban reaches. Mengelkoch et al. (2004) found that 50% of the diet of nestling tree swallows in Minnesota was comprised of aquatic insects, compared to 44% for terrestrial insects, which is consistent with results from our rural reaches. Again, the dense riparian cover of our urban reaches likely explains the variation in aquatic prey contribution to swallows between land-use classes.

Our observation that contaminant biomagnification was higher in adult swallows relative to juveniles aligns with other studies and was consistent with our hypothesis.

Evers et al. (2005), for instance, found that methylmercury (MeHg) concentrations were five to ten times higher in adult tree swallows than in juveniles. Although we did not observe a difference in the dietary composition (aquatic vs. terrestrial) between adult and

93 immature swallows, nestlings might efficiently remove contaminants from their blood as they are deposited into growing feathers (Tsipoura et al.2008). Alternatively, adult swallows may feed on larger-bodied aquatic insects, thereby increasing exposure to aquatic contaminants. For instance, Quinney & Ankney (1985) found that insects 1-6 mm in length made up between 80 and 86% of nestling tree swallow diet in Canada, whereas McCarty & Winkler (1999) found that adult tree swallows actively selected for larger bodied prey, like Odonata, and against small insects (0-3 mm in length).

CONCLUSION

Overall, our results (1) support recent research examining the cross-system trophic linkages between aquatic emergent insects and riparian consumers (Ballinger and

Lake 2006; Fukui et al. 2006; Daley et al. 2011), and (2) implicate landscape alterations as a key factor in regulating aquatic-to-terrestrial contaminant transfers. Our results also provide insight into the role of riparian swallows as integrators of linked aquatic- terrestrial systems within the context of contaminant exposure risk. Specifically, the reliance of swallows upon a mixture of terrestrial and aquatic invertebrate prey makes them highly vulnerable to contaminants from both terrestrial and aquatic systems. Their ability to range freely and utilize multiple prey sources makes them excellent organisms for examination of large, cross-boundary trophic and material transfer processes.

Building upon previous work (Blancher and McNicol 1991; Bishop et al. 1999;

Harris and Elliott 2000; McCarty 2002), we suggest that riparian swallow species may capture system-level mechanisms of contamination. We argue that the trophic dynamics

94 of riparian swallows represent a critical pathway for cross-boundary contaminant transfer and movement across linked aquatic-terrestrial landscapes. Additional research relative to species-specific swallow feeding behavior, movement patterns, and resource utilization will provide valuable insight into examination of multi-system contamination pathways.

The potential role of swallows themselves as vectors of contaminants (e.g., via feces, feathers, and nesting colonies) across both local (e.g., nesting) and broader (i.e., migratory) spatial scales will also be important in further understanding aquatic-terrestrial contaminant fluxes. Additional research related to the mediating role of terrestrial landscapes on transfers of contaminants from aquatic to terrestrial systems will also be critical for improved remediation and long-term conservation of river ecosystems.

Acknowledgements

Research support was provided by state and federal funds appropriated to The Ohio State

University, Ohio Agricultural Research and Development Center. We would like to thank Adam Kautza and Paradzayi Tagwireyi for their help in the field and laboratory.

References

Akamatsu, F. & H. Toda (2011) Aquatic subsidies transport anthropogenic nitrogen to riparian spiders. Environmental Pollution, 159, 1390-1397.

Akamatsu, F., H. Toda & T. Okino (2004) Food source of riparian spiders analyzed by using stable isotope ratios. Ecological Research, 19, 655-662.

Ardia, D. R. (2005) Tree swallows trade off immune function and reproductive effort differently across their range. Ecology, 86, 2040-2046.

Aubin, A., J. P. Bourassa & Pellissi.M (1973) Effective emergence trap for capture of mosquitos. Mosquito News, 33, 251-252.

Ballinger, A. & P. S. Lake (2006) Energy and nutrient fluxes from rivers and streams into terrestrial food webs. Marine and Freshwater Research, 57, 15-28.

Bartel, R. A., N. M. Haddad & J. P. Wright (2010) Ecosystem engineers maintain a rare species of butterfly and increase plant diversity. Oikos, 119, 883-890.

Bigler, W. J., J. H. Jenkins, P. M. Cumbie, G. L. Hoff & E. C. Prather (1975) Wildlife and environmental health - raccoons as indicators of zoonoses and pollutants in southeastern United States. Journal of the American Veterinary Medical Association, 167, 592-597.

Bishop, C. A., N. A. Mahony, S. Trudeau & K. E. Pettit (1999) Reproductive success and biochemical effects in tree swallows (Tachycineta bicolor) exposed to chlorinated hydrocarbon contaminants in wetlands of the Great Lakes and St. Lawrence River basin, USA and Canada. Environmental Toxicology and Chemistry, 18, 263-271.

Blancher, P. J. & D. K. McNicol (1991) Tree swallow diet in relation to wetland acidity. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 69, 2629-2637.

Brasso, R. L. & D. A. Cristol (2008) Effects of mercury exposure on the reproductive success of tree swallows (Tachycineta bicolor). Ecotoxicology, 17, 133-141.

Braun, E. L. 1989. Woody Plants of Ohio: Trees, Shrubs, and Woody Climbers, Native, Naturalized, and Escaped. Columbus, OH: The Ohio State University Press.

Briers, R. A., H. M. Cariss, R. Geoghegan & J. H. R. Gee (2005) The lateral extent of the subsidy from an upland stream to riparian lycosid spiders. Ecography, 28, 165- 170.

Brown, C. R. & M. B. Brown (1996) Coloniality in the cliff swallow: The effect of group size on social behavior. Coloniality in the cliff swallow: The effect of group size on social behavior, xiii+566p.

Brown, C. R., C. M. Sas & M. B. Brown (2002) Colony choice in Cliff swallows: Effects of heterogeneity in foraging habitat. Auk, 119, 446-460.

Cloe, W. W. & G. C. Garman (1996) The energetic importance of terrestrial arthropod inputs to three warm-water streams. Freshwater Biology, 36, 105-114.

Custer, C. M. (2011) Swallows as a Sentinel Species for Contaminant Exposure and Effect Studies. Wildlife Ecotoxicology: Forensic Approaches, 3, 45-91.

Deniro, M. J. & S. Epstein (1978) Influence of diet on distribution of carbon isotopes in animals. Geochimica Et Cosmochimica Acta, 42, 495-506.

Deniro, M.J. & S. Epstein (1981) Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica Et Cosmochimica Acta, 45, 341-351.

Dods, P. L., E. M. Birmingham, T. D. Williams, M. G. Ikonomou, D. T. Bennie & J. E. Elliott (2005) Reproductive success and contaminants in tree swallows

(Tachycineta bicolor) breeding at a wastewater treatment plant. Environmental Toxicology and Chemistry, 24, 3106-3112.

Dreyer, J., D. Hoekman & C. Gratton (2012) Lake-derived midges increase abundance of shoreline terrestrial arthropods via multiple trophic pathways. Oikos, 121, 252- 258.

Echols, K. R., D. E. Tillitt, J. W. Nichols, A. L. Secord & J. P. McCarty (2004) Accumulation of PCB congeners in nestling tree swallows (Tachycineta bicolor) on the Hudson River, New York. Environmental Science & Technology, 38, 6240-6246.

Fitzgerald, W. F., D. R. Engstrom, R. P. Mason & E. A. Nater (1998) The case for atmospheric mercury contamination in remote areas. Environmental Science & Technology, 32, 1-7.

Fukui, D., M. Murakami, S. Nakano & T. Aoi (2006) Effect of emergent aquatic insects on bat foraging in a riparian forest. Journal of Animal Ecology, 75, 1252-1258.

Ghilain, A. & M. Belisle (2008) Breeding success of tree swallows along a gradient of agricultural intensification. Ecological Applications, 18, 1140-1154.

Golondrinas de las Americas (2010). The Golondrinas Handbook. http://golondrinas.cornell.edu/Data_and_Protocol/The%20Golondrinas%20Handb ook%20version%201-8.pdf. Accessed March 24, 2010.

Hagen, E. M. & J. L. Sabo (2012) Influence of river drying and insect availability on bat activity along the San Pedro River, Arizona (USA). Journal of Arid Environments, 84, 1-8.

Hallinger, K. K., K. L. Cornell, R. L. Brasso & D. A. Cristol (2011) Mercury exposure and survival in free-living tree swallows (Tachycineta bicolor). Ecotoxicology, 20, 39-46.

Hamas, M.J. (1994) Belted kingfisher (Cerlye alcyon). In Poole A, Gill F, eds, The Birds of North America, No. 84. The American Ornithologists' Union, Washington, DC, USA.

Hammerschmidt, C. R. & W. F. Fitzgerald (2005) Methylmercury in mosquitoes related to atmospheric mercury deposition and contamination. Environmental Science & Technology, 39, 3034-3039.

Harris, M. L. & J. E. Elliott (2000) Reproductive success and chlorinated hydrocarbon contamination in tree swallows (Tachycineta bicolor) nesting along rivers receiving pulp and paper mill effluent discharges. Environmental Pollution, 110, 307-320.

Hebert, C. E., J. L. Shutt, K. A. Hobson & D. V. C. Weseloh (1999) Spatial and temporal differences in the diet of Great Lakes herring gulls (Larus argentatus): evidence from stable isotope analysis. Canadian Journal of Fisheries and Aquatic Sciences, 56, 323-338.

Henschel, J. R., D. Mahsberg & H. Stumpf (2001) Allochthonous aquatic insects increase predation and decrease herbivory in river shore food webs. Oikos, 93, 429-438.

Iwata, T., S. Nakano & M. Murakami (2003) Stream meanders increase insectivorous bird abundance in riparian deciduous forests. Ecography, 26, 325-337.

Jackson, A. K., D. C. Evers, S. B. Folsom, A. M. Condon, J. Diener, L. F. Goodrick, A. J. McGann, J. Schmerfeld & D. A. Cristol (2011) Mercury exposure in terrestrial birds far downstream of an historical point source. Environmental Pollution, 159, 3302-3308.

Johnson, N. F. & C. A. Triplehorn. 2004. Borror and Delong's Introduction to the Study of Insects. Thomson Brooks/Cole.

Keast, A. (1990) The annual cycle and activity on the breeding grounds in a Canadian broad-leafed deciduous forest bird community relationship to the prey resource base. Keast, a. (Ed.). Biogeography and Ecology of Forest Bird Communities. V+410p. Spb Academic Publishing Bv: the Hague, Netherlands. Illus. Maps, 197- 214.

Langner,H.W., Greene, E., Domenech, R. and M.F. Staats (2012) Mercury and other mining-related contaminants in ospreys along the Upper Clark Fork River,

Montana, USA. Archives of Environmental Contamination and Toxicology, 62, 681-695.

Lord, C. G., K. F. Gaines, C. S. Boring, L. L. Brisbin, M. Gochfeld & J. Burger (2002) Raccoon (Procyon lotor) as a bioindicator of mercury contamination at the US Department of Energy's Savannah River Site. Archives of Environmental Contamination and Toxicology, 43, 356-363.

Luther, D., J. Hilty, J. Weiss, C. Cornwall, M. Wipf & G. Ballard (2008) Assessing the impact of local habitat variables and landscape context on riparian birds in agricultural, urbanized, and native landscapes. Biodiversity and Conservation, 17, 1923-1935.

Lutz M.A. (2008) Procedures for collecting and processing streambed sediment and pore water for analysis of mercury as part of the National Water-Quality Assessment Program. Open File Report 2008-1279.

Malinga, M., P. Szefer & G. W. Gabrielsen (2010) Age, sex and spatial dependent variations in heavy metals levels in the Glaucous Gulls (Larus hyperboreus) from the Bjornoya and Jan Mayen, Arctic. Environmental Monitoring and Assessment, 169, 407-416.

Mattson, D. J. & D. P. Reinhart (1995) Influences of cutthroat trout (Oncorhynchus clarki) on behaviour and reproduction of Yellowstone grizzly bears (Ursus arctos), 1975-1989. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 73, 2072-2079.

Maul, J. D., J. B. Belden, B. A. Schwab, M. R. Whiles, B. Spears, J. L. Farris & M. J. Lydy (2006) Bioaccumulation and trophic transfer of polychlorinated biphenyls by aquatic and terrestrial insects to tree swallows (Tachycineta bicolor). Environmental Toxicology and Chemistry, 25, 1017-1025.

McCarty, J. P. (1997) Aquatic community characteristics influence the foraging patterns of tree swallows. Condor, 99, 210-213.

McCarty (2002) Use of tree swallows in studies of environmental stress. Reviews in Toxicology (Amsterdam), 4, 61-104.

McCarty, J. P. & D. W. Winkler (1999) Foraging ecology and diet selectivity of tree swallows feeding nestlings. Condor, 101, 246-254.

100

Mengelkoch, J. M., G. J. Niemi & R. R. Regal (2004) Diet of the nestling Tree Swallow. Condor, 106, 423-429.

Menzie, C. A. (1980) Potential significance of insects in the removal of contaminants from aquatic systems. Water Air and Soil Pollution, 13, 473-479.

Morrissey, C. A., L. I. Bendell-Young & J. E. Elliott (2004) Linking contaminant profiles to the diet and breeding location of American dippers using stable isotopes. Journal of Applied Ecology, 41, 502-512.

Nakano, S. & M. Murakami (2001) Reciprocal subsidies: Dynamic interdependence between terrestrial and aquatic food webs. Proceedings of the National Academy of Sciences of the United States of America, 98, 166-170.

Oneal, A. S. & J. T. Rotenberry (2009) Scale-dependent habitat relations of birds in riparian corridors in an urbanizing landscape. Landscape and Urban Planning, 92, 264-275.

Ormerod, S. J. & S. J. Tyler (1991a) Exploitation of prey by a river bird, the dipper (Cinclus cinclus) along acidic and circumneutral streams in upland Wales. Freshwater Biology, 25, 105-116.

Ormerod, S.J. & S.J. Tyler (1991b) The influence of stream acidification and riparian land-use on the feeding ecology of gray wagtails (Motacilla cinerea) in Wales. Ibis, 133, 53-61.

Paetzold, A., C. J. Schubert & K. Tockner (2005) Aquatic terrestrial linkages along a braided-river: Riparian arthropods feeding on aquatic insects. Ecosystems, 8, 748- 759.

Pennington, D. N., J. Hansel & R. B. Blair (2008) The conservation value of urban riparian areas for landbirds during spring migration: Land cover, scale, and vegetation effects. Biological Conservation, 141, 1235-1248.

101

Reavie, E. D., T. M. Jicha, T. R. Angradi, D. W. Bolgrien & B. H. Hill (2010) Algal assemblages for large river monitoring: Comparison among biovolume, absolute and relative abundance metrics. Ecological Indicators, 10, 167-177.

Rencher, A. C. 1995. Methods of Multivariate Analysis. New York, NY, U.S.A.: Johny Wiley and Sons, Inc.

Roux, K.E. and P.P. Marra (2007) The presence and impact of environmental lead in passerine birds along an urban to rural land use gradient. Archives of Environmental Toxicology and Chemistry, 53, 261-268.

Sabo, J. L. & M. E. Power (2002) Numerical response of lizards to aquatic insects and short-term consequences for terrestrial prey. Ecology, 83, 3023-3036.

Sanzone, D.M., Meyer, J.L., Marti, E., Gardiner, E.P., Tank, J.L., and N.B. Grimm (2003) Carbon and nitrogen transfer from a desert stream to riparian predators. Oecologia, 134, 238-250.

Sullivan, S. M. P. & K. T. Vierling (2012) Exploring the influences of multiscale environmental factors on the American dipper Cinclus mexicanus. Ecography, 35, 624-636.

Sullivan, S.M.P & K.T. Vierling (Submitted. Ecological Monographs) Deciphering spatial linkages in mountain watersheds with the American dipper (Cinclus mexicanus).

Sullivan, S.M.P., Watzin, M.C., & W.S. Keeton (2007) A riverscape perspective on habitat assocations among riverine bird assemblages in the Lake Champlain Basin, USA. Landscape Ecology, 22, 1169-1186.

Sweeney, B. W. & R. L. Vannote (1982) Population synchrony in mayflies - a predator satiation hypothesis. Evolution, 36, 810-821.

Townes, H. (1972) A light weight malaise trap. Entomological News, 83, 239-247.

Tsipoura, N., J. Burger, R. Feltes, J. Yacabucci, D. Mizrahi, C. Jeitner & M. Gochfeld (2008) Metal concentrations in three species of passerine birds breeding in the Hackensack Meadowlands of New Jersey. Environmental Research, 107, 218- 228.

102

United States Census Bureau (2010). State and county quick facts. http://quickfacts.census.gov/qfd/states/39/39049.html. Accessed July 29, 2012.

United States Environmental Protection Agency (2012). Basic information about mercury (inorganic) in drinking water. http://water.epa.gov/drink/contaminants/basicinformation/mercury.cfm. Accessed July 28, 2012.

Vander Zanden M.J., & and D.M. Sanzone (2004) Food web subsidies at the land-water ecotone. In Food Webs at the Landscape Level (eds G.A. Polis, M.E. Power, G.R. Huxel). University of Chicago Press, Chicago, USA.

Varley, J. D. & P. Schullery (1996) Yellowstone Lake and its cutthroat trout. Science and ecosystem management in the national parks, 49-73.

Vessel, R.D. (1978) Energetics of the belted kingfisher (Megaceryle alcyon). Oregon State University, Corvalis, OR, USA.

Wada, E., H. Mizutani & M. Minagawa (1991) The use of stable isotopes for food web analysis. Critical Reviews in Food Science and Nutrition, 30, 361-371.

Wada, H., D. A. Cristol, F. M. A. McNabb & W. A. Hopkins (2009) Suppressed Adrenocortical Responses and Thyroid Hormone Levels in Birds near a Mercury- Contaminated River. Environmental Science & Technology, 43, 6031-6038.

Walters, D. M., K. M. Fritz & R. R. Otter (2008) The dark side of subsidies: adult stream insects export organic contaminants to riparian predators. Ecological Applications, 18, 1835-1841.

Wayland, M., S. Trudeau, T. Marchant, D. Parker & K. A. Hobson (1998) The effect of pulp and paper mill effluent on an insectivorous bird, the tree swallow. Ecotoxicology, 7, 237-251.

Wright, S.P. (1992) Adjusted p-values for simultaneous inference. Biometrics, 45, 1005- 1013.

Zelt, R.B., & M.R. Johnson (2005) Protocols for Mapping and Characterizing Land Use/Land Cover in Riparian Zones. U.S. Geological Survey Open-File Report 2005-1302.

103

Study Reach Species Land Use Se ̅ (SD) Hg ̅ (SD) (ppb) (ppb) Wetlands TRSW Urban 3711 (3118) 34.3 (26.6) Fawcett TRSE Urban 6569 (5280) 67.4 (51.1) Horseshoe CLSW Urban 3572 (1485) 124.2 (26.0) Berliner TRSW Urban 2512 (1621) 35.5 (44.6) Mosquito NRWS Rural 17506 (3510) 210.5 (34.6) Darby TRSW Rural 2672 (1238) 206.2 (146.6) Gravel Bar BANS Rural 8519 (781) 115 (25.4) Piketon CLSW Rural 1109 (880) 54.0 (40.9) Treefall BANS Rural 12670 (3146) 113.2 (29.5) Last BANS Rural 14268 (4303) 295 (136) Wetlands Sediment Urban 700 31 Horseshoe Sediment Urban 1230 69 Berliner Sediment Urban 2440 99 Mosquito Sediment Rural 1992 137 Gravel Bar Sediment Rural 591 48 Piketon Sediment Rural 419 33 Treefall Sediment Rural 367 33 Last Sediment Rural 494 59

104

Table 3.2. δ13C and δ15N values for riparian swallows, periphyton, terrestrial vegetation, emergent aquatic insects, and terrestrial invertebrates for all study reaches in the Scioto River basin, Ohio, USA. Sites are presented upstream to downstream. Note: CLSW = cliff swallow, BANS = bank swallow, TRSW = tree swallow, NRWS = northern rough- winged swallow.

Stable Isotope Analysis (all in ‰) Study Reach Land Use Swallows Periphyton Terrestrial Vegetation Emergent Aquatic Insects Terrestrial Insects Species δ13C δ15N δ13C δ15N δ13C δ15N δ13C δ15N δ13C δ15N Highbanks Rural TRSW -24.69 10.016 N/A N/A N/A N/A N/A N/A N/A N/A Wetlands Urban TRSW -24.55 11.00 -18.49 6.88 -30.96 2.72 -27.41 11.29 -27.78 11.15 Fawcett Urban TRSW -24.42 10.80 N/A N/A N/A N/A N/A N/A N/A N/A Horseshoe Urban CLSW -23.24 9.04 -17.69 6.14 -30.37 3.08 -27.51 10.04 -27.97 12.31 Berliner Urban TRSW -26.14 12.39 -11.26 7.35 -30.81 4.22 -28.27 11.09 -27.99 12.66 Mosquito Rural NRWS -24.61 11.77 -18.79 8.42 -28.06 4.68 -27.04 11.55 -27.34 11.83 Darby Rural TRSW -24.07 10.00 N/A N/A N/A N/A N/A N/A N/A N/A Gravel Bar Rural BANS -24.42 10.37 -11.68 12.39 -28.00 3.74 -27.78 11.15 -25.09 6.79 Piketon Rural CLSW -23.35 7.91 -21.46 8.66 -31.52 4.53 -27.97 12.31 -25.09 7.74 Treefall Rural BANS -25.15 10.87 -14.82 8.78 -31.53 3.27 -27.99 12.66 -23.83 8.31 Last Rural BANS -23.98 11.13 -13.41 9.64 -30.56 2.89 -27.34 11.83 -25.42 7.64

105

Urbanization Mature Trees Shoreline Habitat Adjacent LULC Metrics Loading r 2 Loading r 2 Loading r 2 % Agriculture -0.38 0.86 0.08 0.01 0.02 0.00 % Shrub Cover (1-3m) 0.34 0.69 0.02 0.00 -0.38 0.22 % Small Tree Cover (3-5m) 0.30 0.54 0.31 0.20 0.05 0.00 % Tall Tree Cover (>5m) -0.01 0.00 0.64 0.86 0.17 0.04 # Standing Dead Trees 0.00 0.00 -0.34 0.24 0.67 0.69 Riparian Forest Width -0.30 0.54 -0.10 0.02 -0.08 0.01 # Bridge Crossings 0.30 0.54 -0.39 0.32 -0.17 0.04 % Impervious Surface 0.38 0.86 -0.23 0.11 -0.04 0.00 # Canopy Layers 0.32 0.61 0.36 0.27 -0.09 0.01 Population Density 0.39 0.91 -0.10 0.02 0.13 0.03 % Overhanging Vegetation 0.26 0.40 0.13 0.04 0.56 0.48

Eigenvalue 5.98 2.10 1.53 % variance 54.4 19.1 14.0

106

Figure 3.1. Eleven study reaches of the Scioto River basin, Ohio, USA. Open circles in popouts indicate study reaches. Circles in right-hand popout represent nest box locations. Popouts are not to scale.

107

Figure 3.2. Relationship between Hg concentrations (log10) in riparian swallows and sediment (ppb). The slope for the regression model was significant (y = 2.95 + 0.46x, R2 = 0.23, P = 0.030).

108

Figure 3.3. Dual isotope plots of δ13C and δ15N values (mean ± 1SE) for the Scioto River basin (Ohio, USA) aquatic and terrestrial food-web components of the Scioto River basin (Ohio, USA) in (a) rural and (b) urban reaches.

109

110

Figure 3.5. Contribution, by percentage of biomass, of aquatic emergent insects to swallow boluses from nest box reaches.

111

15 Figure 3.6. Relationship between swallow Se concentration (log10) and (a) swallow δ N and (b) swallow δ13C. The slopes for the regression models were significant for both δ15N (y = 3.26 + 0.55x, R2 = 0.76, P < 0.001) and δ13C (y = 1.90 - 0.30x, R2 = 0.27, P = 0.032).

112

Figure 3.7. Relationship between Urbanization Axis (i.e., PC1) and (a) Hg concentration (log10) and (b) Se concentration (log10) in riparian swallows. Degree of urbanization increases from -3 to 4. The slope for the regression models were significant for both Hg (y = 4.91-0.11x, R2 = 0.38, P = 0.004) and Se (y = 9.12 – 0.17x, R2 = 0.63, P < 0.001).

113

114

References

Akamatsu, F., Toda, H. & Okino, T. (2004) Food source of riparian spiders analyzed by using stable isotope ratios. Ecological Research, 19, 655-662.

Akamatsu, F. & H. Toda (2011) Aquatic subsidies transport anthropogenic nitrogen to riparian spiders. Environmental Pollution, 159, 1390-1397.

Allan, J.D., Erickson, D.L. & Fay, J. (1997) The influence of catchment land use on stream integrity across multiple spatial scales. Freshwater Biology, 37, 149-161.

Andersen, J. (1995) A comparison of pitfall trapping and quadrat sampling of Carabidae (Coleoptera) on river banks. Entomologica Fennica, 6, 65-77.

Ardia, D. R. (2005) Tree swallows trade off immune function and reproductive effort differently across their range. Ecology, 86, 2040-2046.

Aubin, A., J. P. Bourassa & Pellissi.M (1973) Effective emergence trap for capture of mosquitos. Mosquito News, 33, 251-252.

Ausprey, I.J. & Rodewald, A.D. (2011) Postfledgling survivorship and habitat selection across a rural-to-urban landscape gradient. Auk, 128, 293-302.

115

Bakermans, M.H. & Rodewald, A.D. (2006) Scale-dependent habitat use of Acadian Flycatcher (Empidonax virescens) in central Ohio. Auk, 123, 368-382.

Ballinger, A. & P. S. Lake (2006) Energy and nutrient fluxes from rivers and streams into terrestrial food webs. Marine and Freshwater Research, 57, 15-28.

Bartel, R. A., N. M. Haddad & J. P. Wright (2010) Ecosystem engineers maintain a rare species of butterfly and increase plant diversity. Oikos, 119, 883-890

Baxter, C.V., Fausch, K.D. & Saunders, W.C. (2005) Tangled webs: reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biology, 50, 201- 220.

Bihn, J.H., Verhaagh, M. & Brand, R. (2008) Ecological stoichiometry along a gradient of forest succession: Bait preferences of litter ants. Biotropica, 40, 597-599.

116

Blancher, P.J. & McNicol, D.K. (1991) Tree swallow diet in relation to wetland acidity. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 69, 2629-2637.

Blocksom, K.A. (2011) Development of a regional macroinvertebrate index for large river bioassessment (vol 9, pg 313, 2009). Ecological Indicators, 11, 757-757.

Brasso, R. L. & D. A. Cristol (2008) Effects of mercury exposure on the reproductive success of tree swallows (Tachycineta bicolor). Ecotoxicology, 17, 133-141.

Braun, E. L. 1989. Woody Plants of Ohio: Trees, Shrubs, and Woody Climbers, Native, Naturalized, and Escaped. Columbus, OH: The Ohio State University Press.

Briers, R. A., H. M. Cariss, R. Geoghegan & J. H. R. Gee (2005) The lateral extent of the subsidy from an upland stream to riparian lycosid spiders. Ecography, 28, 165- 170.

Brown, C.R., Sas, C.M. & Brown, M.B. (2002) Colony choice in Cliff swallows: Effects of heterogeneity in foraging habitat. Auk, 119, 446-460.

Burdon, F.J. & Harding, J.S. (2008) The linkage between riparian predators and aquatic insects across a stream-resource spectrum. Freshwater Biology, 53, 330-346.

117

Burton, T.M. (1976) An analysis of the feeding ecology of the salamanders (Amphibia urodela) of the Hubbard Brook Experimental Forest New Hampshire, USA. Journal of Herpetology, 10, 187-204.

Clements, W.H. & Newman, M.C. (2002) Community Ecotoxicology. John Wiley & Sons, Chichester, UK.

Cloe, W. W. & G. C. Garman (1996) The energetic importance of terrestrial arthropod inputs to three warm-water streams. Freshwater Biology, 36, 105-114.

Cover, S. P. & B. L. Fisher. 2007. Ants of North America. University of California Press.

Covich, A.P., Palmer, M.A. & Crowl, T.A. (1999) The role of benthic invertebrate species in freshwater ecosystems. BioScience, 49, 119-127.

Currie, R. S., W. L. Fairchild & D. C. G. Muir (1997) Remobilization and export of cadmium from lake sediments by emerging insects. Environmental Toxicology and Chemistry, 16, 2333-2338.

118

Custer, C. M. (2011) Swallows as a Sentinel Species for Contaminant Exposure and Effect Studies. Wildlife Ecotoxicology: Forensic Approaches, 3, 45-91.

Deniro, M. J. & S. Epstein (1978) Influence of diet on distribution of carbon isotopes in animals. Geochimica Et Cosmochimica Acta, 42, 495-506.

Deniro, M.J. & S. Epstein (1981) Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica Et Cosmochimica Acta, 45, 341-351.

Dorea, J.G. (2008) Elevated PCB levels in anglers and unsuspected transport of pollutants from aquatic food webs into human foods. Environmental Research, 108, 268-268.

Dreyer, J., D. Hoekman & C. Gratton (2012) Lake-derived midges increase abundance of shoreline terrestrial arthropods via multiple trophic pathways. Oikos, 121, 252- 258.

119

Fitzgerald, W.F., Engstrom, D.R., Mason, R.P. & Nater, E.A. (1998) The case for atmospheric mercury contamination in remote areas. Environmental Science & Technology, 32, 1-7.

Fukui, D., M. Murakami, S. Nakano & T. Aoi (2006) Effect of emergent aquatic insects on bat foraging in a riparian forest. Journal of Animal Ecology, 75, 1252-1258.

Ghilain, A. & Belisle, M. (2008) Breeding success of tree swallows along a gradient of agricultural intensification. Ecological Applications, 18, 1140-1154.

Gillespie, R.G. & T. Caraco (1987) Risk-sensitive foraging strategies of two spider populations. Ecology, 68, 887-899.

Golondrinas de las Americas (2010). The Golondrinas Handbook. http://golondrinas.cornell.edu/Data_and_Protocol/The%20Golondrinas%20Handb ook%20version%201-8.pdf. Accessed March 24, 2010.

120

Hagen, E. M. & J. L. Sabo (2012) Influence of river drying and insect availability on bat activity along the San Pedro River, Arizona (USA). Journal of Arid Environments, 84, 1-8.

Hallinger, K. K., K. L. Cornell, R. L. Brasso & D. A. Cristol (2011) Mercury exposure and survival in free-living tree swallows (Tachycineta bicolor). Ecotoxicology, 20, 39-46.

Hamas, M.J. (1994) Belted kingfisher (Cerlye alcyon). In Poole A, Gill F, eds, The Birds of North America, No. 84. The American Ornithologists' Union, Washington, DC, USA.

Hammerschmidt, C. R. & W. F. Fitzgerald (2005) Methylmercury in mosquitoes related to atmospheric mercury deposition and contamination. Environmental Science & Technology, 39, 3034-3039.

121

Henschel, J.R., Mahsberg, D. & Stumpf, H. (2001) Allochthonous aquatic insects increase predation and decrease herbivory in river shore food webs. Oikos, 93, 429-438.

Hering, D. & Plachter, H. (1997) Riparian ground beetles (Coeloptera, Carabidae) preying on aquatic invertebrates: A feeding strategy in alpine floodplains. Oecologia, 111, 261-270.

Herring, G., Ackerman, J.T. & Herzog, M.P. (2012) Mercury Exposure May Suppress Baseline Corticosterone Levels in Juvenile Birds. Environmental Science & Technology, 46, 6339-6346.

Hofer, C., Gallagher, F.J. & Holzapfel, C. (2010) Metal accumulation and performance of nestlings of passerine bird species at an urban brownfield site. Environmental Pollution, 158, 1207-1213.

Iwata, T., Nakano, S. & Murakami, M. (2003) Stream meanders increase insectivorous bird abundance in riparian deciduous forests. Ecography, 26, 325-337.

Jackson, A.K., Evers, D.C., Folsom, S.B., Condon, A.M., Diener, J., Goodrick, L.F., McGann, A.J., Schmerfeld, J. & Cristol, D.A. (2011) Mercury exposure in

122

terrestrial birds far downstream of an historical point source. Environmental Pollution, 159, 3302-3308.

Jackson, J.K. & Fisher, S.G. (1986) Secondary production, emergence, and export of aquatic insects of a Sonoran Desert stream. Ecology, 67, 629-638.

Johnson, N. F. & C. A. Triplehorn. 2004. Borror and Delong's Introduction to the Study of Insects. Thomson Brooks/Cole.

Kato, C., Iwata, T., Nakano, S. & Kishi, D. (2003) Dynamics of aquatic insect flux affects distribution of riparian web-building spiders. Oikos, 103, 113-120.

Kato, C., Iwata, T. & Wada, E. (2004) Prey use by web-building spiders: stable isotope analyses of trophic flow at a forest-stream ecotone. Ecological Research, 19, 633- 643.

Kay, A. (2002) Applying optimal foraging theory to assess nutrient availability ratios for ants. Ecology, 83, 1935-1944.

123

Krombein K.V., H. P. D., Smith D.R., and B.D. Burks. 1979. Catalog of Hymenoptera in America North of Mexico. Washington D.C.: Smithsonian Institution Press.

Laeser, S.R., Baxter, C.V. & Fausch, K.D. (2005) Riparian vegetation loss, stream channelization, and web-weaving spiders in northern Japan. Ecological Research, 20, 646-651.

Lemly, A.D. (2004) Aquatic selenium pollution is a global environmental safety issue. Ecotoxicology and Environmental Safety, 59, 44-56.

Martyniuk, C.J., Feswick, A., Spade, D.J., Kroll, K.J., Barber, D.S. & Denslow, N.D. (2010) Effects of acute dieldrin exposure on neurotransmitters and global gene

124

transcription in largemouth bass (Micropterus salmoides) hypothalamus. Neurotoxicology, 31, 356-366.

McCarty, J. P. (1997) Aquatic community characteristics influence the foraging patterns of tree swallows. Condor, 99, 210-213.

McCarty (2002) Use of tree swallows in studies of environmental stress. Reviews in Toxicology (Amsterdam), 4, 61-104.

McCarty, J. P. & D. W. Winkler (1999) Foraging ecology and diet selectivity of tree swallows feeding nestlings. Condor, 101, 246-254.

Mengelkoch, J.M., Niemi, G.J. & Regal, R.R. (2004) Diet of the nestling Tree Swallow. Condor, 106, 423-429.

Menzie, C. A. (1980) Potential significance of insects in the removal of contaminants from aquatic systems. Water Air and Soil Pollution, 13, 473-479.

Mergler, D., Anderson, H.A., Chan, L.H.M., Mahaffey, K.R., Murray, M., Sakamoto, M. & Stern, A.H. (2007) Methylmercury exposure and health effects in humans: A worldwide concern. Ambio, 36, 3-11.

125

Merna, J.W. (1986) Contamination of stream fishes with chlorinated hydrocarbons from eggs of Great Lakes salmon. Transactions of the American Fisheries Society, 115, 69-74.

Nath, S., H. Anand & P. Haldar (2011) Study of short horned grasshoppers in relation to heavy metals accumulation. Journal of Entomological Research (New Delhi), 35, 291-293.

Ness, J.H., Morris, W.F. & Bronstein, J.L. (2009) For ant-protected plants, the best defense is a hungry offense. Ecology, 90, 2823-2831.

Niemela, J. & D. J. Kotze (2009) Carabid beetle assemblages along urban to rural gradients: A review. Landscape and Urban Planning, 92, 65-71.

Ohio Department of Natural Resources (2000) A Guide to Ohio's Streams. Ohio Chapter American Fisheries Society, Columbus, OH.

Ohio Department of Natural Resources (2011). Olentangy State Scenic River. http://ohiodnr.com/watercraft/sr/tabid/2565/Default.aspx. Accessed December 18, 2011.

Ohio Department of Natural Resources, Division of Forestry (2012) http://www.dnr.state.oh.us/Home/trees/honey_am/tabid/5376/Default.aspx

126