Fertilization and Plant Litter Effects on the Plant and Epigeal Arthropod Communities

Home , Eridantes erigonoides, Monaeses, Myodocha, Myrmoecia, Oncerotrachelus, Pagasa (genus)

FERTILIZATION AND PLANT LITTER EFFECTS ON THE PLANT AND EPIGEAL ARTHROPOD COMMUNITIES

A dissertation submitted to Kent State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

L. Brian Patrick

December 2009

Dissertation written by L. Brian Patrick B.A., University of Kansas, 1998 M.A., University of Kansas, 2001 Ph.D., Kent State University, 2009

______, Chair, Doctoral Dissertation Committee Mark W. Kershner, Ph.D.

______, Members, Doctoral Dissertation Committee James L. Blank, Ph.D.

______Alison J. Smith, Ph.D.

______Lauchlan H. Fraser, Ph.D.

______Randall J. Mitchell, Ph.D.

Accepted by

______, Chair, Department of Biological Sciences James L. Blank, Ph.D.

______, Dean, College of Arts and Sciences John R. D. Stalvey, Ph.D.

TABLE OF CONTENTS

LIST OF FIGURES ……………………………………………………………………..vii

LIST OF TABLES …………………………………………………………………...... x

ACKNOWLEDGMENTS ………………………………………………………………xii

CHAPTER

I. Introduction………………………………………………………………..1 Biodiversity-productivity theory…………………………………..3 Plant litter …………………………………………………………5 Trophic relationships in the epigeal arthropod community ………7 Dissertation organization …………………………………………9 References ……………………………………………………….11

II. Large-scale manipulation of plant litter and fertilizer in a managed successional temperate grassland ………………………………………..23 Abstract…………………………………………………………..23 Introduction………………………………………………………24 Materials and Methods …………………………………………..26 Study site and experimental design ……………………...26 Plant community sampling ……………………………...28 Abiotic sampling ………………………………………...29 Statistical analyses ………………………………………30 Results …………………………………………………………...33 General trends …………………………………………...33 Species-level analyses …………………………………...37 Plant group responses …………………………………...42 Ecosystem response ……………………………………..46 Discussion …………………………………………………….....49 References …………………………………………………….....56

III. Beetle (Coleoptera) responses to fertilization and plant litter in a temperate old-field grassland …………………………………………………….....63 Abstract ……………………………………………………….....63 Introduction ……………………………………………………...64 Materials and methods …………………………………………..67 Study site and experimental design ……………………...67 Plant community sampling ……………………………...69 Beetle community sampling ………………………….....70 Statistical analyses ………………………………………71

iii

Results …………………………………………………………...72 General beetle responses ………………………………...72 Beetle species-level responses …………………………..78 Discussion …………………………………………………….....84 References …………………………………………………….....91 Appendix A ……………………………………………………...98 Appendix B …………………………………………………….106 Appendix C …………………………………………………….109 Appendix D …………………………………………………….110 Appendix E …………………………………………………….111 Appendix F ……………………………………………………..112 Appendix G …………………………………………………….113

IV. Epigeal predator responses to fertilization and plant litter: testing biodiversity theory at the ground level ………………………………...117 Abstract ………………………………………………………...117 Introduction …………………………………………………….118 Materials and methods …………………………………………123 Study site and experimental design …………………….123 Spider community sampling …………………………...125 Statistical analyses ……………………………………..126 Results ………………………………………………………….128 General trends ………………………………………….128 Spider species-level analyses …………………………..139 Aggregate ecosystem-level analyses …………………...146 Discussion ……………………………………………………...150 References ……………………………………………………...159 Appendix A …………………………………………………….168 Appendix B …………………………………………………….171 Appendix C …………………………………………………….172 Appendix D …………………………………………………….173 Appendix E …………………………………………………….174 Appendix F ……………………………………………………..175

V. Review of the Nearctic genus Scyletria Bishop and Crosby (Araneae, Linyphiidae), with a transfer of S. jona to Mermessus O. Pickard- Cambridge ……………………………………………………………...177 Abstract ………………………………………………………...177 Introduction …………………………………………………….177 Materials and methods …………………………………………178 Taxonomy ……………………………………………………...180 Genus Scyletria Bishop and Crosby 1938……………...180 Scyletria inflata Bishop and Crosby 1938 ……………..180

Mermessus jona (Bishop and Crosby 1938) new combination …………………………………………….187 References ……………………………………………………...198 VI. Fertilization effects on predator diversity in a terrestrial detrital food web …………………………………………………………………………..202 Abstract ………………………………………………………...202 Introduction …………………………………………………….203 “Green” world vs. “Brown” world and cascading effects ……...... 206 Materials and methods …………………………………………209 Study site and experimental design …………………….209 Arthropod community sampling ……………………….210 Statistical analyses ……………………………………..212 Results ………………………………………………………….213 Discussion ……………………………………………………...221 References ……………………………………………………...227

VII. Summary and importance of this study ………………………………..237 Future directions ……………………………………………….241 References ……………………………………………………...243

A I. First record of the genus Myrmedonota Cameron (Coleoptera, Staphylinidae) from North America, with descriptions of two new species …………………………………………………………………………..248 Abstract ………………………………………………………...248 Introduction …………………………………………………….248 Materials and methods …………………………………………250 Taxonomy ……………………………………………………...250 Genus Myrmedonota Cameron, 1920 ………………….250 Myrmedonota aidani Maruyama and Klimaszewski, sp. nov. ……………………………………………………..253 Myrmedonota lewisi Maruyama and Klimaszewski, sp. nov. ……………………………………………………..259 Modified key to species of Zyras group of genra in America North of Mexico ………………………………………………………265 References ……………………………………………………...266

A II. Eight new Ohio state records of true bugs (Hemiptera, Heteroptera) from pitfall traps ……………………………………………………………..268 Abstract ………………………………………………………...268 Introduction …………………………………………………….269 Materials and methods …………………………………………269 Collection sites ………………………………………....270 Results and discussion …………………………………………271

Hebridae ………………………………………………..271 Miridae …………………………………………………271 Nabidae ………………………………………………...275 Reduviidae ……………………………………………..275 Rhyparochromidae ……………………………………..279 Thyreocoridae ………………………………………….283 References………………………………………………………285

LIST OF FIGURES

CHAPTER 1. Introduction.

Figure 1. Line graph of the number of journal article titles and “topic” hits containing the word “biodiversity” in a database search of ISI Web of Knowledge …………………………………………………………………………………...... 2

CHAPTER 2. Large-scale manipulation of plant litter and fertilizer in a managed successional temperate grassland

Figure 2. Average plant biomass, Shannon’s H, percent soil organic content and moisture, and PAR ………………………………………………………………35

Figure 3. Ordination diagrams of the first two axes of canonical correspondence analysis (CCA) with the four treatments as environmental dummy variables ….41

Figure 4. Average species richness and average biomass of each growth form plant group within each treatment …………………………………………….....44

Figure 5. Average species richness and average biomass of the native and non- native plant groups within each treatment ………………………………………45

Figure 6. Two-dimensional ordination of ecosystem-level properties from 2005 in 24 experimental plots from nonmetric multidimensional scaling (NMS) ……47

CHAPTER 3. Beetle (Coleoptera) responses to fertilization and plant litter in a temperate old-field grassland

Figure 7. Average effective Shannon’s H’ (eH’) of beetle species in each treatment ………………………………………………………………………...73

Figure 8. Species richness, abundance, and biomass of all beetles, carabids, and staphylinids ……………………………………………………………………...77

Figure 9. Regressions of all beetles, carabids, and staphylinids against abundance, biomass, plant species richness, and standing crop biomass …….....80

Figure 10. Average abundance of selected species by year ………………….....82

CHAPTER 4. Epigeal predator responses to fertilization and plant litter: testing biodiversity theory at the ground level

vii

Figure 11. Average effective Shannon’s H’ (eH’) of spider species in each treatment ……………………………………………………………………….137

Figure 12. Species richness, abundance, and biomass of all spiders, linyphiids, and lycosids …………………………………………………………………….138

Figure 13. Regressions of all spiders, linyphiids, and lycosids against abundance and biomass …………………………………………………………………….140

Figure 14. Regressions of all spiders, linyphiids, and lycosids against plant species richness and standing crop biomass …………………………………...141

Figure 15. Average abundance of selected species by year …………………...143

Figure 16. Two-dimensional ordination of ecosystem-level properties from 2005 from nonmetric multidimensional scaling (NMS) ……………………………..148

CHAPTER 5. Review of the Nearctic genus Scyletria Bishop and Crosby (Araneae, Linyphiidae), with a transfer of S. jona to Mermessus O. Pickard- Cambridge

Figure 17. Scyletria inflata Bishop and Crosby 1938, male palpus…………....185

Figure 18. Mermessus jona (Bishop and Crosby 1938) new combination, male palpus and female epigynum…………………………...... 192

Figure 19. Geographic distribution of Scyletria inflata Bishop and Crosby 1938 and Mermessus jona (Bishop and Crosby 1938) new combination ……………193

CHAPTER 6. Fertilization effects on predator diversity in a terrestrial detrital food web

Figure 20. Abundance, biomass, and species richness of all arthropods, predators, and prey ………………………………………………………………………...216

Figure 21. Regressions of the abundance and species richness against total plant biomass and plant species richness for all arthropods, predators, and prey …...220

APPENDIX 1. First record of the genus Myrmedonota Cameron (Coleoptera, Staphylinidae) from North America, with descriptions of two new species

Figure 22. Habitus of Myrmedonota spp...... 251

Figure 23. Mouthparts of Myrmedonota aidani ...... 254

viii

Figure 24. Terminalia of Myrmedonota aidani ...... 256

Figure 25. Mouthparts of Myrmedonota lewisi ...... 261

Figure 26. Terminalia of Myrmedonota lewisi ...... 263

APPENDIX 2. Eight new Ohio state records of true bugs (Hemiptera, Heteroptera) from pitfall traps

Figure 27. Hebrus burmeisteri distribution north of Mexico …………………272

Figure 28. Pycnoderes obscuratus distribution north of Mexico ……………..273

Figure 29. Pagasa fusca fusca distribution north of Mexico ………………….276

Figure 30. Emesaya brevipennis brevipennis distribution north of Mexico …..277

Figure 31. Oncerotrachelus acuminatus distribution north of Mexico ……….278

Figure 32. Cryphula trimaculata distribution north of Mexico ……………….280

Figure 33. Stygnocoris rusticus distribution north of Mexico ………………...282

Figure 34. Corimelaena pulicaria distribution north of Mexico ……………...284

LIST OF TABLES

CHAPTER 2. Large-scale manipulation of plant litter and fertilizer in a managed successional temperate grassland

Table 1. Results of repeated-measures ANOVA for total biomass and Shannon’s H …………………………………………………………………………………36

Table 2. Proportion of total living biomass within each year for each plant species …………………………………………………………………………...39

Table 3. Results of factorial ANOVA for each of the six plant groups ………...43

Table 4. Results of Multi-response Permutation Procedure (MRPP) on emergent properties for 2005 ………………………………………………………………48

CHAPTER 3. Beetle (Coleoptera) responses to fertilization and plant litter in a temperate old-field grassland

Table 5. Results of the repeated-measure PROC MIXED for each response variable ………………………………………………………………………...... 75

Table 6. Results of the PROC MIXED for each response variable during 2005 ……………………………………………………………………………………76

CHAPTER 4. Epigeal predator responses to fertilization and plant litter: testing biodiversity theory at the ground level

Table 7. Total numbers of each family and species of spider captured during the four year manipulative experiment …………………………………………….131

Table 8. Results of the repeated-measure PROC MIXED for each response variable …………………………………………………………………………135

Table 9. Results of the PROC MIXED for each response variable during 2005 …………………………………………………………………………………..136

Table 10. Results of the repeated-measure PROC MIXED for each species ....145

Table 11. Results of Multi-response Permutation Procedure (MRPP) on emergent properties for 2005 to support NMS analyses ………………………………….151

CHAPTER 6. Fertilization effects on predator diversity in a terrestrial detrital food web

Table 12. Results of the repeated-measures PROC MIXED analyses for each response variable ……………………………………………………………….214

Table 13. Results of the PROC MIXED analyses for each response variable for 2005 …………………………………………………………………………….217

ACKNOWLEDGMENTS

My eternal gratitude goes to my wonderfully patient wife, Traci, and to my two children, Thea and Aidan. I am quite certain that I would have never made it this far in life without their love and support. I cannot even begin to thank my dear wife for all that she has had to endure for the sake of this dissertation, and I have to spend a great deal of time catching up with my kids. However, it was all worth it because we have all endured and finally made it to the end of this arduous process. A simple “thank you” in inadequate, but also a necessity. Traci, Thea, and Aidan, I humbly thank you for all that you have done, are doing, and ever will do for me.

My grandmother, Marjorie Fern Patrick, inspired my love of natural history, while my grandfather, Lucian Greear Patrick, worked the land and helped inspire my appreciation for ecology. My parents, Charles Hiram Patrick and Vicki Jo Patrick, both nurtured my love of the outdoors by taking our family every weekend to the Lake of the

Ozarks to camp, and, occasionally, to the Jack’s Fork river in southern Missouri to camp and canoe. At the time, I don’t think that I knew that I was learning anything, and I certainly didn’t appreciate most of it until later in life. I regret only that neither my parents nor my grandparent lived to see me complete my Ph.D.

I thank my advisors, Mark W. Kershner and Lauchlan H. Fraser, as well as James

L. Blank, and my undergraduate and Master’s advisor Bob Holt. All of you kept your faith in me and encouraged me to keep pecking away at this dissertation. I promised all

xii of you that I would indeed get it done, and now I can finally (and at long, long last) fulfill my promise. Thank you for your support, encouragement, and friendship.

In addition to thanking this next person, I would also like to dedicate this dissertation to him. Richard “Rick” William Bowers III was truly a driving force behind my staying in graduate school and finishing this dissertation. He was my friend, ally, agitator, supporter, partner-in-crime, and spiritual brother. He was plucked from life way too early, and he also did not live to see me finish my Ph.D., though he missed it by only a few months.

I also extend thanks to the innumerable undergrads and grad students, particularly from the Fraser Lab, who helped with various stages of this project. I thank Barb

Andreas for help with plant identifications, Charlie Dondale for mentoring me in the ways of the spiders, Jan Klimaszewski for showing me the ways of the aleocharines,

Yves Bousquet, Pat Bouchard, Serge Leplante, Munetoshi Maruyama, Al Newton, Eric

G. Chapman, Vladimir Gusarov, Ales Smetana, A. Solodovnikov, Anthony Davies, the late Ed C. Becker, and J. H. Frank for help identifying beetles and for providing information pertaining to the biology of several species. I think Steve Chordas for help with the heteropterans, Nadine Dupérré for her help with Scyletria illustrations and descriptions, and probably about a thousand other people that I have forgotten but who very certainly deserve to be in this list!

xiii

CHAPTER 1

INTRODUCTION

Ever since Wilson (1987, 1988) coined the term “biodiversity,” the portmanteau of Lovejoy’s (1980) phrase “biological diversity,” the term has become on one of the most widely used in the field of ecology (Nobis and Wohlgemuth 2004). A search in the

ISI Web of Knowledge database (Thomson Reuters 2009) in the years 1980–2008 for

“biodiversity” by titles only and by topic (searches title, abstract, and author keywords) returned 5,981 journal article titles and 26,566 topic search results (Fig. 1). While the study of diversity has always been a central theme of ecology, it has moved to the forefront because of threats to biodiversity and the emerging knowledge of the effects of biodiversity loss on community and ecosystem properties.

Biodiversity can be affected by a variety of factors, including disturbance (Grime

1979), climate (Nogué et al. 2009), soil fertility (Fraser and Grime 1997, 1998; Fraser

1998; Dybzinski et al. 2008), exotic species invasions (Knops et al. 1999; Kimbro et al.

2009), habitat patch size (Cook et al. 2005; Martinko et al. 2006), anthropogenically mediated nutrient addition (Knops et al. 1999; Haddad et al. 2000, 2001; Tilman et al.

2002a, b; Vitousek et al. 2002; Suding et al. 2005; Patrick et al. 2008a, b), and trophic dynamics (Hutchinson 1959; Hairston et al. 1960; Fretwell 1977; Oksanen et al. 1981;

1 2

Figure 1. Line graph of the number of journal article titles and “topic” hits containing the word “biodiversity” in a database search of ISI Web of Knowledge (Thomson Reuters

2009), showing only years 1987–2008 (no articles found during 1980–1986, and 2009 is an incomplete year at the time of this writing).

700 4500 titles 4000 600 topics

500

s s

3000 c

t o t

400

r r

e 2000

300 e

m u

200 u N 1000 N 100

0 0 1987 1990 1995 2000 2005 2008 Year 3

Fraser and Grime, 1997, 1998; Fraser 1998; Oksanen and Oksanen 2000; Haddad et al.

2000, 2001; Patrick et al. 2008b). Biodiversity can affect ecosystem function (Naeem et al. 1994; Hooper and Vitousek 1997; Hodgson et al. 1998), community and ecosystem resistance and resilience to invasive species (Knops et al. 1999), ecosystem productivity

(Tilman et al. 1996, 1997), nutrient regeneration and recycling (Moore et al. 2004;

Hättenschwiler and Gasser 2005), and ecosystem services (Cummings and Child 2009).

In terrestrial systems, the diversity of arthropods has been shown to be positively correlated to the diversity of plants (Knops et al. 1999; Haddad et al. 2000, 2001;

Schaffers et al. 2008), though the epigeal portion of the arthropod community may not necessarily respond to plant species richness (Patrick et al. 2008b; Schuldt et al. 2008) as it is more closely tied to the detritus-based portion of the food web (Halaj and Wise

2002).

Biodiversity-productivity theory

A major consequence of human activity has been ecosystem eutrophication via anthropogenically induced increases in N fertilization and atmospheric nitrogen deposition (Vitousek et al. 1997; Fenn et al. 2003; Galloway et al. 2003), resulting in significant biodiversity loss and ecosystem function (McCann 2000; Larsen et al. 2005).

manipulated habitats and microcosms (Hector et al. 1999; Tilman et al. 2002a). The resulting species thinning often leads to the rarest species disappearing first and an increase in non-native species (Hector et al. 1999; Tilman et al. 2002a; Suding et al.

2005). A common approach to plant biodiversity studies is to segregate plant species into plant groups (e.g., forbs and graminoids, native and non-native) to test for aggregate responses to treatments (Wilsey and Polley 2006). The rarest species in grassland systems are generally in the functional groups of forbs, non-Poaceae graminoids, and woody plants. Conventional theory (Suding et al. 2005) would predict that these functional groups are then likely to lose the most species as a result of fertilization.

This body of knowledge has coalesced into the biodiversity-productivity theory

(Suding et al. 2005), but it has scarcely been applied beyond plant communities. What is known is that as plant species richness decreases, insect abundance generally increases, particularly herbivorous insect pest species (Elton 1958; Root 1973; Strong et al. 1984), insect species richness decreases (Southwood et al. 1979; Strong et al. 1984; Siemann et al. 1998; Knops et al. 1999; Haddad et al. 2000, 2001; Schaeffers et al. 2008), and fewer predators may be supported by the decreased diversity (Knops et al. 1999; Haddad et al.

2001). As plant species richness increases, the diversity of specialist herbivores is thought to increase because of the higher diversity of plants provides a higher diversity of resources (Southwood et al. 1979; Strong et al. 1984; Siemann et al. 1998; Knops et al.

1999; Haddad 2001), which in turn supports a higher diversity of predators and parasitoids (Hutchinson 1959; Knops et al. 1999; Haddad et al. 2001). However, very little is known of these relationships as they relate to terrestrial nutrient loading. 5

Nitrogen (N) is the primary limiting nutrient in terrestrial systems (Vitousek et al.

2002) and the global N pool is increasing at an alarming rate as a result of human activity

(Vitousek et al. 1997; Fenn et al. 2003; Galloway et al. 2003). This increase in N has been shown to decrease plant species richness while increasing plant biomass (Bobbink et al. 1998; Hector et al. 1999; Tilman et al. 2002a, b), a relationship that has become the basis of the diversity-productivity theory (Suding et al. 2005). This theory has been extended to the animal component of the food web with field experiments that have shown decreased arthropod diversity in plots with augmented N (Knops et al. 1999;

Haddad et al. 2000, 2001; Schaffers et al. 2008).

Plant litter

Plant litter is a fundamental factor affecting plant community structure (Facelli and Pickett 1991; Bosy and Reader 1995; Xiong and Nilsson 1999), and its effects have been investigated in a variety of habitats (for a summary of habitats, see Weltzin et al.

2005). Plant litter influences soil moisture (Hamrick and Lee 1987), acts as a mechanical impediment to germinating seedlings (Facelli 1994), alters light attenuation at the soil surface which affects germination and establishment (Goldberg and Werner 1983; Facelli and Pickett 1991; Weltzin et al. 2005), and provides cover for seed and seedling predators

(Hulme 1996, Edwards and Crawley 1999). In unproductive communities, plant litter may facilitate colonization and ameliorate abiotic conditions (e.g., moisture, soil temperature) necessary for germination and establishment (Hamrick and Lee 1987). In 6

highly productive plant communities, plant litter can increase the dominant plant competitor biomass and litter production, thereby decreasing plant species richness

(Foster and Gross 1997, 1998; Long et al. 2003), whereas removal of plant litter in highly productive communities can moderately increase species richness (Long et al. 2003).

Additionally, plant litter production increases as a result of nitrogen (N) enrichment

(Foster and Gross 1998), which can cause a further reduction in plant species richness

(Carson and Peterson 1990; Foster and Gross 1998; Xiong and Nilsson 1999; Long et al.

2003).

Nutrient loading not only increases plant standing crop biomass, but also plant litter production (Long et al. 2003; Patrick et al. 2008a), which can increase the basal food resource for the detrital community and increase detritivore and epigeal predator abundances (Halaj et al. 2000; Halaj and Wise 2002; Moore et al. 2004). Furthermore, plant litter increases habitat complexity which can also increase arthropod abundance and diversity (Lawton 1983; Strong et al. 1984; Rypstra et al. 1999). However, the detritivore community response can be further complicated because of its sensitivity to litter diversity (Wardle 2002; Moore et al. 2004; Hättenschwiler and Gasser 2005;

Hättenschwiler et al. 2005) and litter quality (Wardle et al. 2006). With nutrient loading decreasing plant diversity (Tilman et al. 2002a; Long et al. 2003; Patrick et al. 2008), litter diversity is also decreased. While more plant litter production could increase detritivore and epigeal predator abundance and biomass (Halaj et al. 2000), a reduction in litter diversity could result in declines in diversity of detritivores and epigeal predators

(Hättenschwiler and Gasser 2005; Wardle 2006), effectively mirroring the aerial 7

community response to nutrient loading. Despite such predictions, only one study has incorporated plant litter effects into biodiversity-productivity theory with specific reference to the epigeal arthropod community (Patrick et al. 2008b). However, that paper did not present data delineating the species-level responses within the epigeal community.

Trophic relationships in the epigeal arthropod community

Most trophic levels (e.g., detritivores, predators) would be expected to exhibit species-level responses to perturbations of terrestrial habitats. Spiders, in particular, are ubiquitous generalist predators that can significantly impact terrestrial food webs (Wise

1993; Wise et al. 1999), and epigeal spiders (e.g., Lycosidae, Linyphiidae) are closely linked to the detritivore community (Wise et al. 1999; Chen and Wise 1999; Wise 2006).

While spider predation affects prey, potentially influencing spider abundance and productivity (Snyder and Wise 2001; Lawrence and Wise 2004; Lensing et al. 2005;

Wise 2004, 2006), the abundance of epigeal spiders is limited ultimately by the abundance of their mainly detritivorous prey via bottom-up forces through the detritus- based portion of the food web (Chen and Wise 1999; Wise et al. 1999; Wise 2004, 2006).

Increasing plant detritus has been shown to increase spider abundance by increasing the quantity of food available to their detritivorous prey (Chen and Wise 1999, Wise et al.

1999; Wise 2004), and by increasing habitat structure for hiding and web building (Uetz

1979, 1991; Rypstra et al. 1999). Such an increase in spider abundance has also been 8

shown to moderately increase the local richness of the spider community (Rypstra et al.

1999).

Even though spider abundance may increase, spider diversity may not increase proportionally because the reduced diversity of plant detritus can limit predator diversity in the detrital food web (Hättenschwiler and Gasser 2005; Wardle 2006). Thus, it is reasonable to expect that predators dependent upon the detritivore food web may have the same response to fertilization as predators more closely associated with the aerial food web. Because plant species richness decreases in mesotrophic to eutrophic systems

(Suding et al. 2005; Patrick et al. 2008a), plant litter diversity also decreases. Even though more plant litter is made, increasing the resource base of the detritivore food web, lower litter diversity begets lower detritivore and detritivore-predator diversity

(Hättenschwiler and Gasser 2005; Wardle 2006). Interestingly, no epigeal spider studies

(that focused strictly on cursorial spiders, e.g., Lycosidae; wolf spiders) have looked at spider diversity response to basal resource manipulation. Moreover, no studies have examined responses of the predominantly epigeal spider family Linyphiidae (wandering sheet/tangle-web builders) that may patrol multiple webs at the ground level (Uetz et al.

1999).

Beetles also are good candidates for assessing the effects of fertilization and plant litter on the epigeal arthropod community. Epigeal beetles (e.g., Carabidae and

Staphylinidae) have been used as indicators of changes in habitat quality because they are sensitive to environmental change (Buse and Good 1993; Loreau 1993; Jonas et al. 2002;

Pohl et al. 2007) and are adequately speciose to examine community responses to 9

experimental manipulations (Loreau 1993). Moreover, detritivores are a significant portion of the diet of carabids (Sunderland 1975; Chiverton and Sotherton 1991), and previous studies (e.g., Snyder and Wise 2001) have reported that carabids can alter trophic cascades based on the abundance of their plant detritus-feeding prey. Ultimately, epigeal beetles are largely separated from the living-plant based portion of the food web and can affect the detritus-based portion of the food web, potentially altering nutrient regeneration into the ecosystem.

Dissertation organization

My dissertation focuses on the effects of fertilization and plant litter on the plant and epigeal arthropod communities in an old-field grassland in northeast Ohio. The dissertation is organized into six main research chapters focusing on the plant, beetle, spider, and overall epigeal arthropod communities to test the efficacy of the biodiversity- productivity hypothesis on established plant communities and on the epigeal arthropod community. Chapter 2 focuses on the plant community, testing the biodiversity- productivity theory with an established plant community, and incorporating plant litter effects into the theory predictions. In Chapters 3 and 4, I examine the epigeal beetle and spider communities, respectively, to determine whether these indicator taxa respond as predicted to biodiversity-productivity theory. Understanding biodiversity means that we must also be able to identify species, and Chapter 5 is a revision of a smaller genus of spiders, Scyletria (Araneae, Linyphiidae), a species of which was captured during the 10

course of my study. Moved to another genus in Chapter 5, the species captured

(Mermessus jona Bishop and Crosby) was only known for the male, and the first described females of this species were captured in pitfall traps during the course of my study. Chapter 6 is a synthesis of the plant and epigeal arthropod data to a community analyses in which specific hypotheses of predator-prey interactions are examined. As the last chapter, Chapter 7 is a brief summary of my dissertation work from Chapters 2–4 and

Chapter 7 to explain how biodiversity-productivity does not adequately include the epigeal arthropod community into its predictions.

References

Bobbink, R., M. Hornung, and J. G. M. Roelofs. 1998. The effects of air-borne nitrogen

pollutants on species diversity in natural and semi-natural European vegetation.

Journal of Ecology 86:717–738.

Bosy, J. L., and R. J. Reader. 1995. Mechanisms underlying the suppression of forb

seedling emergence by grass (Poa pratensis) litter. Functional Ecology 9:635–

639.

Buse, A., and J. E. G. Good. 1993. The effects of conifer forest design and management

on abundance and diversity of rove beetles (Coleoptera: Staphylinidae):

implications for conservation. Biological Conservation 64:67–76.

Carson, W. P., and C. J. Peterson. 1990. The role of litter in an old-field plant

community: impact of litter quantity in different seasons on plant species richness

and abundance. Oecologia 85:8–13.

Chen, B., and D. H. Wise. 1999. Bottom-up limitation of predaceous arthropods in a

detritus-based terrestrial food web. Ecology 80:761–772.

Chiverton, P. A., and N. W. Sotherton. 1991. The effect on beneficial arthropods of the

exclusion of herbicides from cereal crop edges. Journal of Applied Ecology

28:1027–1039.

Cook, W. M., J. Yao, B. L. Foster, R. D. Holt, and L. B. Patrick. 2005. Secondary

succession in an experimentally fragmented landscape: community patterns across

space and time. Ecology 86:1267–1279. 12

Cummings, G. S., and M. F. Child. 2009. Contrasting spatial patterns of taxonomic and

functional richness offers insights into potential loss of ecosystem services.

Philosophical Transactions of the Royal Society B – Biological Sciences

364:1683–1692.

Dybzinski, R., J. E. Fargione, D. R. Zak, D. Fornaro, and D. Tilman. 2008. Soil fertility

increases with plant species diversity in a long-term biodiversity experiment.

Oecologia 158:85–93.

Edwards, G. R., and M. J. Crawley. 1999. Herbivores, seed banks and seedling

recruitment in mesic grassland. Journal of Ecology 87:423–435.

Elton, C. S. 1958. The Ecology of Invasions by Animals and Plants. Methuen, Landon,

UK.

Facelli, J. M. 1994. Multiple indirect effects of plant litter affect the establishment of

woody seedlings in old fields. Ecology 75:1727–1735.

Facelli, J. M., and S. T. A. Pickett. 1991. Plant litter: light interception and effects on an

old-field plant community. Ecology 72:1024–1031.

Fenn, M. E., J. S. Baron, E. B. Allen, H. M. Rueth, K. R. Nydick, L. Geiser, W. D.

Bowman, J. O. Sickman, T. Meixner, D. W. Johnson, and P. Neitlich. 2003.

Ecological effects of nitrogen deposition in the western United States.

BioScience 53:404–420.

Foster, B. L., and K. L. Gross. 1997. Partitioning the effects of plant biomass and litter

on Andropogon gerardii in old-field vegetation. Ecology 78:2091–2104. 13

Foster, B. L., and K. L. Gross. 1998. Species richness in a successional grassland:

effects of nitrogen enrichment and plant litter. Ecology 79:2593–2602.

Fraser, L. H. 1998. Top-down vs bottom-up control influenced by productivity in a

North Derbyshire, UK, dale. Oikos 81:99–108.

Fraser, L. H., and J. P. Grime. 1997. Primary productivity and trophic dynamics

investigated in a North Derbyshire, UK, dale. Oikos 80:499–508.

Fraser, L. H., and J. P. Grime. 1998. Top-down control and its effects on the biomass

and composition of three grasses at high and low soil fertility in outdoor

microcosms. Oecologia 113:239–246.

Fretwell, S. D. 1977. The regulation of plant communities by food chains exploiting

them. Perspectives of Biology and Medicine 20:169–185.

Galloway, J. N., J. D. Aber, J. W. Erisman, S. P. Seitzinger, R. W. Howarth, E. B.

Cowling, and B. J. Cosby. 2003. The nitrogen cascade. BioScience 53:341–355.

Goldberg, D. E., and P. A. Werner. 1983. The effects of size of opening in vegetation

and litter cover on seedling establishment of goldenrods (Solidago spp.).

Oecologia 60:149–155.

Grime, J. P. 1979. Plant Strategies and Vegetation Processes. John Wiley and Sons,

London, UK.

Haddad, N. M., J. Haarstad, and D. Tilman. 2000. The effects of long-term nitrogen

loading on grassland insect communities. Oecologia 124:73–84. 14

Haddad, N. M., D. Tilman, J. Harstaad, M. Ritchie, and J. M. H. Knops. 2001.

Contrasting effects of plant richness and composition on insect communities: a

field experiment. American Naturalist 158:17–35.

Hairston, N. G., F. E. Smith, and L. D. Slobodkin. 1960. Community structure,

population control, and competition. American Naturalist 94:421–425.

Halaj, J., A. B. Cady, and G. W. Uetz. 2000. Modular habitat refugia enhance generalist

predators and lower plant damage in soybeans. Environmental Entomology

29:383–393.

Halaj, J. and D. H. Wise. 2002. Impact of a detrital subsidy on trophic cascades in a

terrestrial grazing web. Ecology 83:3141–3151.

Hamrick, J. L., and J. M. Lee. 1987. Effect of soil surface topography and litter cover on

the germination, survival, and growth of musk thistle (Carduus nutans).

American Journal of Botany 74:451–457.

Hättenschwiler, S., and P. Gasser. 2005. Soil animals alter plant litter diversity effects

on decomposition. Proceedings of the National Academy of Sciences of the USA

102:1519–1524.

Hättenschwiler, S., A. V. Tiunov, and S. Scheu. 2005. Biodiversity and litter

decomposition in terrestrial ecosystems. Annual Review of Ecology, Evolution

and Systematics 36:191–218.

Hector, A., B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G.

Dimitrkopoulos, J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P.

Högberg, K. Huss-Danell, J. Joshi, A. Jumpponen, C. Körner, P. W. Leadley, M. 15

Loreau, A. Minns, C. P. H. Mulder, G. O’Donovan, S. J. Otway, J. S. Pereira, A.

Prinz, D. J. Read, M. Scherer-Lorenzen, E.-D. Schulze, A.-S. D.Siamantziouras,

E. M. Spehn, A. C. Terry, A. Y. Troumbis, F. I. Woodward, S. Yachi, and J. H.

Lawton. 1999. Plant diversity and productivity experiments in European

grasslands. Science 286:1123–1127.

Hodgson, J. G., K. Thompson, P. J. Wilson, and A. Bogaard. 1998. Does biodiversity

determing ecosystem function? The Ecotron experiment reconsidered. Functional

Ecology 12:843–848.

Hooper, D. U., and P. M. Vitousek. 1997. The effects of plant composition and diversity

on ecosystem processes. Science 277:1302–1305.

Hulme, P. E. 1996. Herbivores and the performance of grassland plants: a comparison

of arthropod, mollusk and rodent herbivory. Journal of Ecology 84:43–51.

Hutchinson, G. E. 1959. Homage to Santa Rosalia or why are there so many kinds of

animals? American Naturalist 93:145–159.

Jonas, J. L., M. R. Whiles, and R. E. Charlton. 2002. Aboveground invertebrate

responses to land management differences in a central Kansas grassland.

Environmental Entomology 31:1142–1152.

Kimbro, D. L., E. D. Grosholz, A. J. Baukus, N. J. Nesbitt, N. M. Travis, S. Attoe, and C.

Coleman-Hulbert. 2009. Invasive species cause large-scale loss of native

California oyster habitat by disrupting trophic cascades. Oecologia 160:563–575.

Knops, J. M. H., D. Tilman, N. M. Haddad, S. Naeem, C. E. Mitchell, J. Harstaad, M. E.

Ritchie, K. M. Howe, P. B. Reich, E. Siemann, and J. Groth. 1999. Effects of 16

plant species richness on invasion dynamics, disease outbreaks, insect

abundances, and diversity. Ecology Letters 2:286–293.

Larsen, T. H., N. Williams, and C. Kremen. 2005. Extinction order and altered

community structure rapidly disrupt ecosystem functioning. Ecology Letters

8:538–547.

Lawrence, K. L., and D. H. Wise. 2004. Unexpected indirect effects of spiders on the

rate of litter disappearance in a deciduous forest. Pedobiologia 48:149–157.

Lawton, J. H. 1983. Plant architecture and the diversity of phytophagous insects.

Annual Review of Entomology 28:23–29.

Long, Z. T., C. L. Mohler, and W. P. Carson. 2003. Extending the resource

concentration hypothesis to plant communities: effects of litter and herbivores.

Ecology 84:652–665.

Loreau, M. 1993. Ground beetles in a changing environment: determinants of species

diversity and community assembly. In: Boyle, T. J. B., and C. E. B. Boyle, eds.,

Biodiversity, Temperate Ecosystems, and Global Change. Springer-Verlag,

Berlin, Germany, pp. 77–98.

Lovejoy, T. E. 1980. Forward. In: Soulé, M. E., and B. A. Wilcox, eds., Conservation

Biology—an evolutionary-ecological perspective. Sinauer Associates, Inc.,

Sunderland, MA, USA.

Martinko, E. A., R. H. Hagen, and J. A. Griffith. 2006. Successional change in the insect

community of a fragmented landscape. Landscape Ecology 21:711–721.

McCann, K. S. 2000. The diversity-stability debate. Nature 405:228–233. 17

Moore, J. C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C.

Johnson, K. S. McCann, K. Melville, P. J. Morin, A. D. Rosemond, D. M. Post, J.

L. Sabo, K. M. Scow, M. J. Vanni, and D. H. Wall. 2004. Detritus, trophic

dynamics, and biodiversity. Ecology Letters, 7:584–600.

Naeem, S., L. J. Thompson, S. P. Lawler, J. H. Lawton, and R. M. Woodfin. 1994.

Declining biodiversity can alter the performance of ecosystems. Nature 368:734–

737.

Nobis, M., and T. Wohlgemuth. 2004. Trend words in ecological core journals over the

last 25 years (1978–2002). Oikos 106:411–421.

Nogué, S., V. Rull, and T. Vegas-Vilarrúbia. 2009. Modeling biodiversity loss by global

warming on Pantepui, northern South America: projected upward migration and

potential habitat loss. Climate Change 94:77–85.

Oksanen, L., S. D. Fretwell, J. Arruda, and P. Niemelä. 1981. Exploitation ecosystems

in gradients of primary productivity. American Naturalist 118:240–261.

Oksanen, L., and T. Oksanen. 2000. The logic and realism of the hypothesis of

exploitation ecosystems. American Naturalist 155:703–723.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008a. Large-scale manipulation of

plant litter and fertilizer in a managed successional temperate grassland. Plant

Ecology 197:183–195.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008b. “Brown” world invertebrates

contradict “green” world biodiversity theory. Research Letters in Ecology, vol.

2008, Article ID 694638, 4 pages, doi:10.1155/2008/694638. 18

Pohl, G. R., D. W. Langor, and J. R. Spence. 2007. Rove beetles and ground beetles

(Coleoptera: Staphylinidae, Carabidae) as indicators of harvest and regeneration

practices in western Canadian foothills forests. Biological Conservation 137:294–

307.

Root, R. B. 1973. Organization of plant-arthropod association in simple and diverse

habitats: the fauna of collards (Brassica oleraceae). Ecological Monographs

43:95–124.

Rypstra, A. L., P. E. Carter, R. A. Balfour, and S. D. Marshall. 1999. Architectural

modifications of agricultural habitats and their impact on the spider inhabitants.

Journal of Arachnology 27:371–377.

Schaffers, A. P., I. P. Raemakers, K. V. Sykora, and C. J. F. Ter Braak. 2008. Arthropod

assemblages are best predicted by plant species composition. Ecology 89:782–

794.

Schuldt, A., N. Fahrenholz, M. Brauns, S. Migge-Kleian, C. Platner, and M. Schaefer.

2008. Communities of ground-living spiders in deciduous forests: does tree

species diversity matter? Biodiversity Conservation 17:1267–1284.

Siemann, E., D. Tilman, J. Haarstad, and M. Ritchie. 1998. Experimental tests of the

dependence of arthropod diversity on plant diversity. The American Naturalist

152:738–750.

Snyder, W. E., and D. H. Wise. 2001. Contrasting trophic cascades generated by a

community of generalist predators. Ecology 86:1571–1583. 19

Southwood, T. R. E., V. K. Brown, and P. M. Reader. 1979. The relationships of plant

and insect diversities in succession. Biological Journal of the Linnean Society

12:327–348.

Strong, D. R., J. H. Lawton, and R. Southwood. 1984. Insects on plants: community

patterns and mechanisms. Harvard University Press, Cambridge, MA, USA.

Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.

Milchunas, and S. Pennings. 2005. Functional- and abundance-based

mechanisms explain diversity loss due to N fertilization. Proceedings of the

National Academy of Sciences of the USA 102:4387–4392.

Sunderland, K. D. 1975. The diet of some predatory arthropods in cereal crops. Journal

of Applied Ecology 12:507–515.

Thomson Reuters. 2009. ISI Web of Knowledge. http://www.isiknowledge.com.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002a. Plant diversity and composition:

effects on productivity and nutrient dynamics of experimental grasslands. In:

Loreau, M., S. Naeem, and P. Inchausti, eds., Biodiversity and Ecosystem

Functioning. Synthesis and Perspectives. Oxford University Press, Oxford, UK,

pp. 21–35.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002b. Experimental and observational

studies of diversity, productivity, and stability. In: Kinzig, A. P., S. W. Pacala,

and D. Tilman, eds., The Functional Consequences of Biodiversity. Princeton

University Press, Princeton, NJ, USA, pp. 42–70. 20

Tilman, D., C. L. Lehmen, and K. T. Thompson. 1997. Plant diversity and ecosystem

productivity: theoretical considerations. Proceedings of the National Academy of

Sciences of the USA 94:1857–1861.

Tilman, D., D. Wedin, and J. Knops. 1996. Productivity and sustainability influenced by

biodiversity in grassland ecosystems. Nature 379:718–720.

Uetz, G. W. 1979. The influence of variation in litter habitats on spider communities.

Oecologia 40:29–42.

Uetz, G. W. 1991. Habitat structure and spider foraging. In: Bell, S. S., E. D. McCoy,

and H. R. Mushinsky, eds., Habitat structure: the physical arrangement of

objects in space. Chapman and Hall, London, UK, pp. 325–348.

Uetz, G. W., J. Halaj, and A. B. Cady. 1999. Guild structure of spiders in major crops.

Journal of Arachnology 27:270–280.

Vitousek, P. M., S. Hättenschwiler, L. Olander, and S. Allison. 2002. Nitrogen and

nature. Abio 31:97–101.

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human

domination of Earth’s ecosystems. Science 277:494–499.

Wardle, D. A. 2002. Communities and Ecosystems: Linking the Aboveground and

Belowground Components. Princeton University Press, Princeton, NJ, USA.

Wardle, D. A. 2006. The influence of biotic interactions on soil biodiversity. Ecology

Letters 9:870–886. 21

Wardle, D. A., G. W. Yeates, G. M. Barker, and K. I. Bonner. 2006. The influence of

plant litter diversity on decomposer abundance and diversity. Soil Biology and

Biochemistry 38:1052–1062.

Weltzin, J. F., J. K. Keller, S. D. Bridgham, J. Pastor, P. B. Allen, and J. Chen. 2005.

Litter controls plant community composition in a northern fen. Oikos 110:537–

546.

Wilsey, B. J., and H. W. Polley. 2006. Aboveground productivity and root-shoot

allocation differ between native and introduced grass species. Oecologia

150:300–309.

Wilson, E. O. 1987. An urgent need to map biodiversity. Scientist 1:11–11.

Wilson, E. O., editor. 1988. Biodiversity. National Academy Press, Washington, D. C.,

USA.

Wise, D. H. 1993. Spiders in Ecological Webs. Cambridge University Press,

Cambridge, UK.

Wise, D. H. 2004. Wandering spiders limit densities of a major microbe-detritivore in

the forest-floor food web. Pedobiologia 48:181–188.

Wise, D. H. 2006. Cannibalism, food limitation, intraspecific competition, and the

regulation of spider populations. Annual Review of Entomology 51:441–465.

Wise, D. H., W. E. Snyder, P. Tuntibunpakul and J. Halaj. 1999. Spiders in

decomposition food webs of agroecosystems: theory and evidence. Journal of

Arachnology 27:363–370. 22

Xiong, S., and C. Nilsson. 1999. The effects of plant litter on vegetation: a meta-

analysis. Journal of Ecology 87:984–994.

CHAPTER 2

LARGE-SCALE MANIPULATION OF PLANT LITTER AND FERTILIZER IN

A MANAGED SUCCESSIONAL TEMPERATE GRASSLAND

(This chapter was published August 2008 in Plant Ecology)

Abstract

Plant litter may play an important role in herbaceous plant communities by limiting primary production and influencing plant species richness. However, it is not known how the effect of litter interacts with fertilization. I tested for the role of litter and fertilization in a large-scale experiment to investigate effects on diversity and biomass of plant species, growth forms, native vs. non native groups, and abiotic ecosystem components (e.g., soil moisture, PAR). I manipulated plant litter (removed vs. left in situ) and nutrient availability (NPK-fertilized vs. unfertilized) for four years in 314-m2 plots, replicated six times, in an old-field grassland. While many of my species-level results supported previously published studies and theory, my plant group results generally did not. Specifically, grass species richness and forb biomass was not affected by either fertilization or plant litter. Moreover, plant litter removal significantly

23 24

increased non-native plant species richness. Relative to native plant species, all of my experimental manipulations significantly increased both the biomass and the species richness of non-native plant species. Thus, this grassland system was sensitive to management treatments through the facilitation of non-native plant species. I coupled biotic and abiotic components within a nonmetric multidimensional scaling (NMS) analysis to investigate treatment effects, which revealed that specific treatments altered ecosystem development. These results suggest that fertilization and plant litter may have larger impacts on plant communities and on ecosystem properties than previously understood, underscoring the need for larger-scale and longer-term experiments.

Introduction

highly productive plant communities, plant litter can increase the dominant plant competitor biomass and litter production, thereby decreasing plant species richness

(Foster and Gross 1997, 1998; Long et al. 2003), whereas removal of plant litter in highly productive communities can moderately increase species richness (Long et al. 2003).

Additionally, plant litter production increases as a result of nitrogen (N) enrichment

(Foster and Gross 1998), which can cause a further reduction in plant species richness

(Carson and Peterson 1990; Foster and Gross 1998; Xiong and Nilsson 1999; Long et al.

2003).

The effects of increased N on terrestrial plant communities are well documented, showing declines in plant species richness while augmenting plant biomass across a variety of natural and semi-natural habitats (Bobbink et al. 1998) and experimentally manipulated habitats and microcosms (Hector et al. 1999; Tilman et al. 2002a). The resulting species thinning often leads to the rarest species disappearing first and an increase in non-native species (Hector et al. 1999; Tilman et al. 2002a; Suding et al.

woody plants. Conventional theory (Suding et al. 2005) would predict that these functional groups are then likely to lose the most species as a result of fertilization, though the interacting effects of plant litter and fertilization on plant groups within a naturalized plant community remain largely unexplored. Since litter removal should result in greater light availability at the soil surface, a reduction in mechanical impediment, and fewer seed predators, these conditions may promote seedling emergence and increased plant species richness.

Here I report the results of the first four years of an ongoing field study to investigate the effects of the experimental manipulation of NPK fertilization and plant litter on plant community structure, plant group responses, and abiotic factors; as well as the coupling of biotic (e.g., plant biomass, plant species richness) and abiotic (e.g., PAR, soil moisture, soil organic content) factors. I explore three hypotheses: 1) plant biomass will increase and plant diversity will be reduced (even within growth forms and native vs. non-native groups), 2) plant litter will negatively affect plant species diversity irrespective of fertilization, though the effect will be greater in fertilized plots, and 3) plant litter and fertilization will interact to alter plant community structure and ecosystem properties.

Materials and methods

Study site and experimental design

The study was conducted at the 163.5 ha Bath Nature Preserve (BNP; 41° 10’

36.2” N, 81° 38’ 58.7” W), Bath Township, Summit County, Ohio, USA, in a 16 ha section of an upland former hay meadow. Until 1996, the study site was a hay meadow, harvested one or many times per year. From 1997 through the present, the area has been mown annually by the local township in late August to early September, near or at the end of the growing season, and the mown vegetation has been left on the field. The vegetation is an herbaceous, graminoid community largely dominated by cool-season grasses, e.g., Bromus inermis Leyss., Festuca arundinacea Schreb., Phleum pratense L., and Anthoxanthum odoratum L (Gleason and Cronquist 1991). Aboveground biomass samples taken before the start of experimental manipulations ranged from approximately

500-1000 g m-2 (unpublished data, L.B.P.; mean = 702 g m-2, n = 24, SD = 127), placing it as moderately productive relative to other grassland sites across the U.S. (Sala et al.

1988) and within the upper Midwest (Foster and Gross 1998). The dominant soil type is

Ellsworth silt loam (ElB), which consists of moderately well drained, moderately deep to deep soils formed in silty clay loam or clay loam glacial till of the Wisconsin Age

(Ritchie and Steiger 1974). Soil nutrient analyses conducted before the start of the experiment placed total nitrogen at 0.19 ± 0.02% (n=24) and total phosphorus at 11.2 ±

8.3 kg/ha (n=24), indicating moderate productivity.

In August 2001, twenty-four 20-m diameter circular plots (314 m2) were established. These experimental plots were separated by at least 20 m and were at least

30 m away from any other habitat (e.g., roads, forest). Due to the size of the plots, proximity constraints, and the size of the fields available, plots covered two adjacent 28

fields separated by a rarely used, restricted access single-lane dirt road, resulting in 12 plots in each field. Treatments were applied in a 2 x 2 factorial design of fertilizer (+F = fertilizer added, –F = no fertilizer) and plant litter (-L = litter removed, +L = litter left in situ after yearly mowing) with the control plots characterized as no fertilization and plant litter left in situ (+L/-F), resulting in six replicates per treatment. In April 2002 and continuing each April through 2005, Scotts brand Osmocote 8-9 month Slow Release

Fertilizer 19-6-12 (NPK; Scotts, Marysville, OH, USA) was applied at 200 kg N ha-1 (20 g N m-2) in fertilized plots, well above the Köchy and Wilson (2005) 15 g N m-2 yr-1 threshold necessary to induce a eutrophication effect in grasslands and other habitats. I could not exclude ambient wet/dry atmospheric N deposition, though deposition rates from 1990 to 2005 were relatively low at approximately 10.1 kg N ha-1 yr-1 (1.01 g N m-2 yr-1) at a nearby monitoring site in Lykens (162 km west of my study site), OH, USA, and approximately 9.3 kg N ha-1 yr-1 (0.93 g N m-2 yr-1) at another nearby monitoring site in

Mercer Co. (G. K. Goddard site; 96 km east of my study site), PA, USA (US EPA 2005).

Within two days of annual mowing of the whole site by the local township with a large tractor and brush hog mower (autumn 2001-2004), litter was removed from litter removal treatments using a small 23 hp lawn tractor with a pull-behind 8 hp Agri-Fab Mow-N-

Vac trailer attachment (Agri-Fab, Sullivan, IL, USA).

Plant community sampling

During the second week of August 2002, total plant biomass was sampled in 0.25 m2 quadrats from three randomly chosen locations within each plot, one sample per third of each circular plot (n = 72). Hereafter, “total biomass” refers to standing crop biomass and plant litter biomass together, where all references to “litter” refer to the previous year’s mown vegetation and any vegetation senesced and found within the sampling quadrat after standing crop removal. During the second week of August 2003, total plant biomass samples were again taken in a 0.25 m2 quadrat, though only one randomly chosen sample per plot was taken (n = 24 biomass samples). For 2004 and 2005, standing crop biomass (also referred to as “living biomass”) samples were sorted to species in the field as samples were clipped, with plant litter (also referred to as “litter biomass”) also collected in each quadrat after the standing crop biomass was removed.

Sampling started near the end of the growing season in mid to late July and was completed in early August. For 2004, one standing crop biomass sample was collected from a randomly chosen location within each plot (n = 24). For 2005, three standing crop biomass samples were collected from randomly chosen locations with each plot, one sample per third of each circular plot (n = 72). Each year, collected plant biomass samples were returned to the lab and stored at approximately 6°C until they could be dried at 70°C for 72 hrs, then biomass determined.

Abiotic sampling

On 9 August 2005, five measurements of photosynthetically active radiation

(PAR) were taken at the soil surface every 3 m along a north-south transect through the center of each plot (n = 120 PAR measurements) with a LI-COR LI-190SA Quantum

Sensor averaging µmol s-1 m-2 for 30 s and logged to a LI-COR model LI-1400 data logger. Four days after the last rain event, on 10 August 2005, 2-cm diameter soil plugs were taken to a depth of approximately 20 cm for the same locations at which PAR was measured (n = 120 soil plugs). The wet mass of each soil plug was recorded, and then soil plugs were frozen until they could be dried at 70°C for 10 days. Dry mass was recorded for each soil plug to determine percent soil moisture, then each soil plug was heated at 550°C for four hours, cooled in a desiccator to room temperature, and then mass was recorded to determine percent soil organic content.

Statistical analyses

To analyze trends in all biotic and abiotic response variables among treatments and between years, I used SAS software Version 8.01 (SAS Institute Inc. 1999) to calculate maximum likelihood to generate approximate F-tests in PROC MIXED with

Type III effects based upon the covariance structure of compound symmetry. The various models used the different response variables (total biomass, Shannon’s H diversity, PAR, soil moisture, soil organic content), and predictors used fertilized vs. unfertilized, litter removed vs. litter left in situ, year, and their fully factorial interactions, 31

with year as the repeated predictor. To account for uneven sample sizes among years, the response variables were averaged for each plot each year (n = 24)

Species were classified by growth form: forbs, grasses (Poaceae), non-Poaceae graminoids (i.e., Juncaceae, Cyperaceae), and woody plants. Plants were also classified native and non-native. The classification into “native” or “non-native” was made with

Andreas et al. (2004) and refers to plant species that are native or non-native to Ohio,

USA. To calculate species richness for each group within a plot, each species within a plot was counted only once, even if that species was sampled more than once within a plot, then I summed the total number of species within each group within a plot, yielding n = 24 samples in each group. To calculate the biomass of each group within a plot, I averaged the total biomass of each species across each sampling replicate within a plot

(including zeroes for species not sampled in all sampling locations within a plot), then I summed the total biomass of species within each group within a plot, yielding n = 24 samples in each group. For each of the six groups, I used SAS (SAS Institute Inc. 1999) to calculate maximum likelihood in PROC MIXED with Type III effects, I first used the species richness, then the biomass of each of the six groups as response variables with fertilized vs. unfertilized, and litter removed vs. litter left in situ as the predictor variables. Because the 2004 plant sampling data show similar trends in species abundances and distributions relative to the 2005 plant sampling data, I chose to include only analyses for the 2005 plant sampling period for brevity.

I applied canonical correspondence analysis (CCA) using PC-ORD Version 4.37

(McCune and Mefford 1999) to assess treatment effects in species distributions for 2005. 32

The biotic data used in the analysis were the average biomass of each species within each of the 24 plots, while the environmental data in the analysis were the four treatments as dummy variables. The resulting matrix had the average individual species biomass within a plot for the columns and 24 rows (plots). The CCA used Biplot scaling optimized to species.

To assess treatment effects on aggregate ecosystem properties (properties with both the biotic and abiotic components), I applied nonmetric multidimensional scaling

(NMS; Kruskal 1964) using PC-ORD (McCune and Mefford 1999). For 2005, variables used for each of the 24 plots were average species richness per plot, average standing crop biomass, average litter biomass, average PAR per plot, average percent soil moisture per plot, and average percent soil organic content per plot, resulting in a matrix with six columns and 24 rows (plots). Because 1) NMS is scale sensitive, 2) these variables are on radically different measurement scales, and 3) variables have an enormous range of values between variables, data were transformed to proportions relative to the highest value for each variable (i.e., each value in a column was divided by the largest value in that column, creating a range from 0 – 1 for each column). The NMS analysis was run with Sørensen distance, time as the random seed for the starting configuration, 9999 runs stepping down from 6 to 1 dimensions with the real data, 999 Monte Carlo runs to assess the probability of a similar final stress obtained by chance, and a 0.005 stability criterion.

Additionally for 2005, I used PC-ORD (McCune and Mefford 1999) to run the multi- response permutation procedure (MRPP; Mielke 1984) to test for the hypothesis of no difference among treatments. The MRPP used Sørensen distance with the four 33

treatments as the a priori groupings, resulting in a matrix with six columns (biotic and abiotic variables) and 24 rows (plots) and was calculated with all four treatments together, and for pairwise comparisons between treatments to test for the strength of difference between individual treatments.

Results

General trends

Fertilization and, to a lesser extent, the presence of plant litter significantly increased total plant biomass (Fig. 2a). The repeated PROC MIXED (Table 1) indicated a significant time effect (Year) interacting with fertilization, with fertilization steadily increasing total plant biomass through time (Fig. 2a), though Year by itself had no significant effect. However, the non-significant interaction between fertilization and plant litter (Table 1) indicated that fertilization and plant litter had independent effects on total biomass production. Similarly, fertilization had a significant effect on Shannon’s H

(Fig. 2b), with litter treatment marginally significant (Fig. 2b). In contrast to total plant biomass, Year interactions did not affect Shannon’s H (Table 1).

As expected, 2005 plant litter as a response variable (i.e., litter biomass) was significantly affected by the experimental treatments (F3, 68 = 19.10; P < 0.0001), with mean values and standard errors in parentheses as follows: + litter/- fertilizer 44.26

Fig. 2. A) Average total biomass (g) per 1 m2 within each treatment in each year (2002-

2005). B) Average Shannon’s H per 0.25 m2 within each treatment in each year (2004-

2005). C) Percent Soil Moisture (SM) and percent Soil Organic Content (SOC) within each treatment for 2005. D) PAR in each treatment for 2005. Full results of the repeated-measures PROC MIXED for (a) and (b) are in Table 1. All error bars are ±1

SE. +L indicates litter left in situ, -L indicates litter removed, +F indicates fertilization, -

F indicates no fertilization. 35

A) B) 2002 2.0 2004 1200 2003

² 2005

2004 m

1000 2005 1

H 1.5

s '

e 800

) n

g 1.0

600 n

(

h s

400 S

a 0.5

m o

i 200 B 0 0.0 +L/-F -L/-F +L/+F -L/+F +L/-F -L/-F +L/+F -L/+F Treatment Treatment C) D)

800

t SM

40 ) n

SOC ² 700

d e

n ·

a 600

30 s

r 500

n s

i 400

a m

o 20

μ r

( 300

i l

R i

o 200 A

o 10

100

% % 0 0 +L/-F -L/-F +L/+F -L/+F +L/-F -L/-F +L/+F -L/+F Treatment Treatment

Table 1. Results of repeated-measures ANOVA for total biomass and Shannon’s H.

A significant “Fertilized” effect indicates a significant difference in plots either fertilized (+F; 20 g N m-2 added) or unfertilized (-F), a significant “Litter” effect indicates a significant difference in plots where plant litter was either removed (-L) or left in situ (+L), and a significant “Year” effect indicates a significant difference among sampling years, and Fertilized*Litter, Year*Fertilized, Year*Litter, and

Year*Fertilized*Litter represent their fully factorial interactions. P-values significant at α = 0.05 are denoted by bolding.

df Numerator df Denominator F P

Total Biomass Fertilized 1 20 63.3 < 0.0001 Litter 1 20 6.24 0.0213 Year 3 60 1.92 0.1356 Fertilized*Litter 1 20 1.83 0.1910 Year*Fertilized 3 60 4.82 0.0045 Year*Litter 3 60 0.65 0.5830 Year*Fertilized*Litter 3 60 0.94 0.4268

Shannon's H Fertilized 1 20 9.93 0.0050 Litter 1 20 4.24 0.0526 Year 1 20 0.03 0.8852 Fertilized*Litter 1 20 1.18 0.2896 Year*Fertilized 1 20 0.08 0.7748 Year*Litter 1 20 0.02 0.8873 Year*Fertilized*Litter 1 20 0.81 0.3787

(4.04), - litter/- fertilizer 14.04 (1.22), + litter/+ fertilizer 60.18 (9.18), and – litter/ + fertilizer 15.58 (2.14).

Percent soil moisture (Fig. 2c) was decreased by fertilization (F1, 20 = 17.13, P

= 0.0005), but was not significantly affected by litter (F1, 20 = 1.88, P = 0.1860) and the interaction between litter and fertilization (F1, 20 = 1.84, P = 0.1901). Percent soil organic content (Fig. 2c) was not significantly affected by fertilization (F1, 20 = 1.59,

P = 0.2224) nor litter (F1, 20 = 1.01, P = 0.3273), and there was no significant interaction between litter and fertilization (F1, 20 = 2.15, P = 0.1578). PAR (Fig. 2d) was strongly decreased by fertilization (F1, 20 = 86.13, P < 0.0001), but without significant effects of either litter (F1, 20 = 0.56, P = 0.4628) or the interaction between litter and fertilization (F1, 20 = 0.16, P = 0.6905).

Species-level analyses

Biomass samples for 45 species were collected during 2005 (Table 2) and used for the CCA analysis. The first axis of the CCA (Fig. 3) accounted for 12.4% of the variance in plant species biomass and separated fertilized from unfertilized plots, suggesting that plant species differentially respond to fertilization, with the majority of species in unfertilized plots. This first axis generally separated two of the most abundant graminoids, B. inermis and Dactylis glomerata, and the occasionally abundant forbs (e.g., Galium mollugo, Cirsium arvense from two of the other abundant graminoids, F. arundinacea and A. odoratum, and from the rarer and lower 38

Table 2. Proportion of total living biomass within each year for each plant species.

“Code” refers to the five letter species abbreviation used in CCA ordinations in Fig.

5. “F” indicates forbs, “G” indicates grasses (Poaceae), “S” indicates graminoids other than Poaceae, and “W” indicates woody plant species. “Native” refers to plant species indigenous to Ohio, USA, whereas “Non” refers to plant species not native to

Ohio, USA. Bold text indicates the seven most abundant grass species in the living biomass sampling in 2005.

Species Code Growth Form Native/Non % Biomass Achillea millefolium L. ACHMI F Non 0.0007 Agropyron repens (L.) AGRRE G Non 0.0216 Anthoxanthum odoratum L. ANTOD G Non 0.0439 Apocynum cannabinum L. APOCA F Native 0.0015 Aster pilosus Willd. ASTPI F Native 0.0052 Aster prealtus Poir. ASTPR F Native 0.0080 Bromus inermis Leyss. BROIN G Non 0.2766 Bromus japonicus Thunb. Ex Murr. BROJA G Non 0.0001 Calystegia sepium (L.) CALSE F Native 0.0001 Carex ssp. L. CARSP S Native 0.0094 Cerastium nutans Raf. CERNU F Native 0.0003 Chrysanthemum leucanthemum L. CHRLE F Non 0.0020 Cirsium arvense (L.) Scop. CIRAR F Non 0.0062 Cornus racemosa Lam. CORRA W Native 0.0004 Crataegus ssp. L. CRASP W Native 0.0116 Dactylis glomerata L. DACGL G Non 0.0454 Daucus carota L. DAUCA F Non 0.0005 Euthamia graminifolia (L.) EUTGR F Native 0.0001 Festuca arundinacea Schreb. FESAR G Non 0.2088 Frangula alnus P. Mill. FRAAL W Non 0.0003 Fraxinus americana L. FRAAM W Native 0.0005 Galium mollugo L. GALMO F Non 0.0080 Geum ssp. L. GEUSP F Native 0.0001 Glechoma hederacea L. GLEHE F Non 0.0012 Hieracium aurantiacum L. HIEAU F Non 0.0044 Holcus lanatus L. HOLLA G Non 0.0190 Hypericum punctatum Lam. HYPPU F Native 0.0001 Juncus tenuis Willd. JUNTE S Native 0.0050 Oxailis stricta L. OXAST F Native 0.0003 Panicum ssp. L. PANSP G Native 0.0003 Phleum pratense L. PHLPR G Non 0.0769 Plantago lanceolata L. PLALA F Non 0.0109 Poa pratensis L. POAPR G Non 0.1889 Poa trivialis L. POATR G Native 0.0252 Potentilla canadensis L. POTCA F Native 0.0023 Potentilla recta L. POTRE F Non < 0.0001 Prunella vulgaris L. PRUVU F Native 0.0077 Prunus serotina Ehrh. PRUSE W Native < 0.0001 Ranunculus acris L. RANAC F Non 0.0013 Scirpus georgianus Harper SCIGE S Native 0.0034 Taraxacum officinale Weber TAROF F Non 0.0003 Toxicodendron radicans (L.) Kuntze TOXRA W Native < 0.0001 Trifolium campestre Schreb. TRICA F Non 0.0002 Trifolium pratense L. TRIPR F Non < 0.0001 Viburnum rafinesquianum J.A. Schultes VIBRA W Native 0.0013

Figure 3. Ordination diagrams of the first two axes of canonical correspondence analysis (CCA) with the four treatments as environmental dummy variables. Vectors indicate the direction and strength of correlations between axes scores and the environmental dummy variables, and the percent of variance explained by each axis is noted next to the axis title. Biomass of each of the 45 plant species collected in

2005 with R2 correlations of axis to environment variables for axis 1: +L/-F = 0.080, -

L/-F = 0.634, +L/+F = 0.457, and –L/+F = 0.162, and for axis 2: +L/-F = 0.314, -L/-

F = 0.051, +L/+F = 0.258, and –L/+F = 0.709. See Fig. 2 for key to treatment symbols. The five letter species codes are defined in Table 2.

1.5 CORRA FRAAL A TOXRA GEUSP ASTPR A TRICA A A +L/+F PRUVU

+L/-F 0.5 HIEAU CALSE BROIN CARSP ASTPI PANSP FESAR TAROF RHAFR POAPR PLALA RANAC GLEHE CRASP CHRLE -L/-F HYPPU GALMO DAUCA ANTOD POTCA ACHMI ASTLA DACGL CERNU JUNTE POTRE

Axis 2 (7.9%) Axis PRUSE -0.5 POATR SCIGE CIRAR TRIPR HOLLA PHLPR VIBRA -L/+F AGRRE OXAST APOCA BROJA -1.5 -1.0 0.0 1.0 2.0 Axis 1 (12.4%)

biomass plant species (e.g., Achillea millefolium, Hieracium aurantiacum, Potentilla recta). The second axis (Fig. 3) accounted for 7.9% of the variance in plant species biomass and separated litter removed from litter left in situ plots, suggesting that plant species also respond to differential litter loads. The greatest species responses to plant litter clustered around the fertilized and litter removed (-L/+F) pole. The species-environment Pearson correlations (R2) were 0.882 for the first axis and 0.890 for the second axis.

Plant group responses

Fertilization decreased species richness within three of the plant growth-form groups, i.e., forbs, non-Poaceae graminoids, and woody plant species (Table 3, Fig

3a). The grasses showed no significant responses to any of the predictor variables for species richness (Table 3, Fig. 4a). Forbs were the only growth-form plant group that was significantly affected by litter as species richness increased with litter removal

(Table 3, Fig. 4a). Fertilization increased the biomass of grasses, but decreased the non-Poaceae graminoids (Table 3, Fig. 4b). Thus, while the species richness of grasses was unaffected by either fertilization or litter, the biomass of the grasses were strongly affected by fertilization (Fig. 4b).

Fertilization decreased the species richness for both the native and non-native plant groups (Table 3, Fig. 5a). Litter had no effect on native plant species richness, but the removal of litter increased non-native plant species richness (Table 3, Fig. 5a). 43

Table 3. Results of factorial ANOVA for each of the six plant groups. “Fertilized” treatment is either fertilized (20 g N m-2 added) or unfertilized; Litter treatment is either plant litter removed or left in situ. “Fertilized*Litter” indicates their factorial interaction. For all analyses, df numerator = 1, and df denominator = 20. P-values significant at α = 0.05 are denoted by bolding.

Species Richness Biomass F P F P

Forbs Fertilized 84.44 < 0.0001 2.01 0.1712 Litter 9.88 0.0051 0 0.9694 Fertilized*Litter 0.53 0.4766 0 0.9746

Grasses (Poaceae) Fertilized 0.31 0.5853 109.56 < 0.0001 Litter 0.31 0.5853 0.59 0.4498 Fertilized*Litter 0.31 0.5853 1.12 0.3023

Non -Poaceae Graminoids Fertilized 9.62 0.0056 7.07 0.0151 Litter 0.38 0.5421 0.5 0.4857 Fertilized*Litter 3.46 0.0776 0.02 0.9011

Woody Fertilized 5.07 0.0357 0.89 0.3555 Litter 0.56 0.4616 0.61 0.4442 Fertilized*Litter 0.14 0.7114 1.07 0.3139

Native Fertilized 23.75 < 0.0001 4.06 0.0575 Litter 0.01 0.9427 1.03 0.3212 Fertilized*Litter 2.8 0.1099 0.08 0.7837

Non -Native Fertilized 14.15 0.0012 206.69 < 0.0001 Litter 8.42 0.0088 2.15 0.1586 Fertilized*Litter 0.29 0.597 1.48 0.2374

Figure 4. A) Average species richness of each growth form plant group within each treatment, and B) average biomass of each growth form plant group within each treatment. Non-Grass refers to non-Poaceae graminoids, i.e., Juncaceae and

Cyperaceae. Full results of the factorial ANOVAs are in Table 3. See Fig. 2 for key to treatment symbols.

A) Forbs 12 Non Grass

s Poaceae

s 10

Woody

e n

h 8

i R

e i

c 4

e p

S 2

0 +L/-F -L/-F +L/+F -L/+F Treatment

900 Forbs ² Non Grass

m 800

Poaceae 1

700 r Woody

e 600

) 500

g (

400 s

s 300 a

m 200 o

i 100 B 0 +L/-F -L/-F +L/+F -L/+F

Treatment

Figure 5. A) Average species richness of the native and non-native plant groups within each treatment, and B) average biomass of the native and non-native plant groups within each treatment. Full results of the factorial ANOVAs are in Table 3.

See Fig. 2 for key to treatment symbols.

14 Native Non

s 12

s e

n 10

h c

i 8

s 6

i c

e 4 p

S 2 0 +L/-F -L/-F +L/+F -L/+F Treatment

900 ²

m Native

800 1

Non

r 700 e

p 600

)

g 500

(

s 400 s

a 300 m

o 200 i

B 100 0 +L/-F -L/-F +L/+F -L/+F Treatment

Fertilization increased the biomass of non-native and native plant species, but the effect was much higher for non-native with biomass approximately doubling in fertilized plots (Table 3, Fig. 5b).

Ecosystem response

NMS ordination showed tight clustering of plots into treatments (Fig. 6). The ordination axes explained 50.2% of the variance, with the first axis explaining 40.8% of the variance, and the second axis explaining 9.3% of the variance. The final stress

= 5.83 with a final instability = 0.086, and results of the Monte Carlo simulation indicated that this stress was less than expected by chance (P = 0.001). Following

Clarke (1993), a final stress between 5 – 10 was a very good ordination and did not present any real risk of misinterpretation. The first axis separated fertilized and unfertilized plots with high correlations to living biomass (R2 = 0.829) in the direction of fertilized plots, and correlations to species richness (R2 = 0.582), PAR (R2 =

0.750), and percent soil moisture (R2 = 0.479) in the direction of unfertilized plots, while litter biomass and percent soil organic content were not well correlated (R2 =

0.008 and R2 = 0.053, respectively). The second axis separated litter removed from litter left in situ plots with a good correlation to litter biomass (R2 = 0.551) and only very weak or no correlations to the other five variables: living biomass R2 = 0.001, species richness R2 = 0.015, PAR R2 = 0.007, percent soil moisture R2 = 0.066, and percent soil organic content R2 = 0.028. 47

Figure 6. Two-dimensional ordination of ecosystem-level properties from 2005 in 24 experimental plots from nonmetric multidimensional scaling (NMS) using living biomass, species richness, litter biomass, PAR, percent soil moisture, and percent soil organic content. Vectors indicate the direction and strength of correlations between axis scores and emergent properties (R2 cutoff for joint Biplot = 0.000 to show % soil organic content, R2 of all other vectors is > 0.200) and ordinations are rotated to the dominant axis of living biomass. The percent of variance explained by each axis is noted next to the axis title. See Fig. 2 for key to treatment symbols.

0.6 Treatment +L/-F -L/-F +L/+F 0.2 -L/+F Species Richness % Soil Organic Content PAR Living Biomass % Soil Moisture

-0.2 Axis 2 (9.3%) Axis

-0.6 Litter Biomass

-1.0 -1.0 -0.5 0.0 0.5 1.0 Axis 1 (40.8%)

Table 4. Results of Multi-response Permutation Procedure (MRPP) on emergent properties for 2005. T describes the separation between groups (dissimilarity) and A is the chance-corrected within-group agreement. “All” indicates all four treatments included in the MRPP, and the remainders are MRPP pairwise comparisons of treatments to assess dissimilarity (lower T and higher A). +L indicates litter left in situ, -L indicates litter removed after annual mowing in early autumn, +F indicates fertilization (20 g N m-2) in early spring, -F indicates no fertilization.

Groups T A P All -9.303 0.557 < 0.0001 +L/-F vs. –L/-F -3.531 0.241 0.0067 +L/-F vs. +L/+F -5.952 0.405 0.0004 +L/-F vs. –L/+F -6.419 0.452 0.0004 -L/-F vs. +L/+F -6.169 0.425 0.0004 -L/-F vs. –L/+F -5.574 0.383 0.0005 +L/+F vs. –L/+F -3.563 0.236 0.0060

This strong separation of plots into treatment clusters was supported by MRPP

(Table 4). When all four treatments were run together, the null hypothesis of no difference between treatments was rejected with high within-group agreement and very strong separation between groups. Pairwise comparisons of treatments showed that fertilized plots, while still significantly distinct, were more similar to each other than fertilized treatment plots are to any of the unfertilized treatment plots. The same pattern existed for unfertilized plots, with strong separation of unfertilized plots, yet with lower dissimilarity than when unfertilized plots were compared to fertilized plots. As expected, the maximal differences occurred when extremes of treatments were paired, as in –L/-F vs. +L/+F, and +L/-F vs. –L/+F, indicating that “opposite” treatments radically alter biotic and abiotic components of the local habitat.

Discussion

My results clearly show that plant litter and fertilization alter plant community biomass, plant community structure, and ecosystem properties. The spatial and temporal scale of my study allowed us to assess species-level responses to fertilization and litter treatments within the entire plant community as well as ecosystem-level responses to fertilization and litter treatments.

Supporting my first hypothesis and consistent with other published studies

(e.g., Dyer et al. 1991; Tilman et al. 2002a; Long et al. 2003), fertilization strongly increased plant biomass while generally decreasing plant species and plant group 50

diversity. Fertilization effects on plant biomass were strongest in those plots where litter was left in situ, reflecting not only higher standing crop production, but also plant litter production and accumulation. However, some plant groups responded to fertilization in unpredicted ways.

Forb species were the most speciose plant group overall, including a number of species rare to the site (e.g., Hypericum punctatum Lam., Trifolium pratense L.,

Calystegia sepium (L.)). The loss of forb species due to fertilization was consistent with the abundance-based mechanism of diversity loss due to fertilization (Suding et al. 2005). However, it is striking that fertilization significantly affected only forb species richness, not forb species biomass, contradicting my first hypothesis. Also contradicting conventional theory (e.g., Tilman et al. 2002a; Suding et al. 2005), grass species richness showed no response to fertilization, but grass species biomass was strongly affected by fertilization. This effect was likely due to my experiment manipulating an established plant community with a unique mix of native and non- native grasses and forbs, as opposed to the intentionally seeded/planted plant communities more commonly used in fertilization manipulation experiments (e.g.,

Foster and Gross 1998; Levine et al. 1998). My results suggest that established grassland systems may not respond to fertilization and litter manipulation in ways generally predicted by theories derived from small-scale, short-term experiments.

Consistent with my second hypothesis, removing plant litter augmented

Shannon’s H, though as predicted the effect was dampened in fertilized plots.

However, plant litter did not predictably affect the species richness and/or biomass of 51

some plant groups. Grasses showed no responses to plant litter for either grass species richness or grass species biomass, whereas non-native plant species richness significantly increased in litter removed plots with only a nominal, non-significant biomass response. Thus, while litter removal generally increased species richness overall, this manipulation facilitated non-native species recruitment and retention even though the effects on biomass were marginal at best.

The total biomass measurement includes the litter biomass, therefore it is intuitive that total biomass would be lower in litter-removed plots, explaining the significance of litter in the repeated PROC MIXED. The marginally significant P- value for litter in the repeated PROC MIXED for Shannon’s H underscores the notion that litter itself affects plant diversity, though litter did not significantly affect any of the abiotic factors. Contrary to my results, Weltzin et al. (2005) found that the removal of litter in a northern fen affected abiotic factors, including increased availability of light and soil temperature. Perhaps litter did not affect abiotic factors in my grassland because of the higher productivity of the field when compared to the fen community, and the dense mat-forming properties of the dominant grasses. I conclude that fertilization is the strongest determining factor for total plant biomass production and abiotic conditions, though litter should be considered when assessing plant species diversity, especially in consideration of the response of forbs and non- native plant species.

In support of my third hypothesis, plant litter and fertilization interacted to induce a shift in community structure as shown by fine-scale analyses of species 52

responses to treatments, with the most abundant grasses, particularly B. inermis in fertilized plots and F. arundinacea in unfertilized plots, separating along the first

CCA axis (Fig. 3). The large majority of plant species lie along the unfertilized portion of the first axis, bolstering the assertion that fertilization decreases plant diversity (Grime 1979; Tilman et al. 2002b). The distinct separation along the fertilization axis of F. arundinacea and B. inermis, the two most abundant grasses

(see Table 2), indicates a stronger response to fertilization elicited by B. inermis, a species known to more efficiently utilize N when compared with F. arundinacea (Eck et al. 1981). In my study, Poa pratensis, my third most abundant grass, was intermediate to B. inermis and F. arundinacea along the fertilization axis, showing a weak affiliation with fertilization. The relationship between P. pratensis and fertilization appears to be somewhat equivocal, as Pennings et al. (2005) report that the abundance of P. pratensis decreased in five of nine fertilization experiments.

Indeed, P. pratensis was ubiquitous in my system and, therefore, its placement near the centroid of the CCA was not surprising.

The experimental plots cluster very tightly into distinct aggregates of ecosystem and plant community properties, separated along the first axis by fertilization and along the second axis by plant litter (Fig. 6). Living biomass again associates with fertilized plots, while species richness and light attenuation (PAR) were higher in unfertilized plots, consistent with other published studies that included abiotic factors (e.g., Foster and Gross 1998). Percent soil moisture increased in the direction of unfertilized plots, a surprising result given that higher litter production 53

and lower light attenuation at the soil level would intuitively seem more conducive to increased soil moisture (Hamrick and Lee 1987); however, this may be explained by greater plant biomass in fertilized plots having a greater water demand, thus leading to soil drying. The greatest scatter within a treatment is in the control group (+L/-F), reflecting the spatial heterogeneity expected in an old field. Interestingly, fertilized plots exhibited extremely tight clustering relative to unfertilized plots, indicating strong within-group similarity and reduced heterogeneity.

The measurement and analysis of the plant community and associated abiotic properties demonstrates strong treatment effects. These highly differentiated treatments may affect ecosystem function (e.g., sequestering carbon), an effect likely to increase in magnitude through time as the communities’ responses to treatments mature. While each of these biotic properties has been shown to respond individually to fertilization (e.g., Carson and Peterson 1990; Tilman et al. 2002a, b; Long et al.

2003), this is the first time that these three biotic ecosystem properties have been used in a multivariate ordination to explicitly determine whether they can define discrete and distinct plant communities and their associated abiotic properties. Most of the previous work on the effects of fertilization and plant litter has focused on a single, dominant response species or a small collection of species within the entire community over a single growing season (e.g., Foster and Gross 1997, 1998; Violle et al. 2006). I know of no other long-term, temperate system studies that have manipulated both fertilization and plant litter in an established plant community at the 54

same (or greater) scale. However, I realize that my study has some distinct differences when compared to previous work.

My use of an NPK fertilizer, as opposed to N-only fertilizer, is likely to have induced a stronger response to fertilization due to the added P and K. Nevertheless, my results were generally consistent with other studies that used NPK fertilizers (e.g.,

Carson and Barrett 1988, Turkington et al. 2002), N-only fertilizers (e.g., Tilman et al. 2002b, Long et al. 2003) as well as other studies that simulated N-only atmospheric deposition (e.g., Throop 2005). Further, my running definition of litter

(see Methods) includes the vegetation mown in the previous year and not removed in my litter removal treatment, potentially altering the nutritional quality of the litter relative to naturally senesced vegetation, and the physical structure of the litter as it lay after mowing (e.g., Semmartin et al. 2004). Because the timing of the mowing was determined by the local township, litter from the annual mowing accumulated earlier than might normally be expected for my region of the USA. Were the mowing to stop, the site would very quickly yield to encroaching woody vegetation typical of early secondary succession (Cook et al. 2005).

This study expands our understanding of the long-term effects of fertilization and plant litter on plant species and specific plant groups, confirms some aspects of previous work (e.g., Tilman et al. 2002b), and extends the scope of scientific knowledge regarding the effects of terrestrial ecosystem eutrophication on ecosystem- level processes. Some plant groups did not respond to my treatments as predicted by theory, which underscores the need for larger-scale and longer-term experiments in 55

natural systems. When viewed through multivariate ordination, the realized differences in biotic and abiotic ecosystem properties emerged, and these properties indicate that each treatment could potentially alter the trajectory of grassland ecosystem dynamics.

References

Andreas, B. K., J. J. Mack, and J. S. McCormac. 2004. Floristic Quality Assessment

Index (FQAI) for vascular plants and mosses for the State of Ohio. Ohio

Environmental Protection Agency, Division of Surface Water, Wetland

Ecology Group, Columbus.

Bobbink, R., M. Hornung, and J. G. M. Roelofs. 1998. The effects of air-borne

nitrogen pollutants on species diversity in natural and semi-natural European

vegetation. Journal of Ecology 86:717–738.

Bosy, J. L., and R. J. Reader. 1995. Mechanisms underlying the suppression of forb

seedling emergence by grass (Poa pratensis) litter. Functional Ecology

9:635–639.

Carson, W. P., and G. W. Barrett. 1988. Succession in old-field plant communities:

effects of contrasting types of nutrient enrichment. Ecology 69:984–994.

Carson, W. P., and C. J. Peterson. 1990. The role of litter in an old-field plant

community: impact of litter quantity in different seasons on plant species

richness and abundance. Oecologia 85:8–13.

Clarke, K. R. .1993. Non-parametric multivariate analyses of changes in community

structure. Australian Journal of Ecology 18:117–143.

Cook, W. M., J. Yao, B. L. Foster, R. D. Holt, and L. B. Patrick. 2005. Secondary

succession in an experimentally fragmented landscape: community patterns

across space and time. Ecology 86:1267–1279. 57

Dyer, M. I., C. L. Turner, and T. R. Seastedt. 1991. Mowing and fertilization effects

on productivity and spectral reflectance in Bromis inermus plots. Ecological

Applications 1:443–452.

Edwards, G. R., and M. J. Crawley. 1999. Herbivores, seed banks and seedling

recruitment in mesic grassland. Journal of Ecology 87:423–435.

Eck, H. V., G. C. Wilson, and T. Martinez. 1981. Tall fescue, Festuca arundinacea,

and smooth brome, Bromis inermis, 2. Effects of nitrogen fertilization and

irrigation regimes on quality. Agronomy Journal 73:453–456.

Facelli, J. M. 1994. Multiple indirect effects of plant litter affect the establishment of

woody seedlings in old fields. Ecology 75:1727–1735.

Facelli, J. M., and S. T. A. Pickett. 1991. Plant litter: light interception and effects

on an old-field plant community. Ecology 72:1024–1031.

Fenn, M. E., J. S. Baron, E. B. Allen, H. M. Rueth, K. R. Nydick, L. Geiser, W. D.

Bowman, J. O. Sickman, T. Meixner, D. W. Johnson, and P. Neitlich. 2003.

Ecological effects of nitrogen deposition in the western United States.

BioScience 53:404–420.

Foster, B. L., and K. L. Gross. 1997. Partitioning the effects of plant biomass and

litter on Andropogon gerardii in old-field vegetation. Ecology 78:2091–2104.

Foster, B. L., and K. L. Gross. 1998. Species richness in a successional grassland:

effects of nitrogen enrichment and plant litter. Ecology 79:2593–2602. 58

Galloway, J. N., J. D. Aber, J. W. Erisman, S. P. Seitzinger, R. W. Howarth, E. B.

Cowling, and B. J. Cosby. 2003. The nitrogen cascade. BioScience 53:341–

355.

Gleason, H. A., and A. Cronquist. 1991. Manual of Vascular Plants of Northeastern

United States and Adjacent Canada, 2nd ed. New York Botanical Garden,

New York, NY, USA.

Goldberg, D. E., and P. A. Werner. 1983. The effects of size of opening in

vegetation and litter cover on seedling establishment of goldenrods (Solidago

spp.). Oecologia 60:149–155.

Grime, J. P. 1979. Plant Strategies and Vegetation Processes. John Wiley and Sons,

London, UK.

Hamrick, J. L., and J. M. Lee. 1987. Effect of soil surface topography and litter

cover on the germination, survival, and growth of musk thistle (Carduus

nutans). American Journal of Botany 74:451–457.

Hayek, L. C., and M. A. Buzas. 1997. Surveying Natural Populations. Columbia

University Press, New York, NY, USA.

Hector, A., B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G.

Dimitrkopoulos, J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P.

Högberg, K. Huss-Danell, J. Joshi, A. Jumpponen, C. Körner, P. W. Leadley,

M. Loreau, A. Minns, C. P. H. Mulder, G. O’Donovan, S. J. Otway, J. S.

Pereira, A. Prinz, D. J. Read, M. Scherer-Lorenzen, E.-D. Schulze, A.-S.

D.Siamantziouras, E. M. Spehn, A. C. Terry, A. Y. Troumbis, F. I. 59

Woodward, S. Yachi, and J. H. Lawton. 1999. Plant diversity and

productivity experiments in European grasslands. Science 286:1123–1127.

Hulme, P. E. 1996. Herbivores and the performance of grassland plants: a

comparison of arthropod, mollusk and rodent herbivory. Journal of Ecology

84:43–51.

Köchy, M., and S. D. Wilson. 2005. Variation in nitrogen deposition and available

soil nitrogen in a forest-grassland ecotone in Canada. Landscape Ecology

20:191–202.

Kruskal, J. B. 1964. Nonmetric multidimensional scaling: a numerical method.

Psychometrika 29:115–129.

Larsen, T. H., N. Williams, and C. Kremen. 2005. Extinction order and altered

community structure rapidly disrupt ecosystem functioning. Ecology Letters

8:538–547.

Levine, J. M., J. S. Brewer, and M. D. Bertness. 1998. Nutrients, competition and

plant zonation in a New England salt marsh. Journal of Ecology 86:285–292.

Long, Z. T., C. L. Mohler, and W. P. Carson. 2003. Extending the resource

concentration hypothesis to plant communities: effects of litter and herbivores.

Ecology 84:652–665.

McCann, K. S. 2000. The diversity-stability debate. Nature 405:228–233.

McCune, B., and M. J. Mefford. 1999. Multivariate Analysis of Ecological Data

Version 4.37. MjM Software, Gleneden Beach, OR, USA. 60

Mielke, P. W., Jr. 1984. Meteorological applications of permutations techniques

based on distance functions. In: Krishnaiah, P. R., and P. K. Sen, eds.,

Handbook of Statistics, Vol. 4., Elsevier, St. Louis, MO, USA, pp. 813–830.

Pennings, S. C., C. M. Clark, E. C. Cleland, S. L. Collins, L. Gough, K. L. Gross, D.

G. Milchunas, and K. N. Suding. 2005. Do individual plant species show

predictable responses to nitrogen addition across multiple experiments?

Oikos 110:547–555.

Ritchie, A., and J. R. Steiger. 1974. Soil Survey of Summit County Ohio. United

States Department of Agriculture, Soil Conservation Service in cooperation

with Ohio Department of Natural Resources, Division of Lands and Soil and

the Ohio Agricultural Research and Development Center.

Sala, O. E., W. J. Warton, L. A. Joyce, and W. K. Lauenroth. 1988. Primary

production of the central grasslands region of the United States. Ecology

69:40–45.

SAS Institute, Inc. 1999. SAS System for Windows. SAS Institute, Inc., Cary, NC,

USA.

Semmartin, M., M. R. Aguiar, R. A. Distel, A. S. Moretto, and C. M. Ghersa. 2004.

Litter quality and nutrient cycling affected by grazing-induced species

replacements along a precipitation gradient. Oikos 107:148–160.

Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.

Milchunas, and S. Pennings. 2005. Functional- and abundance-based 61

mechanisms explain diversity loss due to N fertilization. Proceedings of the

National Academy of Sciences 102:4387–4392.

Throop, H. L. 2005. Nitrogen deposition and herbivory affect biomass production

and allocation in an annual plant. Oikos 111:91–100.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002a. Plant diversity and

composition: effects on productivity and nutrient dynamics of experimental

grasslands. In: Loreau, M., S. Naeem, and P. Inchausti, eds., Biodiversity and

Ecosystem Functioning. Synthesis and Perspectives. Oxford University Press,

Oxford, UK, pp. 21–35.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002b. Experimental and

observational studies of diversity, productivity, and stability. In: Kinzig, A.

P., S. W. Pacala, and D. Tilman, eds., The Functional Consequences of

Biodiversity. Princeton University Press, Princeton, NJ, USA, pp. 42–70.

Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of

fertilization and herbivory on the herbaceous vegetation of the boreal forest in

north-western Canada: a 10-year study. Journal of Ecology 90:325–337.

US EPA. 2005. Clean air status and trends network (CAST-NET).

http://www.epa.gov/castnet/

Violle, C., J. Richarte, and M. Navas. 2006. Effects of litter and standing biomass on

growth and reproduction of two annual species in a Mediterranean old-field.

Journal of Ecology 94:196–205. 62

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human

domination of Earth’s ecosystems. Science 277:494–499.

Weltzin, J. F., J. K. Keller, S. D. Bridgham, J. Pastor, P. B. Allen, and J. Chen. 2005.

Litter controls plant community composition in a northern fen. Oikos

110:537–546.

Wilsey, B. J., and H. W. Polley. 2006. Aboveground productivity and root-shoot

allocation differ between native and introduced grass species. Oecologia

150:300–309.

Xiong, S., and C. Nilsson. 1999. The effects of plant litter on vegetation: a meta-

analysis. Journal of Ecology 87:984–994.

CHAPTER 3

BEETLE (COLEOPTERA) RESPONSES TO FERTILIZATION AND PLANT

LITTER IN A TEMPERATE OLD-FIELD GRASSLAND

Abstract

Recent studies in terrestrial ecosystems have shown that increased nutrient loading led to increased arthropod abundance and biomass and decreased arthropod diversity. However, these studies have focused on the portion of the food web most closely associated with living plants (i.e., the ―aerial‖ arthropod community) – none of these studies have explicitly examined the effects of nutrient loading, plant diversity, and plant litter on the ground-level (―epigeal‖) arthropod community. To test whether these nutrient loading effects extend to the detritus-based epigeal community and to individual species within this community, I used pitfall traps to sample the epigeal beetle community for four years within 24 large (314 m2) plots in which I manipulated nutrient loading (NPK fertilizer) and plant litter. I measured the diversity, abundance, biomass, and community structure responses of the entire epigeal beetle community and of carabid and staphylinid beetles, as well as the abundance and biomass responses of the six most common individual beetle species in a temperate-latitude grassland. As with previous

63 64

studies that sampled the aerial arthropod community, I hypothesized that increased nutrient loading would increase epigeal beetle abundance and biomass but decrease epigeal beetle diversity. Contrary to predictions, total beetle species richness, diversity, and abundance were significantly increased as a result of fertilization, while beetle biomass was moderately decreased as a result of fertilization. Also contrary to predictions, plant litter moderately decreased these same beetle community variables.

Staphylinid beetles showed the strongest responses to fertilization, with two species,

Meronera venustula and the exotic and parthenogenetic Mocyta fungi, colonizing nearly exclusively on fertilized plots after two years of experimental manipulations.

Surprisingly, feather-winged beetles (Ptiliidae) also strongly responded to fertilization and may also be good candidates for predicting epigeal beetle responses to nutrient loading. I based my hypotheses on previous studies that investigated the aerial arthropod community and my results did not support any of the hypotheses. My results clearly emphasize that the epigeal community requires more investigation and integration into conventional biodiversity-productivity theory.

Introduction

A major consequence of human activity has been an increase in the global nitrogen (N) pool through fertilization and increased atmospheric N deposition (Vitousek et al. 1997; Fenn et al. 2003; Galloway et al. 2003). Galloway et al. (2003) reported an

11-fold increase in reactive N from 1860 to 2000, citing anthropogenic factors (e.g., 65

industrial and agricultural reactive N production and utilization) as the reason for the dramatic increase. The effects of increased N on terrestrial plant communities are well documented across a variety of natural and semi-natural habitats (Bobbink et al. 1998), as well as in experimentally manipulated habitats and microcosms (Hector et al. 1999;

Tilman et al. 2002; Suding et al. 2005; Patrick et al. 2008), and include decreased plant species richness, increased standing crop biomass, and plant community composition limited to a few dominant species. Decreased species richness and increased abundance resulting from nutrient loading have also been documented for terrestrial arthropods, particularly those species most closely linked to the living-plant portion of the food web

(Knops et al. 1999; Haddad et al. 2000, 2001). This ―eutrophication effect‖ (Fenn et al.

2003) can result in significant biodiversity loss and the potential decline in important ecosystem functions, such as ecosystem stability (McCann 2000; Larsen et al. 2005).

Previous studies have focused on the portions of the food web closely tied to living plant material by sampling the ―aerial‖ arthropod community with sweep nets

(Knops et al. 1999; Haddad et al. 2000, 2001). While there is evidence to support the eutrophication effect on aerial arthropod diversity, it is not known how nutrient loading affects the epigeal (ground-level) arthropod community. The epigeal arthropod community is more closely tied to the detritus-based portion of the food web (Halaj and

Wise 2002). This detrital community increases overall biodiversity and facilitates effective nutrient regeneration and recycling into an ecosystem and thereby influences ecosystem functioning (Moore et al. 2004; Hättenschwiler and Gasser 2005), however it remains an understudied food web component. 66

Nutrient loading not only increases plant standing crop biomass, but also plant litter production (Long et al. 2003; Patrick et al. 2008), which can increase the basal food resource for the detrital community and increase detritivore and epigeal predator abundances (Halaj et al. 2000; Halaj and Wise 2002; Moore et al. 2004). Furthermore, plant litter increases habitat complexity which can also increase arthropod abundance and diversity (Lawton 1983; Strong et al. 1984). However, the detritivore community response can be further complicated because it can be sensitive to litter diversity

(Hättenschwiler and Gasser 2005). With nutrient loading decreasing plant diversity

(Tilman et al. 2002; Long et al. 2003; Patrick et al. 2008), litter diversity is also decreased. While more plant litter production could increase detritivore and epigeal predator abundance and biomass (Halaj et al. 2000), a reduction in litter diversity could result in declines in diversity of detritivores and epigeal predators (Hättenschwiler and

Gasser 2005), effectively mirroring the aerial community response to nutrient loading.

Despite such predictions, no studies have incorporated plant litter effects with biodiversity-productivity theory, nor have any studies investigated these effects on either the epigeal arthropod community or on species-level responses within the epigeal community.

Beetles are good candidates for assessing the effects of fertilization and plant litter on the epigeal arthropod community. Epigeal beetles (e.g., Carabidae and Staphylinidae) have been used as indicators of changes in habitat quality because they are sensitive to environmental change (Buse and Good 1993; Loreau 1993; Jonas et al. 2002; Pohl et al.

2007) and are adequately speciose to examine community responses to experimental 67

manipulations (Loreau 1993). Moreover, detritivores are a significant portion of the diet of carabids (Sunderland 1975; Chiverton and Sotherton 1991), and previous studies (e.g.,

Snyder and Wise 2001) have reported that carabids can alter trophic cascades based on the abundance of their plant detritus-feeding prey. Ultimately, epigeal beetles are largely separated from the living-plant based portion of the food web and can affect the detritus- based portion of the food web, potentially altering nutrient regeneration into the ecosystem.

Here I report the results of a four year study that investigated the effects of experimentally manipulating NPK fertilization and plant litter on the epigeal beetle community. I measured the diversity, abundance, biomass, and community structure responses of the entire epigeal beetle community, carabid and staphylinid beetles, and individual beetle species in a temperate-latitude grassland. Based on the findings of previous studies that sampled the aerial arthropod community responses to nutrient loading (e.g., Knops et al. 1999; Haddad et al. 2000, 2001), I tested four hypotheses: as a result of fertilization, 1) beetle biomass and abundance will increase while beetle diversity decreases, 2) beetle diversity will decrease as plant diversity decreases, 3) beetle diversity will decrease as plant biomass increases, and 4) plant litter will moderately increase beetle diversity, though this effect will be dampened in fertilized plots.

Materials and methods

Study site and experimental design 68

The study was done at the 163.5 ha Bath Nature Preserve (BNP; 41° 10’ 36.2‖ N,

81° 38’ 58.7‖ W), Bath Township, Summit County, Ohio, USA, in a 16 ha section of grassland. Until the early 1980s, the study site was a hay meadow, harvested one or many times per year. Since then, the area has been mown annually in late August to early September, and the mown vegetation has been left on the field. The dominant vegetation is an herbaceous, graminoid community largely dominated by cool-season C3 grasses, e.g., Bromus inermis Leyss., Festuca arundinacea Schreb., Phleum pratense L., and Anthoxanthum odoratum L. The site is moderately productive relative to other grasslands within the upper Midwest and across the U.S. (Patrick et al. 2008). The dominant soil type is Ellsworth silt loam (ElB), which consists of moderately well drained, moderately deep to deep soils formed in silty clay loam or clay loam glacial till of the Wisconsin Age (Ritchie and Steiger 1974).

During August 2001, twenty-four 20-m diameter circular plots (314 m2) were established. These experimental plots were separated by at least 20 m and were at least

30 m away from any other habitat. Treatments were applied in a 2 x 2 factorial design of fertilizer (+F = fertilizer added, -F = no fertilizer) and plant litter (-L = litter removed, +L

= litter left in situ after yearly mowing) with the control plots characterized as no fertilization and plant litter left in situ (+L/-F), resulting in six replicates per treatment.

Hereafter, all references to ―litter‖ refer to the previous year’s mown vegetation and any vegetation senesced and found within the sampling quadrat after standing crop removal.

In April 2002 and continuing each April through 2005, Scotts brand Osmocote 8-9 month 69

Slow Release Fertilizer 19-6-12 (NPK; Scotts, Marysville, OH, USA) was applied at 20 g

N m-2 in fertilized plots, well above the Köchy and Wilson (2005) 15 g N m-2 yr-1 threshold necessary to induce a eutrophication effect in grasslands and other habitats. I could not exclude ambient wet/dry atmospheric N deposition, though deposition rates from 1990 to 2005 were relatively low at approximately 1.01 g N m-2 yr-1 at a nearby monitoring site in Lykens (162 km west of my study site), OH, USA, and approximately

0.93 g N m-2 yr-1 at another nearby monitoring site in Mercer Co. (G. K. Goddard site; 96 km east of my study site), PA, USA (US EPA 2005). Within two days of annual mowing of the whole site by the local township with a large tractor and brush hog mower (autumn

2001-2004), litter was removed from litter-removal treatments using a small 23 hp lawn tractor with a pull-behind 8 hp Agri-Fab Mow-N-Vac trailer attachment (Agri-Fab,

Sullivan, IL, USA).

Plant community sampling

A detailed description of plant species richness and diversity, plant biomass, plant community structure, and litter biomass in these plots is presented in Patrick et al. (2008).

Fertilized plots had low species richness and were dominated by Bromus inermis, Poa pratensis, and Dactylus glomerata, while unfertilized plots were mainly dominated by

Festuca arundinacea, Poa trivialis, Anthoxanthum odoratum, and Phleum pratense.

Moreover, unfertilized plots had high species richness of forbs, including Daucus carota, 70

Hieracium aurantiacum, and Ranunculus acris, as well as non-Poaceae graminoids and woody plants.

Beetle community sampling

Beetles were collected using four pitfall traps in each of the 24 experimental plots

2004) and continuing through mid to late August, traps were alternately left open for two weeks and closed for 2 weeks. This resulted in three sampling periods each year during

2002, 2003, and 2005. During 2004, only the second and third sampling periods were collected. When closed, the plastic cups were removed, the contents collected and preserved in 70% EtOH, and the PVC sleeve was tightly capped. While pitfall traps do not capture all beetles in the community, they are an effective sampling technique for determining the relative abundance and species richness of epigeal beetles (Greenslade

1964; Phillips and Cobb 2005). Beetles were identified to species or morphospecies when possible, otherwise to family, and exact numbers within in each trap were recorded, 71

then dried at 70°C for 72 hrs to determine species biomass to the nearest 0.0001 g.

Lacking sufficient numbers captured within a trap, some extremely small species did not register a biomass, and their biomass was recorded as ―0.0000 g.‖

Statistical analyses

I tested the responses of the abundance, biomass, species richness (SR), and effective Shannon’s diversity (eH’, where H’ is the Shannon diversity index) of all epigeal beetles (―Coleoptera‖), carabid and staphylinid beetles, and the six most abundant species to fertilization, plant litter, and the interaction of fertilization and litter. I used eH’ to correct for differences in species richness that might have resulted from differential beetle abundances (Ricklefs and Miller 2000; Haddad et al. 2000). To calculate the average SR within a plot, I summed the total number of beetle species caught in each trap, then averaged this SR for each of the four traps within a plot within a sampling period

(including zeroes for traps where no beetles were captured), then averaged this SR across sampling periods in a year, yielding n = 24 samples within each year. The same method was used to calculate the average abundance, biomass, and eH’ within a plot within a year, also yielding n = 24 samples within a plot within a year.

To analyze trends per year and per treatment in abundance, biomass, SR, and eH’ I used SAS software version 8.01 (SAS Institute, Inc. 1999) to calculate a repeated- measure maximum-likelihood analysis in PROC MIXED with Type III effects based upon the covariance structure of compound symmetry. The various models used the 72

different response variables (biomass, SR, eH’, abundance), and for the predictor variables used fertilized vs. unfertilized, litter removed vs. litter left in situ, year, and their factorial interactions, with year as the repeated predictor.

Results

General beetle responses

8694 beetles representing 172 species in 31 families (Appendix A) were caught during 14784 trap nights over the four years of this study. In the repeated-measures

PROC MIXED analysis, beetle diversity, corrected for abundance with eH’, was significantly affected by fertilization (F1, 80 = 51.85, P < 0.0001) and by year (F3, 80 =

36.38, P < 0.0001), but not by litter (F1, 80 = 2.45, P = 0.1217). As a result, the factorial interaction between fertilization and year was significant (F3, 80 = 10.79, P < 0.0001), but interactions between fertilization and litter (F3, 80 = 2.39, P = 0.1264), litter and year (F3,

80 = 1.65, P = 0.1855), and the fully factorial interaction of fertilization by litter by year

(F3, 80 = 1.53, P = 0.2140) were not significant. Because of the year interactions, I tested for treatment effects within a year (Fig.7) and used SAS software version 8.01 (SAS

Institute, Inc. 1999) to calculate a maximum-likelihood analysis in PROC MIXED with

Type III effects based upon the covariance structure of compound symmetry, eH’ as the response variable, and fertilization, litter, and the factorial interaction of fertilization and litter as predictor variables. During 2002, neither fertilization (F1, 20 = 0.44, P = 0.5157) 73

Figure 7. Average effective Shannon’s H’ (eH’) of beetle species in each treatment; the letter ―a‖ above a year denotes significance at α < 0.05 for fertilization. Open circles (○) and ―+L/-F‖ represent the control treatment plots of unfertilized and litter left in situ, open triangles (∆) and ―-L/-F‖ represent unfertilized and litter removed plots, filled circles (●) and ―+L/+F‖ represent fertilized and litter left in situ plots, and filled triangles

(▲) and ―-L/+F‖ represent fertilized and litter removed plots.

18 a +L/-F 15 -L/-F

H +L/+F

s ' a

n 12 -L/+F

n a

h 9

i t

c 6

f E 3

0 2002 2003 2004 2005 Year 74

nor litter (F1, 20 = 1.07, P = 0.3125) were significant, though their interaction (F1, 20 =

8.76, P = 0.0077) was significant. During 2003, neither fertilization (F1, 20 = 3.82, P =

0.0648), litter (F1, 20 = 0.29, P = 0.5964) nor their interaction (F1, 20 = 1.27, P = 0.2722) were significant. During 2004, fertilization (F1, 20 = 36.01, P < 0.0001) was significant, but neither litter (F1, 20 = 0.27, P = 0.6060) nor their interaction (F1, 20 = 2.77, P = 0.1116) were significant. During 2005, fertilization (F1, 20 = 23.04, P < 0.0001) was significant, but neither litter (F1, 20 = 2.79, P = 0.1107) nor their interaction (F1, 20 = 0.39, P = 0.5381) were significant.

The two dominant families captured were Staphylinidae and Carabidae, with 3881

(44.6% of all beetles captured) specimens from 59 species and 1713 (19.7% of all beetles captured) specimens from 51 species, respectively. In the repeated-measures PROC

MIXED analyses, Coleoptera (all beetles) SR, abundance, and biomass were significantly affected by fertilization and year (Table 5, Fig. 8A–C), and litter significantly affected

Coleoptera abundance and biomass (Table 5, Fig. 8B–C). For Coleoptera, significant interactions were detected, especially for interactions with year (Table 5). Because of the year interactions, I tested for treatment effects within a year and used SAS software version 8.01 (SAS Institute, Inc. 1999) to calculate a maximum-likelihood analysis in

PROC MIXED with Type III effects based upon the covariance structure of compound symmetry, and the predictor variables were SR, abundance, and biomass of all beetles together (Coleoptera) and the two dominant beetle families, Carabidae and Staphylinidae, while the independent variables were fertilization, litter, and their interaction. By 2005, carabid SR, abundance, and biomass responded primarily to litter (Table 6, Fig. 8D–F), 75

Table 5. Results of the repeated-measure PROC MIXED for each response variable.

Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter

(L), year (Y), and (*) their fully factorial interactions.

Response Variable F L Y F*L F*Y L*Y F*L*Y

Coleoptera SR F1, 80 = 62.32 F1, 80 = 3.16 F3, 80 = 37.40 F1, 80 = 0.73 F3, 80 = 13.24 F3, 80 = 2.15 F3, 80 = 2.35 P < 0.0001 P = 0.0794 P < 0.0001 P = 0.3947 P < 0.0001 P = 0.1004 P = 0.0781

Coleoptera F1, 80 = 57.50 F1, 80 = 4.67 F3, 80 = 36.47 F1, 80 = 0.16 F3, 80 = 26.25 F3, 80 = 5.22 F3, 80 = 3.76 abundance P < 0.0001 P = 0.0336 P < 0.0001 P = 0.6928 P < 0.0001 P = 0.0024 P = 0.0140

Coleoptera F1, 80 = 16.64 F1, 80 = 9.54 F3, 80 = 19.15 F1, 80 = 5.33 F3, 80 = 1.49 F3, 80 = 4.65 F3, 80 = 0.67 biomass P = 0.0001 P = 0.0028 P < 0.0001 P = 0.0236 P = 0.2233 P = 0.0048 P = 0.5710

Carabidae SR F1, 80 = 0.37 F1, 80 = 4.02 F3, 80 = 43.39 F1, 80 = 0.46 F3, 80 = 1.83 F3, 80 = 3.11 F3, 80 = 2.29 P = 0.5467 P = 0.0484 P < 0.0001 P = 0.4979 P = 0.1493 P = 0.0308 P = 0.0846

Carabidae F1, 80 = 0.64 F1, 80 = 4.72 F3, 80 = 36.43 F1, 80 = 0.11 F3, 80 = 0.53 F3, 80 = 4.53 F3, 80 = 1.14 abundance P = 0.4267 P = 0.0329 P < 0.0001 P = 0.7390 P = 0.6633 P = 0.0055 P = 0.3393

Carabidae F1, 80 = 17.93 F1, 80 = 8.91 F3, 80 = 23.89 F1, 80 = 5.12 F3, 80 = 2.94 F3, 80 = 5.59 F3, 80 = 0.96 biomass P < 0.0001 P = 0.0038 P < 0.0001 P = 0.0263 P = 0.0380 P = 0.0016 P = 0.4176

Staphylinidae SR F1, 80 = 151.22 F1, 80 = 1.74 F3, 80 = 47.61 F1, 80 = 0.01 F3, 80 = 26.29 F3, 80 = 0.97 F3, 80 = 1.53 P < 0.0001 P = 0.1904 P < 0.0001 P = 0.9077 P < 0.0001 P = 0.4110 P = 0.2128

Staphylinidae F1, 80 = 190.68 F1, 80 = 9.63 F3, 80 = 71.03 F1, 80 = 4.58 F3, 80 = 64.09 F3, 80 = 8.05 F3, 80 = 8.10 abundance P < 0.0001 P = 0.0026 P < 0.0001 P = 0.0353 P < 0.0001 P < 0.0001 P < 0.0001

Staphylinidae F1, 80 = 0.71 F1, 80 = 5.19 F3, 80 = 1.46 F1, 80 = 0.19 F3, 80 = 2.23 F3, 80 = 0.37 F3, 80 = 1.68 biomass P = 0.4013 P = 0.0254 P = 0.2317 P = 0.6661 P = 0.0915 P = 0.7731 P = 0.1783

Table 6. Results of the PROC MIXED for each response variable during 2005. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter (L), and (*) their factorial interaction. Results for 2002–2004 are in Appendix B.

Response Variable F L F*L

Coleoptera SR F1, 20 = 30.91, P < 0.0001 F1, 20 = 3.52, P = 0.0752 F1, 20 = 1.37, P = 0.2551

Coleoptera abundance F1, 20 = 49.21, P < 0.0001 F1, 20 = 8.03, P = 0.0103 F1, 20 = 3.90, P = 0.0623

Coleoptera biomass F1, 20 = 4.15, P = 0.0551 F1, 20 = 7.17, P = 0.0145 F1, 20 = 1.40, P = 0.2498

Carabidae SR F1, 20 = 0.90, P = 0.3553 F1, 20 = 6.60, P = 0.0183 F1, 20 = 1.76, P = 0.2001

Carabidae abundance F1, 20 = 0.04, P = 0.8438 F1, 20 = 5.78, P = 0.0261 F1, 20 = 0.62, P = 0.4385

Carabidae biomass F1, 20 = 5.24, P = 0.0331 F1, 20 = 7.48, P = 0.0127 F1, 20 = 1.95, P = 0.1782

Staphylinidae SR F1, 20 = 81.73, P < 0.0001 F1, 20 = 1.99, P = 0.1737 F1, 20 = 1.67, P = 0.2113

Staphylinidae abundance F1, 20 = 149.43, P < 0.0001 F1, 20 = 14.89, P = 0.0010 F1, 20 = 12.76, P = 0.0019

Staphylinidae biomass F1, 20 = 4.53, P = 0.0458 F1, 20 = 2.54, P = 0.1267 F1, 20 = 1.04, P = 0.3197 77

Figure 8. Species richness, abundance, and biomass of all beetles (A–C), carabids (D–F), and staphylinids (G–I). Definitions of symbols and abbreviations for treatments are given in Figure 7, while the letters above each year denote significance at α < 0.05 for ―a‖

= fertilization, ―b‖ = litter, and ―c‖ = the interaction of fertilization and litter. Full results of the PROC MIXED procedures are given in Table 7 (2005 only) and Appendix B

(2002–2004 only).

A) B) C)

a, b b s 12 +L/-F 35 0.20

s a

)

e e

-L/-F g

n c

30 (

n s

c +L/+F

i a s

r 9 0.15 d

a n

s -L/+F

o i

c a 20

b e

6 0.10

p a

r a s

15 r

e e

a t a a

t c

p r

c p o

e 10 a, c

o t

3 e 0.05

p e

o o

5 o

C o

C 0 0 0.00 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year

D) E) F)

s 3 7 b 0.20 a, b

s b

)

n e

6 g

h c (

c n

i s

a 0.15

s R

d 5

2 n

u m

i b

4 o

0.10

e p

3 e

a S

a a

d d

i a i e 1

c b b a 2

a a, c

a d

r 0.05 i

r c

b a

a 1

r a

C 0 0 0.00 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year

G) H) I)

a s a, b, c

s 6 22 0.012

)

e g

n a

(

n s

c 18

i a

r d

0.009

a n

s 15

u m

e 4

i b

a o

c i

e 12

e p

e 0.006

s a

d a

i 9

e d

n l

2 a i

l y

i 6

n h

i 0.003 h

l a

t a

h 3

S a t 0 0 0.000 S 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year

while staphylinid SR, abundance, and biomass primarily responded to fertilization (Table

6, Fig. 8G–I). The notable exception for staphylinids was abundance, which significantly responded to all predictor variables (Table 6), though this response was only in fertilized plots and the significance of the interaction resulted from the clear separation of fertilized plots with litter left in situ vs. removed (Fig. 8H).

Species richness of all beetles (Coleoptera), carabids, and staphylinids were correlated with their abundance (Fig. 9A–C), though staphylinids showed the clearest response to fertilization (Fig. 9C), which, in turn, drove the same trend for Coleoptera

(Fig. 9A). Biomass and SR for both Coleoptera and carabids were not correlated (Fig.

9D–E), though staphylinid biomass and SR were highly correlated (Fig. 9F). Plant SR and standing crop biomass were inversely related to Coleoptera and staphylinid SR (Fig.

9G, I–J, L), though carabid SR had no relationship with either plant SR (Fig. 9H) or standing crop biomass (Fig. 9K).

Beetle species-level analyses

The most abundant beetle was the staphylinid Meronera venustula Erichson, with

1493 specimens captured (17.2% of all beetles, 38.5% of staphylinids). Virtually absent during the first two years of the study, M. venustula showed a strong response to fertilization with nearly zero specimens captured in unfertilized plots during all years of the study (Fig. 10A; see Appendix G for results of the PROC MIXED analyses). During

2005, M. venustula became the most abundant beetle species and had a significant response to fertilization and litter removal (Fig. 10A). Its absence in unfertilized plots

63 79

Figure 9. Regressions of all beetles, carabids, and staphylinids (left to right) against abundance (A–C), biomass (D–F), plant species richness (G–I), and standing crop biomass (J–L). Symbols are defined in Figure 7, and data presented are for 2005 (data for 2002–2004 are found in Appendices C–F). (*), (**), and (***) indicate significance at

P < 0.0001, P = 0.003, and P = 0.049, respectively, for the R2 values.

A) B) C)

14 4 s

s s

s 8

e n

n 12 R²=78.3%*

n h

h c

i r

i 3

r 6

10 s

R²=85.5%*

e i

e R²=93.3%*

i c

c 8

e p

e 2 4

p s

6 s

a d

e i

t 4

d n

i 2 i

p 1

e h

2 r

C t C 0 0

0 S 0 5 10 15 20 25 30 35 40 45 0 2 4 6 8 10 0 5 10 15 20 25 Coleoptera abundance Carabidae abundance Staphylinidae abundance

D) E) F)

4 s

s 14

s s

s 8

n n

12 n

i r

i 3 R²=16.5%***

r 6

s e

i R²=54.6%*

c c

8 c

e e

R²=0.5% e

2 p 4

s s

e i

t 2

4 d i

1 n

r h

l 2

a p

o 0

C t C 0

0 S 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.0 0.1 0.2 0.3 0.000 0.005 0.010 0.015 0.020 0.025 Coleoptera biomass (g) Carabidae biomass (g) Staphylinidae biomass (g)

G) H) I)

4 s

s s

s 14 8

e n

n h

n 12

h c

r i

i 3

r 6

i c

c c

8 e

e p

e 2 4

p s

R²=34.2%** s s

r a

R²=0.0% d

i e

4 d R²=47.3%*

t i

1 n 2

p l

r h

e 2

a p

a C t 0 C 0 0 0 5 10 15 20 0 5 10 15 20 S 0 5 10 15 20 Plant species richness Plant species richness Plant species richness

J) K) L)

4 s

s s

s 14 8

s e

n h

n 12

c r

i 3

r 6

s e

c e

c 8

e p

e 2 4

s s

6 R²=6.9%

a a

e d

4 i t

1 n 2 R²=71.5%*

b l

y a

o R²=51.6%*

r h

e 2

a C t 0 C 0

0 S 0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200 0 300 600 900 1200 Standing crop biomass (g / m²) Standing crop biomass (g / m²) Standing crop biomass (g / m²)

63 81

Figure 10. Average abundance of selected species by year. Definitions of symbols and abbreviations for treatments are given in Figure 7, while the letters above each year denote significance at α < 0.05 for ―a‖ = fertilization, ―b‖ = litter, and ―c‖ = the interaction of fertilization and litter. Full results of the PROC MIXED procedures are given in Appendix G.

A) Meronera venustula B) Mocyta fungi

a 14 3 +L/-F a, b, c 12 -L/-F

10 +L/+F 2

e c

c -L/+F n

n 8

n n

6 u

u b

b 1 A A 4 a 2 a 0 0 2002 2003 2004 2005 2002 2003 2004 2005 Year Year

C) Myrmedonota aidanii D) Tomoderus constrictus

5 1.6 a

1.2

e c

c 3

n a

a 0.8

n u

u 2

b A A 0.4 1

0.0 0 2002 2003 2004 2005 2002 2003 2004 2005 Year Year

E) Poecilus lucublandus F) Ptiliidae

2.0 3.5 a 3.0 a

1.5 2.5

e c

a, b c n

n 2.0 a

1.0 a

d n

n 1.5

u b

b a A 0.5 A 1.0 0.5 a 0.0 0.0 2002 2003 2004 2005 2002 2003 2004 2005 Year Year

63 83

and its response to litter removal caused a significant interaction between fertilization and litter (see Appendix G).

The fifth most abundant species was the parthenogenetic and exotic staphylinid

Mocyta fungi (Gravenhorst), with 471 specimens captured (5.4% of all beetles, 12.1% of staphylinids). Like M. venustula, Mo. fungi was virtually absent during the first two years of the study, then became very abundant only in fertilized plots (Fig. 10B). Indeed, through four years of sampling only five specimens were ever captured in unfertilized plots (Appendix A). These two species accounted for nearly a quarter of all beetles captured and over half of all staphylinids captured, yet their abundance was largely restricted to the last two years of the study and almost exclusively in fertilized plots.

With the sixth most abundant species (Fig. 10C), the newly described staphylinid

Myrmedonota aidani Maruyama and Klimaszewski (Maruyama et al. 2008), over a quarter of all beetles captured and 61.5% of all staphylinids were in only 3 species, though the latter did not have any significant responses to treatments (see Appendix G).

The second most abundant beetle was Tomoderus constrictus (Say), with 981 specimens captured (11.3% of all beetles; Fig. 10D). This was the only species in this family (Anthicidae) captured, and over half (547 specimens) were captured during a single year (2003) and had a significant response to fertilization in which the species was highly abundant in unfertilized plots (Fig. 10D; see Appendix G for results of the PROC

MIXED analyses). An additional third (319 specimens) were caught only during the first year (2002), resulting in 88.3% of this species’ captures occurred during only the first two years of the study. 84

The carabid beetle Poecilus lucublandus was the fourth most abundant species with 485 specimens captured (5.6% of all beetles, 28.3% of carabids), but did not have a significant response to any treatments (Fig. 10E; see Appendix G for results of the PROC

MIXED analyses). The tiny and poorly studied feather-winged beetles (family Ptiliidae) were the third most abundant family and the fourth most abundant ―species,‖ with 595 specimens captured (6.8% of all beetles). The tiniest of all beetles, they significantly responded to fertilization during all years of the study (Fig. 10F; Appendix G), though their abundance was highest during the last two years of the study; during the four years of sampling only five specimens were captured in unfertilized plots (Appendix A). It is unfortunate that these beetles are so poorly known as their diversity was striking (LBP, pers. obs.), though they could not be sorted confidently even to morphospecies.

Discussion

I found that beetle diversity increased with increasing beetle abundance in fertilized plots, and therefore my first hypothesis was not supported. The result is contrary to previous studies in which arthropod diversity decreased as abundance increased (e.g., Knops et al. 1999; Haddad et al. 2000, 2001). While beetle abundance and species richness were highest in fertilized plots, beetle biomass was highest in unfertilized plots, particularly those with litter removed. This dichotomy is likely driven by the differences in the size of the species captured– the most abundant species in fertilized plots, most notably the staphylind species M. venustula and Mo. fungi, are very 85

small (circa 2 mm), whereas those in unfertilized plots, particularly the carabid species P. lucublandus, are quite large (circa 12 mm) with a biomass more than 200 times that of the aforementioned abundant yet small staphylinids.

Despite the abundance of M. venustula and Mo. fungi during the latter two years of the study, eH’ rose dramatically in fertilized plots. My measure of diversity, eH’, normalizes species richness by abundance (Ricklefs and Miller 2000; Haddad et al.

2000). Thus, even if several species are caught, a single, dominant species in a sample would decrease eH’ to the point of approximately equating it with less speciose yet more evenly distributed samples. The increased diversity (eH’) in fertilized plots indicated that as the abundance of M. venustula and Mo. fungi increased, the number of other species and the abundance of these additional species also increased, a result opposite to work that sampled the aerial arthropod community (e.g., Haddad et al. 2001). Moore et al.

(2004) posited that the detritus-based community may be independent of the living-plant based portion of the food web. Here, I provide empirical support for that hypothesis.

Also in contrast with previous studies, I reject my second and third hypotheses because beetle diversity increased as plant species richness decreased and plant biomass increased. While previous theory and studies have documented that the diversity of herbivores and predators is dependant on plant community diversity (Hutchinson 1959;

Root 1973; Hunter and Price 1992; Knops et al. 1999), these studies did not consider the epigeal community. Rather, they focused on the portion of the food web most closely tied to living plants, where increases in plant diversity resulted in increased specialist herbivore diversity, driving up overall arthropod species diversity. However, a 86

significant portion of the diets of my most abundant beetles, carabids and staphylinds, is comprised of detritivores (Sunderland 1975; Chiverton and Sotherton 1991; Snyder and

Wise 2001). Thus, it would seem that these epigeal and generalist-predator beetles were responding to changes in the community only indirectly tied to the living plant community. This disjunct between the living-plant and detritus-based arthropod community compartments underscores the need to better understand the detritus-based community.

Finally, despite documented effects of increased habitat complexity on arthropod abundances and diversity (e.g., Lawton 1983; Halaj et al. 2000), my results do not support my fourth hypothesis. By the fourth year of this study, beetle diversity was highest in my litter removal treatments. Beetle diversity increased in fertilized plots, but was highest in litter-removed plots. Rather than moderately increasing beetle diversity, litter moderately decreased beetle diversity in both fertilized and unfertilized plots.

Perhaps plant litter and vegetation impeded the movement of carabids and staphylinids and possibly created unsuitable habitat for larger-bodied species (see Honek 1988). This habitat-structure effect on epigeal beetles seemed to change the predator guild composition to favor smaller-bodied epigeal beetles, potentially altering the abundance and diversity of their detritivorous prey and releasing some prey from predation pressure

(Halaj and Wise 2002). While my study did not address this portion of the food web, nutrient regeneration and recycling through the ecosystem could be significantly altered by these changes in the epigeal predator community. 87

Staphylinid beetles were most sensitive to my treatments. Compared with carabids, staphylinids may respond to finer scale changes in habitat quality (Pohl et al.

2007). Based on my results, staphylinids are the better candidate family for assessing overall epigeal arthropod responses to nutrient loading and plant litter. Carabids were significantly sensitive to plant litter, though this result was likely weighted by the strong response of carabids to litter removal in unfertilized plots, while staphylinid responses to litter were obvious in fertilized plots. Staphylinid species richness and abundance were low in unfertilized plots, and no detectable response to litter was indicated in unfertilized plots. However, fertilization elicited a strong response from staphylinids, and I suggest that considering a species-level taxonomic resolution within this family will best represent the general response of the entire epigeal beetle community.

Surprisingly and not previously documented, ptiliids actually responded strongly to fertilization. These tiny, poorly known beetles are generally found in leaf litter and rotting organic matter (Sörensson 2003). In future studies, this family may also prove to be a favorable indicator of epigeal beetle responses to nutrient loading. The anthicid T. constrictus had a significant negative response to fertilization during 2003, its year of highest abundance. The one year peak in abundance indicates that 2003 was likely an outbreak year, though I am unsure of what environmental controls act on this species to promote such instances and I was unable to detect any community changes (e.g., crashes or peaks in other epigeal beetle species) induced by this outbreak. Given that T. constrictus did not significantly respond to my treatments during other years of the study, the sensitivity of this species to my treatments during non-outbreak years is questionable. 88

The diversity and community structure of beetles and other arthropods have been shown to be sensitive to plot size (Martinko et al. 2006). The large size of my experimental plots integrated important determinants of the within-plot plant communities, including spatial heterogeneity (De Boeck et al. 2006), leaching of nutrients from litter (Berendse 1998), local nutrient cycling (Hooper and Vitousek 1998), and the translocation of nutrients within clumping and clonal plants (Hutchings and

Bradbury 1986), which are the primary growth forms of my dominant graminoids

(Patrick et al. 2008). These spatial factors are also important to epigeal beetles because of their vagility and their need to find suitable food; the larger plot sizes more realistically emulate natural habitat patches of varying quality and can support higher insect diversity (Martinko et al. 2006). Other studies that examined the effects of nutrient loading on arthropod communities had plot sizes ranging from 9 m2 – 169 m2 (e.g.,

Knops et al. 1999; Haddad et al. 2001), making my experimental plots (314 m2) nearly twice as large—an important factor when considering the vagility of some beetle species.

However, I realize that my study has some distinct differences when compared to previous work.

My use of an NPK fertilizer, as opposed to N-only fertilizer, is likely to have induced a stronger response to fertilization due to the added P and K. Nevertheless, my plant results (see Patrick et al. 2008) were generally consistent with other plant studies that used NPK fertilizers (e.g., Carson and Barrett 1988; Turkington et al. 2002) and N- only fertilizers (e.g., Haddad et al. 2000; Tilman et al. 2002), which allowed us to formulate my epigeal beetle hypotheses on the same bases as previous studies that 89

investigated the responses of arthropods to nutrient loading. Further, my running definition of litter (see Methods) includes the vegetation mown in the previous year and not removed from litter left in situ treatment plots, potentially altering the nutritional quality of the litter relative to naturally senesced vegetation, and the physical structure of the litter as it lay after mowing (e.g., Semmartin et al. 2004). Because the timing of the mowing was determined by the local township, litter from the annual mowing accumulated earlier than might normally be expected for this region of the USA.

However, were the mowing to stop, the site would very quickly yield to encroaching woody vegetation typical of early secondary succession.

I based my hypotheses on previous studies that investigated the aerial arthropod community and, surprisingly, I rejected all of my hypotheses. Epigeal beetles, particularly staphylinids, responded opposite to predictions based on the results from the aerial portion of the food web in which arthropod diversity decreased with the decreased plant species richness and increased plant biomass as a result of increased nutrient loading. Rather, epigeal beetle diversity increased with the decreased plant species richness and increased plant biomass that resulted from increased nutrient loading

(Patrick et al. 2008). Moreover, plant litter effects have been largely ignored in biodiversity-productivity studies (Patrick et al. 2008), and my experimental design has integrated these effects into an investigation of the epigeal beetle community. This study expands our knowledge of arthropod responses to nutrient loading by considering a portion of the food web more closely tied to detritus. This is an understudied portion of the food web and a poorly understood component of ecosystems. My results clearly 90

emphasize that the epigeal community requires more investigation and integration into conventional biodiversity-productivity theory.

References

Berendse, F. 1998. Effects of dominant plant species on soil during succession in

nutrient-poor ecosystems. Biogeochemistry 42:73–88.

Bobbink, R., M. Hornung, and J. G. M. Roelofs. 1998. The effects of air-borne nitrogen

pollutants on species diversity in natural and semi-natural European vegetation.

Journal of Ecology 86:717–738.

Buse, A., and J. E. G. Good. 1993. The effects of conifer forest design and management

on abundance and diversity of rove beetles (Coleoptera: Staphylinidae):

implications for conservation. Biological Conservation 64:67–76.

Carson, W. P. and G. W. Barrett. 1988. Succession in old-field plant communities:

effects of contrasting types of nutrient enrichment. Ecology 69:984–994.

Chiverton, P. A., and N. W. Sotherton. 1991. The effect on beneficial arthropods of the

exclusion of herbicides from cereal crop edges. Journal of Applied Ecology

28:1027–1039.

De Boeck, H. J., I. Nijs, C. M. H. M. Lemmens, and R. Ceulemans. 2006. Underlying

effects of spatial aggregation (clumping) in relationships between plant diversity

and resource uptake. Oikos 113:269–278.

Fenn, M. E., J. S. Baron, E. B. Allen, H. M. Rueth, K. R. Nydick, L. Geiser, W. D.

Bowman, J. O. Sickman, T. Meixner, D. W. Johnson, and P. Neitlich. 2003.

Ecological effects of nitrogen deposition in the western United States.

BioScience 53:404–420. 92

Galloway, J. N., J. D. Aber, J. W. Erisman, S. P. Seitzinger, R. W. Howarth, E. B.

Cowling, and B. J. Cosby. 2003. The nitrogen cascade. BioScience 53:341–355.

Greenslade, P. J. M. 1694. Pitfall trapping as a method for studying populations of

Carabidae (Coleoptera). Journal of Animal Ecology 33:301–310.

Haddad, N. M., J. Haarstad, and D. Tilman. 2000. The effects of long-term nitrogen

loading on grassland insect communities. Oecologia 124:73–84.

Haddad, N. M., D. Tilman, J. Harstaad, M. Ritchie, and J. M. H. Knops. 2001.

Contrasting effects of plant richness and composition on insect communities: a

field experiment. American Naturalist 158:17–35.

Halaj, J., A. B. Cady, and G. W. Uetz. 2000. Modular habitat refugia enhance generalist

predators and lower plant damage in soybeans. Environmental Entomology

29:383–393.

Halaj, J. and D. H. Wise. 2002. Impact of a detrital subsidy on trophic cascades in a

terrestrial grazing web. Ecology 83:3141–3151.

Hättenschwiler, S., and P. Gasser. 2005. Soil animals alter plant litter diversity effects

on decomposition. Proceedings of the National Academy of Sciences of the USA

102:1519–1524.

Hector, A., B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G.

Dimitrkopoulos, J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P.

Högberg, K. Huss-Danell, J. Joshi, A. Jumpponen, C. Körner, P. W. Leadley, M.

Loreau, A. Minns, C. P. H. Mulder, G. O’Donovan, S. J. Otway, J. S. Pereira, A.

Prinz, D. J. Read, M. Scherer-Lorenzen, E-D. Schulze, A-S. D. Siamantziouras, 93

E. M. Spehn, A. C. Terry, A. Y. Troumbis, F. I. Woodward, S. Yachi, J. H.

Lawton. 1999. Plant diversity and productivity experiments in European

grasslands. Science 286:1123–1127.

Honek, A. 1988. The effect of crop density and microclimate on pitfall catches of

Carabidae, Staphylinidae (Coleoptera) and Lycosidae (Araneae) in cereal fields.

Pedobiologia 32:233–242.

Hooper, D. U. and P. M. Vitousek, 1998. Effects of plant competition and diversity on

nutrient cycling. Ecological Monographs 68:121–149.

Hunter, M. D., and P. W. Price. 1992. Playing chutes and ladders: heterogeneity and the

relative roles of bottom-up and top-down forces in natural communities. Ecology

73:724–732.

Hutchings, M. J. and I. K. Bradbury. 1986. Ecological perspectives on clonal perennial

plants. BioScience 36:178–182.

Hutchinson, G. 1959. An homage to Santa Rosalia, or why are there so many kinds of

animals? American Naturalist 93:145–159.

Jonas, J. L., M. R. Whiles, and R. E. Charlton. 2002. Aboveground invertebrate

responses to land management differences in a central Kansas grassland.

Environmental Entomology 31:1142–1152.

Knops, J. M. H., D. Tilman, N. M. Haddad, S. Naeem, C. E. Mitchell, J. Harstaad, M. E.

Ritchie, K. M. Howe, P. B. Reich, E. Siemann, and J. Groth. 1999. Effects of

plant species richness on invasion dynamics, disease outbreaks, insect

abundances, and diversity. Ecology Letters 2:286–293. 94

Köchy, M., and S. D. Wilson. 2005. Variation in nitrogen deposition and available soil

nitrogen in a forest-grassland ecotone in Canada. Landscape Ecology 20:191–

202.

Larsen, T. H., N. Williams, and C. Kremen. 2005. Extinction order and altered

community structure rapidly disrupt ecosystem functioning. Ecology Letters

8:538–547.

Lawton, J. H. 1983. Plant architecture and the diversity of phytophagous insects.

Annual Review of Entomology 28:23–29.

Long, Z. T., C. L. Mohler, and W. P. Carson. 2003. Extending the resource

concentration Hypothesis to plant communities: effects of litter and herbivores.

Ecology 84:652–665.

Loreau, M. 1993. Ground beetles in a changing environment: determinants of species

diversity and community assembly. In: Boyle, T. J. B., and C. E. B. Boyle, eds.,

Biodiversity, Temperate Ecosystems, and Global Change. Springer-Verlag,

Berlin, Germany, pp. 77–98.

Martinko, E. A., R. H. Hagen, and J. A. Griffith. 2006. Successional change in the insect

community of a fragmented landscape. Landscape Ecology 21:711–721.

Maruyama, M., L. B. Patrick, and J. Klimaszewski. 2008. First record of the genus

Myrmedonota Cameron (Coleoptera, Staphylinidae) from North America, with

descriptions of two new species. Zootaxa 1716:35–43.

McCann, K. S. 2000. The diversity-stability debate. Nature 405:228–233. 95

Moore, J. C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C.

Johnson, K. S. McCann, K. Melville, P. J. Morin, A. D. Rosemond, D. M. Post, J.

L. Sabo, K. M. Scow, M. J. Vanni, and D. H. Wall. 2004. Detritus, trophic

dynamics, and biodiversity. Ecology Letters, 7:584–600.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008. Large-scale manipulation of

plant litter and fertilizer in a managed successional temperate grassland. Plant

Ecology 197:183–195.

Phillips, I. D., and T. P. Cobb. 2005. Effect of habitat structure and lid transparency on

pitfall catches. Environmental Entomology 34:875–882.

Pohl, G. R., D. W. Langor, and J. R. Spence. 2007. Rove beetles and ground beetles

(Coleoptera: Staphylinidae, Carabidae) as indicators of harvest and regeneration

practices in western Canadian foothills forests. Biological Conservation 137:294–

307.

Ricklefs, R. E., and G. L. Miller. 2000. Ecology. Freeman, New York, NY, USA.

Ritchie, A., and J. R. Steiger. 1974. Soil Survey of Summit County Ohio. United States

Department of Agriculture, Soil Conservation Service in cooperation with Ohio

Department of Natural Resources, Division of Lands and Soil and the Ohio

Agricultural Research and Development Center.

Root, R. B. 1973. Organization of a plant-arthropod association in simple and diverse

habitats: the fauna of collards (Brassica oleraceae). Ecological Monographs

43:95–124. 96

SAS Institute, Inc. 1999. SAS System for Windows. SAS Institute, Inc., Cary, NC,

USA.

Semmartin, M., M. R. Aguiar, R. A. Distel, A. S. Moretto, and C. M. Ghersa. 2004.

Litter quality and nutrient cycling affected by grazing-induced species

replacements along a precipitation gradient. Oikos 107:148–160.

Snyder, W. E., and D. H. Wise. 2001. Contrasting trophic cascades generated by a

community of generalist predators. Ecology 86:1571–1583.

Sörensson, M. 2003. New records of featherwing beetles (Coleoptera: Ptiliidae) in North

America. Coleopterists Bulletin 57:369–381.

Strong, D. R., J. H. Lawton, and R. Southwood. 1984. Insects on plants: community

patterns and mechanisms. Harvard University Press, Cambridge, MA, USA.

Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.

Milchunas, and S. Pennings. 2005. Functional- and abundance-based

mechanisms explain diversity loss due to N fertilization. Proceedings of the

National Academy of Sciences of the USA 102:4387–4392.

Sunderland, K. D. 1975. The diet of some predatory arthropods in cereal crops. Journal

of Applied Ecology 12:507–515.

Tilman, D., J. Knops, D. Wedin, P. Reich. 2002. Experimental and observational studies

of diversity, productivity, and stability. In: Kinzig, A. P., S. W. Pacala, and D.

Tilman, eds., The Functional Consequences of Biodiversity. Princeton University

Press, Princeton, NJ, USA, pp. 42–70. 97

Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of

fertilization and herbivory on the herbaceous vegetation of the boreal forest in

north-western Canada: a 10-year study. Journal of Ecology 90:325–337.

US EPA. 2005. Clean air status and trends network (CAST-NET).

http://www.epa.gov/castnet/

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human

domination of Earth’s ecosystems. Science 277:494–499.

Appendix A. Total numbers of each family and species of beetles captured during the four year manipulative experiment. ―+L/-F‖ represents the control treatment of unfertilized plots with litter left in situ, ―-L/-F‖ represents the unfertilized with litter removed treatment, ―+L/+F‖ represents the fertilized with litter left in situ treatment, ―-

L/+F‖ represents the fertilized with litter removed treatment, and ―Total‖ represents the total number caught.

Family/Species +L/-F -L/-F +L/+F -L/+F

Anthicidae Tomoderus constrictus (Say) 373 341 99 168

Byrrhidae 6 5 7 16

Calydiidae 0 1 0 0

Cantharidae 1 1 0 1

Carabidae 359 538 338 478 Agonum ferreum Haldeman 0 0 1 0 Agonum gratiosum Mannerheim 1 1 21 26 Agonum metallescens (LeConte) 0 1 0 0 Agonum nutans (Say) 17 20 7 6 Agonum palustre Goulet 0 0 0 1 Agonum punctiforme (Say) 0 0 6 11 Amara familiaris (Duftschmid) 0 1 0 17 Amara flebilis (Casey) 3 5 12 55 Amara lunicollis Schiodte 32 8 24 66 Amara pallipes Kirby 5 22 22 40 Amphasia interstitialis (Say) 0 0 2 0 Amphasia sericea (T.W. Harris) 1 3 1 1 Anisodactylus harrisii LeConte 0 0 0 1 Anisodactylus nigerrimus (Dejean) 1 1 1 0 Anisodactylus rusticus Say 0 0 1 0 Badister notatus Haldeman 1 0 0 0 Bembidion affine Say 0 1 1 4 Bembidion impotens Casey 0 0 0 1 Bembidion mimus Hayward 0 0 0 1 Bembidion rapidum (LeConte) 0 0 5 5 Bembidion transparens (Gebler) 0 0 1 0 Chlaenius lithophilus Say 0 1 0 0 Chlaenius pusillus Say 0 0 0 1 Chlaenius tricolor Dejean 12 5 2 0 Cincindela punctulata Oliver 0 1 0 0 Colliuris pensylvanica (L.) 0 0 0 1 Cyclotrachelus sodalis (LeConte) 19 140 5 10 Dicealus elongatus Bonelli 17 30 10 4

100

(Appendix A continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Carabidae continued

Dyschirius globulosus (Say) 0 0 0 1 1 Harpalus compar LeConte 0 1 0 0 1 Harpalus fulgens Csiki 0 3 3 2 8 Harpalus herbivagus Say 0 1 0 1 2 Harpalus longicollis LeConte 1 1 0 0 2 Harpalus pennsylvanicus (DeGeer) 1 2 1 0 4 Harpalus somnulentus Dejean 3 9 3 5 20 Ophonus puncticeps (Stephens) 1 8 0 1 10 Oxypselaphus pusillus (LeConte) 0 0 3 6 9 Paratachys pumilus Dejean 3 2 6 2 13 Poecilus lucublandus (Say) 162 96 121 106 485 Pterostichus atratus (Newman) 46 107 26 54 233 Pterostichus caudicalis (Say) 0 0 1 0 1 Pterostichus commutabilis (Motschulsky) 7 7 2 8 24 Pterostichus femoralis (Kirby) 3 0 28 12 43 Pterostichus luctuosus (Dejean) 0 2 0 0 2 Pterostichus melanarius (Illiger) 1 0 0 0 1 Pterostichus permundus (Say) 7 18 11 9 45 Pterostichus stygicus (Say) 2 1 2 3 8 Scarites quadriceps Chaudoir 0 8 0 0 8 Scarites subterraneus Fabricius 11 17 0 3 31 Selenophorus opalinus (LeConte) 0 1 0 0 1 Stenolophus ochropezus (Say) 2 14 9 14 39

Chrysomelidae 4 0 4 3 11 Charidotella bicolor (Fabricius) 0 0 1 0 1 Chrysomelidae 4 0 3 3 10

Coccinellidae 3 4 5 6 18

Colydiidae 0 1 0 0 1

Cryptophagidae 10 9 84 114 217

101

(Appendix A continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Curculionidae 31 71 137 48 287 Barypeithes pellucidus (Boheman) 25 57 130 33 245 Sphenophorus sp. 0 1 0 0 1 Otiorhynchus rugosostriatus (Goeze) 0 0 1 0 1 Phyxelis rigidus (Say) 2 1 3 2 8 Sphenophorus minimus Hart 0 5 1 0 6 Sphenophorus parvulus Gyllenhal 3 5 1 9 18 Sphenophorus zeae Walsh 0 2 1 4 7 Tychius picirostris (Fabricius) 1 0 0 0 1

Dytiscidae 0 0 1 1 2 Copelatus glyphicus (Say) 0 0 0 1 1 Ilybius biguttulus (Germar) 0 0 1 0 1

Elateridae 38 125 8 22 193 Aeolus amabilis (LeConte) 5 8 0 0 13 Agriotes mancus (Say) 0 0 1 1 2 Elateridae 33 116 7 21 177 Hemicrepidius bilobatus (Say) 0 1 0 0 1

Eucinetidae 0 0 0 1 1

Histeridae 18 0 6 8 32 Euspilotus assimilis (Paykull) 1 0 0 0 1 Histeridae 17 0 5 7 29 Margarinotus egregius (Casey) 0 0 1 0 1 Margarinotus immunis (Erichson) 0 0 0 1 1

Hydrophilidae 0 3 45 27 75 Anacaena limbata (Fabricius) 0 0 0 1 1 Cercyon connivens Fall 0 0 0 1 1 Cercyon occalatus (Say) 0 0 36 9 45 Cercyon praetextatus (Say) 0 1 4 7 12 Enochrus perplexus (LeConte) 0 1 0 0 1 Helophorus linearis LeConte 0 1 0 0 1 Helophorus marginicollis Smetana 0 0 1 1 2

102

(Appendix A continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Hydrophilidae continued

Paracymus subcupreus (Say) 0 0 0 1 1 Tectosternum naviculare (Zimmermann) 0 0 4 7 11

Lampyridae 1 3 1 0 5

Languriidae 0 1 7 3 11

Latridiidae 43 63 53 41 200

Leiodidae 8 7 6 6 27 Leiodidae 7 7 6 1 21 Ptomophagus sp. 1 1 0 0 5 6

Monotomidae 0 0 9 8 17

Nitidulidae 0 1 2 1 4

Phalacridae 8 6 84 78 176

Ptiliidae 3 2 330 260 595

Scarabaeidae 18 27 33 18 96 Aphodius ruricola Melsheimer 5 13 17 3 38 Onthophagus hecate (Panzer) 10 13 13 13 49 Onthophagus orpheus canadensis (Fabricius) 0 0 1 0 1 Phyllophaga futilis (LeConte) 3 1 1 0 5 Popillia japonica (Newman) 0 0 1 2 3

Scydmaenidae 5 5 1 2 13

Silphidae 20 24 7 7 58 Necrophila americana (L.) 17 12 3 3 35 Nicrophorus marginatus Fabricius 0 1 1 2 4 Nicrophorus orbicollis Say 3 8 3 1 15 Nicrophorus tomentosus Weber 0 1 0 0 1

103

(Appendix A continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Silphidae continued

Oiceoptoma inaequale (Fabricius) 0 2 0 0 2 Oiceoptoma noveboracense (Forster) 0 0 0 1 1

Silvanidae 0 1 8 12 21 Telephanus velox (Haldeman) 0 1 8 12 21

Sphindidae 1 0 0 0 1

Staphylinidae 237 337 1366 1941 3881 Acrotona sp. 1 20 27 14 26 87 Acrotona sp. 2 0 1 4 6 11 Acrotona sp. 3 0 10 2 1 13 Acylophorus pronus Erichson 0 0 0 1 1 Aleochara curtula (Goeze) 17 3 3 10 33 Aleochara gracilicornis Bernhauer 0 0 17 11 28 Aleochara sculptiventris (Casey) 0 0 1 0 1 Aleochara sp. 1 0 0 1 0 1 Aleocharini sp. 0 0 0 1 1 Amischa analis (Gravenhorst) 0 0 8 2 10 Amischa sp. 1 9 4 14 10 37 Amischa sp. 2 0 0 0 1 1 Apocellus sphaericollis (Say) 9 1 0 4 14 Atheta aemula (Erichson) 0 0 1 0 1 Atheta sp. 1 1 0 0 2 3 Atheta sp. 2 2 15 6 3 26 Atheta sp. 3 0 0 0 2 2 Athetini sp. 1 0 0 2 0 2 Athetini sp. 2 0 0 0 1 1 Autalia rivularis (Gravenhorst) 0 0 0 1 1 Autalia sp. 1 0 0 2 1 3 Bolitobius sp. 1 0 0 1 0 1 Calodera sp. 1 0 0 1 0 1 Falagria dissecta Erichson 0 0 1 5 6 Homaeotarsus pallipes (Gravenhorst) 2 4 8 1 15 Liogluta sp. 1 0 0 0 2 2

104

(Appendix A continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Staphylinidae continued

Meotica sp. 1 0 0 1 0 1 Meronera venustula Erichson 15 31 555 892 1493 Mocyta fungi (Gravenhorst) 1 4 167 299 471 Mycetoporus sp. 1 0 0 0 1 1 Myrmedonota aidani Maruyama and Klimaszewski 64 101 130 129 424 Oxypoda perexilis Casey 4 7 54 52 117 Paederinae 2 9 18 23 52 Paederus iowensis Casey 44 51 89 125 309 Philhygra clemens (Casey) 1 0 4 3 8 Philonthus asper Horn 0 1 47 67 115 Philonthus carbonarius (Gravenhorst) 0 0 2 5 7 Philonthus cognatus Stephens 0 0 2 14 16 Philonthus lomatus Erichson 4 5 73 70 152 Philonthus monaeses Smetana 0 0 1 1 2 Philonthus sericans (Gravenhorst) 0 0 30 45 75 Philonthus sericinus Horn 0 0 0 4 4 Philonthus validus Casey 0 0 1 1 2 Platydracus immaculatus (Mannerheim) 5 4 1 0 10 Platydracus maculosus (Gravenhorst) 0 10 3 8 21 Platydracus mysticus (Erichson) 8 18 4 4 34 Platydracus zonatus (Gravenhorst) 0 2 1 1 4 Pselaphinae 6 3 1 2 12 Quedius capucinus (Gravenhorst) 0 0 2 0 2 Quedius laticollis Gravenhorst 3 4 5 8 20 Scaphidiinae 3 1 2 6 12 Scopaeus sp. 1 0 0 0 2 2 Staphylinidae 9 14 57 57 137 Steninae 6 1 3 8 18 Stenus annularis Erichson 0 0 1 1 2 Stethusa dichroa (Gravenhorst) 0 1 3 0 4 Tachyporinae 1 1 16 11 29 Tachyporus elegans Horn 1 3 7 11 22 Tasgius melanarius Heer 0 1 0 0 1

105

(Appendix A continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Tenebrionidae 1 1 0 1 3 Neatus tenebrioides Beauv. 1 1 0 1 3

Throscidae 3 14 0 0 17

106

Appendix B. Results of the PROC MIXED analysis for each response variable during

2002–2004. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter (L), and (*) their factorial interaction.

2002

Response Variable F L F*L

Coleoptera SR F1, 20 = 0.86, P = 0.3660 F1, 20 = 1.01, P = 0.3275 F1, 20 = 5.12, P = 0.0349

Coleoptera abundance F1, 20 = 0.00, P = 0.9601 F1, 20 = 0.08, P = 0.7769 F1, 20 = 0.53, P = 0.4760

Coleoptera biomass F1, 20 = 2.80, P = 0.1098 F1, 20 = 1.65, P = 0.2137 F1, 20 = 5.77, P = 0.0261

Carabidae SR F1, 20 = 0.00, P = 1.000 F1, 20 = 0.93, P = 0.3464 F1, 20 = 7.59, P = 0.0122

Carabidae abundance F1, 20 = 0.13, P = 0.7256 F1, 20 = 0.69, P = 0.4161 F1, 20 = 5.82, P = 0.0256

Carabidae biomass F1, 20 = 1.27, P = 0.2725 F1, 20 = 2.05, P = 0.1673 F1, 20 = 5.84, P = 0.0253

Staphylinidae SR F1, 20 = 2.59, P = 0.1231 F1, 20 = 0.77, P = 0.3904 F1, 20 = 0.34, P = 0.5649

Staphylinidae abundance F1, 20 = 0.22, P = 0.6474 F1, 20 = 0.02, P = 0.8958 F1, 20 = 0.06, P = 0.8103

Staphylinidae biomass F1, 20 = 0.13, P = 0.7207 F1, 20 = 1.04, P = 0.3192 F1, 20 = 1.84, P = 0.1901

107

(Appendix B continued)

2003

F L F*L

Coleoptera SR F1, 20 = 4.82, P = 0.0401 F1, 20 = 0.45, P = 0.5117 F1, 20 = 2.39, P = 0.1378

Coleoptera abundance F1, 20 = 0.04, P = 0.8393 F1, 20 = 0.20, P = 0.6591 F1, 20 = 1.59, P = 0.2225

Coleoptera biomass F1, 20 = 10.38, P = 0.0043 F1, 20 = 2.40, P = 0.1369 F1, 20 = 5.61, P = 0.0281

Carabidae SR F1, 20 = 0.87, P = 0.3612 F1, 20 = 0.12, P = 0.7297 F1, 20 = 3.07, P = 0.0951

Carabidae abundance F1, 20 = 0.98, P = 0.3350 F1, 20 = 0.31, P = 0.5846 F1, 20 = 3.43, P = 0.0788

Carabidae biomass F1, 20 = 8.40, P = 0.0089 F1, 20 = 1.88, P = 0.1855 F1, 20 = 4.76, P = 0.0412

Staphylinidae SR F1, 20 = 10.74, P = 0.0038 F1, 20 = 0.32, P = 0.5768 F1, 20 = 0.67, P = 0.4222

Staphylinidae abundance F1, 20 = 14.30, P = 0.0012 F1, 20 = 1.10, P = 0.3060 F1, 20 = 0.52, P = 0.4797

Staphylinidae biomass F1, 20 = 0.69, P = 0.4173 F1, 20 = 1.43, P = 0.2451 F1, 20 = 2.10, P = 0.1624

108

(Appendix B continued)

2004

F L F*L

Coleoptera SR F1, 20 = 42.75, P < 0.0001 F1, 20 = 0.55, P = 0.4650 F1, 20 = 1.28, P = 0.2714

Coleoptera abundance F1, 20 = 67.53, P < 0.0001 F1, 20 = 2.14, P = 0.1590 F1, 20 = 1.91, P = 0.1825

Coleoptera biomass F1, 20 = 16.42, P = 0.0006 F1, 20 = 0.03, P = 0.8702 F1, 20 = 0.12, P = 0.7290

Carabidae SR F1, 20 = 2.82, P = 0.1088 F1, 20 = 0.09, P = 0.7702 F1, 20 = 0.24, P = 0.6269

Carabidae abundance F1, 20 = 3.38, P = 0.0807 F1, 20 = 0.35, P = 0.5589 F1, 20 = 0.23, P = 0.6356

Carabidae biomass F1, 20 = 20.31, P = 0.0002 F1, 20 = 0.00, P = 0.9468 F1, 20 = 0.31, P = 0.5862

Staphylinidae SR F1, 20 = 83.03, P < 0.0001 F1, 20 = 0.26, P = 0.6165 F1, 20 = 0.94, P = 0.3430

Staphylinidae abundance F1, 20 = 64.73, P < 0.0001 F1, 20 = 0.15, P = 0.7010 F1, 20 = 0.24, P = 0.6286

Staphylinidae biomass F1, 20 = 0.00, P = 0.9825 F1, 20 = 0.57, P = 0.4607 F1, 20 = 0.26, P = 0.6134

109

Appendix C. Coleoptera (A–C), carabid (D–F), and staphylinid (G–I) species richness regressed against their abundances during each year, 2002–2004. Definitions of symbols are given in Figure 7. (*) indicates significance at P < 0.0001 for the R2 values.

A) B) C)

8 s

s 8

s s s 2002 8

2003 s e

e 2004 e

7 n 7

n n

h 7

c i

6 i

6 r

R²=83.2%*

s e

e 5 i

5 e

i 5

c c

R²=63.7%* c

e e 4 e

4 p 4

R²=48.7%* s

a 3 a a 3

3 r

e t t 2

p 2

2 p

e l 1 e

e 1

l l

1 o

o C

0 C C 0 0 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 Coleoptera abundance Coleoptera abundance Coleoptera abundance

D) E) F)

2 s

2 s 2004

s s 2003 e

s 2002

e n

e R²=78.5%*

r i

e c

1 R²=88.4%* p

s e

s R²=92.2%* 1

r C a 0 0

C 0 1 2 3 4 5 0 1 2 3 0 Carabidae abundance 0 1 2 3 Carabidae abundance Carabidae abundance G) H) I)

s 2

s e

2002 s

n s

s 4

h s

4 e 2004

h n

r 2003

e r

i 3 s

c s

R²=59.4%* e

c i

p 1 c

e R²=78.2%*

e 2

e s

p 2

i a

e R²=73.6%*

d n

i 1 y

1 i

p h

t 0

p S

0 t 0

a t 0 1 2 3 4 5 6 S 0 2 4 6 8 10 S 0 2 4 6 Staphylinidae abundance Staphylinidae abundance Staphylinidae abundance

110

Appendix D. Coleoptera (A–C), carabid (D–F), and staphylinid (G–I) species richness regressed against their biomass during each year, 2002–2004. Definitions of symbols are given in Figure 7. (*), (**), (***), (****) and (******) indicate significance at P <

0.0001, P = 0.008, P = 0.034, P = 0.016, and P = 0.018, respectively, for the R2 values.

A) B) C)

8 s

s 8

s s 8 s 2003

s 2004 e 2002 e

e 7 7

n n

n 7

h h

c c

i i

c 6 6

r r

r 6

s s

s e

e 5 5

i i

e 5

c c

e e

e 4 4 p 4 p

p R²=27.9%**

s s

a a 3 3

a 3

r r

e e

t t e R²=2.2%

t 2 p

2 p 2

p o

o R²=18.9%***

e e

l l e 1 1

l 1

o o

C C C 0 0 0 0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.02 0.04 0.06 0.08 0.10 Coleoptera biomass (g) Coleoptera biomass (g) Coleoptera biomass (g)

D) E) F)

2 3

s s

s 2

s s

2002 s 2004

e e

e 2003

R²=23.5%****

s s

s R²=22.8%****** 2

e e

1 e 1

p p

s R²=79.6%*

e e

e 1

C C 0 C 0 0 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.00 0.02 0.04 0.06 0.08 0.00 0.02 0.04 0.06 0.08 0.10 Carabidae biomass (g) Carabidae biomass (g) Carabidae biomass (g)

G) H) I)

s s

2 s s

e 2003 4 e

e 2002 3 2004

c i

e i

2 i

e p p R²=2.6% p 2

1 s

R²=1.3% s

e a

1 a

d i i

i 1 n

R²=2.3% n

n i

p p

a 0

S S

S 0 0.000 0.005 0.010 0.015 0.020 0.000 0.003 0.006 0.009 0.012 0.000 0.003 0.006 0.009 0.012 0.015 Staphylinidae biomass (g) Staphylinidae biomass (g) Staphylinidae biomass (g)

111

Appendix E. Coleoptera (A), carabid (B), and staphylinid (C) species richness regressed against 2004 plant species richness. Definitions of symbols are given in Figure 7. (****) and (*******), indicate significance at P = 0.016 and P = 0.011, respectively, for the R2 values.

A) B)

8 3 s

2004 s s

s 2004 e

7 e

c i

6 i

s 2 e

5 e

c e

4 e

p s

R²=23.8%**** s

a 3 e

r 1 R²=2.9%

t i

p 2

e l

1 a

o C C 0 0 0 5 10 15 20 0 5 10 15 20 Plant species richness Plant species richness

s 4

s e

n 2004

i r

c e

p 2

a R²=26.1%*******

i n

i 1

a t

S 0 0 5 10 15 20 Plant species richness

112

Appendix F. Coleoptera (A), staphylinid (B), and carabid (C) species richness regressed against 2004 standing crop biomass. Definitions of symbols are given in Figure 7.

(*****) indicates significance at P = 0.001 for the R2 values.

A) B)

s s

8 s 4 s

e 2004 e

2004 n

r 6 3

s e

i R²=37.6%*****

e e

4 p 2 p

R²=38.6%***** s

e t

2 n 1

a t C 0 0 0 300 600 900 1200 1500 1800 S 0 300 600 900 1200 1500 1800 Standing crop biomass (g / m²) Standing crop biomass (g / m²)

3 s

s 2004

s 2

e 1

a R²=0.0%

r a

C 0 0 300 600 900 1200 1500 1800 Standing crop biomass (g / m²)

113

Appendix G. Results of the repeated-measures and yearly PROC MIXED analysis for the individual species presented in Fig. 10. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α <

0.05. The predictor variables are fertilization (F), litter (L), year (Y), and (*) their factorial interactions.

114

Species F L Y F*L F*Y L*Y F*L*Y

Meronera venustula

Repeated-measures F1, 80 = 84.59 F1, 80 = 4.62 F3, 80 = 44.10 F1, 80 = 3.72 F3, 80 = 41.08 F3, 80 = 4.29 F3, 80 = 4.06 P < 0.0001 P = 0.0345 P < 0.0001 P = 0.0574 P < 0.0001 P = 0.0074 P = 0.0097

2002 F1, 20 = 1.32 F1, 20 = 0.68 -- F1, 20 = 0.68 ------P = 0.2634 P = 0.4208 -- P = 0.4208 ------

2003 F1, 20 = 26.10 F1, 20 = 1.44 -- F1, 20 = 0.48 ------P < 0.0001 P = 0.2448 -- P = 0.4958 ------

2004 F1, 20 = 13.95 F1, 20 = 0.04 -- F1, 20 = 0.13 ------P = 0.0013 P = 0.8529 -- P = 0.7265 ------

2005 F1, 20 = 61.81 F1, 20 = 5.49 -- F1, 20 = 4.98 ------P < 0.0001 P = 0.0296 -- P = 0.0372 ------

Mocyta fungi

Repeated-measures F1, 80 = 9.18 F1, 80 = 0.01 F3, 80 = 3.99 F1, 80 = 0.00 F3, 80 = 3.86 F3, 80 = 0.26 F3, 80 = 0.27 P = 0.0033 P = 0.9273 P = 0.0106 P = 0.9868 P = 0.0124 P = 0.8520 P = 0.8483

2002 F1, 20 = 0.91 F1, 20 = 3.64 -- F1, 20 = 0.00 ------P = 0.3517 P = 0.0710 -- P = 1.000 ------

2003 F1, 20 = 2.50 F1, 20 = 0.10 -- F1, 20 = 0.10 ------P = 0.1295 P = 0.7551 -- P = 0.7551 ------

2004 F1, 20 = 2.27 F1, 20 = 0.27 -- F1, 20 = 0.32 ------P = 0.1472 P = 0.6099 -- P = 0.5780 ------

2005 F1, 20 = 6.35 F1, 20 = 0.17 -- F1, 20 = 0.15 ------P = 0.0204 P = 0.6838 -- P = 0.6980 ------

115

(Appendix G continued)

Species F L Y F*L F*Y L*Y F*L*Y

Myrmedonota aidani

Repeated-measures F1, 80 = 0.91 F1, 80 = 0.17 F3, 80 = 1.94 F1, 80 = 0.18 F3, 80 = 0.54 F3, 80 = 0.55 F3, 80 = 0.05 P = 0.3436 P = 0.6831 P = 0.1295 P = 0.6759 P = 0.6540 P = 0.6480 P = 0.9854

2002 F1, 20 = 0.04 F1, 20 = 0.18 -- F1, 20 = 0.05 ------P = 0.8440 P = 0.6764 -- P = 0.8249 ------

2003 F1, 20 = 1.63 F1, 20 = 1.05 -- F1, 20 = 0.09 ------P = 0.2168 P = 0.3182 -- P = 0.7633 ------

2004 F1, 20 = 0.21 F1, 20 = 1.00 -- F1, 20 = 0.67 ------P = 0.6538 P = 0.3285 -- P = 0.4221 ------

2005 F1, 20 = 2.61 F1, 20 = 0.01 -- F1, 20 = 0.08 ------P = 0.1217 P = 0.9252 -- P = 0.7784 ------

Tomoderus constrictus

Repeated-measures F1, 80 = 8.56 F1, 80 = 0.04 F3, 80 = 8.41 F1, 80 = 0.46 F3, 80 = 4.79 F3, 80 = 0.14 F3, 80 = 0.07 P = 0.0045 P = 0.8518 P < 0.0001 P = 0.5006 P = 0.0040 P = 0.9360 P = 0.9750

2002 F1, 20 = 0.02 F1, 20 = 0.32 -- F1, 20 = 0.32 ------P = 0.8772 P = 0.5777 -- P = 0.5777 ------

2003 F1, 20 = 8.31 F1, 20 = 0.00 -- F1, 20 = 0.10 ------P = 0.0092 P = 0.9453 -- P = 0.7547 ------

2004 F1, 20 = 2.09 F1, 20 = 0.64 -- F1, 20 = 0.35 ------P = 0.1639 P = 0.4315 -- P = 0.5626 ------

2005 F1, 20 = 3.41 F1, 20 = 0.05 -- F1, 20 = 0.01 ------P = 0.0796 P = 0.8302 -- P = 0.9146 ------

116

(Appendix G continued)

Species F L Y F*L F*Y L*Y F*L*Y

Poecilus lucublandus

Repeated-measures F1, 80 = 0.59 F1, 80 = 2.20 F3, 80 = 14.60 F1, 80 = 0.87 F3, 80 = 0.89 F3, 80 = 0.63 F3, 80 = 1.78 P = 0.4450 P = 0.1423 P < 0.0001 P = 0.3527 P = 0.4499 P = 0.5981 P = 0.1571

2002 F1, 20 = 0.05 F1, 20 = 0.21 -- F1, 20 = 2.91 ------P = 0.8224 P = 0.6540 -- P = 0.1035 ------

2003 F1, 20 = 0.04 F1, 20 = 0.90 -- F1, 20 = 0.33 ------P = 0.8511 P = 0.3532 -- P = 0.5748 ------

2004 F1, 20 = 7.98 F1, 20 = 5.71 -- F1, 20 = 2.31 ------P = 0.0105 P = 0.0268 -- P = 0.1439 ------

2005 F1, 20 = 0.02 F1, 20 = 0.50 -- F1, 20 = 1.42 ------P = 0.8778 P = 0.4864 -- P = 0.2466 ------

Ptiliidae

Repeated-measures F1, 80 = 55.67 F1, 80 = 1.14 F3, 80 = 8.46 F1, 80 = 1.09 F3, 80 = 8.63 F3, 80 = 0.70 F3, 80 = 0.72 P < 0.0001 P = 0.2890 P < 0.0001 P = 0.2989 P < 0.0001 P = 0.5573 P = 0.5455

2002 F1, 20 = 8.47 F1, 20 = 2.84 -- F1, 20 = 3.97 ------P = 0.0086 P = 0.1075 -- P = 0.0602 ------

2003 F1, 20 = 40.49 F1, 20 = 0.28 -- F1, 20 = 0.44 ------P < 0.0001 P = 0.6018 -- P = 0.5150 ------

2004 F1, 20 = 27.60 F1, 20 = 2.08 -- F1, 20 = 2.08 ------P < 0.0001 P = 0.1646 -- P = 0.1646 ------

2005 F1, 20 = 15.24 F1, 20 = 0.04 -- F1, 20 = 0.03 ------P = 0.0009 P = 0.8347 -- P = 0.8565 ------

CHAPTER 4

EPIGEAL PREDATOR RESPONSES TO FERTILIZATION AND PLANT

LITTER: TESTING BIODIVERSITY THEORY AT THE GROUND LEVEL

Abstract

Recent studies of terrestrial ecosystem eutrophication have focused on the portion of the food web most closely associated with living plants (i.e., the ―aerial‖ arthropod community). These studies have formed the biodiversity-productivity theory in which nutrient loading resulted in increase arthropod abundance and biomass, but decreased arthropod diversity. However, none of these studies have explicitly examined the effects of nutrient loading, plant diversity, and plant litter on the ground-level (―epigeal‖) arthropod community. To test whether these nutrient loading effects extend to the detritus-based epigeal community and to individual species within this community, I used pitfall traps to sample the epigeal spider community for four years within 24 large (314 m2) plots in which I manipulated nutrient loading (NPK fertilizer) and plant litter. I measured the diversity, abundance, biomass, and community structure responses of the entire spider beetle community and of wolf spiders (Lycosidae) and penny spiders

(Linyphiidae), as well as the abundance and biomass responses of the six most common

117 118

individual spider species in a temperate-latitude grassland. As with previous studies that sampled the aerial arthropod community, I hypothesized that increased nutrient loading would increase epigeal spider abundance and biomass but decrease epigeal spider diversity. Contrary to predictions, total spider species richness, diversity, and biomass were not significantly affected by fertilization, while spider abundance was significantly increased as a result of fertilization. Also contrary to predictions, plant litter did not affect any of these variables. Penny spiders showed the strongest responses to fertilization, with significantly increased abundance and biomass, and, contrary to predictions, increased species richness in fertilized plots. Wolf spiders responded more closely to predictions, with increased abundance and biomass, but without any significant change in species richness. One particular wolf spider, Pardosa moesta, significantly increased in abundance in fertilized plots, with so many captures in fertilized plots during the final two years to make it the most frequently captured spider species of the entire experiment. My results indicate that the epigeal spider community does not respond as would be predicted by biodiversity-productivity theory. This underscores the need to integrate the largely detritus-based epigeal community into current biodiversity- productivity theory.

Introduction

Increased human activity has resulted in a significant increase in the global nitrogen (N) pool through fertilization and increased atmospheric N deposition (Vitousek

119

et al. 1997; Fenn et al. 2003; Galloway et al. 2003). The effects of increased N are well documented in terrestrial plant communities across a range of natural and semi-natural habitats (Bobbink et al. 1998), as well as in experimentally manipulated habitats and mesocosms (Hector et al. 1999; Tilman et al. 2002; Suding et al. 2005; Patrick et al.

2008a). Typical plant community responses include decreased plant species richness, increased standing crop biomass, and the limitation of community composition to a few dominant species. This research has provided substantive support for biodiversity- productivity theory which predicts declines in local and regional richness as one moves from mesotrophic to eutrophic systems (Grime 1973; McCann 2000; Worm and Duffy

2003; Suding et al. 2005; Chalcraft et al. 2008). Similarly, increased nutrient loading has been linked with decreased species richness and increased abundance of terrestrial arthropods, particularly those species most closely linked to the living-plant portion of the food web (Knops et al. 1999; Haddad et al. 2000, 2001; but see Patrick et al. 2008b).

This ―eutrophication effect‖ (Fenn et al. 2003) can result in significant biodiversity loss and the potential decline in important ecosystem functions, such as ecosystem stability

(McCann 2000; Larsen et al. 2005).

Previous studies of N loading have focused on the portions of the food web closely tied to living plant material, e.g., sweep net sampling of the ―aerial‖ arthropod community associated with the upper portions (e.g., stems) of plants (Knops et al. 1999;

Haddad et al. 2000, 2001). While there is evidence to support the eutrophication effect on aerial arthropod diversity, less is known regarding how nutrient loading affects the epigeal (ground-level) arthropod community. In one recent study, diversity of epigeal

120

invertebrates increased in fertilized sites (Patrick et al. 2008b), opposite to what was observed for aerial arthropods. This differential response by the epigeal arthropod community may result from it being more closely tied to the detritus-based portion of the food web (Halaj and Wise 2002). Ultimately, this detrital community increases overall biodiversity and facilitates effective nutrient regeneration and recycling in terrestrial ecosystems, thereby influencing ecosystem function (Moore et al. 2004; Hättenschwiler and Gasser 2005; Wardle 2006), ecosystem services (Barrios 2007), and plant species diversity (Mazzoleni et al. 2007). Despite the important role it may play, the epigeal arthropod community remains an understudied food web component (Wardle 2002;

Hättenschwiler et al. 2005; Cross et al. 2006), particularly in biodiversity-productivity theory.

Hättenschwiler et al. 2005) and litter quality (Wardle et al. 2006). With nutrient loading decreasing plant diversity (Tilman et al. 2002; Long et al. 2003; Patrick et al. 2008), litter diversity is also decreased. While more plant litter production could increase detritivore

121

and epigeal predator abundance and biomass (Halaj et al. 2000), a reduction in litter diversity could result in declines in diversity of detritivores and epigeal predators

(Hättenschwiler and Gasser 2005; Wardle 2006), effectively mirroring the aerial community response to nutrient loading. Despite such predictions, only one study has incorporated plant litter effects into an experimental test of biodiversity-productivity theory with specific reference to the epigeal arthropod community (Patrick et al. 2008b).

However, that paper did not present data delineating the species-level responses within the epigeal community.

Most terrestrial trophic levels (e.g., detritivores, predators) would be expected to exhibit species-level responses to habitat perturbations. Spiders, in particular, are ubiquitous generalist predators (Wise 1993) that can significantly impact terrestrial food webs (Wise et al. 1999), and epigeal spiders (e.g., Lycosidae, Linyphiidae) are closely linked to the detritivore community (Wise et al. 1999; Chen and Wise 1999; Wise 2006).

While spider predation affects prey populations, potentially influencing spider abundance and productivity (Snyder and Wise 2001; Lawrence and Wise 2004; Lensing et al. 2005;

Thus, increasing plant detritus can increase spider abundance by increasing the quantity of food available to their detritivorous prey which increases detritivore abundance (Chen and Wise 1999, Wise et al. 1999; Wise 2004), which may affect different spider species differently. Further, increased detritus also enhances habitat structure for hiding and web

122

building (Uetz 1979, 1991; Rypstra et al. 1999), which can also moderately increase the local richness of the spider community (Rypstra et al. 1999), also differentially affecting individual spider species.

(Suding et al. 2005; Patrick et al. 2008a), plant litter diversity also decreases. Even though more plant litter is produced, increasing the resource base of the detritivore food web, lower litter diversity likely begets lower detritivore and detritivore-predator diversity (Hättenschwiler and Gasser 2005; Wardle 2006). Interestingly, no epigeal spider studies (that focused strictly on cursorial spiders, e.g., Lycosidae; wolf spiders) have looked at spider diversity response to basal resource manipulation. Moreover, no studies have examined responses of the predominantly epigeal spider family Linyphiidae

(wandering sheet/tangle-web builders) that may patrol multiple webs at the ground level

(Uetz et al. 1999).

Here I report the results of a four year study that investigated the response of the epigeal spider community to experimental manipulations of NPK fertilization and plant litter availability in a temperate-latitude grassland. I measured the diversity, abundance, biomass, and community structure responses of the entire epigeal spider community, the

123

spider families Lycosidae and Linyphiidae, and the dominant individual spider species.

My goal was to integrate the detrital food web into biodiversity-productivity theory through insight gained from the responses of predators that rely largely on the detritivore food web. Based on previous studies that sampled the aerial arthropod community responses to nutrient loading (e.g., Knops et al. 1999; Haddad et al. 2000, 2001), I tested four hypotheses: as a result of fertilization, 1) spider biomass and abundance will increase while spider species richness decreases, 2) spider species richness will decrease as plant species richness decreases, 3) spider species richness will decrease as plant biomass increases, and 4) the presence of plant litter will moderately increase spider species richness, though this effect will be dampened in fertilized plots.

Materials and methods

Study site and experimental design

The study was done at the 163.5 ha Bath Nature Preserve (BNP; 41° 10’ 36.2‖ N,

124

and Anthoxanthum odoratum L. The site is moderately productive relative to other grasslands within the upper Midwest and across the U.S. (Patrick et al. 2008a). The dominant soil type is Ellsworth silt loam (ElB), which consists of moderately well drained, moderately deep to deep soils formed in silty clay loam or clay loam glacial till of the Wisconsin Age (Ritchie and Steiger 1974).

During August 2001, twenty-four 20-m diameter circular plots (314 m2) were established. These experimental plots were separated by at least 20 m and were at least

30 m away from any other habitat. Treatments were applied in a 2 x 2 factorial design of fertilizer (+F = fertilizer added, -F = no fertilizer) and plant litter (-L = litter removed, +L

= litter left in situ after yearly mowing) with control plots characterized as no fertilization and plant litter left in situ (+L/-F), resulting in six replicates per treatment. Hereafter, all references to ―litter‖ refer to the previous year’s mown vegetation and any vegetation senesced and found within the sampling quadrat after standing crop removal. In April

2002 and continuing each April through 2005, Scotts brand Osmocote 8-9 month Slow

Release Fertilizer 19-6-12 (NPK; Scotts, Marysville, OH, USA) was applied at 20 g N m-

2 in fertilized plots, well above the Köchy and Wilson (2005) 15 g N m-2 yr-1 threshold necessary to induce a eutrophication effect in grasslands and other habitats. I could not exclude ambient wet/dry atmospheric N deposition, though deposition rates from 1990 to

2005 were relatively low at ~1.01 g N m-2 yr-1 at a nearby monitoring site in Lykens (162 km west of my study site), OH, USA, and ~0.93 g N m-2 yr-1 at another nearby monitoring site in Mercer Co. (G. K. Goddard site; 96 km east of my study site), PA,

USA (US EPA 2005). Within two days of annual mowing of the whole site by the local

125

township with a large tractor and brush hog mower (autumn 2001-2004), litter was removed from litter-removal treatments using a small 23 hp lawn tractor with a pull- behind 8 hp Agri-Fab Mow-N-Vac trailer attachment (Agri-Fab, Sullivan, IL, USA).

Spider community sampling

Spiders were collected using four pitfall traps in each of the 24 experimental plots

(n = 96 total pitfall traps). Within each plot, a single trap was placed 5 m from the center of the plot at each of four magnetic compass directions (northeast, northwest, southeast, and southwest). Each trap consisted of a 10 cm diameter, 18 cm tall PVC sleeve into which a 710-mL plastic cup was inserted and filled to ~4 cm with a 50/50 water/propylene glycol mixture. To deter trap raiders (e.g., microtine mammals), prevent captured spiders from climbing out of the trap, and prevent precipitation from directly flooding the trap, an 8-cm powder funnel with a base enlarged to ~ 3 cm was inserted and a 15 cm x 15 cm board was placed over each trap, leaving ~ 3 cm clearance. Staring in mid to late May (mid July during 2004) and continuing through mid to late August, traps were alternately left open for two weeks and closed for two weeks. This resulted in three sampling periods each year during 2002, 2003, and 2005. During 2004, only the second and third sampling periods were collected. When each two-week sampling period ended, the plastic cups were removed, the contents collected and preserved in 70% EtOH, and the PVC sleeve was tightly capped. While pitfall traps do not capture all spiders in the

126

community, they are an effective sampling technique for determining the relative abundance and species richness of epigeal spiders (Greenslade 1964; Phillips and Cobb

2005). Spiders captured in each trap were identified to species for all mature specimens, and to family for all immature specimens, and exact numbers of species/families within in each trap were recorded, and dried at 70°C for 72 hrs to determine species-specific biomasses to the nearest 0.0001 g. Lacking sufficient numbers captured within a trap, some extremely small species did not register a biomass, and their biomass was recorded as ―0.000001 g.‖

Statistical analyses

I tested the effects of fertilization, plant litter, and the interaction of fertilization and litter on the abundance, biomass, species richness (SR), and effective Shannon’s diversity (eH’, where H’ is the Shannon diversity index) of 1) all mature spiders

(Araneae), 2) lycosid and linyphiid spiders, and 3) on the abundance and biomass of the six most abundant spider species. I used eH’ to correct for differences in species richness that might have resulted from differential spider abundances (Ricklefs and Miller 2000;

Haddad et al. 2000). To calculate the average SR within a plot, I summed the total number of spider species caught in each trap, then averaged this SR for each of the four traps within a plot within a sampling period (including zeroes for traps where no spiders were captured), then averaged these SRs for each plot across sampling periods in a year, yielding n = 24 samples within each year. The same method was used to calculate the

127

average abundance, biomass, and eH’ within a plot within a year, also yielding n = 24 samples within each year. Correlations and regressions of these spider responses with plant species richness (plant SR) and standing crop biomass utilized data from Patrick et al (2008a).

To analyze trends per year and per treatment in abundance, biomass, SR, and eH’ I used SAS software version 8.01 (SAS Institute, Inc. 1999) to calculate a repeated- measures maximum-likelihood analysis in PROC MIXED with Type III effects based upon the covariance structure of compound symmetry. The various models used the different response variables (biomass, SR, eH’, abundance), and for the predictor variables used: fertilized vs. unfertilized, litter removed vs. litter left in situ, year, and their factorial interactions, with year as the repeated predictor. When year was detected as a significant effect for a response variable, I tested for treatment effects within a year and used SAS software version 8.01 (SAS Institute, Inc. 1999) to calculate a maximum- likelihood analysis in PROC MIXED with Type III effects based upon the covariance structure of compound symmetry, with fertilization, litter, and the factorial interaction of fertilization and litter as predictor variables.

To assess treatment effects on aggregate biotic and abiotic components in my system, I applied nonmetric multidimensional scaling (NMS; Kruskal 1964) using PC-

ORD (McCune and Mefford 2006). For 2005, variables used for each of the 24 plots were average spider species richness per plot and four variables used in a previously published analysis (Patrick et al. 2008a): average plant litter biomass, average PAR per plot, average percent soil moisture per plot, and average percent soil organic content per

128

plot, resulting in a matrix with five columns and 24 rows (plots). The same analysis was run a second time with of the same variables except Linyphiidae species richness replaced spider species richness. Because 1) NMS is scale sensitive, 2) these variables are on radically different measurement scales, and 3) variables have an enormous range of values between variables, data were transformed to proportions relative to the highest value for each variable (i.e., each value in a column was divided by the largest value in that column, creating a unitless range from 0 – 1 for each column). The NMS analysis was run with Sørensen distance, time as the random seed for the starting configuration,

9999 runs stepping down from 5 to 1 dimensions with the real data, 999 Monte Carlo runs to assess the probability of a similar final stress obtained by chance, and a 0.005 stability criterion. Additionally for 2005 and to support NMS analyses with stable results, I used PC-ORD (McCune and Mefford 2006) to run the multi-response permutation procedure (MRPP; Mielke 1984) to test for the hypothesis of no difference among treatments. The MRPP used Sørensen distance with the four treatments as the a priori groupings, resulting in a matrix with five columns (biotic and abiotic variables) and 24 rows (plots) and was calculated with all four treatments together, and for pairwise comparisons between treatments to test for the strength of difference between individual treatments.

Results

General trends

129

A total of 13174 spiders from 14 families were captured during 14784 trap nights.

Of this total, 2515 spiders were immature and from 11 families, while the remaining

10659 spiders were mature and from 94 species and 12 families (Table 7). Lycosidae was the most commonly captured spider family, with 6577 mature specimens (61.7% of all mature spiders captured) from 20 species, while Linyphiidae was the second most commonly captured spider family with 3200 mature specimens (30.0 % of all mature spiders captured) from 34 species. Together these two families represented 9777 (91.7% of all mature spiders captured) specimens from 54 species (57.4% of all species captured). In the repeated-measures PROC MIXED analysis, spider diversity, corrected for abundance with eH’, was significantly affected by fertilization and by year, but not by litter (Table 8). The factorial interaction between fertilization and year, fertilization and litter, litter and year, and the fully factorial interaction of fertilization by litter by year were not significant (Table 8). Because of the significance of year (Fig. 11), I tested for treatment effects for spider diversity within each year. During 2002 and 2003, fertilization was significant, but neither litter nor the interaction of fertilization and litter were significant (Appendix A). During 2004 (Appendix A) and 2005 (Table 9), neither fertilization, litter, nor their interaction were significant.

In the repeated-measures PROC MIXED analyses, fertilization significantly affected Araneae (all spiders) abundance but not Araneae SR or Araneae biomass (Table

8, Fig. 12A–C). Moreover, fertilization effects were significant for Linyphiidae SR, abundance, and biomass (Table 8, Fig. 12D–F), as well as for Lycosidae SR and

130

Table 7. Total numbers of each family and species of spider captured during the four year manipulative experiment. ―+L/-F‖ represents the control treatment of unfertilized plots with litter left in situ, ―-L/-F‖ represents the unfertilized with litter removed treatment, ―+L/+F‖ represents the fertilized with litter left in situ treatment, ―-L/+F‖ represents the fertilized with litter removed treatment, and ―Total‖ represents the total number caught.

131

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Agelenidae Agelinid spiderling 1 0 0 0 1

Araneidae Araneid spiderlings 1 2 0 0 3

Clubionidae 15 15 4 3 37 Clubiona abbottii L. Koch 0 0 1 0 1 Clubiona kastoni Gertsch 6 9 1 1 17 Clubionid spiderlings 9 6 2 2 19

Corinnidae 90 71 29 26 216 Castianeira gertschi Kaston 3 5 0 0 8 Castianeira longipalpa (Hentz) 1 0 0 1 2 Castianeira variata Gertsch 1 0 1 0 2 Corinnid spiderling 1 0 0 0 1 Meriola decepta Banks 10 13 3 6 32 Phrurotimpus borealis (Emerton) 0 0 1 0 1 Scotinella britcheri (Petrunkevitch) 0 0 0 2 2 Scotinella fratrella (Gertsch) 71 42 24 17 154 Scotinella madisonia Levi 3 11 0 0 14

Dictynidae Cicurina arcuata Keyserling 0 0 2 1 3

Gnaphosidae 139 104 77 77 397 Drassyllus creolus Chamberlin and Gertsch 12 12 3 2 29 Drassyllus depressus (Emerton) 63 57 20 30 170 Gnaphosa parvula Banks 43 19 44 37 143 Gnaphosid spiderlings 21 16 8 7 52 Litopyllus temporarius Chamberlin 0 0 2 1 3

Hahniidae 5 17 4 2 28 Hahniid spiderlings 2 4 0 1 7 Neoantistea agilis (Keyserling) 0 1 2 0 3 Neoantistea magna (Keyserling) 2 4 2 1 9 Neoantistea riparia (Keyserling) 1 8 0 0 9

132

(Table 7 continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Linyphiidae 717 532 1131 1058 3438 Agyneta sp. 1 0 1 0 0 1 Agyneta sp. 2 1 0 0 0 1 Agyneta sp. 3 0 0 0 1 1 Allomengea dentisetis (Grube) 0 0 1 1 2 Bathyphantes pallidus (Banks) 127 64 336 232 759 Centromerus cornupalpis (O. P.- Cambridge) 4 0 11 4 19 Ceraticelus similis (Banks) 0 1 1 0 2 Ceratinopsis laticeps Emerton 0 0 0 1 1 Diplostyla concolor (Wider) 3 8 117 42 170 Eridantes erigonoides (Emerton) 242 126 195 225 788 Erigone autumnalis Emerton 56 51 33 45 185 Erigone dentigera O. P.-Cambridge 0 1 1 2 4 Grammonota gentilis Banks 0 0 0 1 1 Grammonota inornata Emerton 20 51 13 17 101 Halorates plumosus (Emerton) 31 20 150 192 393 Islandiana flaveola (Banks) 11 12 8 7 38 Linyphiid spiderlings 31 53 90 64 238 Maso sundevalli (Westring) 1 0 0 0 1 Meioneta fabra (Keyserling) 10 13 6 9 38 Meioneta micaria (Emerton) 5 4 2 0 11 Meioneta unimaculata (Banks) 85 52 61 72 270 Mermessus entomologicus (Emerton) 0 1 0 0 1 Mermessus jona (Bishop and Crosby) 9 9 8 2 28 Mermessus tridentatus (Emerton) 3 2 1 3 9 Mermessus trilobatus (Emerton) 38 26 55 56 175 Neriene clathrata (Sundevall) 23 8 2 6 39 Oedothorax trilobatus (Banks) 0 0 2 5 7 Tennesseellum formicum (Emerton) 0 1 0 0 1 Tenuiphantes tenuis (Blackwall) 0 0 0 5 5 Walckenaeria directa (O. P.-Cambridge) 3 0 6 3 12 Walckenaeria palustris Millidge 0 0 1 0 1 Walckenaeria sp. 1 1 0 0 0 1 Walckenaeria sp. 2 0 0 1 0 1 Walckenaeria spiralis (Emerton) 12 28 30 63 133 Walckenaeria tibialis (Emerton) 1 0 0 0 1

133

(Table 7 continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Liocranidae Agroeca pratensis Emerton 1 0 0 0 1

Lycosidae 1913 2021 2422 2276 8632 Allocosa funerea (Hentz) 19 15 2 14 50 Hogna helluo (Walckenaer) 0 6 1 1 8 Lycosid spiderlings 760 648 254 393 2055 Pardosa milvina (Hentz) 0 3 0 0 3 Pardosa modica (Blackwall) 2 1 6 3 12 Pardosa moesta Banks 344 233 1444 1177 3198 Pardosa saxatilis (Hentz) 140 132 16 73 361 Pirata aspirans Chamberlin 0 1 0 0 1 Pirata canadensis Dondale and Redner 0 1 0 0 1 Pirata giganteus Gertsch 0 0 0 1 1 Pirata insularis Emerton 7 27 10 22 66 Pirata minutus Emerton 365 638 538 395 1936 Pirata sedentarius Montgomery 0 1 1 10 12 Rabidosa punctulata (Hentz) 2 3 0 0 5 Rabidosa rabida (Walckenaer) 4 3 0 0 7 Schizocosa avida (Walckenaer) 33 94 4 9 140 Schizocosa bilineata (Emerton) 80 61 25 34 200 Schizocosa crassipalpata Roewer 113 108 43 71 335 Trochosa ruricola (De Geer) 24 16 36 37 113 Trochosa terricola Thorell 18 29 42 36 125 Varacosa avara (Keyserling) 2 1 0 0 3

Philodromidae Ebo latithorax Keyserling 0 1 0 0 1

Salticidae 63 75 16 21 175 Ghelna barrowsi (Kaston) 1 2 0 1 4 Ghelna canadensis (Banks) 4 3 0 2 9 Ghelna castanea (Hentz) 2 0 0 0 2 Marpissa lineata (C. L. Koch) 7 3 1 2 13 Myrmarachne formiciaria (De Geer) 0 1 0 1 2 Neon avalonus Gertsch and Ivie 1 0 0 0 1

134

(Table 7 continued)

Family/Species +L/-F -L/-F +L/+F -L/+F Total

Salticidae continued Neon nelli Peckham and Peckham 7 12 1 2 22 Neon plutonus Gertsch and Ivie 25 35 7 7 74 Salticid spiderlings 5 11 5 3 24 Sarinda hentzi (Banks) 0 1 0 0 1 Talavera minuta (Banks) 11 7 2 3 23

Tetragnathidae 51 46 16 22 135 Glenognatha foxi (McCook) 9 10 3 8 30 Pachygnatha autumnalis Marx 17 14 9 6 46 Pachygnatha clerki Sundevall 0 0 0 1 1 Pachygnatha xanthostoma C. L. Koch 0 0 1 0 1 Tetragnatha laboriosa Hentz 1 0 0 0 1 Tetragnathid spiderlings 24 22 3 7 56

Thomisidae 19 54 14 20 107 Thomisid spiderlings 13 27 6 11 57 Xysticus bicuspis Keyserling 1 0 0 0 1 Xysticus canadensis Gertsch 1 1 0 0 2 Xysticus ferox (Hentz) 4 25 7 9 45 Xysticus fraternus Banks 0 1 0 0 1 Xysticus luctans (C. L. Koch) 0 0 1 0 1

135

Table 8. Results of the repeated-measure PROC MIXED for each response variable.

Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter

(L), year (Y), and (*) their fully factorial interactions.

Response Variable F L Y F*L F*Y L*Y F*L*Y

H' e F1, 80 = 6.55 F1, 80 = 0.76 F3, 80 = 79.98 F1, 80 = 2.21 F3, 80 = 1.39 F3, 80 = 2.33 F3, 80 = 0.21 P = 0.0124 P = 0.3863 P < 0.0001 P = 0.1413 P = 0.2518 P = 0.0804 P = 0.8909

Araneae SR F1, 80 = 0.98 F1, 80 = 0.50 F3, 80 = 87.14 F1, 80 = 1.49 F3, 80 = 3.35 F3, 80 = 1.61 F3, 80 = 0.27 P = 0.3264 P = 0.4818 P < 0.0001 P = 0.2255 P = 0.0231 P = 0.1935 P = 0.8434

Araneae F1, 80 = 39.64 F1, 80 = 0.99 F3, 80 = 54.93 F1, 80 = 0.66 F3, 80 = 5.23 F3, 80 = 0.06 F3, 80 = 0.83 abundance P < 0.0001 P = 0.3218 P < 0.0001 P = 0.4199 P = 0.0024 P = 0.9817 P = 0.4795

Araneae F1, 80 = 0.04 F1, 80 = 1.63 F3, 80 = 38.14 F1, 80 = 2.25 F3, 80 = 1.46 F3, 80 = 1.61 F3, 80 = 2.02 biomass P = 0.8445 P = 0.2052 P < 0.0001 P = 0.1376 P = 0.2322 P = 0.1937 P = 0.1175

Linyphiidae SR F1, 80 = 28.05 F1, 80 = 0.19 F3, 80 = 53.05 F1, 80 = 2.82 F3, 80 = 9.67 F3, 80 = 1.90 F3, 80 = 0.39 P < 0.0001 P = 0.6624 P < 0.0001 P = 0.0968 P < 0.0001 P = 0.1371 P = 0.7615

Linyphiidae F1, 80 = 20.90 F1, 80 = 1.53 F3, 80 = 17.25 F1, 80 = 0.72 F3, 80 = 6.58 F3, 80 = 0.08 F3, 80 = 0.24 abundance P < 0.0001 P = 0.2202 P < 0.0001 P = 0.3999 P = 0.0005 P = 0.9717 P = 0.8659

Linyphiidae F1, 80 = 24.22 F1, 80 = 3.47 F3, 80 = 27.11 F1, 80 = 0.00 F3, 80 = 6.73 F3, 80 = 1.49 F3, 80 = 0.29 biomass P < 0.0001 P = 0.0663 P < 0.0001 P = 0.9470 P = 0.0004 P = 0.2224 P = 0.8291

Lycosidae SR F1, 80 = 5.43 F1, 80 = 1.94 F3, 80 = 86.81 F1, 80 = 0.16 F3, 80 = 1.11 F3, 80 = 0.56 F3, 80 = 0.40 P = 0.0223 P = 0.1678 P < 0.0001 P = 0.6876 P = 0.3491 P = 0.6452 P = 0.7539

Lycosidae F1, 80 = 22.78 F1, 80 = 0.05 F3, 80 = 29.13 F1, 80 = 2.09 F3, 80 = 0.90 F3, 80 = 0.01 F3, 80 = 0.38 abundance P < 0.0001 P = 0.8222 P < 0.0001 P = 0.1520 P = 0.4456 P = 0.9989 P = 0.7674

Lycosidae F1, 80 = 0.04 F1, 80 = 1.60 F3, 80 = 31.46 F1, 80 = 2.23 F3, 80 = 1.57 F3, 80 = 1.53 F3, 80 = 1.84 biomass P = 0.8502 P = 0.2102 P < 0.0001 P = 0.1392 P = 0.2024 P = 0.2139 P = 0.1468

136

Table 9. Results of the PROC MIXED for each response variable during 2005. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter (L), and (*) their factorial interaction. Results for 2002–2004 are in Appendix A.

Response Variable F L F*L

H' e F1, 20 = 1.03, P = 0.3217 F1, 20 = 2.42, P = 0.1352 F1, 20 = 0.50, P = 0.4894

Araneae SR F1, 20 = 0.19, P = 0.6694 F1, 20 = 3.06, P = 0.0957 F1, 20 = 0.05, P = 0.8249

Araneae abundance F1, 20 = 35.20, P < 0.0001 F1, 20 = 0.03, P = 0.8759 F1, 20 = 2.80, P = 0.1100

Araneae biomass F1, 20 = 0.75, P = 0.3963 F1, 20 = 1.98, P = 0.1744 F1, 20 = 2.56, P = 0.1250

Linyphiidae SR F1, 20 = 20.76, P = 0.0002 F1, 20 = 1.60, P = 0.2207 F1, 20 = 0.07, P = 0.7940

Linyphiidae abundance F1, 20 = 8.99, P = 0.0071 F1, 20 = 0.03, P = 0.8550 F1, 20 = 0.02, P = 0.8879

Linyphiidae biomass F1, 20 = 9.75, P = 0.0054 F1, 20 = 3.25, P = 0.0876 F1, 20 = 0.32, P = 0.5755

Lycosidae SR F1, 20 = 3.94, P = 0.0610 F1, 20 = 2.12, P = 0.1610 F1, 20 = 0.37, P = 0.5495

Lycosidae abundance F1, 20 = 9.34, P = 0.0062 F1, 20 = 0.00, P = 0.9910 F1, 20 = 2.15, P = 0.1585

Lycosidae biomass F1, 20 = 0.86, P = 0.3656 F1, 20 = 1.86, P = 0.1875 F1, 20 = 2.34, P = 0.1421

137

Figure 11. Average effective Shannon’s H’ (eH’) of spider species in each treatment; the letter ―a‖ above a year denotes significance at α < 0.05 for fertilization. Open circles (○) and ―+L/-F‖ represent the control treatment plots of unfertilized and litter left in situ, open triangles (∆) and ―-L/-F‖ represent unfertilized and litter removed plots, filled circles (●) and ―+L/+F‖ represent fertilized and litter left in situ plots, and filled triangles

(▲) and ―-L/+F‖ represent fertilized and litter removed plots.

+L/-F '

H -L/-F

5 s

' +L/+F n

o 4 -L/+F n n a a a

h 3

e v

i 2

e f

f 1 E 0 2002 2003 2004 2005 Year

138

Figure 12. Species richness, abundance, and biomass of all spiders (A–C), linyphiids (D–

F), and lycosids (G–I). Definitions of symbols and abbreviations for treatments are given in Figure 11, while the letter ―a‖ above a year denotes significance at α < 0.05 for fertilization.

A. B. C.

8 25 0.14

s +L/-F a

s ) e 0.12

-L/-F e

n (

c 20

6 n c

+L/+F s i

a 0.10

r s

-L/-F a s

n 15

e m u

i a a 0.08 o

b a

c i

4 a

e b

e 0.06 e

s 10

n a n 0.04 a

2 a

r r

n 5

A A

a 0.02

r A 0 0 0.00 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year D. E. F. a 3.5 10 0.0025

s a

s )

e a

e g

n 3.0

(

h a

a n 8 0.0020

r d

2.5

s n

m u e a

i 6 0.0015

b i

c 2.0

s a

1.5 a

4 0.0010

e i

h h

d 1.0

y y h 0.0005

n 2

i p

0.5 i

y L

n i

L 0.0 0 0.0000 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year G. H. I. a

4 15 0.14

)

e g

n 0.12

(

h 12

s c

i 3 a a s

r 0.10

e u

i 9 o

b 0.08

2 a

p e

e 0.06

a a

a s

s 0.04

o o

i 1 a

c c

s 3

y o

L 0.02

c y L 0 0 0.00 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year

139

abundance (Table 8, Fig. 12G–H), but not for Lycosidae biomass (Table 8, Fig. 12I). All response variables were significantly affected by year (Table 8), and Araneae SR and abundance, and Linyphiidae SR, abundance, and biomass had significant fertilization and year interactions. PROC MIXED analyses for each year resulted in fertilization as the only significant predictor variable (Table 9, Appendix A) in any given year, notably for

2005 always significantly affecting abundance, though also significantly affecting

Linyphiidae SR and biomass.

Neither Araneae SR nor Lycosidae SR were significantly correlated with abundance, but Linyphiidae SR was well correlated with abundance (Fig. 13A–C). As with abundance, biomass was only correlated to Linyphiidae SR (Fig. 13D–F). While

Araneae SR was not significantly correlated to plant SR (Fig. 14A), Linyphiidae SR was negatively correlated with plant SR and Lycosidae SR was positively correlated with plant SR (Fig. 14B–C). Araneae SR was also not correlated with standing crop biomass

Fig. 14D), while Linyphiidae SR was positively correlated and Lycosidae SR was negatively correlated with standing crop biomass (Fig. 14E–F).

Spider species-level analyses

The most abundant spider was Pardosa moesta Banks, with 3198 mature specimens captured (30.0% of all mature spiders, 48.6% of mature lycosids). Pardosa moesta was easily the most captured spider in fertilized plots while being virtually absent from unfertilized plots (Fig. 15A), and had a strong response to fertilization and year,

140

Figure 13. Regressions of all spiders, linyphiids, and lycosids (left to right) against abundance (A–C) and biomass (D–F). Symbols are defined in Figure 11, and data presented are for 2005 (data for 2002–2004 are found in Appendices B–C).

A. B.

10 s 5 s

s 2005

s n

e +L/-F h

n 8 4

i r

c -L/-F

r s

+L/+F

e s

6 i 3

e c

i -L/+F

s p R² = 11.2% P = 0.110 2

s 4

a R² = 73.4% P < 0.001

i e

h 1

n 2

i A 0 L 0 0 5 10 15 20 25 0 2 4 6 8 10 12 14 16 Araneae abundance Linyphiidae abundance C. D.

5 10

n n

h 8

4 h

s s

e 6 e i 3

i R² = 0.2% P = 0.833

p s

s 4 2

R² = 3.6% P = 0.371

d e

i 2 n

s 1

A y

L 0 0 0 5 10 15 20 0.00 0.05 0.10 0.15 0.20 0.25 Araneae biomass (g) Lycosidae abundance E. F.

s 5

n h

4 h 4

e e i 3

i 3 c

c R² = 0.8% P = 0.680

2 2

d d

i R² = 39.6% P < 0.001

i i

1 s

h 1

i L

L 0 0 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.00 0.05 0.10 0.15 0.20 0.25 Lycosidae biomass (g) Linyphiidae biomass (g)

141

Figure 14. Regressions of all spiders, linyphiids, and lycosids (left to right) against plant species richness (A–C) and standing crop biomass (D–F). Symbols are defined in Figure

11, and data presented are for 2005 (data for 2002–2004 are found in Appendices D–E).

A. B.

10 s 5

e h n 4

8 c

e s

6 i 3

s p R² = 7.6% P = 0.193 2

s 4

a R² = 16.0% P = 0.053

i e

2 h 1

i A 0 L 0 0 5 10 15 20 0 5 10 15 20 Plant species richness Plant species richness C. D.

5 10

e n

n 8 h

4 h

s 6

e i

3 i

e e

p R² = 40.8% P < 0.001

p 4 R² = 0.0% P = 0.985

d e

i 2 n

s 1

c A y 0 L 0 0 5 10 15 20 250 500 750 1000 1250 1500 Standing crop biomass (g) Plant species richness E. F.

5 s

s 5

n h

h 4 4

s s

e 3

e i

i 3

e p

p R² = 18.1% P = 0.038 s

s 2

i i

i 1

1 s h

R² = 39.6% P < 0.001 o

n i

L 0

L 0 250 500 750 1000 1250 1500 250 500 750 1000 1250 Standing crop biomass (g) Standing crop biomass (g)

142

Figure 15. Average abundance of selected species by year. Definitions of symbols and abbreviations for treatments are given in Figure 11, while the letters above each year denote significance at α < 0.05 for ―a‖ = fertilization, ―b‖ = litter, and ―c‖ = the interaction of fertilization and litter.

143

A. Pardosa moesta B. Pirata minutus

12 a 5 c

10 4 e

e 8

c c

n 3

n a a a a

6 d

n u

u 2 b

b 4

A A 2 1

0 0 2002 2003 2004 2005 2002 2003 2004 2005 Year Year C. Eridantes erigonoides D. Bathyphantes pallidus

2.0 3.5 a a 3.0

1.5 2.5

c n

n 2.0 a

a 1.0

d n

n 1.5

b A

A 1.0 0.5 0.5 0.0 0.0 2002 2003 2004 2005 2002 2003 2004 2005 Year Year E. Halorates plumosus F. Schizocosa avida a, b, c 3.5 a 1.4 3.0 1.2

2.5 1.0

e e

c c n

n 2.0 0.8

a a

d d n

n 1.5 0.6

u u b

b a A A 1.0 0.4 a 0.5 0.2 0.0 0.0 2002 2003 2004 2005 2002 2003 2004 2005 Year Year

144

resulting in a significant fertilization by year interaction (Table 10),. A smaller lycosid,

Pirata minutus Emerton, was the second most abundant spider with 1936 mature specimens captured (18.2% of all mature spiders, 29.4% of mature lycosids) but did not seem to specifically and consistently respond to a specific treatment (Fig. 15B), though year and the fertilization by litter interaction were significant (Table 10). Together, these two species accounted for nearly half (48.2%) of all mature spiders captured, and over three-quarters (78%) of all mature lycosid spiders.

The third most abundant spider was the linyphiid Eridantes erigonoides

(Emerton), with 788 mature specimens captured (7.4% of all mature spiders, 24.6% of mature linyphiid spiders). Similar to Pi. minutus, E. erigonoides did not consistently respond to any particular treatment in any particular year (Fig. 15C; Appendix F), but did have a significant response to year in the repeated-measures analysis (Table 10).

However, the fourth most abundant spider, the linyphiid Bathyphantes pallidus (Banks) with 759 mature specimens captured (7.1% of all mature spiders, 23.7% of linyphiid spiders), strongly responded to fertilization, year, and the interaction between fertilization and year, and a marginally significant response to litter in the repeated-measures analysis

(Fig. 15D; Table 10). Virtually absent from the study site during the first two years of the study, B. pallidus became a fairly common spider in fertilized plots during the final two years of the study, with a weak affiliation with plots where litter was left in situ (Fig.

15D; Appendix F). Also virtually absent from the site during the first two years of the study, the fifth most abundant spider, the linyphiid Halorates plumosus (Emerton) with

393 specimens captured (3.7% of all mature spiders, 12.3% of mature linyphiid spiders),

145

Table 10. Results of the repeated-measure PROC MIXED for each species. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter (L), year (Y), and (*) their fully factorial interactions.

Pardosa F1, 80 = 47.89 F1, 80 = 1.51 F3, 80 = 7.86 F1, 80 = 0.26 F3, 80 = 3.76 F3, 80 = 0.34 F3, 80 = 0.13 moesta P < 0.0001 P = 0.2232 P = 0.0001 P = 0.6112 P = 0.0141 P = 0.7968 P = 0.9394

Pirata F1, 80 = 0.29 F1, 80 = 0.61 F3, 80 = 7.91 F1, 80 = 7.18 F3, 80 = 1.83 F3, 80 = 0.89 F3, 80 = 1.53 minutus P = 0.5909 P = 0.4366 P = 0.0001 P = 0.0090 P = 0.14486 P = 0.4515 P = 0.2124

Bathyphantes F1, 80 = 26.09 F1, 80 = 4.26 F3, 80 = 27.15 F1, 80 = 0.18 F3, 80 = 7.84 F3, 80 = 1.01 F3, 80 = 0.30 pallidus P < 0.0001 P = 0.0422 P < 0.0001 P = 0.6711 P = 0.0001 P = 0.3906 P = 0.8235

Eridantes F1, 80 = 0.56 F1, 80 = 1.28 F3, 80 = 17.34 F1, 80 = 4.32 F3, 80 = 1.04 F3, 80 = 1.68 F3, 80 = 0.22 erigonoides P = 0.4570 P = 0.2612 P < 0.0001 P = 0.0409 P = 0.3773 P = 0.1773 P = 0.8826

Halorates F1, 80 = 8.00 F1, 80 = 0.06 F3, 80 = 7.11 F1, 80 = 0.22 F3, 80 = 3.93 F3, 80 = 0.23 F3, 80 = 0.22 plumosus P = 0.0059 P = 0.8003 P = 0.0003 P = 0.6387 P = 0.0114 P = 0.8719 P = 0.8799

Schizocosa F1, 80 = 22.76 F1, 80 = 7.32 F3, 80 = 15.75 F1, 80 = 5.26 F3, 80 = 11.10 F3, 80 = 5.56 F3, 80 = 4.78 avida P < 0.0001 P = 0.0083 P < 0.0001 P = 0.0245 P < 0.0001 P = 0.0016 P = 0.0041

146

also strongly responded to fertilization and year, which resulted in a significant interaction between fertilization and year (Fig. 15E; Table 10).

One of the most abundant of the larger spiders, the lycosid Schizocosa avida

(Walckenaer) with 140 mature specimens captured (1.3% of all mature spiders, 2.1% of mature lycosid spiders), was virtually absent from fertilized plots (Fig. 15F). In repeated- measures PROC MIXED analysis (Table 10) as well as the PROC MIXED analysis for

2005 (Appendix F), all variables were significant. The species was most commonly captured in litter removed treatments, particularly in unfertilized plots with litter removed

(-L/-F).

Aggregate ecosystem-level analyses

The NMS ordination with Araneae species richness showed clustering of plots into treatments (Fig. 16A), and the ordination axes explained 78.9% of the variance, with the first axis explaining 44.8% of the variance, and the second axis explaining 34.1% of the variance. The final stress = 6.23 with a final instability = 0.099, and results of the

Monte Carlo simulation indicated that this stress was less than expected by chance (P =

0.001). Following Clarke (1993), a final stress between 5 – 10 was a very good ordination and did not present any real risk of misinterpretation. The first axis separated fertilized and unfertilized plots with high correlations to PAR (R = -0.912) and soil moisture (R = -0.612) in the direction of unfertilized plots, and correlations to species richness (R = 0.347) and percent soil organic content (R = 0.446) in the direction of

147

Figure 16. Two-dimensional ordination of ecosystem-level properties from 2005 from nonmetric multidimensional scaling (NMS) using plant litter biomass, PAR, percent soil moisture, percent soil organic content, and (A) Araneae species richness or (B)

Linyphiidae species richness (Lycosidae did not produce s stable result). Vectors indicate the direction and strength of correlations between axis scores and emergent properties (R2 cutoff for joint Biplot = 0.100) and ordinations are rotated to the dominant axis of fertilization. The percent of variance explained by each axis is noted next to the axis title. Ellipses are hand drawn and are added only for visual convenience. See Fig.

11 for key to treatment symbols.

148

1.5 Plant Litter

0.5

Soil Moisture Soil Organic Content

Spider SR Axis 2 (34.1%) -0.5 PAR

+L/-F -L/-F +L/+F -1.5 -L/+F -1.5 -0.5 0.5 1.5 Axis 1 (44.8%)

0.5

Linyphiidae SR Soil Organic Content Soil Moisture PAR

-0.5 Axis 2 (21.2%)

+L/-F -L/-F Plant Litter +L/+F -1.5 -L/+F -1.5 -0.5 0.5 1.5 Axis 1 (71.7%)

149

fertilized plots, while plant litter biomass was not well correlated (R = -0.127). The second axis separated litter removed from litter left in situ plots with a good correlation to plant litter biomass (R = 0.916) in the direction of litter left in situ plots and PAR (R = -

0.529) in the direction of litter removed plots. Only weak or no correlations to the other three variables: spider species richness R = -0.230, percent soil moisture R = -0.144, and percent soil organic content R = -0.034.

While the NMS ordination with Lycosidae did not produce a stable result, the

NMS ordination with Linyphiidae species richness showed clustering of plots into treatments (Fig. 16B), and the ordination axes explained 92.9% of the variance, with the first axis explaining 71.7% of the variance, and the second axis explaining 21.2% of the variance. The final stress = 8.89 with a final instability = 0.056, and results of the Monte

Carlo simulation indicated that this stress was less than expected by chance (P = 0.002).

The first axis separated fertilized and unfertilized plots with high correlations to PAR (R

= -0.953) and soil moisture (R = -0.491) in the direction of unfertilized plots, and correlations to linyphiid species richness (R = 0.815) and percent soil organic content (R

= 0.466) in the direction of fertilized plots, while plant litter biomass was not well correlated (R = 0.183). The second axis separated litter removed from litter left in situ plots with a good correlation to plant litter biomass (R = 0.942) in the direction of litter left in situ plots and weakly correlated to linyphiid species richness (R = -0.359) and percent soil organic content (R = -0.297) in the direction of litter removed plots, and no correlations to the other two variables: PAR R = -0.039 and percent soil moisture R = -

0.034.

150

For both NMS ordinations, the separation of plots into treatment clusters was supported by MRPP (Table 11). When all four treatments were run together, the null hypothesis of no difference between treatments was rejected with high within-group agreement and very strong separation between groups. Pairwise comparisons of treatments showed that fertilized plots, while still significantly distinct, were more similar to each other than fertilized treatment plots are to any of the unfertilized treatment plots.

The same pattern existed for unfertilized plots, with strong separation of unfertilized plots, yet with lower dissimilarity than when unfertilized plots were compared to fertilized plots. As expected, the maximal differences occurred when extremes of treatments were paired, as in –L/-F vs. +L/+F, and +L/-F vs. –L/+F, indicating that

―opposite‖ treatments significantly alter biotic and abiotic components of the local habitat.

Discussion

Spider abundance increased as a result of fertilization, but neither spider biomass nor spider species richness were significantly affected, and therefore my first hypothesis was not supported. This result is contrary to previous studies in which arthropod diversity in fertilized plots decreased as abundance increased (e.g., Knops et al. 1999;

Haddad et al. 2000, 2001). However, my null result for the overall spider community is likely the results of a canceling effect of the responses of the two dominant spider families, the wolf spiders (Lycosidae) and the penny spiders (Linyphiidae). Wolf spider

151

Table 11. Results of Multi-response Permutation Procedure (MRPP) on emergent properties for 2005 to support NMS analyses (Fig. 15). T describes the separation between groups (dissimilarity) and A is the chance-corrected within-group agreement.

―All‖ indicates all four treatments included in the MRPP, and the remainders are MRPP pairwise comparisons of treatments to assess dissimilarity (lower T and higher A). +L indicates litter left in situ, -L indicates litter removed, +F indicates fertilization, -F indicates no fertilization.

Groups T A P

All Spiders All -9.892 0.421 < 0.0001 +L/-F vs. –L/-F -3.352 0.164 0.0097 +L/-F vs. +L/+F -5.838 0.332 0.0006 +L/-F vs. –L/+F -6.014 0.414 0.0010 -L/-F vs. +L/+F -6.470 0.424 0.0006 -L/-F vs. –L/+F -5.643 0.313 0.0009 +L/+F vs. –L/+F -4.999 0.313 0.0023

Linyphiidae All -10.055 0.437 < 0.0001 +L/-F vs. –L/-F -3.236 0.159 0.0102 +L/-F vs. +L/+F -6.395 0.373 0.0005 +L/-F vs. –L/+F -6.445 0.442 0.0008 -L/-F vs. +L/+F -6.501 0.435 0.0006 -L/-F vs. –L/+F -5.842 0.338 0.0010 +L/+F vs. –L/+F -5.188 0.284 0.0022

152

abundance was indeed significantly affected by fertilization, but wolf spider biomass and species richness were not affected. As nearly two-thirds of all spiders captured were wolf spiders, the response of this family drove the patterns found in the overall spider data, though it should be noted that my pitfall trap sampling method may have more bias towards wolf spiders due to their cursorial habit.

On the other hand, fertilization increased the abundance, biomass, and species richness of linyphiid spiders during the final two years of the study. The species richness response was actually completely opposite of the predictions of biodiversity-productivity theory (Haddad et al. 2000, 2001; Suding et al. 2005). This was likely a result of a bottom-up food web response to fertilization by the main penny spider food source, collembolans. Collembolans did not respond to the treatments despite the increased abundance of plant litter (unpublished data), with the basal resource for the collembolan prey being the bacteria that aid in the breakdown of plant litter. Thus, while the basal resource likely increased, the primary consumer of that resource, collembolans, did not, but the primary collembolan predator did increase in abundance and diversity. It is therefore feasible to propose that a top-down food web effect by linyphiid spiders limited collembolan abundance, ultimately enhancing their own abundance and diversity.

The differences in responses between wolf and penny spiders are likely the results of different foraging behaviors. Spider guilds are based primarily on foraging behavior, habitat preferences, and web type (Uetz et al. 1999). Based upon this classification system, wolf spiders are considered ground-running spiders and linyphiid spiders are characterized as wandering sheet/tangle web weavers. Wolf spiders are classic epigeal

153

wandering spiders that actively hunt for prey and are largely restricted to hunting in two- dimensional space. Thus, habitat structure (i.e., the physical structure of the surrounding environment, including plant litter and living plant material) may enhance wolf spider hunting success by providing additional hiding places for ambush hunting and for lairs

(Rypstra et al. 1999; Halaj et al. 2000). However, too much habitat structure can also increase predation and intraguild predation risk through increased density responses to habitat structure while inhibiting movement (Wise 2006; Rypstra et al. 2007). These factors reduce both abundance and hunting success, thereby reducing the numbers of wolf spiders in an area. In my study, the increased habitat structure that resulted from fertilization seemed to moderately (but not significantly) reduce wolf spider species richness while increasing wolf spider abundance. However, this increased abundance was likely due to the population explosion in fertilized plots of the medium-sized wolf spider, P. moesta (see below).

Penny spiders rely more upon webs for prey capture, sometimes maintaining and patrolling multiple webs (Uetz et al. 1999). While these webs are generally constructed at or close to the ground level, the webs can enhance prey capture space to include a portion of a third dimension. Moreover, increased habitat structure can provide additional structure for web building (Rypstra et al. 1999). Thus, while fertilized plots significantly reduced wolf spider species richness probably due to the enhanced habitat structure that impeded foraging ability, these plots may have provided the tiny web building linyphiid spiders the habitat structure to flourish because of the increased structure for web building, and, thus, increased prey capture rates.

154

A medium-sized wolf spider, P. moesta thrived in fertilized plots probably due to decreased intraguild predation by larger wolf spiders (e.g., S. avida) that became less abundant in fertilized plots. Moreover, there was increased abundance of potential prey in fertilized plots (Patrick et al. 2008b). These two factors together likely released P. moesta from competition and predation, resulting in an increased abundance in fertilized plots. However, these microhabitat changes likely caused the decreased abundance of S. avida in fertilized plots, as the increased habitat structure likely impeded this species’ foraging abilities. Pirata minutus was one of the smallest wolf spiders captured at my site, and could have benefited from the increased habitat structure in a similar way to P. moesta. While it was the second most abundant spider captured during the course of my study, it did not significantly respond to a single habitat, except for unfertilized plots with litter removed during 2005. Interestingly, Pi. minutus was observed on several occasions in the jaws of P. moesta and S. avida, making Pi. minutus a victim of intraguild predation.

Spider species richness was not significantly correlated with either plant species richness or standing crop biomass, therefore I reject the second and third hypotheses.

However, both dominant spider families responded to fertilization (by the fourth year of the study) in distinctly different ways. Wolf spiders followed predictions of current biodiversity-productivity theory, with decreased species richness associated with decreased plant species richness and increased standing crop biomass. While fertilization increased wolf spider abundance, wolf spider species richness was more closely tied to

155

plant species richness and therefore decreased as nutrient loading into the system increased.

Finally, despite documented effects of increased habitat structure on arthropod abundances and diversity (e.g., Lawton 1983; Halaj et al. 2000), particularly for spiders

(e.g., Uetz 1991; Rypstra et al. 1999; Halaj et al. 2000), my results do not support my fourth hypothesis. Plant litter had no significant effect on spider species richness. Most studies of spider responses to plant litter have been conducted in plant monocultures in agroecosystems (e.g., Rypstra et al. 1999). These managed ecosystems tend to have much higher disturbance and more bare ground than would be expected from a grassland.

Thus, increased refugia via plant litter additions to these agroecosystems would certainly provide more habitat than the existing bare ground, so it is perhaps not surprising that there have been stronger responses to plant litter in agroecosystems.

Analysis of the spider community and associated abiotic variables demonstrated strong treatment effects. These highly differentiated treatments are likely to have a strong effect on ecosystem properties (e.g., nutrient cycling, carbon sequestering), an effect likely to increase through time as the treatment plots further mature. Spiders have been shown to affect detritivore abundance (Wise et al. 1999), thereby indirectly altering nutrient cycling within the system (Chen and Wise 1999). The results of my ordinations clearly showed that my plots responded to my treatments and that the spider community affected ecosystem-level processes. The long-term implications are unknown, but it is clear that the trajectories of each treatment are significantly different and may impact ecosystem function and services. To my knowledge, this is the first time that these biotic

156

and abiotic factors have been coupled in a multivariate ordination to explicitly determine whether they can define discrete and distinct predator communities and their associated abiotic properties in the context of the biodiversity-productivity theory. Most previous work (e.g., Haddad et al. 2000, 2001) did not attempt to associate the invertebrate community with abiotic changes resulting from fertilization, and I know of no other studies that coupled fertilization and plant litter effects to test predictions of biodiversity- productivity theory.

The diversity and community structure of spiders and other arthropods are sensitive to plot size (Martinko et al. 2006). The large size of my experimental plots integrated important determinants of the within-plot plant communities, including spatial heterogeneity (De Boeck et al. 2006), leaching of nutrients from litter (Berendse 1998), local nutrient cycling (Hooper and Vitousek 1998), and the translocation of nutrients within clumping and clonal plants (Hutchings and Bradbury 1986), which are the primary growth forms of my dominant graminoids (Patrick et al. 2008a). These spatial factors are also important to epigeal spiders because of their vagility and their need to find suitable food; the larger plot sizes more realistically emulate natural habitat patches of varying quality and can support higher insect diversity (Martinko et al. 2006). Other studies that examined the effects of nutrient loading on arthropod communities had plot sizes ranging from 9 m2 – 169 m2 (e.g., Knops et al. 1999; Haddad et al. 2001), making my experimental plots (314 m2) nearly twice as large—an important factor when considering the vagility of some spider species.

157

However, I realize that my study has some distinct differences when compared to previous work. My use of an NPK fertilizer, as opposed to N-only fertilizer, is likely to have induced a stronger response to fertilization due to the added P and K. Nevertheless, my plant results (see Patrick et al. 2008a) were generally consistent with other plant studies that used NPK fertilizers (e.g., Carson and Barrett 1988; Turkington et al. 2002) and N-only fertilizers (e.g., Haddad et al. 2000; Tilman et al. 2002), which allowed me to formulate my epigeal spider hypotheses on the same bases as previous studies that investigated the responses of arthropods to nutrient loading. Further, my running definition of litter (see Methods) includes the vegetation mown in the previous year and not removed from litter left in situ treatment plots, potentially altering the nutritional quality of the litter relative to naturally senesced vegetation, and the physical structure of the litter as it lay after mowing (e.g., Semmartin et al. 2004). Because the timing of the mowing was determined by the local township, litter from the annual mowing accumulated earlier than might normally be expected for this region of the USA.

However, were the mowing to stop, the site would very quickly yield to encroaching woody vegetation typical of early secondary succession.

My study underscores the disjunct between conventional and plant-based biodiversity-productivity theory and the animal component of the food web, particularly epigeal predators. This portion of the food web is more closely associated with the quality of its basal resource (plant litter) than with the diversity of that resource (Cross et al. 2006; Seeber et al. 2008). This starkly contrasts the more aerial portion of the food web that is more dependent on living plants, where specialist herbivores can be affected

158

by plant diversity more than by plant quality. Ultimately, the loss of plant species with increased nutrient loading may result in the loss of arthropod herbivores and their specialist predators and parasites. However, the increases in diversity may be balanced by the epigeal community and its different resource base.

159

References

Barrios, E. 2007. Soil biota, ecosystem services and land productivity. Ecological

Economics 64:269–285.

Berendse, F. 1998. Effects of dominant plant species on soil during succession in

nutrient-poor ecosystems. Biogeochemistry 42:73–88.

Bobbink, R., M. Hornung, and J. G. M. Roelofs. 1998. The effects of air-borne nitrogen

pollutants on species diversity in natural and semi-natural European vegetation.

Journal of Ecology 86:717–738.

Carson, W. P. and G. W. Barrett. 1988. Succession in old-field plant communities:

effects of contrasting types of nutrient enrichment. Ecology 69:984–994.

Chalcraft, D. R., S. B. Cox, C. Clark, E. E. Cleland, K. N. Suding, E. Weiher, and D.

Pennington. 2008. Scale-dependent responses of plant biodiversity to nitrogen

enrichment. Ecology 89:2165–2171.

Chen, B., and D. H. Wise. 1999. Bottom-up limitation of predaceous arthropods in a

detritus-based terrestrial food web. Ecology 80:761–772.

Clarke, K. R. 1993. Non-parametric multivariate analyses of changes in community

structure. Australian Journal of Ecology 18:117–143.

Cross, W. F., J. B. Wallace, A. D. Rosemond, and S. L. Eggert. 2006. Whole-system

nutrient enrichment increases secondary production in a detritus-based system.

Ecology 87:1556–1565.

160

De Boeck, H. J., I. Nijs, C. M. H. M. Lemmens, and R. Ceulemans. 2006. Underlying

effects of spatial aggregation (clumping) in relationships between plant diversity

and resource uptake. Oikos 113:269–278.

Draney, M. L., and D. J. Buckle. 2005. Linyphiidae. In: Ubick, D. P., P. Paquin, P. E.

Cushing, and V. Roth, eds., Spiders of North America: an identification manual.

American Arachnological Society, pp. 124–161.

Fenn, M.E., J. S. Baron, E. B. Allen, H. M. Rueth, K. R. Nydick, L. Geiser, W. D.

Bowman, J. O. Sickman, T. Meixner, D. W. Johnson, and P. Neitlich. 2003.

Ecological effects of nitrogen deposition in the western United States.

BioScience 53:404–420.

Foster, B. L., and K. L. Gross. 1998. Species richness in a successional grassland:

effects of nitrogen enrichment and plant litter. Ecology 79:2593–2602.

Galloway, J.N., J. D. Aber, J. W. Erisman, S. P. Seitzinger, R. W. Howarth, E. B.

Cowling, and B. J. Cosby. 2003. The nitrogen cascade. BioScience 53:341–355.

Greenslade, P. J. M. 1694. Pitfall trapping as a method for studying populations of

Carabidae (Coleoptera). Journal of Animal Ecology 33:301–310.

Grime, J. P. 1973. Control of species density in herbaceous vegetation. Journal of

Environmental Management 1:151–167.

Haddad, N. M., J. Haarstad, and D. Tilman. 2000. The effects of long-term nitrogen

loading on grassland insect communities. Oecologia 124:73–84.

161

Haddad, N. M., D. Tilman, J. Harstaad, M. Ritchie, and J. M. H. Knops. 2001.

Contrasting effects of plant richness and composition on insect communities: a

field experiment. American Naturalist 158:17–35.

Halaj, J., A. B. Cady, and G. W. Uetz. 2000. Modular habitat refugia enhance generalist

predators and lower plant damage in soybeans. Environmental Entomology

29:383–393.

Halaj, J. and D. H. Wise. 2002. Impact of a detrital subsidy on trophic cascades in a

terrestrial grazing web. Ecology 83:3141–3151.

Hättenschwiler, S., and P. Gasser. 2005. Soil animals alter plant litter diversity effects

on decomposition. Proceedings of the National Academy of Sciences of the USA

102:1519–1524.

Hättenschwiler, S., A. V. Tiunov, and S. Scheu. 2005. Biodiversity and litter

decomposition in terrestrial ecosystems. Annual Review of Ecology, Evolution

and Systematics 36:191–218.

Hector, A., B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G.

Dimitrkopoulos, J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P.

Högberg, K. Huss-Danell, J. Joshi, A. Jumpponen, C. Körner, P. W. Leadley, M.

Loreau, A. Minns, C. P. H. Mulder, G. O’Donovan, S. J. Otway, J. S. Pereira, A.

Prinz, D. J. Read, M. Scherer-Lorenzen, E-D. Schulze, A-S. D. Siamantziouras,

E. M. Spehn, A. C. Terry, A. Y. Troumbis, F. I. Woodward, S. Yachi, J. H.

Lawton. 1999. Plant diversity and productivity experiments in European

grasslands. Science 286:1123–1127.

162

Hooper, D. U. and P. M. Vitousek, 1998. Effects of plant competition and diversity on

nutrient cycling. Ecological Monographs 68:121–149.

Hutchings, M. J. and I. K. Bradbury. 1986. Ecological perspectives on clonal perennial

plants. BioScience 36:178–182.

Knops, J. M. H., D. Tilman, N. M. Haddad, S. Naeem, C. E. Mitchell, J. Haarstad, M. E.

Ritchie, K. M. Howe, P. B. Reich, E. Siemann, and J. Groth. 1999. Effects of

plant species richness on invasions dynamics, disease outbreaks, insect

abundances and diversity. Ecology Letters 2:286–293.

Köchy, M., and S. D. Wilson. 2005. Variation in nitrogen deposition and available soil

nitrogen in a forest-grassland ecotone in Canada. Landscape Ecology 20:191–

202.

Kruskal, J. B. 1964. Nonmetric multidimensional scaling: a numerical method.

Psychometrika 29:115–129.

Larsen, T. H., N. Williams, and C. Kremen. 2005. Extinction order and altered

community structure rapidly disrupt ecosystem functioning. Ecology Letters

8:538–547.

Lawrence, K. L., and D. H. Wise. 2004. Unexpected indirect effects of spiders on the

rate of litter disappearance in a deciduous forest. Pedobiologia 48:149–157.

Lawton, J. H. 1983. Plant architecture and the diversity of phytophagous insects.

Annual Review of Entomology 28:23–29.

163

Lensing, J. R., S. Todd, and D. H. Wise. 2005. The impact of altered precipitation on

spatial stratification and activity-densities of springtails (Collembola) and spiders

(Araneae). Ecological Entomology 30:194–200.

Long, Z. T., C. L. Mohler, and W. P. Carson. 2003. Extending the resource

concentration Hypothesis to plant communities: effects of litter and herbivores.

Ecology 84:652–665.

Martinko, E. A., R. H. Hagen, and J. A. Griffith. 2006. Successional change in the insect

community of a fragmented landscape. Landscape Ecology 21:711–721.

Mazzoleni, S., G. Bonanomi, F. Giannino, M. Rietkerk, S. C. Dekker, and F. Zucconi.

2007. Is plant biodiversity driven by decomposition processes? An emerging new

theory on plant diversity. Community Ecology 8:103–109.

McCann, K.S. 2000. The diversity-stability debate. Nature 405:228–233.

McCune, B., and M. J. Mefford. 2006. Multivariate Analysis of Ecological Data

Version 5.14. MjM Software, Gleneden Beach, OR, USA.

Mielke, P. W., Jr. 1984. Meteorological applications of permutations techniques based

on distance functions. In: Krishnaiah, P. R., and P. K. Sen, eds., Handbook of

Statistics, Volume 4. Elsevier, St. Louis, MO, USA, pp. 813–830.

Moore, J. C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C.

Johnson, K. S. McCann, K. Melville, P. J. Morin, A. D. Rosemond, D. M. Post, J.

L. Sabo, K. M. Scow, M. J. Vanni, and D. H. Wall. 2004. Detritus, trophic

dynamics, and biodiversity. Ecology Letters, 7:584–600.

164

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008a. Large-scale manipulation of

plant litter and fertilizer in a managed successional temperate grassland. Plant

Ecology 197:183–195.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008b. ―Brown‖ world invertebrates

contradict ―green‖ world biodiversity theory. Research Letters in Ecology, vol.

2008, Article ID 694638, 4 pages, doi:10.1155/2008/694638.

Phillips, I. D., and T. P. Cobb. 2005. Effect of habitat structure and lid transparency on

pitfall catches. Environmental Entomology 34:875–882.

Platnick, N.I. 2007. The World Spider Catalog, version 9.0. American Museum of

Natural History, online at

http://research.amnh.org/entomology/spiders/catalog/index.html.

Ricklefs, R. E., and G. L. Miller. 2000. Ecology. Freeman, NY, USA.

Ritchie, A., and J. R. Steiger. 1974. Soil Survey of Summit County Ohio. United States

Department of Agriculture, Soil Conservation Service in cooperation with Ohio

Department of Natural Resources, Division of Lands and Soil and the Ohio

Agricultural Research and Development Center.

Rypstra, A. L., P. E. Carter, R. A. Balfour, and S. D. Marshall. 1999. Architectural

modifications of agricultural habitats and their impact on the spider inhabitants.

Journal of Arachnology 27:371–377.

Rypstra, A. L., J. M. Schmidt, B. D. Reif, J. DeVito, and M. H. Persons. 2007.

Tradeoffs involved in site selection and foraging in a wolf spider: effects of

substrate structure and predation risk. Oikos 116:853–863.

165

Sala, O. E., W. J. Warton, L. A. Joyce, and W. K. Lauenroth. 1988. Primary production

of the central grasslands region of the United States. Ecology 69:40–45.

SAS Institute Inc. 1999. SAS System for Windows. SAS Institute Inc., Cary, NC, USA.

Seeber, J., G. U. H. Seeber, R. Langel, S. Scheu, and E. Meyer. 2008. The effect of

macro-invertebrates and plant litter of different quality on the release of N from

litter to plant on alpine pastureland. Biology and Fertility of Soils 44:783–790.

Semmartin, M., M. R. Aguiar, R. A. Distel, A. S. Moretto, and C. M. Ghersa. 2004.

Litter quality and nutrient cycling affected by grazing-induced species

replacements along a precipitation gradient. Oikos 107:148–160.

Snyder, W. E., and D. H. Wise. 2001. Contrasting trophic cascades generated by a

community of generalist predators. Ecology 86:1571–1583.

Strong, D. R., J. H. Lawton, and R. Southwood. 1984. Insects on plants: community

patterns and mechanisms. Harvard University Press, Cambridge, MA, USA.

Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.

Milchunas, and S. Pennings. 2005. Functional- and abundance-based

mechanisms explain diversity loss due to N fertilization. Proceedings of the

National Academy of Sciences of the USA 102:4387–4392.

Tilman, D., J. Knops, D. Wedin, P. Reich. 2002. Experimental and observational studies

of diversity, productivity, and stability. In: Kinzig, A. P., S. W. Pacala, and D.

Tilman, eds., The Functional Consequences of Biodiversity. Princeton University

Press, Princeton, NJ, USA, pp. 42–70.

166

Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of

fertilization and herbivory on the herbaceous vegetation of the boreal forest in

north-western Canada: a 10-year study. Journal of Ecology 90:325–337.

Uetz, G. W. 1979. The influence of variation in litter habitats on spider communities.

Oecologia 40:29–42.

Uetz, G. W. 1991. Habitat structure and spider foraging. In: Bell, S. S., E. D. McCoy,

and H. R. Mushinsky, eds., Habitat structure: the physical arrangement of

objects in space. Chapman and Hall, London, UK, pp. 325–348.

Uetz, G. W., J. Halaj, and A. B. Cady. 1999. Guild structure of spiders in major crops.

Journal of Arachnology 27:270–280.

US EPA. 2005. Clean air status and trends network (CAST-NET).

http://www.epa.gov/castnet/

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human

domination of Earth’s ecosystems. Science 277:494–499.

Wardle, D. A. 2002. Communities and Ecosystems: Linking the Aboveground and

Belowground Components. Princeton University Press, Princeton, NJ, USA.

Wardle, D. A. 2006. The influence of biotic interactions on soil biodiversity. Ecology

Letters 9:870–886.

Wardle, D. A., G. W. Yeates, G. M. Barker, and K. I. Bonner. 2006. The influence of

plant litter diversity on decomposer abundance and diversity. Soil Biology and

Biochemistry 38:1052–1062.

167

Wise, D. H. 1993. Spiders in Ecological Webs. Cambridge University Press,

Cambridge, UK.

Wise, D. H. 2004. Wandering spiders limit densities of a major microbe-detritivore in

the forest-floor food web. Pedobiologia 48:181–188.

Wise, D. H. 2006. Cannibalism, food limitation, intraspecific competition, and the

regulation of spider populations. Annual Review of Entomology 51:441–465.

Wise, D. H., W. E. Snyder, P. Tuntibunpakul and J. Halaj. 1999. Spiders in

decomposition food webs of agroecosystems: theory and evidence. Journal of

Arachnology 27:363–370.

Worm, B., and J. E. Duffy. 2003. Biodiversity, productivity, and stability in real food

webs. Trends in Ecology and Evolution 18:628–632.

168

Appendix A. Results of the PROC MIXED analysis for each response variable during

2002

Response Variable F L F*L

H' e F1, 20 = 4.58, P = 0.0449 F1, 20 = 0.00, P = 0.9784 F1, 20 = 1.87, P = 0.1867

Araneae SR F1, 20 = 3.53, P = 0.0750 F1, 20 = 0.06, P = 0.8038 F1, 20 = 0.83, P = 0.3736

Araneae abundance F1, 20 = 0.13, P = 0.7214 F1, 20 = 0.53, P = 0.4742 F1, 20 = 0.29, P = 0.5954

Araneae biomass F1, 20 = 0.18, P = 0.6782 F1, 20 = 0.20, P = 0.6609 F1, 20 = 0.14, P = 0.7097

Linyphiidae SR F1, 20 = 0.47, P = 0.5020 F1, 20 = 1.43, P = 0.2455 F1, 20 = 0.18, P = 0.6737

Linyphiidae abundance F1, 20 = 0.15, P = 0.7030 F1, 20 = 3.60, P = 0.0724 F1, 20 = 0.52, P = 0.4809

Linyphiidae biomass F1, 20 = 0.98, P = 0.3333 F1, 20 = 0.01, P = 0.9350 F1, 20 = 0.49, P = 0.4906

Lycosidae SR F1, 20 = 0.33, P = 0.5694 F1, 20 = 0.18, P = 0.6759 F1, 20 = 0.54, P = 0.4722

Lycosidae abundance F1, 20 = 1.49, P = 0.2361 F1, 20 = 0.01, P = 0.9165 F1, 20 = 0.91, P = 0.3505

Lycosidae biomass F1, 20 = 0.42, P = 0.5242 F1, 20 = 0.09, P = 0.7734 F1, 20 = 0.09, P = 0.7645

169

(Appendix A continued)

2003

F L F*L

H' e F1, 20 = 11.04, P = 0.0034 F1, 20 = 2.06, P = 0.1666 F1, 20 = 0.06, P = 0.8138

Araneae SR F1, 20 = 8.26, P = 0.0094 F1, 20 = 0.74, P = 0.3988 F1, 20 = 0.01, P = 0.9246

Araneae abundance F1, 20 = 5.00, P = 0.0369 F1, 20 = 0.35, P = 0.5621 F1, 20 = 0.01, P = 0.9119

Araneae biomass F1, 20 = 0.82, P = 0.3758 F1, 20 = 0.15, P = 0.7055 F1, 20 = 0.05, P = 0.8318

Linyphiidae SR F1, 20 = 0.51, P = 0.4841 F1, 20 = 3.50, P = 0.0760 F1, 20 = 1.56, P = 0.2265

Linyphiidae abundance F1, 20 = 0.73, P = 0.4044 F1, 20 = 2.05, P = 0.1678 F1, 20 = 2.56, P = 0.1249

Linyphiidae biomass F1, 20 = 1.35, P = 0.2585 F1, 20 = 0.08, P = 0.7826 F1, 20 = 0.75, P = 0.3953

Lycosidae SR F1, 20 = 3.45, P = 0.0778 F1, 20 = 0.19, P = 0.6682 F1, 20 = 0.56, P = 0.4612

Lycosidae abundance F1, 20 = 7.36, P = 0.0134 F1, 20 = 0.01, P = 0.9201 F1, 20 = 0.30, P = 0.5892

Lycosidae biomass F1, 20 = 1.38, P = 0.2532 F1, 20 = 0.14, P = 0.7129 F1, 20 = 0.04, P = 0.8487

170

(Appendix A continued)

2004

F L F*L

H' e F1, 20 = 1.06, P = 0.3150 F1, 20 = 0.04, P = 0.8348 F1, 20 = 2.09, P = 0.1641

Araneae SR F1, 20 = 2.70, P = 0.1162 F1, 20 = 0.02, P = 0.8828 F1, 20 = 1.65, P = 0.2139

Araneae abundance F1, 20 = 16.44, P = 0.0006 F1, 20 = 0.27, P = 0.6103 F1, 20 = 0.34, P = 0.5674

Araneae biomass F1, 20 = 4.72, P = 0.0419 F1, 20 = 0.00, P = 0.9581 F1, 20 = 0.05, P = 0.8259

Linyphiidae SR F1, 20 = 23.06, P = 0.0001 F1, 20 = 0.01, P = 0.9161 F1, 20 = 2.56, P = 0.1251

Linyphiidae abundance F1, 20 = 25.68, P < 0.0001 F1, 20 = 0.39, P = 0.5395 F1, 20 = 0.98, P = 0.3333

Linyphiidae biomass F1, 20 = 13.16, P = 0.0017 F1, 20 = 0.58, P = 0.4566 F1, 20 = 0.08, P = 0.7795

Lycosidae SR F1, 20 = 0.02, P = 0.8851 F1, 20 = 0.06, P = 0.8098 F1, 20 = 0.00, P = 0.9616

Lycosidae abundance F1, 20 = 5.74, P = 0.0265 F1, 20 = 0.06, P = 0.8152 F1, 20 = 0.00, P = 0.9793

Lycosidae biomass F1, 20 = 4.58, P = 0.0449 F1, 20 = 0.01, P = 0.9345 F1, 20 = 0.14, P = 0.7123

171

Appendix B. Araneae (A–C), linyphiid (D–F), and lycosid (G–I) species richness regressed against their abundances during each year, 2002–2004. Open circles (○) represent the control treatment plots of unfertilized and litter left in situ, open triangles

(∆) represent unfertilized and litter removed plots, filled circles (●) represent fertilized and litter left in situ plots, and filled triangles (▲) represent fertilized and litter removed plots.

A. B. C.

8 10 10

s s

2002 s 2003 2004

s s

e e

n n

h 8 8

h h

c 6

c c

i i

r r

s s

e 6 6

e e

i i

c c

4 c

e e

p p

s s

s 4 4

e e

a a

2 a

e e e R² = 44.4% P < 0.001

n 2 2 n R² = 18.5% P = 0.036 n

a R² = 3.1% P = 0.411

a a

r r

A A 0 A 0 0 0 3 6 9 12 15 0 5 10 15 20 25 0 5 10 15 20 25 Araneae abundance Araneae abundance Araneae abundance

D. E. F.

s s

2 s 2 4

s s

e e

e 2003 2004

n n

h h

c c

i i

r r

r 3

s s

e e

i i

c c

e e

1 e 1 2

p p

s s

e e

a a

R² = 38.1% P < 0.001 a

d d

i R² = 78.9% P < 0.001 i i R² = 75.9% P < 0.001

i 1

i i

h h

p p

y y

n n

i i

L L 0 L 0 0 0 1 2 3 4 5 0 1 2 3 4 5 0 2 4 6 8 Linyphiidae abundance Linyphiidae abundance Linyphiidae abundance G. H. I.

s s

s 3 3

s s

e 2003 2004

e e

n n

h h

c c

i i

r r

s 2 s

s 2 2

e e

i i

c c

e e

p p

s s

e e

1 e 1 R² = 8.1% P = 0.177 1

a a

d d

i i

s s

R² = 11.1% P = 0.112 s R² = 24.0% P = 0.015

o o

c c

y y

L L 0 L 0 0 0 5 10 15 0 5 10 15 20 0 5 10 15 Lycosidae abundance Lycosidae abundance Lycosidae abundance

172

Appendix C. Araneae (A–C), linyphiid (D–F), and lycosid (G–I) species richness regressed against their biomass during each year, 2002–2004. Open circles (○) represent the control treatment plots of unfertilized and litter left in situ, open triangles (∆) represent unfertilized and litter removed plots, filled circles (●) represent fertilized and litter left in situ plots, and filled triangles (▲) represent fertilized and litter removed plots.

A. B. C.

6 6 8

s s

s 2003 2004

s s

e e

n n

n 5 5

h h

c 6

c c

i i

r r

4 4

s s

e e

i i

c c

c 3 3 4

e e

p p

s s

s R² = 34.1% P = 0.003

2 e R² = 13.9% P = 0.073 e

e R² = 23.1% P = 0.017

a a

a 2

e e

n n

n 1 1

a a

r r

A A A 0 0 0 0.000 0.025 0.050 0.075 0.000 0.025 0.050 0.075 0.000 0.025 0.050 0.075 Araneae biomass (g) Araneae biomass (g) Araneae biomass (g)

D. E. F.

s s

s 2 2 4

s s

s 2003

e e

e 2004

i i

r 3

e e

e 1 1 2

p p

p R² = 38.1% P < 0.001 R² = 13.0% P = 0.084

a d

d 1

i i

i R² = 35.9% P = 0.002

n i i 0

i 0 0

L L 0.0000 0.0002 0.0004 0.0006 0.0000 0.0002 0.0004 0.0006 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 Linyphiidae biomass (g) Linyphiidae biomass (g) Linyphiidae biomass (g) G. H. I.

3 s 3 s

s 3

s s

s 2003 2004

e e

n n

h h

c c

i i

r r

2 2 2

s s

e e

i i

c c

e e

p p

s s

s R² = 25.7% P = 0.011

1 1 1

e e

e R² = 23.9% P = 0.015

a a

d d

i i

s s

R² = 41.6% P < o

o o

c c

y 0 y

y 0 0.001 0

L L 0.000 0.025 0.050 0.000 0.025 0.050 0.075 0.000 0.025 0.050 0.075 Lycosidae biomass (g) Lycosidae biomass (g) Lycosidae biomass (g)

173

Appendix D. Araneae (A), linyphiid (B), and lycosid (C) species richness regressed against 2004 plant species richness. Open circles (○) represent the control treatment plots of unfertilized and litter left in situ, open triangles (∆) represent unfertilized and litter removed plots, filled circles (●) represent fertilized and litter left in situ plots, and filled triangles (▲) represent fertilized and litter removed plots.

A. B.

8 s 4 s

s 2004 2004

s e

c h

i 3 r

c 6

s e

i e

c 4 2

e a

a i

2 i 1 e

h R² = 4.7% P = 0.308

n R² = 0.5% P = 0.734

A i

0 L 0 0 5 10 15 20 0 5 10 15 20 Plant species richness Plant species richness C.

3 s

e n

s 2

e 1

s o

c R² = 1.6% P = 0.554 y

L 0 0 5 10 15 20 Plant species richness

174

Appendix E. Araneae (A), linyphiid (B), and lycosid (C) species richness regressed against 2004 standing crop biomass. Open circles (○) represent the control treatment plots of unfertilized and litter left in situ, open triangles (∆) represent unfertilized and litter removed plots, filled circles (●) represent fertilized and litter left in situ plots, and filled triangles (▲) represent fertilized and litter removed plots.

A. B.

10 s 4 s s 2004

2004 e

h n

8 c

h i

r 3

e s

6 i

i e

c 2

s p

e d

a 1 i

i R² = 37.0% P = 0.002 e

2 h n

R² = 6.1% P = 0.246 p

n i A 0

0 L 250 500 750 1000 1250 1500 1750 250 500 750 1000 1250 1500 1750 Standing crop biomass (g) Standing crop biomass (g) C.

5 s

s 2004

e n

h 4

e 3 i

c R² = 0.0% P = 0.938

e p

s 2

a d

i 1

o c

y 0 L 250 500 750 1000 1250 1500 1750 Standing crop biomass (g)

175

Appendix F. Results of the yearly PROC MIXED analysis for the individual species presented in Fig. 14. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter (L), (F*L) their factorial interaction.

Response Variable F L F*L

Pardosa moesta

2002 F1, 20 = 3.59, P = 0.0726 F1, 20 = 0.28, P = 0.6056 F1, 20 = 0.35, P = 0.5603

2003 F1, 20 = 10.68, P = 0.0038 F1, 20 = 0.10, P = 0.7505 F1, 20 = 0.04, P = 0.8466

2004 F1, 20 = 9.60, P = 0.0057 F1, 20 = 0.07, P = 0.7965 F1, 20 = 0.02, P = 0.8974

2005 F1, 20 = 25.38, P < 0.0001 F1, 20 = 1.51, P = 0.2331 F1, 20 = 0.29, P = 0.5989

Pirata minutus

2002 F1, 20 = 0.01, P = 0.9261 F1, 20 = 0.04, P = 0.8530 F1, 20 = 1.57, P = 0.2251

2003 F1, 20 = 0.86, P = 0.3637 F1, 20 = 0.00, P = 0.9630 F1, 20 = 2.01, P = 0.1716

2004 F1, 20 = 1.71, P = 0.2062 F1, 20 = 1.20, P = 0.2855 F1, 20 = 0.46, P = 0.5047

2005 F1, 20 = 3.16, P = 0.0906 F1, 20 = 2.52, P = 0.1278 F1, 20 = 5.51, P = 0.0293

Bathyphantes pallidus

2002 F1, 20 = 0.42, P = 0.5249 F1, 20 = 1.44, P = 0.2435 F1, 20 = 0.21, P = 0.6489

2003 F1, 20 = 3.19, P = 0.0893 F1, 20 = 3.19, P = 0.0893 F1, 20 = 0.98, P = 0.3330

2004 F1, 20 = 12.47, P = 0.0021 F1, 20 = 0.66, P = 0.4247 F1, 20 = 0.01, P = 0.9178

2005 F1, 20 = 12.86, P = 0.0018 F1, 20 = 3.92, P = 0.0617 F1, 20 = 0.74, P = 0.4004

176

(Appendix F continued)

Response Variable F L F*L

Eridantes erigonoides

2002 F1, 20 = 0.63, P = 0.4373 F1, 20 = 1.99, P = 0.1742 F1, 20 = 1.03, P = 0.3232

2003 F1, 20 = 0.46, P = 0.5065 F1, 20 = 1.33, P = 0.2621 F1, 20 = 1.24, P = 0.2784

2004 F1, 20 = 0.70, P = 0.4132 F1, 20 = 1.37, P = 0.2558 F1, 20 = 3.38, P = 0.0809

2005 F1, 20 = 3.71, P = 0.0684 F1, 20 = 1.30, P = 0.2677 F1, 20 = 1.68, P = 0.2093

Halorates plumosus

2002 F1, 20 = 3.18, P = 0.0896 F1, 20 = 1.62, P = 0.2172 F1, 20 = 1.62, P = 0.2172

2003 F1, 20 = 0.34, P = 0.5639 F1, 20 = 0.34, P = 0.5639 F1, 20 = 1.12, P = 0.3034

2004 F1, 20 = 6.06, P = 0.0231 F1, 20 = 0.12, P = 0.7288 F1, 20 = 0.01, P = 0.9079

2005 F1, 20 = 4.91, P = 0.0384 F1, 20 = 0.19, P = 0.6650 F1, 20 = 0.23, P = 0.6385

Schizocosa avida

2002 F1, 20 = 3.10, P = 0.0935 F1, 20 = 0.57, P = 0.4592 F1, 20 = 0.57, P = 0.4592

2003 F1, 20 = 10.14, P = 0.0047 F1, 20 = 1.13, P = 0.3013 F1, 20 = 2.21, P = 0.1530

2004 F1, 20 = 3.46, P = 0.0776 F1, 20 = 0.38, P = 0.5421 F1, 20 = 0.38, P = 0.5421

2005 F1, 20 = 14.61, P = 0.0011 F1, 20 = 6.42, P = 0.0198 F1, 20 = 5.18, P = 0.0340

CHAPTER 5

REVIEW OF THE NEARCTIC GENUS SCYLETRIA BISHOP AND CROSBY

(ARANEAE, LINYPHIIDAE), WITH A TRANSFER OF S. JONA TO

MERMESSUS O. PICKARD-CAMBRIDGE

(This chapter was published April 2008 in Zootaxa)

Abstract

The genus Scyletria Bishop and Crosby 1938 is reviewed and reduced to its type species, Scyletria inflata Bishop and Crosby 1938, by transfer of the only other species in the genus, Scyletria jona Bishop and Crosby 1938, to Mermessus O. Pickard-Cambridge

1899. The male of S. inflata is re-described, the female of M. jona (Bishop and Crosby

1938) new combination is described for the first time, and the male is re-described.

Introduction

Bishop and Crosby (1938) erected the genus Scyletria for two erigonine linyphiid spiders, Scyletria jona and the type species S. inflata. The two species were placed

177 178

together in the genus Scyletria “because of the similarity in the structure of the embolic division of the male palpus” (Bishop and Crosby 1938:89). However, I find that the two species differ significantly and that it is necessary to transfer S. jona to the genus

Mermessus O. Pickard-Cambridge 1899, creating the new combination Mermessus jona

(Bishop and Crosby 1938). The transfer of this species is based on the generic characterization of Eperigone Crosby and Bishop 1928 by Millidge (1987) and of

Mermessus by Miller (2007), who synonymized Eperigone under Mermessus.

Paquin and Dupérré (2003) illustrated both sexes of S. inflata and Dupéré [sic] et al. (2006) formally described the female of S. inflata, noting that the female paratype of

Savignia birostra (Chamberlin and Ivie 1947) strongly resembled that of S. inflata. I examined the paratype specimen of S. birostra and conclude that it was incorrectly identified and is a female specimen of S. inflata.

Materials and methods

Specimens were examined in 95% ethanol under a Leica MZ95 dissection microscope or a Wild Leitz M5A dissection microscope. For illustrations, specimens were examined in 70% ethanol under a SMZ-U Nikon dissection microscope. A Nikon

Coolpix 950 digital camera attached to the microscope was used to photograph all the structures to illustrate. The digital photos were used to trace proportions and the illustrations were detailed and shaded by referring to the structure under the microscope.

For the study of the embolic division, the male palps were placed for ~10 minutes in 179

warm KOH, washed in 70% alcohol, mounted on a slide in lactic acid and observed under an AmScope XSG Series T-500 compound microscope. Female genitalia were excised using a sharp entomological needle and transferred to lactic acid to clear non- chitinous tissues. A temporary lactic acid mount was used to examine the genitalia under the compound microscope. The structure was photographed and illustrated as explained above. All measurements were made with a micrometer ruler fitted to the eyepiece of the microscope. When possible, I measured at least 5 individuals of each sex. Holding locations for the examined materials are indicated in parenthesis at the end of each record. Specimens examined were from the Canadian National Collection of Insects,

Arachnids, and Nematodes, Ottawa, ON, Canada (CNC), the personal collection of the first author (LBP), the American Museum of Natural History, New York, NY, USA

(AMNH), the Denver Museum of Nature and Science, Denver, CO, USA (DMNS), the

Field Museum of Natural History, Chicago, IL, USA (FMNH), or the United States

National Museum, Washington, DC (USNM). For each location, the latitude and longitude (indicated in brackets) are given in decimal degrees and should be considered approximate. Abbreviations used for the terminology of the male and female genitalia follow Hormiga (2000), Miller and Hormiga (2004), and Miller (2007): ARP, anterior radical process; AT, anterior tooth of radix; CD, copulatory duct; DP, dorsal plate of epigynum; DSA, distal suprategular apophysis; E, embolus; EM, embolic membrane; FD, fertilization duct; MT, median tooth of radix; P, paracymbium; PT, protegulum; PTA, palpal tibial apophysis; R, radix; R Out, radix outgrowth; S, spermatheca; SD, sperm duct; SPT, suprategulum; ST, subtegulum; T, tegulum; TmI, position of trichobothrium 180

on tibia I; TmIV, position of trichobothrium on metatarsus IV; TP, tailpiece of radix; VP, ventral plate of epigynum.

Taxonomy

Family Linyphiidae Blackwall 1859

Genus Scyletria Bishop and Crosby 1938

Scyletria Bishop and Crosby 1938:89 (part); Buckle et al. 2001:141 (part); Draney and

Buckle 2005:127 (part); Platnick 2007 (part).

Type species. Scyletria inflata Bishop and Crosby 1938, by original designation.

Included species. S. inflata Bishop and Crosby 1938. The only other species originally included was S. jona Bishop and Crosby 1938, which is here transferred to

Mermessus (see below).

Diagnosis. See the below species diagnosis of S. inflata.

Scyletria inflata Bishop and Crosby 1938

(Fig. 17, 19)

Scyletria inflata Bishop and Crosby 1938:89, Pl. 7, Figs. 72–74 (male); Bélanger and

Hutchinson 1992:38; Aitchison-Benell and Dondale 1992:224; Paquin et al. 2001:19; 181

Paquin and Dupérré 2003:118, Figs. 1233–1235 (male, female); Draney and Buckle

2005:155, Figs. 35.276, 35.308 (male); Dupéré [sic] et al. 2006:152, Figs. 20, 21

(female).

Cephalethus birostrum Chamberlin and Ivie 1947:30, Fig. 21. Female paratype only, misidentified.

Savignia birostra: Buckle et al. 2001:139 (part); Platnick 2007 (part).

Type specimens. Holotype m#, EXAMINED. UNITED STATES OF

AMERICA: New York: Raquette Lake, [43.9°N, 74.6°W], June 11, 1927, coll. C.R.

Crosby (AMNH). Specimen in poor condition, with all legs from femur to tarsus missing, left palpus missing, and cephalothorax separated from abdomen.

Paratype f#, Cephalethus birostrum. EXAMINED. Alaska: Matanuska [61°N,

149°W], 23 May, 1945, colls. J.C. Chamberlin and Alan Linn (AMNH).

Other material examined. UNITED STATES OF AMERICA: North

Carolina: Summit of Mt. Mitchell [35.77°N, 82.26°W], 12 October 1923, 3 m#m#

(AMNH). CANADA: Alberta: Elkwater Lake, Cypress Hills Provincial Park

[49.65°N, 110.30°W], 7 August 1978, Sedges, coll. E.E. Lindquist, 1 f# (CNC); Elkwater

Lake, Cypress Hills Provincial Park [49.65°N, 110.30°W], 5–15 August 1978, coll. E.E.

Lindquist, 1 f# (CNC). Manitoba: Cowan Creek, Duck Mountain Provincial Park

[52.02°N, 100.65°W], 2 June 1980, emergence trap, coll. Flannagan, 1 f# (CNC); Riding

Mountain National Park: Swanson Spring [50.68°N, 99.82°W], 20 June 1979, coll. D.B.

Lyons, 1 m# (CNC); Riding Mountain National Park: Jackfish Creek [50.75°N, 182

100.82°W], 15 August 1979, sedges, colls. J and M Redner, 1 f# (CNC); Riding

Mountain National Park: Bison enclosure [55.47°N, 98.45°W], 30 May–19 June 1979, coll. D.B. Lyons, 1 m# (CNC); Riding Mountain National Park: North Shore Drive

[50.65°N, 99.97°W], 2 August 1979, coll. S.J. Miller, 2 m#m# (CNC); Riding Mountain

National Park: Wasagaming [50.65°N, 102.62°W], 29 August 1979, moss, colls. J. and

M. Redner, 1 m# 1 f# (CNC). New Brunswick: 25 km SW of Bathurst [47.62°N,

65.68°W], 19–30 June 1984, Balsam Fir foliage, coll. B.L. Cadogan, 1 f# (CNC); Green

River, 30 miles N of Edmunston [47.32°N, 68.15°W], 2–9 July 1963, litter, coll. T.R.

Renault, 3 m#m# (CNC); Priceville, 12 miles N of Boiestown [46.52°N, 66.28°W], 22

August 1968, under stone, coll. T.R. Renault, 1 f# (CNC); Green River, 30 miles N of

Edmunston [47.32°N, 68.15°W], 26 June 1968, coll. T.R. Renault, 1 f# (CNC);

Fredericton [45.97°N, 65.45°W], 9–12 July 1970, Balsam Fir foliage, coll. T.R. Renault,

1 m# (CNC). Newfoundland: The Arches [50.12°N, 57.65°W], 7 August 1984, in grass and moss, coll. G. Costello, 1 f# (CNC). Northwest Territories: Fort Simpson

[61.82°N, 121.35°W], 15 June 1972, litter, coll. A. Smetana, 1 m# 1 f# (CNC). Nova

Scotia: Cape Breton Highlands National Park: Lone Shieling [46.82°N, 60.80°W], pan trap, 18 June–11 July 1983, coll. R. Vockeroth, 1 m# (CNC); Cape Breton Highlands

National Park: sandy beach [46.82°N, 60.80°W], 22 June 1983, coll. Y. Bousquet, 1 f#

(CNC); Cape Breton Highlands National Park: Pleasant Bay [46.82°N, 60.80°W], 6–17

June 1984, coll. L. Masner, 1 m# (CNC); Coldbrook, Kings County [45.07°N, 64.58°W],

8 July 1958, apple foliage, coll. C.D. Dondale, 1 f# (CNC). Ontario: Spruce River, 42 mile N of Hurkett [49.28°N, 88.85°W], 17 August 1972, moss, coll. E.E. Lindquist, 1 f# 183

(CNC). Québec: St. Méthode, near Lac St. Jean [48.73°N, 72.42°W], 13 July 1982, litter, colls. C.D. Dondale and J. Redner, 1 f# (CNC); St. Hippolyte [45.93°N, 79.02°W],

10 June 1975, coll. R. Rochon, 1 f# (CNC); St. Hippolyte [45.93°N, 79.02°W], 25 June

1975, coll. R. Rochon, 1 m# (CNC); Îles de la Madeleine: Grosse Île [47.5°N, 61.67°W],

1 July 1985, sedges, coll. L. LeSage, 1 f# (CNC).

Diagnosis. It is difficult to diagnose Scyletria properly until its nearest relatives are known. The data available in older original descriptions are often insufficient and do not provide enough details with regard to the embolic division configuration and sclerites. Furthermore, as Miller and Hormiga (2004:425) remarked, "Identification of sclerites in the linyphiid embolic division appears to be more difficult than first thought.”

As such, the diagnosis and the description of the genus Scyletria are represented by the type species characteristics provided below.

Scyletria inflata is distinguished from other erigonine species by the following characters: male palpal tibia with two apophyses separated by a deep fissure (Fig. 17D); embolic division with long tailpiece bearing a sharp, transparent spine basally (Fig. 17C), radix folded, bearing a large outgrowth (R Out) with fringed edges (Figs. 17A, 17C).

Epigynum characterized by a pair of blunt prominences separated by a deep indentation

(Dupéré [sic] et al. 2006: Figs. 20–21).

Description. Male: Male from Cape Breton Highlands National Park, Nova

Scotia, Canada: total length 1.91 mm; carapace 0.91 mm long, 0.62 mm wide. Carapace dark orange, evenly convex at sides, steeply ascending from posterior margin to dorsal groove, then convex to posterior eye row, lacking lobes and pits; clypeus somewhat 184

Figure 17. Scyletria inflata Bishop and Crosby 1938. A) male palpus, ventral view, B) male palpus, retrolateral view, C) expanded male palpus, schematic illustration, D) male palpal tibia, dorsal view. Scale bars = 0.1 mm.

185

A B

C D 186

protruding. Posterior eye row procurved; anterior eye row recurved; eyes subequal in diameter; anterior lateral eyes and posterior lateral eyes touching. Chelicerae yellow, with about 18 stridulatory ridges; promargin with 5 small teeth, and retromargin with 3 denticles. Sternum mid-brown, shiny. Abdomen dark grey. Legs yellowish; tibial dorsal macrosetae 2221; TmI circa 0.40; TmIV absent; coxa IV with stridulatory pick. Palpal tibia black, with 2 dorsal apophyses of about equal length (Fig. 17D); mesal of these apophyses bifid at tip; apophyses separated by deep fissure (Fig. 17D). Paracymbium broad, flat, with distal hook, proximal end bearing two setae (Fig. 17B); embolic division with radix and tailpiece without break (Fig. 17C); radix long and folded, with large outgrowth (R out) with fringed edges (Fig. 17C); tailpiece long, sac-shaped, with shallow excavation on mesal margin, bearing a long, sharp, transparent spine basally (Fig. 17C); embolus short and flat; anterior radical process and embolic membrane present (Figs.

17A, 18).

Female: Female from Ste. Méthode, QC, Canada: total length 1.83 mm; carapace

0.87 mm long, 0.46 mm wide. Carapace, sternum, abdomen and legs essentially as in male. Cheliceral retromargin with 10 minute denticles. A full description with illustrations of the female is given in Dupéré [sic] et al. (2006: Figs. 20–21).

Variation. Males: Six males gave the following (mean ± 1 SD): total length

1.64 ± 0.16 mm; carapace 0.81 ± 0.06 mm long, 0.60 ± 0.03 mm wide. Carapace dull yellow to dark yellow, dull reddish in some specimens. Sternum often suffused with dark grey. Abdomen usually dark. Legs yellow to pale orange. 187

Females: Five females gave the following (mean ± 1 SD): total length 1.65 ±

0.16 mm; carapace 0.68 ± 0.05 mm long, 0.48 ± 0.05 mm wide (from Dupéré [sic] et al.

2006). See Dupéré [sic] et al. 2006 for additional details.

Natural history. Captured in a variety of habitats (summarized in Dupéré [sic] et al. 2006), the species is an epigeal spider from the boreal region of North America with extension southward in the Appalachian Mountains. Little else is known of its natural history.

Distribution. Fig. 19. USA: Alaska, North Carolina, New York. CANADA:

Alberta, Manitoba, New Brunswick, Newfoundland, Northwest Territories, Nova Scotia,

Ontario, Québec, Saskatchewan.

Mermessus jona (Bishop and Crosby 1938) new combination

(Figs. 18, 19)

Scyletria jona Bishop and Crosby 1938:90, pl. 7, figs. 75, 76 (male); Kaston 1976:25

(male, not female); Buckle et al. 2001:141; Draney and Buckle 2005:153, fig. 35.293

(male); Platnick 2007.

Type specimen. Holotype m#, TYPE SPECIMEN LOST. UNITED STATES

OF AMERICA: New York: Ithaca, [42.4°N, 76.5°W], 17 May 1924, coll. “in stomach of brook trout by H. J. Pack” (AMNH). Missing from type vial and presumed lost and/or destroyed. 188

Other material examined. UNITED STATES OF AMERICA: Connecticut:

Middlesex, CT [41.52°N, 72.71°W], 7 Feb. 1951, coll. PFB [P. F. Bellinger], det. Kaston

1951, 1 m# (USNM). Illinois: Cook Co., Swallow Cliff Woods, NW of Palos Park, south site [41.68°N, 87.86°W], 30 May–12 June 1996, carrion trap (squid), elev. 215 m, colls. “M. Thayer et al.,” det. M. Draney, 1 m# (FMNH); Cook Co., Swallow Cliff

Woods, NW of Palos Park, south site [41.68°N, 87.86°W], 12–26 June 1996, mini-FIT, elev. 215 m, colls. “M. Thayer et al.,” det. M. Draney, 1 m# (FMNH); Cook Co.,

Swallow Cliff Woods, NW of Palos Park, south site [41.68°N, 87.86°W], 3–17 April

1997, pitfall trap, elev. 215 m, colls. P. Parrillo and J. Pulizzi, det. M. Draney, 1 m#

(FMNH); Cook Co., Swallow Cliff Woods, NW of Palos Park, south site [41.68°N,

87.86°W], 17 April–1 May 1997, pitfall trap, elev. 215 m, coll. P. Parrillo, det. M.

Draney, 1 m# (FMNH); Cook Co., Swallow Cliff Woods, NW of Palos Park, north site

[41.68°N, 87.87°W], 16–30 May 1997, mini-FIT, elev. 215 m, colls. M. Thayer, A.

Varsek, and J. Louderman, det. M. Draney, 1 m# (FMNH); Cook Co., Swallow Cliff

Woods, NW of Palos Park, north site [41.68°N, 87.87°W], 17 April–1 May 1997, mini-

FIT, elev. 215 m, colls. M. Thayer, A. Varsek, and J. Louderman, det. M. Draney, 1 m#

(FMNH); Cook Co., Swallow Cliff Woods, NW of Palos Park, north site [41.68°N,

87.87°W], 17 April–1 May 1997, pitfall trap, elev. 215 m, coll. P. Parrillo, det. M.

Draney, 1 m# (FMNH); Cook Co., Swallow Cliff Woods, NW of Palos Park, north site

[41.68°N, 87.87°W], 1 May 1997, Berlese funnel, log and leaf litter, elev. 215 m, colls.

M. Thayer, A. Varsek, and J. Louderman, det. M. Draney, 1 m# (FMNH); Cook Co.,

Cherry Hill Woods (W of), NW of Palos Park, north site [41.68°N, 87.88°W], 27 June– 189

10 July 1997, pitfall trap, elev. 210 m, coll. P. Parrillo, det. M. Draney, 2 m#m#

(FMNH); Cook Co., Swallow Cliff Woods, NW of Palos Park, south site [41.68°N,

87.86°W], 26 May–9 June 1998, pitfall trap, elev. 215 m, colls. “M. Thayer et al.,” det.

M. Draney, 1 m# (FMNH); Cook Co., Swallow Cliff Woods, NW of Palos Park, south site [41.68°N, 87.86°W], 31 March–14 April 1999, pitfall trap, elev. 215 m, colls. J.

Louderman and A. Antov, det. M. Draney, 2 m#m# (FMNH); Cook Co., Swallow Cliff

Woods, NW of Palos Park, south site [41.68°N, 87.86°W], 31 March–14 April 1999, pitfall trap, elev. 215 m, colls. J. Louderman and A. Antov, det. M. Draney, 24 m#m#

(FMNH); Cook Co., Cherry Hill Woods (W of), NW of Palos Park, north site [41.68°N,

87.88°W], 14 April 1999, Berlese funnel, log and leaf litter, elev. 210 m, colls. M.

Thayer and A. Varsek, det. M. Draney, 1 m# (FMNH). New York: Jamesville,

Onondaga Co., 3 miles N of Route 173 [42.98°N, 76.07°W], 9 May 1981, old field, coll.

D. Sillman, 4 m#m# 3 f#f# (CNC). Ohio: Bath Nature Preserve, north of Akron

[41.18°N, 81.65°W], 14–28 May 2003, plot B1 trap 3, from pitfall in grassland, coll. L.B.

Patrick, det. C.D. Dondale and L.B. Patrick, 1 m# 1 f# (CNC); Bath Nature Preserve,

Bath Township, Summit Co. [41.18°N, 81.65°W], 14–28 May 2003, plot B2 trap 4, from pitfall in grassland, coll./det. L.B. Patrick, 1 m# (LBP); Bath Nature Preserve, Bath

Township, Summit Co. [41.18°N, 81.65°W], 30 June–14 July 2005, plot B1 trap 2, pitfall trap, upland old-field grassland, coll./det. L.B. Patrick, 2 m#m# (LBP); Bath Nature

Preserve, Bath Township, Summit Co. [41.18°N, 81.65°W], 14–28 May 2003, plot B4 trap 1, pitfall trap, upland old-field grassland, coll./det. L.B. Patrick, 1 f# (LBP); Bath

Nature Preserve, Bath Township, Summit Co. [41.18°N, 81.65°W], 13–27 July 2004, 190

pitfall trap, upland old-field grassland, coll./det. L.B. Patrick, 3 m#m# (DMNS); Bath

Nature Preserve, Bath Township, Summit Co. [41.18°N, 81.65°W], 14–28 May 2003, pitfall trap, upland old-field grassland, coll./det. L.B. Patrick, 3 m#m# (AMNH).

CANADA: Ontario: 18 km E of Gananoque [44.33°N, 76.17°W], 12 May–9 June

1977, pitfall trap in old field, colls. C.D. Dondale and J. Redner, 5 m#m# (CNC);

Wainfleet Marsh, 8 km S of Welland [42.92°N, 79.3°W], 14–20 June 1988, coll. A.

Stirling, 1 m# (CNC).

Etymology. The type specimen was found in the stomach of a brook trout. The specific epithet is a biblical reference to the tale of Jonah and the whale, making it a proper name in apposition. Thus, the apparent feminine gender of the specific epithet does not change with the masculine genus Mermessus.

Diagnosis. This species is properly transferred to the genus Mermessus based on the generic diagnosis by Miller (2007), wherein males have a prolateral excavation of the radix and a free ventrally recurved tailpiece (Figs. 18A, 18D), an embolic division that is as tall as it is long, a median tooth of the radix (Fig. 18D), the absence of anterior radical process, lack of a series of teeth around the margin of the carapace common to the closely related genus Erigone, and no palpal patellar tooth. Females do not have the terminally divided ventral plate of the epigynum (Fig. 18F) typical of most species of Mermessus

(Millidge, 1987; Miller, 2007), but do share an undivided ventral plate with other, atypical epigyna within the genus (e.g., M. entomologicus (Emerton 1911), M. index

(Emerton 1914), and M. indicabilis (Crosby and Bishop 1928); see Miller, 2007 and

Millidge, 1987). 191

Figures 18. Mermessus jona (Bishop and Crosby 1938) new combination. A) male palpus, ventral view, B) male palpus, retrolateral view, C) male palpal tibia, dorsal view,

D) expanded male palpus, schematic illustration, E) spermathecae, ventral view, F) epigynum, ventral view, G) spermathecae, dorsal view. Scale bars = 0.1 mm.

192

B A

G 193

Figure 19. Geographic distribution of Scyletria inflata Bishop and Crosby 1938 (filled circles) and Mermessus jona (Bishop and Crosby 1938) new combination (crosses).

194

Distinguishing this species from other species of Mermessus are the deep groove of the tibial apophysis that potentially gives the appearance of two apophyses (Fig. 18C), the large and quadrate paracymbium (Figs. 18A, 18B), the long copulatory ducts creating a “U” shape and terminating at the spermathecae in a short lateral turn towards the midline of the epigynum (Fig. 18F), and the extremely small size.

Description. Male: from the Bath Nature Preserve, Summit Co., Ohio, USA: total length 1.08 mm; carapace 0.48 mm long, 0.33 mm wide; carapace smooth, shiny, lacking pits and lobes, dull yellow with diffuse pattern of very light orange radiating from midline; anterior portion around eye region and clypeus slightly darker than remaining carapace; 3 erect setae along midline; sternum strongly concave and light yellow with sparse setae; endites with sparse setae on ventral surface, and blunt and lightly sclerotized along anterior margin. In lateral view, carapace level in anterior two-thirds, then gently sloping to pedicel in posterior third; carapace margin whitish along posterior half.

Anterior eye row procurved, with eyes closely situated together; anterior median eyes small, just less than half the diameter of anterior lateral eyes. Anterior lateral eyes and posterior lateral eyes with reflective tapetum, while anterior median eyes and posterior median eyes apparently lack a tapetum. Posterior eye row slightly procurved; posterior median eyes large, oval, and separated by at least the diameter of anterior median eyes; anterior lateral eyes and posterior lateral eyes nearly touching. Clypeus with small, erect seta just below anterior median eyes. Chelicerae light orange with sparse setae, each with a single lightly sclerotized spur on antero-prolateral face in distal fourth of the chelicera, and with 5–8 stridulatory ridges; cheliceral promargin and retromargin each with five 195

denticles. Abdomen unicolor, nearly white, with short sparse setae; colulus one quarter the length of anterior spinnerets; posterior lateral spinnerets and anterior lateral spinnerets conical, with posterior lateral spinnerets subequal in length to anterior lateral spinnerets.

Legs dull yellow, slightly darkening distally; coxa IV with stridulatory pick situated distally on retrolateral side; TmI 0.45; TmIV absent; dorsal tibial macrosetae 2221.

Femur and patella of palpus normal and light yellow with stridulatory pick at base of femur; tibia darker than patella; tibia (Fig. 18C), cup-shaped, with dorsal portion rising to a heavily sclerotized apophysis that terminates in two blunt teeth (prolateral view) subequal in length. In mesal view, tibia with a deep groove which runs from the separation between the two blunt teeth of the tibial apophysis to the base of the cymbium.

Paracymbium large, quadrate, smooth, with distinct spur on dorsal margin, with blunt hook on ventral margin, and with proximal end bearing five setae (Figs. 18A, 18B).

Embolus lightly sclerotized, with its blackened tip protected by a translucent embolic membrane (Fig. 18D). Embolus situated distally on radix (R) (Fig. 18D; the “embolic division” in Bishop and Crosby, 1938), radix with mesal tooth (“ventral projection” of

Millidge, 1987), and with large hooked tailpiece (“median projection” of Millidge, 1987)

(Fig. 18A, 18D). Lateral to the hooked tailpiece is the heavily sclerotized, spoon-shaped distal suprategular apophysis (Figs. 18A, 18B), which may easily be confused for a portion of radix.

Female: from 18 km E of Gananoque, Ontario, Canada: total length 1.08 mm; carapace 0.42 mm long, 0.35 mm wide. Description is as for the male with the following deviations: chelicerae lacking spurs; cheliceral retromargin with 7 denticles. 196

Epigynum (Fig. 18F) weakly sclerotized, with copulatory openings situated side- by-side at base of epigynum and barely covered by the undivided ventral plate of epigynum; copulatory ducts well sclerotized and laterally arching away from copulatory openings towards the spermathecae (Figs. 18E, 18F), terminating at the spermathecae in a short, lateral turn towards midline of the epigynum (Fig. 18F). Spermathecae broadly separated and situated anteriad to anterior margin of ventral plate (Figs. 18E, 18F); fertilization ducts oriented mesally. Epigynal dorsal plate small and weakly developed

(Fig. 18G).

Variation. Males: Six males gave the following (mean ± 1 SD): total length 1.03

± 0.10 mm; carapace 0.52 ± 0.06 mm long, 0.39 ± 0.06 mm wide. Carapace dull yellow or whitish to light orange; sternum light yellow to whitish with sparse setae.

Females: Three females gave the following (mean ± 1 SD): total length 1.05 ±

0.09 mm; carapace 0.43 ± 0.05 mm long, 0.33 ± 0.05 mm wide.

Natural history. Little is known of the natural history of this species. Other than the lost holotype specimen and a few specimens from Illinois, specimens of this species have been caught largely in pitfall traps, generally in open habitats such as grasslands, though the Illinois specimens were caught in a degraded oak savannah. Thus, M. jona is likely an epigeal spider of grassland and oak savannah habitats. Nearly all documented specimens are males, leading us to conclude that males are relatively mobile.

Remarks. In Draney and Buckle (2005), this species replaces Scyletria at couplet

179. Mermessus jona keys to "Eperigone" index in Millidge's (1987) key. Kaston (1976) 197

reported also capturing female specimens of M. jona, which I examined and determined not to be female M. jona specimens.

Distribution. Fig. 19. Until recently, the species was only known from the type locality in New York, as well as Connecticut (Kaston 1976) and possibly Maryland

(Muma 1944), though Muma's material could not be found for verification. Specimens have since been collected in Ohio and Illinois in the USA, as well as in Ontario in

Canada.

198

References

Aitchison-Benell, C. W., and C. D. Dondale. 1992. A checklist of Manitoba spiders

(Araneae) with notes on geographic relationships. Le Naturaliste Canadien

117(4):215–237.

Bélanger, G., and R. Hutchinson. 1992. Liste annotée des Araignées (Araneae) du

Québec. Pirata 1(1):2–119.

Bishop, S. C., and C. R. Crosby. 1938. Studies in American spiders: miscellaneous

genera of Erigoneæ, Part II. Journal of the New York Entomological Society

46:55–107.

Blackwall, J. (1859) Descriptions of newly-discovered spiders captured by James Yate

Johnson, Esq., in the island of Madeira. Annals and Magazine of Natural History

3(4):255-267.

Buckle, D. J., D. Carrol, R. L. Crawford, and V. D. Roth. 2001. Linyphiidae and

Pimoidae of America north of Mexico: Checklist, synonymy, and literature. In:

Paquin, P., and D. J. Buckle, eds., Contributions à la connaissance des Araignées

(Araneae) d’Amérique du Nord. Fabreries, Supplément 10, pp. 89–191.

Chamberlin, R. V., and W. Ivie. 1947. The spiders of Alaska. Bulletin of the University

of Utah, 37(10), Biological Series 10:5–103.

Crosby, C. R. and S. C. Bishop. 1928. Revision of the spider genera Erigone,

Eperigone, and Catabrithorax (Erigoneae). New York State Museum Bulletin

278:1–73. 199

Denis, J. 1949. Notes sur les Érigonides. XVI. Essai sur la détermination des femelles

d’érigonides. Bulletin de la Société d’Histoire naturelle de Toulouse 83:129–158.

Draney, M. L., and D. J. Buckle. 2005. Linyphiidae. In: Ubick, D. P., P. Paquin, P. E.

Cushing, and V. Roth, eds., Spiders of North America: an identification manual.

American Arachnological Society, pp. 124–161.

Dupéré [sic], N., P. Paquin, and D. J. Buckle. 2006. Have you seen my mate?

Descriptions of unknown sexes of some North American species of Linyphiidae

and Theridiidae. Journal of Arachnology 34:142–158.

Emerton, J. H. 1882. New England spiders of the family Theridiidae. Transactions of

the Connecticut Academy of Arts and Sciences 6:1–86.

Emerton, J. H. 1911. New Spiders from New England. Transactions of the Connecticut

Academy of Arts and Sciences 16:383–407.

Emerton, J. H. 1914. New spiders from the neighborhood of Ithaca. Journal of the New

York Entomological Society 22:262–264.

Hormiga, G. 2000. Higher level phylogenetics of erigonine spiders (Araneae,

Linyphiidae, Erigoninae). Smithsonian Contributions to Zoology 609:1–160.

Kaston, B. J. 1976. Supplement to the spiders of Connecticut. Journal of Arachnology

4:1–72.

Miller, J. A. 2007. Review of erigonine spider genera in the Neotropics (Araneae:

Linyphiidae, Erigoninae). Zoological Journal of the Linnean Society,

149(Supplement 1):1–272. 200

Miller, J. A., and G. Hormiga. 2004. Clade stability and the addition of data: A case

study from erigonine spiders (Araneae: Linyphiidae, Erigoninae). Cladistics

20:385–442.

Millidge, A. F. 1977. The conformation of the male palpal organs of linyphiid spiders,

and its application to the taxonomic and phylogenetic analysis of the family

(Araneae: Linyphiidae). Bulletin of the British Arachnological Society 4:1–60.

Millidge, A. F. 1987. The erigonine spiders of North America. Part 8. The genus

Eperigone Crosby and Bishop (Araneae, Linyphiidae). American Museum

Novitates 2885:1–75.

Muma, M. H. 1944. A report on Maryland spiders. American Museum Novitates

1257:1–14.

Paquin, P., N. Dupérré, and R. Hutchinson. 2001. Liste révisée des Araignées (Araneae)

du Québec. In: Paquin, P., and D. J. Buckle, eds., Contributions à la

connaissance des Araignées (Araneae) d’Amérique du Nord. Fabreries,

Supplément 10, pp. 5–87.

Paquin, P., and N. Dupérré. 2003. Guide d’identification des Araignées (Araneae) du

Québec. Fabreries, Supplément 11.

Pickard-Cambridge, O. 1899. Arachnida. Araneidea. Biologia Centrali-Americana,

Zoology 1:289–304.

Platnick, N. I. 2007. The world spider catalog, version 8.0. American Museum of

Natural History, online at

http://research.amnh.org/entomology/spiders/catalog/index.html. 201

Simon, E. 1884. Les Arachnides de France. Tome cinqième, deuxième partie, contenant

la Famille des Theridionidae (suite). Librairie Encyclopédique de Roret, Paris,

France, pp. 180–885.

CHAPTER 6

FERTILIZATION EFFECTS ON PREDATOR DIVERSITY IN A TERRESTRIAL

DETRITAL FOOD WEB

Abstract

Biodiversity can affect a variety of factors (e.g., ecosystem function, resistance and resilience to exotic species invasion, nutrient cycling) be affected by other factors

(e.g., disturbance, climate, eutrophication). Increased nitrogen (N) inputs into terrestrial ecosystems have been shown to have a eutrophication effect in which plant and insect biomass and abundance increased while diversity in these taxa decreased, and this body of work has been formalized into biodiversity-productivity theory. However, biodiversity-productivity theory has not been tested in the largely detritus-based epigeal arthropod community, a component of the ecosystem that can significantly affect nutrient cycling and ecosystem function. In a four-year study in which nutrients (via fertilization) and plant litter (via removal) were manipulated, I tested the bottom-up predictions of biodiversity-productivity theory by measuring the abundance, biomass, and species richness of epigeal predators (ground and rove beetles, spiders, and centipedes) and their primarily detritivorous prey (collembolans, flies, and pillbugs). I found that the epigeal

202 203

arthropod community did not respond to fertilization as predicted by biodiversity- productivity theory, but instead as would be predicted by top-down trophic cascading effects. Opposite the predictions of biodiversity-productivity theory, predator abundance, biomass, and diversity increased, while the detritivorous prey abundance and biomass did not significantly increase in any year, though the diversity of detritivorous prey did increase. These results imply that the epigeal arthropod community is more influenced by the quantity and quality of the basal resource (plant litter) than by the diversity of that resource. Thus, biodiversity-productivity theory should be revised to include the fundamentally different responses of the epigeal community.

Introduction

Biodiversity can be affected by a variety of factors, including disturbance (Grime

1979), climate (Nogué et al. 2009), soil fertility (Fraser and Grime 1997, 1998; Fraser

1998; Dybzinski et al. 2008), exotic species invasions (Knops et al. 1999; Kimbro et al.

2009), habitat patch size (Cook et al. 2005; Martinko et al. 2006), anthropogenically mediated nutrient addition (Knops et al. 1999; Haddad et al. 2000, 2001; Tilman et al.

2002a, b; Vitousek et al. 2002; Suding et al. 2005; Patrick et al. 2008a, b), and trophic dynamics (Hutchinson 1959; Hairston et al. 1960; Fretwell 1977; Oksanen et al. 1981;

Fraser and Grime, 1997, 1998; Fraser 1998; Oksanen and Oksanen 2000; Haddad et al.

2000, 2001; Patrick et al. 2008b). In turn, biodiversity can affect ecosystem function

(Naeem et al. 1994; Hooper and Vitousek 1997; Hodgson et al. 1998), community and

204

ecosystem resistance and resilience to invasive species (Knops et al. 1999), ecosystem productivity (Tilman et al. 1996, 1997), nutrient regeneration and recycling (Moore et al.

2004; Hättenschwiler and Gasser 2005), and ecosystem services (Cummings and Child

2009). In terrestrial systems, arthropod diversity is positively correlated with plant diversity (Knops et al. 1999; Haddad et al. 2000, 2001; Schaffers et al. 2008). However, the epigeal portion of the arthropod community is more closely tied to the detritus-based portion of the food web (Halaj and Wise 2002), and may differ in its response to plant species richness relative to aerial (stem- and leaf-associated) arthropods (Patrick et al.

2008b; Schuldt et al. 2008) in terrestrial habitats.

As plant species richness decreases, insect abundance generally increases and insect species richness decreases (Southwood et al. 1979; Strong et al. 1984; Siemann et al. 1998; Knops et al. 1999; Haddad et al. 2000, 2001; Schaeffers et al. 2008) and fewer predators may be supported by the decreased prey diversity (Knops et al. 1999; Haddad et al. 2001). Underlying these patterns is the idea that as plant species richness increases, the diversity of specialist herbivores increases as a function of increased resource diversity (Southwood et al. 1979; Strong et al. 1984; Siemann et al. 1998; Knops et al.

1999; Haddad 2001), in turn supporting a higher diversity of predators and parasitoids

(Hutchinson 1959; Knops et al. 1999; Haddad et al. 2001).

Further, plant species richness and biomass are strongly influenced by nitrogen

(N) availability (Tilman et al 2002a, b, Suding et al 2005; Patrick et al. 2008a). While N is the primary limiting nutrient in terrestrial systems (Vitousek et al. 2002), the global N pool is increasing at an alarming rate as a result of human activity (Vitousek et al. 1997;

205

Fenn et al. 2003; Galloway et al. 2003). This increase in N has been shown to decrease plant species richness while increasing biomass (Bobbink et al. 1998; Hector et al. 1999;

Tilman et al. 2002a, b), a relationship that has become the basis of the biodiversity- productivity theory (Suding et al. 2005). Predictions of this theory have been extended to the animal component of the food web with field experiments that have shown decreased diversity of the aerial portion of the arthropod community in plots with augmented N

(Knops et al. 1999; Haddad et al. 2000, 2001; Schaffers et al. 2008).

Another factor that can affect arthropod diversity is the quantity and quality of plant litter. Nutrient loading not only increases plant standing crop biomass, but also plant litter production (Long et al. 2003; Patrick et al. 2008a), which can increase the basal food resource for the detrital community, resulting in increased detritivore and epigeal predator abundances (Halaj et al. 2000; Halaj and Wise 2002; Moore et al. 2004).

Furthermore, plant litter increases habitat complexity, which can also increase arthropod abundance and diversity (Lawton 1983; Strong et al. 1984). However, the detritivore community response is further complicated because it can be sensitive to litter diversity

(Hättenschwiler and Gasser 2005). With nutrient loading decreasing plant diversity

(Tilman et al. 2002a, b; Long et al. 2003; Patrick et al. 2008a), litter diversity is also decreased. While increased plant litter production could increase detritivore and epigeal predator abundance and biomass (Halaj et al. 2000), a reduction in litter diversity could result in declines in diversity of detritivores and epigeal predators (Hättenschwiler and

Gasser 2005), effectively mirroring the aerial community response to nutrient loading

(e.g., Knops et al. 1999; Haddad et al. 2000, 2001). Despite such predictions, no studies

206

have incorporated plant litter effects with biodiversity-productivity theory, nor have any studies investigated these effects on either the epigeal arthropod community as a whole or on species-level responses within the epigeal arthropod community.

“Green” world vs. “brown” world and cascading effects

In the portion of the food web more reliant on living, standing crop plant material—the ―green‖ portion of the food web (sensu Patrick et al. 2008b), decreased plant species richness may result in increased abundance of insect pest species (Elton

1958; Root 1973; Strong et al. 1984; Haddad et al. 2001) and lower insect species richness (Southwood et al. 1979; Strong et al. 1984; Siemann et al. 1998; Knops et al.

1999; Haddad et al. 2000, 2001) because specialist pest herbivores can more easily find their host plants and the host plants occur with greater frequency and generalist predator abundances are likely lower (Root 1973; Long et al. 2003). If I extend the ideas of the previous sentence to the litter-based epigeal arthropod community—the ―brown‖ portion of the food web (sensu Patrick et al. 2008b), then reduced diversity of plant litter should decrease detritivore diversity, but could increase detritivore abundance. Because epigeal predators are more closely tied to the detritivore community (Halaj and Wise 2002), decreased detritivore diversity could decrease the diversity of epigeal predators. Thus, applying the green world hypotheses to the brown world predicts decreased detritivore and epigeal predator diversity, but increased detritivore abundance.

207

However, detritivores may also be affected by litter quality, thereby increasing detritivore biomass and diversity that facilitates a bottom-up effect on the food web

(Ilieva-Makulec et al. 2006; Greenwood et al. 2007). Thus, increased litter quality and quantity could increase detritivore abundance, biomass, and diversity, thereby supporting more predators, and potentially a greater diversity of predators. If bottom-up effects are the primary mechanism, then increased litter quality and quantity from anthropogenically mediated fertilization should additively increase detritivore and epigeal predator abundance, biomass, and diversity. However, if the green world principles famously put forth by Hairston et al. (1960) are the primary mechanism, then increased plant litter quality and quantity should not necessarily lead to an increase in detritivores, but should lead to higher predator abundances in a classic top-down cascade. In this scenario, predator abundance, biomass, and diversity increases while detritivore abundance and biomass does not (but detritivore diversity could increase).

Here I report the results of a four year study that investigated the effects of experimentally manipulating NPK fertilization and plant litter on the epigeal arthropod community. I determined the diversity, abundance, and biomass responses of the entire epigeal arthropod community, the dominant epigeal predators (spiders, ground and rove beetles, and centipedes), and the dominant epigeal and detritivorous prey (pillbugs, collembolans, and flies) in a temperate-latitude grassland. I am testing two main hypotheses. The first hypothesis is that biodiversity-productivity theory adequately explains the relationships between epigeal arthropod diversity and fertilization.

Effectively, this would be driven by bottom-up mechanisms in which nutrients mediate

208

changes in the abundance and biomass of trophic levels from the basal resource upwards.

The epigeal arthropod community is closely tied to detritus (Halaj and Wise 2002), which may not respond in the same manner as the green world portions of the food web (Moore et al. 2004; Patrick 2008b) upon which biodiversity-productivity theory is based. For this hypothesis to be supported, arthropod abundance and biomass would increase in fertilized plots and arthropod diversity would decrease (Knops et al. 1999; Haddad et al. 2000,

2001).

The second hypothesis is that fertilization and plant litter (the basal resource in the detritivore portion of the food web) will enhance a top-down control in which predator abundance, biomass, and diversity increase while detritivorous prey biomass and abundance will not significantly change. The increased predator load will exert a top- down control in which detritivorous prey abundance and biomass will not significantly change as a result of either plant litter or fertilization. Rosenzweig (1971) proposed that nutrient enrichment could release prey from density-dependence (i.e., abundance can increase above pre-nutrient levels), but this would also release predators from density- dependence. This could create an unstable equilibrium in which both trophic levels ultimately go locally extinct (Rosenzweig 1971, 1973; Luckinbill 1973), but would more likely result in a higher predator abundance and lower or unchanged prey abundance

(Roy and Chattopadhyay 2007), providing evidence for the top-down trophic cascade of

Hairston et al. (1960). For the top-down hypothesis to be supported, there would be no significant changes in detritivore abundance and biomass in fertilized plots, but there would be significant increases in predator abundance, biomass, and diversity.

209

Materials and methods

Study site and experimental design

The study was done at the 163.5 ha Bath Nature Preserve (BNP; 41° 10’ 36.2‖ N,

During August 2001, twenty-four 20-m diameter circular plots (314 m2) were established. These experimental plots were separated by at least 20 m and were at least

30 m away from any other habitat. Treatments were applied in a 2 x 2 factorial design of fertilizer (+F = fertilizer added, -F = no fertilizer) and plant litter (-L = litter removed, +L

= litter left in situ after yearly mowing) with the control plots characterized as no fertilization and plant litter left in situ (+L/-F), resulting in six replicates per treatment.

210

Hereafter, all references to ―litter‖ refer to the previous year’s mown vegetation and any vegetation senesced and found within the sampling quadrat after standing crop removal.

In April 2002 and continuing each April through 2005, Scotts brand Osmocote 8-9 month

Slow Release Fertilizer 19-6-12 (NPK; Scotts, Marysville, OH, USA) was applied at 20 g

2001-2004), litter was removed from litter-removal treatments using a small 23 hp lawn tractor with a pull-behind 8 hp Agri-Fab Mow-N-Vac trailer attachment (Agri-Fab,

Sullivan, IL, USA).

Arthropod community sampling

Arthropods were collected using four pitfall traps in each of the 24 experimental plots (n = 96 total pitfall traps). Within each plot, a single trap was placed 5 m from the center of the plot at each of four magnetic compass directions (northeast, northwest, southeast, and southwest). Each trap consisted of a 10 cm diameter, 18 cm tall PVC

211

sleeve into which a 710-mL plastic cup was inserted and filled to approximately 4 cm with a 50/50 water/propylene glycol mixture. To deter trap raiders (e.g., microtine mammals), prevent captured spiders from climbing out of the trap, and prevent precipitation from directly flooding the trap, an 8-cm powder funnel with a base enlarged to ~3 cm was inserted and a 15 cm x 15 cm board was placed over each trap, leaving ~3 cm clearance. Starting in mid to late May (mid July during 2004) and continuing through mid to late August, traps were alternately left open for two weeks and closed for two weeks. This resulted in three sampling periods each year during 2002, 2003, and 2005.

During 2004, only the second and third sampling periods were collected. When closed, the plastic cups were removed, the contents collected and preserved in 70% EtOH, and the PVC sleeve was tightly capped. While pitfall traps do not capture all arthropods in the community, they are an effective sampling technique for determining the relative abundance and species richness of epigeal arthropods (Greenslade 1964; Phillips and

Cobb 2005). Arthropods were identified to species when possible, otherwise to the lowest possible taxonomic level, and the numbers within in each trap were recorded of each taxonomic identification unit (TIU), i.e., the lowest level of identification for a given group of arthropods. Each TIU within a trap was dried at 70°C for 72 hrs, following which the biomass of each TIU was determined to the nearest 0.0001 g was determined. Lacking sufficient numbers captured within a trap, some extremely small

TIUs did not register a biomass, and their biomass was recorded as ―0.000001 g‖ unless caught in sufficient numbers to register a biomass.

212

Statistical analyses

I tested the responses of the abundance, biomass, and TIU richness of all arthropods (also referred to as ―overall‖ abundance/biomass/species richness), of predators, and of prey to fertilization, plant litter, and the interaction of fertilization and plant litter. The ―predator‖ group included only the numerically dominant predators, primarily spiders, ground and rove beetles, and centipedes. The ―prey‖ group consisted of the numerically dominant detritivorous prey, primarily pillbugs (Isopoda), collembolans (Collembola), and flies (Diptera), known to be common prey for these predators. To calculate the average TIU richness within a plot, I summed the total number of TIUs caught in each trap, then averaged this TIU for each of the four traps within a plot within a sampling period (including zeroes for traps where nothing was captured), then averaged this TIU across sampling periods in a year, yielding n = 24 samples within each year. The same method was used to calculate the average abundance and biomass within a plot within a year, also yielding n = 24 samples within a plot within a year.

To analyze trends per year and per treatment in abundance, biomass, and TIU richness, I used SAS software version 8.01 (SAS Institute, Inc. 1999) to calculate a repeated-measure maximum-likelihood analysis in PROC MIXED with Type III effects based upon the covariance structure of compound symmetry. The various models used the different response variables (biomass, TIU richness, abundance), and for the predictor variables used fertilized vs. unfertilized, litter removed vs. litter left in situ, year, and

213

their factorial interactions, with year as the repeated predictor. When year was detected as a significant effect for a response variable, I tested for treatment effects within a year and used SAS software version 8.01 (SAS Institute, Inc. 1999) to calculate a maximum- likelihood analysis in PROC MIXED with Type III effects based upon the covariance structure of compound symmetry, with fertilization, litter, and the factorial interaction of fertilization and litter as predictor variables.

Results

A total of 176783 arthropods were caught during 14784 trap nights. Focusing on the most abundant TIUs, 16326 were classified as epigeal predators, including 10659 mature spiders (Araneae), 5594 predatory beetles from the two most dominant families

(Coleoptera: Carabidae, Staphylinidae), and 73 centipedes (Chilopoda). Additionally,

104365 were classified as detritivorous prey, including 64563 pillbugs (Isopoda), 34230 collembolans (Collembola), and 5572 flies from the two most dominant families

(Diptera: Sphaeroceridae, Phoridae).

For the abundance and biomass of all captured arthropods, fertilization was the only significant predictor variable in the repeated-measures PROC MIXED analyses

(Table 12). Overall species richness was significantly affected by fertilization, year, and the interaction of fertilization and year (Table 12). While there was no significant effect of year for either abundance or biomass, fertilization did significantly affect abundance during 2004 and 2005 (Figs. 20A, B). This result for overall biomass is interesting

214

Table 12. Results of the repeated-measures PROC MIXED analyses for each response variable. Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter (L), year (Y), and (*) their fully factorial interactions.

Response Variable F L Y F*L F*Y L*Y F*L*Y

Predator F1, 80 = 95.75 F1, 80 = 1.81 F3, 80 = 102.69 F1, 80 = 0.03 F3, 80 = 22.88 F3, 80 = 3.73 F3, 80 = 0.75 abundance P < 0.0001 P = 0.1826 P < 0.0001 P = 0.8578 P < 0.0001 P = 0.0145 P = 0.5253

Predator F1, 80 = 17.90 F1, 80 = 19.85 F3, 80 = 89.93 F1, 80 = 12.81 F3, 80 = 5.07 F3, 80 = 12.29 F3, 80 = 4.13 biomass P < 0.0001 P < 0.0001 P < 0.0001 P = 0.0006 P = 0.0029 P < 0.0001 P = 0.0089

Predator SR F1, 80 = 23.52 F1, 80 = 4.82 F3, 80 = 136.02 F1, 80 = 0.25 F3, 80 = 13.27 F3, 80 = 3.75 F3, 80 = 0.87 P < 0.0001 P = 0.0310 P < 0.0001 P = 0.6208 P < 0.0001 P = 0.0141 P = 0.4615

Prey F1, 80 = 6.81 F1, 80 = 1.68 F3, 80 = 9.46 F1, 80 = 1.08 F3, 80 = 0.89 F3, 80 = 0.54 F3, 80 = 0.40 abundance P = 0.0108 P = 0.1985 P < 0.0001 P = 0.3022 P = 0.4492 P = 0.6574 P = 0.7565

Prey F1, 80 = 10.62 F1, 80 = 2.00 F3, 80 = 10.94 F1, 80 = 2.90 F3, 80 = 0.07 F3, 80 = 1.75 F3, 80 = 0.40 biomass P = 0.0016 P = 0.1610 P < 0.0001 P = 0.0923 P = 0.9762 P = 0.1632 P = 0.7569

Prey SR F1, 80 = 95.75 F1, 80 = 1.81 F3, 80 = 102.69 F1, 80 = 0.03 F3, 80 = 22.88 F3, 80 = 3.73 F3, 80 = 0.75 P < 0.0001 P = 0.1826 P < 0.0001 P = 0.8578 P < 0.0001 P = 0.0145 P = 0.5253

Overall F1, 80 = 14.23 F1, 80 = 0.65 F3, 80 = 2.54 F1, 80 = 1.10 F3, 80 = 1.72 F3, 80 = 0.43 F3, 80 = 0.41 abundance P = 0.0003 P = 0.4212 P = 0.0622 P = 0.2972 P = 0.1700 P = 0.7327 P = 0.7445

Overall F1, 80 = 7.51 F1, 80 = 0.07 F3, 80 = 0.46 F1, 80 = 1.53 F3, 80 = 0.43 F3, 80 = 0.34 F3, 80 = 0.18 biomass P = 0.0076 P = 0.7991 P = 0.7140 P = 0.2197 P = 0.7324 P = 0.7967 P = 0.9110

Overall SR F1, 80 = 32.45 F1, 80 = 2.19 F3, 80 = 40.55 F1, 80 = 0.14 F3, 80 = 8.63 F3, 80 = 1.54 F3, 80 = 0.46 P < 0.0001 P = 0.1432 P < 0.0001 P = 0.7071 P < 0.0001 P = 0.2113 P = 0.7142

215

Figure 20. Abundance, biomass, and species richness of all arthropods (A–C), predators

(D–F), and prey (G–I). The letters above a year denotes significance at α < 0.05 of the

PROC MIXED analysis (Table 13) for that year for ―a‖ = fertilization, ―b‖ = plant litter, and ―c‖ = the interaction between fertilization and plant litter, respectively. Open circles

(○) and ―+L/-F‖ represent the control treatment plots of unfertilized and litter left in situ, open triangles (∆) and ―-L/-F‖ represent unfertilized and litter removed plots, filled circles (●) and ―+L/+F‖ represent fertilized and litter left in situ plots, and filled triangles

(▲) and ―-L/+F‖ represent fertilized and litter removed plots.

A. B. C.

250 a 1.25 40 s

a s )

e a, b e

g 1.00

(

200 h 30

a a n

0.75 s

b i

150 c 20

l l

0.50 p

v v

100 r 10 O

0.25 e

v O 50 0.00 0 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year D. E. F.

50 a, b 0.30 a, b, c 16

s a, b

)

e e

g 0.25

n (

c 40

h n

s 12

d 0.20

n s

30 m

u e

i a

i c

a 0.15

b 8

r a

o s

20 o

a t

r a

a 0.10

a o

t e

e 4

r r

10 d

P 0.05

r P 0 0.00 0 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year G. H. I.

175 0.5 4.25 a

150 s 4.00

s e

) 0.4

e n

a g

c (

h 3.75

125

n c

s a

d 0.3 a 3.50

100 s

u m

c i

a 3.25

75 e

0.2

r 3.00 r

50 y

e P

0.1 r

25 P 2.75 0 0.0 2.50 2002 2003 2004 2005 2002 2003 2004 2005 2002 2003 2004 2005 Year Year Year

202 217

Table 13. Results of the PROC MIXED analyses for each response variable for 2005.

Given are the F values with degrees of freedom, and the resulting P values, where bolded values indicate significance at α < 0.05. The predictor variables are fertilization (F), litter

(L), and (F*L) their factorial interaction.

Response Variable F L F*L

Predator abundance F1, 20 = 75.69, P < 0.0001 F1, 20 = 7.28, P = 0.0138 F1, 20 = 0.97, P = 0.3371

Predator biomass F1, 20 = 10.31, P = 0.0044 F1, 20 = 20.04, P = 0.0002 F1, 20 = 8.31, P = 0.0092

Predator SR F1, 20 = 33.42, P < 0.0001 F1, 20 = 11.96, P = 0.0025 F1, 20 = 1.66, P = 0.2124

Prey abundance F1, 20 = 2.47, P = 0.1317 F1, 20 = 0.16, P = 0.6972 F1, 20 = 0.24, P = 0.6328

Prey biomass F1, 20 = 2.82, P = 0.0189 F1, 20 = 0.57, P = 0.4593 F1, 20 = 0.08, P = 0.7779

Prey SR F1, 20 = 50.56, P < 0.0001 F1, 20 = 0.00, P = 1.0000 F1, 20 = 0.01, P = 0.9334

Overall abundance F1, 20 = 8.85, P = 0.0075 F1, 20 = 0.62, P = 0.4418 F1, 20 = 0.32, P = 0.5797

Overall biomass F1, 20 = 0.70, P = 0.4132 F1, 20 = 0.59, P = 0.4497 F1, 20 = 0.90, P = 0.3536

Overall SR F1, 20 = 27.42, P < 0.0001 F1, 20 = 5.11, P = 0.0351 F1, 20 = 0.52, P = 0.4773

218

because fertilization significantly affected biomass in the repeated-measures PROC

MIXED analysis, but not significantly during any one year (Table 12), indicating that the effect of fertilization on biomass is significant over the duration of the experiment, but not necessarily in any given year. Overall species richness did not significantly differ during 2002 and 2003, then differed significantly during 2004 as a result of fertilization and during 2005 as a result of both fertilization and litter (Table 12, Fig. 20C).

Predator abundance was significantly affected by fertilization, year, and the interactions between fertilization and year and between litter and year in the repeated- measures PROC MIXED analysis (Table 12). Predator abundance was clearly significantly affected by fertilization, with predator abundance increasing each year, with litter also becoming significant during 2005 (Fig. 20D). Predator biomass was significant for all predictor variables in the repeated-measures PROC MIXED analysis (Table 12), with fertilization, litter, and their interaction significantly affecting predator biomass during 2005 (Fig. 20E). The unusually high biomass of predators in unfertilized plots with litter removed (Fig. 20E) was due to an unusually high number of captures of one of the largest predators, the wolf spider Schizocosa avida (Walckenaer), in a single plot for two consecutive trapping periods during 2005. In the repeated-measures PROC MIXED analysis, predator species richness was significantly affected by fertilization, litter, year, and the interactions between fertilization and year and between litter and year (Table 12).

However, predator species richness was not affected by any interactions between fertilization and litter in the repeated-measures PROC MIXED analysis (Table 12).

219

Moreover, predator species richness was significantly affected by fertilization and by litter, but not their interaction, during 2005 (Fig. 20F).

The abundance, biomass, and species richness of prey were significantly affected by fertilization and year, and the interactions between fertilization and year and between litter and year additionally affected prey species richness in the repeated-measures PROC

MIXED analysis (Table 12). While fertilization did significantly affect prey abundance during 2004 (Fig. 20G), no predictor variables significantly affected prey biomass during any individual year of the experiment (Fig. 20H). Prey species richness was strongly and significantly affected by fertilization during 2004 and 2005.

During 2005, the overall arthropod abundance was significantly correlated to total plant biomass (Fig. 21A), while overall arthropod species richness was significantly and positively correlated to total plant biomass and negatively to plant species richness (Figs.

21B, C). Predator abundance and species richness were both significantly and positively correlated with total plant biomass (Figs. 21D, E), but significantly and negatively correlated to plant species richness during 2005 (Fig. 21F). Prey abundance was not significantly correlated to total plant biomass, but prey species richness was significantly and positively correlated to total plant biomass during 2005 (Figs. 21G, H). Moreover, prey species richness was significantly and negatively correlated to plant species richness during 2005 (Fig. 21I).

220

Figure 21. Regressions of the abundance and species richness against total plant biomass and plant species richness for all arthropods (A–C), predators (D–F), and prey (G–I).

Symbols are defined in Figure 20, and data presented are for 2005.

A. B. C.

300 40 40

s s

e e

e 250

n n

h h

n 30 R² = 27.6% P = 0.008

c c

i i

a r

200 r 30

n s s

e e

i i

b 150 20

c c

e e

p p

a s

100 s 20

l l

e 10 R² = 34.0% P = 0.003

a a

v r

50 r

O e

R² = 26.3% P = 0.010 e

v v O 0 O 0 10 250 500 750 1000 1250 250 500 750 1000 1250 0 5 10 15 20 Total plant biomass (g / m²) Total plant biomass (g / m²) Plant species richness D. E. F.

60 16 16

s s s

s R² = 13.5% P = 0.078

e e

e n

50 n

c h

h 14

c c

i i

r r

d 40 12

s s

n 12

e e

i i

c c

a 30

e e

r p

p 10

s s

8 r

20 r

o o

t t

e a

a 8

r d 10 d

P R² = 33.2% P = 0.003 e

R² = 56.1% P < 0.001 e

r r P 0 P 4 6 250 500 750 1000 1250 250 500 750 1000 1250 0 5 10 15 20 Total plant biomass (g / m²) Total plant biomass (g / m²) Plant species richness G. H. I.

200 5 5

s s s

s R² = 53.7% P < 0.001

e e

e n

150 n

h h

n 4 4

c c

i i

r r

n s s

u e

100 e

i i

c c

e e

3 3

p p

e s s

r 50

y y

e e r

r R² = 63.2% P < 0.001 P R² = 15.0% P = 0.062 P 0 2 2 250 500 750 1000 1250 250 500 750 1000 1250 0 5 10 15 20 Total plant biomass (g / m²) Total plant biomass (g / m²) Plant species richness

221

Discussion

My results clearly support a top-down trophic cascade response, as would be predicted by Hairston et al. (1960), and current biodiversity-productivity theory (Suding et al. 2005) was not supported. As the quantity (Patrick et al. 2008a) and quality (Pan et al. in press) of the basal resource (plant litter) increased, so did the abundance, biomass, and diversity of predators, while the abundance and biomass of the detritivorous prey did not significantly change. Biodiversity-productivity theory would predict increased abundance and biomass of predators, but it would not predict increased predator diversity, nor would it predict increased diversity of the detritivorous prey. The epigeal arthropod community of my field site was likely subject to top-down cascading effects, whereas biodiversity-productivity theory did not apply to this portion of the community.

I contend that plant litter quality is more important than plant litter diversity for the epigeal food web. Despite lower plant litter diversity, the diversity of the main detritivores (collembolans, flies, and pillbugs) increased, as did the abundance, biomass, and diversity of the main predators (ground and rove beetles, spiders, and centipedes).

Fertilization induced increased plant litter production (Patrick et al. 2008a), and this plant litter has been shown to have significantly higher nutrient content than would be found in unfertilized plots (Pan et al. in press). This is consistent with some other studies (e.g.,

Cross et al. 2006; Srivastava et al. 2009), though these studies did not measure detritivore abundance (or biomass or diversity), particularly in the context of the top-down cascading effects of Hairston et al. (1960). Thus, my study is unique in its demonstration

222

of top-down trophic cascade effects (sensu Hairston et al. 1960) to a terrestrial epigeal community reliant on plant litter as the basal resource.

My result that arthropod species richness was negatively correlated to plant species richness directly contrasts with conclusions of other related studies that found arthropod species richness was positively correlated with plant species richness (Murdoch et al. 1972; Siemann et al. 1998; Knops et al. 1999; Haddad et al. 2000, 2001, Schaffers et al. 2008). While also showing that insect species richness was negatively correlated to plant species richness despite N fertilization, Siemann (1998) notes that this is most likely related to the increased abundance of insects in fertilized plots because of the increased resource base (i.e., increased plant production). Indeed, this may be the case with my study as I also had higher abundance of overall arthropods. Thus, even though species richness increased in fertilized plots, it might have been a result of simple increased probabilities of capture because of the increased numbers of arthropods, i.e., if there are more arthropods to catch, the chances of catching rarer species increases. However, this principle did not generally hold true with the detritivorous prey species, which did not significantly increase in abundance in fertilized plots, but certainly did experience higher species richness in fertilized plots. This was likely due the fertilization increasing the quality of their mainly plant detritus resource base. This increase in resource quality likely supported a more diverse detritivore community (Cross et al. 2006; Srivastava et al. 2009), which supported a greater abundance and diversity of predators, as predicted by top-down cascading effects (Hairston et al. 1960; Costamagna and Landis 2006).

Moreover, Siemann (1998) did look at detritivores, though these comprised only a small

223

portion (approximately 15–18%) of his captures, whereas detritivores comprised over half (approximately 59%) of the overall captures in this study, so the two studies are not easily compared.

Despite the overall increased abundance of arthropods, biomass was not significantly higher in fertilized plots in any given year, though fertilization did have a significant effect on biomass in the repeated-measures ANOVA. It is interesting to note that the effects of fertilization on biomass were opposite in predators and their detritivorous prey, with predator biomass slightly higher in unfertilized plots, while the detritivorous prey biomass was slightly (but not significantly in any given year) higher in fertilized plots. These contrasting responses resulted in the null response to fertilization for the overall biomass of arthropods. Additionally, while the abundance of predators increased, the biomass of predators did not, indicating that the average size of the predators decreased with increasing abundance. This surprising result indicates that the physical changes in habitat structure in fertilized plots favored smaller predators, possibly due to a higher density of plant shoots and, in litter left in situ plots, higher quantities of plant litter. While plant litter may increase predator abundance and diversity (Rypstra et al. 1999, 2007), the studies that found this relationship fundamentally differed from my system in that they were annual plant, monoculture agroecosystems with large amounts of bare ground—very different from the perennial graminoids of my system.

The significant response of predator biomass to litter removal (Fig. 29E), particularly in the unfertilized plot with litter removed, was caused by an unusually high number of captures of the second largest predator, the wolf spider Schizocosa avida

224

(Araneae, Lycosidae), most of which were caught in a single plot. The influence of this species is only for biomass in this unfertilized plot with litter removed, and therefore was not influential on the significance of litter for predator abundance and species richness during 2005. Both predator abundance and predator species richness not only responded to fertilization, but also to plant litter with litter removal resulting in increases in these response variables. This is a very surprising result given that numerous previous studies have linked predator diversity to increased habitat structure and plant litter (e.g., Lawton

1983; Strong et al. 1984; Uetz 1991; Rypstra et al. 1999; Halaj et al. 2000). The effects of the fertilization were very strong on the plant community of this system, with large changes in plant and litter production (Patrick et al. 2008a). This may have resulted in a microhabitat unsuitable for larger arthropod predators because of physical impediments to movement, vision, and hunting.

The epigeal predators that dominated this system rely on detritivores for significant portions of their diets (Sunderland 1975; Chiverton and Sotherton 1991).

While the overall effect of fertilization across the four years of the experiment did increase prey abundance and biomass (Table 12), the results within any given year were not significant. With more predators in fertilized plots, this might have been due to higher predation rates in fertilized plots. However, the sustained rate of prey abundance and biomass suggests that the nutrient enriched detrital resource base (i.e., plant litter) in fertilized plots might have attracted additional detritivorous prey from outside the plot.

Thus, rather than a sustained detritivore community at equilibrium, this might have been a resource source area that not only sustained a detritivore community, but also attracted

225

emigrants, effectively maintaining a constant inflow of prey for the predator community.

Conversely, many of the predators captured are fairly vagile and are wandering predators.

Thus, the fertilized plots may not have been their permanent residence, but rather an area that was foraged occasionally foraged, i.e., any increased detritivore community acts as an attractant for predators from outside the plot. However, were this the case, it would seem that the entire suite of predators captured in the unfertilized plots would have also been captured with equal (or greater) frequency in the fertilized plots. This did not happen, as evidenced by the increase in abundance and diversity of predators in fertilized plots without a change in the biomass of predators captured in fertilized plots. Thus, the largest predators were not frequently captured in fertilized plots and it is more likely that the epigeal community of fertilized plots were indeed somewhat autonomous relative to the epigeal community of unfertilized plots.

This study demonstrates that the epigeal arthropod community responds very differently than the portion of the arthropod community more dependent on living plant material. When released from the density-dependent factor of nutrient limitation, detritivore diversity increases but detritivore abundance and biomass remain relatively unchanged. This was likely due to the increased abundance and diversity of predators, which acted on the detritivores as would be expected in a top-down cascade. Thus, it is detrital quality that has a larger control on the detritus-based epigeal community and this component of the food web can alter ecosystem level processes, such as nutrient regeneration and ecosystem function (Moore et al. 2004). Greater care must be taken to

226

incorporate the epigeal community into our understanding of terrestrial ecosystem function, particularly as it relates to biodiversity.

227

References

Bobbink, R., M. Hornung, and J. G. M. Roelofs. 1998. The effects of air-borne nitrogen

pollutants on species diversity in natural and semi-natural European vegetation.

Journal of Ecology 86:717–738.

Chen, B., and D. H. Wise. 1999. Bottom-up limitation of predaceous arthropods in a

detritus-based terrestrial food web. Ecology 80:761–772.

Chiverton, P.A., and N. W. Sotherton. 1991. The effect on beneficial arthropods of the

exclusion of herbicides from cereal crop edges. Journal of Applied Ecology

28:1027–1039.

Cook, W. M., J. Yao, B. L. Foster, R. D. Holt, and L. B. Patrick. 2005. Secondary

succession in an experimentally fragmented landscape: community patterns across

space and time. Ecology 86:1267–1279.

Costamagna, A. C., and D. A. Landis. Predators exert top-down control of soybean aphid

across a gradient of agricultural management systems. Ecological Applications

16:1619–1628.

Cross, W. F., J. B. Wallce, A. D. Rosemond, and S. L. Eggert. Whole-system nutrient

enrichment increases secondary production in a detritus-based system. Ecology

87:1556–1565.

Cummings, G. S., and M. F. Child. 2009. Contrasting spatial patterns of taxonomic and

functional richness offers insights into potential loss of ecosystem services.

228

Philosophical Transactions of the Royal Society B – Biological Sciences

364:1683–1692.

Dybzinski, R., J. E. Fargione, D. R. Zak, D. Fornaro, and D. Tilman. 2008. Soil fertility

increases with plant species diversity in a long-term biodiversity experiment.

Oecologia 158:85–93.

Elton, C. S. 1958. The Ecology of Invasions by Animals and Plants. Methuen, Landon,

UK.

Fenn, M. E., J. S. Baron, E. B. Allen, H. M. Rueth, K. R. Nydick, L. Geiser, W. D.

Bowman, J. O. Sickman, T. Meixner, D. W. Johnson, and P. Neitlich. 2003.

Ecological effects of nitrogen deposition in the western United States.

BioScience 53:404–420.

Fraser, L. H. 1998. Top-down vs bottom-up control influenced by productivity in a

North Derbyshire, UK, dale. Oikos 81:99–108.

Fraser, L. H., and J. P. Grime. 1997. Primary productivity and trophic dynamics

investigated in a North Derbyshire, UK, dale. Oikos 80:499–508.

Fraser, L. H., and J. P. Grime. 1998. Top-down control and its effects on the biomass

and composition of three grasses at high and low soil fertility in outdoor

microcosms. Oecologia 113:239–246.

Fretwell, S. D. 1977. The regulation of plant communities by food chains exploiting

them. Perspectives of Biology and Medicine 20:169–185.

Galloway, J. N., J. D. Aber, J. W. Erisman, S. P. Seitzinger, R. W. Howarth, E. B.

Cowling, and B. J. Cosby. 2003. The nitrogen cascade. BioScience 53:341–355.

229

Greenslade, P. J. M. 1964. Pitfall trapping as a method for studying populations of

Carabidae (Coleoptera). Journal of Animal Ecology 33:301–310.

Greenwood, J. L., A. D. Rosemond, J. B. Wallace, W. F. Cross, and H. S. Weyers. 2007.

Nutrients stimulate leaf breakdown rates and detritivore biomass: bottom-up

effects via heterotrophic pathways. Oecologia 151:637–649.

Grime, J. P. 1979. Plant Strategies and Vegetation Processes. John Wiley and Sons,

London.

Haddad, N. M., J. Haarstad, and D. Tilman. 2000. The effects of long-term nitrogen

loading on grassland insect communities. Oecologia 124:73–84.

Haddad, N. M., D. Tilman, J. Harstaad, M. Ritchie, and J. M. H. Knops. 2001.

Contrasting effects of plant richness and composition on insect communities: a

field experiment. The American Naturalist 158:17–35.

Hairston, N. G., F. E. Smith, and L. D. Slobodkin. 1960. Community structure,

population control, and competition. American Naturalist 94:421–425.

Halaj, J., A. B. Cady, and G. W. Uetz. 2000. Modular habitat refugia enhance generalist

predators and lower plant damage in soybeans. Environmental Entomology

29:383–393.

Halaj, J. and D. H. Wise. 2002. Impact of a detrital subsidy on trophic cascades in a

terrestrial grazing web. Ecology 83:3141–3151.

Hättenschwiler, S., and P. Gasser. 2005. Soil animals alter plant litter diversity effects

on decomposition. Proceedings of the National Academy of Sciences of the USA

102:1519–1524.

230

Hector, A., B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G.

Dimitrkopoulos, J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P.

Högberg, K. Huss-Danell, J. Joshi, A. Jumpponen, C. Körner, P. W. Leadley, M.

Loreau, A. Minns, C. P. H. Mulder, G. O’Donovan, S. J. Otway, J. S. Pereira, A.

Prinz, D. J. Read, M. Scherer-Lorenzen, E.-D. Schulze, A.-S. D.Siamantziouras,

E. M. Spehn, A. C. Terry, A. Y. Troumbis, F. I. Woodward, S. Yachi, and J. H.

Lawton. 1999. Plant diversity and productivity experiments in European

grasslands. Science 286:1123–1127.

Hodgson, J. G., K. Thompson, P. J. Wilson, and A. Bogaard. 1998. Does biodiversity

determing ecosystem function? The Ecotron experiment reconsidered. Functional

Ecology 12:843–848.

Hooper, D. U., and P. M. Vitousek. 1997. The effects of plant composition and diversity

on ecosystem processes. Science 277:1302–1305.

Hutchinson, G. E. 1959. Homage to Santa Rosalia or why are there so many kinds of

animals? American Naturalist 93:145–159.

Ilieva-Makulec, K., I. Olejniczak, and M. Szanser. 2006. Response of soil micro- and

mesofauna to diversity and quality of plant litter. European Journal of Soil

Biology 42(Supplement):S244–S249.

Kimbro, D. L., E. D. Grosholz, A. J. Baukus, N. J. Nesbitt, N. M. Travis, S. Attoe, and C.

Coleman-Hulbert. 2009. Invasive species cause large-scale loss of native

California oyster habitat by disrupting trophic cascades. Oecologia 160:563–575.

231

Knops, J. M. H., D. Tilman, N. M. Haddad, S. Naeem, C. E. Mitchell, J. Harstaad, M. E.

Ritchie, K. M. Howe, P. B. Reich, E. Siemann, and J. Groth. 1999. Effects of

plant species richness on invasion dynamics, disease outbreaks, insect

abundances, and diversity. Ecology Letters 2:286–293.

Köchy, M., and S. D. Wilson. 2005. Variation in nitrogen deposition and available soil

nitrogen in a forest-grassland ecotone in Canada. Landscape Ecology 20:191–

202.

Lawton, J. H. 1983. Plant architecture and the diversity of phytophagous insects.

Annual Review of Entomology 28:23–29.

Long, Z. T., C. L. Mohler, and W. P. Carson. 2003. Extending the resource

concentration Hypothesis to plant communities: effects of litter and herbivores.

Ecology 84:652–665.

Luckinbill, L. S. 1973. Coexistence in laboratory populations of Paramecium Aurelia

and its predator Didinium nasutum. Ecology 54:1320–1327.

Martinko, E. A., R. H. Hagen, and J. A. Griffith. 2006. Successional change in the insect

community of a fragmented landscape. Landscape Ecology 21:711–721.

Moore, J. C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C.

Johnson, K. S. McCann, K. Melville, P. J. Morin, A. D. Rosemond, D. M. Post, J.

L. Sabo, K. M. Scow, M. J. Vanni, and D. H. Wall. 2004. Detritus, trophic

dynamics, and biodiversity. Ecology Letters 7:584–600.

Murdoch, W. W., F. C. Evans, and C. H. Peterson. 1972. Diversity and pattern in plants

and insects. Ecology 53:819–829.

232

Naeem, S., L. J. Thompson, S. P. Lawler, J. H. Lawton, and R. M. Woodfin. 1994.

Declining biodiversity can alter the performance of ecosystems. Nature 368:734–

737.

Nogué, S., V. Rull, and T. Vegas-Vilarrúbia. 2009. Modeling biodiversity loss by global

warming on Pantepui, northern South America: projected upward migration and

potential habitat loss. Climate Change 94:77–85.

Oksanen, L., S. D. Fretwell, J. Arruda, and P. Niemelä. 1981. Exploitation ecosystems

in gradients of primary productivity. American Naturalist 118:240–261.

Oksanen, L., and T. Oksanen. 2000. The logic and realism of the hypothesis of

exploitation ecosystems. American Naturalist 155:703–723.

Pan, J. J., B. Widner, D. Ammerman, R. E. Drenovsky. In press. Plant community and

tissue chemistry responses to fertilizer and litter nutrient manipulations in a

temperate grassland. Plant Ecology.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008a. Large-scale manipulation of

plant litter and fertilizer in a managed successional temperate grassland. Plant

Ecology 197:183–195.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008b. ―Brown‖ world invertebrates

contradict ―green‖ world biodiversity theory. Research Letters in Ecology, vol.

2008, Article ID 694638, 4 pages, doi:10.1155/2008/694638.

Phillips, I. D., and T. P. Cobb. 2005. Effect of habitat structure and lid transparency on

pitfall catches. Environmental Entomology 34:875–882.

233

Ritchie, A., and J. R. Steiger. 1974. Soil Survey of Summit County Ohio. United States

Department of Agriculture, Soil Conservation Service in cooperation with Ohio

Department of Natural Resources, Division of Lands and Soil and the Ohio

Agricultural Research and Development Center.

Root, R. B. 1973. Organization of plant-arthropod association in simple and diverse

habitats: the fauna of collards (Brassica oleraceae). Ecological Monographs

43:95–124.

Rosenzweig, M. L. 1971. Paradox of enrichment: destabilization of exploitative

ecosystems in ecological time. Science 171:385–387.

Rosenzweig, M. L. 1973. Exploitation in three trophic levels. American Naturalist

107:275–294.

Roy, S., and J. Chattopadhyay. 2007. The stability of ecosystems: a brief overview of

the paradox of enrichment. Journal of Biosciences 32:421–428.

Rypstra, A. L., P. E. Carter, R. A. Balfour, and S. D. Marshall. 1999. Architectural

modifications of agricultural habitats and their impact on the spider inhabitants.

Journal of Arachnology 27:371–377.

Rypstra, A. L., J. M. Schmidt, B. D. Reif, J. DeVito, and M. H. Persons. 2007.

Tradeoffs involved in site selection and foraging in a wolf spider: effects of

substrate structure and predation risk. Oikos 116:853–863.

SAS Institute Inc. 1999. SAS System for Windows. SAS Institute Inc., Cary, NC, USA.

234

Schaffers, A. P., I. P. Raemakers, K. V. Sykora, and C. J. F. Ter Braak. 2008. Arthropod

assemblages are best predicted by plant species composition. Ecology 89:782–

794.

Schuldt, A., N. Fahrenholz, M. Brauns, S. Migge-Kleian, C. Platner, and M. Schaefer.

2008. Communities of ground-living spiders in deciduous forests: does tree

species diversity matter? Biodiversity Conservation 17:1267–1284.

Siemann, E. 1998. Experimental tests of effects of plant productivity and diversity on

grassland arthropod diversity. Ecology 79:2057–2070.

Siemann, E., D. Tilman, J. Haarstad, and M. Ritchie. 1998. Experimental tests of the

dependence of arthropod diversity on plant diversity. American Naturalist

152:738–750.

Southwood, T. R. E., V. K. Brown, and P. M. Reader. 1979. The relationships of plant

and insect diversities in succession. Biological Journal of the Linnean Society

12:327–348.

Srivastava, D. S., B. J. Cardinale, A. L. Downing, J. E. Duffy, C. Jouseau, M. Sankaran,

and J. P. Wright. 2009. Diversity has stronger top-down than bottom-up effects

on decomposition. Ecology 90:1073–1083.

Strong, D. R., J. H. Lawton, and R. Southwood. 1984. Insects on plants: community

patterns and mechanisms. Harvard University Press, Cambridge, MA, USA.

Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.

Milchunas, and S. Pennings. 2005. Functional- and abundance-based

235

mechanisms explain diversity loss due to N fertilization. Proceedings of the

National Academy of Sciences of the USA 102:4387–4392.

Sunderland, K. D. 1975. The diet of some predatory arthropods in cereal crops. Journal

of Applied Ecology 12:507–515.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002a. Plant diversity and composition:

effects on productivity and nutrient dynamics of experimental grasslands. In:

Loreau, M., S. Naeem, and P. Inchausti, eds., Biodiversity and Ecosystem

Functioning. Synthesis and Perspectives. Oxford University Press, Oxford, pp.

21–35.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002b. Experimental and observational

studies of diversity, productivity, and stability. In: Kinzig, A. P., S. W. Pacala,

and D. Tilman, eds., The Functional Consequences of Biodiversity. Princeton

University Press, Princeton, pp. 42–70.

Tilman, D., C. L. Lehmen, and K. T. Thompson. 1997. Plant diversity and ecosystem

productivity: theoretical considerations. Proceedings of the National Academy of

Sciences of the USA 94:1857–1861.

Tilman, D., D. Wedin, and J. Knops. 1996. Productivity and sustainability influenced by

biodiversity in grassland ecosystems. Nature 379:718–720.

Uetz, G. W. 1991. Habitat structure and spider foraging. In: Bell, S. S., E. D. McCoy,

and H. R. Mushinsky, eds., Habitat structure: the physical arrangement of objects

in space. Chapman and Hall, London, UK, pp. 325–348.

236

US EPA. 2005. Clean air status and trends network (CAST-NET).

http://www.epa.gov/castnet/

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human

domination of Earth’s ecosystems. Science 277:494–499.

Vitousek, P. M., S. Hättenschwiler, L. Olander, and S. Allison. 2002. Nitrogen and

nature. Abio 31:97–101.

CHAPTER 7

SUMMARY AND IMPORTANCE OF THIS STUDY

In the portion of the food web more reliant on living, standing crop plant material—the ―green‖ portion of the food web (sensu Patrick et al. 2008a), decreased plant species richness may result in increased abundance of insect pest species (Elton

1958; Root 1973; Strong et al. 1984; Haddad et al. 2001) and lower insect species richness (Southwood et al. 1979; Strong et al. 1984; Siemann et al. 1998; Knops et al.

1999; Haddad et al. 2000, 2001). Nutrient loading not only increases plant standing crop biomass, but also plant litter production (Long et al. 2003; Patrick et al. 2008b), which can increase the basal food resource for the detrital community, resulting in increased detritivore and epigeal predator abundances (Halaj et al. 2000; Halaj and Wise 2002;

Moore et al. 2004). Furthermore, plant litter increases habitat complexity, which can also increase arthropod abundance and diversity (Lawton 1983; Strong et al. 1984). These important factors affect both biodiversity and trophic interactions, and my experiment has measured these effects in one of the largest spatial-scale studies of its kind.

The plant community largely responded as predicted by biodiversity-productivity theory, but some notable aspects of the plant community did not. The loss of forb species due to fertilization was consistent with the abundance-based mechanism of diversity loss

237 238

due to fertilization (Suding et al. 2005). However, it is striking that fertilization significantly affected only forb species richness, not forb species biomass, contradicting biodiversity-productivity theory. Also contradicting conventional theory (e.g., Tilman et al. 2002a; Suding et al. 2005), grass species richness showed no response to fertilization, but grass species biomass was strongly affected by fertilization. This was likely due to my experiment using an established plant community, where the main experiments involved with the formulation of the biodiversity-productivity theory (Suding et al. 2005) were seeded communities (e.g., Tilman et al. 2002a, b). Because N enrichment via atmospheric deposition and fertilization most often affects established plant communities, my results more likely reflect plant community responses in temperate latitude grasslands.

Epigeal arthropod community responses to fertilization were mixed. When looking at beetles and spiders separately, taxa that are typically used as indicators, fertilization increased abundance and biomass as predicted by biodiversity-productivity theory, but species richness was not affected as predicted. Beetle species richness actually increased in fertilized plots, as did the species richness of linyphiid spiders.

Moreover, the detritivorous prey of the beetles and spiders did not increase in biomass or abundance as would be expected by biodiversity-productivity theory. Further contradicting theory, the detritivorous prey also increased species richness. These surprising results are likely due to a bottom-up enrichment (via fertilization and a resulting increased in nutrient-enriched plant litter) in the food web that supported a greater diversity of predators, in turn resulting in no significant changes in prey 239

(detritivore) biomass and abundance because of a top-down effect of the predators

(Hairston et al. 1960; Oksanen et al 1981; Oksanen and Oksanen 2000).

An additional component of my dissertation was biodiversity discovery. Chapter

5 provided the first description of the female of Mermessus jona, one of the smallest spiders in North America (~ 1 mm total length), an expansion of its range, and a state record of the spider for Ohio. Additionally, Appendices 1 and 2 include a description of a new species of staphylinid beetle, Myrmedonota aidani, and documentation of several new state records of heteropterans, respectively. In addition to exploring ecological hypotheses, ecological studies provide unique opportunities to also document and catalogue biodiversity.

The overall size and scope of the experiment is notable. Most studies of the effects of fertilization and plant litter have focused on a single, dominant species or small collection of species within the entire plant community, only secondarily considering species richness and community structure (e.g., Foster and Gross, 1997, 1998; Long et al.

2003; Violle et al. 2006). Further, most studies have been at small scales, with experimental plots ranging from 0.49 m2 (e.g., Levine et al. 1998) to 25 m2 (e.g.,

Turkington et al. 2002), while two experiments had larger experimental plots of 100 m2, one in tundra (Gough and Hobbie 2003) and one in a successional old-field (Carson and

Barrett 1988). To my knowledge, the largest experimental plots to explicitly manipulate fertilization to investigate the effects on plant community structure were 2500 m2 in arctic tundra heath (Grellman 2002) and 4000 m2 in a mid-successional old-field

(Bakelaar and Odum 1978), though these two experiments neither fertilized for more than 240

the first year of the experiment nor manipulated plant litter in any way. Thus, with experimental plots of 314 m2, my study is historically the third largest experiment of its kind and presently the largest to manipulate fertilization and plant litter on an annual basis.

The spatial and temporal scale of my study allowed me to also assess both the species-level responses to my treatments within the entire plant community and the ecosystem-level responses to my treatments. The longer duration of my study allowed the largely perennial plant community to more fully respond to my treatments, an aspect often lacking in shorter term studies in habitats dominated by perennial species (Cook et al. 2005). The large size of my experimental plots integrated important determinants of the within-plot plant communities, including spatial heterogeneity (De Boeck et al. 2006), leaching of nutrients from litter (Berendse 1998), local nutrient cycling (Hooper and

Vitousek 1998), and the translocation of nutrients within clumping and clonal plants

(Hutchings and Bradbury 1986), which are the primary growth forms of our dominant graminoids.

My dissertation expands our understanding of anthropogenic influences on habitats by exploring portions of the community often ignored or underappreciated by previous studies. Moreover, my dissertation work helps to expand biodiversity- productivity theory to include established communities at a large scale. Trophic interaction theory is commonly tested on the living food web; here I show that top-down control may be an important component within the detritivore food chain.

241

Future directions

My study examined only the first four years of the experimental manipulations of fertilization and plant litter. Inclusion of the ―green‖ world arthropods with a concurrent sampling of the epigeal community would help with direct comparisons between the

―green‖ world and the ―brown‖ world arthropod communities. Moreover, an assessment of the floral differences has already yielded promising results (Pan et al. in press), and more detailed studies of seed size, quantity, and viability are needed to better understand the potential evolutionary consequences of anthropogenic nutrient loading. Another potential direction would be to examine the below-ground processes of the soil biota

(e.g., soil nematodes, earthworms, and possibly the soil bacteria and fungi) and belowground plant dynamics (e.g., plant resource allocation to above- vs. below-ground plant parts). The soil biology of a system is often overlooked as a ―black box‖ that is difficult to study, but it is an important part of nutrient cycling and plant community dynamics.

Thus, an examination of this component of the system could help elucidate the fine-scale mechanisms of terrestrial eutrophication.

My dissertation has helped define species new to science, filling in some small pieces of the puzzle in biodiversity inventorying. This type of work is important and must continue. Ironically, the USA has some of the top taxonomists in the world, yet our own fauna is often overlooked as research money pours in to fund studies in the tropics.

However, as my study has shown, even the temperate latitudes still have undiscovered species and surprising diversity. Such species inventories mean the most when they stem 242

from ecological studies because it then becomes possible to also better understand the biology of species, not just their collections sites and names. As I look to expand on my dissertation work, I do so with an eye towards also understanding the biology of biodiversity, not just the taxonomy.

243

References

Bakelaar, R. G. and E. P. Odum. 1978. Community and population level responses to

fertilization in an old-field ecosystem. Ecology 59:660–665.

Berendse, F. 1998. Effects of dominant plant species on soil during succession in

nutrient-poor ecosystems. Biogeochemistry 42:73–88.

Carson, W. P. and G. W. Barrett. 1988. Succession in old-field plant communities:

effects of contrasting types of nutrient enrichment. Ecology 69:984–994.

Cook, W. M., J. Yao, B. L. Foster, R. D. Holt, and L. B. Patrick. 2005. Secondary

succession in an experimentally fragmented landscape: community patterns across

space and time. Ecology 86: 1267-1279.

De Boeck, H. J., I. Nijs, C. M. H. M. Lemmens, and R. Ceulemans. 2006. Underlying

effects of spatial aggregation (clumping) in relationships between plant diversity

and resource uptake. Oikos 113:269–278.

Elton, C. S. 1958. The Ecology of Invasions by Animals and Plants. Methuen, Landon,

UK.

Foster, B. L., and K. L. Gross. 1997. Partitioning the effects of plant biomass and litter

on Andropogon gerardii in old-field vegetation. Ecology 78:2091–2104.

Foster, B. L., and K. L. Gross. 1998. Species richness in a successional grassland:

effects of nitrogen enrichment and plant litter. Ecology 79:2593–2602. 244

Gough, L. and S. E. Hobbie. 2003. Responses of moist non-acidic arctic tundra to

altered environment: productivity, biomass, and species richness. Oikos 103:204–

216.

Grellman, D. 2002. Plant responses to fertilization and exclusion of grazers on an arctic

tundra heath. Oikos 98:190–204.

Haddad, N. M., J. Haarstad, and D. Tilman. 2000. The effects of long-term nitrogen

loading on grassland insect communities. Oecologia 124:73–84.

Haddad, N. M., D. Tilman, J. Harstaad, M. Ritchie, and J. M. H. Knops. 2001.

Contrasting effects of plant richness and composition on insect communities: a

field experiment. The American Naturalist 158:17–35.

Halaj, J., A. B. Cady, and G. W. Uetz. 2000. Modular habitat refugia enhance generalist

predators and lower plant damage in soybeans. Environmental Entomology

29:383–393.

Halaj, J. and D. H. Wise. 2002. Impact of a detrital subsidy on trophic cascades in a

terrestrial grazing web. Ecology 83:3141–3151.

Hooper, D. U., and P. M. Vitousek. 1998. Effects of plant competition and diversity on

nutrient cycling. Ecological Monographs 68:121–149.

Hutchings, M. J., and I. K. Bradbury. 1986. Ecological perspectives on clonal perennial

plants. BioScience 36:178–182.

Knops, J. M. H., D. Tilman, N. M. Haddad, S. Naeem, C. E. Mitchell, J. Harstaad, M. E.

Ritchie, K. M. Howe, P. B. Reich, E. Siemann, and J. Groth. 1999. Effects of 245

plant species richness on invasion dynamics, disease outbreaks, insect

abundances, and diversity. Ecology Letters 2:286–293.

Lawton, J. H. 1983. Plant architecture and the diversity of phytophagous insects.

Annual Review of Entomology 28:23–29.

Levine, J. M., J. S. Brewer, and M. D. Bertness. 1998. Nutrients, competition and plant

zonation in a New England salt marsh. Journal of Ecology 86:285–292.

Long, Z. T., C. L. Mohler, and W. P. Carson. 2003. Extending the resource

concentration hypothesis to plant communities: effects of litter and herbivores.

Ecology 84: 652–665.

Moore, J. C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C.

Johnson, K. S. McCann, K. Melville, P. J. Morin, A. D. Rosemond, D. M. Post, J.

L. Sabo, K. M. Scow, M. J. Vanni, and D. H. Wall. 2004. Detritus, trophic

dynamics, and biodiversity. Ecology Letters 7:584–600.

Oksanen, L., S. D. Fretwell, J. Arruda, and P. Niemelä. 1981. Exploitation ecosystems

in gradients of primary productivity. American Naturalist 118:240–261.

Oksanen, L., and T. Oksanen. 2000. The logic and realism of the hypothesis of

exploitation ecosystems. American Naturalist 155:703–723.

Pan, J. J., B. Widner, D. Ammerman, R. E. Drenovsky. In press. Plant community and

tissue chemistry responses to fertilizer and litter nutrient manipulations in a

temperate grassland. Plant Ecology. 246

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008a. ―Brown‖ world invertebrates

contradict ―green‖ world biodiversity theory. Research Letters in Ecology, vol.

2008, Article ID 694638, 4 pages, doi:10.1155/2008/694638.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008b. Large-scale manipulation of

plant litter and fertilizer in a managed successional temperate grassland. Plant

Ecology 197:183–195.

Root, R. B. 1973. Organization of plant-arthropod association in simple and diverse

habitats: the fauna of collards (Brassica oleraceae). Ecological Monographs

43:95–124.

Siemann, E., D. Tilman, J. Haarstad, and M. Ritchie. 1998. Experimental tests of the

dependence of arthropod diversity on plant diversity. American Naturalist

152:738–750.

Southwood, T. R. E., V. K. Brown, and P. M. Reader. 1979. The relationships of plant

and insect diversities in succession. Biological Journal of the Linnean Society

12:327–348.

Strong, D. R., J. H. Lawton, and R. Southwood. 1984. Insects on plants: community

patterns and mechanisms. Harvard University Press, Cambridge, MA, USA.

Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.

Milchunas, and S. Pennings. 2005. Functional- and abundance-based

mechanisms explain diversity loss due to N fertilization. Proceedings of the

National Academy of Sciences of the USA 102:4387–4392. 247

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002a. Plant diversity and composition:

effects on productivity and nutrient dynamics of experimental grasslands. In:

Loreau, M., S. Naeem, and P. Inchausti, eds., Biodiversity and Ecosystem

Functioning. Synthesis and Perspectives. Oxford University Press, Oxford, pp.

21–35.

Tilman, D., J. Knops, D. Wedin, and P. Reich. 2002b. Experimental and observational

studies of diversity, productivity, and stability. In: Kinzig, A. P., S. W. Pacala,

and D. Tilman, eds., The Functional Consequences of Biodiversity. Princeton

University Press, Princeton, pp. 42–70.

Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of

fertilization and herbivory on the herbaceous vegetation of the boreal forest in

north-western Canada: a 10-year study. Journal of Ecology 90:325–337.

Violle, C., J. Richarte, and M. Navas. 2006. Effects of litter and standing biomass on

growth and reproduction of two annual species in a Mediterranean old-field.

Journal of Ecology 94:196–205.

APPENDIX 1

FIRST RECORD OF THE GENUS MYRMEDONOTA CAMERON

(COLEOPTERA, STAPHYLINIDAE) FROM NORTH AMERICA, WITH

DESCRIPTIONS OF TWO NEW SPECIES

(This appendix was published in February 2008 in Zootaxa)

Abstract

The lomechusine genus Myrmedonota Cameron is recorded from North America for the first time. Myrmedonota is diagnosed, and two new species, M. aidani Maruyama and Klimaszewski, sp. nov., and M. lewisi Maruyama and Klimaszewski, sp. nov., are described and illustrated with new bionomical information.

Introduction

The genus Myrmedonota Cameron, 1920 (tribe Lomechusini Fleming, subtribe

Myrmedoniina Thomson) is represented by 11 species from Malaysia, Singapore,

Indonesia and Papua-New Guinea. According to the original description of the type

248 249

species of Myrmedonota (Cameron, 1920), the type series were collected with small ants, so it is presumed to be a myrmecophile. Though bionomics of the most other species are unknown, one species is also reported as a myrmecophile associated with Papyrius nitidus (Mayr, 1862) (Hymenoptera, Formicidae, Dolichoderinae) (Kistner, 2003), and another species is known as a termitophile associated with Schedorhinotermes sp.

(Isoptera, Rhinotermitidae) (Bourguignon and Roisin, 2006).

The original diagnosis of Myrmedonota was not well defined (Cameron, 1920), and the description is not useful for discriminating it from the other lomechusine genera.

Probably, the affiliations of most species have been based on the similarities of their facies to the type species and the small body size in comparison to the other lomechusine members. Bourguignon and Roisin (2006) mentioned some diagnostic features of the genus, but they are rather general features that are shared by other lomechisine genera, e.g., Myrmoecia Mulsant and Rey, 1874, Pella Stephens, 1835, and Zyras Stephens, 1835

(s. lat.).

Recently, we have found two undescribed species of the Lomechusini in the

United States. After close examination of the specimens and comparison with the type species, they are found to be members of the genus Myrmedonota. In this paper, we redefine the genus based on the type species and new material from North America, provide external and internal illustrations of diagnostic structures for the first time, and describe the new species with the bionomical information.

250

Materials and methods

Technical procedures and terminology adopted here follow Maruyama (2006).

The following abbreviations are used for measurements in the descriptions: BL, body length; FBL, fore body length; HW, head width; EL, eye length; AL, antennal length; PL, pronotal length; PW, pronotal width; ELW, elytral width; HTL, hind tibial length. The specimens are deposited in the following collections: the Insectarium René-Martineau,

Canadian Forest Service (LFC), the Canadian National Collection of Insects, Arachnids, and Nematodes (CNC), the Field Museum of Natural History (FMNH), and the private collection of Munetoshi Maruyama (cMM).

Taxonomy

Genus Myrmedonota Cameron, 1920

See, Kistner (2003) for references, list of species, and key to species.

Type species: Myrmedonota cingulata Cameron, 1920, by monotypy.

Diagnosis. This genus is characterised and distinguished from the other lomechusine genera by a combination of the following character states: 1) body surface finely punctated; 2) head with occipital suture; 3) pronotum transverse, >1.5 wider than long; 4) setation on abdomen sparse to moderate; 5) cardo of maxilla covers bases of stipes and lacinia (Fig. 23A); 6) lacinia extremely narrowed and parallel-sided (Fig.

251

Figure 22. Habitus of Myrmedonota spp. A) M. aidani, B) M. lewsi.

A B

252

23A); 7) mentum almost as long as wide (Figs. 23B, 25A); 8) apodeme of labium with medial projection (Figs. 23C, 25B); 9) 1st segment of labial palpus longer than 2nd segment (Figs. 23C, 25B); 10) each lobe of ligula with 2 setulae (Figs. 23C, 25B).

Comments. Myrmedonota is placed in the subtribe Myrmedoniia of the tribe

Lomechusini by the combination of the following character states: elongated lacinia; mesosternal process shorter than metasternal process; tarsal formula 4, 5, 5; base of capsule of aedeagal median lobe partially covers the compressor plate.

This is the first record of Myrmedonota from North America. The diagnosis is based on an examination of the type species M. cingulata Cameron, 1920; 2 m# m# and 2 f#f# (Ulu Gombak, Selangor, Malaysia, 8–15 VI 2005, coll. M. Maruyama, by flight interception traps); the present new species, and the original descriptions of M. papyriomyrmecis Kistner, 2003 and M. termitophila Bourguignon and Roisin, 2006.

In North America, the members of this genus are similar to those of Pella in general appearance, but can be distinguished by the small body size (in Pella, > 3.5 mm) and the mouthpart character states: 5) to 10) of the above. This genus is keyed out as

Pella in the key of Klimaszewski et al (2006); see, “Modified Key to Species of Zyras

Group of Genera in America North of Mexico (Klimaszewski et al (2005, p. 708).” In

Ashe’s (2001) “Key to the Nearctic genera of Lomechusini (pp. 310-311)”, Pella and

Myrmedonota are keyed out to “Zyras.”

253

Myrmedonota aidani Maruyama and Klimaszewski, sp. nov.

Figs. 22A, 23–24

Type series. Holotype, m#, UNITED STATES OF AMERICA: Ohio: Summit

County, Bath Nature Preserve [41.18°N, 81.65°W], 20 VI 2003–4 VII 2003, coll. L. B.

Patrick (LFC). Paratypes: 5 m#m#, 5 f#f#, same data as holotype (LFC, FMNH); 5 males,

3 f#f#, 2 sex?, same data but 30 VI 2005 - 14 VII 2005 (LFC, CNC, cMM); 3 m#m#, 1 f#, same data but 28 VII 2005–11 VIII 2005 (cMM).

Etymology. Dedicated to Aidan C. Patrick, son of the collector of the original series, L. Brian Patrick.

Diagnosis. This species is closely similar to M. lewisi in facies and body size but may be easily distinguished by the mostly brown body color, the morphology of the genital organs, especially the wider dorsal bridge of the median lobe of aedeagus and the

S-shaped spermatheca. This species can easily be separated from the other known species of Myrmedonota by having 5 macrosetae along the lateral margin of the pronotum.

Description. Body slender and subparallel (Fig. 22A). Brown in ground colour; head, 5th to 11th segments of antennae, 5th to 8th segments (sometimes also median areas of 4th and 5th segments) of abdomen blackish brown; elytra paler, but more or less infuscate laterally and sometimes around scutellum. Head (Fig. 22A) widest at eyes; surface finely reticulated, moderately covered with setae; setae moderately long, as long as those on pronotum and elytra; length of eyes 0.45–0.46 times as long as head width.

254

Figure 23. Mouthparts of Myrmedonota aidani. A) maxilla, B) mentum, C) labium.

A B C 255

Figure 24. Terminalia of Myrmedonota aidani. A) male 8th tergite male, B) male 8th sternite, C) median lobe of aedeagus, lateral view, D) median lobe of aedeagus, ventral view, E) female 8th tergite, F) female 8th sternite, G) spermatheca.

256

B A

F E

C D 257

Mentum (Fig. 23B) trapezoidal, with basal margin emarginate. Labium (Fig. 23C) with about 30 minute medial pseudopores. Antennae (Fig. 22A) shorter than head, pronotum and elytra combined; 1st segment much shorter than 2nd and 3rd combined; 2nd segment about 0.7 times as long as 3rd; 3rd segment about 0.8 times as long as 1st; 4th to 10th segments almost as long as wide except for stem of each segment; 11th segment conical, longer than 1st. Pronotum (Fig. 22A) subelliptical, 1.37–1.44 times as wide as long, widest just after anterior margin; surface finely punctured, finely reticulated among punctures, densely covered with setae, with 5 long macrosetae along anterior to lateral margins; lengths of macrosetae variable, anterolateral one longest. Scutellum with surface smooth, moderately covered with short setae. Elytra (Fig. 22A) widened apicad; surface finely punctured, finely reticulated among punctures, densely covered with setae, with 3 small macrosetae laterally. Legs short; hind tibia 0.81–0.84 times as long as elytral width.

Abdomen subparallel-sided, slightly narrower than elytra, widest around 4th and 5th segments; surface smooth; 3rd to 7th tergites almost glabrous, but with a low of setae and macrosetae along posterior margins; 8th tergite (Figs. 24A, 24E) with 5 macrosetae; 9th tergite with 4 macrosetae; 10th tergite with posterior margin slightly emarginate, with 4 macrosetae.

Male: Eighth tergite (Fig. 24A) with posterior margin truncate, its truncate apex crenate, and with a protrusion laterally; 8th sternite (Fig. 24B) with posterior margin rounded, with 8 macrosetae; 9th sternite with posterior margin rounded. Aedeagus (Figs.

24C, 24D) somewhat tear-shaped in ventral view; apical lobe gently curved ventrad in lateral view, pointed at apex in lateral and ventral views; basal ridge convex. 258

Female: Eighth tergite (Fig. 24E) with posterior margin truncate; 8th sternite

(Fig. 24F) with 7 macrosetae; sensory setae of 8th sternite generalised, almost the same as the other setae in shape. Spermatheca (Fig. 24G) curved twice, S-shaped.

Measurements. BL, ca. 2.6–3.0; FBL, ca. 1.2–1.4; HW, 0.49–0.52; EL,

0.220–0.237; AL, 1.03–1.13; PL, 0.43–0.47; PW, 0.59–0.67; ELW, 0.75–0.86; HTL,

0.63–0.68.

Comments. This species, as well as the next species, are keyed out in first couplet of the key (Kistner, 2003) in having 5 macrosetae along the lateral margin of the pronotum.

Bionomics. This species was readily caught in pitfall traps with a 50/50 solution of water/propylene glycol. The habitat was an annually mown old-field grassland dominated by the European cool season grasses, Bromus inermis Leyss., Festuca arundinacea Schreb., Phleum pratense L., and Anthoxanthum odoratum L. The species was captured frequently in anthropogenically disturbed areas, especially with moderate to high plant productivity, and with moderate to high plant litter accumulation.

The type series specimens of M. aidani were caught with ants (Formicidae), though the exact species of ants were not determined, only the subfamily. Captured with the type series of this species were several specimens in each of the following subfamilies: Mymicinae, Dolichoderinae, and Formicinae. Additionally, a single specimen of the subfamily Ponerinae (presumably Ponera pennsylvanica Buckley) was captured with the holotype of M. aidani. Additional specimens of M. aidani not in the type series were captured with the following ant species: Lasius neoniger Emery, Lasius 259

alienus (Foerster), Lasius umbratus (Nylander), Brachymyrmex depilis Emery,

Prenolepis imparis (Say), Formica nitidiventris Emery, Myrmica latifrons Stärcke,

Myrmica americana Weber, Myrmica fracticornis Forel, Stenamma brevicorne (Mayr),

Stenamma impar Forel, Solenopsis molesta (Say), Aphaenogaster rudis complex

(Enzmann), and Ponera pennsylvanica Buckley. Further, M. aidani of the type series were captured with the following beetle species: Amara lunicollis Schiodte and Poecilus lucublandus (Say) in the family Carabidae, Barypeithes pallucidus (Boheman) in the family Curculionidae, and Apocellis sphaericollis (Say), Falagria dissecta Erichson,

Philhygra clemens (Casey), Amischa sp., and Meronera venustula Erichson in the family

Staphylinidae.

This species, as well as the next species, is most probably myrmecophilous, though none of the type series of the next species was collected with ants. Most species of the Lomechusini are considered to be myrmecophilous (Maruyama, 2006), and the present new species are closely similar to the type species, which has been presumed to be a myrmecophile. Further field investigation is needed for confirmation of their myrmecophily.

Myrmedonota lewisi Maruyama and Klimaszewski, sp. nov.

Figs. 22B, 25–26

Type series. Holotype, m#, UNITED STATES OF AMERICA: Indiana: Clark 260

County, Burns Hollow, ca 2 mi E Borden, 30 VII 2006, J. Lewis (LFC). Paratypes. 2 m#m#, 8 f#f#, 1 sex?, same data as holotype (LFC, CNC, FMNH, cMM).

Etymology. Named in honour of Jerry Lewis, the collector of the original series.

Diagnosis. This species is closely similar to M. aidani in facies and body size but may be easily distinguished by the basically blackish brown body color, the character states of the genital organs, especially the narrower dorsal bridge of the median lobe of the aedeagus and the V-shaped spermatheca. This species can easily be separated from the other known species of Myrmedonota by having 5 macrosetae along the lateral margin of the pronotum.

Description. Body slender and subparallel (Fig. 22B). Blackish brown in ground colour; head, 1st to 3rd segments of antennae, shoulders and posterior margins of elytra yellowish brown. Head (Fig. 22B) widest at eyes; surface finely reticulated, moderately covered with setae; setae moderately long, as long as those on pronotum and elytra; length of eyes 0.38–0.40 times as long as head width. Mentum (Fig. 25A) trapezoidal, with basal margin shallowly emarginated. Labium (Fig. 25B) with about 15 medial pseudopores. Antennae (Fig. 22B) shorter than head, pronotum and elytra combined; 1st segment much shorter than 2nd and 3rd combined; 2nd segment about 0.8 times as long as

3rd; 3rd segment about 0.7 times as long as 1st; 4th to 10th segments almost as long as wide except for stem of each segment; 11th segment conical, longer than 1st. Pronotum (Fig.

22B) subelliptical, 1.54–1.60 times as wide as long, widest just after anterior margin; surface finely punctured, finely reticulated among punctures, densely covered with setae, with 5 long macrosetae along anterior to lateral margins; lengths of macrosetae almost

261

Figure 25. Mouthparts of Myrmedonota lewisi. A) mentum, B) labium.

A B

262

Figure 26. Terminalia of Myrmedonota lewisi. A) male 8th tergite, B) male 8th sternite, C) median lobe of aedeagus, lateral view, D) median lobe of aedeagus, ventral view, E) female 8th tergite, F) female 8th sternite, G) spermatheca.

263

A B

C D 264

same. Scutellum with surface smooth, moderately covered with short setae. Elytra (Fig.

22B) widened apicad; surface finely punctured, finely reticulated among punctures, densely covered with setae, with 3 distinct macrosetae laterally. Legs short; hind tibia

0.72–0.75 times as long as elytral width. Abdomen subparallel-sided, slightly narrower than elytra, widest around 4th and 5th segments; surface smooth; 3rd to 7th tergites almost glabrous, but with a low of setae and macrosetae along posterior margins; 8th tergite (Figs.

26A, 26E) with 5 macrosetae; 9th tergite with 4 macrosetae; 10th tergite with posterior margin slightly emarginate, with 4 macrosetae.

Male: Eighth tergite (Fig. 26A) with posterior margin emarginate, its emarginate apex crenate, and protruded laterally; 8th sternite (Fig. 26B) with posterior margin rounded, with 7 macrosetae; 9th sternite with posterior margin truncated. Aedeagus (Figs.

26C, 26D) somewhat tear-shaped in ventral view; apical lobe gently curved ventrad in lateral view, pointed at apex in lateral and ventral views; basal ridge convex.

Female: Eighth tergite (Fig. 26E) with posterior margin almost truncate; 8th sternite (Fig. 26F) with 7 macrosetae; sensory setae of 8th sternite generalised, almost the same as the other setae in shape. Spermatheca (Fig. 26G) curved once.

Measurements. BL, ca. 2.3–2.6; FBL, ca. 1.0–1.1; HW, 0.48–0.50; EL,

0.175–0.181; AL, 0.88–0.99; PL, 0.35–0.40; PW, 0.56–0.61; ELW, 0.68–0.76; HTL,

0.51–0.56.

Comments. This species, as well as the previous species, are keyed out in first couplet of the key (Kistner, 2003) in having 5 macrosetae along the lateral margin of the pronotum. 265

Bionomics. All the specimens were attracted to paint thinner set in a cave.

Modified Key to Species of Zyras Group of Genera in America North of Mexico

(Klimaszewski et al, 2005, p. 708)

- [p. 708, lines 33-36]; body larger, usually > 3.5 mm; mouthparts: cardo of maxilla do not covers bases of stipes and lacinia; lacinia narrowed apicad; mentum transverse; apodeme of labium without medial projection; 1st segment of labial palpus almost as long as 2nd segment; each lobe of ligula with 4 campaniform sensillae (sockets of setulae)

(Pella)………………….8

- [ditto]; body smaller, usually < 3.0 mm; mouthparts: cardo of maxilla covers bases of stipes and lacinia; lacinia extremely narrowed and parallel-sided; mentum almost as long as wide; apodeme of labium with medial projection; 1st segment of labial palpus longer than 2nd segment; each lobe of ligula with 2 setulae (Myrmedonota)…………..16

16 Body color mostly brown; dorsal bridge of median lobe of aedeagus wide (Fig. 24C); spermatheca S-shaped (Fig. 24G)………………….M. aidani

- Body color basically blackish brown; dorsal bridge of median lobe of the aedeagus narrow (Fig, 26C); spermatheca V-shaped (Fig. 26G)…………….M. lewisi

266

References

Ashe, J. S. 2001. Keys to the tribes and genera of Nearctic Aleocharinae. In: Arnett, R.

H., and M. C. Thomas, eds., American Beetles. 1. Archostemata, Myxophaga,

Adephaga, Polyphaga: Staphyliniformia. CRC Press, Boca Raton, FL, USA, pp.

299–374.

Bourguignon, T., and Y. Roisin. 2006. A new genus and three new species of

termitophilous staphylinids (Coleoptera: Staplylinidae) associated with

Schedorhinotermes (Isoptera: Rhinotermitidae) in New Guinea. Sociobiology

48:395–407.

Cameron, M. 1920. New species of Staphylinidae from Singapore. Part III.

Transactions of the Entomological Society, London 1920:212-274.

Kistner, D. H. 2003. A new species of Myrmedonota from Papua-New Guinea with

the first specific host record (Coleoptera: Staphylinidae, Lomechusini).

Sociobiology 42:519–532.

Klimaszewski, J, G. Pelletier, M. Maruyama, and P. Hlavac. 2005. Canadian species

of the Zyras group of genera and review of the types from America north of

Mexico (Coleoptera, Staphylinidae, Aleocharinae). Revue Suisse de Zoologie

122(3):703–733.

Mayr, G. 1862. Myrmecologische Studien. Verhandlungen der k. k.

zoologish-botanischen Gessellschaft in Wien 12:649–776.

Maruyama, M. 2006. Revision of the Palearctic species of the myrmecophilous genus 267

Pella (Coleoptera, Staphylinidae, Aleocharinae). National Science Museum

Monographs, 32. National Science Museum, Tokyo, Japan.

Mulsant, M. E., and C. Rey. 1874. Histoire naturalle des coléoptères de France:

Brévipennes, Aléochariens, suite. Deyrolle, Paris, France.

Stephens, J. F. 1835. Illustrations of British Entomology; or, a synopsis of indigenous

insects; containing their generic and specific distinctions; with an account of

their metamorphoses, times of appearance, localities, food, and economy, as far

as practicable. Mandibulata. 5:369–448. Baldwin and Cradock, London, UK.

APPENDIX 2

EIGHT NEW OHIO STATE RECORDS OF TRUE BUGS (HEMIPTERA,

HETEROPTERA) FROM PITFALL TRAPS

(This appendix was published January 2008 (2009) in The Great Lakes Entomologist)

Abstract

Thirty one species of true bugs (Hemiptera, Heteroptera) were collected from pitfall traps set in old-field grasslands in Summit and Portage Counties, Ohio, between

2002 and 2005. Of these, eight were new state records including: Corimelaena pulicaria

(Thyreocoridae), Cryphula trimaculata (Rhyparochromidae), Emesaya brevipennis brevipennis (Reduviidae), Hebrus burmeisteri (Hebridae), Oncerotrachelus acuminatus

(Reduviidae), Pagasa fusca fusca (Nabidae), Pycnoderes obscuratus (Miridae), and

Stygnocoris rusticus (Rhyparochromidae). Current distribution maps (north of Mexico) and bionomical information are provided for each species.

268 269

Introduction

Documentation of the true bugs (Hemiptera, Heteroptera) of the United States and

Canada is an ongoing effort. The most recent comprehensive sources of information on distribution are the catalog of North American Heteroptera (Henry and Froeschner 1988) and the checklist of Canadian species (Maw et al. 2000). Records are lacking for many common species that would be expected to occur in Ohio (Henry and Froeschner 1988).

Between 2002 and 2005, pitfall traps were used to capture insects and investigate the influence of management practices (e.g., mowing, fertilizer use) on grasslands in the

Bath Nature Preserve, Summit County Ohio (Patrick et al. 2008a, b). Although only a small fraction of the total insect community, 31 bug species were collected during the study, eight of which are new state records.

The purpose of this paper is to document the eight species as new state records for

Ohio. Additionally, I provide current distribution maps (north of Mexico) for all eight species because new records have been published since the catalog by Henry and

Froeschner (1988).

Materials and methods

Bugs were captured in pitfall traps and preserved in 80% ethanol. Wheeler

(1983), Henry and Froeschner (1988), McPherson (1992), Wheeler (1992), Maw et al.

(2000), Williams (2000), Chordas et al. (2005), Chordas and Kovarik (in press) were used as distributional references. Blatchley (1926), Knight (1941), McPherson (1982),

270

and Hilsenhoff (1986) were used as taxonomic references. Voucher specimens were deposited in the entomological collections at Kent State University (Kent, Ohio).

Duplicate specimens of state record taxa were deposited in the first author’s personal collection (SWAC collection Columbus, Ohio). Additionally, voucher specimens of

Pycnoderes obscuratus Knight, 1926 (Miridae) were deposited in the United States

National Museum (USNM, Washington, D.C).

Collection sites

All species, except for the Hebrus species, were collected from the following location: Ohio, Summit County, Bath Township, Bath Nature Preserve [41 10' 36.2" N :

81 38' 58.7" W], pitfall traps in old-field grassland, coll. L.B. Patrick. Dates (or date ranges) of collection are listed individually for each species. Patrick et al. (2008a) provided detailed habitat characterization, as well as specific and vegetative parameters of the location where these hemipterans were collected, and Patrick et al. (2008b) provided details on the pitfall trap sampling procedure. The Hebrus species was collected from a pitfall trap placed in the riparian zone of a constructed wetland pool on the Arthur and Margaret Herrick Aquatic Ecology Research Facility at Kent State

University, Portage County, Ohio [41 8'16.77"N : 81 20'22.32"W] (Lauffer 2004).

271

Results and Discussion

Hebridae

Hebrus burmeisteri Lethierry & Severin, 1896. Although considered a semiaquatic bug, this species was captured in a pitfall trap that was close to the water’s edge of a man-made wetland. Previously recorded for Michigan, Kentucky, and

Pennsylvania (Fig. 27). H. burmeisteri was expected to occur in Ohio. A single specimen was collected on 30 July 2003.

Miridae

Pycnoderes obscuratus Knight, 1926. No species of this genus had been reported previously from Ohio. This is not surprising as the mirids are not well known for Ohio, and the species of this genus are apparently rarely encountered throughout their known ranges. There are few records of any of the Pycnoderes species for the eastern

United States (Thomas Henry, USNM, personal communication). Four specimens were collected: 11-25 August 2004 (1 specimen), 1-15 July 2005 (2 specimens), and 29 July -

12 August 2005 (1 specimen). These Ohio records represent a western range extension and only the 3rd state (in addition to Pennsylvania and Virginia [Thomas Henry, personal communication]) with records of this species (Fig. 28).

272

Figure 27. Hebrus burmeisteri distribution north of Mexico. Canada: Nova Scotia,

Ontario, Quebec, Saskatchewan. United States: Florida, Georgia, Illinois, Iowa, Kansas,

Kentucky, Maryland, Massachusetts, Michigan, Missouri, New Hampshire, New Jersey,

Ohio, Pennsylvania, South Carolina, Virginia, Wisconsin.

273

Figure 28. Pycnoderes obscuratus distribution north of Mexico. United States: Ohio,

Pennsylvania, Virginia

274

This species was difficult to identify as there are few mirid keys that include Great

Lakes fauna and none that are specific to Ohio; Watson’s (1928) Miridae of Ohio lacks a key. Pycnoderes obscuratus was not included in Knight’s (1941) monograph of the

Miridae of Illinois or in Blatchley’s (1926) key, although Blatchley did mention it in the text summary of P. balli. Pycnoderes obscuratus originally was described as a variety of

P. balli, from which this species is clearly different. Because my specimens did not fit any Pycnoderes species in Blatchley’s (1926) key or any of his limited descriptions, I sent my specimens to Thomas Henry (USNM) for identification. A further confounding item was that all four of my Ohio specimens were partly brachypterous, a character not mentioned by Blatchley (1926).

Other bugs captured along with P. obscuratus in the same pitfall traps were three species of Rhyparochromidae: Myodocha serripes Olivier, 1811, Ptochiomera nodosa

Say, 1832, and Stygnocorus rusticus (Fallen, 1807). Other bugs captured in different pitfall traps during the same collection period and in the same sample plot were

Corimelaena pulicaria (Germar, 1829) (Thyreocoridae) and Cryphula trimaculata

(Distant, 1882) (Rhyparochromidae). Pycnoderes obscuratus was captured only in unfertilized plots; three of the four specimens were captured in plots with the plant litter removed (see Patrick et al. 2008a, b for treatment details).

275

Nabidae

Pagasa fusca fusca (Stein, 1857). This damsel bug is widespread in North

America (Fig. 29) and was expected for Ohio. One specimen was collected 30 June - 14

July 2005. No other bugs were captured with P. fusca fusca in the same pitfall trap, although Cryphula trimaculata (Rhyparochromidae) was captured in other pitfall traps during the same collection period and in the same sample plot.

Reduviidae

Emesaya brevipennis brevipennis (Say, 1928). This slender assassin bug occurs throughout the Great Lakes area, across the eastern half of the United States and into

Canada, with one disjunct record from California (Fig. 30). Ohio was within the known range of this species, and its occurrence was expected. Two specimens were collected 24

July - 7 August 2003 and 18 July - 1 August 2004. The only bug captured with E. brevipennis brevipennis in the same pitfall trap was Stygnocorus rusticus. No other bugs were captured in other pitfall traps during the same collection period and in the same sample plot.

Oncerotrachelus acuminatus (Say, 1832). Although there are no records of this species from Canada, it is now reported from all of the states bordering the Great Lakes except Wisconsin (Fig. 31). Previously recorded from Michigan, Indiana, and

Pennsylvania, this species was anticipated to occur in Ohio. Three specimens were

276

Figure 29. Pagasa fusca fusca distribution north of Mexico. Canada: Alberta, British

Columbia, Labrador, Manitoba, Newfoundland, Northwest Territories, Nova Scotia,

Ontario, Quebec, Saskatchewan, Yukon. United States: Arizona, California, Colorado,

Idaho, Illinois, Indiana, Kansas, Louisiana, Maine, Minnesota, Mississippi, Missouri,

Nebraska, Ohio, Pennsylvania, New York, South Dakota, Texas, Utah, Wisconsin.

277

Figure 30. Emesaya brevipennis brevipennis distribution north of Mexico. Canada:

Ontario. United States: Arkansas, California, Connecticut, Florida, Georgia, Illinois,

Indiana, Iowa, Kansas, Maryland, Massachusetts, Michigan, Missouri, New Jersey, New

York, North Carolina, Ohio, Oklahoma, Pennsylvania, Rhode Island, Texas.

278

Figure 31. Oncerotrachelus acuminatus distribution north of Mexico. United States:

Alabama, Delaware, Florida, Illinois, Indiana, Kansas, Maryland, Massachusetts,

Michigan, Minnesota, Missouri, New Jersey, New York, North Carolina, Ohio,

Oklahoma, Pennsylvania, South Carolina, Texas.

279

collected 1-15 June 2005 (2 specimens) and 30 June - 14 July 2005 (1 specimen). No other bugs were captured with O. acuminatus in the same pitfall traps, although

Myodocha serripes and Stygnocorus rusticus were captured in different pitfall traps during the same collection period and in the same sampling plot.

Rhyparochromidae

Cryphula trimaculata (Distant, 1882). This species has been reported through the mid-central United States, the Great Lakes region, and scattered locations along the eastern United States (Fig. 32) with an Arkansas record recently reported (Chordas and

Kovarik, in press). Numerous specimens (102) were captured during this study and during every month from May through August. Although this bug was common in the pitfall traps in this study, the first author has not encountered it in sweep-net or beating collections from various locations throughout Ohio over the past decade.

Bugs captured with C. trimaculata were Isthmocoris piceus (Say, 1832)

(Geocoridae) and Pseudopachybrachius basalis (Dallas, 1852) (Rhyparochromidae).

Other bugs captured in different pitfall traps during the same collection period and in the same sampling plot were Amnestus spinifrons (Say, 1825) (Cydnidae), Blissus leucopterus (Say, 1832) (Blissidae), C. pulicaria, Mormidea lugens (Fabricius, 1775)

(Pentatomidae), M. serripes, P. fusca fusca, Phlegyas abbreviatus (Uhler, 1876)

(Pachygronthidae), P. obscuratus, and S. rusticus. Nearly half (48) of the 102 specimens

280

Figure 32. Cryphula trimaculata distribution north of Mexico. Canada: Ontario. United

States: Arkansas, Colorado, Connecticut, Florida, Illinois, Indiana, Iowa, Kansas,

Massachusetts, Missouri, New Jersey, New York, North Carolina, Ohio, South Carolina.

281

were captured in unfertilized plots with plant litter removed; the remaining specimens were evenly distributed among other treatments.

Stygnocoris rusticus (Fallen, 1807). A non-indigenous, Palearctic species recorded from the United States in the early 1900s (Wheeler 1983), it now is known from several northern states and across Canada (Fig. 33). It feeds on fallen seeds from composites (Asteraceae) and other plants (Sweet 1964, Wheeler 1992), an activity that increased the likelihood of capturing them in pitfall traps. Numerous specimens (51) were captured, and S. rusticus was encountered in pitfall traps each year from June through August.

Other bugs captured with S. rusticus during the same collection period and in the same pitfall traps were C. pulicaria, E. brevipennis brevipennis, I. piceus, M. serripes,

Neopamera bilobata (Say, 1832) (Rhyparochromidae), P. basalis, and P. obscuratus.

Other bugs captured with S. rusticus in other pitfall traps during the same collection period and same sampling plot were C. trimaculata, a Nabis nymph, and O. acuminatus.

Of the 51 specimens of S. rusticus, 42 were captured in sampling plots from which plant litter had been removed (23 and 19 specimens in unfertilized and fertilized plots, respectively). Given that S. rusticus feeds on fallen seeds of composites (Wheeler 1992), and virtually all forbs (including composites) had been eliminated from all fertilized plots

(Patrick et al. 2008a), it is interesting that nearly as many S. rusticus were encountered in fertilized plots (22 specimens) as unfertilized plots (29 specimens). Apparently, these bugs are utilizing various food sources.

282

Figure 33. Stygnocoris rusticus distribution north of Mexico. Canada: Alberta, British

Columbia, Nova Scotia, Ontario, Prince Edward Island, Quebec. United States:

Connecticut, Illinois, Maine, Michigan, New York, Ohio, Pennsylvania, Vermont, West

Virginia, Wisconsin, Washington.

283

Thyreocoridae

Corimelaena pulicaria (Germar, 1839). Although only recorded for Michigan and Pennsylvania bordering Ohio (Fig. 34), it is a common and widespread species that was expected for Ohio. My two specimens were captured during 1 - 15 July 2005. The only other bug captured with C. pulicaria in the same pitfall trap during the same collection period was S. rusticus. Other bugs captured with C. pulicaria in different pitfall traps during the same collection period and in the same sampling plot were P. obscuratus, C. trimaculata and M. serripes.

284

Figure 34. Corimelaena pulicaria distribution north of Mexico. Canada: Alberta,

British Columbia, Manitoba, Nova Scotia, Ontario, Quebec, Saskatchewan. United

States: Arkansas, California, Colorado, Connecticut, Florida, Georgia, Illinois, Iowa,

Kansas, Louisiana, Maine, Maryland, Massachusetts, Michigan, Minnesota, Mississippi,

Missouri, Nebraska, New Jersey, New York, Ohio, Oregon, Pennsylvania, Rhode Island,

South Dakota, Texas, Vermont, Virginia, Wisconsin.

285

References

Blatchley W. S. 1926. Heteroptera or true bugs of eastern North America, with especial

reference to the faunas of Indiana and Florida. Nature Publishing Company,

Indianapolis, IN, USA.

Chordas, S. W. III, H. W. Robison, E. G. Chapman, B. G. Crump, and P. W. Kovarik.

2005. Fifty-four state records of true bugs (Hemiptera: Heteroptera) from

Arkansas. Journal of the Arkansas Academy of Sciences 59:43–50.

Chordas, S. W. III, and P. W. Kovarik. (In press). Two Lygaeoidea (Hemiptera),

Ischnodemus slossonae and Cryphula trimaculata, new for Arkansas, U.S.A.

Journal of the Arkansas Academy of Sciences.

Henry, T. J., and R. C. Froeschner. 1988. Catalog of the Heteroptera, or true bugs, of

Canada and the continental United States. E.J. Brill, New York, NY, USA.

Hilsenhoff, W. L. 1986. Semiaquatic Hemiptera of Wisconsin. The Great Lakes

Entomologist 19(1):7–19.

Knight, H. H. 1941. The plant bugs, or Miridae, of Illinois. Illinois Natural History

Survey Bulletin 22:1–234.

Lauffer, M. B. 2004. The role of aquatic subsidies in wetland riparian food webs.

Thesis (M.S.), Kent State University.

Maw, H. E. L., R. G. Foottit, K. G. A. Hamilton, and G. G. E. Scudder. 2000. Checklist

of the Hemiptera of Canada and Alaska. NRC Research Press, Ottawa, Ontario,

Canada.

286

McPherson, J. E. 1982. The Pentatomoidea (Hemiptera) of Northeastern North America

with emphasis on the fauna of Illinois. Southern Illinois University Press,

Carbondale, IL, USA.

McPherson, J. E. 1992. The assassin bugs of Michigan. The Great Lakes Entomologist

25(1):25–31.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008a. Large-scale manipulation of

plant litter and fertilizer in a managed successional temperate grassland. Plant

Ecology 197(2):183–195.

Patrick, L. B., L. H. Fraser, and M. W. Kershner. 2008b. “Brown” world invertebrates

contradict “green” world biodiversity theory. Research Letters in Ecology. Vol.

2008, Article ID 694638, 4 pages, doi:10.1155/2008/694638.

Sweet, M. H. 1964. The biology and ecology of the Rhyparochrominae of New England

(Heteroptera: Lygaeidae). Part I, II. Entomological Americana 43:1–124, 44:1–

201.

Watson, S. A. 1928. The Miridae of Ohio. Ohio Biological Survey Bulletin 16:1–44.

Wheeler, A. G., Jr. 1983. Stygnocoris rusticus: new records in eastern North America

with a review of its distribution (Hemiptera-Heteroptera: Lygaeidae).

Entomology News 94(4):131–135.

Wheeler, A. G., Jr. 1992. Holarctic insects adventive in Michigan: New and additional

records (Homoptera, Heteroptera, Coleoptera, Neuroptera). The Great Lakes

Entomologist 25(2):99–106.

287

Williams, A. H. 2000. Wisconsin Cydnidae (Hemiptera: Heteroptera). The Great Lakes

Entomologist 33(2):161–164.