Development of Best-Practices for Conserving Beneficial Insects Within Iowa's Agricultural Landscape Kelly Ann Gill Iowa State University

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Digital Repository @ Iowa State University

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2013 Development of best-practices for conserving beneficial insects within Iowa's agricultural landscape Kelly Ann Gill Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Agriculture Commons, Ecology and Evolutionary Biology Commons, and the Natural Resources and Conservation Commons

Recommended Citation Gill, Kelly Ann, "Development of best-practices for conserving beneficial insects within Iowa's agricultural landscape" (2013). Graduate Theses and Dissertations. 13048. https://lib.dr.iastate.edu/etd/13048

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Development of best-practices for conserving beneficial insects within Iowa’s

agricultural landscape

Kelly Ann Gill

A thesis submitted to the graduate faculty in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE

Major: Entomology

Program of Study Committee:

Matthew O’Neal, Major Professor

Lisa Schulte Moore

Mary Harris

Iowa State University

Ames, Iowa

2013

ii TABLE OF CONTENTS

THESIS ABSTRACT v

CHAPTER 1. GENERAL INTRODUCTION AND 1 LITERATURE REVIEW Thesis Organization 1 Introduction and Literature Review 1 Ecosystem Services 1 Beneficial Insects and Their Value to Agriculture 2 Pollinators 3 Natural enemies 5 Resource Requirements for Beneficial Insects 6 Declining Biodiversity in Iowa 8 Where Can Biodiversity Exist in Iowa’s Agroecosystem? 9 Objectives and Hypotheses 12 Chapter Two 12 Chapter Three 12 References Cited 13

CHAPTER 2. QUALITY OVER QUANTITY: BUFFER STRIPS 18 CAN BE IMPROVED WITH SELECT NATIVE PLANT SPECIES Abstract 18 Introduction 19 Materials and Methods 23 Site Description 23 Experimental Design 23 Buffer Strips on Organic Farms 24 Simple Plant Communities 25 Diverse Plant Communities 25 The “CP-IA mixture” 26 The “MSU Best Bet mixture” 26 Forb-only Plant Communities 27 Plant Establishment 27 Switchgrass (Panicum virgatum L.) 27 Alfalfa (Medicago sativa L.) 28 Willow (Salix matsudana Koidzumi) 28 Corn (Zea mays L.) 28 Native perennial plant plugs 28 Field and Plot Maintenance 29 Plant Measurements 30 Arthropod Collection 32 Arthropod Identification and Guild Assignment 33 Arthropod Community Composition 34

iii

Statistical Analyses 36 Results 40 Plant Diversity 40 Flower Abundance 40 Arthropod Community Composition 41 Pollinator Diversity 42 Natural Enemy Diversity 44 Pollinator Abundance 46 Natural Enemy Abundance 47 Activity-density 48 Model Comparisons 48 Producer Perception of Existing Buffer Strips 49 Discussion 50 References Cited 58 Tables 64 Figure Legends 76 Figures 77

CHAPTER 3. DO BUFFER STRIPS CONTRIBUTE TO THE 80 BIODIVERSITY OF ORGANIC FARMS? Abstract 80 Introduction 81 Materials and Methods 85 Study Site Selection 85 Study Site Description 85 Arthropod Collection 89 Arthropod Identification and Guild Assignment 91 Arthropod Community Composition 92 Activity-density 93 Landscape Measurements 93 Statistical Analyses 94 Results 95 Arthropod Community Composition 95 Pollinator Community Composition 97 Natural Enemy Community Composition 99 Activity-density 102 Distance From Buffer Strip 103 Discussion 104 References Cited 112 Tables 117 Figure Legends 130 Figures 131

CHAPTER 4. GENERAL CONCLUSIONS 134 Chapter Two 134 Chapter Three 136 Recommendations 137

ACKNOWLEDGEMENTS 138

APPENDIX 140

THESIS ABSTRACT

The annual value of crop pollination and biological control of plant pests provided by beneficial insects is estimated to exceed $20 billion to United States crop production alone. Beneficial insects that supply these services to agricultural lands are threatened by limited or suboptimal resources due to the loss of biodiversity in agroecosystems, which is a growing concern in agricultural states like Iowa. Conservation practices are recommended to address a multitude of conservation concerns related to Iowa’s declining natural resources; however, guidelines for best practices that conserve beneficial insects are not well defined. Due to the valuable relationship of beneficial insects and successful crop production, there is a need for developing best practices that conserve beneficial insects within Iowa’s agricultural landscape. The first objective was to design mixtures of native perennial plants that range in diversity and resource availability and evaluate these different plant communities as candidates for buffer strips that attract and provide resources for beneficial insects. The second objective was to evaluate the insect community in non-crop buffer strips already established on organic farms and in the adjacent organic crops and conventional row crops. This research seeks to identify mixtures of native perennial plants optimized with resources attractive to pollinators and natural enemies and to determine if these mixtures can enhance Iowa’s buffer strips to conserve beneficial insects and protect their services. Best-practices for conserving beneficial insects can be adapted for different regions, land uses, and habitat restoration scenarios beyond the study system used for this research.

1 CHAPTER 1. GENERAL INTRODUCTION AND LITERATURE REVIEW

Thesis Organization

The scope of this research encompasses the development of best-management practices that can provide resources to beneficial insects within Iowa’s agricultural landscape. This thesis is organized in four chapters. Chapter one contains a literature review and an introduction that will provide background information and addresses the significance of the research presented in subsequent chapters. Chapter two will report how plant communities in buffer strips can be improved for conserving beneficial insects with mixtures of select native plant species. Chapter three describes the beneficial insect community in non-crop buffer strips already existing on organic farms, as well as the adjacent organic crops and conventional row crops. Chapter four will provide a brief summary of the conclusions from the research presented in this thesis, followed by acknowledgements. The appendix includes a table of the beneficial insect taxa collected across the candidate buffer strips (pertains to chapter two).

Introduction and Literature Review

Ecosystem Services.

Complex ecological processes through which ecosystems, and the species existing within, function to sustain and provision human life (Daily 1997, MEA 2005, Swinton et al. 2006, Zhang et al. 2007) are referred to as ecosystem services. These benefits to humans are realized through the maintenance of biodiversity and the production of

2 ecosystem goods and services (Daily 1997). The different types of goods and services ecosystems supply to humans are organized into four subcategories of services: provisioning, regulating, supporting, and cultural (MEA 2005). This research focuses on provisioning and regulating services, in agroecoysytems. Although supporting and cultural services have a role in agricultural systems, they are beyond the framework of the research presented in subsequent chapters and are not discussed in depth (see Daily

1997 for chapters regarding ecosystem services not included within).

Provisioning ecosystem services include the production of goods such as food, fuel, fiber, pharmaceuticals, and genetic resources. Ecosystems managed for agricultural production supply the majority of provisioning ecosystem services. However, managed landscapes depend on ecosystem services derived from natural ecosystems as inputs (e.g. regulating ecosystem services). These inputs determine the success and quality of provisioning ecosystem services and the goods obtained from agricultural production

(MEA 2003, Müller et al 2005, Zhang et al. 2007). In this way, agricultural landscapes both supply and demand ecosystem services.

Beneficial Insects and Their Value to Agriculture.

Beneficial insects provide regulating ecosystem services to agriculture such as pollination and the natural regulation of plant pests. These insect-derived services are critical to the ecological balance and economical profitability of agricultural production, and in turn, food security. Provisioning services, such as the production of food, fuel, and fiber provide humans with direct benefits by means of tangible goods that are of marketable value; however, it is more difficult to assign a value to the regulating services

(pollination and natural regulation of pests) that enable the production of such ecosystem goods. It is noteworthy to mention that this research aims to enhance insect-derived ecosystem services from a conservation perspective (i.e. enhancing beneficial insects in agricultural landscapes that provide ecosystem services to crops), but aspects related to insect-derived services were not directly measured. Best-practices to conserve beneficial insects must be evaluated for their ability to attract beneficial insects, and be neutral to pests, before the impact on the level of services provided through this type of conservation can be assessed. Despite these challenges, a growing interest in developing a framework for determining the value of insect-derived ecosystem services exists

(Swinton 2008). In later sections, estimates pertaining to the dollar-value of insect- derived ecosystem services that pollinators and natural enemies provide to crop production in the United States (US) are presented.

Pollinators. Insect pollinators, specifically, encompass many species of flower- visiting insects that forage on flowering plants to obtain plant-provided food (nectar, pollen). Flower-visiting insects have the potential to transfer male gametes (contained in pollen) to the female gametes while foraging, resulting in pollination, which can then lead to fertilization (fusion of male and female gametes). Insect-mediated pollination is an essential step in reproduction for the majority of the world’s flowering plants, including numerous cultivated plant species. Many crops depend on pollination for seed production and fruit set to achieve economically viable yields (Kevan et al. 1983, Klein et al 2007). Globally, an estimated 35% of crop production is a result of insect pollination

(Klein et al. 2007). The European honey bee (Apis meliffera L.) is responsible for the

4 majority of crop pollination services, and the annual value of pollination by A. meliffera estimated to be worth $14.6 billion to US crop production (Morse and Calderone 2000).

In addition to honey bees, there are over 4,000 species of social and solitary non-

Apis bees in North America and most of these are wild species that nest in the ground, hollow stems, or excavated cavities (Michener 2000). Non-Apis bees also are important pollinators of crops, especially for crops in which honey bees are inefficient pollinators

(e.g. alfalfa, squash). With recent declines in honey bee colonies, these species (wild or managed) are increasingly being recognized for their contribution to US crop production, which is estimated to be worth $3.1 billion annually (Losey and Vaughan 2006). A few non-Apis species are managed for crop pollination. Examples of managed non-Apis species include bumble bees, Bombus impatiens Cresson (Hymenoptera: Apidae) managed for cranberry (Vaccinium spp.) and greenhouse tomato (Solanum lycopersicum

L.) pollination, alfalfa leafcutting bees Megachile rotundata F. (Hymenoptera:

Megachilidae) managed for alfalfa (Medicago sativa L.) pollination, and Osmia spp.

(Hymenoptera: Megachilidae) managed for pollination of several tree fruit species

(Delaplane and Mayer 2000).

Although bees are considered the most effective insect-pollinator of most plant species, other insects have been recognized for their contributions to pollination. Flower- visiting flies (Diptera) have been documented as proficient pollinators of several crops including carrot (Dacus carota L.), mustard (Brassica spp.), leek, (Allium ampeloprasum

L.), and almond (Prunis dulcis (Mill.) D. A. Webb) (Courtney 2009, Klein et al. 2012).

The presence of a diverse assemblage of insect-pollinators (e.g. different combinations of honey bees, bumble bees, solitary non-Apis bees, and flies) have shown to be

5 complimentary to plant reproductive success when compared to the presence of a single, yet abundant, bee species (Fontaine et al. 2006, Greenleaf and Kremen 2006, Hoehn et al.

2008, Winfree et al. 2007, Thies et al. 2011, Klein et al. 2012).

Natural Enemies. Insect predators and parasitoids that attack and feed on other insects; particularly, insect pests of plants are considered natural enemies. Through this type of feeding, natural enemies contribute to a type of pest regulation referred to as natural biological control. Natural enemies responsible for 33% of the natural pest control in cultivated systems, which is estimated to be worth an additional $4.5 billion dollars annually to US agriculture (Hawkins et al. 1999, Losey and Vaughn 2006).

Predaceous natural enemies belong to several insect orders (~ 20) and are generally characterized as free-living, mobile, larger than their insect prey, and able consume several prey-items throughout their life cycle (DeBach and Rosen 1991). By contrast, parasitoids mainly belong to two orders (Hymenoptera and Diptera) and their host ranges are considered to be more specialized than that of predators (Obrycki and Strand 1996).

Free-living adult parasitoids seek out a host, and depending on the parasitoid species, parasitize different life stages of their host (i.e. egg, larva, pupa, adult). Parasitoids can lay an egg (solitary) or several eggs (gregarious) on or within their host and the immature parasitoid(s) feed on their host to complete development, kill their host, and emerge as free-living adult (DeBach and Rosen 1991). Despite the fact that natural enemies feed on insects, many require plant-provided food during certain stages, or throughout, their life cycle.

In agricultural landscapes, natural enemies have the potential to prevent crop pests from reaching economically damaging levels. Predators and parasitoids can suppress or

6 delay pest population growth by contributing to pest mortality, causing high pest pressure to be asynchronous with the crop growth stages that are most vulnerable to herbivore damage (DeBach and Rosen 1991). When diverse assemblages of natural enemies are present, pest control is thought to be more effective due to differing phenology that can lead to the pest being attacked by natural enemies throughout the growing season

(Cardinale et al. 2003). Beyond natural biological control, natural enemies can be manipulated as part of integrated pest management programs through the importation and establishment of exotic natural enemy species (classical biological control), direct manipulation of populations (augmentative biological control), and, more pertinent to this research, through manipulation of their environment (conservation biological control).

Resource Requirements for Beneficial Insects.

Biodiversity is essential to the complex ecological processes that result in ecosystem services. This includes biodiversity within beneficial insect communities and the biodiversity among resources that allow beneficial insects to fulfill their functional role in agricultural landscapes (Foley et al. 2005). Pollinators and natural enemies depend on plant-provided resources such as nectar, pollen, alternate prey, refuge and shelter, overwintering sites, and nesting materials (Westrich 1996, Menalled et al. 1999,

Elliott et al. 2002, Steffan-Dewenter et al. 2002, Landis et al. 2000, 2005; Ricketts et al.

2008, Klein et al 2007, Kremen et al. 2007, Williams and Kremen 2007, Zhang et al.

2007, Kwaiser and Hendrix 2008, Tscharntke et al. 2008, Wackers et al. 2008, Le Féon et al. 2011). The composition and configuration of plant communities providing these resources are important to population dynamics of beneficial insects, including plant

7 communities found within a cropped field, surrounding cultivated land, and in the larger surrounding landscape matrix (Landis et al. 2000).

Within a cropped field, some of these resources can be provided by cultivated plant species. For example, mass-flowering crops such as canola (Brassica napus L.) has shown to provide floral resources to pollinators (Westphal et al. 2003) and natural enemies (Bowie et al. 1999). Additionally, common pests of cultivated plants such as aphids (Hemiptera: Aphididae) provide insect prey and hosts for over 30 species of predators and eight species of parasitoids (Rutledge et al. 2004). Beyond serving as a prey, aphids excrete honeydew (sugary exudates) on crop plants. For aphidophagous predators such as lady beetles (Coleoptera: Coccinellidae), honeydew is a high energy food source frequently encountered in cultivated fields. Honeydew, as a food source, has shown to extend the survival, allow for modest reproduction, and intensify foraging activity of lady beetles in both larval and adult stages (Wackers et al. 2008, Lundgren

2009). Furthermore, nectar and honeydew have shown to be important nutritional resources necessary for survival and egg maturation in parasitoids (Vollhardt et al. 2010).

Although a multitude of resources for beneficial insects can exist within cropped fields, the ephemeral nature of cropping systems limits the availability of resources, both spatially and temporally, that beneficial insects need to persist in the long term (Landis et al. 2005). Moreover, agricultural systems may be unfavorable for beneficial insects even when resources are available due to disturbances from management regimes and environmental conditions. Particularly, unfavorable are the non-target effects on beneficial insect populations resulting from exposure to broad spectrum pesticides

(Harwood et al. 2009).

Declining biodiversity in Iowa.

Natural ecosystems of savanna and wetland, interspersed among vast expanses of tallgrass prairie, once accounted for the land cover of the US Midwest, including the state of Iowa. During the past 200 years, changes in land use have altered the composition and configuration of Iowa’s land cover (Samson and Knopf 1994, Smith 1998). As recently as the early to mid 1800’s, tallgrass prairie vegetation covered 80% of Iowa, but when

Euro-American settlers began to occupy Iowa, the area of cultivated land increased.

During early settlement, traditional farming practices included planting multiple, small fields to different crops (e.g. row crops, grains, hay, and grasses). Despite the loss of native biodiversity, the heterogeneous cropping systems contributed agricultural biodiversity to the landscape. Advances in technology during the 20th century led to the development of conventional farming practices and the decline of agricultural biodiversity (Tscharntke et al. 2005). This is evident in Iowa where assessments of landscape change show crop diversity declined as the number of large fields of row crop monocultures steadily increased (Brown and Schulte 2011).

The expansion of agriculture has allowed significant gains in agricultural productivity (e.g. increased crop yields) and Iowa is now one of the leading producers of corn (Zea mays L.) and soybean (Glycine max (L.) Merr.) in the US. Annual monocultures of corn and soybean row crops account for more than 75% of Iowa’s land cover (IDALS 2011), consequently, Iowa is currently in last place (ranked 50 out of the

50 states) for the amount natural vegetation (0.1%) present in a state (Klopatek et al.

1979, Samson and Knopf 1994, Smith 1998). The removal of native vegetation, loss of perennial cover, simplification of cropping systems, expansion of annual monocultures,

9 and continued use of external inputs have been repeatedly blamed for the depletion of biodiversity in many agricultural landscapes. It has been identified that losses in biodiversity can adversely effect ecological processes and ecosystem functions essential to the health of both natural and managed ecosystems (Daily 1997, Altieri 1999, Green et al. 2005, Tscharntke et al. 2005, Rusch 2010).

Where Can Biodiversity Exist in Iowa’s Agroecosystem?

Biodiversity can exist within cropping systems and in the land surrounding cultivated fields. Organic crop production can be one agricultural practice that promotes biodiversity at the farm level. According to the USDA National Organic Standards

Board (1995), organic agriculture is defined as “an ecological production management system that promotes and enhances biodiversity, biological cycles, and soil biological activity. It is based on minimal use of off-farm (external) inputs and on management practices that restore, maintain, or enhance ecological harmony…The primary goal of organic agriculture is to optimize the health and productivity of interdependent communities of soil life, plants, animals, and people.” There are many reports in the scientific literature that organic agriculture, and associated practices, may counteract or reduce the negative effects of conventional agriculture, including the loss of biodiversity

(Altieri 1999, Foley et al. 2005, Tscharntke et al. 2012). Although Iowa’s agricultural landscape is dominated by conventionally managed row crop monocultures, there is a growing trend in the number of certified organic farms in the state. Iowa currently ranks fifth in the nation in the number of certified organic farms (467) and 2011 sales of organic goods in the state reached $60.7 million (USDA NASS 2011). Many of these

10 organic farms have a diverse cropping system and grow several types of fruits, vegetables, and herbs.

Organic farmers typically use management techniques that garner both economic benefits (i.e. profitable crop yields) as well as ecological benefits (i.e. reduce environmental stressors for sustainable delivery of multiple ecosystem services).

Organically managed farms are not disturbance free, although limited in the types of pesticides permitted, the intensity of farming practices (tillage, crop rotation, etc.) is not insignificant (Kovach et al. 1992). In this way, organic farms contribute to some amount of ecological disturbance (e.g. harvesting vegetation) that can also be unfavorable to beneficial insect species sensitive to such disturbances. Although many benefits may be derived from organic practices at the farm level, organic farms in Iowa account for only a small area of the total farmland. Surrounding habitats and landscape-level influences may be controlling the population dynamics of species existing at these small scales.

Beneficial insects also require biodiversity outside of cultivated fields. Many studies have demonstrated that patches of non-crop vegetation within agricultural landscapes are particularly important for determining the ability of beneficial insect communities to persist near agricultural fields before, during, and after periods when insect-derived ecosystem services are valuable to crops (Landis et al. 2000, Coll and

Guershon 2002, Bianchi et al. 2006, Isaacs et al. 2009). Evidence shows beneficial insect abundance and diversity increases on sites with greater land complexity, due to the accessibility of food sources and habitat, when compared to less complex, monoculture landscapes (Zhang et al. 2007). However, in simplified landscapes like Iowa’s, these resource patches become less frequent and lose connectivity, which contributes to the

11 degradation of ecological infrastructure that permits the dispersal and colonization of beneficial insects (Marino and Landis 1996, Bianchi et al. 2006). It has been recognized that species unable to persist in such altered landscapes are often replaced with species able to thrive under the prevailing field conditions and the replacements are often invasive, pest species (Corbin and D’Antonio 2004). Long-term consequences include local extinction, loss of a species from a local area, and functional extinction, reduction of a species to the extent where it is no longer able to play a significant role in ecosystem function (MEA 2005). The current challenge is to minimize tradeoffs between meeting the food, fuel, and fiber demands of a growing population and negative impacts on the biodiversity that drives ecosystem services (Foley et al. 2005, Robertson and Swinton,

Sandhu et al. 2010).

In many agricultural landscapes, biodiversity must be reestablished through conservation practices such as habitat restoration. Financial and technical support is available to assist eligible farmers with many conservation practices through programs such as the Conservation Stewardship Program (CSP) and the Environmental Quality

Incentives Program (EQIP) (USDA NRCS 2012). Although these conservation programs address a multitude of conservation concerns related to Iowa’s declining natural resources, guidelines for best-practices that conserve beneficial insects are not well defined. Due to the valuable relationship of beneficial insects and successful crop production, there is a need for best-practices that conserve beneficial insects within

Iowa’s agricultural landscape.

Objectives and Hypotheses

Chapter Two.

Establish and evaluate plant communities that vary in plant diversity and resource availability as candidates for buffer strips that attract beneficial insects.

I hypothesize that the diversity and abundance of beneficial insects will be greatest in

diverse plant communities with continuous availability floral resources; intermediate

in plant communities reduced in plant species richness and floral resources; and,

lowest in in simple plant communities composed of single-species.

I hypothesize that diverse plant communities can be further optimized to attract

beneficial insects by mixing plant species that are individually attractive to multiple

groups of beneficial insects.

Chapter Three.

Determine the diversity, abundance, and activity of beneficial arthropods in buffer strips already existing on organic farm and compared to what is found in the adjacent land-use types.

I hypothesize that the abundance, diversity, and activity of beneficial insects will be

greatest in buffer strips, intermediate in organic vegetable crops, and lowest in

conventional row crops.

I hypothesize that the abundance, diversity, and activity of pollinators and natural

enemies within organic farms and row crops will decrease as distance from buffer

increases.

References Cited

Altieri, M. A. 1999. The ecological role of biodiversity in agroecosystems. Agr.. Ecosys. Environ. 74: 19-31.

Bianchi, F. J. J. A., C. J. H. Booij, and T. Tscharntke. 2006. Sustainable pest regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural pest control. Proc. R. Soc. Lond. Biol. 273: 1715-1727.

Bowie, M. H., G. M. Gurr, Z. Hossain, L. R. Baggen, and C. M. Frampton. 1999. Effects of distance from field edge on aphidophagous insects in a wheat crop and observations on trap-design and placement. Int. J. Pest Manage. 45: 69-73.

Brown, P. W. and L. A. Schulte. 2011. Agricultural landscape change (1937-2002) in three townships in Iowa, USA. Landscape Urban Plan. 100: 202-212.

Cardinale, B. J., C. T. Harvey, K. Gross, and A. R. Ives. 2003. Biodiversity and biocontrol: emergent impacts of a multi-enemy assemblage on pest suppression and crop yield in an agroecosystem. Ecol. Lett. 6: 857-865.

Coll, M. and M. Guershom. 2002. Omnivory in terrestrial arthropods: mixing plant and prey diets. Annu. Rev. Entomol. 47: 267-297.

Corbin, J. D. and C. M. D’Antonio. 2004. Competition between native perennial and exotic annual grasses: implications for an historical invasion. Ecology. 85: 1273-1283.

Courtney, G. W., T. Pape, J. H. Skevington, and B. J. Sinclair. 2009. Biodiversity of Diptera, pp. 185-222. In: R. Foottit and P. Adler (eds.), Insect biodiversity: science and society.. Blackwell publishing, Chichester, West Sussex, UK.

Daily, G. C. 1997. Nature’s services: societal dependence on natural ecosystems. Island Press, Washington, D. C. , USA.

DeBach, P. and D. Rosen. 1991. Biological control by natural enemies. 2nd ed. Cambridge University Press, New York, NY, USA.

Delaplane K. S. and D. F. Mayer. 2000. Crop pollination by bees. CABI Publishing, New York, NY, USA.

Elliott, N. C., R. W. Kieckhefer, G. J. Michels, and K. L. Giles. 2002. Predator abundance in alfalfa fields in relation to aphids, within-field vegetation, and landscape matrix. Environ. Entomol. 31: 253:260.

Foley, J. A., R. DeFries, G. P. Asner, C. Barford, G. Bonan, S. R. Carpenter, F. Stuart, M. T. Coe, G. C. Daily, H. K. Gibbs, J. H. Helkowski, T. Holloway, E. A. Howard, C. J. Kucharik, C. Monfreda, J. A. Patz, I. C. Prentice, N. Ramankutty, and P. K. Snyder. 2005. Global consequences of land use. Science. 309: 570-574.

Fontaine, C. I., J. Meriguet, and M. Loreau, 2006. Functional diversity of plant- pollinator interaction webs enhances the persistence of plant communities. PLoS Biol 4(1): e1. doi:10.1371/journal.pbio.0040001.

Green, R. E., S. J. Cornell, J. P. S. Sharlemann, and A. Balmford. 2005. Farming and the fate of wild nature. Science. 307:550-555.

Greenleaf, S. S., and C. Kremen. 2006. Wild bee species increase tomato production and respond differently to surrounding land use in Northern California. Biol. Conserv. 133: 81-87.

Hawkins, B.A., N. J. Mills, M. A. Jervis, and P. W. Price. 1999. Is the biological control of insects a natural phenomenon? Oikos 86: 493-506.

Hoehn, P. T., T. Tscharntke, J. M. Tylianakis, and I. Steffan-Dewenter. 2008. Functional group diversity of bee pollinators increases crop yield. Proc. R. Soc. Lond. Biol. 275: 2283-2291.

Isaacs, R., J. Tuell, A. K. Fiedler, M. M. Gardiner, and, D. A. Landis. 2009. Maximizing arthropod-mediated ecosystem services in agricultural landscapes: the role of native plants. Front Ecol. Environ. 7: 196-203.

Kevan, P. G. and H. G. Baker. 1983. Insects as flower visitors and pollinators. Annu. Rev. Entomol. 28: 407-453.

Klein, A. M., C. Brittian, S. D. Hendrix, R. Thorp, N. Williams, and C. Kremen. 2012. Wild pollination services to California almond rely on semi-natural habitat. J. Appl. Ecol. 49: 723-732.

Klein, A. M., B.Vaissière, J. H. Cane, I. Steffan-Dewenter, S. A. Cunningham, C. Kremen, and T. Tscharntke. 2007. Importance of crop pollinators in changing landscapes for world crops. Proc. R. Soc. Lond. Biol. 274: 303-313

Kovach, J., C. Petzoldt, J. Degni, and J. Tette. 1992. A method to measure the environmental impact of pesticides. New York’s Food and Life Sciences Bulletin, No. 139. Cornell University, Ithaca, NY.

Regetz, and T. H. Ricketts. 2007. Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecol. Letters. 10: 299-314.

Kwaiser, K. S., and S. D. Hendrix. 2008. Diversity and abundance of bees (Hymenoptera: Apiformes) in native and ruderal grasslands of agriculturally dominated landscapes. Agr. Ecosys. Environ. 124: 200-204.

Landis, D. A., F. D. Menalled, A. C. Costamagna, and T. K. Wilkinson. 2005. Manipulating plant resources to enhance beneficial arthropods in agricultural landscapes. Weed Sci. 53: 902-908.

Landis, D. A., S. D. Wratten, and G. M. Gurr, 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Annu. Rev. Entomol. 45: 175-201.

Losey J. E., and M. Vaughan. 2006. The economic value of ecological services provided by insects. Bioscience. 56: 311-323.

Le Féon, V., F. Burel, R. Chifflet, M. Henry, A. Ricroch, B. E. Vaissière, J. Baudry. 2011. Solitary bee abundance and species richness in dynamic agricultural landscapes. Agr. Ecosys. Environ. (In press).

Lundgren, J. L. 2009. Nutritional aspects of non-prey foods in the life histories of predaceous Coccinellidae. Biol. Control. 51: 294-305.

Menalled, F. D., P. C. Marino, S. H. Gage, and D. A. Landis. 1999. Does agricultural landscape structure affect parasitism and parasitoid diversity? Ecol. Appl. 9: 634-641.

Michener, C. D. 2000. The Bees of the world. John Hopkins University Press, Baltimore, Maryland USA.

Millennium Ecosystem Assessment (MA). 2005. Ecosystems and human well-being: Biodiversity synthesis. World Resources Institute, Washington, DC.

Morse, R. A. and N. W. Calderone. 2000. The value of honey bees as pollinators of U. S. crops in 2000. Bee Culture. 128: 15 pp insert.

Strand, M. R. and J. T. Obrycki. 1996. Host specificity of insect parasitoids and predators. BioScience. 46: 422-429.

Ricketts, T. H., J. Regetz, I. Steffan-Dewenter, S. A. Cunningham, C. Kremen, A. Bogdanski, B. Gemmill-Herren, S. S. Greenleaf, A. M. Klein, M. M. Mayfield, L. A. Morandin, A. Ochieng, and B. F. Viana. 2008. Landscape effects on crop pollination services: are there general patterns. Ecol. Lett. 11: 499-515.

Robertson, G. P. and S. M. Swinton. 2005 Reconciling agricultural productivity and environmental integrity: a grand challenge for agriculture. Front. Ecol. Environ. 3: 38- 46.

Rusch, A., M. Valantin-Morison, J. Sarthou, and J. Roger-Estrade. 2010. Biological control of insect pests in agroecosystems: effects of crop management, farming systems, and seminatural habitats at the landscape scale: a review, pp. 219-259. In D. L. Sparks (ed.), Advances in agronomy (1st ed., vol. 109). Academic Press, San Diego, CA, USA.

Rutledge, C. E., R. J. O’Neil, T. B. Fox and D. A. Landis. 2004. Soybean aphid predators and their use in integrated pest management, Ann. Entomol. Soc. Am. 97: 240-248.

Samson, F. and F. Knopf. 1994. Prairie conservation in North America. BioScience. 44: 418-421.

Sandhu, H. S., S. D. Wratten, R. Cullen. 2010. Organic agricultural and ecosystem services. Environ. Sci. and Policy 13: 1-7.

Smith, D. D. 1998. Iowa prairie: original extent and loss, preservation and recovery attempts. J. Iowa Acad. Sci. 105: 94-108.

Steffan-Dewenter, I., U. Munzenberg, C. Burger, C. Thies, and T. Tscharntke. 2002. Scale-dependent effects of landscape context on three pollinator guilds. Ecology. 83: 1421:1432.

Swinton, S. M., F. Lupi, G. P. Robertsom, and D. A. Landis. 2006. Ecosystem services from agriculture: looking beyond the usual suspects. Am. J. Agr. Econ. 5: 1160-1166.

Swinton, S. 2008. Reimaging farms as managed ecosystems. Choices. 23: 28-31.

Thies, C., S. Haenke, C. Scherber, J. Bengtsson, R. Bommarco, L. W. Clement, P. Ceryngier, C. Dennis, M. Emmerson, V. Gagic, V. Hawro, J. Liira, W. W. Weisser, C. Winqvist, and T. Tscharntke. 2011. The relationship between agricultural intensification and biological control: experimental tests across Europe. Ecol. Appl. 21: 2187-2196.

Tscharntke, T., A. M. Klein, A. Kruess, I. Steffan-Dewenter, and C. Thies. 2005. Landscape perspectives on agricultural intensification and biodiversity – ecosystem service management. Ecol. Lett. 8: 857-874.

Tscharntke, T., R. Bommarco, Y. Clough, T. O. Crist, D. Kleijn, T. A. Rand, J. M. Tylianakis, S. van Nouhuys, and S. Vidal. 2008. Conservation biological control and enemy diversity on a landscape scale. Biol. Control. 45: 238-253.

Tscharntke, T., Y. Clough, T. C. Wagner, L. Jackson, I. Perfecto, J. Vansermeer, A. Whitbread. 2012. Global food security, biodiversity conservation and the future if agricultural intensification. Biol. Conserv. 151: 53-59.

United States Department of Agriculture National Agricultural Statistics Service (USDA NASS). 2011. Certified organic production survey. Available online: http://www.nass.usda.gov/Surveys/Organic_Production_Survey. Last visited on 12 January 2013.

United States Department of Agriculture National Organic Standards Board (USDA NOSB). 1995. Available online: http://www.nal.usda.gov/afsic/pubs/ofp/ofp.shtml. Last visited on 12 January 2013.

USDA, NRCS (United States Department of Agriculture) The Natural Resources Conservation Service (NRCS), 2012. NRCS Conservation Programs. Available online at http://www.nrcs.usda.gov/wps/portal/nrcs/main/national/programs/. Last visited on 26 December 2012.

Volhardt, I. M. G., F. J. J. A. Bianchi, F. L. Wakers, C. Thies, and T. Tscharntke. 2010. Spatial distribution of flower vs. honeydew resources in cereal fields may affect aphid parasitism. Biol. Control. 53: 204-213.

Wackers, F. L., P. C. J. van Rijn, G. E. Heimpel. 2008. Honeydew as a food source for natural enemies: making the best of a bad meal, Biol. Control. 45: 176-184.

Westphal, C., I. Steffan-Dewenter, and T. Tscharntke. 2003. Mass flowering crops enhance pollinator densities at a landscape scale. Ecol. Lett. 6: 961-965.

Westrich, P. 1996. Habitat requirements of central European bees and the problems of partial habitats. In Matheson, Buchmann, O'Toole, Westrich, and Williams (eds.). The Conservation of Bees, pp. 1-16. Academic Press Inc., USA.

Winfree, R., N. M. Williams, J. Dushoff, and C. Kremen. 2007. Native bees provide insurance against ongoing honey bee losses. Ecol. Lett. 10: 1105-1113.

Williams, N. M. and C. Kremen. 2007. Resource distribution among habitats determine solitary bee offspring production in a mosaic landscape. Ecol. Appl. 17: 910- 921.

Zhang W., Ricketts T. H., C. Kremen, K. Carney, S. M. Swinton. 2007. Ecosystem services and dis-services to agriculture. Ecol. Econ. 64: 253-260

CHAPTER 2. QUALITY OVER QUANTITY: BUFFER STRIPS CAN BE

IMPROVED WITH SELECT NATIVE PLANT SPECIES

A paper submitted to Environmental Entomology

K. A. GILL, 1 R. COX 2 AND M. E. O’NEAL1

1Department of Entomology, Iowa State University, Ames, IA 50011

2 Centro Internacional del Mejoramiento de Maíz y Trigo (CIMMYT), Km. 45, Carretera,

Mexico-Veracruz, El Batán, Texcoco, Edo. de México, CP 56130 México

Abstract

Native plants attractive to beneficial insects may improve the value of buffer strips by contributing to local biodiversity and enhancing the delivery of insect-derived ecosystem services. To determine best management practices for designing buffer strips that conserve beneficial insects, nine plant communities were evaluated. We hypothesized that the diversity and abundance of beneficial insects will be: (1) greatest in diverse plant communities with continuous availability of floral resources; (2) intermediate in plant communities with reduced plant species richness and availability of floral resources; and, (3) lowest in plant communities composed of a single species.

More diverse and abundant beneficial insect communities were observed in diverse and forb-only plant communities compared to simple plant communities, especially in treatments composed of species attractive to beneficial insects compared to compositions recommended for prairie restoration. Model comparisons revealed significant positive relationships between the diversity and abundance of beneficial guilds with plant

community diversity and the number of flowers blooming among treatments. These

differences suggest: (1) plant communities that dominate existing buffer strips are not

optimal for conserving beneficial insects; (2) adding flowering perennial species

improves buffer strips as habitat for beneficial insects; (3) buffer strips can be further

optimized by intentionally combining the most attractive native species even at modest

levels of plant diversity; and, (4) plant mixtures recommended for conservation programs

and traditional prairie reconstruction may not contain the number or density of the most

attractive native species necessary to support beneficial insects from multiple guilds.

Keywords: habitat management, floral provisioning, multifunctional landscapes,

biodiversity-ecosystem function, pollination, biological control.

Introduction

The annual value of insect-derived ecosystem services to agriculture is estimated to be at least $22 billion in the United States (US) with $4.5 billion attributed to biological control of pests by natural enemies and $3.1 billion to pollination by native bees (Losey and Vaughn 2006). In addition to wild, native insects, the annual value of pollination services from the introduced European honey bee (Apis meliffera L.) to US agriculture is estimated at an additional $14.6 billion (Morse and Calderone 2000).

Plant-derived resources such as, nectar, pollen, alternative prey, refuge, nesting substrates, and overwintering sites surrounding cultivated land positively influence the diversity and abundance of many natural enemies (Elliot et al. 2002, Landis et al. 2005,

Zhang et al. 2007, Tscharntke et al. 2008, Wackers et al. 2008) and bees (Westrich 1996,

Steffan-Dewenter et al. 2002, Klein et al 2007, Kwaiser and Hendrix 2008, Le Féon et al.

2011). Patches of non-crop vegetation within agricultural landscapes can provision such resources, allowing beneficial insect communities to persist near agricultural fields before, during, and after periods when insect-derived ecosystem services are provided to annual crops (for reviews, see Landis et al. 2000, Bianchi et al. 2006, Isaacs et al. 2009).

Plant provided resources are exploited at varying times and levels across different beneficial insect guilds (pollinators, predators, parasitoids) making the season-long availability of non-crop vegetation an important component of agricultural landscapes.

The loss of biodiversity in simplified agricultural landscapes can result in the uneven spatial and temporal distribution of the resources beneficial insect populations require to survive, and to effectively deliver ecosystem services to surrounding crops

(Landis et al. 2005). The loss native biodiversity is evident across many portions of the

Midwest US, including Iowa. Historically (early to mid-1800’s), the state of Iowa was dominated (~ 79.5 %) by tallgrass prairie ecosystems, but during the last 150 years most of Iowa’s native vegetation has replaced by agricultural systems, and now less than 0.1% of native, prairie vegetation remains in patches across the state (Samson and Knopf 1994,

Smith 1998). Many of the perennial plant species found in prairies are attractive to beneficial insects (Fiedler and Landis 2007, Tuell et al. 2008, Frank et al. 2008), though variation exits among individual species. The extent to which prairies can contribute to the conservation of beneficial insects and their delivery of insect-derived ecosystem services likely depends upon the species composition of the plant community (Fiedler and Landis 2007a, 2007b; Tuell et al. 2008, Isaacs et al. 2009).

Native plants commonly found in prairies can be a valuable component for inclusion within managed landscapes if there is a desire to increase the delivery of ecosystem services. Buffer strips are areas of planted or naturally occurring vegetation typically recognized for their role in soil and water conservation practices (e.g. grass filter strips, riparian buffer plantings) (Clark and Reeder 2007). Although the inclusion of buffer strips can provide value to adjacent cropland, periods of high commodity prices may not justify the removal of land committed to crop production to establish non-crop habitat. Organic agriculture limits the use of synthetic inputs (pesticides and fertilizers); therefore, ecosystem services like the biological control of insect pests are of great value

(Zehnder et al. 2007). In contrast to the voluntary, incentive-based implementation of most conservation buffer strip practices, requirement § 205.202 (USDA NOP 2009) states that buffers are required of organic producers seeking certification from the United

States Department of Agriculture (USDA) when organically managed land is adjacent to land not under organic management. The primary purpose of this mandatory buffer zone is to mitigate the drift of pesticide and genetically modified pollen from adjoining land.

Given the evidence for both field and landscape-level influence on the provision of biological pest suppression and pollination, these services can potentially be improved by manipulating the composition of plant communities (Swinton et al. 2007) within the buffers required for organic certification. Guidelines that aim to maintain naturally occurring vegetation or reintegrate non-crop plants can vary by region, by cropping system, and differing conservation goals. The addition of perennial, flowering plants in conservation plantings (e.g. establishment of hedgerows and shrubs in field margins) increases biodiversity and can garner additional benefits compared to less diverse

22 plantings (e.g. grass or weedy compositions) (Pywell et al. 2005, Hannon and Sisk 2009,

Boutin et al. 2011). Compatibility of buffer strips with crop production is important, and considerations need to include monitoring the response of herbivorous insect species to ensure economically important pest species do not benefit from these resources

(Ambrosino et al. 2006, Lavandero et al. 2006, Fiedler and Landis 2007). Several plant compositions have been proposed for habitat conservation, but for many conservation programs (e.g. Conservation Prairie (CP) mixtures) the focus is on the management of general wildlife habitat, and not on the needs of beneficial insects.

Combinations of native plant species that are both locally adapted and attractive to beneficial insects may improve the value of buffer strips for both organic and conventional agriculture. Our goal is to determine best management practices for designing and establishing perennial, multi-species buffers that are compatible with agricultural landscapes, contribute to local biodiversity, and can lead to improved ecosystem function. In the following two-year study, we evaluated nine plant communities that varied in plant diversity and resource availability as candidates for buffer strips that conserve beneficial insects. We expect these habitats to vary in their ability to attract and conserve beneficial insects. Specifically, we tested the hypotheses that the diversity and abundance of beneficial insects will be: (1) greatest in diverse plant communities with a continuous availability of floral resources; (2) intermediate in plant communities reduced in plant species richness and availability of floral resources; and,

(3) lowest in in simple plant communities composed of single-species. In addition, we hypothesize that diverse plant communities can be further optimized to attract and

23 conserve beneficial insects by mixing plant species that are individually attractive to multiple groups of beneficial insects.

Materials and Methods

Site Description.

The study site was established at Iowa State University’s Field Extension

Education Laboratory (FEEL) located in Boone County, Iowa (42°00.318’N

93°47.272’W). The site is a 17 ha demonstration farm divided into multiple plots devoted to crop related research. Adjacent fields as well as the surrounding landscape were composed of corn and soybean crops. On 23 June 2009, 36 garden-style plots measuring 2 by 2 m were constructed using 5.08 by 15.24 cm pressure treated lumber.

These plots were distributed along a 55 by 24m bare-soil field in a grid formation of four blocks (oriented west to east) with nine plots per block.

Experimental Design.

Nine buffer treatments were designed intentionally to have plant communities that vary in floral abundance and complexity, plant species richness, and growth habits.

These buffer treatments represent both current options employed within certified organic farms as well as options recommended for enhancing the diversity and abundance of beneficial insects. Plots were randomly assigned to the nine buffer treatments with four replications using a randomized complete block design.. To describe species within all

treatments, plant nomenclature, taxonomy, and characteristics follow PLANTS database

(USDA, NRCS 2011).

Buffer Strips on Organic Farms.

To determine what plant compositions best represented those commonly used in

buffer strips in Iowa and to evaluate organic producer’s perception regarding the benefits

from established buffer strips, we surveyed organic producers within the state. In 2008,

we mailed surveys to all organic producers who were certified under the three most

common certifying agencies in Iowa: Midwest Organic Services Association, Inc.

(MOSA), Organic Crop Improvement Association (OCIA), and the Iowa Department of

Agriculture and Land Stewardship (IDALS). Producer names and contact information

were collected directly from the certifier (IDALS and MOSA) or through a database on

the certifier’s website (OCIA). A letter of introduction to the project was sent to all

producers on 16 April 2008, followed by another letter of explanation and a six-page

survey on 28 April 2008. The written survey was a mixture of multiple-choice, yes/no,

fill in the blank, and open-ended questions. Producers were given no monetary incentive

for participating in the written survey. The survey was constructed to address issues

related to the implementation, perception, and utility of buffer strips on organic farms.

Questions in the survey were based on four general areas: (1) buffer strip composition and management; (2) producer perception of buffer strips; (3) perception of relationship

between pest management and buffer strips; and, (4) potential of producers to alter buffer

strip practices.

From this survey, we determined typical plant compositions of buffers on organic farms in Iowa (Table 1). This informed our design of the simple buffer treatments included in the study and comparisons with more diverse buffers constructed from perennial, flowering plants.

Simple Plant Communities.

To test our hypotheses regarding the impact of buffers comprised of a single plant species on the diversity and abundance of beneficial insects, four buffer treatments were established (Table 2). Individual plant species in simple plant communities were selected using the following considerations: (1) crop and non-crop species identified as common plant types used in organic buffer strips based on results from our survey of organic producers in Iowa; (2) species with growth requirements adapted or compatible with local agricultural field conditions (e.g., full-sun, non-invasive), and; (3) species that have ecological and economic benefits in addition to potential beneficial insect conservation

(e.g., erosion control, crops harvested and sold as conventionally produced, and species that may be used or sold as forage).

Diverse Plant Communities.

We hypothesized that the diversity and abundance of beneficial insects will be greater in diverse plant communities with continuous availability of floral resources, compared to the simple compositions currently used in buffer strips. Diverse plant communities were composed with mixtures of grasses and forbs, and plant species were selected if they met the following criteria: (1) species native to the north-central region of the US; (2) species that, in combination, provide flowers throughout the growing season;

(3) species with low to moderate aggressive growth, and; (4) species that were commercially available in locally adapted genotypes (ecotypes). In addition, we hypothesized that diverse plant communities can be further optimized to attract and conserve beneficial insects by mixing plant species that are individually attractive to multiple groups of beneficial insects. Therefore, one of the two diverse buffer treatments was designed by incorporating plants attractive to beneficial insect into the plant community composition.

The “CP-IA mixture” We established a diverse buffer treatment based on recommendations from the Iowa Natural Resources Conservation Services (USDA,

NRCS 2010). The name CP-IA refers to the 14 species used to create the experimental mixture that were selected from a larger, commercially available conservation prairie

(CP) seed mixture recommended for prairie restorations in Iowa (IA) as well as habitat set-asides as part of the Conservation Reserve Program (CRP) (USDA NRCS 2010)

(Table 3). The species in this particular mixture are recommended for conservation

(Practice CP25) targeting the restoration of rare and declining habitats (i.e., Iowa’s native tallgrass prairie). Goals of this practice include increasing plant diversity and providing habitat and food to wildlife. While the description also indicates, in general, healthy prairie habitats can be a source of flowers for pollinating insects (USDA NRCS 2010), this mixture was not designed with the primary goal of increasing the diversity and abundance of beneficial insects; its ability to do so has not been tested.

The “MSU Best Bet mixture” A combination of plant species were selected to increase the diversity and abundance of beneficial insects using species identified by

Fiedler and Landis (2007) and Tuell et al. (2008) as being highly attractive to beneficial

27 insects. Fiedler and Landis (2007) and Tuell et al. (2008) evaluated plant species individually and certain species were considered highly attractive to natural enemies and bees; with relatively low attractiveness to pest species. Twelve species we selected to create the MSU Best Bet mixture (named due to the origin of the work conducted at

Michigan State University; details available at nativeplants.msu.edu) (Table 4).

Forb-only Plant Communities.

Three buffer treatments were established as forb-only compositions (Table 5) to assess the response of beneficial insects to plant communities with a reduction in plant species richness and resource availability, while still maintaining a community of plants attractive to beneficial insects. We hypothesized that beneficial insect diversity and abundance will be intermediate in plant communities reduced in plant species richness and seasonal availability of floral resources. The selection criteria for these species were consistent with (1-4) of the diverse plantings, with a focus on the most attractive forbs from the MSU Best Bet mixture. The most species rich of these plant communities, referred to as the “MSU5” mixture, contained five species of forbs which provided flowering resources from two or more species blooming throughout the growing season.

Treatments referred to as the “MSU3” and “MSU2” mixtures were systematic reductions of the MSU5 mixture designed by reducing the phenological overlap in of species in bloom.

Plant Establishment.

Switchgrass (Panicum virgatum L.) plugs (one-year old plants, Ion Exchange

Inc., Harpers Ferry, IA) were transplanted by hand on 21 April 2010. Plugs were planted

at a density of one plug per 929 sq cm resulting in 25 plants per replicate plot. Plugs

were positioned 15.24 cm from the plot borders on all sides and 30.48 cm spacing was

maintained between plants.

Alfalfa (Medicago sativa L.) seed was purchased locally (Brekke’s Town and

Country Store, Ames, IA) and sown on 9 April 2010. Seed was hand broadcast using the

standard rate of 8.16 to 9.07 kg per ha resulting in 0.009 kg (9.07 g) of seed per replicate

plot. Due to the small amount of seed being used, the seeds for each plot were weighed,

portioned, and combined with coarse sand to add bulk to the material to ensure an even

distribution when broadcasting.

Willow (Salix matsudana Koidzumi) cuttings were taken from established

willow stands (Small Potatoes Farm, Minburn, IA) in February of 2010. Once the root

mass adequately developed, the shrubs were obtained from the farm and transplanted on

21 April 2010. Willow shrubs ranging from 60.96 to 91.44 cm in height were planted at

a density of three shrubs per replicate in a triangle formation with 121.92 cm spacing

between each shrub.

Corn (Zea mays L.) seed (DEKALB® DKC 61-72 Roundup Ready® Corn) was

sown by hand on 7 May 2010 and 11 May 2011 with three rows per replicate plot, 15.24

cm plant spacing and 76.2 cm row spacing resulting in a density of 35 plants per replicate

plot. In 2010, some corn plants did not establish and were replanted on 1 June 2010.

Native perennial plant plugs (one-year old seedlings, Ion Exchange Inc.,

Harpers Ferry, IA) were used to establish diverse and forb-only plant communities. All plugs were transplanted by hand on 16 September 2009. These five plant communities

(MSU2, MSU3, MSU5, CP-IA, MSU Best Bet) were planted earlier than other treatments

29 to allow for the establishment time required by forb species. Plugs were planted at a density of 25 plugs per plot and individual species placement within plots was kept the same across replications to reduce within treatment variation among replicates. Plugs were positioned 15.24 cm from the plot borders on all sides and 30.48 cm spacing was maintained between plants. Among all perennial treatments, taller plants were placed to the north of low growing species to avoid shading.

Field and Plot Maintenance.

A 4 m distance was maintained between each plot in all directions to allow for mowing between plots. In late October 2009, all plots were mulched with clippings of clean oat straw to protect seedlings (plugs) from frost or animal damage. The straw was removed in early April 2010 before the establishment of single-plant treatments. Annual ryegrass (Lolium multiflorum Lam.) was sown as ground cover between plots on 24 May

2010 and mowed throughout the sampling period. Plots were not mulched with straw after the 2010 growing season, as a thatch layer from first year plant material was left in plots. In each plot, the thatch layer that accumulated was subsequently maintained with minimal disturbance from the 2010 growing season throughout the 2011 sampling period to provide nesting habitat for stem-nesting bee species. In both years, weedy species growing in the field between plots were removed by hand and plots were continuously hand-weeded throughout the growing season to maintain species composition with special attention to weed removal immediately before insect sampling.

Plant Measurements.

Several measurements were taken in each buffer treatment to describe the relationship between plant characteristics and beneficial insect communities, and to determine if specific features of the plant communities account for variation in beneficial insect diversity and abundance among the buffer treatments. Measurements include plant diversity, flower abundance, percent ground cover, and canopy height.

The number of plants and plant species were counted monthly in each plot.

Simpson’s Diversity Index (1/D) was calculated for each plot based on final measurements taken in August of 2010 and 2011 (Simpson 1949), to represent the maximum, end-of-season plant diversity and to account for the annual establishment of corn plants. For each year, resulting diversity values were summed among replicates and mean plant diversity was calculated per buffer treatment (n = 4). Simpson’s diversity indices were calculated using the “vegan” package version 2.0-1 in R version 2.14.1

(Oksanen et al. 2011, R Development Core Team 2011).

Buffer treatments were designed to achieve variation in the amount and timing of floral resources. To determine if we achieved this variation, the number of flowers and species blooming were counted two times per month, coinciding with arthropod collection. Flower abundance was measured for plots containing conspicuous flowers

(i.e. those with conspicuous petals and sepals); therefore, corn and switchgrass were not measured. Although, early blooming species may provide important floral resources at a time when others are scarce, the early bloom period of willows preceded the annual establishment of corn plants and, therefore, our sampling period, so willows also were excluded from floral measurements.

For buffer treatments containing plants with conspicuous flowers, flower

abundance was measured by counting the number of individual, open (blooming) flowers

on each plant. Individual flowers were defined as, flower heads for Asteraceae spp. and

Geraniaceae spp., umbels per cluster for Asclepiadaceae spp., solitary flowers for

Ranunculaceae spp., and spikes and racemes for Scrophulariaceae and Fabaceae spp. For each year, flower data were summed across the six sample dates among replicates and

mean flower abundance was calculated based on the total number of observations (n =

24) per buffer treatment.

Additional measurements of plant communities including plant height (cm) and

percent ground coverage were recorded monthly in each plot. Five random subsamples

per plot were taken to measure percent ground cover by tossing a 30.48 by 30.48 cm

quadrat into plots and visually estimating the proportion of ground covered by vegetation

within each quadrat. Percentages for each toss per plot were summed and averaged over

the total number of estimates recorded in each buffer treatment. The height (cm) of each

plant was measured to the tallest point. Mean canopy height was calculated as the sum of

all plant heights per plot over the total plant heights of each buffer treatment. Each year,

mean percent ground cover and mean canopy height were estimated for each buffer

treatment, based on final measurements taken in August. These measurements were used

to represent the maximum, end-of-season height and ground cover and to account for the

annual establishment of corn plants.

Arthropod Collection.

Arthropod (insect and spider) communities were sampled in each plot throughout the 2010 and 2011 growing season (June, July, and August). Two different trapping methods were deployed to account for multiple feeding guilds (Rebek et al. 2005, Schmidt et al. 2008) including an active sampling method (vacuum sampling) to account for insects residing on plant foliage and visiting flowers. A passive sampling method (yellow sticky traps) was used to account for species active at different times of the day and those sensitive to plant disturbance by the vacuum. As such, the sticky traps assess activity- density, whereas the vacuum estimates arthropod abundance on plants.

To describe the foliar dwelling, flower visiting, and more sedentary arthropod community we sampled with a vacuum using methods adapted from Fiedler and Landis

(2007). A fine mesh, white paint strainer was placed over the air intake on a gas-powered leaf blower (Troy-Bilt, Model# TB320BV) and vegetation in each plot was vacuumed for

30 s while moving continuously around each plot to contact the foliage and flowers on all sides. The mesh strainer with the sample was then removed, placed into a clear plastic, resalable bag. An unused, clean mesh strainer was used for sampling subsequent plots.

Vacuum sampling occurred during the first and third week of each month during the sampling period with no less than 12 days between sampling events. To ensure high insect activity and consistency among samples, vacuum sampling was restricted to mid- day during favorable weather conditions (warm, sunny days, no cloud cover < 30% and with wind gusts < 5 mph). After each sampling event, insects were transported to the lab and frozen until processed.

Unbaited yellow sticky trap (Pherocon AM®, GEMPLER’S, Madison, WI) were used to measure the activity-density of mobile insects. Traps were deployed for 5 d and plots were not vacuum sampled during this time period. One trap per plot was fastened to a wooden stake located in the center of each plot. The yellow sticky traps containing samples were then collected, placed into a clear plastic, resalable bag, transported back to the lab, and frozen for future identification.

Arthropod Identification and Guild Assignment.

When possible, insects were identified to species. Spiders were identified to order

(Araneae). When species identification could not be resolved, individuals were identified to the lowest taxonomic unit possible or organized into morphospecies, and given a unique identifier for reference and classification of duplicates. Following identification individuals were classified to guilds; herbivores, predators, parasitoids, pollinators, detritivores, fungivores, and “other” based on species accounts described in the identification keys and literature reviewed. The group referred to as “other” includes species with non-feeding adults, blood-feeders, and unresolved feeding habits. Insects occupying different guilds in different stages of their life cycles (e.g. herbivores, predators and parasitoids) were classified based on feeding behaviors of their immature stages. For this study, the pollinator guild was defined by insects associated with mellitophilous syndromes (Hymenoptera: Apoidea) therefore restricting this category to managed A. meliffera and wild, non-Apis bee species.

Arthropod Community Composition.

We initially described the species composition of the entire insect community among buffer treatments to determine differences in overall biodiversity compared to diversity within guilds. Further analyses focused on the diversity, abundance, and density of activity of beneficial guilds that provide either biological control (a combination of predators and parasitoids) or pollinator (bees).

To describe the diversity of beneficial insect communities in each buffer treatment, species richness was measured as the number of taxonomic units in each vacuum sample. For each year, the resulting values for each sample were pooled across the six sampling dates per replicate plot and seasonal abundance was calculated as the average number of unique species per treatment. Mean species richness was summarized separately for each beneficial guild for each plot. Data used to describe beneficial insect diversity for natural enemies and bees was limited to taxonomic units identified to species or classified as morphospecies. Spiders were only identified to order (Araneae) and as such not included in estimates of diversity. However, we did include spiders in all estimates of the abundance of natural enemies. Taxa with undetermined identifications were omitted from measures of diversity (species richness) due to their unresolved identifications, but were included in the analyses that tested differences in abundances across buffer treatments. Diversity indices of species richness were calculated using the

“vegan” package version 2.0-1 in R version 2.14.1 (Oksanen et al. 2011, R Development

Core Team 2011).

The number of individuals in each vacuum sample was used to describe the overall abundance of insects in each plot. For each year, the resulting values for each sample were summed across the six sampling dates among replicates and means were calculated based on the total number of observations (n=24) per buffer treatment. Mean abundance was summarized separately for herbivores, predators, parasitoids, pollinators, detritivores, and fungivores within each buffer treatment. All guilds were included to describe the proportion of the insect community comprised by each group. Analyses focused on pollinators (bees) and beneficial guilds that could be a source of biological control, including parasitoids and predatory insects; spiders were also included as predators.

To describe the activity-density for mobile natural enemies among buffer treatments, each buffer treatment was sampled with a yellow sticky trap. For each year, the resulting values for each sample were summed across the six sampling dates among replicates and means were calculated based on the total number of observations (n=24) per buffer treatment. Means were summarized separately for a subset of natural enemies accounting for taxa that were more abundant on yellow sticky trap samples than vacuum samples based on the assumption that some more mobile taxa were able to escape active sampling (vacuum) and were underrepresented in vacuum samples. These data included the abundance of individuals in the following orders and families, Diptera; Syrphidae,

Dolichopodidae, Empididae, and Tachinidae; Coleoptera; Coccinellidae, and Neuroptera:

Chrysopidae and Hemerobiidae.

Statistical Analyses.

A paired t-test was used to test for differences between expected diversity and observed diversity (Simpson’s Diversity Index 1/D) of plant communities within a buffer treatment. For expected diversity, 1/D was calculated as if all species in each plot established as planned. This was compared to observed diversity, 1/D (calculated for each plot based on species that actually established (PROC TTEST, SAS software version 9.2 SAS Institute 2008, 1/D calculated using the “vegan” package version 2.0-1 in R version 2.14.1, Oksanen et al. 2011, R Development Core Team 2011). To test for variation in the means for response variables, observed plant diversity, flower abundance, percent ground cover, and canopy height among different buffer treatments, general linear model analysis of variance (ANOVA) was used. This model included treatment

(nine buffer treatments) and block (four replicate plots) as fixed effects. For flower abundance, the model describe above also included time (six sampling dates) as a fixed effect. When differences in plant measurements data were indicative of experimental treatment differences, a post hoc mean comparisons test was performed using least significant differences (LSD), Student-Newman-Kuels (SNK) procedure (α = 0.05)

(PROC GLM, SAS software version 9.2 SAS Institute 2008). Plant measurement data were analyzed separately for 2010 and 2011 to account for variation between years.

Results of analyses pertaining to flower abundance, percent ground cover, and canopy height reported separately by year, but values for plant diversity across treatments did not vary between years and are only reported once (using 2010 data) to represent both 2010 and 2011.

Multiple hypotheses related to the relationship between plant communities (i.e. the nine buffer treatments) and insect diversity and abundance were tested, and several of these hypotheses employed a subset of treatments. All buffer treatments were included to test the null hypothesis that the insect community (both diversity and abundance) did not vary among the nine different buffer treatments. In all procedures described below, data were analyzed separately for 2010 and 2011 to account for variation between years.

To test the null hypothesis that there was no difference in the community composition of insects across buffer treatments, we computed the Bray-Curtis dissimilarity matrix and differences among treatments were visualized by using non- metric multidimensional scaling (nMDS). This exploratory analysis was performed separately for 2010 and 2011 using a random starting configuration and the ordinations for each year were plotted in two dimensions with the final configurations being reached after two iterations for 2010 data and seven iterations for 2011 data. Stress for the final solutions was 0.02 (2%) and 0.067 (6.7%) for 2010 and 2011, respectively, which is considered ideal for ecological (species abundances) data (Clarke 1993, McCune and

Grace 2002). The closer the points (i.e. treatments) are in multidimensional space the more similar the diversity of the insect community. Multiple-response permutation procedure (MRPP) was then performed based on the Bray-Curtis dissimilarity matrix for each year to determine if the variation in overall insect community differed significantly across buffer treatments by quantifying the difference in species composition using 1,000 permutations. In addition we fitted the abundance of species within each guild as regressed arrow vectors to explain the correlation between the fitted variables and the ordination (arrow length indicates amount of proportional correlation) and arrows point

38 in the direction of the most rapid change of increasing abundance for species belonging to each group. The “vegan” package version 2.0-1 in R version 2.14.1 (Oksanen et al. 2011,

R Development Core Team 2011) was used to conduct nMDS, MRPP, and vector fitting procedures.

To test the null hypothesis that there was no difference in the diversity of beneficial insects across buffer treatments a general linear model ANOVA was used to analyze diversity data (species richness). This model included treatment (nine buffer treatments) and block (four replicate plots) as fixed effects. When differences in diversity data were indicative of experimental treatment effects, a post hoc mean comparisons test was performed with the LSD SNK procedure (α = 0.05) to identify differences in the number of beneficial insect species for each guild (α = 0.05) (PROC

GLM, SAS software version 9.2 SAS Institute 2008).

To test the null hypothesis that there was no difference in the abundance of beneficial insects among plant community types, general linear ANOVA model was used with fixed effects of treatment (nine buffer treatments), block (four replicate plots), time

(six sampling events), and time and time by treatment interactions. Random effects included the block by treatment interaction (i.e. plots). (PROC GLM, SAS software version 9.2 SAS Institute 2008). Analyses focused on natural enemies and bee- pollinators according to our hypotheses; however, herbivores were also included in an analysis of abundance data to determine if buffer treatments vary in their attractiveness to herbivores, particularly pest species. When differences in abundance data were indicative of treatment effects, a post hoc mean comparisons test was performed data using the LSD SNK procedure (α = 0.05) to identify which treatments have significant

effects on mean beneficial insect abundances for each guild (PROC GLM, SAS software

version 9.2 SAS Institute 2008).

We determined which plant characteristics explained the most variation in both

the diversity and abundance of natural enemies and bees (collected with a vacuum) using

multiple linear regression analysis and Akaike’s Information Criterion for model

selection, adjusted for sample size (AICc; Burnham and Anderson 2002). For each year, explanatory variables examined were plant diversity (Simpson’s Diversity 1/D), flower abundance, percent ground cover, and canopy height. Response variables in the models included diversity and abundance for the natural enemy and bees. We report the “best-fit model” (the model with the minimum AICc value) and “competing models” (any model for the same response variable having an AICc value with a difference less than two is

considered strongly supported) (Burnham and Anderson 2002). Models with differences

in AICc values greater than two (compared to the best-fit model) were considered to weakly support these data and are not shown. For each best-fit and competing model we present the response variable, number of variables in the model (K), minimum AICc

value, differences (ΔAICc), Akaike weights as an estimate of the relative likelihood of a

2 given model against all other models (ωi), the adjusted R values, and a list of

explanatory variables associated with the model. Model selection was performed using

the “AICcmodavg” package version 1.24 in R version 2.14.1 (Mazerolle 2012, R

Development Core Team 2011).

Results

Plant Diversity.

There was no evidence of a significant difference in the expected and observed mean plant diversity (Simpson’s Diversity 1/D) within our established buffer treatments

(P > 0.5). However, a few replicates had a slightly lower observed diversity than expected diversity due to a few plant species that did not establish. Plants that did not establish (treatment, replicates) include Canada wildrye (in the CP-IA and MSU Best Bet, rep 1), Indiangrass (CP-IA, rep 1 and 2), meadow zizia (MSU Best Bet, rep 3), New

England Aster (MSU3, rep 2), prairie ironweed (MSU Best Bet, rep 2), and sideoats grama, (CP-IA, rep 4). As expected, we observed a significant difference in plant diversity among the nine buffer treatments (F = 40.16; df = 8,35; P < 0.0001) (Table 6).

Flower Abundance.

The number of flowers among all treatments with conspicuous flowers increased from 2010 to 2011. The abundance of floral resources available per vacuum sampling event varied significantly across buffer treatments in both years (2010: F = 13.61; df =

8,215; P < 0.0001, 2011: F = 12.86; df = 8,215; P < 0.0001) (Table 6). The MSU Best

Bet, MSU3, and MSU2 had significantly more flowers in bloom per sampling event compared to the MSU5 and CP-IA mixtures and alfalfa (Table 6). There was significant differences in canopy height and percent ground cover among buffer treatments in 2010

(canopy height: F = 14.47; df = 1,35; P < 0.0001, ground cover: F = 41.70; df = 1,35; P

< 0.0001) and 2011 (canopy height: F = 14.04; df = 1,35; P < 0.0001, ground cover: F =

41.08; df = 1,35; P < 0.0001). Despite these differences, these characteristics were not

significant factors in our subsequent regression analysis (see results in section below).

Arthropod Community Composition.

Using the vacuum, we collected a total of 14,632 in 2010 and 22,261 insects in

2011. Samples collected in 2010 were primarily composed of herbivores (59%), followed by beneficial insects comprised of predators, parasitoids, and bees (28% pooled). Detritivores, fungivores, and “other” accounted for the remaining 13% of the total insect community. Herbivores remained the dominant guild in 2011 samples, accounting for 73% of the vacuum collected insects and the proportion of beneficial groups decreased to (17%) relative to the total. Detritivores, fungivores, and “other” accounted for the remaining 10% of the total insect community.

In both years, alfalfa plots experienced an outbreak of the potato leafhopper,

Empoasca fabae Harris (Hemiptera: Cicadellidae). This single species was the most common herbivore in both 2010 (22% of all herbivores and 13% of the total community) and 2011 (47% of all herbivores and 35% of the total community). Leafhopper abundance in our experiment may not represent how outbreaks of E. fabae are typically in an on-farm setting, i.e. cutting of infested alfalfa. We did not mow the alfalfa plots during this experiment, allowing the populations of E. fabae to persist. Other than E. fabae, no other economic pests were observed. When E. fabae is omitted, the recalculated ratios for each guild relative to the total are more analogous between years

(2010: herbivores (53%) and beneficial groups (33%), 2011: herbivores (60%) and beneficial groups (24%)).

Insect communities in vacuum samples varied significantly by year (estimated by

LS-means) for natural enemies (P < 0.0001), bees (P < 0.0001), and herbivores (P =

0.0001). Due to the variation in insect abundance per vacuum sample by year, data were

analyzed separately for 2010 and 2011. We used nMDS plots based on Bray-Curtis

dissimilarities to show the configuration of treatments superimposed into “species space,”

and for each year the nMDS plots reveal differences in species compositions suggesting

insect communities are different across buffer treatments (Figure 1). The variation in

species composition was significantly different across treatments in 2010 (P < 0.0001)

and 2011 (P = 0.0009) based on the MRPP analysis. Furthermore, for both years, vector

arrows indicate treatments that beneficial species are significantly correlated with, and

their abundances consistently increase in the directions of the MSU Best Bet, MSU5, and

MSU3 mixtures. The communities in single-species treatments, especially corn and

willow, have the least number of shared species in comparison to all other treatments.

Pollinator Diversity.

Over the course of the study, we observed a pollinator (i.e. mellitophilous

hymenoptera) community composed of 24 taxonomic units representing five families

(Table 7). Bee communities were consistently more diverse in buffer treatments with

multiple plant species (diverse and forb-only mixtures). On a per plot basis, we found two to three times more taxa of bees in diverse and forb-only treatments than the number

of bee taxa observed in plots composed of only a single plant species. Furthermore, in

the multiple-species treatments, several bee species were collected in consecutive years,

whereas we did not observe any reoccurring species in the single-species treatments. The

43 bee community was most diverse (18 taxonomic units) in the MSU Best Bet mixture compared to all other buffer treatments. Among CP-IA and all the forb-only mixtures

(MSU5, MSU3 and MSU2), bee communities were equally diverse, varying only by one species.

During each year, we observed significant differences in bee diversity among the treatments. The mean number of species vacuum collected per plot varied significantly across buffer treatments in 2010 (F = 6.25; df = 8, 35; P = 0.0002) and 2011 (F = 5.73; df

= 8,35; P = 0.0004). Bee diversity was lowest in the treatments comprised of only one species; no bees were captured during 2010 in the corn treatments and in 2011 in either willow or switchgrass treatments. In 2010 we observed the most bee species per plot within the MSU Best bet mix; in 2011 we did not observe significant differences among any of the treatments with multiple plant species (Table 7) (See appendix, Table 1. for the presence and absence of all bee taxa across treatments).

The majority of the bees collected among the buffer treatments were species native to North America. Exceptions include only a few introduced species such as the domesticated honey bee (A. meliffera) and the semi-domesticated alfalfa leafcutting bee,

Megachile rotundata F., both found only in 2011. M. rotundata was observed only in the

MSU Best Bet mixture while A. meliffera was observed in MSU5, MSU Best Bet, and

CP-IA mixtures. The majority of the taxa (79%) we captured were ground nesting bees, but at least one cavity nesting (includes stem, wood, and preexisting holes) species was represented in all families except for in Andrenidae. The ground nesting species we observed exhibit different levels of sociality ranging from annual eusocial, Bombus spp., communal (Agapostemon spp.), solitary (Melissodes spp.), and variations thereof

(Halictidae spp.), whereas the cavity nesting species we observed are all solitary nesters

(Packer et al. 2007). With few exceptions, most of the species collected among buffer treatments are considered common or locally abundant in our region (Michener 2000,

Packer et al. 2007; see also http://www.discoverlife.org [Ascher and Pickering 2012] for geographic distribution maps).

Natural Enemy Diversity.

Over the course of the study, we observed a natural enemy community (predators and parasitoids) composed of 87 taxonomic units representing 41 families (Table 7).

Overall, natural enemy communities were more diverse in buffers with multiple plant species, specifically, the MSU5, MSU Best Bet, CP-IA mixtures. On a per plot basis, we found two to four times the mean number of taxa observed in plots with multiple plant species than those composed of only a single plant species. Like bees, natural enemy taxa were more commonly shared between years within diverse and forb-only treatments compared to simple treatments. The natural enemy community was most diverse (59 taxonomic units) in the MSU5 mixture, however several other treatments (excluding corn and willow) were equally species rich, varying by ten or fewer species (Table 7).

During each year, we observed significant differences in the diversity of natural enemies among the treatments. The mean number of species per plot varied significantly across buffer treatments in 2010 (F = 10.22; df = 8, 35; P < 0.0001) and 2011 (F = 8.35; df = 8,35; P < 0.0001). In 2010, we observed significantly fewer natural enemy taxa within corn and willow treatments, but no significant differences were observed among the remaining treatments. In 2011, we observed significantly more natural enemy taxa

within alfalfa and the MSU Best Bet and MSU5 mixtures compared to the other

treatments (Table 7). During 2010, parasitoids accounted for a greater proportion of

natural enemy taxa compared to predators (60 and 40% respectively), and in 2011

predators were slightly more dominant than parasitoids (51 and 49% respectively) (See

appendix, Table 1, for the presence and absence of all natural enemy taxa across

treatments).

With few exceptions, the natural enemies collected among buffer treatments are

considered widely distributed and common across our region. Most of the natural enemy

species captured were generalists such as Orius insidiosus (Say) (Hemiptera:

Anthocoridae), Nabis spp. (Hemiptera: Nabidae), Harmonia axyridis (Pallas)

(Coleoptera: Coccinellidae), Tachinid spp. (Diptera: Tachinidae), and Pteromalid spp.

(Hymenoptera: Pteromalidae); however, some are omnivorous, and many may

supplement their diet with plant-derived foods, such as pollen and nectar (e.g.

Coleomegilla maculata [DeGeer], Syrphidae spp., free-living adult parasitoids)

(Triplehorn and Johnson 2005). These taxa mentioned above have been considered

biological control agents of agronomic insect pests including the soybean aphid (Aphis

glycines Matsumura) (Rutledge et al. 2004, Costamagna et al. 2008), potato leafhopper

(E. fabae) (Östman and Ives 2003, Wiser Erlandson and Obrycki 2010), and European

corn borer (Ostrinia nubilalis (Say)) (Musser and Shelton 2003). In several studies, the

same natural enemies mentioned above show a positive association with plant community diversity and flowering plants used in beneficial insectary plantings or maintained in field margins (Colley and Luna 2000, Harmon et al. 2000, Fiedler and Landis 2007b,

Lundgren et al. 2009, Lundgren 2009, Al-Dobai et al. 2012).

Pollinator Abundance.

Over both years, we observed a pollinator (i.e. mellitophilous hymenoptera) community comprised of 325 bees. During each year, we observed significant differences in the abundance of bees among treatments. The mean number of bees per vacuum sample varied significantly across buffer treatments in 2010 (F = 6.47; df =

8,215; P < 0.0001) and 2011 (F = 4.33; df = 8,215; P < 0.0001). During 2010 we did not capture a single bee in the corn treatments; in 2011 we did not capture a single bee in either willow or switchgrass treatments. Bees were consistently more abundant in buffer treatments with multiple plant species, from which we captured three to four times the mean number of individuals than in simple, single-species treatments on a per sample basis. Furthermore, across multiple species treatments, noticeably more bees were found within MSU forb-only and Best Bet treatments compared to the CP-IA mixture (Fig. 2).

Overall, the most bees (75) were collected from the MSU Best Bet mixture.

Lasioglossum spp. (Hymenoptera: Halictidae) in the Dialictus subgenus was the most abundant bees collected in 2010. This group was present in all treatments except for alfalfa. The greatest number of Lasioglossum species (24) was observed in the MSU

Best Bet mixtures and MSU3 mixtures accounting for 44 and 50% of the total bee abundance in these buffer treatments, respectively. In contrast, Bombus griseocollis

(DeGeer) (Hymenoptera: Apidae) were the most abundant bees collected in 2011. B. griseocollis were only present in samples from the MSU Best Bet and forb-only buffer treatments, but the greatest number of individuals (14) was observed in the MSU2 mixture accounting for 46% of the total bee abundance in the MSU2 buffer treatment.

Natural Enemy Abundance.

During the course of our study, we collected 7,520 natural enemies (predators and parasitoids). Among the natural enemies, predators accounted for a greater portion of the community abundance than parasitoids (57 and 43% respectively). Natural enemies were consistently more abundant in the multiple plant species and alfalfa from which we collected two to 10 times the number of natural enemies per sample than the remaining single-species treatments. The greatest number of natural enemies (1602) was observed in the MSU Best Bet mixture. The mean number of natural enemies per vacuum sample varied significantly across buffer treatments in 2010 (F = 9.15; df = 8,215; P = 0.0008) and 2011(F = 8.79; df = 8,215; P < 0.0001). In 2010 we observed significantly more natural enemies per sample within the MSU Best bet mix compared to all other treatments; in 2011 we did not observe significant differences among the treatments with multiple plant species and alfalfa (Fig. 3).

In both years, O. insidiosus was the most abundant predator and was present in samples from all buffer treatments, but the greatest number of individuals (523) was observed in the MSU Best Bet mixture accounting for 36% of the total natural enemy community in this buffer treatment. Pteromalids were the most abundant parasitoid family and were present in samples from all buffer treatments, but the greatest number of individuals (384) was observed in the MSU5 mixture comprising 34% of the total natural enemy community in this treatment.

Activity-density.

In 2010, we collected a total of 4,701 natural enemies with yellow sticky traps and parasitoids accounted for a greater proportion of natural enemy abundance than predators

(46 and 53% respectively). In 2011, fewer natural enemies were observed (1654), but the proportion of parasitoids and predators was similar to 2010 (55 and 44% respectively).

However, in both years, certain taxa were significantly more abundant on yellow sticky traps compared to those collected with the vacuum (P < 0.05, data not shown), including species of the following: Diptera: Syrphidae, Dolichopodidae, Empididae, and

Tachinidae; Coleoptera: Coccinellidae, and Neuroptera: Chrysopidae and Hemerobiidae.

Despite the differences between sampling methods, we did not observe significant differences in the diversity or abundance of the aforementioned natural enemies among the buffer treatments in either year (analysis not shown).

Model Comparisons.

Beneficial insect diversity and abundance exhibited positive relationships with several of the plant characteristics measured among buffer treatments. All best fit and competing models were significant (P < 0.05). During 2010, we observed a significant positive relationship between bee species richness and plant diversity and the number of flowers in bloom (Table 8). In 2010, there was evidence for a competing model for explaining bee abundance. In 2011, the variables in the best-fit models for both bee species richness and abundance were reduced to a significant positive relationship with the number of flowers, and no competing models. During both years we observed a significant positive relationship between the species richness and abundance of natural

enemies and plant diversity, the number of flowers in bloom, and ground cover in the

best-fit models; in 2011, there was evidence for a competing model for natural enemy

abundance. In addition the to positive relationships, we observed a significant negative

relationship between natural enemy species richness and canopy height for the best-fit

model in 2010 and competing model in 2011.

Producer Perception of Existing Buffer Strips.

In our survey of Iowa’s organic producers, the most commonly identified service

provided by buffers was that the buffers fulfill the requirement needed for organic

certifications (Table 9). Beyond this, small percentages of row and horticulture crop

producers identified one or more ecosystem service. Within the category of ecosystem

services, an insect-related comment was made by only one producer. In addition, the

general importance of buffers to producers as a part of organic standards was surveyed,

specifically asking producers, “Do you think buffers should continue to be a requirement

in the USDA Organic Standards?” A majority of both row crop (83.7%) and horticulture

crop (72.7%) farmers responded “yes”.

Producer perceptions of buffer strips were determined during interviews by asking

about the benefits provided and harm perceived by farmers to be caused by buffers (Table

10). Compared to the survey data that showed “certification” as the most frequently

mentioned benefit of buffers, the benefit most commonly mentioned in interviews among

all producers was the category “ecosystem services” – which in interviews included erosion control, wildlife, water filtration, beneficial insects, and carbon sequestration.

Producer perceptions of pest management and its relation to buffer strip usage were assessed by asking two questions: “How have pest management issues affected your usage of buffer strips?” and “Do you feel your buffer strip has any effect on pest management on your farm?” In both questions, the majority of producers responded negatively, suggesting they do not draw connections between pest management and buffer strips or make decisions based on this connection (Table 11 and 12).

Discussion

Over the course of our two-year study, we compared insect communities across nine treatments composed of different plant communities that are used as buffers on organic farms in Iowa or are candidates for buffer strips that optimize insect-derived ecosystem services. We tested the hypotheses that the diversity and abundance of beneficial insects will be: (1) greatest in diverse plant communities with continuous availability of floral resources; (2) intermediary in plant communities reduced in species richness and availability of floral resources; and, (3) lowest in simple plant communities composed of single-species. In addition, we proposed that optimizing buffer strips with native plants attractive to multiple guilds and species would be the “Best Bet” for achieving long-term benefits.

Overall, our results suggest that: (1) plant communities that currently dominate the buffer strips on organic farms (at least within Iowa) may not be optimal for conserving beneficial insects; (2) the addition of flowering perennial species can improve buffer strips as habitats for beneficial insects; (3) native perennial plant communities can

51 be further optimized by intentionally selecting combinations of the most attractive native species even at modest levels of plant diversity; and, (4) plant mixtures recommended for conservation programs and traditional prairie reconstruction may not contain the number of the most attractive native species at densities necessary to attract and conserve beneficial insects from multiple guilds.

We hypothesized that beneficial insect diversity and abundance would be limited in simple (single-species) plant communities compared to moderately diverse (forb-only) and diverse (mixed forb and grass) plant communities. Our results agree with our expectations; however, we also observed variation in beneficial insect communities among simple plant communities. This indicates that some single-species compositions may be more suitable habitats than other single-species compositions. Among the simple plant communities, bee and natural enemy communities were more diverse and abundant in perennial plant communities compared to annual plantings of corn, suggesting that perennial buffer strips may be more hospitable refuges for beneficial insects than ephemeral plant communities. Beneficial insect abundance was similar in willow, an introduced perennial shrub, and corn. Although the early (spring) flowering period of willow may be an important supply of food resources for insects active in early spring, when other floral resources are scarce, we did not sample during this period. Willow lacks characteristics (e.g., season-long floral resources, vegetative ground cover) our analyses show to be important to beneficial insect communities. This further indicates that even though the perennial plants may have advantages over annual plants, certain plant characteristics should be considered to support beneficial insects throughout the season.

Switchgrass is a native perennial grass commonly used in conservation programs

(USDA, NRCS 2012) and it is being explored for bioenergy (Prochnow et al. 2009). In

addition, the survey of organic producers in Iowa indicated that grass is the dominant

plant type currently established in the majority of buffer strips. The results of our study

indicate that switchgrass monocultures may not be effective for increasing beneficial

insect abundance or diversity. In both years, natural enemy communities in switchgrass

did not significantly differ from corn and willow. Bee communities did not significantly

differ among all simple plant communities and bees were absent in switchgrass in 2011.

In contrast to corn and willow, switchgrass shared some characteristics (e.g., a greater

percentage of ground cover) that, as indicated by our analyses, have a positive

relationship with the diversity and abundance of natural enemies. However, when

considering multiple beneficial guilds, switchgrass, like corn and willow, lacks other

components (e.g., species diversity, floral resources) found in plant communities where

more diverse and abundant beneficial insect communities were consistently observed,

making switchgrass a sub-optimal candidate for buffer strips.

Gardiner et al. (2010) compared the beneficial insects communities in larger

(minimum 2 ha) fields of corn, switchgrass, and mixed prairie and found that beneficial

groups (e.g., bees and lady beetles) varied significantly among plant communities with

bee abundance and lady beetle species richness greater in switchgrass monocultures and

mixed prairie polycultures compared to corn. Our study demonstrated similar results in regards to plant communities composed of mixtures of prairie species; however, we did not see any significant differences between beneficial insect communities observed in corn and switchgrass. Our results may differ here because our switchgrass plots were

53 intensively managed as monocultures, and in the Gardiner et al. (2010) study switchgrass fields also contained varying amounts of other species depending on how each field was managed.

In contrast to the other simple buffer treatments, alfalfa had several characteristics

(e.g., percent ground cover, floral resources) that, as indicated by our analyses, have a positive relationship with the diversity and abundance of beneficial insect communities.

Natural enemy communities were significantly more diverse and abundant in alfalfa compared to corn and willow in both years, and additionally to switchgrass in 2011.

However, the same was true for herbivores in alfalfa in both years of our study. Alfalfa can provide multiple resources for beneficial insects, however the management regimes used in Iowa to control undesired pests, such as E. fabae, populations below economic thresholds include insecticide applications and early alfalfa harvest (Lefko et al. 1999).

Insecticides are not compatible with organic production and early harvest can remove habitat and prey, and therefore natural enemies. We did not manage our plots in this manner and uncut alfalfa plots became infested with E. fabae in both years.

Unmanaged pest populations in our alfalfa plots may be partially responsible for recruiting natural enemy populations in abundances not typically found in alfalfa when pests are controlled, therefore overestimating its ability to attract natural enemies. In addition, bee diversity and abundance in alfalfa was not significantly different from those plant communities where bees were completely absent (corn in 2010; willow and switchgrass in 2011), limiting the potential for the delivery of multiple insect-derived services to the surrounding landscape. Despite these results, alfalfa can be an attractive option when it doubles as a harvestable forage crop. This may apply to a subset of

54 organic farmers, and in this situation, benefits may increase when alfalfa is harvested in strips (see Weiser et al. 2003), such that not all habitat for beneficial insects is being harvested at once.

The MSU Best Bet mixture accumulated more beneficial insect species in both years overall. In regards to bees, halictids were particularly diverse and abundant in the

MSU Best Bet mixture. Several species of halictids are responsible for pollinating crops that require or benefit from insect-mediated pollination; examples include field-grown tomato (Greenleaf and Kremen 2006), watermelon (Kremen et al. 2004), and canola

(Morandin et al. 2005). Furthermore, many halictids are able to nest in areas of sparsely vegetated ground, are multi-voltine, and exhibit communal nesting behavior, while others, exhibit sociality at more complex levels (Packer et al. 2007). Therefore, attracting this group may lead to the pollination of crops across multiple bloom periods. Plant communities that accommodate an array of behaviors associated with cosmopolitan groups such as halictids may also be able to support more conserved behaviors exhibited by other groups.

In 2011, the most abundant bee species in MSU Best Bet and forb-only mixtures was the bumblebee, B. griseocollis. Bombus species are also known as pollinators of the crops listed above for halictids, and are especially effective pollinators of crops that require sonication. Halictids spp. and B. griseocollis (as well as other Bombus spp.) were consistently abundant in the MSU Best Bet even when their communities fluctuated between years among the other plant communities. Several species of parasitoids

(Braconids and Pteromalids), and predatory O. insidiosus were also observed in the MSU

Best Bet mixture and, were considerably abundant. These natural enemies, like the bees

observed in this mixture, have a well-established role in agroecosystems. These natural

enemies can attack a range of herbivorous insect pests. Overall, the diverse and abundant

beneficial insect communities described in the MSU Best Bet mixture can provide a suite

of ecosystem services that compliment organic crop production, as well as conventional

production systems.

In all observations, the MSU Best Bet mixture, outperformed the most species

rich plant community included in our study, the CP-IA mixture. In several instances, the

beneficial insect communities exhibited greater or equivalent diversity and abundance in

the forb-only treatments compared to the CP-IA mixture. Surprisingly, beneficial insect

communities in the CP-IA mixture (composed of 14 plant species) did not differ

significantly from some of the simple plant communities composed of only one plant

species.

The difference in beneficial insect communities between diverse treatments, CP-

IA and the MSU Best Bet mixtures, occurred despite the two plant communities being

composed of native grass and forb species. Typically, prairie mixtures have a greater

proportion of grass compared to forbs; however, we manipulated these ratios to contain a

greater proportion of forbs in both the MSU Best Bet (24 % forbs and 76% grass) and

CP-IA (32% forbs and 68% grass). This forb-rich ratio was used to optimize mixtures so

each diverse treatment had a season-long bloom period where the availability and accessibility of floral resources was maximized to accommodate a range of insect species. Despite the forb-rich plantings and continuous bloom periods common to each diverse mixture, beneficial insects preferred the MSU Best Bet mixture. This may be because the forbs in the MSU mixture produced a greater number of flowers creating a

56 more attractive floral display. The differences we observed between the MSU Best Bet and forb-only mixtures compared to the CP-IA mixtures reinforce the importance of the decision-making process needed for targeted conservation efforts. Carefully selecting the composition, considering the characteristics of individual plant species, and manipulating the density of these species in prairie plant mixtures can be essential to conserving beneficial insects at the farm scale where small additions of a few specific species can maximize benefits. For example, in a landscape-scale study conducted in the same eco- region as our research site, attractive forb species similar to those in the MSU Best Bet mixture (Aster laevis L., Ratibida pinnata (Vent.), Silphium perfoliatum (L.), and Zizia aurea (L.) W.D.J. Koch) were documented as present in a large (> 1,619 ha) reconstructed prairie embedded in cropland comprised of a corn-soybean rotation

(Schmidt et al. 2011). Despite the prairies close proximity to cultivated fields, no increase in natural enemy abundance or diversity was observed in adjacent crops suggesting that the densities of these species found in traditional prairie restorations may not be optimal for enhancing both biological diversity and functional diversity in a landscape altered for agricultural production.

The importance of conserving beneficial insects to maintain many ecological processes is being increasingly recognized and previous work has focused on the relationship of insect-derived services and successful crop production. In a review of habitat management literature, Fiedler et al. (2008) compiled data synthesizing habitat management field studies and associated publications to determine the extent to which recent studies focus solely on habitat management that promotes conservation biological control and found that many of these studies focus on describing the relationship of

57 natural enemies to specific plants. However, from 165 plant species documented among

34 habitat management-centered studies, the majority of research was conducted using exotic plant species.

Gaps in research regarding the establishment of native plants and their advantages over exotic species for enhancing multiple ecosystem services, including those mediated by beneficial insects, are being increasingly accounted for in the scientific literature

(Fiedler and Landis 2007; Tuell et al. 2008; Frank et al. 2008). Despite this trend, only a portion of the organic farmers in the survey identified ecosystem services as a benefit of their buffer (Table 11). This is an important indication of the need to communicate these benefits, not only to farmers, but also to policy writers and representatives of conservation agencies who consider incentives that promote the adoption of management practices to acquire multiple ecosystem services from buffer strips. Buffer strips have the potential to provide benefits beyond simply meeting requirements for organic certification.

In summary, the results from our field experiments indicate plant communities that dominate existing buffer strips and lands designated for conservation are not optimal for beneficial insects. Adding flowering perennial species can improve buffer strips as habitat for beneficial insects, especially bee pollinators. Moreover, buffer strips can be further optimized by incorporating attractive native species even at modest levels of plant diversity (e.g. MSU forb-only mixtures), such that flowering resources are available throughout the summer. In conclusion, successful habitat creation for beneficial insects appears to be a product of plant resource quality, a high density of attractive native species.

References Cited

Al-Dobai, S., S. Reitz, and J. Sivinski. 2012. Tachinidae (Diptera) associated with flowering plants: estimating floral attractiveness. Biol. Control. 61: 230-239.

Ambrosino, M. D., J. M. Luna, P. C. Jepson, and S. D. Wratten. 2006. Relative frequencies of visits to selected insectary plants by predatory hoverflies (Diptera: Syrphidae), other beneficial insects, and herbivores. Environ. Entomol. 35: 394-400.

Ascher, J. S., and J. Pickering. 2012. Discover Life bee species guide and world checklist (Hymenoptera: Apoidea: Anthophila). Available online at http://www.discoverlife.org. Last visited on 10 October 12.

Boutin, C., A. Baril, S. K. McCabe, P. A. Martin, and M. Guy. 2011. The value of woody hedgerows for moth diversity on organic and conventional farms. Environ. Entomol. 40: 560-569.

Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach, 2nd ed. Springer, New York, NY.

Clark, W. R., and K. F. Reeder. 2007. Agricultural buffers and wildlife conservation: a summary about linear practices, pp. 45-55. In J. B. Haufler (ed.), Fish and wildlife response to farm bill conservation practices, technical review 07-1. (USDA) United States Department of Agriculture (NRCS) Natural Resource Conservation service, and (FSA) Farm Service Agency in partnership with The Wildlife Society, Bethesda, MD.

Clarke, K. R. 1993. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 18: 117-143.

Colley, M. R., and J. M. Luna. 2000. Relative attractiveness of potential beneficial insectary plants to aphidophagous hoverflies (Diptera: Syrphidae). Environ. Entomol. 29: 1054-1059.

Costamagna, A. C., D. A. Landis, and M. J. Brewer. 2008. The role of natural enemy guilds in Aphis glycines suppression. Biol. Control. 45: 368-379.

Ehrenfeld, J. G., and L. A. Toth. 1997. Restoration ecology and the ecosystem perspective. Restor. Ecol. 5: 307-317.

Fiedler, A. K., and D. A. Landis. 2007a. Attractiveness of Michigan native plants to arthropod natural enemies and herbivores. Environ. Entomol. 36: 751-765.

Fiedler, A. K., and D. A. Landis. 2007b. Plant characteristics associated with natural enemy abundance at Michigan native plants. Environ. Entomol. 36: 878-886.

Fiedler A. K., D. A. Landis, and S. D. Wratten. 2008. Maximizing ecosystem services from conservation biological control: the role of habitat management. Biol. Control 45: 254-271.

Frank S. D., P. M. Shrewsbury, and O. Esiekpe. 2008. Spatial and temporal variation in natural enemy assemblages on Maryland native plant species. Environ. Entomol. 37: 478-486.

Feener, D. H., and B. V. Brown. 1997. Diptera as parasitoids. Annu. Rev. Entomol. 42: 73-97.

Gardiner, M. A., J. K. Tuell, R. Isaacs, J. Gibbs, J. S. Ascher, D. A. Landis. 2010. Implications of three biofuel crops for beneficial arthropods in agricultural landscapes. Bioenergy. Res. 3: 6-19

Greenleaf, S. S., and C. Kremen. 2006. Wild bee species increase tomato production and respond differently to surrounding land use in Northern California. Biol. Conserv. 133: 81-87.

Grissell, E. E. and M. E. Schauff. 1997. A handbook of the families of Nearctic Chalcidoidea (Hymenoptera), 2nd ed., revised. Entomological Society of Washington, Washington, DC.

Hannon L. E., and T. D. Sisk. 2009. Hedgerows in an agri-natural landscape: potential habitat value for native bees. Biol. Conserv. 142: 2140:2154.

Harmon, J. P., A. R. Ives, J. E. Losey, A. C. Olson, and K. S. Rauwald. 2000. Coleomegilla maculata (Coleoptera: Coccinellidae) predation on pea aphids promoted by proximity to dandelions. Oecologia. 125: 543-548.

Kremen, C., N. M. Williams, R. L. Bugg, J. P. Fay, and R. W. Thorp. 2004. The area requirements of an ecosystem service: crop pollination by native bee communities in California. Ecol. Lett. 7: 1109-1119.

Landis, D. A., F. D. Menalled, A. C. Costamagna, and T. K. Wilkinson. 2005. Manipulating plant resources to enhance beneficial arthropods in agricultural landscapes. Weed Sci. 53: 902-908.

Landis, D. A., S. D. Wratten, and G. M. Gurr, 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Annu. Rev. Entomol. 45: 175-201.

Lavandero, B. I., S. D. Wratten, R. K. Didham, and G. M. Gurr. 2006. Increasing floral diversity for selective enhancement of biological control agents: a double-edge sward? Basic Appl. Ecol. 7: 236-243.

Lefko, S. A., M. E. Rice, and L. P. Pedigo. 1999. Producer perception and pest management practices in Iowa alfalfa. J. Prod. Agric. 12: 257-263.

Losey J. E., and M. Vaughan. 2006. The economic value of ecological services provided by insects. Bioscience. 56: 311-323.

Lundgren, J. C., K. A. G. Wyckhuys, and N. Desneux. 2009. Population responces by Orius insidiosus to vegetational diversity. BioControl. 54: 135:142.

Lundgren, J. G. 2009. Nutritional aspects of non-prey foods in the life histories of predaceous Coccinellidae. Biol. Control. 51: 294-305

Mazerolle, M. J., 2012. AICcmodavg: model and multimodel inference based on (Q)AIC(c). R package version 1.25. (http://CRAN.R-project.org/package= AICcmodavg).

Michener, C. D. 2000. The Bees of the world. John Hopkins University Press, Baltimore, Maryland USA.

McCune B., and J. Grace 2002. Analysis of ecological communities. MjM Software Design. Gleneden Beach, OR.

Morandin, L. A., and M. L. Winston. 2005. Wild bee abundance and seed production in conventional, organic, and genetically modified canola. Ecol. Appl. 15: 871-881.

Morse, R. A. and N. W. Calderone. 2000. The value of honey bees as pollinators of U. S. crops in 2000. Bee Culture. 128: 15 pp insert.

Musser, F. R., and A. M. Shelton. 2003. Predation of Ostrinia nubilalis (Lepidoptera: Crambidae) eggs in sweet corn by generalist predators and the impact of alternative foods. Environ. Entomol. 32: 1131-1138.

Naeem, S. 2006. Biodiversity and ecosystem functioning in restored ecosystems: extracting principles for a synthetic perspective, pp. 210-237. In D. A. Flak, M. A. Palmer, and J. B. Zedler (eds.), Foundations of restoration ecology. Island Press, Washington D. C., USA.

Michigan State University (MSU), Native plants and ecosystem services. 2012. Available online at http://nativeplants.msu.edu. Last visited on 17 June 2012.Oksanen,

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O’Hara, G. L. Simpson, P. Solymos, H. H. Stevens, and H. Wagner. 2011. Vegan: community ecology package. R package version 2.0-1. (http://CRAN.R-project.org/package=vegan).

Östman, Ö. and A. R. Ives. 2003. Scale-dependent indirect interactions between two prey species through a shared predator. Oikos. 102: 505-514.

Packer, L., J.A. Genaro, and, C.S. Sheffield. 2007. The Bee Genera of Eastern Canada. Canadian Journal of Arthropod Identification No. 3, 25 September 2007, available online at http://www.biology.ualberta.ca/bsc/ejournal/pgs03/pgs_03.html, doi: 10.3752/cjai.2007.03. Last visited on 26 June 2012.

Palmer, M. A., R. F. Ambrose, and N. L. Poff. 1997. Ecological theory and community restoration ecology. Restor. Ecol. 5: 291-300.

Prochnow, A., M. Heiermann, M. Plöchl, T. Amon, and P.J. Hobbs. 2009. Bioenergy from permanent grassland – A review: 2. Combustion. Bioresource Technology. 21:4945-4954.

Pywell, R. F., E. A. Warman, C. Carvell, T. H. Sparks, L. V. Dicks, D. Bennett, A. Wright, C. N. R. Critchley, A. Sherwood. 2005. Providing foraging resources for bumblebees in intensively farmed landscapes. Biol. Conserv. 121, 479–494.

R Development Core Team. 2011. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. (http://www.R- project.org).

Rebek, E. J., C. S. Sadof, and L. M. Hanks. 2005. Manipulating the abundance of natural enemies in ornamental landscapes with floral resource plants. Biol. Control. 33: 203-216.

Rutledge, C. E., R. J. O’Neil, T. B. Fox, and D. A. Landis. 2004. Soybean aphid predators and their use in integrated pest management. Ann. Entomol. Soc. Am. 97: 240-248.

Samson, F. and F. Knopf. 1994. Prairie conservation in North America. BioScience. 44: 418-421.

SAS Institute. 2008. SAS/STAT, Version 9.2. SAS Institute, Cary, NC.

Schmidt, N. P., M. E. O’Neal, and P. M. Dixon. 2008. Aphidophagous predators in Iowa soybean: a community comparison across multiple years and sampling methods. Ann. Entomol. Soc. Am. 101: 341-350.

Schmidt, N. P., M. E. O’Neal, and L. A. Schulte Moore. 2011. Effects of grassland habitat and plant nutrients on soybean aphid and natural enemy populations. Environ. Entomol. 40: 260-272.

Simpson, E. H. 1949. Measure of diversity. Nature. 163: 688-688.

Smith, D. D. 1998. Iowa prairie: original extent and loss, preservation and recovery attempts. J. Iowa Acad. Sci. 105: 94-108.

Steffan-Dewenter, I., U. Munzenberg, C. Burger, C. Thies, and T. Tscharntke. 2002. Scale-dependent effects of landscape context on three pollinator guilds. Ecology. 83: 1421:1432.

Swinton, S. M., F. Lupi, G. P. Robertson, S. K. Hamilton. 2007. Ecosystem services and agriculture: cultivating agricultural ecosystems for diverse benefits. Ecol. Econ. 64: 245-52.

Tull, J. K., A. K. Fiedler, D. A. Landis, and R. Isaacs. 2008. Visitation by wild and managed bees (Hymenoptera: Apoidea) to Eastern U.S. native plants for use in conservation programs. Environ. Entomol.. 37: 707-718.

Triplehorn, C. A. and N. F. Johnson. 2005. Borror and DeLong’s Introduction to the study of insects, 7th ed. Thomson Brooks/Cole, Belmont, CA.

USDA (United States Department of Agriculture), 2009. National Organic Program, Production and Handling Preamble. Available online at http://www.ams.usda.gov/AMSv1.0/ams.fetchTemplateData.do?template=TemplateN&n avID=NOSBlinkNOSBMeetings&rightNav1=NOSBlinkNOSBMeetings&topNav=&left Nav=&page=NOPOrganicStandards&resultType=&acct=nopgeninfo. Last visited on 27 June 2012.

USDA (United States Department of Agriculture), 2010. Restoration of rare and declining habitat program fact sheet, CRP Practice CP25. Natural Resource Conservation Service, Washington DC.

USDA (United States Department of Agriculture), 2012. The PLANTS Database Natural Resource Conservation Service, National Plant Data Team, Greensboro, NC. Available online at http://plants.usda.gov. Last visited on 27 June 2012.

Wackers, F. L., P. C. J. van Rijn, G. E. Heimpel. 2008. Honeydew as a food source for natural enemies: making the best of a bad meal. Biol. Control. 45: 176-184.

Wiser Erlandson L. A., and J. J. Obrycki 2010. Predation of immature and adult Empoasca fabae (Harris) (Hemiptera: Cicadellidae) by three species of predatory insects. J Kansas Entoml. Soc. 83: 1-6.

Weiser, L. A., J. J. Obrycki, and K. L. Giles. 2003. Within-field manipulation of potato leafhopper (Homoptera: Cicadellidae) and insect predator populations using an uncut alfalfa strip. J. Econ. Entomol. 96: 1184-192.

Zehnder, G., G. G. Gurr, S. Kuhne, M. R. Wade, S. D. Wratten, and E. Wyss. 2006. Arthropod pest management in organic crops. Annu. Rev. Entomol. 52: 57-80.

Zhang W., Ricketts T. H., C. Kremen, K. Carney, S. M. Swinton. 2007. Ecosystem services and dis-services to agriculture. Ecol. Econ. 64: 253-260

Table 1. Common plant types used in organic buffer zones by organic row- crop and horticultural-crop producers Percenta Plant species/Plant categoryb Row crops (47) Horticulture (13) All Grasses 52.50 92.90 60.00 Legumes 34.40 57.10 38.70 Shrubs or trees 6.60 28.60 13.30 Row Crop 9.80 0.00 12.00 Prairie or CRPc 14.80 7.10 6.70 Other 4.90 14.30 6.70 aPercentages calculated for plant categories total to over 100% because this was an open ended survey question and producers could list multiple species used in their buffers. Numbers in parentheses indicate the number of organic producers that responded to this question. bActual question asked in survey was “What do you plant and maintain in your buffer strips?” Plants used by growers within buffer strips were collapsed by the most general category used in individual producer’s responses. cCRP refers to land enrolled in the conservation reserve program.

Table 2. Species selected for simple buffer treatments and their associated characteristics Common namea Scientific name Duration Growth habitb Switchgrass Panicum virgatum L. Perennial WS-GR Alfalfa Medicago sativa L. Perennial LG-FB Willow Salix matsudana Koidzumi Perennial SH Corn Zea mays L. Annual WS-GR aSingle-species compositions were selected for simple habitats to represent common plant types currently used in organic buffer zones. Species are listed in order (greatest to least) according to the total percentage (“All”) of buffers composed with each plant category, based on information obtained in the survey of organic producers in Iowa (Table 1). bGrowth habit codes indicate functional groups: WS-GR = warm-season graminoid; LG-FB = leguminous forb; SH = shrub.

Table 3. Species selected for CP-IA mixture treatments and their associated characteristics Common namea Scientific name Bloom timeb Growth habitc Spotted geranium Geranium maculatum L. May-June FB Pale purple coneflower Echinacea pallida (Nutt.) Nutt. June-July FB Blackeyed susan Rudbeckia hirta L. June-Aug. FB Smooth oxeye Heliopsis helianthoides (L.) Sweet June-Aug. FB Culver’s root Veronicastrum virginicum (L.) Farw. June-Aug. FB Purple prairie clover Dalea purpurea Vent. June-Aug. LG-FB Canada wildrye Elymus canadensis L. July-Sept. CS-GR Sideoats grama Bouteloua curtipendula (Michx.) Torr. July-Sept. WS-GR Big bluestem Andropogon gerardi Vitman Aug.-Sept. WS-GR Indiangrass Sorghastrum nutans (L.) Nash Aug.-Sept. WS-GR

Rough dropseed Sporobolus clandestinus (Biehler) Hitchc. Aug.-Sept. WS-GR 66 Prairie sage Artemisia ludoviciana Nutt. Aug.-Oct. SS Short’s aster Symphyotrichum shortii (Lindl.) G.L.Nesom Aug.-Oct. FB Stiff goldenrod Oligoneuron rigidum (L.) Small Aug.-Oct. FB aThe CP-IA mixture includes 14 native perennial species, five grasses, eight forbs and a shrublet which are a subset of species selected from the Iowa Ecotype CP25 conservation prairie mixture that in combination, bloom throughout the season. bSpecies are ordered by bloom periods (earliest-latest for Iowa) when conspicuous flowers, or, for grasses, inflorescences are present. The duration of flowering can be from three weeks to three months depending on the species and environmental conditions. cGrowth habit codes indicate functional groups: FB = forb/herb; LG-FB = leguminous forb; CS-GR = cool- season graminoid; WS-GR = warm-season graminoid; SS = sub shrub.

Table 4. Species selected for MSU Best Bet mixture treatments and their associated characteristics. Common namea Scientific name Bloom timeb Growth habitc Canadian anemone Anemone canadensis L. May-June FB Meadow ziziad Zizia aptera (A. Gray) Fernald May-June FB Pinnate prairie coneflower Ratibida pinnata (Vent.) Barnhart June-Aug. FB Swamp milkweed Asclepias incarnata L. June-Aug. FB Switchgrass Panicum virgatum L. July-Aug. WS-GR Canada wildrye Elymus canadensis L. July-Sept. CS-GR Common boneset Eupatorium perfoliatum L. July-Sept. FB Cup plant Silphium perfoliatum L. July-Sept. FB Prairie ironweedd Vernonia fasciculata Michx. July-Sept. FB Little bluestem Schizachyrium scoparium (Michx.) Nash Aug.-Oct. WS-GR

New England aster Symphyotrichum novae-angliae (L.) G.L. Nesom Aug.-Oct. FB 67 Smooth blue aster Symphyotrichum laeve (L.) Á. Löve & D. Löve Aug.-Oct. FB aThe MSU Best Bet mixture includes12 native perennial species, three grasses and nine forbs, selected based on those individually rated “Best” for relative attractiveness to either (or both) arthropod natural enemies and wild and managed bees in evaluations by Fiedler and Landis (2007) and Tuell et al. (2008). Selection was further restricted to Iowa ecotypes, that in combination, bloom throughout the season. bSpecies are ordered by bloom periods (earliest-latest for Iowa) when conspicuous flowers, or, for grasses, inflorescences are present. The duration of flowering can be from three weeks to three months depending on the species and environmental conditions. cGrowth habit codes indicate functional groups: FB = forb/herb; CS-GR = cool-season graminoid; WS-GR = warm-season graminoid. dVernonia missurica and Zizia aurea used by Fiedler and Landis (2007) were not available as plugs through our local provider and were replaced with V.fasciculata and Z. aptera; similar species in the same genus.

68 Table 5. Species included in forb-only mixtures Common namea Scientific nameb MSU5 MSU3 MSU2 Meadow zizia Zizia aptera X X X Swamp milkweed Asclepias incarnata X X Pinnate prairie coneflower Ratibida pinnata X Cup plant Silphium perfoliatum X X X New England aster Symphyotrichum novae-angliae X aForbs from the MSU Best Bet mixture were selected to create three additional treatments. The planting density of each forb-only mixture remained the same as the diverse plantings, but species richness was reduced. Bloom periods and growth habits are as in Table 3. X indicates species present in each forb-only mixture. bAuthors for species as in Table 4.

Table 6. Mean ± SEM for plant characteristics measured among buffer types during 2010 and 2011 Canopy Ground Flower Canopy Ground Flower Buffer Diversitya height (cm)b cover (%)c numberd height (cm)b cover (%) numberd Corn 1 ± 0e 159 ± 13b 20 ± 4ef ---- 156 ± 13b 18 ± 4d ---- Willow 1 ± 0e 135 ± 46a 5 ± 0.2d ---- 297 ± 49a 4 ± 1f ---- Switchgrass 1 ± 0e 115 ± 10b 74 ± 6ab ---- 119 ± 10b 74 ± 6ab ---- Alfalfa 1 ± 0e 55 ± 4b 85 ± 3a 8 ± 2b 60 ± 4b 81 ± 3a 9 ± 3b MSU2 2 ± 0d 104 ± 11b 37 ± 11de 64 ± 12a 115 ± 21b 37 ± 10c 70 ± 13a MSU3 3 ± 0.1c 89 ± 29b 45 ± 4dc 52 ± 13a 91 ± 26b 47 ± 4bc 58 ± 15a MSU5 5 ± 0.1b 79 ± 11b 68 ± 6abc 30 ± 7a 99 ± 14b 71 ± 6ab 31 ± 7a MSU Best Bet 12 ± 0.3a 117 ± 32b 53 ± 5bcd 67 ± 14a 109 ± 6b 54 ± 6bc 68 ± 14a CP-IA 12 ± 0.4a 77 ± 15b 60 ± 9abcd 23 ± 4b 83 ± 6b 61 ± 9abc 24 ± 4b Means within columns followed by common letters are not significantly different at P < 0.05. aPlant diversity is the mean value of Simpson’s diversity index (1/D), values increase as diversity increases. Diversity 69 values were the same for both years and are reported once to represent both 2010 and 2011. bCanopy represent end-of-season canopy height measured in August data during 2010 and 2011. cPercent ground cover and canopy height (cm) represent end-of season conditions with means calculated using August data in 2010 and 2011 to account for the annual establishment of corn treatments. dFlower number is the mean abundance of flowers per buffer type per sampling event.

Table 7. Total and mean ± SEM number of taxonomic units per treatment for bees and natural enemies vacuum collected in 2010 and 2011 Switch- MSU Taxa Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA Bees Totala 1 2 3 6 13 14 13 18 14 2010 0d 1 ± 0.6d 1 ± 0.7d 2 ± 0.9cd 5 ± 1.4bc 5 ± 1.6bc 7 ± 1.5b 10 ± 1.2a 6 ± 1 2011 0.3 ± 0.2b 0b 0b 1 ± 0.2b 4 ± 0.6a 4 ± 0.6a 2 ± 1.2ab 3 ± 1.2ab 1 ± 0.4ab

Natural enemies Totala 37 32 50 58 50 51 59 57 57 2010 17 ± 2b 10 ± 1b 36 ± 1a 43 ± 6a 36 ± 4a 38 ± 6a 40 ± 3a 48 ± 5a 45 ± 3 2011 16 ± 4d 15 ± 3d 25 ± 4cd 46 ± 4a 22 ± 0.6cd 28 ± 2bcd 36 ± 2abc 40 ± 5a 27 ± 2bcd Means within columns followed by common letters are not significantly different at P < 0.05.

Species captured in both years were only counted once for totals. See Appendix for a list of bee and natural enemy species and their 70 presence/absence per treatment per year. aTotal refers to the total number of species/unique morphospecies within a guild summed across all samples for 2010 and 2011.

Table 8. Summary of model selection statistics used for evaluating the influence of plant community characteristics on the species richness and abundance of bees and natural enemies vacuum collected in 2010 and 2011. a 2 b Response variable Ki AICc ΔAICc ωi Adj. R Variables in model 2010 Bees Richness 4 18.20 0.00 0.99 0.54 Plant diversity, Flower number Abundance 4 35.97 0.00 0.58 0.35 Flower number, Plant diversity Abundance 5 36.65 0.68 0.41 0.29 Flower Number 2011 Bees Richness 3 -77.00 0.00 0.99 0.60 Flower number Abundance 3 -23.77 0.00 0.99 0.56 Flower number

2010 Natural enemies Richness 6 44.42 0.00 0.69 0.80 Plant diversity, Flower number, Ground cover (Canopy height) Richness 5 46.09 1.67 0.30 0.78 Plant diversity, Ground cover

Abundance 5 204.14 0.00 0.88 0.57 Plant diversity, Flower number, Ground cover 71 2011 Natural enemies Richness 5 63.46 0.00 0.44 0.37 Plant diversity, Flower number, Ground cover Richness 4 63.83 0.37 0.36 0.33 Plant diversity, Ground cover (Canopy height) Abundance 5 183.20 0.00 0.56 0.41 Plant diversity, Flower number, Ground cover Abundance 4 184.21 1.01 0.34 0.36 Plant diversity, Ground cover

Ki = number of variables in each model, AICc = Akaike’s Information Criterion adjusted for sample size, ΔAICc = difference in AICc score between best-fit and competing models, ωi = Akaike weights; an estimate of the relative likelihood of a given model against all other models, and Adj. R2 = R-square adjusted for the number of terms in the model. a Where response variables are listed twice within guilds and years, the first model is the best-fit model based on the minimum 2 AICc and greater Adj. R values. Competing models are listed second and only competing models with ΔAICc < 2 are shown. bPlant diversity is Simpson’s 1/D, flower number is mean flower abundance per sampling event. Variables in parentheses indicate a negative relationship. (For details regarding plant characteristics see Methods and Materials: plant measurements; see Table 6 for plant characteristics comparisons across treatments).

72 Table 9. Summary of producer perception of services accrued from buffer strips Percenta Servicea Row crops (45) Horticulture (11) All Meets requirements 67.2 57.1 65.3 Ecosystems services 19.7 21.4 20.0 Reduces pesticide drift 13.1 7.1 12.0 Turn strip, driving alley 11.5 14.3 12.0 None 6.6 0.0 5.3 Reduces cross-pollination 6.6 0.0 5.3 Used or sold as feed 6.6 7.1 6.7 Decreases income 1.6 0.0 1.3 Trees/nuts 1.6 0.0 1.3 Improved aesthetics 1.6 0.0 1.3 aActual question asked in survey was “What services do buffer strips meet in farm production?” Responses by producers were collapsed by the most general category used in producer descriptions. Numbers in parentheses indicate the number of organic producers that responded to this question.

73 Table 10. Benefits and harm from buffer strips identified by producers during interviews Percenta Row crops (11) Horticulture (3) All Benefitsa Turn strip 33.3 33.3 33.3 Ecosystem service 50.0 66.7 53.3 Contamination prevention 41.7 33.3 40.0 Certification 25.0 00.0 20.0 Animal Feed 8.3 33.3 13.3 Harma Financial/can’t farm 33.3 0.0 26.7 Gophers/animals 25.0 0.0 20.0 Weeds 8.3 33.3 13.3 None 41.7 66.7 46.7 aActual questions asked, “What positive effects does your buffer have on production?” and “What negative effects does your buffer have on production?” Responses by producers were collapsed by the most general category used in producer descriptions. Numbers in parentheses indicate the number of organic producers that responded to this question.

74 Table 11. Producer’s perceptions of how pest management affects buffer strip usage Percenta Categories of responsea Row Crops (45) Horticulture (11) All None 70.0 72.7 70.6 Don’t know 7.5 0.0 5.9 Animal related pest problems 15.0 9.1 13.7 Insect related pests or benefits 2.5 18.2 5.9 Management responses 5.0 0.0 3.9 aActual question asked in survey was “How have pest management issues affected your usage of buffer strips?” Responses by producers were collapsed by the most general category used in producer descriptions. Numbers in parentheses indicate the number of organic producers that responded to this question.

75 Table 12. Producers’ perception of how buffer strips effect pest management on farm Percenta Categories of response Row Crops (49) Horticulture (12) All No 65.3 58.3 52.0 Yes 22.5 16.7 21.3 Don’t know 8.2 8.3 8.2 Other 6.1 16.7 8.2 aActual question asked in survey was “Do you feel your buffer strip has any effect on pest management on your farm? Responses by producers were collapsed by the most general category used in producer descriptions. Numbers in parentheses indicate the number of organic producers that responded to this question.

76 Figure Legends

Fig 1. Non-metric multidimensional scaling ordinations (nMDS plots) of the species composition based on the Bray-Curtis dissimilarity indices for (a) 2010 and (b) 2011 vacuum samples depicting the configuration of treatments in relation to the community dissimilarity; treatments in close proximity have a species composition that is more similar than treatments separated by greater distances. Guild vectors represented by the plotted arrows indicate the strength in which the species composing each guild are correlated with treatments and the arrow points in the direction of the most rapid change in increasing abundances. Correlations were significant at P ≤ 0.05.

Fig 2. Mean ± SEM abundance of bees across buffer types collected in (a) 2010 and (b) 2011 per plot. Means with common letters are not significantly different.

Fig 3. Mean ± SEM abundance of natural enemies across buffer types collected in (a) 2010 and (b) 2011 per plot. Means with common letters are not significantly different.

Fig. 1

Fig. 2

Fig. 3

80 CHAPTER 3. DO BUFFER STRIPS CONTRIBUTE TO THE BIODIVERSITY

OF ORGANIC FARMS?

A paper to be submitted to Environmental Entomology

K. A. GILL AND M. E. O’NEAL1

1Department of Entomology, Iowa State University, Ames, IA 50011

Abstract

Organic farming is thought to be a source of biodiversity to agricultural landscapes by means of heterogeneous cropping systems and required land management practices. Iowa currently ranks fifth in the Unities States in the number of certified organic farms. Certified organic farms are required to create buffer zones on organically managed land to mitigate drift of pesticide and genetically modified pollen from adjoining land that is not managed organically. Typically, these buffer zones are strips of perennial, non-crop vegetation. Patches of non-crop vegetation provide resources to beneficial insects, allowing them to persist near agricultural fields. Our goal is to document the community of beneficial insects in buffer strips and compare them to what is found in the adjacent organically managed farms and conventionally managed row crops. We hypothesize that the abundance, diversity, and activity of pollinators and natural enemies will be greatest in buffer strips, intermediate in organically managed farms, and lowest in conventionally managed row crops. We hypothesize that the abundance, diversity, and activity of pollinators and natural enemies within organic farms and row crops will decrease as distance from buffer strip increases. The abundance of

81 beneficial insects was greatest in buffer strips, intermediary in the organic farms, and

lowest in the conventional row crops.

Keywords: pollinator, natural enemy, organic, biodiversity, conservation

Introduction

Beneficial insects contribute significant ecosystem services to agricultural lands.

Insect-derived ecosystem services such as pollination and biological control of plant pests

by natural enemies is crucial to successful agricultural production (Zhang et al. 2007,

Caballero-López et al. 2012). The annual value of pollination to United States (US)

agriculture is estimated to be $14.6 billion for pollination services provided by the

managed European honey bee, Apis meliffera L., (Hymenoptera: Apidae) (Morse and

Calderone 2000). Many species of wild bees contribute to 15% of US crop pollination

services, which is valued at $3 billion per year (Losey and Vaughn 2006). Moreover,

insect natural enemies (predators and parasitoids) are reported to be responsible for 33%

of the natural pest control in cultivated systems which, is estimated to be worth an

additional $4.5 billion dollars to US agriculture annually (Hawkins et al. 1999, Losey and

Vaughn 2006).

Despite the valuable relationship of beneficial insects and crop production, the

biodiversity that drives these relationships is declining as a consequence of agricultural

intensification and landscape simplification (Foley et al. 2005, MEA 2005). The lack of

biodiversity limits the quantity and quality of plant-derived resources available to beneficial insects. Limiting resources can have negative consequences on the diversity

82 and abundance of beneficial insects, and their ability to consistently fulfill their functional role in managed landscapes (Foley et al. 2005).

Nectar, pollen, refuge from agricultural disturbances, nesting materials, and overwintering sites are examples of plant-derived resources that positively influence diversity and abundance in beneficial insect communities (Westrich 1996, Menalled et al.

1999, Elliott et al. 2002, Steffan-Dewenter et al. 2002, Landis et al. 2000, 2005; Ricketts et al. 2006, Klein et al 2007, Kremen et al. 2007, Williams and Kremen 2007, Zhang et al. 2007, Kwaiser and Hendrix 2008, Tscharntke et al. 2008, Wackers et al. 2008, Le

Féon et al. 2011). Patches of non-crop vegetation within agricultural landscapes can provide such resources, allowing beneficial insect communities to persist near agricultural fields before, during, and after periods when insect-derived ecosystem services are valuable to crops (Landis et al. 2000, Bianchi et al. 2006, Isaacs et al. 2009).

Methods for reintroducing biodiversity to agriculturally dominated landscapes include the diversification within cropping systems (e.g., crop rotations, polycultures, cover crops) and establishing diverse vegetation surrounding cultivated areas (e.g., field margins, hedgerows, buffer strips) (Altieri 1999, Landis et al. 2005)

Organic farming and associated practices may counteract, or reduce, the negative effects of conventional agriculture, including the loss of biodiversity (Altieri 1999, Foley et al. 2005, Tscharntke et al. 2012). Although there is variation by taxa, recent reviews conclude that, in general, biodiversity is higher on organic versus conventional farms and his response is consistent for most invertebrates, especially insects (Bengtsson et al.

2005, Hole et al. 2005, Letourneau and Bothwell 2008, Ponce et al. 2011). As noted by

Hole et al. (2005), one problem with determining the impact of organic farming on

83 biodiversity is the variation in definitions of organic agriculture. This is especially true if efforts to address declining biodiversity within agroecosystems attempt to include and evaluate practices from organic production systems (e.g. buffer strips, fallow margins, cover crops). More recently, Letourneau and Bothwell (2008) reviewed studies that compared organic and conventional farms (including aforementioned Hole et al. [2005] and Bengtsson et al. [2005]). This review concluded that, due to the increase in organic farming and support for sustainable agricultural practices, additional research in these agroecosystems should address the relationship between on-farm biodiversity and enhancing ecosystem services for farmers from both natural and managed ecosystems.

In the US, farmers who want to sell farm products as organic must meet a national certification standard required by the United States Department of Agriculture (USDA)

National Organic Program (NOP). Included in this certification process is a requirement

(§ 205.202 USDA NOP 2009) that buffer zones be in place between the organically managed land and adjacent farmland not under organic management. The primary purpose of this mandatory buffer zone is to mitigate the drift of pesticide and genetically modified pollen from adjoining land that is not managed organically. However, these buffer zones can also be managed as strips of non-crop vegetation (i.e. buffer strips) to serve as a refuge for biodiversity, including beneficial insects. To what extent buffer strips contribute to the differences in biodiversity between organic farms and conventional agriculture is unclear.

The state of Iowa is one of the leading producer’s of corn, Zea mays L., and soybean, Glycine max (L.) Merr., in the US. During 2011, Iowa planted 5.5 million ha of corn and 4 million ha of soybean, representing 75% of the farmland in Iowa’s landscape

84 (IDALS 2011), the majority of which is not grown organically. Despite this, the state of

Iowa currently ranks fifth in the US in the number of certified organic farms (467)

(USDA NASS 2011). The majority of these organic farms are small, accounting for only

24,790 of the 12.5 million ha of farmland in Iowa, and many produce a variety of vegetables. Although organic agriculture is limited in the types of pesticides permitted, the intensity of farming practices (tillage, crop rotation, etc.) is not insignificant (Kovach et al. 1992). These organic farms are embedded in a landscape dominated by conventional corn and soybean row crop, which generally lack biodiversity especially with regard to beneficial insects (Altieri 1999, Landis et al. 2000, 2005). To what extent buffer strips, located between organic and conventional farms contribute beneficial insects to either cropping system, is not known.

Our goal is to document the community of beneficial arthropods in buffer strips and compare them to what is found in the adjacent organically managed farms and conventionally managed row crops. We hypothesize that the abundance, diversity, and activity of pollinators and natural enemies will be greatest in buffer strips, intermediate in organically farms, and lowest in conventionally managed row crops. Additionally, we hypothesize that the abundance, diversity, and activity of pollinators and natural enemies within organic farms and row crops will decrease as distance from buffer strip increases.

85 Materials and Methods

Study Site Selection.

Each study site was selected using various site-level and field-level criteria. Site- level criteria were used to obtain replicate study blocks where three adjacent land-use types (i.e. organic farms, buffer strips, and conventional row crops) were already established in the landscape. To limit sources of variation beyond the land-use types, four sites were selected that shared similar landscape features that surrounded each study site (see the Landscape Measurements section of this chapter for how this was quantified). Organic farms and associated buffer strips were land-use types essential to our questions, so we first identified organic producers who were willing to collaborate with this research. Using the available organic farms, we narrowed candidates to organic farms with similar cropping systems that included multiple crops such as vegetables, fruits, and herbs. We further narrowed sites to organic farms that had non-crop plant communities established in their buffers. The last criterion was that each organic farm

(and buffer strip) adjoined a conventionally (i.e. not under organic management) field of corn or soybean row crops (hereafter, row crops).

Study Site Description.

Four replicate study sites met the above selection criteria and were used in both years of our study, 2010 and 2011. Sites were located in four Iowa counties within a 120 km radius of Iowa State University, Ames, Iowa, US (Table 1). The minimum distance between sites measured approximately 16 km and the maximum distance between sites

86 approximately 96 km. Each of the four study sites included the aforementioned adjacent

land-use types (see Study Site Selection).

theOrganic landowner farms was also were the lead privately operator of the owned production production system. Each systems. farmer For all organic

organicallyfarms, grew approximately 30 to 50 different crops (Table 2), which required

multiple plantings and harvests across several months (April – November). The size and

configuration of the four organic farms within each site varied, with the total parcel size

ranging from approximately 3 to 32 ha (Table 1). On average, the amount of land

specifically devoted to organic crop production at any one time was approximately 2.5

ha. Pastureland, livestock, greenhouse production, production of ornamental plants, or

cover crops accounted for land use in the remaining hectares not planted to organic crops.

Data collection at organic farms focused on portions of the parcel devoted to organic crop

production (vegetables, fruits, herbs) sold for human consumption.

Three out of four organic farms produced certified organic products with

certifications obtained through accredited certifying agencies (Table 1) (USDA, NOP

2002). One organic farm was not certified, but shared characteristics and organic

management practices of those that were certified. Pest management practices employed

among organic farms over the course of the growing season included cover-cropping,

crop rotations, green manures, and composting. In compliance with organic

management, all four organic farms prohibited the use of genetically modified (GM) crops, synthetic pesticides, and substances not approved for USDA certified organic production.

87 Although buffer strips are part of the land use on organic farms, we treated them separately from crop vegetation sampled because these areas contained uncultivated, perennial vegetation not harvested as crops. Each buffer strip in our study adjoined land not managed organically, such that buffer strip edges were shared on one side by an organic farm and on the opposite side by row crops. Although the plant species within buffer strips varied among sites, many buffers were composed of species that shared growth habits (Table 3). To compare buffer strip composition among sites, we collected plant data within the areas of buffer strips in which arthropod sampling occurred. These data were collected at each site by randomly tossing a quadrat (30.48 by 30.48 cm) into buffer strips and visually identifying plants within the quadrat. At each site, 16 random quadrat samples were collected (four subsamples near each trap location within buffer strips). The plants observed in quadrat samples were recorded and categorized according to growth habits defined in the PLANTS database (USDA, NRCS 2013; http://plants.usda.gov/growth_habits_def.html). In the event that a quadrat landed in a location where it contacted (on any side) vegetation that could not be realistically sampled using this method (i.e., trees, large shrubs) the plants were recorded as if they were within the quadrat. In 2010, we conducted quadrat sampling in June, July, and

August, coinciding with arthropod sampling dates. Since our traps were located in the same position in 2011 as in 2010, quadrat sampling was limited to July in 2011 as we did not observe plants with growth habits that were not previously identified. The configurations of all buffer strips were linear. Beyond the standardized sampling area used throughout this study, the total length of buffer strips ranged among sites from buffer strips maintained along property lines (formerly fence lines) along the length of

88 certain field edges to buffer strips maintained along the majority of the organic farm perimeter.

Row crops were either owner-operated or operated by individual farmers renting the parcel. Row-crop fields were managed conventionally and each farmer grew corn and soybeans in a rotation (i.e. alternating each crop every other year at the field-level) according to standard production methods (Table 2). In 2010, two row-crop fields produced corn (Dallas Co. and Polk Co.), and the other two soybeans (Poweshiek Co. and

Story Co.) and the opposite in 2011. The growing season for conventional fields covered a similar time span for corn (planting dates: 25 April through 18 May; harvest dates: 5

October through 9 November) and soybean (planting dates: 8 May through 2 June; harvest dates: 28 September through 20 October). In contrast to the diverse cropping- system on organic farms, these crops were grown in monocultures, each planted and harvested once throughout the season. Hereafter these conventionally managed corn and soybean rotations are referred to as row crops.

The size and configuration of the area planted to row crops at each site varied, with the total area ranging from approximately 12 to 40 ha, most of which was solely devoted to corn or soybeans (i.e. no cover-crops or intercropping methods were used, nor did conservation plantings exist within the cropped area). Row crops were harvested for bulk sale.

Among row crops, common management practices included low or no-tillage and application of synthetic pesticides to seed and foliage. Other in-depth, histories of specific practices employed in row crops beyond the time frame of this study were not available. However, when we received permission to sample row crops, individuals

89 currently responsible for making management decisions agreed to share information regarding pesticide application (specifically the no-entry period and re-entry period) prior to each field visit. This was to ensure personal safety, to avoid moving unwanted chemical residue to adjacent organic farms, and to avoid compromising the quality of our data collection.

Arthropod Collection.

To describe the arthropod community, we collected arthropods using a line- transect method following a stratified random sampling plan. Per site, we established four parallel line-transects, and each transect measured 450 m extending across all three fields using the following procedure. A random starting point was selected within an area of the buffer strip such that the opposite sides adjoined either an organic farm or conventional field. A location in the centermost portion of the buffer strip was recorded as the first of four sampling point in the buffer strip and as a reference point to measure from when constructing the remaining transects. From the outermost edge of the buffer strip, into the organic farm and row crop, sampling points were marked at 50 m intervals to a maximum distance of 200 m into each farm; this was repeated for a total of four transects, each parallel to the existing transect with approximately 50 m between each transect.

The arthropod (insects and spiders) community was sampled along these transects in June, July, and August of 2010 and 2011. Two different trapping methods were deployed to account for multiple guilds (Rebek et al. 2005, Schmidt et al. 2008); an active sampling method (vacuum sampling) to account for arthropods residing on

90 vegetation; and, a passive sampling method (yellow sticky traps) to account for activity- density, including species active at different times of the day, highly mobile species, and those sensitive to plant disturbance.

The vacuum sampling was done with gas-powered leaf blower (Troy-Bilt,

Model# TB320BV) modified (per Fiedler and Landis 2007) to capture foliar-dwelling, flower-visiting, and more sedentary arthropods. Samples were collected within a fine mesh paint strainer placed over the air intake when the leaf blower was set to vacuum.

To ensure consistency among samples, vacuum sampling was restricted to mid-day during favorable weather conditions (warm, sunny days, limited cloud cover < 30% and with wind gusts < 5 mph). Samples were collected along the transects at each point

(center of buffer and 50 m interval from the buffer edge) by directly vacuuming vegetation at each sampling point along each transect. The vegetation in buffer strips, organic farms, and row crops was vacuumed for 30 s while moving continuously, contacting the foliage and flowers within 1 m on either side for the transect line. The mesh strainer and the sample contained within was then removed, placed into a clear plastic, resealable bag, and replaced with an unused, clean mesh strainer for subsequent sampling. Samples were transported to the lab and frozen until processed.

Unbaited yellow sticky traps (Pherocon AM®, GEMPLER’S, Madison, WI) were used to measure the activity-density of mobile arthropods. Traps were positioned along transects at each measured point (center of buffer and 50 m interval from the buffer edge) by fastening the trap to a wooden stake adjusted such that the bottom of the trap was at the height of the vegetation. The yellow sticky traps remained in the fields for approximately 5 d and were then collected, transported back to the lab, and frozen for

91 future identification. Not all sites could be sampled on the same day, but during each month, sites were sampled within the same week except for one instance when the

Poweshiek Co. site was sampled 10 d later than the other three) due to extended suboptimal weather conditions. All three land-use types that comprise each site were always sampled during the same day or span of days depending on the trap type. Due to the difficulty of identifying individuals on yellow sticky traps, analyses of species richness and species evenness were performed using data collected in vacuum samples only.

Arthropod Identification and Guild Assignment.

When possible, we identified insects to species and other arthropods (spiders) to order (Araneae). The lowest taxonomic unit possible was assigned when species identification could not be resolved. These individuals were then organized into morphospecies and given a unique identifier for reference and classification of duplicates.

Following identification, individuals were classified to guilds; natural enemies, pollinators, detritivores, fungivores, and “other” based on species accounts described in the identification keys used and reviewed literature. The group referred to as “other” includes species with non-feeding adults, blood-feeders, and unresolved feeding habits.

For this study, we focused on pollinators associated with two different floral traits including melittophilous syndromes (bee-pollinated flowers) and myophilous syndromes

(fly-pollinated flowers). Specifically, our community analyses included domesticated A. meliffera, wild bee species (non-Apis, multiple families), and syrphid flies (Diptera:

Syrphidae). Hereafter, referred to as bees and syrphids. Analyses pertaining to natural enemies included both predators and parasitoids. As adults, syrphids species are known

92 to visit a variety of flowering plants to feed on nectar (for a list of floral hosts see; Tooker et al. 2006) and their contribution to pollination has been documented (Kevan and Baker

1983, Ssymank et al. 2008). However, unlike immature bees, the immature stages of syrphids are not provisioned with pollen and syrphids vary by species in their immature feeding habits. Species considered pollinators in this study either prey on insects

(subfamily Syrphinae) or consume organic matter (subfamily Eristalinae) as larvae.

Species that are predators as larvae are also accounted for as natural enemies in this study.

Arthropod Community Composition.

To understand the arthropod community throughout the entire study system, we report the composition for the overall community (i.e. collected across all agricultural land-use types and sites) observed in each year. Communities specific to each field type are further summarized by measures of total and proportional abundance by guild. Total abundance, calculated for each year, describes the total number of individuals summed across all sites and the different fields within sites for each year. Proportional abundance in each guild is presented as a percentage of the total abundance for each year.

For descriptions of beneficial arthropod communities at the field-level, vacuum data were used to summarize measures of abundance and diversity of pollinators (bees and syrphids) and natural enemies (predators and parasitoids). Analyses compared abundance, species richness, and evenness for each beneficial guild on a per sample basis. Species richness and evenness estimated the diversity of beneficial guilds. Species richness (S) was calculated as the total number of unique taxa per guild observed per

93 vacuum sample. Pielou’s evenness index (Peilou 1969) also referred to as the Shannon evenness measure (Magurran 2004) estimated evenness (equitability), described as the extent of the representation by equal numbers of individuals of the different species of a given community. Evenness values (J) were calculated for each guild as the ratio of observed diversity to maximum diversity per vacuum sample with J = H'/log (S) where H' is the resulting value of the Shannon diversity index measure (1949) and S is species richness (as described above). Evenness values (J) were calculated for each guild per vacuum sample. Diversity indices of species richness (S) and evenness (J) were calculated using the “vegan” package version 2.0-1 in R version 2.14.1 (Oksanen et al.

2011, R Development Core Team 2011).

Activity-density.

To describe the activity-density of arthropods among different fields, we used data collected with yellow sticky traps. These traps include a visual component and intercept mostly mobile arthropods. Bees were captured infrequently with yellow sticky traps; therefore, measures of activity-density were only reported for natural enemies and select taxa within this guild. Activity-density was reported as a density of individuals per sample (i.e., mean number per trap).

Landscape Measurements.

To limit sources of variation beyond the land-use types, we selected four sites for study that shared similar surrounding landscape. To confirm that the landscape around the study sites were similar, the land-cover types were quantified by identifying land- cover types at a 2 km radius surrounding each site with CropScape using the 2007

94 Cropland Data Layer (CDL) (USDA NASS RDD 2007). Among the landscapes

surrounding each study site, a total of 15 land cover types were identified by the CDL.

Agricultural land-cover categories included corn, soybean, oats (Avena sativa L.), alfalfa

(Medicago sativa L.), “other hay/non-alfalfa,” and “other crops” (vegetables, fruits,

herbs). Non-agricultural land cover included deciduous forest, herbaceous grassland,

woody wetland, herbaceous wetland, clover/wildflower, and turf/sod. Non-vegetation

land cover included water, urban development, and barren areas. Simpson’s diversity

index was then used to estimate the diversity of land cover surrounding each site based

on the number and amount (area in ha) of all identified categories listed above (Simpson

1949). Simpson’s index values were similar (ranging from 0.41 to 0.53) among sites, as

the dominate land cover surrounding all was corn and soybean, together accounting for

71 to 87% of surrounding land cover. These estimates were used on a one-time basis to

identify any outstanding differences among our sites. Since we did not observe

significant differences in the land cover among our sites, site was defined as a random

blocking factor in the statistical analysis.

Statistical Analyses.

To test multiple hypotheses related to the abundance and diversity of pollinators and natural enemies across space (i.e. among and within land-use types) we analyzed data

collected via vacuum sampling. To test hypotheses related to the activity-density of

natural enemies, we analyzed data collected via yellow sticky traps. Statistical analyses

were performed using mixed model analysis of variance (ANOVA), and to account for

unbalanced data the Kenward-Roger option was used to approximate denominator

95 degrees of freedom (PROC MIXED, SAS software version 9.3 SAS Institute 2010). A

partially nested (hierarchical) linear mixed effects model was used to analyze these data.

The model included fixed effects of land-use type (organic farm, buffer strip,

conventional row crop), distance (specifically chosen sampling points nested within

organic farms and row crops at increasing distances from buffer strips), time (sampling

month), and the interaction of field and time. Blocks (four study sites) and the interaction

of block and sampling points were included as random effects. To test the null

hypothesis that there was no difference in abundance, diversity, and activity of beneficial

arthropods among organic farms, buffer strips, and row crops, we used sample means, calculated across all replicate observations pooled over time. To test the null hypothesis that there was no difference in abundance, diversity, and activity of beneficial arthropods among distances, we calculated means at each level of distance within organic farms and row crops. Specific comparisons of means for fixed effects were tested with t-tests generated through pairwise comparisons of multiple means performed using least squares means (LS-means) analyses with Tukey-Kramer adjustments. Initially year was included in the model, but to account for variation observed between study years, we performed subsequent analyses separately by year.

Results

Arthropod Community Composition.

Using the vacuum, we collected 56,818 arthropods in 2010 based on season-long

totals (i.e. summed across all samples). In 2011, the number of individuals collected in

96 vacuum samples decreased by more than half that observed during the previous year with a season-long of 28,279 arthropods. Despite the noticeable decrease in the season-long total number arthropods collected during the second year of our study, arthropod abundance calculated at the level of guild remained proportionally consistent relative to the total abundance observed in each year. In 2010, the arthropod community was primarily composed of herbivores, which accounted for 68% of the total abundance.

Beneficial arthropod guilds (including natural enemy and pollinator guilds combined) accounted for 26% of the total community. Detritivores, fungivores, and “other” arthropod groups accounted for the remaining 6% of the 2010 season-long total. In 2011, herbivores remained the dominant guild accounting for 55% of the total abundance.

Despite the lower abundance of total arthropods captured in 2011, the percentage of arthropods belonging to beneficial guilds (22%) was proportionally similar to that observed in 2010. Detritivores, fungivores, and “other” groups comprised the remaining

23% of the season-long total observed in 2011 vacuum samples. Due to the variations between years in the community composition and activity-density across arthropod guilds in both vacuum and yellow sticky trap samples, beneficial arthropod communities and their relationship to the different aspects of our study system were analyzed and reported separately for each year. Furthermore, describing the arthropod communities within beneficial guilds separately for each year may reveal certain members of the community that are particularly robust or sensitive to year-to-year variation in our study system.

97 Pollinator Community Composition.

Using the vacuum, we collected 2,171 pollinators during our study. Syrphids accounted for the greatest proportion of the pollinator community compared to bees in

2010 (bees 22% and syrphids 78%) and 2011 (bees 31% and syrphids 69%). Overall, we collected a total of 519 bees and 1,652 syrphids across all three land-use types during both years. The abundance of pollinators was greater in 2010 than 2011 (bees: F = 27.68; df =

1, 287; P < 0.000, syrphids: F = 55.75; df = 1,287; P < 0.0001) per vacuum sample. The percent decrease in pollinators abundance was noticeably larger for syrphids, with a 69% decrease in the second year of our study, compared to a 56% decrease observed for bees.

The abundance of pollinators varied significantly across land use types in 2010

(bees: F = 10.69; df = 2, 143; P = 0.0124, syrphids: F = 38.82; df = 2,143; P < 0.0001) and 2011(bees: F = 11.96; df = 2, 143; P = 0.0081, syrphids: F = 24.66; df = 2,143; P <

0.0001). Over the course of the study, the greatest numbers of pollinators (1,451) were collected from buffer strips. Pollinator abundance was lowest in row crops in each year, especially for bees, with only 20 individual bees observed in row-crop fields during the entire study. As a result, the mean bee abundance in row crops equates to a fraction of a bee per sample (Figure 1). On a per sample basis, bee and syrphid abundance in buffer strips and organic farms ranged from three to over 10 times that observed in row crops.

During 2010, we did not observe significant differences in bee abundance per sample between organic farms and buffer strips. In 2011, we observed significantly more bees in buffer strips compared to all other land use types. In both years, syrphid abundance was significantly greater in buffer strips than all other land-use types (Figure 1).

98 Over the course of our study, we observed a pollinator community composed of

34 taxonomic units representing six families (Table 4). Using the vacuum, we collected a total of 25 bee species and nine syrphid species across all three land-use types during both years. The species richness of pollinators was greater in 2010 than 2011 (bees: F =

33.07; df = 1,287; P < 0.0001; syrphid flies: F = 66.90; df = 1, 287; P < 0.0001).

The pollinator community varied significantly in species richness across land use types in 2010 (bees: F = 6.06; df = 2,143; P = 0.0364, syrphids: F = 33.70; df = 2, 143; P

< 0.0001) and 2011 (bees: F = 9.92; df = 2,143; P = 0.0125, syrphids: F = 6.47; df =

2,143; P = 0.0143). Over the course of the study, the greatest number of pollinator species was observed in organic farms, although this diversity (in terms of richness and evenness) was not always significantly greater than what was observed in the buffer strip

(Table 5). In 2010, bee diversity (both richness and evenness) was significantly greater in organic farms compared to buffer strips and row crops. In 2011, bee species richness was significantly lower in row crops, but we did not observe significant differences in bee species richness between organic farms and buffer strips. We did not observe a significant difference in the diversity of syrphids (both richness and evenness) between buffer strips and organic farms in either year. In both years, row crops had the lowest diversity of syrphids.

The evenness values, especially in 2011, did not indicate that the pollinators were composed of species equally abundant among the land use types (Table 4). The evenness values for bee and syrphid communities observed in row crops were approaching or equal to zero, suggesting random occurrences of species (i.e. a community composed mostly of

99 singletons), however we did not observe significant differences in evenness among all land-use types (Table 4).

Over the course of the study, bee species belonging to the family Halictidae dominated the pollinator community and accounted for 76% of the total bee abundance.

Not only were halictids the most abundant among bees, but also the most diverse among all pollinators. We observed more halictids (17 species) than the total number of species in all the other pollinator families combined. For bees, the other families we collected were represented by a minimum of one and maximum of three different species belonging to those families. Similarly, the syrphid community had two dominant species,

Toxomerus marginatus (Say) and Toxomerus geminatus (Say), which accounted for 59% of the total syrphid abundance. Despite the close proximity of the three land-use types sampled, only five of the 25 bee species collected were observed in all three land use- types; however, the majority of syrphid species were shared among all land-use types

(Table 4). Bees and syrphid species collected from row crops were mostly singletons.

Additionally, unlike the organic vegetable fields and buffer strips where there were several species unique to these areas, all species found in row crops were also observed in at least one other land-use type.

Natural Enemy Community Composition.

Using vacuum sampling, we collected 17,562 natural enemies during our study.

Predators accounted for the greatest proportion of the natural enemy community compared to parasitoids in 2010 (predators 69% and parasitoids 31%) and 2011

(predators 73% and parasitoids 27%). Overall, we collected a total of 12,342 predators

100 and 5,220 parasitoids across all three land-use types during both years. The abundance of

natural enemies was greater in 2010 than 2011 (predators: F = 19.97; df = 1,287, P <

0.0001, parasitoids: F = 11.80; df = 1,287; P < 0.0001) per vacuum sample. In total, 50%

fewer natural enemies were observed in 2011 than 2010.

The abundance of natural enemies varied significantly across all land-use types in

2010 (predators: F = 39.22; df = 2,143, P = 0.0004; parasitoids: F = 17.24; df = 2,143; P

= 0.0033) and 2011 (predators: F = 17.37; df = 2,143; P = 0.0032, parasitoids: F = 12.94;

df = 2,143; P = 0.0067). Over the course of the study, the greatest numbers of natural

enemies (8,334) were collected from buffer strips. On a per sample basis, predator and

parasitoid abundance observed in buffer strips and organic farms ranged from two to 12

times greater than that observed in row crops. In both years, we observed natural enemy

abundance that was significantly greater in buffer strips compared to all other land-use

types in both years (Figure 2).

Over the course of our study, we observed a natural enemy community composed

of 110 taxonomic units representing 50 families (Table 6). Using the vacuum, we

collected 53 species (representing 30 families) of predators and 69 parasitoid species

(representing 24 families) across all three land-use types during both years. The species richness of natural enemies was greater in 2010 than 2011 (predators: F = 19.85; df = 1,

479; P < 0.0001; parasitoids: F = 37.61; df = 1,479; P < 0.0001). The majority of species

(110 of 113) comprising the natural enemy community observed throughout the study were accounted for in 2010.

The natural enemy community varied significantly in species richness across land- use types in 2010 (predators: F = 7.37; df = 2,143; P = 0.0242, parasitoids: F =

101 11.86; df = 2,143; P = 0.0082) and 2011 (predators: F = 10.53; df = 2,143; P = 0.0109, parasitoids: F = 11.88; df = 2,143; P = 0.0082). Over the course of the study, the greatest number of natural enemy species was observed in organic farms, although this diversity

(in terms of richness and evenness) was not always significantly greater than what was observed in buffer strips (Table 5). In both years, predator species richness was significantly greater in organic farms compared to buffer strips and row crops. On a per sample basis, there were two to four times the number predator and parasitoid species in samples collected from organic farms than observed in samples from buffer strips and row crops. Parasitoid species richness was also significantly greater in organic farms compared to all other land-use types in 2010; however, we did not observe significant differences in parasitoid species richness across all land-use types in 2011.

Throughout the study, we observed natural enemy communities with evenness that varied significantly among the land use types in 2010 (predator: F = 8.45; df = 2,143;

P = 0.0180, parasitoids F = 8.49; df = 2,143; P = 0.0178) and 2011 (predators: F = 10.53; df = 2,143; P = 0.0109, parasitoids: F = 10.66; df = 2,143; P = 0.0106). In both years, predator and parasitoid evenness was significantly lower in row crops compared to organic farms and buffer strips. The evenness values for natural enemy communities observed in row crops were approaching or equal to zero, indicating random occurrences of species (Table 5).

Over the course of the study, the most abundant predators were spiders

(Arachnidia: Araneae), Condylostyulus spp. (Diptera: Dolichopodidae), and Orius insidiosus (Say) (Hemiptera: Anthocoridae). The most abundant parasitoids were braconids (Hymenoptera: Braconidae) in the subfamily Microgastrinae and tachinids

102 (Diptera: Tachinidae). When summed together, the abundance of these five natural

enemies accounted for 48% of the total natural enemy abundance observed throughout

the study. Of these natural enemies, the predators occurred in all three land-use types

with the greatest abundance consistently observed in buffer strips in both years.

Microgastrinae spp. were only collected in 2010 across all land-use types. The most

species rich family of predators was lady beetles (Coleoptera: Coccinellidae), with eight

species observed throughout the study. In each year, the greatest number of coccinellid

species were found in samples collected in organic farms; however, the abundance of

each species was generally low. Braconids comprised the most species rich family of

parasitoid natural enemies with more than 20 species/morphospecies observed throughout

the study. Braconids also were consistently more species rich in samples collected in

organic farms. These natural enemies are mostly generalists that feed on several species

within, or among, different orders.

Activity-density.

Using yellow sticky traps, we collected a total of 31,749 arthropods throughout

the study (Table 7). Herbivore and beneficial arthropod guilds were proportionally

equivalent based on season-long totals across all land use types in 2010 (herbivores 48%

and beneficial groups 52%) and 2011 (herbivores 41% and beneficial groups 59%). Only

12 total bees were captured with this method; therefore, our analyses focused on natural enemies. The density of natural enemies observed on yellow sticky traps was greater in

2011 than 2010 (predators: F = 26.94; df = 1, 287 P < 0.0001, parasitoids: F = 54.27; df

= 1, 287 P < 0.0001). Unlike the abundance of arthropods estimated by the vacuum,

103 which indicate a decline from 2010 to 2011, the activity-density of arthropods increased

from 11,169 in 2010 to nearly two times that (20,580) on yellow sticky traps in 2011.

The activity-density of natural enemies varied significantly across all land-use types in 2010 (predators: F = 12.57; df = 2,143; P = 0.0071, parasitoids: F = 5.33; df =

2,143; P = 0.0467) and 2011 (predators: F = 20.56; df = 2,143; P = 0.0021, parasitoids: F

= 19.19; df = 2,143; P < 0.0001). Natural enemy activity-density was significantly greater in buffer strips in both years compared to all other land-use types (Figure 3). On a per sample basis, the number of predators and parasitoids observed in buffer strips was two to nine times greater than that observed in either organic farms or row crops.

Distance From Buffer Strip.

Contrary to our hypothesis, the abundance of pollinators did not vary with sample location for vacuum sampling in 2010 (bees: F = 1.61; df = 3,106; P = 0.0695, syrphids:

F = 0.41; df = 3,106; P = 0.6648) or 2011 (bees: F = 0.43; df = 3,106; P = 0.6541, syrphids: F = 0.71; df = 3,106; P = 0.3070). We expected the abundance of pollinators within organic farms and row crops to be greater, at sampling points closer to the edge of buffer strips, however these results show that differences could not be detected at this scale. The same was observed for pollinator species richness (bees: F = 0.46; df = 3,106;

P = 0.7640, syrphids: F = 0.85; df = 3,106; P = 0.4293) or 2011 (bees: F = 0.11; df =

3,106; P = 0.8922, syrphids: F = 0.08; df = 3,106; P = 0.9267). Contrary to our hypothesis, the abundance of natural enemies did not vary with sample location in vacuum samples in 2010 (predators: F = 1.19; df = 3,106; P = 0.3070, parasitoids: F =

0.95, df = 3,106, P = 0.3883) or 2011 (predators: F = 2.27; df = 3,106; P = 0.1087,

104 parasitoids: F = 1.03, df = 3,106, P = 0.3169). Natural enemy activity-density did not vary with sample location in yellow sticky trap samples in 2010 (predators: F = 0.17; df =

3,106; P = 0.8467, parasitoids: F = 0.23, df = 3,106, P = 0.7920) or 2011 (predators: F =

0.15; df = 3,106; P = 0.8600, parasitoids: F = 0.19, df = 3,106, P = 0.8300). We expected the abundance and activity-density of natural enemies within organic farms and row crops to be greater, at sampling points closer to the edge of buffer strips, however these results show that differences could not be detected at this scale. The same was observed for natural enemy species richness (predators: F = 0.27; df = 3,106; P = 0.8460, parasitoids: F = 0.70; df = 3,106; P = 0.4999) or 2011 (predators: F = 1.02; df = 3,106; P

= 0.3624, parasitoids: F = 3.17; df = 3,106; P = 0.0941).

Discussion

In general, the abundance of beneficial insects was greatest in buffer strips, intermediary in the organic farms and lowest in the conventional row crop fields. This trend was strongest and most consistent for natural enemies, in which the difference between the buffer strips and either crop type was often several times the difference between each crop. The difference in pollinator abundance was not as great; in fact, during 2010 we observed similar abundance of bees in organic farms and buffer strips. One curious observation between the two sampling methods is the difference in the year-to-year trends. Based on vacuum sampling we observed a decrease in natural enemy abundance between 2010 and 2011, however based on yellow sticky traps we observed an increase in natural enemy activity density from 2010 to 2011. To what extant the actual

105 abundance varied between these two years is not clear, but these results do suggest that the sampling methods are reporting different aspects of the natural enemy communities.

Despite these differences, the general conclusion from both methods is that more beneficial arthropods were found in the buffer strips than either crop type.

In many studies, comparisons between organic and conventional land-use have only considered the relationship between beneficial insects and the cropped areas or did not distinguish between cropped areas and non-crop margins. If one considers the buffer strips evaluated in our study, as a component of the organic farm, then these results are consistent with previous studies that indicate that organic farms possess greater biodiversity than conventional farms. However, the abundance of this biodiversity was greatest in buffer strips on these organic farms. To what extent these buffer strips are a source of this biodiversity for organic farms is not clear, as the difference in the diversity of beneficial insects among the three land-use types was not as clear as the differences in abundance. Although the diversity of beneficial insects was lowest in the row crop, we occasionally observed greater species richness (depending upon the year and guild) in the organic farm then the buffer. This pattern is counter to our initial hypothesis

Krauss et al. (2011) stated that insect diversity and abundance often differs between field edges and field centers and suggested that considering these elements separately may be valuable in determining relationship among biodiversity, land-use, and ecosystem services. In their study, the diversity and abundance of pollinators, abundance of predators, and predator-prey ratios were consistently greater in organic farms than conventional farms, and within each land-use type, on filed edges compared to centers

(see also; Holzschuh et al. 2011). Furthermore, Holzschuh et al. (2011) report the

106 benefits associated with organic farms was positively associated with bee species

richness and abundance in fallow strips at both local and landscape scales. These studies

are among the first to report (in relation to reviews in Hole et al. [2005] and Bengtsson et

al. [2005]), comparisons between organic and conventional land-use, that also account for differences between crop and adjacent non-crop elements. However, in each of these studies, field sites per land-use type were separate entities (i.e., organic farms and conventional farms did not share edges and paired land-use types were separated by a range of distances). Similarily, by including buffer strips as a separate land-use component of our study we observed a factor that contributes to this biodiversity. Our results suggest the practice of using a buffer strip between the organic and conventional farm is essential if organic farms are to be considered a source of biodiversity within an agroecosystem.

Other studies show non-crop habitat in close proximity to cultivated fields consists of resources important in determining the abundance and diversity of beneficial arthropods in agricultural landscapes (Landis et al. 2000, Bianchi et al. 2006, Isaacs et al.

2009). Our results suggest that buffer strips can be such a source of such non-crop vegetation in Iowa. Additionally, our results suggest that diverse cropping systems found on organic farms can also be important resources for beneficial arthropods, especially regarding our results pertaining to pollinators. Overall, pollinators were more abundant in buffer strips, but greater pollinator diversity was observed on organic farms. Several bee species were unique to organic farms (9), but the majority of species were shared between organic farms and buffer strips. By contrast, none of the bees collected during our study were unique to row crops and were also observed in at least one other land-use

107 type in either or both years. One explanation for this is that bees were using both buffer strips and row crops throughout the season. Buffers varied at each farm in the abundance, diversity, bloom time, and configuration of flowering plants. The floral resources from crops on organic farms may have been more attractive to pollinators, especially bees. The bees species observed in this study were mostly generalists and with foraging distances that would not limit them to one land use types considering the scale of our study (Gathmann and Tscharntke 2002, Zurbuchen et al. 2010). Although non- crop flowering plants may have been available in buffer strips, bees exhibit the behavior of flower constancy, in which they prefer to focus their attention on one type of flower during a period of time (Waser 1986, Waser and Ollerton 2006) allowing efficient collection of floral rewards. At times, flowering crops may have been preferred by bees because the configuration of large patches of a single flowering species facilitates flower constancy and this configuration of flowering plants was not always available in buffer strips. With the results of this study, it is evident that bees observed in our study utilized both organic farms and buffer strips more than row crops. However, a diverse community of bees has been observed in Iowa’s soybean fields during soybean flowering stages (personal observation) and the abundance of bees in row crops observed in this study could be underestimated; more intense sampling during times when row crops are flowering or shedding pollen could provide greater insight on how pollinators are using row crops. Several species of halictid bees, Agapostemon virescens F., Halictus confuses

Smith, and Lasioglossum (Dialictus) spp. were abundant in both years and found in all three land-use types suggesting that these species many be more robust to disturbances and year to year variation, or are simply more abundant, in agricultural landscapes

108 The species richness of syrphids has shown to be significantly related to the abundance of flowering plants and semi-natural habitat within 500 and 1000 m, but these effects were evident only where there were large areas of semi-natural habitat (Kleijn and

Van Langevelde 2006). This may mean that buffer strips are not large enough to sustain syrphids on agricultural land; however, we observed that syrphids were significantly more abundant in buffer strips than in either crop type. Despite their abundance in buffer strips, surprisingly, the number of syrphid species was equivalent in buffer strips and row crops and only greater in organic farms by one species. This may be explained by the availability of other resources such as flowering crops and oviposition sites near larval food sources found in organic farms and row crops. Overall, this contributes to evidence that buffer strips benefit pollinators by providing resources when others are unavailable in nearby crops and buffer strips also may act as a source of pollinators when the resources in cropped fields are accessible.

We consistently observed the greatest abundance and activity-density of natural enemies in buffer strips throughout this study. Many other studies have noted that natural enemies are abundant, but also diverse in non-crop field margins (Dennis and Fry 1992,

Thies and Tscharntke 1999, Grez and Prado 2000). By contrast, we did not consistently observe a more diverse natural enemy community in buffer strips. These differing results may be a product of the varying diversity among cropping systems as many of these comparisons are between field margins and crops grown in monoculture. One explanation for greater natural enemy species richness observed on organic farms rather than buffer strip is that the biodiversity on organic farms in our study offer resources to a greater number of natural enemy species. This agrees with studies and metaanalyses that

109 show biodiversity increases on organic farms (Bengtsson et al. 2005, Hole et al. 2005,

Ponce et al. 2011).

A common theme among these studies is that natural enemy abundance in field

margins can result higher densities in adjacent crop (Dennis and Fry 1992, Thies and

Tscharntke 1999, Grez and Prado 2000). In this way, our study also suggests that buffer

strips may be a source of beneficial insects colonizing adjacent crops. Many of the most

abundant predators vacuum collected from vegetation in buffer strips were also abundant

in crop vegetation including exotic lady beetle, Harmonia axyridis (Pallas), green lacewings (Chrysopidae: Neuroptera), and tachinids (Table 6 and 7). Furthermore, the activity-density of these species was evident in all three land-use types suggesting that these species are moving among the adjacent fields (Table 7). This was not the case for some of the other most abundant natural enemies particularly with Orius insidiosus, being abundant in vacuum samples across all three land-use types, but virtually absent in measures of activity-density in yellow sticky traps.

Despite the apparent benefits of the biodiversity within organic farms and buffer strips, the low abundance and very low evenness of the beneficial insect community observed in the row crop, indicated that the movement of this community from buffer strips to row crops is limited. Furthermore, we observed significant year-to-year variation in beneficial insects communities. This variation suggests that improvements can be made to ensure long-term benefits from buffer strips by enhancing the resilience of beneficial insect communities. In Iowa’s simplified landscape the pre-existing non- crop habitat may not be composed of the optimal density of flowering plants that together provide floral resources throughout the growing season. The presence and bloom period

110 of flowering plants was inconsistent among buffer strips observed in this study (personal observation), making these particular plant communities suboptimal for beneficial arthropods. Efforts to increase the abundance and diversity of beneficial insects within this agroecosystem could include the use of plant communities with more suitable plants, particularly native species that bloom throughout the season (Landis et al. 2000, Isaacs et al. 2009). Previous research has evaluated and rated plants on an individual basis for their attractiveness to beneficial arthropods as well as arthropod plant pest (Fiedler and

Landis 2007, Tuell et al. 2008). Many of these plants are prairie plants (forbs and grasses) native to Iowa.

With the research presented in chapter two, we identified plant communities designed with mixtures of native prairie plants in which beneficial insects were consistently more diverse and abundant than plant communities commonly found in

Iowa’s buffer strips. We propose that reestablishing these plants in buffer strips can improve quality of non-crop plant communities to better conserve beneficial insects. The quality of the plant resources (e.g. forage and nesting sites) in non-crop areas has shown to be important for pollinators (Pots et al. 2003; 2005). Studies in mixed prairies have shown that the predator-to-prey ratios increase with plant diversity (Haddad et al. 2009) and Gardiner et al. (2010) found that coccinellid diversity was positively correlated with floristic diversity in reconstructed mixed prairies.

In addition to the composition of buffer strips, efforts to improve the configuration of plant communities are more complicated. Fragmented landscapes have been reported to cause declines in diversity (Klein et al. 2002, Steffan-Dewenter et al.

2002). Although beneficial insect populations may be enhanced by farming and habitat

111 restoration practices at the field scale, surrounding habitats and landscape-level

influences may be controlling the population dynamics of species existing at these small

scales. This theory may account for the differential results pertaining to abundance and

diversity in our study because species-dependent responses have been shown to vary in

agricultural landscapes that differ in levels of landscape connectivity (Williams and

Kremen 2007). Studies further demonstrate that this phenomenon is restricted to

simplified landscapes and composition of local communities, such as those we observed

on organic farms, are greatly influences by surrounding landscapes, which acts a source

pool for local species (Bengtsson et al. 2005, Hole et al. 2005, Ponce et al. 2011).

In conclusion, this study provided insight on beneficial arthropod communities across different land use types. Developing best practices that are compatible with modern and changing landscapes is important for improving buffer strips and safeguarding the ecosystem services provided by beneficial insects. Practices that favor beneficial insects may be particularly valuable if organic crop production continues to increase in Iowa, because of the economic and ecological benefits organic farmers can receive through the delivery of inset-derived ecosystem services. However, developing best practices that conserve beneficial arthropods are not limited to Iowa’s organic farmers.

112 References Cited

Altieri, M. A. 1999. The ecological role of biodiversity in agroecosystems. Agr. Ecosys. Environ. 74: 19-31.

Bengtsson, J., J. Ahnström, and A. Weibull. 2005. The effects of organic agriculture on biodiversity and abundance: a meta-analysis. J. Appl. Ecol. 42: 261-269.

Caballero-López, B., R. Bommarco, J. M. Blanco-Moreno, F. X. Sans, J. Pujade- Villar, M. Rundlöf, H. G. Smith. 2012. Aphids and their natural enemies are differently affected by habitat features at local and landscape levels. Biol. Control. 63: 222-229.

Dennis, P. and G. L. A. Fry. 1992. Field margins: Can they enhance natural enemy population densities and general arthropod diversity on farmland? Agr. Ecosys. Environ. 40:95-115.

Fiedler, A. K., and D. A. Landis. 2007. Attractiveness of Michigan native plants to arthropod natural enemies and herbivores. Environ. Entomol. 36: 751-765.

Gathmann, A. and T. Tscharntke. 2002. Foraging ranges of solitary bees. J. Anim. Ecol. 71: 757-764.

Grez, A. A. and E. Prado. 2000. Effect of plant patch shape and surrounding vegetation on the dynamics of predatory coccinellids and their prey (Brevicoryne brassicae (Hemiptera: Aphididae). Environ. Entomol. 29: 1244-1250.

Haddad, N. M., G. M. Crutsinger, K. Gross, J. Haarstad, J. M. H. Knops, D. Tilman. 2009. Plant species loss decreases arthropod diversity and shifts trophic structure. Ecol. Lett. 12:1029-39.

113 Hawkins, B.A., N. J. Mills, M. A. Jervis, and P. W. Price. 1999. Is the biological control of insects a natural phenomenon? Oikos 86: 493-506.

Hole, D. G., A. J. Perkins, I. H. Alexander, P. V. Grice, and A. D. Evans. 2005. Does organic farming benefit biodiversity? Biol. Conserv. 122:113-130.

Iowa Department of Agriculture and Land Stewardship (IDALS). 2011. Iowa agricultural quick facts. Available online: http://www.iowaagriculture.gov/quickfacts.asp. Last visited on 14 January 2013.

Kevan, P. G. and H. G. Baker. 1983. Insects as flower visitors and pollinators. Annu. Rev. Entomol. 28: 407-453.

Kleijn, D. and F. van Langevelde. 2006. Interacting effects of landscape context and habitat quality on flower visiting insects in agricultural landscapes. Basic and Appl. Ecol. 7: 201-214.

Kremen, C., N. M. Williams, M. A. Aizen, B. Gemmill-Harren, G. LeBuhn, R. Minckley, L. Packer, S. G. Potts, T. Roulston, I. Steffan-Dewenter, D. P. Vazquez, R. Winfree, L. Adams, E. E. Crone, S. S. Greenlead, T. H. Keitt, A. M. Klein, J. Regetz, and T. H. Ricketts. 2007. Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecol. Lett. 10: 299-314.

Landis, D. A., F. D. Menalled, A. C. Costamagna, and T. K. Wilkinson. 2005. Manipulating plant resources to enhance beneficial arthropods in agricultural landscapes. Weed Sci. 53: 902-908.

114 Landis, D. A., S. D. Wratten, and G. M. Gurr. 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Annu. Rev. Entomol. 45: 175-201.

Losey J. E., and M. Vaughan. 2006. The economic value of ecological services provided by insects. Bioscience. 56: 311-323.

Mugurran, A. E. 2004. Measuring biological diversity. Blackwell Publishing company, Malden, MA, USA.

Millennium Ecosystem Assessment. 2005. Ecosystems and human well-being: Biodiversity synthesis. World Resources Institute, Washington, DC.

Menalled, F. D., P. C. Marino, S. H. Gage, and D. A. Landis. 1999. Does agricultural landscape structure affect parasitism and parasitoid diversity? Ecol. Appl. 9: 634-641.

Morse, R. A. and N. W. Calderone. 2000. The value of honey bees as pollinators of U. S. crops in 2000. Bee Culture. 128: 15 pp insert.

Ponce, C., C. Bravo, D. García de León, M. Magaña, and J. Carols Alonso. 2011. Effects of organic farming on plant and arthropod communities: A case study in Mediterranean dryland cereal. Agr. Ecosys. Environ. 141: 193-201.

Potts, S. G., B. Vulliamy, A. Dafni, G. Ne`eman, and P. Wilmer. 2003. Linking bees and flowers: how do floral communities structure pollinator communities. Ecology. 84: 2628-2642.

R Development Core Team. 2011. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. (http://www.R- project.org).

Rebek, E. J., C. S. Sadof, and L. M. Hanks. 2005. Manipulating the abundance of natural enemies in ornamental landscapes with floral resource plants. Biol. Control. 33: 203-216.

Ricketts, T. H., J. Regetz, I. Steffan-Dewenter, S. A. Cunningham, C. Kremen, A. Bogdanski, B. Gemmill-Herren, S. S. Greenleaf, A. M. Klein, M. M. Mayfield, L. A.

115 Morandin, A. Ochieng, and B. F. Viana. 2008. Landscape effects on crop pollination services: are there general patterns. Ecol. Lett. 11: 499-515.

SAS Institute. 2010. SAS/STAT, Version 9.2. SAS Institute, Cary, NC.

Schmidt, N. P., M. E. O’Neal, and P. M. Dixon. 2008. Aphidophagous predators in Iowa soybean: a community comparison across multiple years and sampling methods. Ann. Entomol. Soc. Am. 101: 341-350.

Simpson, E. H. 1949. Measure of diversity. Nature. 163: 688-688.

Steffan-Dewenter, I., U. Munzenberg, C. Burger, C. Thies, and T. Tscharntke. 2002. Scale-dependent effects of landscape context on three pollinator guilds. Ecology. 83: 1421:1432.

Ssymank, A., C. A. Kearns, T. Pape, and F. C. Thompson. 2008. Pollinating flies (Diptera): a major contribution to plant diversity and agricultural production. Biodiversity. 9: 86-89.

Thies, C. and T. Tscharntke. 1999. Landscape structure and biological control in agroecosystems. Science. 285: 893:895.

Tooker, J. F., M. Hauser, and L. M. Hanks. 2006. Floral host plants of Syrphidae and Tachinidae (Diptera) of Central Illinois. Ann. Entomol. Soc. A. 99: 96-112.

116 Tscharntke, T., Y. Clough, T. C. Wagner, L. Jackson, I. Perfecto, J. Vandermeer, A. Whitbread. 2012. Global food security, biodiversity conservation and the future if agricultural intensification. Biol. Conserv. 151: 53-59.

(USDA NASS RDD) United States Department of Agriculture National Statistics Service, Research and Development Division. 2007. Cropland data layer. Available online: http://www.nass.usda.gov/research/Cropland/SARS1a.htm. Last visited on 12 December 2012.

(USDA NOP) United States Department of Agriculture National Organic Program 2009. National Organic Program, Production and Handling Preamble. Available online http://www.ams.usda.gov/AMSv1.0/ams.fetchTemplateData.do?template=TemplateN&n avID=NOSBlinkNOSBMeetings&rightNav1=NOSBlinkNOSBMeetings&topNav=&left Nav=&page=NOPOrganicStandards&resultType=&acct=nopgeninfo. Last visited on 27 June 2012.

(USDA NRCS PLANTS) United States Department of Agriculture, Natural Resource Conservation Service, PLANTS Database. 2013. National Plant Data Team, Greensboro, NC. Available online: http://plants.usda.gov. Last visited on 4 January 2013.

Wackers, F. L., P. C. J. van Rijn, G. E. Heimpel. 2008. Honeydew as a food source for natural enemies: making the best of a bad meal. Biol. Control. 45: 176-184.

Waser, N. M. 1986. Flower constancy: definition, cause, and the nature of species boundaries. Am. Nat. 127: 593.

Waser, N. M. and J. Ollerton. 2006. Plant-pollinator interactions: from specialization to generalization. University of Chicago Press, Chicago, IL, USA.

Williams, N. M. and C. Kremen. 2007. Resource distribution among habitats determine solitary bee offspring production in a mosaic landscape. Ecol. Appl. 17: 910- 921.

Zhang, W., Ricketts T. H., C. Kremen, K. Carney, and S. M. Swinton. 2007. Ecosystem services and dis-services to agriculture. Ecol. Econ. 64: 253-260.

Zurbuchen, A., L. Landert, J. Klaiber, A. Müller, S. Hein, and S. Dorn. 2010. Maximum foraging range in solitary bees: only few individuals have the capability to cover long foraging distances. Biol. Conserv. 143: 669-676.

117 Table 1. Organic farm characteristics County Coordinates Area (ha)a Yearsb Certifying agencyc Dallas 41°46'39.76''N 93°59'50.50''W 4 11 IDALS Polk 41°45'23.69''N 93°48'44.67''W 8 17 IDALS Poweshiek 41°45'34.39''N 92°42'39.97''W 32 6 MOSA Story 41°58'44.57''N 93°31'54.25''W 1 14 -- aTotal area of parcel is the area devoted to organic crop production ranges from 0.8 to 4 ha at any one time during the growing season and other portions of land were used for livestock, pasture, cover crops, and fallow. bYears operating refers to the number of years organic farms were under current ownership and includes periods of organic management before and after certification or recertification dates. cCertifying agencies include IDALS, Iowa Department of Agricultural and Land Stewardship and MOSA, Midwest Organic Services Association. No certification indicated by (--). Organic farms obtained certification one or more years prior to the study, but all were managed organically at least three years prior to certification.

118 Table 2. Crops present in organic farms and row crop fields shown as the percent of occurrence among four sites Common Name Genusa Organic farm (%)b Row crop (%)b Arugula Eruca Mill. 75 -- Asparagus Asparagus L. 100 -- Beet Beta L. 100 -- Broccoli Brassica L. 100 -- Brussels sprouts Brassica L. 75 -- Bok choy Brassica L. 75 -- Cabbage Brassica L. 100 -- Carrot Daucus L. 100 -- Cauliflower Brassica L. 100 -- Celery Apium L. 25 -- Collards Brassica L. 75 -- Cornc Zea L. 50 -- Cornd Zea L -- 50 Cucumber Cucumis L. 100 -- Daikon Raphanus L. 50 -- Eggplant Solanum L. 100 -- Garlic Allium L. 100 -- Ginger Zingiber Mill. 25 -- Gourd Cucurbita L. 25 -- Green bean Phaseolus L. 100 -- Green onion Allium L. 100 -- Hop pepper Capsicum L. 75 -- Kale Brassica L. 75 -- Kohlrabi Brassica L. 100 -- Leek Allium L. 100 -- Lettuce Lactuca L. 100 -- Mustard greens Brassica L. 50 -- Okra Abelmoschus (L.) Moench 50 -- Onion Allium L 100 -- Parsnip Pastinaca L. 50 -- Pea Pisum L. 100 -- Potato Solanum L. 100 -- Pumpkin Cucurbita L. 75 -- Radish Raphanus L. 100 -- Raspberry Rubus L. 75 -- Rhubarb Rheum L. 50 --

119 Table 2. continued Common Name Genusa Organic farm (%)b Row crop (%)b Rutabaga Brassica L. 25 -- Shallot Allium L. 75 -- Spinach Spinacia L. 75 -- Soybeanse Glycine L. 50 -- Soybeansf Glycine L. -- 50 Strawberry Fragaria L. 50 -- Summer squash Cucurbita L. 75 -- Sweet pepper Capsicum L. 100 -- Sweet potato Ipomoea (L.) Lam. 75 -- Swiss chard Beta L. 75 -- Tomato Solanum L 100 -- Turnips Brassica L. 100 -- Watermelon Citrullus (Thumb.) Matsum. & Nakai 25 -- Winter squash Cucurbita L. 75 -- Zucchini Cucurbita L. 100 -- Fresh herb mixed 75 -- Fresh flower mixed 75 -- aIncludes multiple varieties within the listed genus. b Numbers in this column refer to the percent occurrence of each cultivated species on organic farms in row crops that contain grew the associated crop type in both years (organic farms) or either year (conventional farms). c Sweet corn d Field corn e Edamame f Field soybean

120 Table 3. Vegetation categories present in buffer strips and the percent of occurrence among four buffer strips Growth habita Occurrence (%)b Coniferous treec 75 Deciduous treed Open canopy 50 Closed canopy 50 Forb/herbe Legume 25 Wildflower 75 Graminoidf Sedge 25 Sod-forming grass 75 Tussock/bunchgrass 100 Other grass 100 Shrubg 75 Subshrubg 100 a Growth habits categorized plant species based on similar characterizes using the USDA PLANTS database. b Percent occurrence is the proportion of sites where plants belonging to corresponding growth habits were observed in quadrat samples within areas of buffer strips sampled during this study. Other types may exist in any of the buffer strips beyond the defined, standardized sampling area. c Coniferous trees occurred individually or in small clusters that did not form closed canopies. dDeciduous tress that occurred individually or in small clusters (open canopy) and in larger stands (closed canopies) e Forb/herbs included leguminous species and non-- leguminous species. Since buffer strips were not consistently managed for weeds, this growth habit also includes species that are considered weedy (e.g., wild parsnip, Pastinaca sativa L.) fGraminoids include cool and warm--season bunchgrass, sod- forming grass / turf, sedges, and other grass-like species with unresolved identifications. f Shrubs are perennial woody species less than 4 to 5 m in height and subshrubs are low-growing shrubs less than 1 m tall.

121 Table 4. Total abundance of pollinator taxa vacuum collected per land use type in 2010 and 2011 Abundance 2010 (2011)a Taxa Organic farm Buffer strip Row crop HYMENOPTERA Andrenidae Andrena wilkella Kirby 5 (3) 0 (24) -- -- Calliopsis andreniformis Smith 5 (1) ------Perdita sp. 1 (0) ------Apidae Apis meliffera L. 13 (0) ------Bombus bimaculatus Cresson 6 (0) 12 (13) -- -- Melissodes trinodus Robertson 2 (0) ------Colletidae Hylaeus affinis (Smith) 1 (0) 12 (0) -- -- Halictidae Agapostemon virescens F. 27 (5) 11 (0) 4 (2) Augochlora pura Say 2 (2) -- -- 1 (0) Augochlorella aurata (Smith) 4 (0) ------Dieunomia sp. -- -- 10 (0) -- -- Halictus confuses Smith 33 (8) 62 (26) 1 (1) Halictus ligatus Say 4 (0) 36 (0) 1 (0) Halictus rebicondus (Christ) 2 (2) 0 (12) 2 (0) Halictus tripartitus Cockerell 0 (1) ------Lasioglossum s.str. sp. 27 (0) -- -- 0 (1) Lasioglossum (Dialictus) spp./groups L. (Dialictus) albipenne (Robertson) 1 (0) ------L. (Dialictus) viridatum (Lovell) 1 (0) ------L. (Dialictus) morpho sp. 1 27 (6) 12 (35) 2 (1) L. (Dialictus) morpho sp. 2 10 (0) 10 (0) 2 (0) L. (Dialictus) morpho sp. 3 4 (0) ------L. (Dialictus) morpho sp. 4 2 (0) ------L. (Dialictus) morpho sp. 5 2 (0) ------Nomia sp. 4 (2) -- -- 0 (2) Megachilidae Megachilie rotundata F. 0 (1) 0 (10) -- --

122 Table 4. continued Abundance 2010 (2011)a Taxa Organic farm Buffer strip Row crop DIPTERA Syrphidae Eristalinaeb Eristalis tenax L. 8 (8) 47 (36) 2(10) Orthonevra nitida (Wiedemann) 8 (0) 62 (0) -- -- Syritta pipiens L. 5 (0) -- -- 1 (0) Syrphinaec Melanostoma mellinum L. 17 (2) 32 (0) 1 (0) Paragus sp. 7 (1) 30 (0) 5 (0) Platycheirus sp. 5 (0) 70 (0) 1 (1) Sphaerophoria sp. 2 (0) 12 (0) 2 (0) Toxomerus geminatus (Say) 39 (8) 40 (156) 57 (17) Toxomerus marginatus (Say) 222 (55) 621 (60) 2 (0) Total bees 183 (31) 165 (120) 13 (7) Total syrphids 313 (74) 914 (252) 7 (28) GRAND TOTAL 601 1451 119 aAbundance is the total number of individual bee or syrphid taxa vacuum collected per land use type per year. Values not enclosed in parentheses are for data collected in 2010 and values within parentheses for 2011. bIncludes species captured as adults that are non-predatory as larvae c Includes species captured as adults that are predators as larvae (see Methods and Materials: Arthropod Identification and Guild Assignment)

Table 5. Total and mean ± SEM species richness and evenness of beneficial insect guilds per vacuum sample across three land use types in 2010 and 2011 2010 2011

Organic farm Buffer strip Row crop Organic farm Buffer strip Row crop

Bee Total taxaa 22 8 7 10 6 5 Richnessb 9 ± 2a 1 ± 0.3b 1 ± 0.5b 2 ± 0.40a 1 ± 0.20a 0.3 ± 0.20b Evennessc 0.8 ± 0.04a 0.5 ± 0.03b 0.1 ± 0.01b 0.1 ± 0.01ab 0.3 ± 0.01a 0b Syrphids Total taxa 9 8 8 5 3 3 Richness 6 ± 0.2a 4 ± 0.2a 3 ± 0.1b 3 ± 0.1a 2 ± 0.1ab 1 ± 0.08b Evenness 0.3 ± 0.03a 0.4 ± 0.06a 0.1± 0.2b 0.3 ± 0.01a 0.1 ± 0.01ab 0b 123 Predators Total taxa 47 37 40 37 21 22 Richness 8 ± 0.40a 4 ± 0.30b 4 ± 0.40b 4 ± 0.30a 2 ± 0.30b 1 ± 0.10b Evenness 0.7 ± 0.03a 0.7 ± 0.04a 0.4 ± 0.04b 0.5 ± 0.04a 0.3 ± 0.06a 0.1 ± 0.02b Parasitoids Total taxa 46 31 41 15 13 12 Richness 5 ± 0.03a 3 ± 0.30b 2 ± 0.23b 1 ± 0.10a 1 ± 0.20a 0.3 ± 0.06b Evenness 0.5 ± 0.03a 0.6 ± 0.06a 0.1 ± 0.03b 0.1 ± 0.02a 0.2 ± 0.06a 0b aTotal taxa are the season-long number of unique taxa per guild observed in each land use type in each year. bRichness (S) is mean species richness within guilds per sample for each land use type and year. cEvenness (Pielou’s evenness index J') is the mean value of J' within guilds for each land use type and year. Values for J' are constrained between 0 and 1, where 1 = a community where species are equally abundant, and 0 = a community where species are randomly distributed (i.e., higher values of J' indicate greater guild diversity within a land use regime).

124 Table 6. Total abundance of pollinators by taxa vacuum collected per land use type for each year Abundance 2010 (2011) Taxa Organic farm Buffer strip Row crop PREDATORS ARANEAE 451 (281) 2281(1512) 209 (88) HEMIPTERA Anthocoridae Orius insidiosus (Say) 171 (60) 256 (204) 143 (10) Berytidae Neoneides muticus (Say) 27 (7) 128 (36) 3 (1) Nabidae Nabis sp. 96 (74) 227 (276) 34 (13) Pentatomidae Podisus maculiventris (Say) 17 (4) 69 (0) 12 (2) Phymantidae Phymanta sp. 11 (2) 10 (0) 3 (1) Reduviidae 4 (16) 58 (24) 2 (1) Saldidae 0 (1) -- -- 0 (1) COLEOPTERA Cantharidae Chauliognathus marginatus F. 27 (27) 33 (192) 4 (5) Chauliognathus pensylvanicus (DeGeer) 81 (32) 195 (0) 3 (0) Cantharis rotundicollis Say 11 (1) 12 (0) 2 (0) Podabrus s.l. 1 (0) 21 (0) -- -- Trypherus latipennis (Germar) 3 (0) 24 (0) -- -- Carabidae Carabidae morpho sp. 1 17 (5) 0 (24) 1 (23) Carabidae morpho sp. 2 1 (0) 12 (0) -- -- Carabidae morpho sp. 3 3 (0) -- -- 1 (0) Carabidae morpho sp. 4 4 (0) ------Cleridae 3 (0) 0 (36) 1 (0) Coccinellidae Brachiacantha ursina F. 2 (0) ------Coccinella septempunctata L. 3 (1) 0 (36) 2 (0) Coleomegilla maculata (DeGeer) 36 (17) 10 (36) 8 (0) Cycloneda munda (Say) 26 (0) 12 (0) 3 (1) Harmonia axyridis (Pallas) 30 (14) 101 (0) 19 (7) Hippodamia convergens Guerin 0 (2) 0 (12) 1 (2) Hippodamia parenthesis (Say) 2 (1) -- -- 1 (0) Hyperaspis undulata (Say) 3 (0) -- -- 1 (0) Lampyridae 35 (3) 46 (120) 18 (4) Melyridae 0 (2) 11 (0) -- -- Staphylinidae 4 (4) 36 (0) 4 (0)

125 Table 6. continued Abundance 2010 (2011) Taxa Organic farm Buffer strip Row crop NEUROPTERA Chrysopidae Chrysoperla sp. 55 (10) 139 (108) 15 (12) Chrysopidae nymph 18 (7) 67 (24) 11 (7) Hamerobiidae 0 (9) 0 (12) 1 (5) DIPTERA Asilidae 1 (1) 22 (24) 3 (1) Dolichopodidae Condylostylus spp. 233 (39) 859 (156) 149 (16) Empididae Empis sp. 20 (14) -- -- 2 (1) Rhagionidae Rhagio morpho sp. 1 7 (0) 10 (0) 12 (0) Rhagio morpho sp. 2 2 (0) 42 (0) 2 (0) Syrphidaea Melanostoma mellinum L. 17 (2) 32 (0) 1 (0) Paragus sp. 7 (1) 30 (0) 5 (0) Platycheirus sp. 5 (0) 70 (0) 1 (1) Sphaerophoria sp. 2 (0) 12 (0) 2 (0) Toxomerus geminatus (Say) 39 (8) 40 (60) 2 (0) Toxomerus marginatus (Say) 222 (55) 621 (156) 57 (17) HYMENOPTERA Crabronidae 2 (3) 22 (0) 2 (0) Pompilidae Pompilidae morpho sp.1 3 (4) 10 (12) -- -- Pompilidae morpho sp.2 0 (1) ------Sphecidae 2 (1) 11 (0) 3 (0) Vespidae 6 (2) ------ODONATA Calopterygidae 6 (2) 10 (12) -- -- Coenagrionidae 153 (77) 55 (0) 13 (0) Libellulidae -- -- 33 (0) -- -- PARASITOIDS HYMENOPTERA Ceraphronoidea Megaspilidae 4 (0) ------Ceraphronidae 1 (0) -- -- 1 (0) Chalcidoidea Chalcidae 0 (39) 12 (24) 2 (2) Encyrtidae Encyrtidae morpho sp. 1 11 (0) 119 (0) 6 (0)

126 Table 6. continued Abundance 2010 (2011) Taxa Organic farm Buffer strip Row crop (Encyrtidae continued) Encyrtidae morpho sp. 2 12 (0) ------Eulophidae Eulophidae morpho sp.1 30 (1) 418 (0) 19 (0) Eulophidae morpho sp.2 6 (0) -- -- 3 (0) Eupelmidae 3 (0) 12 (0) 4 (0) Eurytomidae Eurytomidae morpho sp. 1 47 (0) 56 (0) 1 (1) Eurytomidae morpho sp. 2 7 (0) -- -- 2 (0) Pteromalidae Pteromalidae morpho sp. 1 54 (8) 176 (0) 12 (1) Pteromalidae morpho sp. 2 64 (0) 58 (0) 7 (0) Pteromalidae morpho sp. 3 -- -- 13 (0) -- -- Mymaridae 1 (0) -- --1 (0) Chalcidoid morpho sp. 1 2 (34) 0 (192) 1 (6) Chalcidoid morpho sp. 2 1 (0) 0 (12) 1 (0) Chalcidoid morpho sp. 3 31 (54) 0 (120) 1 (4) Cynipoidea Figitidae 19 (0) 64 (0) 8 (0) Eucoilinae sp. 14 (0) 35 (0) 1 (0) Ichneumonoidea Braconidae Agathininae spp. 8 (0) 46 (0) 6 (0) Aphidius spp. 26 (2) 46 (0) 8 (0) Heterospilus eurostae ------1 (0) Cheloninae sp. 1 (0) ------Macrocentrus sp. 3 (0) ------Microgastrinae spp. 118 (0) 309 (0) 41 (1) Opiinae sp. ------1 (0) Rogadinae sp. 8 (0) 12 (0) 1 (0) Braconidae morpho sp. 1 8 (21) 12 (228) 2 (3) Braconidae morpho sp. 2 4 (4) 48 (72) 2 (1) Braconidae morpho sp. 3 7 (0) 55 (0) 5 (0) Braconidae morpho sp. 4 10 (6) 0 (12) 0 (1) Braconidae morpho sp. 5 -- -- 10 (0) -- -- Braconidae morpho sp. 6 -- -- 10 (0) -- -- Braconidae morpho sp. 7 -- -- 10 (0) -- -- Braconidae morpho sp. 8 1 (0) ------Braconidae morpho sp. 9 1 (0) ------Braconidae morpho sp. 10 14 (62) 12 (408) 3 (4) Braconidae morpho sp. 11 ------3 (0) Braconidae morpho sp. 12 6 (0) ------

127 Table 6. continued Abundance 2010 (2011) Taxa Organic farm Buffer strip Row crop (Ichneumonoidea continued) Ichneumonidae Cremastinae morpho sp.1 ------1 (0) Cremastinae morpho sp.2 ------4 (0) Cremastinae morpho sp.3 2 (0) -- -- 6 (0) Ichneumonidae morpho sp. 1 11 (5) 44 (36) 6 (1) Ichneumonidae morpho sp. 2 1 (2) -- -- 2 (0) Ichneumonidae morpho sp. 3 6 (0) -- -- 2 (0) Ichneumonidae morpho sp. 4 ------2 (0) Ichneumonidae morpho sp. 5 16 (13) 59 (60) 2 (0) Ichneumonidae morpho sp. 6 1 (5) 10 (36) -- -- Ichneumonidae morpho sp. 7 3 (0) 25 (0) -- -- Platygastroidea Platygastridae 2 (0) 12 (0) -- -- Scelionidae spp. 7 (0) 43 (0) 2 (0) DIPTERA Conopidae 23 (0) 256 (0) 8 (0) Tachinidae 171 (5) 643 (12) 47 (5) Other/undet.b 13 (0) 82 (12) 4 (0) Total predators 1887 (790) 5627 (3072) 757 (219) Total parasitoids 770 (261) 2707 (1224) 228 (30) GRAND TOTAL 3698 12,630 1234 aAbundance is the total number of individual predator and parasitoid taxa vacuum collected per land use type per year. Values not enclosed in parentheses are for data collected in 2010 and values within parentheses for 2011. b Includes species captured as adults that are predators as larvae (see Methods and Materials: Arthropod Identification and Guild Assignment) cOther/undetermined refers to taxa including parasitic Apoidea (four species from three families) and parasitic Apocrita (two species from two families). It also includes two (Ichneumonoidea) taxa that could not be identified due to the poor condition of the specimen. There are (at minimum) six unique morphospecies from five families that were collapsed into this category.

128 Table 7. Abundance of natural enemies by taxa and in total collected on yellow sticky traps per land use type for each year Abundance 2010 (2011)a Taxa Organic farm Buffer strip Row crop PREDATORS ARANEAE 9 (14) 56 (100) 11 (8) HEMIPTERA Anthocoridae Orius insidiosus (Say) 0 (2) 0 (8) 1 (3) Nabidae Nabis sp. 0 (1) 8 (2) 0 (1) COLEOPTERA Anthicidae 1 (0) 8 (0) -- -- Cantharidae Chauliognathus marginatus F. -- -- 0 (9) 5 (0) Chauliognathus pensylvanicus (DeGeer) 175 (68) 136 (230) 29 (40) Cantharis rotundicollis Say 1 (0) 0 (15) 14 (2) Podabrus s.l. 1 (0) ------Carabidae 1 (6) 8 (19) 2 (1) Coccinellidae Coccinella septempunctata L. 9 (30) 16 (95) 2 (15) Coccinella trifasciata L. -- -- 8 (0) -- -- Coleomegilla maculata (DeGeer) 10 (5) 8 (64) 24 (26) Cycloneda munda (Say) 8 (4) 128 (13) 5 (0) Harmonia axyridis (Pallas) 99 (30) 224 (91) 42 (11) Hippodamia convergens Guerin 6 (16) 0 (43) 1 (12) Hippodamia parenthesis (Say) 0 (1) 0 (11) -- -- Hippodamia tredecimpunctata L. 1 (0) -- -- 1 (1) Hippodamia variegata Goeze -- -- 1 (0) -- -- Lampyridae 77 (57) 120 (213) 27 (14) Staphylinidae 0 (4) 0 (2) 0 (0) NEUROPTERA Chrysopidae Chrysoperla spp. 63 (48) 216 (215) 30 (65) Hamerobiidae 7 (5) 8 (25) 1 (4) DIPTERA Asilidae 1 (3) 0 (25) 1 (1) Dolichopodidae 244 (519) 552 (1858) 55 (123) Empididae 13 (0) 0 (27) -- -- Syrphidaeb Melanostoma mellinum L. 0 (8) 0 (3) 0 (2) Toxomerus marginatus (Say) 109 (116) 288 (539) 143 (102)

129 Table 7 continued. Abundance 2010 (2011)a Taxa Organic farm Buffer strip Row crop PARASITOIDS HYMENOPTERA Chalcidoidea 292 (773) 1360 (4325) 147 (538) Chalcidae spp. 10 (0) 8 (0) 3 (0) Ichneumonoidea Braconidae Aphidius spp. 1 (0) -- -- 0 (1) Microgastrinae spp. 11 (0) 16 (0) 14 (1) Braconidae spp. 0 (2) 0 (4) 0 (1) Ichneumonidae spp. 49 (18) 112 (63) 31 (13) Cynipoidea Figitidae Eucoilinae sp. -- -- 40 (0) -- -- DIPTERA Conopidae 6 (0) ------Tachinidae 27 (90) 64 (297) 19 (12) Total predators 836 (940) 1784 (3607) 395 (431) Total parasitoids 396 (883) 1600 (4689) 214 (566) GRAND TOTAL 3055 11,680 1,606 aAbundance is the total number of individual predator and parasitoid taxa collected on yellow sticky traps per land use type per year. Values not enclosed in parentheses are for data collected in 2010 and values within parentheses for 2011. bIncludes only species captured as adults that are predators as larvae (see Arthropod Identification and Guild Assignment in Methods and Materials)

130 Figure Legends

Fig 1. Mean ± SEM abundance of bees and syrphids per vacuum sample on organic farms, buffer strips, and row crops in (a) 2010 and (b) 2011. For bees, means are with common lowercase letters are not significantly different. For syrphids, means with common capital letters are not significantly different.

Fig 2. Mean ± SEM abundance of predators and parasitoids per vacuum sample on organic farms, buffer strips, and row crops in (a) 2010 and (b) 2011. For predators, means are with common lowercase letters are not significantly different. For parasitoids, means with common capital letters are not significantly different.

Fig 3. Mean ± SEM activity-density of predators and parasitoids per yellow sticky trap on organic farms, buffer strips, and row crops in (a) 2010 and (b) 2011. For predators, means are with common lowercase letters are not significantly different. For parasitoids, means with common capital letters are not significantly different.

131

Fig. 1

132

Fig. 2

133

Fig. 3.

134 CHAPTER 4. GENERAL CONCLUSIONS

The goal of this research was to develop best-practices for designing buffer strips that attract and conserve beneficial insects in Iowa’s agricultural landscape. To achieve this goal, I conducted two field-based research experiments where the objectives were to:

1) evaluate plant communities that vary in complexity as candidates for buffer strips that attract and conserve beneficial insects, and 2) describe the beneficial insect communities in buffer strips already existing on organic farm and compare them to what is found in the adjacent organically managed farms and conventionally managed row crops. Although we identify buffer strips on organic farms for the prospective implementation of these best practices, I learned that this research has a much broader application - designing multifunctional landscapes to promote the sustainability of Iowa’s managed and natural ecosystems.

Chapter Two

Based on our results, we determined that thoughtfully selecting native plant species based on plant resource quality and high density of the most attractive native species are provide the “Best Bet” mixture for beneficial insect conservation.

In 2010 and 2011, nine plant communities that range in species richness, growth habits, and attractiveness to beneficial insects were established as treatments and evaluated for their ability to attract and conserve beneficial insects. Among the plant communities, we observed significant differences in beneficial insect diversity and abundance and diversity in both years. Beneficial insect diversity and abundance was lowest in simple plantings

135 composed of species that currently dominate the buffer strips on organic farms in Iowa.

This suggests these plant communities may not be optimal for conserving beneficial insects and that there are opportunities to improve buffer strips for this purpose.

Beneficial insect diversity and abundance was positively related to characteristics found in more diverse treatments such as plant species diversity and the abundance of floral resources, which were limited in simple plant communities.

The addition of flowering perennial plants, such as native prairie species that bloom throughout the season, can improve buffer strips for beneficial insects. However, not all prairie plant mixtures are composed with species that beneficial insects find most attractive. This was observed in the most diverse mixture we evaluated (CP-IA), which is recommended for conservation programs and traditional prairie reconstruction. Although this mixture was composed of 14 species (forbs and grass) there were instances when insect abundance and diversity in this mixture did not differ significantly from some of the simple plant communities composed of only one plant species.

Native perennial plant communities can be further optimized by intentionally designing mixtures with native species, even at modest levels of plant diversity (i.e. forb- only mixtures), that are attractive to beneficial insects. This was most evident in the

MSU Best Bet mixture, designed with 12 species of native prairie plants specifically selected for their attractiveness to beneficial insects. In all observations the MSU Best

Bet mixture outperformed the CP-IA. In several instances, the beneficial insect communities exhibited greater or equivalent diversity and abundance in the forb-only treatments compared to the CP-IA mixture. Carefully selecting the composition, considering the characteristics of individual plant species, and manipulating the density

136 of these species in prairie plant mixtures can be essential to conserving beneficial insects at the farm scale where small additions of a few specific species can maximize benefits

In conclusion, habitat provisioning for beneficial insects appears to be a product of plant resource quality, a high density of the most attractive native species, and necessarily a product of quantity, a habitat made up of many native plant species.

Chapter Three

Based on our results, we determined that the crop and non-crop vegetation may provide complementary resources for beneficial insects within Iowa’s agricultural landscape. In 2010 and 2011, we compared the abundance, diversity, and activity-density of beneficial insect communities among three adjacent land use types including organic farms, non-crop buffer strips, and conventional row crops. Overall, the abundance and activity-density of beneficial insects was consistently greater in buffer strips. This suggests that buffer strips can harbor beneficial insect communities in agricultural landscapes. Beneficial insect abundance and activity-density was intermediary in organic farms and lowest in row crops, indicating that beneficial insects are moving between buffer strips and organic farm and the activity between buffer strips and row crops is limited. Furthermore, there were species that were unique to the buffer strip or the organic farm and many other species were shared between the two. This was not the case in row crops, as species observed in row crops were also found in either buffer strips, organic farms, or both.

137 Differences pertaining to the diversity (species richness and evenness) of

beneficial insects among the three land-use types were not as clear as that observed for

abundance. The diversity of beneficial insects was lowest in the row crop; however,

there were instances where certain guilds exhibited greater species richness in organic farms than observed in the buffer strips.

Furthermore, we observed significant year- to- year variation in beneficial insects communities among all three land-use types. Although there are many factors that cause

insect communities to fluctuate over time, improvements that enhance the quality of

resources in buffer strips may increase the resiliency of beneficial insect communities to

disturbances present in agricultural landscapes.

Recommendations

With the research presented in chapter one, we identified plant communities,

particularly the MSU Best Bet mixture, in which beneficial insects were consistently

more diverse and abundant than plant communities commonly found in Iowa’s buffer

strips. The pre-existing non-crop habitat in buffer strips may not be composed of the

density of flowering attractive to beneficial insects. In conclusion, we proposed that the

MSU Best Bet mixture can be used for best-practices that aim to provide resources for

beneficial insects in Iowa’s agricultural landscape.

138 ACKNOWLEDGEMENTS

I thank the Department of Entomology at Iowa State University for assisting me with my education, research, and professional development. The decision to attend ISU was difficult because it meant moving away from my family and friends in Pennsylvania.

I am extremely thankful for the love of my life, Chris, who moved from Pennsylvania to

Iowa with me. Chris, I appreciate all you have sacrificed to join me in this adventure. I am also grateful for my family, especially my parents, Steve and Marguerite. Without their encouragement, support, and love I would not be who I am, or where I am, today.

Mom, I wish you were here now to see the end result and celebrate with me; you are my guardian angel and your life lessons will never be forgotten.

I am greatly indebted to Dr. Matthew O’Neal for guiding me through my research efforts and providing me with a multitude of opportunities to explore my interests. Matt, you have been an inspiration, and I thank you for not letting my hard work go unnoticed.

Working with you has truly been a fulfilling experience. I also, thank my committee members Drs. Lisa Schulte and Mary Harris. It has been an honor to be able to have two women scientists as role models to assist me throughout this process. Thank you for all of your contributions and time spent reading and critiquing my work. A special thank you to goes to recent ISU alumna and co-author, Rachael Cox, for contributing the survey of Iowa’s organic farmers and her valuable insight to chapter two.

To all of my fellow graduate students and co-workers in the Soybean Entomology

Lab and beyond, thank you for all of your help as co-workers and friends. A special thanks to past and present lab mates: Nicholas Schmidt, Michael McCarville, Adam

139 Varenhorst, Rene Hessel, and Joe Wheelock for being there for me and making this a fun and enjoyable learning experience. I owe many thanks to the ISU undergraduate hourly workers who assisted me with all of the field research and grunt work: Joe Wheelock,

Kate Russell, Jake Smith, and Zach Jameson, without your dedication and hard work this research would not have been possible. I would like to a acknowledge all of our collaborators, especially the farmers that allowed me to conduct research on their farms:

Rick and Stacy Hartmann, Angela Tedesco and Ben Saunders, Gary Guthrie, and Andrew and Melissa Dunham. I have learned so much from all of you. Additionally, I am grateful for research conducted at Michigan State University by Drs. Doug Landis and

Anna Fiedler, as some of this research is an extension of their important findings.

Finally, my assistantship, this research, and associated activities would not be possible without funding from Iowa State University and The Leopold Center for

Sustainable Agriculture. Thank you for investing in this research and me.

For all the those acknowledged here, it is hard to find the perfect words to express how grateful I am for all of your contributions to this research, my education, and my life experiences. You are all truly wonderful people.

Table 1. The presencea and absenceb of beneficial insect taxa collected in each buffer strip treatment and by year in 2010 and 2011 vacuum samples Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 POLLINATORS HYMENOPTERA Andrenidae Calliopsis andreniformis Smith ------O X - - X X X O Calliopsis nebraskensis Crawford ------X X - Perdita sp. ------X X - Apidae Apis meliffera L. ------O O O - O

Bombus griseocollis DeGeer ------XO O XO XO - - X O APPENDIX Bombus impatients Cresson ------O ------O Ceratina calcerata Robertson ------X X XO XO X X O 140

Melissodes agilis Cresson ------O - - - O Melissodes trinodus Robertson ------X - - X X - - X -

Colletidae Colletes inaequalis Say ------X X X X - Hylaeus affinis (Smith) - - - - X X XO XO X - - - - X O Halictidae Agapostemon texanus Cresson ------O - O Agapostemon virescens F. ------X XO X X XO X X O Augochlora pura Say - - - - X X X X X X X X - Augochlorella aurata (Smith) ------O - - - - - O Halictus confuses Smith ------O XO XO XO XO X X O Halictus ligatus Say - - X - - X X X X X XO X O Halictus rebicondus (Christ) ------X O XO - - XO X X O

Table 1. continued Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 (Halictidae continued) Lasioglossum (Dialictus) spp./groups L. (Dialictus) pruinosum (Robertson) ------X - - X - - X - L. (Dialictus) viridatum (Lovell) O ------X X - - X X X O L. (Dialictus) morpho sp. 1 - - X X - - XO XO XO XO XO X O L. (Dialictus) morpho sp. 2 ------X X - - X - PREDATORS ARANEAE XO XO XO XO XO XO XO XO XO X O HEMIPTERA Anthocoridae Orius insidiosus (Say) XO X X XO XO XO XO XO XO X O Berytidae 141 Neoneides muticus (Say) O XO XO XO X X XO XO XO X O Nabidae Nabis spp. XO O XO XO XO X XO XO XO X O Pentatomidae Podisus maculiventris (Say) X X XO X X X X X XO X O Phymantidae Phymanta sp. ------O O O O - O Reduviidae X - - O XO O - - - - O XO X O COLEOPTERA Cantharidae Chauliognathus marginatus F. O - - O O O O O O O - O Chauliognathus pensylvanicus O ------O O XO XO X X O (DeGeer) Cantharis rotundicollis Say ------XO - - X - - X X O

Table 1. continued Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 (Cantharidae continued) Rhagonycha sp. ------O O O - - O - O Carabidae Carabidae morpho sp. 1 ------X - - O O X X O Carabidae morpho sp. 2 - - - - X ------X - - - - X - Calleida sp. ------X ------X - Cleridae - - - - O O XO O O O O X O Coccinellidae Coleomegilla maculata (DeGeer) ------O X X O Cycloneda munda (Say) - - XO O XO - - O XO XO O X Harmonia axyridis (Pallas) XO XO X XO X XO X XO XO X O Lampyridae O X XO XO XO XO XO XO XO X O 142 Staphylinidae ------O ------O NEUROPTERA Chrysopidae Chrysoperla sp. XO XO XO XO O O XO XO XO X O Hamerobiidae XO O O XO XO O XO XO XO X O DIPTERA Asilidae - - - - O ------O - O Chamaemyiidae ------O - - O ------O Dolichopodidae Dolichopus spp. - - X X XO X X X X X X O Condylostylus spp. O O XO XO XO XO XO XO XO X O Empididae Empis spp. X XO X XO XO XO XO XO XO X O Rhagionidae Rhagio sp. O - - O O O - - O O O - O

Table 1. continued Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 (Diptera continued) Syrphidae Melanostoma mellinum L. - - O X - - X X XO XO - - X O Sphaerophoria scripta L. ------X X - - - - X - Toxomerus geminatus (Say) - - O - - X X XO XO XO X X O Toxomerus marginatus (Say) XO XO XO XO XO XO XO XO XO X O Syrphidae morpho sp. 1 ------O X - - X O HYMENOPTERA Crabronidae O ------X - - XO - - X O Pompilidae - - - - O ------O Vespidae Polistes fuscatus F. - - - - O - - - - X XO - - XO X O 143 PARASITOIDS DIPTERA Tachinidae O XO XO XO XO XO XO XO XO X O HYMENOPTERA Ceraphronoidea Ceraphrononidae - - X X - - - - X X - - X X - Megaspilidae ------X X - - X X - - X - Chalcidoidea Chalcidae ------X - - X X O X X O Encyrtidae Encyrtidae morpho sp. 1 - - - - X XO XO X XO X XO X O Encyrtidae morpho sp. 2 - - - - X O X - - - - X - - X - Encyrtidae morpho sp. 3 - - O - - O ------O - O Encyrtidae morpho sp. 4 ------O ------O

Table 1. continued Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 (Chalcidoidea continued) Eulophidae Elasmus sp. ------O ------O Eulophidae morpho sp. 1 - - O XO XO X XO XO XO XO X O Eulophidae morpho sp. 2 ------Eupelmidae X - - X X - - X XO X X X O Eurytomidae Eurytomidae morpho sp. 1 X O X XO X X XO XO XO X O Eurytomidae morpho sp. 2 - - O ------O - - O - - - O Pteromalidae Pteromalidae morpho sp. 1 O O XO XO XO XO XO XO XO X O Pteromalidae morpho sp. 2 X O XO X X X XO X X X O 144 Mymaridae - - X XO XO X X XO X X X O Chalcidoid morpho sp. 1 O O XO O XO O O XO XO X O Chalcidoid morpho sp. 2 - - X ------X - Cynipoidea Figitidae Eucoilinae sp. X O XO XO XO XO XO X X X O Neralsia sp. X - - XO X X X X X XO X O Ichneumonoidea Braconidae Aleiodes sp. X - - X X XO O X X O X O Alysiinae sp. ------O - - - - - O Aphidius spp. XO XO XO XO O XO XO XO XO X O Bracon sp. ------X X - - X - Heterospilus eurostae Viereck X - - X X X - - X - - X X -

Table 1. continued Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 (Braconidae continued) Meteorus sp. X - - X X XO XO XO XO X X O Microgastrinae spp. X XO XO XO XO X XO X XO X O Opiinae sp. X - - X X - - X X - - X X - Braconidae morpho sp. 1 XO O O XO XO O O XO X X O Braconidae morpho sp. 2 X - - X - - - - X - - X - - X - Braconidae morpho sp. 3 O - - O O O O O O O - O Ichneumonidae Centeterus sp. - - - - X ------X X - Cremastinae morpho sp. 1 ------X ------X X X - Cremastinae morpho sp. 2 ------X ------X - Cremastinae morpho sp. 3 ------X X X X - 145 Mesochorinae sp. - - - - X X X O - - X - - X O Hybrizontinae sp. ------X X ------X - Phaeogenes sp. ------X - - - - X - Ichneumonidae morpho sp. 1 O X O O O O O O O X O Ichneumonidae morpho sp. 2 ------X ------X X - Ichneumonidae morpho sp. 3 ------X ------X - Ichneumonidae morpho sp. 4 ------XO XO ------X O Platygastroidea Platygastridae Innostemma sp. ------X - - X - Scelioninae sp. X - - X X X X X X X X - Synopeas sp. ------O ------O Platygastridae morpho sp. 1 O - - O O O X XO XO O X O

Table 1. continued Switch- MSU Taxonomic unit Corn Willow grass Alfalfa MSU2 MSU3 MSU5 Best Bet CP-IA 2010 2011 (Hymenoptera continued) Proctotrupoidea Heloridae Helorus sp. ------O ------O Diaprioidea ------X - - - - X - - - - X - Mymarommatoidea ------X ------O X O Undetermined parasitoid wasp sp. ------X - - X - Symbols in each treatment column indicate the presence of species within a buffer type. X = collected in 2010 only; O = collected in 2011 only; XO = collected in both years, and dashed lines = species not found. 146