Native Bee Visitation on Florida Native Wildflowers

NATIVE BEE VISITATION ON FLORIDA NATIVE WILDFLOWERS

KATHARINE D. BUCKLEY

A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2011

To those who knew I was weird and encouraged me to study bugs anyway

ACKNOWLEDGMENTS

I thank my advisor, Dr. Jamie Ellis, and committee members Drs. Jaret Daniels and Glenn Hall for all their advice and guidance. I thank Dr. David Jarzen for the

generous gift of the use of his laboratory, supplies and expertise in palynology. I thank

James Colee for helping me with statistics. I also thank my fellow graduate students

Eddie Atkinson, Jason Graham, Pablo Herrera, and Dr. J. Akers Pence as well as research technicians Jeanette Klopchin, Mark Dykes, Michelle Kelley, Jonnie Dietz and

Ben King for their help and support. I especially thank my family for their love and encouragement of my desire to study bugs, including letting me keep a tarantula in my bedroom.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS...... 4

LIST OF TABLES...... 7

LIST OF FIGURES...... 8

ABSTRACT ...... 9

CHAPTER

1 INTRODUCTION ...... 11

2 MATERIALS AND METHODS ...... 22

General Site Information...... 22 Site Preparation and Establishment ...... 23 Site Maintenance...... 24 Vegetation Survey ...... 24 Bee Survey ...... 25 Primary Bee Surveys...... 25 Supplementary Bee Surveys...... 26 Pollen Survey...... 27 Statistical Analysis ...... 29

3 RESULT S ...... 43

Overall Bee Attraction to Native Florida Wildflowers...... 43 Principle Overlapping Flowers...... 43 Spring Time Period...... 44 Late Summer/Early Fall Time Period...... 44 Late Fall Time Period ...... 45 Wildflower Visitation by Bee Tribe ...... 45 Apini (Apidae)...... 45 Spring time period...... 45 Augochlorini (Halictidae)...... 46 Bombini (Apidae)...... 46 Principle overlapping flowers ...... 46 Late summer/early fall time period ...... 47 Eucerini (Apidae)...... 47 Late fall time period...... 47 Halictini (Halictidae)...... 48 Principle overlapping flowers ...... 48 Spring time period...... 48 Late summer/early fall time period ...... 48

Late fall time period...... 49 Megachilini (Megachilidae)...... 49 Principle overlapping plants ...... 49 Late summer/early fall time period ...... 49 Late fall time period...... 50 Xylocopini (Apidae) ...... 50 Principle overlapping flowers ...... 50 Late summer/early fall time period ...... 50 Pollen Washes...... 51 Pollen Wash Means ...... 51 Pollen Wash Percentage Likelihood...... 51 Pollen on Bumble Bees ...... 51 Pollen on Halictus poeyi ...... 52

4 DISCUSSION ...... 79

Benefits of Wildflower Mixes ...... 79 Native Bee Use of Wildflower Species...... 79 Developing a Florida Wildflower Mix That Is Used By Native Bees ...... 81 Study Limitations ...... 83 Further Investigations ...... 84 Conclusion ...... 88

LIST OF REFERENCES ...... 90

BIOGRAPHICAL SKETCH...... 99

LIST OF TABLES

Table page

1-1 Examples of bee forage studies ...... 21

2-1 Wildflowers assessed in the study...... 32

2-2 Wildflower drought tolerance and months of bloom time ...... 33

2-3 Wildflower sources, % pure seed, % germination rate, and ecotype/source of the wildflowers...... 34

2-4 Grams of wildflower seed per plot ...... 35

2-5 Total bees observed and collected...... 36

3-1 Bee visitation of the tested wildflower species...... 53

3-2 Bee tribe visitation by plant species...... 54

3-3 Pollen wash data ...... 54

3-4 Pollen wash percentage likelihoods...... 54

3-5 Bumble bee pollen wash means...... 55

3-6 Bumble bee pollen wash percentage likelihoods...... 55

3-7 Halictus poeyi pollen wash means...... 55

3-8 Halictus poeyi pollen wash percentage likelihoods...... 55

4-1 Bee-plant associations from literature ...... 89

LIST OF FIGURES

Figure page

2-1 Map of University of Florida Plant Science Research and Extension Unit near Citra, Florida...... 37

2-2 Plot spacing at the wildflower sites...... 38

2-3 Layout of plots ...... 39

2-4 Grouped Aster-Like pollen at 400x magnification under a light microscope ...... 40

2-5 Gaillardia pulchella pollen at 400x magnification under a light microscope...... 40

2-6 Helianthus angustifolius pollen at 400x magnification under a light microscope...... 41

2-7 Chamaecrista fasciculata pollen at 400x under a light microscope...... 41

2-8 Monarda punctata pollen at 400x under a light microscope...... 42

2-9 Example frame locations (in grey) for a vegetation survey ...... 42

3-1 Bee visitation (LS-mean number of bees/100 blooms/minute) on the principle six wildflower species ...... 56

3-2 Bee visitation (LS-mean number of bees/100 blooms/minute) on the spring blooming flowers...... 1

3-3 Bee visitation (LS-mean number of bees/100 blooms/minute) on the late summer/early fall blooming flowers ...... 60

3-5 Apini visitation (LS-mean number of bees/100 blooms/minute)...... 63

3-6 Augochlorini overall flower visitation (LS-mean number of bees/100 blooms/minute)...... 65

3-7 Bombini visitation (LS-mean number of bees/100 blooms/minute)...... 1

3-8 Eucerini visitation (LS-mean number of bees/100 blooms/minute) on late fall blooms...... 69

3-9 Halictini visitation (LS-mean number of bees/100 blooms/minute) ...... 1

3-10 Megachilini visitation (LS-mean number of bees/100 blooms/minute)...... 1

3-11 Xylocopini visitation (LS-mean number of bees/100 blooms/minute) ...... 1

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science

NATIVE BEE VISITATION ON FLORIDA NATIVE WILDFLOWERS

Katharine D. Buckley

August 2011

Chair: James D. Ellis Major: Entomology and Nematology

Declining pollination services in Florida is a cause for concern as several

economically-valuable crops in Florida (blueberries, watermelons and other cucurbits,

citrus, avocados, etc.) require pollination to some degree. Declines in bee populations

associated with pollinating many of these crops have been correlated strongly with the

loss of forage. Consequently, investigators outlining bee conservation programs

recommend improving forage availability for bees using several management schemes, the most successful of which has been planting seeds from wildflower species known to attract bees.

The goal of my study was to identify and quantify the attractiveness of several native Florida wildflower species to native bees for use in native bee conservation. To that end I surveyed bees on 20 different wildflowers species planted in 5 mixes at 4 different sites to compare bee visitation rates. Ten of the wildflower species

(Chamaecrista fasciculata (Michaux), Coreopsis basalis (A. Dietrich), Coreopsis lanceolata Linnaeus, Coreopsis leavenworthii Torrey & A. Gray, Gaillardia pulchella

Fougeroux de Bondaroy, Helianthus angustifolius Linnaeus, Monarda punctata

Linnaeus, Rudbeckia hirta Linnaeus, Trifolium incarnatum Linnaeus and Trifolium

repens (Linnaeus) were significantly more attractive to bees than were the other species. The attraction of these wildflower species to bees was not uniform and varied according to time of year, time of day and tribe to which a bee belonged. Overall, G. pulchella, C. lanceolata, and R. hirta were the most attractive wildflower species to the bees observed in this project.

The data I collected highlight what native Florida wildflower species are attractive to bees in general and to some bee tribes specifically. This information can be used by the general public and/or farmers interested in providing forage for local bee pollinator species. Combined with extension and outreach materials and programming, the data provided in this study hopefully will aid the conservation and restoration of native bee habitat.

CHAPTER 1 INTRODUCTION

Since The Forgotten Pollinators was published in 1996 (Buchmann & Nabhan

1996) there have been an increasing number of published studies on pollinators. This is due, in part, to the decline in the number of managed and feral European honey bee colonies in North America and bumble bee populations and range in North America and

Europe (Corbet et al. 1991; Watanabe 1994; Kearns et al. 1998; Kremen & Ricketts

2000; Steffan-Dewenter 2005; Goulson et al. 2005; Biesmeijer et al. 2006; Grixti et al.

2009; Aizen & Harder 2009). Butterflies also have been the topic of research and media attention due to declining populations of several widely distributed and well-known species such as the monarch butterfly and multiple previously common species in

Europe (Van Dyck et al. 2009).

Taken together, these declines have prompted discussion about a pollinator crisis.

While some scientists are skeptical about such a crisis, many remain concerned

(Ghazoul 2005a; Steffan-Dewenter et al. 2005; Ghazoul 2005b; Klein et al. 2007; Kluser

& Peduzzi 2007; National Research Council of the National Academies 2007; Murray et al. 2009; among others). To address the issue of pollinator decline in North America the

National Research Council investigated the issue, then published a book detailing their recommendations (National Research Council of the National Academies 2007). The research presented here touches on a few of their goals for wild pollinator conservation:

“Conservation and restoration of habitat…

 Basic information on the resource requirements of a wider variety of pollinator species is needed to improve habitat management…

 Encourage the stewards of a wide range of urban and rural areas to adopt pollinator-friendly practices and also to encourage information exchange and outreach…

 Identify and implement practices that promote the availability of diverse commercial and wild pollinators.”

Animal pollinators are estimated to pollinate 90% of all flowering plants

(Buchmann & Nabhan 1996), and 75% of crop plants either need pollinators or produce better yields when pollinated, accounting for nearly a third of the human diet (Klein et al.

2007). Many animals, including butterflies, moths, flies, beetles, birds, bats, mammals,

and even reptiles are pollinators. Hymenoptera, specifically bees, undisputedly remain

the most efficient pollinators because immature and adult bees feed solely from flowers.

Honey bees are responsible for providing most of the pollination services in

modern agriculture (MacGregor 1976). Unfortunately, honey bees in North America and

globally are under threat from pests, parasites, diseases, pesticides, poor management

practices, low honey prices, and bad publicity due to Africanized honey bees, among

many other stressors (National Research Council of the National Academies 2007).

More bee pollinated crops are grown today than ever before, and honey bee

populations no longer can meet the demand (Allen-Wardell et al. 1998; Aizen & Harder

2009; Aizen et al. 2009).

Fortunately, native bees (non-Apis species of bees) can help supplement honey

bee pollination services. In highly heterogeneous landscapes, native bees can perform

most pollination services (Winfree et al. 2008). Crop fields close to natural areas can be

pollinated almost entirely by native bees in some circumstances (Kremen et al. 2004).

Pollination services provided by honey bees and native bees together can increase fruit

set in several crops (Greenleaf & Kremen 2006). For some crops, like onions and

squash, native bees are more efficient pollinators than honey bees (Parker 1982;

Canto-Aguilar & Parra-Tabla 2000), and can be managed in high enough numbers to pollinate the target crops adequately (Shipp et al. 1994).

In many cases however, providing one managed native bee species probably is not the best solution to the ongoing pollination crisis. A better solution, especially in wild systems, is the maintenance and propagation of a diversity of pollinators rather than one or a few species (Klein et al. 2003; Fontaine et al. 2006). High diversity in pollinator species is best because populations of individual bee species have been shown to vary considerably from year to year (Williams et al. 2001), making reliance on any one species risky. Bee communities experience variations not only during the year (different spring/summer/fall compositions are common), but also year to year with different dominant species in a multiple year cycle (Williams et al. 2001).

In the interest of increasing or conserving diversity in native bee populations, it is important to understand bee biology and ecology (Freitas et al. 2009). Bees are very diverse in their life histories, especially their food and nesting habits. As such, there are several different issues that need to be addressed in a conservation or pollinator diversity management plan. These include genetic structure, nest habitat availability, bee health as affected by external factors such as pesticides and various pests/pathogens, and food resource availability.

All bees have an unusual genetic structure in common, a structure that makes them more susceptible to inbreeding: males are haploid and females are diploid (termed

“haplodiploidy”). Because male bees contain only a single copy of DNA while females have two, males only contribute half as much genetic diversity within the population compared to that contributed by diploid females. Also, the likelihood of a fertilized egg

being homozygous at the sex-determining locus increases as a bee population

decreases, thus increasing the probability of a diploid male being produced. Diploid

males cause female mortality directly (as they would have otherwise been female) and

indirectly because their offspring are essentially sterile (Zayed 2009).

Haplodiploidy is not the only unusual genetic structure associated with bees. While

most bees are solitary, many are social. Higher degrees of sociality decrease the

number of reproductively active individuals in the total population as numbers of non-

reproductive workers increase. Fewer reproductives lead to lower genetic diversity

within the population, thus increasing the chances of inbreeding depression (Zayed

2009). In bee conservation and management programs, genetic diversity is an indicator

of healthy bee populations.

The problem with controlling bee genetics is that to date genetic testing cannot

replace a pedigree for predicting inbreeding (Grueber et al. 2011). Since many bee

species of conservation importance often have cryptic nesting habits as well as

relatively unknown life histories, especially mating systems (Zayed 2009), pedigrees for

wild bees currently are very difficult to determine.

Because of the difficulties associated with monitoring genetics, the monitoring of nest sites can be done to give some indication of the health of bee populations (Frankie et al. 1998). Creating suitable nesting habitat is one of the three most common methods currently used in bee conservation and management programs (the other two being pesticide use reduction and increasing bee forage plants). Several native bee species are managed successfully for crop pollination by providing suitable nesting habitat.

Managed native species include the blue orchard bee (Osmia lignaria), the alkali bee

(Nomia melanderi), the alfalfa leafcutter bee (not technically native, but a notable managed solitary species) (Megachile rotundata), and several bumble bee species

(Bombus spp.) (Johansen et al. 1978; Peterson et al. 1992; Bosch & Kemp 2001;

Velthuis & van Doorn 2006). These bees all are managed by inducing populations of the bees to nest where pollination services are needed.

Many other bee species can be trap-nested in bamboo, different sized straws, and holes drilled into wood (Krombein 1967). However, most bee species nest in the ground

(Michener 2007) so other management schemes are necessary. One of the most recommended methods of encouraging ground nesting bees to nest in an area is to keep the ground as undisturbed as possible by limiting tilling (Vaughan et al. 2004).

Furthermore, one can create soil mounds or banks in a bee’s preferred nesting soil type since sites with bare soil are encouraged (Vaughan et al. 2004). Keeping a record of where bees are nesting from year to year or how many bees are using trap-nest sites can give one a reasonable estimation of bee populations in an area.

Bee health is another factor one must consider when developing a conservation or pollinator diversity program. Bee health is affected by many external factors (weather, climate change, pesticides, poor nutrition, predators and parasites, etc.) perhaps the most damaging of which are pesticides. Many common insecticides are known to affect native bee populations (Johansen 1977; Kevan 1999; Desneux et al. 2007). Even at sub-lethal levels, some insecticides have been shown to cause navigating problems in adult bees, leading to higher forager mortality, decreased food availability in social colonies and lower fecundity among solitary bees whose level of reproduction is limited by the amount of food for which they can forage in their lifetimes (Desneux et al. 2007;

Winfree 2010). Other pesticides, including some common fungicides, have been associated with direct mortality in honey bees (Desneux et al. 2007), which typically serve as a surrogate for native bees in 50% lethal dose (LD50) tests required by the EPA

before a pesticide can be registered (OECD 2010). To a lesser extent (largely for

research purposes) bumble bees, alfalfa leafcutter bees and alkali bees also have been

used in insecticide testing (Johansen 1977; Drescher & Geusen-Pfister 1991;

Thompson & Hunt 1999). Johansen (1977) showed that all species tested displayed a

different LD50 to the same chemicals, thus calling into question the idea that honey bees

can be used as a model for the effects of pesticides on other pollinator species and

pollinator communities (Thomson & Hunt 1999; Barmaz et al. 2010).

Herbicides and weed management also may affect bee populations negatively

(Johansen 1977). Mowing weedy roadsides causes a noticeable decline in bee

populations because the weeds’ flowers provide forage when nearby crops are no

longer flowering (Benedek 1996). Similarly, herbicide use also can reduce available

forage for pollinators in an area. Since many solitary bees are constrained by the

distance they can fly and have foraging seasons that outlast a given crops’ flowering

period, bee populations near crop fields where intensive weed management is practiced

often suffer. From the standpoint of pesticides (herbicides included), conservation and

management practices for native bees include reduced pesticide use as an important

component (Vaughan et al. 2004).

Though genetic structure, nest habitat availability and bee health are important

considerations in native bee conservation schemes, the availability of food resources

(i.e. nectar and pollen) is of major importance, and, therefore, the subject of my thesis.

Bees are flower obligates, i.e. bees acquire their food exclusively from flowers. Adult bees use nectar as their carbohydrate source and actively gather pollen which is fed to bee larva as their protein source (Michener 2007). Some bees also collect volatile floral oils, which may be used as mating pheromones (Eltz et al. 1999). This mutualism between flowers and bees has led to the theory that the abundance and diversity of bees in an area can be estimated with a combination of pan traps near flowers as well as active netting from flowers (LeBuhn et al. 2003, Matteson et al. 2008). Due to bees’ often cryptic nesting habitat, collecting from flowers is usually the most reliable method of monitoring bee populations (Roulston et al. 2007).

Bee dependence on flowers for food also has led to the theory that planting flowers that bees actively utilize could improve their populations. Declines in bee populations have been correlated strongly with the loss of forage, especially in bumble

bees (Benedek 1996; Goulson & Darvill 2004; Biesmeijer et al. 2006; Carvell et al.

2006; Kleijn & Raemakers 2008; Decourtye 2010). These declines have been especially

prevalent with the current tendency towards agricultural intensification (Kim et al. 2006).

Therefore a major tool used in bee conservation programs has been increasing forage

for bees using several different management schemes, the most successful of which

has been sowing plant mixes known to attract bees (Carvell et al. 2004; Pywell et al.

2006; Carvell et al. 2007; Potts et al. 2009). It has not been conclusively demonstrated

that planting wildflowers increases bee populations and diversity. It has, however, been

shown that bee visitations and diversity within a crop strongly increases with proximity

to ‘natural habitat’, although actual crop production showed only a weak increase

(Ricketts et al. 2008). Increasing suitable habitat in agricultural areas has been shown

to increase population of several other groups, so providing flowers for bees would likely

increase their numbers as well (Goulson 2003). The subsequent problem that arises lies

in determining which flowering plants are beneficial to various local bee communities in

attempting to restore this ‘natural habitat’.

Researchers in Europe, and especially Great Britain, have pioneered most of the studies on bee plant preference (Carvell et al. 2004; Pywell et al. 2005; Pywell et al.

2006; Carvell et al. 2007; Potts et al. 2009). The majority of these studies include bees and plants that are not native to North America and in some cases are considered invasive weeds (Winfree 2010). To date, the only studies of native bee forage preference in the United States have been in Washington (Patien et al. 1993), California

(Vaughn et al. 2004; Frankie et al. 2005), Michigan (Tuell et al. 2008), and New Jersey

(Williams et al. 2009).

Many of these studies, including those in Europe, include information regarding the bees surveyed on specific plants or in a specific area without considering the amount and density of available blooms (Winfree 2010). The problem with this strategy is that it tests bees’ use of flowers, but not their actual preference for certain flowers

(Winfree 2010). A more attractive but less abundant plant may have fewer visits than a less attractive but highly abundant plant. Consequently, studies that do not include vegetation information such as survey data or standardized plots of plants cannot confirm bee preference of a particular flower, rather only bee use of that flower.

There is a solution to this to this dilemma. Plant surveys should be conducted when determining bee preference for a given plant because the information can be used to estimate the relative abundance of flowers on which the bees are being found

(LeBuhn et al. 2003). There are a few examples of investigators who have used this method (Johnson 1980; Kells et al. 2001; Williams 2005; Williams et al. 2010). Another technique that can be used to estimate bee preference for a given plant is to grow different plant species at standard densities and conduct bee surveys on these standardized plots (Tuell et al. 2008).

While Patien et al. (1993), Vaughn et al. (2004), Frankie et al. (2005), Tuell et al.

(2008), and Williams et al. (2009) did survey a large number of wildflowers native to

North America, many of the plants and/or bees surveyed were not native to central

Florida. For concerned Floridians this is a problem because wildflowers native to Florida may support native pollinators better than would non-native plants (Frankie et al. 2005).

Native plants are less likely to behave invasively, less likely to support invasive pest insects, and more likely to be adapted to local conditions (Rieff 2011). Consequently these beneficial qualities will make the general public more likely to use native plants, especially since they likely will require less fertilizer, pesticides and watering than non- native plants (Florida’s Water Management Districts 2009).

Though certain plants are known to be highly attractive to bees, different plants bloom at different times throughout the year. Furthermore, local bee communities change throughout the year (although eusocial species often are found year round, which is why they benefit most from year round forage). As such, anyone developing a native seed mix that would include seeds of plants attractive to a wide variety of bee pollinators needs to consider the bloom period of tested plants as well as the abundance of bees seasonally.

Despite multiple studies on the composition of the bee community in Florida

(Graenicher 1930; Pascarella et al. 1999; Deyrup et al. 2002; Hall & Ascher 2010), and publicly accessible keys to Florida bees with flower associations (Pascarella 2008;

Ascher & Pickering 2010), no comparative study of native bee flower preference has been conducted in Florida, or anywhere else in the south-eastern United States.

Declining pollination services in Florida is cause for concern, as several economically- valuable crops in Florida (blueberries, watermelons and other cucurbits, citrus, avocados, etc.) require pollination to some degree (MacGregor 1976). Conservation in general is also a concern. For example, Florida’s main ecosystem, pine flatwoods, has a diverse plant understory including many wildflower species which require insect pollination to reproduce (Myers & Ewel 1990; Whitney et al. 2004). Therefore, the conservation of native Florida bees, which likely are a major pollinating group in this ecosystem, should remain a high priority.

The goal of the current study is to identify and quantify the attractiveness of several native Florida wildflower species to native bees. I hypothesize that different common Florida bee species will exhibit varied preferences for a host of wildflower species. Further, I expect bee preference for certain plants to change during the year. I hope that this information will contribute to the development of a pollinator seed mixture that would support the greatest number and diversity of bees throughout the year. Such a mix potentially would provide year-round forage for native bees in otherwise florally depauperate areas such as agriculture-intense regions. My long term goal is that the information I present herein could be used to develop conservation practices for native bees, thus protecting the pollination services they provide.

Table 1-1. Examples of bee forage studies Location Study Reference California Analyzed native versus non-native bee forage plants Williams et al. and New – bees did not prefer non-native and non-native 2010 Jersey abundance did not affect bee diversity Tested attractiveness of common ornamental plants Frankie et al. California to different bee groups for bee population survey 2009 purposes in urban areas Tested seven grassland management strategies for UK bumble bee response – responded best to sown Potts et al. 2009 flowers 43 native perennials tested for bee preference for Michigan Tuell et al. 2008 use in bee conservation Bee visitation rates across organic and conventional New Jersey farms - suggest landscape effects important and bee Winfree et al. and conservation best focused on intensively farmed 2008 Pennsylvania areas for best effects Compared Environmental Stewardship schemes for bumble bee use - recommend legume flower mix Carvell et al. UK with greater seasonal bloom diversity and/or diverse 2007 perennial forage plants Comparison of annual and perennial mixes - Carvell et al. UK different bee species preferred different mixes 2006 Comparison of seed mixes on bumble bee diversity Pywell et al. UK and abundance - wildflower and forage plant mixes 2006 performed best Four management strategies for field margins Pywell et al. UK compared for bumble bees - wildflower seed mix 2005 significantly better Field margin effects on bumble bee diversity - flowers over grass, most important are a few flower Bäckman & Finland species' abundance during bumble bee reproduction Tiainen 2002 period Six flowers planted at various times and ratios, Carreck & UK evaluated for different pollinating insect preferences Williams 2002 Five treatments of field margins indicate sown mixes UK Meek et al. 2002 better than natural regeneration Bees prefer natural regeneration over cropped field UK Kells et al. 2001 margins 21 herbaceous bee forage plants evaluated for Patien et al. Washington planting near cranberry farms - honey bee and 1993 bumble bee preferences differ

CHAPTER 2 MATERIALS AND METHODS

General Site Information

Five different wildflower mixes were planted (Table 2-1) at four sites at the

University of Florida Plant Science Research and Extension Unit (PSREU) near Citra,

Florida (N29°24'36.28" W82°10'11.90"). There were six plots at every site: (1) control,

(2) basic annual flower mix, (3) basic perennial flower mix, (4) diverse annual flower

mix, (5) diverse perennial flower mix and (6) a mixture of annual and perennial flowers.

The wildflower species were selected based on potential attractiveness to pollinators,

their status as regionally native or naturalized, seed availability, low cost, similar

management, and representing a diversity of bloom times (spring/summer/fall) (Table 2-

2, Table 2-3). Eighteen species of native Florida wildflowers were selected along with

three naturalized wildflowers: Phlox drummondii Hooker, Trifolium incarnatum Linnaeus

and T. repens Linnaeus (Table 2-1). Coreopsis basalis (A. Dietrich), C. lanceolata

Linnaeus, C. tinctoria Nuttall, Gaillardia pulchella Fougeroux de Bondaroy, Monarda

punctata Linnaeus, Helianthus angustifolius Linnaeus, P. drummondii and Rudbeckia hirta Linnaeus were also specifically chosen because they are commonly planted along

Florida roadsides as part of the Florida Department of Transportation’s wildflower program. A few of these species either never grew or produced too few flowers to attract bees significantly.

The four sites at the PSREU were > 1000 m from one another (Figure 2-1). All sites were prepared during the fall of 2009. Each site was ~75 × 21 m, containing the six plots which were 15 × 3 meters each and spaced 10 m from one another (Figure 2-

2). The type of wildflower seed mix planted was randomized among the plots within a site.

1) Site 1 - N29˚24.621' W082˚10.828' (Figure 2-3a)

2) Site 2 - N29˚24.163' W082˚10.124' (Figure 2-3b)

3) Site 3 - N29˚24.610' W082˚08.360' (Figure 2-3c)

4) Site 4 - N29˚24.227' W082˚08.912' (Figure 2-3d).

Site Preparation and Establishment

I followed methods suggested by Florida wildflower experts (Janet Grabowski,

USDA, NRCS Florida state office; M.J. Williams, USDA, NRCS Florida state office; and

Jeff Norcini, Oecohort, LLC) for planting the 5 seed mixes at each site (Norcini &

Aldritch 2004). Site preparation began September 15, 2009. Plots were measured and

laid out at all sites following which the plots were treated with glyphosate

(TouchdownTM) at 3 qt/acre mixed with Induce® pH, a non-ionic surfactant, at 1 pt/acre.

After seven days, the Bahia grass in the plots was cut with a Hiniker 5700 Flail-chopper

which cuts the grass to 5-8 cm tall and vacuums up the cuttings. On 3 November, the plots were treated with glyphosate again. Later that week, herbicide resistant perennials, including Mexican Tea (Chenopodium ambrosioides Linnaeus), Hairy Indigo

(Indigofera hirsuta Linnaeus), Benghal Dayflower (Commelina benghalensis Linnaeus) and Nutsedge (Cyperus spp.), were removed with shovels. On 17 November, the plots were burned to remove the remaining dead thatch. Although the thatch burned well, the hard roots of the Bahia grass in many places did not, forming thick mats in some places.

It was determined that a Lely Rod Weeder (Lely USA, Pella, IA) would be able to break

up the mats while disturbing the soil as little as possible. Several passes from different

directions were necessary to break up the root masses.

Seeds were mixed beforehand at densities per acre based on manufacturer’s recommendations (Table 2-4). Chamaecrista fasciculata (Michaux) and Baptisia alba

(Linnaeus) were scarified in tumblers with sandpaper to improve the chances of germination. All seeds were weighed carefully and individually bagged in preparation for planting. On 8 December, sites once again were raked with the Lely Rod Weeder and rolled with a 16 foot cylindrical roller. The seeds were mixed with 19L sand and spread

over their respective plots using Scotts Pro Turf Professional SS2 Drop Spreaders (The

Scotts Company LLC, Marysville, OH). Plots then were pressed with the roller in an effort to keep the seeds from blowing away. Sites were irrigated with 1.25 to 2.5 cm of water, once a week when < 0.6 cm of rain fell. Irrigation aided in seedling establishment.

Site Maintenance

Little maintenance was necessary once the wildflowers were established. Wild

radishes (Raphanus raphanistrum Linnaeus) were weeded from Site 1 early in the

season when they were growing in high densities. On 10 August, I removed large perennial weeds (most notably Hairy Indigo (Indigofera hirsuta) which resembled small trees) from all plots except the control. A few weeks later, it was determined that

Partridge Pea (Chamaecrista fasciculata) had established too densely and was shading out the other wildflower species in the annual mixes. Consequently, the Partridge Pea was thinned by roughly 25%.

Vegetation Survey

Vegetation surveys were conducted the day prior to the beginning of the intensive bee surveys, except for the first vegetation survey which was conducted the week prior

to the first bee survey due to the length of time it took to perform. During the first

vegetation survey, all plant stems, percent vegetation cover, and number of blooms in

10, 1 × ½ m frames were noted for each plot at all sites. The frames were placed randomly in 10 different areas delineated within a plot (Figure 2-9). This level of replication was necessary because of the performance of varying plant species from one side of a plot to the other, this likely due to different amounts of watering or a slight slope throughout the plot causing differences in soil moisture and nutrients (Elzinga et al. 1998). During subsequent surveys, only the number of blooms and percent bloom coverage were noted.

Bee Survey

Primary Bee Surveys

The primary bee surveys were conducted at all sites every third week. Nine

surveys were conducted from 11 May - 4 November 2010. Each survey lasted two days,

with two teams consisting of an observer and a recorder responsible for one site both

days. The surveys involved observing bees on each plot at each site for 10 minutes,

followed by 10 minutes of collecting bees from the plot. This was done once in the

morning and once in the early afternoon every sampling day. For each plot sampling,

the recorder would record all information for the observer, leaving the observer’s hands

and eyes free. The recorder also was responsible for noting the weather conditions, and

recording time, date, location, and sampling data for each survey (Figure 2-7).

During the observation period, the observer would walk down the sides of the plot

and tell the recorder which bees they saw (Table 2-5) and which flowers the bees were

visiting. The recorder would write down the flower and bee type per the observer’s

comments. The observer walked the perimeter of the plot twice during the 10 minute

observation period. During the first trip around the plot, the observer would identify

every new bee sighted. During the second trip, he/she only noted novel bees (species

not seen during the first trip) or increases in the frequency of a previously observed bee

species. For example, if 15 Bombus impatiens were seen during the first trip and 17

during the second, the greater of the two numbers was used.

For each ten minute collection period, the observer randomly chose a side of the

plot from which to collect in the morning, which was alternated during the afternoon

sampling period. This was due to the possibility of some of the flowers becoming

damaged from the action of netting bees in the morning, so sides were alternated to

avoid bias. Bees were captured using standard 45 cm diameter butterfly nets and were

placed immediately into kill jars prepared with cyanide. After the collection period was

over, the collected bees were placed into labeled vials and frozen in preparation for

subsequent pinning and identification. At the beginning of the season, most bees seen

on the wildflowers were captured as reference specimens. After several samplings, I

discovered that many of the bees being captured were one species, Halictus poeyi,

which could easily be sight identified. After that, only one specimen of H. poeyi was

caught during each sampling period, although samples of all other bees were collected

for identification and for bee community composition determination (Table 2-5).

Supplementary Bee Surveys

Supplementary bee surveys were conducted every other week except those

weeks on which the primary bee surveys fell. These surveys were conducted over all

sites in one day by a team of two individuals. Unlike the primary bee surveys, the

supplementary surveys were conducted by surveying each plant species for visiting

bees for a total of 10 minutes at every site across all plots.

Similar to the primary surveys, the team consisted of a recorder and an observer.

For ten minutes the observer would walk around all plots within a site containing a

targeted plant species. Any bees seen on that plant during the ten minute observation period were recorded (Table 2-5). During this time, voucher specimens of unknown species were captured (pinned and identified) (Table 2-5) as were bees seen carrying large pollen loads (saved for the pollen survey).

Bee species were identified using Michener (1994) and Pascarella (2010). All other insects were identified to family using Johnson & Triplehorn (2004). Voucher specimens are maintained at the Honey Bee Research and Extension Lab from both surveys.

Pollen Survey

To determine pollen use by the bees, I collected and pressed samples from each blooming wildflower species throughout the study. At the end of the study, I created reference slides of the pollen for each species in order to create a pollen reference collection. To do this, I used the preparation technique for modern pollen and spores as outlined in Jarzen 2008. Permanent reference slides were created using the glycerine jelly techniques also outlined in Jarzen 2008.

The pollen preparation method was as follows:

 The reference material (a single flower) was placed in 15 cc centrifuge tubes and half filled with 5% KOH.

 The tubes were placed in a hot water bath at about 90°C for two to three minutes and then removed.

 The tube’s contents were stirred with a glass rod and then poured through a fine meshed plastic screen into a clean 15 cc centrifuge tube.

 The 15 cc tubes were centrifuged at 1500 rpm for two to three minutes, decanted, and the contents in the tube were washed twice with distilled water and once with glacial acetic acid. To conduct a wash (from this point forward), the washing liquid was poured into the tube until the total volume reached approximately a centimeter from the top, unless otherwise specified. The tubes then were centrifuged and decanted in preparation for the next wash.

 Once the washes were complete, the tubes were half filled with acetolysis mixture (one part concentrated sulfuric acid to nine parts acetic anhydride) and placed in a hot water bath for 3-5 minutes.

 The tubes were removed and allowed to cool before centrifuging and decanting into a large beaker under running water.

 The contents of the tube were washed once with glacial acetic acid, twice with distilled water, and once with a 1:1 mixture of glycerin and distilled water. After decanting, the tubes were allowed to drain upside down overnight.

 The next day, a small cube of glycerine jelly was placed in the bottom of the tube and the tube placed in hot water bath until the glycerine jelly was completely melted. The resulting product was stirred thoroughly.

 One drop of the glycerine mix was placed onto a pre-cleaned microscope slide and a cover slide was pressed gently on the drop using toothpicks. The remaining glycerine mix was stored in individual vials at room temperature.

 The slides were allowed to air dry for at least 24 hours before they were sealed using Sally Hansen “Hard as Nails”® clear nail polish (Coty LLC, Uniondale, NY).

Bees visibly carrying pollen were captured in individual vials during the supplementary sampling for pollen collection analysis. They were stored in individual vials in a freezer until identification and pollen washing could be performed.

The bees were washed according to the pollen washing procedure outlined by

Ellis and Delaplane (2008). A stock solution of 45 parts by volume 70% ethyl alcohol, 3 parts by volume liquid soap, and 2 parts by volume basic fuchsin at 5% concentration was prepared. This was added to the vials containing individual bees in quantities as follows: small bees (i.e., Lasioglossum spp.) = 0.125 ml; medium bees (i.e., large

Halictus poeyi to Melissodes communis) = 0.250 ml; and large bees (i.e., Xylocopa and

Bombus spp.) = 0.500 ml. After the addition of the stock solution, the vials were closed and vortexed for 1.5 minutes.

As the pollen was not analyzed immediately following the addition of stock solution, the vials always were stirred using a vortex for at least 15 seconds prior to

analysis. To determine the volume and type of pollen that individual bees were carrying, a drop of the pollen suspension was loaded into a Bright-Line® Reichert hemacytometer

(manufacturer, city, state). All pollen grains within the center square millimeter were identified, counted and recorded. Four samples were counted per bee. The four samples were averaged and, following the hemacytometer’s instructions, (1) multiplied by 10,000, (2) then multiplied by the milliliters of stock solution added to the bee’s vial.

The resulting number is the approximate pollen load of that bee. Pollen was identified

(Kapp et al. 2000) to one of six different categories: Grouped Aster-like (Coreopsis spp. and Rudbeckia hirta) (Figure 2-4), Gaillardia pulchella (Figure 2-5), Helianthus angustifolius (Figure 2-6), Chamaecrista fasciculata (Figure 2-7), Monarda punctata

(Figure 2-8) and Unknown.

These groups of pollen were relatively easy to distinguish from one another. Most were very different shapes. For example, M. punctata pollen was spheroid with six furrows, while C. fasciculata pollen was a flattened oblong shape with two furrows on each side. All Asteraceae pollen in this study was a spiky ball, which is very typical of the family. Some were indistinguishable to me, hence the Aster-Like group. However, H. angustifolius had a significantly larger pollen grain size than the other Asteraceae in the study, and G. pulchella had a very distinctive two part spike. For comparative purposes,

all included figures of pollen were photographed at exactly the same settings and

magnification (Figure 2-4 to 2-8).

Statistical Analysis

To facilitate data analysis, the bee survey data was paired with the vegetation survey data to create the standard unit of measure: # bees/100 blooms/minute (number of bees visiting 100 blooms of a particular wildflower species during one minute) which

was the dependent variable. The overall bee survey data (grouping all bees together independent of bee species) were analyzed using a generalized linear model (GLM) procedure with an exponential distribution (SAS Institute, 2010), recognizing plant species, the estimated number of blooms for that plant species, and the overall number of species in bloom in the plot as the independent variables. Furthermore, the GLM procedure was run by morning and afternoon samplings as it had been noted previously during sampling that time may affect bee species use of various wildflowers. The overall bee survey data then was analyzed by three time periods; spring (May though the first week of June), late summer/early fall (end of July to mid-October), late fall (beginning of

October to early November) and a fourth period when the six longest blooming and most attractive plant species overlapped completely in bloom. These species included

Chamaecrista fasciculata, Coreopsis basalis, Coreopsis lanceolata, Coreopsis leavenworthii, Gaillardia pulchella, and Rudbeckia hirta. For each time period, only flowers that bloomed throughout the entire period were included in the analysis.

Furthermore, I determined which wildflowers were used most by each bee tribe, the lowest taxonomic level that could be sight identified reasonably during monitoring.

Where possible (i.e. when enough data were collected), bee tribe use of the various wildflower species was analyzed by time period using the GLM procedure already described (SAS Institute, 2010).

For the pollen data, the means and standard error are reported; along with the percent likelihood that a bee caught on a species of wildflower will be carrying pollen from that flower species. The two groups most captured (H. poeyi and Bombus spp.)

were analyzed separately and their means, standard errors, and percent likelihood that

they carried the pollen from the flower which they were captured were determined.

Table 2-1. Wildflowers assessed in the study Scientific Name Common Name Family Sig. Asclepias tuberosa Linnaeus Butterfly Milkweed Asclepiadaceae No Baptisia alba (Linnaeus) White Indigo Fabaceae No Chamaecrista fasciculata Partridge Pea Fabaceae Yes (Michaux) Coreopsis basalis (A. Dietrich) Dye Flower Asteraceae Yes Coreopsis lanceolata Linnaeus Lanceleaf Coreopsis Asteraceae Yes Coreopsis leavenworthii Torrey & Leavenworth's Coreopsis Asteraceae Yes A. Gray Coreopsis tinctoria Nuttall Golden Tickseed Asteraceae Yes Eryngium yuccifolium Michaux Rattlesnake Master Apiaceae No Gaillardia pulchella Fougeroux Blanketflower Asteraceae Yes de Bondaroy Helianthus angustifolius Narrow-leaved Sunflower Asteraceae Yes Linnaeus Ipomopsis rubra (Linnaeus) Standing Cypress Polemoniaceae No Liatris spicata (Linnaeus) Spiked Gayfeather Asteraceae No Monarda punctata Linnaeus Spotted Bee Balm Lamiaceae Yes Phlox drummondii Hooker Drummond Phlox Polemoniaceae No Rudbeckia hirta Linnaeus Black Eyed Susan Asteraceae Yes Salvia coccinea Buc’hoz ex Tropical Sage Lamiaceae No Etlinger Solidago fistulosa Miller Pinebarren Goldenrod Asteraceae No Trifolium incarnatum Linnaeus Crimson Clover Fabaceae No Trifolium repens Linnaeus Osceola Clover Fabaceae No Vernonia gigantea (Walter) Giant Ironweed Asteraceae No Osceola Clover is a cultivar of White Dutch Clover. “Sig.” implies either that a flower could not be used in the analysis (no = the species either did not produce any blooms or did not attract enough bees to be used in statistical analysis) or that it could be used in the analysis (yes).

Table 2-2. Wildflower drought tolerance and months of bloom time Drought Bloom Time Plant Tolerance JFMAMJ J A S O ND Asclepias tuberosa High XXXX X Baptisia alba High XXX Chamaecrista fasciculata Medium X X X X X Coreopsis basalis High XXXX Coreopsis lanceolata High XXXXX X Coreopsis leavenworthii Low X X X X X X X X Coreopsis tinctoria Low XXX Eryngium yuccifolium Low XXXX Gaillardia pulchella High X X X X X X X X Helianthus angustifolius Low X X Ipomopsis rubra High XXX Liatris spicata Low XX X Monarda punctata High X X X X Phlox drummondii High XXXX Rudbeckia hirta High X X X X X X X Salvia coccinea Medium XXXXXXX X X X Solidago fistulosa High X X Trifolium incarnatum Low XXX Trifolium repens Medium XX Vernonia gigantea Medium X X X

Table 2-3. Wildflower sources, % pure seed, % germination rate, and ecotype/source of the wildflowers Scientific Name Wildflower Source % Pure Germination Ecotype/ Seed Rate Source Asclepias tuberosa ERNST Seeds 99.61% 80% CA Baptisia alba Florida Wildflower 98% 80% FL Cooperative Chamaecrista fasciculata ERNST Seeds 99.22% 21% FL Coreopsis basalis Florida Wildflower 98% 65% FL Cooperative Coreopsis lanceolata Florida Wildflower 95% 65% FL Cooperative Coreopsis leavenworthii Florida Wildflower 85% 55% FL Cooperative Coreopsis tinctoria ERNST Seeds 93.21% 62% NC Eryngium yuccifolium ERNST Seeds 79.42% 1% FL Gaillardia pulchella Florida Wildflower 89% 85% FL Cooperative Helianthus angustifolius Florida Wildflower 95% 55% FL Cooperative Ipomopsis rubra Florida Wildflower 95% 75% FL Cooperative Liatris spicata ERNST Seeds 87.09% 40% FL Monarda punctata ERNST Seeds 99.89% 87% SC Phlox drummondii Florida Wildflower 99% 65% FL Cooperative Rudbeckia hirta ERNST Seeds 99.80% 93% NC Salvia coccinea Wildseed Farms 99.97% 91% TX Solidago fistulosa ERNST Seeds 14.66% 14% FL Trifolium incarnatum Adams-Briscoe Seed 98% 80% GA Company Trifolium repens Adams-Briscoe Seed 99.65% 90% GA Company Vernonia gigantea Florida Wildflower 85% 45% FL Cooperative

Table 2-4. Grams of wildflower seed per plot Plant AB AD PB PD Combo Asclepias tuberosa 18.00g Baptisia alba 4.01g 10.04g 4.62g Chamaecrista fasciculata 80.70g 40g 70.20g Coreopsis basalis 5.04g 2.52g 2.22g Coreopsis lanceolata 7.57g Coreopsis leavenworthii 1.26g 0.76g Coreopsis tinctoria 0.78g Eryngium yuccifolium 10.04g Gaillardia pulchella 12.60g 6.05g 11.50g Helianthus angustifolius 14.13g Ipomopsis rubra 4.88g Liatris spicata 6.38g 1.79g 1.79g Monarda punctata 0.50g Phlox drummondii 5.55g Rudbeckia hirta 4.70g 2.90g 1.74g Salvia coccinea 7.01g 5.04g Solidago fistulosa 5.04g 5.04g 5.04g Trifolium incarnatum 100.90g 100.90g Trifolium repens 15.20g 15.10g 15.20g Vernonia gigantea 5.42g 0.72g 2.17g Mix abbreviations refer to Perennial Basic (PB), Perennial Diverse (PD), Annual Basic (AB), Annual Diverse (AD), and Combination of Annual and Perennial (Combo) wildflower mixes.

Table 2-5. Total bees observed and collected Tribe Obs. Genus Obs. Coll. Species Obs. Coll Anthidiini 1 Anthidiellum 11Anthidiellum (Loyolanthidium) perplexum (Smith) 1 1 Apini 87 Apis 87 9 Apis mellifera Linnaeus 87 9 Augochlorini 17 Augochlorella 2 Augochlorella aurata (Smith) 2 Augochloropsis 2 Augochloropsis anonyma (Cockerell) 2 Bombini 694 Bombus 694 60 Bombus (Pyrobombus) bimaculatus Cresson 2 Bombus (Cullumanobombus) fraternus (Smith) 19 1 Bombus (Pyrobombus) impatiens Cresson 191 35 Bombus (Thoracobombus) pensylvanicus (DeGeer) 130 22 Colletini 3 Colletes 33Colletes thysanelle Mitchell 3 3 Epeolini Epeolus 3 Epeolus glabratus Cresson 1 Epeolus pusillus Cresson 2 Triepeolus 9 Triepeolus lunatus (Say) 7 Triepeolus rufithorax Graenicher 2 Eucerini 163 Melissodes 16 Melissodes communis Cresson 16 Svastra 27 Svastra aegis (LaBerge) 2 26 Svastra petulca (Cresson) 1 Halictini 3437 Agapostemon 137 18 Agapostemon splendens (Lepeletier) 137 18 Halictus 3278 648 Halictus (Odontalictus) poeyi Lepeletier 3278 648 Lasioglossum 22 10 Lasioglossum (Dialictus) nymphale (Smith) 1 Lasioglossum (Dialictus) pectorale (Smith) 1 Lasioglossum (Dialictus) puteulanum Gibbs 1 Sphecodes 2 Sphecodes heraclei Robertson 2 Megachilini 230 Coelioxys 12 Megachile 229 59 Megachile (Acentron) albitarsis Cresson 4 33 Megachile (Litomegachile) brevis Say 2 Megachile (Litomegachile) mendica Cresson 16 Megachile (Argyropile) parallela Smith 1 Megachile (Leptorachis) petulans Cresson 2 Megachile (Sayapis) policaris Say 1 Megachile (Litomegachile) texana Cresson 4 Nomadini Nomada 1 Nomada fervida Smith 1 Nomiinae 1 Dieunomia 11Dieunomia heteropoda (Say) 1 1 Xylocopini 121 Xylocopa 121 6 Xylocopa (Schonnherria) micans Lepeletier 43 5 Xylocopa (Xylocopoides) virginica (Linnaeus) 5 1 Abbreviations refer to Observed (Obs.) and Collected & Identified (Coll.) bees. For some bee tribes, the identified bee species in the last Obs. column are only a subset of the observed bees in that tribe. Bees in bold were those included in the statistical analysis .

1 4 1

1 2 3 4

Figure 2-1. Map of University of Florida Plant Science Research and Extension Unit near Citra, Florida. The four sites are labeled and separated by >1000 m. Permission courtesy of Daniel L. Colvin.

15m

68m

10m 75m

22m

Figure 2-2. Plot spacing at the wildflower sites.

A B C D PD AD AD Combo

Control PB Control Control

PB Control PB AD

AD PD AB PD

AB AB Combo AB

N N N Combo Combo PD PB N

Figure 2-3. Layout of plots within: A) Site 1, B) Site 2, C) Site 3, D) Site 4. Note direction north. Abbreviations marking wildflower mixes in plots are as follows: AB = Annual Basic, AD = Annual Diverse, Combo = Combination of Annual and Perennial mixes, PB = Perennial Basic, PD = Perennial Diverse. Control is the control plot in which no seeds were sown.

A B

Figure 2-4. Grouped Aster-Like pollen at 400x magnification under a light microscope. Examples: A) Coreopsis lanceolata, B) Coreopsis leavenworthii, C) Rudbeckia hirta.

Figure 2-5. Gaillardia pulchella pollen at 400x magnification under a light microscope.

Figure 2-6. Helianthus angustifolius pollen at 400x magnification under a light microscope.

Figure 2-7. Chamaecrista fasciculata pollen at 400x under a light microscope.

Figure 2-8. Monarda punctata pollen at 400x under a light microscope.

Figure 2-9. Example frame locations (in grey) for a vegetation survey. The large square outlined in black is a single plot. The frames are 1 × ½ m PVC rectangles placed randomly in one of ten areas within the plot.

CHAPTER 3 RESULTS

Overall Bee Attraction to Native Florida Wildflowers

In order to determine the native Florida wildflower species most attractive to bees,

I analyzed bee attraction to the six dominant blooming species (those species that had the longest overlapping bloom throughout the study) as well as bee attraction to groups of wildflowers blooming in three different time periods. Regarding the latter, the attraction of bees to seasonally dominant flowers is important to know when attempting to provide year-round floral resources for bees.

Principle Overlapping Flowers

Six of the wildflower species planted attracted large numbers of bees and bloomed

from the first week of June through mid-October. These six species were Chamaecrista

fasciculata, Coreopsis basalis, Coreopsis lanceolata, Coreopsis leavenworthii,

Gaillardia pulchella and Rudbeckia hirta. These were the only wildflowers to bloom for an extended period of time as well as consistently attract bees. For this study, I analyzed the number of bee visits/100 blooms/minute for the time period when all six species had overlapping bloom periods. Some of the wildflowers were significantly more attractive to bees than were others (F=8, df=5, 524, P<0.01). Of the six main wildflower species, bees visited G. pulchella significantly more than all of the other species with the exception of C. lanceolata (Figure 3-1A). During morning hours, C. lanceolata and

G. pulchella were visited by significantly more bees than were C. fasciculata and C. leavenworthii (F=3, df=5, 265, P<0.01; Figure 3-1B). In the afternoon, bees visited G. pulchella significantly more than any other wildflower except C. basalis (F=10, df=5,

251, P<0.01; Figure 3-1C).

Spring Time Period

The spring sampling time during which the bloom period for most flowers

overlapped occurred from the second week of May through the first week of June. This

was the only time during which the clovers (Trifolium incarnatum and T. repens)

bloomed. Additional blooming plants included C. basalis, C. lanceolata, C. tinctoria, G. pulchella and R. hirta. Bees visited G. pulchella significantly more than all other plants except T. repens (F=4, df=6, 89, P<0.01; Figure 3-2A). During morning hours, G. pulchella had the highest bee visitation rates, but its rates were not significantly different from those of C. lanceolata, C. tinctoria or R. hirta (F=3, df=6, 37, P=0.02; Figure 3-2B).

In the afternoon, G. pulchella and T. repens were visited by significantly more bees than were the other wildflowers (F=2, df=6, 43, P=0.05; Figure 3-2C).

Late Summer/Early Fall Time Period

The late summer/early fall period time period lasted from the end of July through mid-October and is when the Dotted Horsemint (Monarda punctata) produced its main bloom. C. fasciculata, C. leavenworthii, G. pulchella and R. hirta also were blooming during this period. Overall, bees visited G. pulchella significantly more than any flower

except R. hirta (F=16, df=4, 326, P<0.01; Figure 3-3A). During AM hours, the flowers

were visited equally by bees with the exception of M. punctata which was visited

significantly less by bees than were any other wildflower species (F=13, df=4, 167,

P<0.01; Figure 3-3B). During the afternoon, G. pulchella was visited by bees

significantly more than were the other wildflower species (F=8, df=4, 152, P<0.01;

Figure 3-3C).

Late Fall Time Period

The late fall time period lasted from early October through early November and was when the Narrow-leaved Sunflower (Helianthus angustifolius) bloomed. C. fasciculata, C. leavenworthii, G. pulchella, M. punctata, and R. hirta were in bloom as well. G. pulchella, H. angustifolius, and R. hirta were visited most by bees, though not more so statistically than C. fasciculata (F=9, df=5, 142, P<0.01; Figure 3-4A). Bees did not prefer any wildflower species over another one during the am hours (F=2, df=4, 62,

P=0.18). During the afternoon, bees preferred to visit G. pulchella, H. angustifolius and

R. hirta significantly more than the other test wildflowers (F=8, df=4, 72, P<0.01; Figure

3-4B).

Wildflower Visitation by Bee Tribe

Apini (Apidae)

In Florida, the Apini tribe consists of only one species, Apis mellifera Linnaeus or

the Western honey bee. Honey bees visited some wildflower species more than others

throughout the entire sampling season, though no clear patterns were observed (F=26,

df=6, 643, P<0.01; Figure 3-5A).

Spring time period

Although I occasionally saw honey bees at my sites throughout the year, they only visited the sites enough to analyze flower visitation rates for the spring sampling period.

Honey bees did not visit the wildflower sites enough in summer/fall to analyze those time periods. During spring, honey bees visited C. lanceolata significantly more so than

the other flower species (F=533, df=2, 69, P<0.01, Figure 3-5B). This pattern remained

true statistically during the morning sampling period (F=387, df=2, 30, P<0.01, Figure 3-

5C).

Augochlorini (Halictidae)

Augochlorini is the tribe that contains the green sweat bees. In Florida, it is composed of the genera Augochlora, Augochloropsis, and Augochlorella. While these bees are common, I did not see them in sufficient numbers at any time of the year to run seasonal analyses, so only an overall test (all wildflowers over all time periods) could be performed. The overall analysis did indicate bee visitation preference for some of the flowers though the pattern was hard to interpret due to the amount of absence data

(F=594, df=4, 643, P<0.01; Figure 3-6). That said, Augochlorini bees were recorded from the following plant species: C. fasciculata, C. lanceolata, C. leavenworthii, and G. pulchella.

Bombini (Apidae)

In Florida, the Bombini tribe consists of six different bumble bee species. Of these six species, approximately 99% of all bumble bee sightings at my research sites were either Bombus impatiens Cresson or B. pensylvanicus (DeGeer). Bombus fraternus

(Smith), B. griseocollis (DeGeer) and B. bimaculatus (Cresson) accounted for the remainder of the bumble bee sightings. B. variabilis Cresson, the sixth and very

uncommon bumble bee parasite species, was never seen. Bumble bees were rarely

present on my research plots in the spring and the late fall, but they were present in

large numbers during the overlapping bloom period of the principle six wildflowers, and

in the late summer/early fall bloom period.

Principle overlapping flowers

For the principle six flower species, bumble bees exhibited a very clear preference

(F=21, df=3, 423, P<0.01). C. fasciculata was visited by bumble bees more than were

the other flowers overall (Figure 3-7A) and during AM hours (F=9, df=2, 220, P<0.01;

Figure 3-7B). In the afternoon, bumble bees visited C. fasciculata and G. pulchella

similarly to one another but significantly more so than the other flowers (F=1494, df=2,

195, P<0.01; Figure 3-7C).

Late summer/early fall time period

During late summer/early fall, C. fasciculata was the wildflower most visited by

bumble bees (F=10072, df=2, 276, P<0.01, Figure 3-7D). During AM hours, bumble

bees visited C. fasciculata significantly more than the other wildflower species (F=1617,

df=2, 145, P<0.01, Figure 3-7E). In the afternoon, C. fasciculata, G. pulchella, and M. punctata were visited more by bumble bees than were the other flowers (F=8114, df=2,

124, P<0.01, Figure 3-7F).

Eucerini (Apidae)

The Eucerini or long-horned bees are a common group in Florida. The species

caught most regularly on my plots were Melissodes communis Cresson and Svastra

aegis (LaBerge). Infrequently found in the spring on my plots, there was insufficient data

to perform an analysis for the principle six or the late summer/early fall bloom periods.

Late fall time period

The long-horned bees were very common in the late fall during the H. angustifolius

bloom. In the overall test, H. angustifolius was visited most (F=11, df=2, 106, P<0.01;

Figure 3-8A). In the morning the long-horned bees visited H. angustifolius significantly

more as well (F=245, df=1, 50, P<0.01; Figure 3-8B). The long-horned bees did not

demonstrate a flower visitation preference during the afternoon observation period (F=2,

df=2, 50, P=0.12).

Halictini (Halictidae)

The Halictini tribe contained the most commonly found bee on my research plots, the primitively eusocial Halictus poeyi Lepeletier. Agapostemon splendens (Lepeletier), a solitary green sweat bee, is also a member of this tribe and was seen frequently.

Principle overlapping flowers

In the overall test for the principle six wildflower species, Halictini visited G.

pulchella significantly more than all the other wildflowers except C. basalis and C.

lanceolata (F=15, df=5, 423, P<0.01; Figure 3-9A). In the morning sampling period, only

C. fasciculata was not visited significantly by Halictini (F=8, df=5, 220, P<0.01; Figure 3-

9B). During the afternoon sampling periods, G. pulchella was visited significantly more

than were the other flower species with the exception of C. basalis (F=13, df=5, 195,

P<0.01; Figure 3-9C).

Spring time period

In the overall test in the spring, Halictini visited G. pulchella the most numerically,

though its visitation rate was not significantly different from those of C. basalis, C.

lanceolata or R. hirta (F=4, df=5, 69, P<0.01; Figure 3-9D). Similarly, in the morning, C.

lanceolata and G. pulchella were visited significantly more than all but C. basalis and R.

hirta (F=5, df=5, 30, P<0.01; Figure 3-9E). In the afternoon, Halictini visited C. basalis,

C. lanceolata, G. pulchella and T. repens significantly more than the other wildflower

species (F=4, df=5, 30, P=0.01; Figure 3-9F).

Late summer/early fall time period

In the overall test for the late summer/early fall, Halictini visited G. pulchella the

most, though not significantly more so than R. hirta (F=35, df=4, 276, P<0.01; Figure 3-

9G). In the morning, C. leavenworthii, G. pulchella and R. hirta were significantly more

visited than the other wildflowers (F=22, df=4, 145, P<0.01; Figure 3-9H). In the afternoon, G. pulchella, C. leavenworthii, and R. hirta were visited by Halictini more than

was M. punctata (F=4, df=4, 124, P<0.01; Figure 3-9I).

Late fall time period

In the late fall overall test, G. pulchella, H. angustifolius and R. hirta were visited

significantly more than C. leavenworthii (F=4, df=3, 106, P=0.02; Figure 3-9J). In the

morning and afternoon tests, bee visitation across all wildflower species was similar

(F=2, df=3, 50, P=0.22; F=2, df=3, 50, P=0.17).

Megachilini (Megachilidae)

The Megachilini are a tribe of leafcutter bees. The majority of the leafcutter bees

caught and identified were Megachile albitarsis Cresson and M. mendica Cresson,

although smaller numbers of other native species also were identified. It should be

noted that the leafcutter bees used my sites for not only nectar and pollen, but also

were observed cutting pieces of petals off G. pulchella for their nests. The leafcutter

bees were not common in the spring, but were all other times of the year.

Principle overlapping plants

For the principle six wildflowers, leafcutters visited G. pulchella most (F=5, df=5,

423, P<0.01; Figure 3-10A). In the morning, C. fasciculata was visited significantly less

than were the other flowers (F=3, df=5, 220, P=0.03; Figure 3-10B). The analysis of

afternoon flower visitation by leafcutter bees in the afternoon could not be performed

due to insufficient data.

Late summer/early fall time period

In the late summer/early fall overall test, G. pulchella was most visited, though not

significantly more so than C. leavenworthii (F=6, df=3, 276, P<0.01; Figure 3-10C). Bee

visitation of flowers in the morning and afternoon did not vary significantly (morning:

F=2, df=3, 145, P=0.21; afternoon: F=2, df=2, 124, P=0.10).

Late fall time period

The overall and morning tests were not significant (F=1, df=3, 106, P=0.42; F=0, df=3, 50, P=0.99), but the afternoon test for late fall was. In the late fall afternoons

leafcutter bees visited G. pulchella and H. angustifolius significantly more than they visited the other wildflower species (F=1555, df=1, 50, P<0.01; Figure 3-10D).

Xylocopini (Apidae)

The Xylocopini tribe in Florida consists of only two species of carpenter bees:

Xylocopa micans Lepeletier and X. virginica (Linnaeus). The majority of the carpenter

bees I saw on my research plots were X. micans. I rarely saw carpenter bees on my

plots in the spring and late fall, which probably was due to when the flowers they

preferred in my sites bloomed.

Principle overlapping flowers

For the principle six wildflower species, carpenter bees visited C. fasciculata

significantly more than all other wildflowers (F=958, df=2, 423, P<0.01; Figure 3-11A), a trend that held true during morning observations (F=2962, df=1, 220, P<0.01; Figure 3-

11B). Carpenter bee visitation rates of flower did not vary among wildflower species in the afternoon (F=0, df=1, 195, P=0.98).

Late summer/early fall time period

In the late summer/early fall overall test, C. fasciculata and M. punctata were the

most significantly attractive wildflowers to carpenter bees (F=16466, df=1, 276, P<0.01;

Figure 3-11C). In the morning carpenter bees visited C. fasciculata more than any other

flower (F=∞, df=4, 145, P<0.01; Figure 3-11D). The afternoon analysis could not be

performed due to low bee flower visitation rates during the afternoon observations. Bee

tribe attraction by season and time of day (Table 3-1) and bee tribe by plant species

(Table 3-2) are summarized.

Pollen Washes

Pollen Wash Means

Only bees visibly carrying pollen were collected, so the number of pollen grains on

each bee was high. Furthermore, the amount of pollen carried by a bee varied by bee

species and plant species (Table 3-3).

Pollen Wash Percentage Likelihood

The pollen wash percentage likelihood data indicate the likelihood that a bee

collected from a particular wildflower species actually has pollen from that wildflower

species on it. This is probably a better indicator of given bee’s pollen collection habits

because some plants naturally produce less pollen than others. For example, oak (a

wind pollinated plant) produces large amounts of pollen, which bees will collect actively.

Many legumes, such as C. fasciculata, do not produce much pollen as they are

pollinated exclusively by insects. As such, they rely on insects to carry their pollen from flower to flower. In general, bees captured on a particular wildflower species had a high likelihood of carrying pollen from that species (Table 3-4).

Pollen on Bumble Bees

Bumble bees carrying pollen were caught only on C. fasciculata. Surprisingly, most of the pollen they carried was not from Chamaecrista, but was from unknown

sources (Table 3-5). Additionally, the bumble bees captured from Chamaecrista did not

always carry Chamaecrista pollen, and a quarter of the time had aster pollen on them

despite only rare sightings of bumble bees on aster flowers (Table 3-6).

Pollen on Halictus poeyi

Halictus poeyi, in comparison to bumble bees, were caught only rarely on C.

fasciculata, and instead usually were caught on Coreopsis spp., R. hirta and G.

pulchella. Most of the pollen carried by H. poeyi was from the Asteraceae flowers, with a

few unknown pollen grains and no Chamaecrista pollen grains (Table 3-7). A third of the

bees caught on non-Gaillardia Asteraceae plants had Gaillardia pollen on them, and almost three quarters of the bees caught on Gaillardia had other Asteraceae pollen on them as well (Table 3-8). It should be further noted that in the field, H. poeyi frequently

were observed flying from Gaillardia to Coreopsis and back again.

Table 3-1. Bee visitation of the tested wildflower species Time Period Tribe Main Six Spring Late Summer to Early Fall Late Fall Overall Apini NA C. lanceolata NA NA NA Augochlorini NA NA NA NA C. fasciculata, C. lanceolata, C. leavenworthii, G. pulchella Bombini C. fasciculata , NA C. fasciculata , NA NA G. pulchella M. punctata , G. pulchella Eucerini NA NA NA H. angustifolius NA Halictini C. basalis , C. basalis , C. leavenworthii , G. pulchella , NA C. lanceolata , C. lanceolata , G. pulchella , H. angustifolius , C. leavenworthii , G. pulchella , R. hirta R. hirta G. pulchella , T. repens R. hirta Megachilini C. basalis , NA G. pulchella G. pulchella , NA C. lanceolata , H. angustifolius C. leavenworthii , G. pulchella , R. hirta Xylocopini C. fasciculata NA C. fasciculata , NA NA M. punctata Flower species listed are those significantly attractive to that tribe of bees. Superscripts next to species names mean the following: = significance differences were present in the overall test, = significant differences were present in the morning test, = significant differences were present in the afternoon test. NA means data were not available.

Table 3-2. Bee tribe visitation by plant species Apini Augochlorini Bombini Eucerini Halictini Megachilini Xylocopini C. fasciculata X X X C. basalis X X C. lanceolata X X X X C. leavenworthii X X X G. pulchella X X X X H. angustifolius X X X M. punctata X X R. hirta X X T. repens X X indicates significant visitation in at least one time period.

Table 3-3. Pollen wash data Pollen found Plant from which bees were captured on the bee Aster Gaillardia Chamaecrista Helianthus Aster 17575±2381(29) 4024±1164(49) 1864±970(31) 155468±154428(4) Gaillardia 840±386(29) 8647±2094(49) 100±59(31) 9843±8818(4) Chamaecrista 0±0(29) 0±0(49) 14798±2744(31) 0±0(4) Monarda 0±0(29) 0±0(49) 665±665(31) 0±0(4) Helianthus 0±0(29) 31±26(49) 0±0(31) 60000±49339(4) Unknown 775±538(29) 459±173(49) 142862±45147(31) 1250±510(4) Data are the mean ± standard error (n) of the estimated number of pollen grains found on an individual bee captured on a given wildflower species.

Table 3-4. Pollen wash percentage likelihoods Pollen found Plant from which the bees were captured on the bee Aster Gaillardia Chamaecrista Helianthus Aster 100±9%(29) 73±7%(49) 22±9%(31) 75±25%(4) Gaillardia 34±9%(29) 95±7%(49) 9±9%(31) 75±25%(4) Chamaecrista 0±9%(29) 0±7%(49) 87±9%(31) 0±25%(4) Monarda 0±9%(29) 0±7%(49) 3±9%(31) 0±25%(4) Helianthus 0±9%(29) 4±7%(49) 0±9%(31) 75±25%(4) Unknown 41±9%(29) 32±7%(49) 100±9%(31) 75±25%(4) Data are mean percent likelihood ± standard error (n).

Table 3-5. Bumble bee pollen wash means Pollen found on Plant from which the bees were captured the bee Chamaecrista Aster 1793±1119(23) Gaillardia 54±54(23) Chamaecrista 14538±3193(23) Unknown 168070±59828(23) Data are the mean ± standard error (n) of the estimated number of pollen grains found on an individual bumble bee (Bombus spp.) captured on Chamaecrista fasciculata.

Table 3-6. Bumble bee pollen wash percentage likelihoods Pollen found Plant from which the bees were captured on the bee Chamaecrista Aster 26±10%(23) Gaillardia 4±10%(23) Chamaecrista 91±10%(23) Unknown 100±10%(23) Data are mean percent likelihood ± standard error (n).

Table 3-7. Halictus poeyi pollen wash means Pollen found Plant from which the bees were captured on the bee Aster Gaillardia Chamaecrista Aster 19170±2468(26) 3344±1174(37) 8281±8281(2) Gaillardia 625±317(26) 4797±774(37) 625±625(2) Chamaecrista 0±0(26) 0±0(37) 0±0(2) Helianthus 0±0(26) 42±35(37) 0±0(2) Unknown 168±47(26) 135±46(37) 937±313(2) Data are the mean ± standard error (n) of the estimated number of pollen grains found on an individual Halictus poeyi captured on a given wildflower species.

Table 3-8. Halictus poeyi pollen wash percentage likelihoods Pollen found Plant from which the bees were captured on the bee Aster Gaillardia Chamaecrista Aster 100±9%(26) 73±8%(37) 50±35%(2) Gaillardia 35±9%(26) 95±8%(37) 50±35%(2) Chamaecrista 0±9%(26) 0±8%(37) 0±35%(2) Helianthus 0±9%(26) 5±8%(37) 0±35%(2) Unknown 38±9%(26) 24±8%(37) 100±35%(2) Data are mean percent likelihood ± standard error (n).

c b a,b b,c a b

Based on 3767 observations of bees

c a,b,c a b,c a a,b

Based on 2554 observations of bees

Figure 3-1. Bee visitation (LS-mean number of bees/100 blooms/minute) on the principle six wildflower species, i.e., those with the longest overlapping bloom period. LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) overall bee visitation; B) bee visitation during AM hours; C) and bee visitation during PM hours

c a,b b b a b

Based on 1213 observations of bees

Figure 3-1. Continued

A b,c a c d a,b b,c b

Based on 1319 observations of bees

b a,b a,b a a,b c b

Based on 861 observations of bees

Figure 3-2. Bee visitation (LS-mean number of bees/100 blooms/minute) on the spring blooming flowers. LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) overall bee visitation; B) bee visitation during AM hours; C) and bee visitation during PM hours

b b b,c a c d a

Based on 458 observations of bees

Figure 3-2. Continued

c b,c a d a,b

Based on 1663 observations of bees

a a a b a

Based on 1093 observations of bees

Figure 3-3. Bee visitation (LS-mean number of bees/100 blooms/minute) on the late summer/early fall blooming flowers. LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) overall bee visitation; B) bee visitation during AM hours; C) and bee visitation during PM hours

c b a c b

Based on 570 observations of bees

Figure 3-3. Continued

a,b b a a c a,b

Based on 284 observations of bees

d b a a c a

Based on 157 observations of bees

Figure 3-4. Bee visitation (LS-mean number of bees/100 blooms/minute) on the late fall blooming flowers. LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) overall bee visitation; B) bee visitation during PM hours.

b b b f d b a e a,b a,b c

Based on 85 observations of Apini bees

e a b a e f g dc

Based on 12 observations of Apini bees

Figure 3-5. Apini visitation (LS-mean number of bees/100 blooms/minute). LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) overall bee visitation; B) overall Spring bee visitation; C) Spring bee visitation during AM hours

e a b a e f g c d

Based on 6 observations of Apini bees

Figure 3-5. Continued

a c a a e a f g a d b

Based on 14 observations of Augochlorini bees

Figure 3-6. Augochlorini overall flower visitation (LS-mean number of bees/100 blooms/minute). LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors.

a d c e b c

Based on 694 observations of Bombini bees

a d c e b b

Based on 592 observations of Bombini bees

Figure 3-7. Bombini visitation (LS-mean number of bees/100 blooms/minute). LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) Principle six overlapping flowers overall bee visitation; B) Principle six overlapping flowers bee visitation by AM; C) Principle six overlapping flowers bee visitation by PM; D) Late summer/early fall overall bee visitation; E) Late summer/early fall overall bee visitation by AM; F) Late summer/early fall overall bee visitation by PM

a d b e a c

Based on 102 observations of Bombini bees

a e c b d

Based on 596 observations of Bombini bees

Figure 3-7. Continued

a e c b d

Based on 503 observations of Bombini bees

a c a a b

Based on 93 observations of Bombini bees

Figure 3-7. Continued

d b a c

Based on 97 observations of Eucerini bees

d b a c

Based on 41 observations of Eucerini bees

Figure 3-8. Eucerini visitation (LS-mean number of bees/100 blooms/minute) on late fall blooms. LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) overall bee visitation; B) bee visitation by AM.

c a,b a,b b a b

Based on 2617 observations of Halictini bees

b a a a a a

Based on 1674 observations of Halictini bees

Figure 3-9. Halictini visitation (LS-mean number of bees/100 blooms/minute). LS-Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) principle six flowers overall bee visitation; B) principle six flowers bee visitation by AM; C) principle six flowers bee visitation by PM; D) Spring blooming flowers overall bee visitation; E) Spring blooming flowers bee visitation by AM; F) Spring blooming flowers bee visitation by PM; G) Late summer/early fall blooming flowers overall bee visitation; H) Late summer/early fall blooming flowers bee visitation by AM; I) Late summer/early fall blooming flowers bee visitation by PM; J) Late fall blooming flowers overall bee visitation

c a,b b b a b

Based on 943 observations of Halictini bees

a,b a,b c a a,b,c d b,c

Based on 1252 observations of Halictini bees

Figure 3-9. Continued

a,b a b,c a a,b d c

Based on 824 observations of Halictini bees

a a b a b c a

Based on 428 observations of Halictini bees

Figure 3-9. Continued

c b a c a,b

Based on 748 observations of Halictini bees

b a a b a

Based on 388 observations of Halictini bees

Figure 3-9. Continued

a,b a a b a

Based on 360 observations of Halictini bees

b a a a

Based on 141 observations of Halictini bees

Figure 3-9. Continued

A c a,b b b a b

Based on 149 observations of Megachilini bees

b a a a a a

Based on 63 observations of Megachilini bees

Figure 3-10. Megachilini visitation (LS-mean number of bees/100 blooms/minute). LS- Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) principle six flowers overall bee visitation; B) principle six flowers bee visitation by AM; C) late summer/early fall blooming flowers overall bee visitation; D) late fall blooming flowers bee visitation by PM

c a,b a d b

Based on 112 observations of Megachilini bees

D c a a b

Based on 22 observations of Megachilini bees

Figure 3-10. Continued

a d b e b c

Based on 121 observations of Xylocopini bees

a d b e f c

Based on 110 observations of Xylocopini bees

Figure 3-11. Xylocopini visitation (LS-mean number of bees/100 blooms/minute). LS- Means with the same letter are not different at P ≤ 0.05. Plus/minus bars are standard errors. The graphs are A) principle six flowers overall bee visitation; B) principle six flowers bee visitation by AM; C) late summer/early fall blooming flowers overall bee visitation; D) late summer/early fall blooming flowers bee visitation by AM

a c c a b

Based on 112 observations of Xylocopini bees

a e d c b

Based on 104 observations of Xylocopini bees

Figure 3-11. Continued

CHAPTER 4 DISCUSSION

Benefits of Wildflower Mixes

Multiple investigators have shown that one of the best methods of attracting and

improving populations of native pollinators on agricultural field margins is with sown

wildflowers and bee forage plants (Bäckman & Tiainen 2002; Meek et al. 2002; Pywell

et al. 2005; Pywell et al. 2006; Carvell et al. 2007; Potts et al. 2009). However, not all wildflower species likely attract pollinators equally nor can one expect them to work well in all situations. For example, a wildflower species that attracts pollinators well in parts of Arizona would not do the same in Florida intuitively. To understand the attraction of native bee pollinators to Florida wildflowers, I developed a study to test the following two hypotheses: (1) different wildflower species would be attractive to different bee groups and (2) bee attraction to Florida wildflower species would vary throughout the year. Both hypotheses were supported by my research.

The data presented herein support work by other investigators who have shown outside of Florida that different groups of bees, such as honey bees and bumble bees, preferentially visit different wildflower species (Patien et al. 1993; Carvell et al. 2006;

Tuell et al. 2008; Frankie et al. 2009). Previous investigators also have shown that time of year can be a factor in pollinator preference (Bäckman & Tiainen 2002; Carreck &

Williams 2002).

Native Bee Use of Wildflower Species

Data are available on bee communities of Florida (Graenicher 1930; Pascarella et al. 1999; Deyrup et al. 2002; Hall & Ascher 2010) and bee-plant associations

(Graenicher 1930; Mitchell 1960; Balduf 1962; Mitchell 1962; Hardin et al. 1972;

Buchmann & Nabhan 1996; Deyrup et al. 2002; Pascarella 2008). Though the data are survey, rather than comparative data, they coincide with my results in most cases.

These known associations from the literature are summarized in Table 4-1.

For example Svastra aegis, a long horned bee, is known to be an oligolege on

Asteraceae (a composite family), with a strong preference to Helianthus species (Hurd et al. 1980). My data suggested that even among Asteraceae members, which Eucerini were seen visiting in my study, the bees in this tribe preferred to visit H. angustifolius.

Halictus poeyi also is known to frequent Asteraceae (Deyrup et al. 2002;

Pascarella 2008). My data demonstrate that H. poeyi prefers Asteraceae flowers,

especially G. pulchella which they preferred over other Asteraceae species during multiple time periods.

Similarly, bumble bees are listed as polylectic with strong preferences towards

Lamiaceae (mint family), Fabaceae (legume family) and Asteraceae (Pascarella 2008).

Otherwise, they are known to visit a large variety of plants (Deyrup et al. 2002). In my results, bumble bees preferred C. fasciculata (Fabaceae) most, followed by G. pulchella

(Asteraceae), then M. punctata (Lamiaceae).

In a few cases, my data were inconsistent with literature reports. Xylocopa micans, the carpenter bee, is known to be polylectic and visits a large variety of flowers (Balduf

1962; Hardin et al. 1972; Deyrup et al. 2002; Pascarella 2008). My data suggest a preference for only one or two flowers (C. fasciculata and M. punctata). Asteraceae flowers were seldom visited by X. micans despite the reported carpenter bee use of multiple Asteraceae species (Deyrup et al. 2002; Pascarella 2008). However, it should be noted that only the flower species growing in my plots were compared in their

visitation by bees. It is entirely possible, even likely, that flowers outside of my plots

were more attractive than the ones in my plots. This is especially likely in the case of

Apis mellifera, which were notably rarely seen in the plots, despite several apiaries

located within 1000 meters of all of my sites. The notable lack of several bee groups at

specific times of the year (such as bumble bees in the spring) leads me to believe that

either none of the wildflowers in those time periods were attractive to those bee groups

or that there were other, far more attractive, wildflowers blooming somewhere in the

vicinity. This is one of the main reasons why I think any future studies should include a

far larger number of wildflower species.

I believe that surveys of bee communities and their floral hosts are a vital first step

in the conservation of bees and pollination services. Controlled comparison tests of bee

flower species use should follow general surveys. They reveal new levels of interactions

between bees and flowers, and narrow our search for good bee forage plants to plants

that the bees actually prefer to use.

Developing a Florida Wildflower Mix That Is Used By Native Bees

One goal of this research was to recommend wildflower species that can be used to support native bee populations throughout a complete season. Based on the results, a seed mixture of Chamaecrista fasciculata, Coreopsis basalis, Coreopsis lanceolata,

Coreopsis leavenworthii, Gaillardia pulchella, Helianthus angustifolius, Monarda punctata, Rudbeckia hirta, Trifolium incarnatum and Trifolium repens would provide forage for the greatest number and diversity of bees for the longest period of time.

Of the twenty wildflower species tested in this project, only these ten plus

Coreopsis tinctoria either bloomed in sufficient numbers or attracted enough bees to be used in the statistical tests. This does not mean that the flowers which failed to bloom

are of no use to bee pollinators. Several species did not bloom because they are

perennials that begin to bloom only during the second year. For wildflower plantings

allowed to grow more than one season, these might in fact be ideal flowers to use to

attract bee pollinators. However, I was unable to test this assumption.

Furthermore, some of the wildflower species that did not bloom exhibited lower

germination rates or failed to germinate altogether. In many of these cases, the

germination rate of the seeds was known to be low beforehand; however the cost of the seeds was high, thus prohibiting my ability to compensate for the lower germination rate by simply planting more seeds. For example, only two Solidago fistulosa Miller

(Pinebarren Goldenrod) bloomed at all and then for only a few weeks in late fall. Despite this, one of these plants attracted an eighth tribe of bees that I did not mention in the results section: Colletini. Three of these bees from the genus Colletes were seen on the

goldenrod. In natural settings, goldenrod species are often an important late fall floral

source. It is also a perennial, and individuals that did not bloom this year may have

been overlooked.

The wildflower species that I list above were visited by significant numbers of bees

during at least one time period, but also bloomed well the first year they were planted.

For these reasons I recommend them not only as good bee forage plants, but also as

plants that are easy to grow, hardy and produce a large number of blooms.

I can also recommend them in places besides Florida. Most of the plants used in

this study are found outside of Florida (with the possible exception of Coreopsis leavenworthii which was until recently considered entirely endemic to Florida). Asclepias tuberosa Linnaeus, C. lanceolata, C. tinctoria, G. pulchella, M. punctata, R. hirta, T.

repens and T. incarnatum are all found from Florida to Canada to California. Most of the rest of the plants are common throughout the eastern United States (USDA, NRCS

2011). Similarly, all of the bee tribes, and even individual bee species are widespread throughout the United States. For these reasons my study may have applicability beyond Florida.

Study Limitations

Along with low seed germination rates for some tested wildflower species, there were several other limitations to this project. First, the study lasted only one field season. Bee communities have been shown to change drastically from year to year

(Kremen et al. 2004). As such, a number of new species may have been found in the wildflower plots had it run a second season. However, a bloom period of one year is indicative of what might occur if these wildflowers are sown on an agricultural field

margin. In this case, only annual wildflowers could be planted beside a field that will be

used only for one season or planted in its place during the fallow year.

Another limitation of the study is that most bee identification occurred in the field

and was accomplished by the sight identification of bees. Some bees such as the

bumble bees, carpenter bees and honey bees are large, common and easily sight

identified. With minor training, a field technician could sight identify these three species.

Other groups of bees, notably many Eucerini and Augochlorini, require a microscope

and dichotomous key to identify to genus or a taxonomic level of greater resolution. To

account for variability between field technicians, I recognized bee tribe as the best level

of resolution I could use in my study. Reference specimens were captured and

identified, thus allowing me to corroborate the sight identifications.

Further Investigations

While conducting my experiments, I recognized a number of additional studies that could be conducted. One would be a comparison between two of the main methods used to compare bee visitation rates: the vegetation survey method and the standard area method. Both the vegetation survey method and standard area method provide valuable information. Both provide a measure of bloom density. The vegetation survey method provides a bloom density based on bloom availability in a mixed wildflower planting while in the standard area method, wildflower species are planted in uniform areas, individually by species with a goal of standardizing the number of blooms available per unit area. Determining bloom density is important for estimating bee use of flowers. For example, if flower A is very abundant and frequently visited by bees, and flower B is equally abundant but not visited by bees, flower A can be assumed to be more attractive to bees. If flower A is very abundant and frequently visited by bees but flower B is not abundant, then flower B’s apparent unattractiveness to bees could be due to its rarity or a true lack of preference by the bees for it. Bloom density estimates allow for standardization of the average bee use of flowers.

The vegetation survey method is what I used in this study. In this method, bloom density is estimated from repeated vegetation surveys. The vegetation survey accurately measures the resulting flower mosaic in a plot regardless of the seed composition of the original common seed mixture. However, sampling personnel must be able to identify visiting bees and the target plants for this survey method to be useful.

Furthermore, an observer can overlook plant species that do not germinate in large numbers or favor larger plants that may out shade smaller ones (C. fasciculata for

example shaded out many of the other species in the annual mix plots until it was

thinned).

The standard area method, on the other hand, does not mimic agricultural field

margin conditions. In a standard area experiment, plant species are planted separately

in individual, standard sized plots. When each plant is labeled accurately, the observer

does not need to be able to identify each plant since the plant is labeled. Raw data received from this type of survey do not need to be standardized with vegetation survey data, but can be used as is. Since the survey typically is conducted using a smaller numbers of plants, the plants need to be in peak blooming conditions to increase study accuracy. My attempt to conduct this type of survey, per my original proposal, was

unsuccessful due to the poor performance of the plants at the study sites. Ideally, the

study sites should have been located in well-maintained gardens and established to

mimic ornamental gardens rather than agricultural field margins. This would have made

possible higher plant survival rates, a more standardized number of blooms, and greater

numbers and quality of blooms. Viewed in total, vegetation surveys (like those used in

the current study) are best used for agricultural field margin studies while standard area

surveys are more suited to address garden vegetation plantings. As a future

investigation, a comparison of data on bee flower preference collected using both of

these methods would highlight the pros and cons of both methods and may indicate the

effects of plant density and distribution on bee visitation (Cresswell 1997).

Further questions were raised as a result of the implementation of this study. First,

what is the best way to sample bee preference for flowering plants? Pan traps, which

commonly are used to sample for smaller bee species (Roulston et al. 2007), could not

be used in this study as they do not provide flower use information, especially when flowers occur in mixed plantings. Consequently, I had to use sweep netting and sight identification techniques to survey bees. While more labor intensive than pan trapping, netting has been proven to be approximately as effective (Roulston et al. 2007) or more effective at surveying bee communities than pan trapping (Cane et al. 2000; Wilson et al. 2008).

Another question resulting from this study dealt with temporal sampling techniques, i.e. what length of time is sufficient to determine bees’ use of wildflowers, but short enough to avoid repeat sampling of bees? Coupled with sampling time, one must consider how many sites/plots/flower species can be surveyed accurately in a day and what manpower is needed to conduct the survey. The ten minute observation period used in this survey seemed to work well, but a comparison of different observation periods should be conducted in the future.

Additional temporal sampling issues arose. I found an interaction between time of day and bee’s flower visitation. Bee plant use habits changed from morning to afternoon samplings (Table 3-1). It is well known that some bee begin foraging earlier in the morning than others (bumble bees, for example, are well known to be one of the first flower visitors). However, it was a surprise to find that most of the bees were more active on flowers in the morning than in the afternoon. It would be interesting to see if this holds true in other studies, especially ones taking place in different climates. In addition, data collection rarely began before 09:00 and usually ended by 16:30. To that end, it is possible that bees foraging before 09:00 or after 16:30 were sampled inadequately. This is true especially in the summer months when sunrise was occurred

before 06:00. Future investigators could study bee species richness and diversity throughout an entire day.

Future investigators also could conduct studies that include more wildflower species. Only commercially available wildflower species were used in my study. Future studies could include plant species that are not commercially available. The seeds of many flowering plants can be hand-picked for an experiment. If such species are shown to be attractive to bees, then seed companies may begin offering them commercially.

Additionally, crop plants may serve as a good addition to any further comparison

studies. Some wildflowers may compete in their attractiveness to crop plants, which

goes against the purpose of planting wildflowers to increase crop production. This is

one of the reasons why the bloom period of different species is so important. When

choosing wildflower species for pollinators of a specific crop it is important to choose

wildflowers that would bloom before and after, but not during, the crop’s bloom. This is

another type of interaction that has yet to be adequately studied.

Additionally, studies should be conducted in other areas and climates to address

unique bee and wildflower communities. Such studies could include woody plants as

well. Bushes that provide good bee forage would be ideal for hedgerows and home

gardens. Furthermore, trees can provide more permanent bee forage, and often are

responsible for winter and early spring nectar and pollen sources. Perennials often have

energy stores and deep root systems that annuals do not have (Mauseth 2009). This

allows them to bloom during droughts or times of the year that annuals cannot.

These investigations may be used to further the study of pollination ecology. It would be useful to better standardize methods used in pollination ecology and may lead

to streamlined methods that take advantage of bee and plant biology. Knowing the pros and cons of research methods would allow for incoming researchers to better design their experiments in this quickly expanding field of study. More streamlined methods would also help to make studies with greater numbers of species and types of plants possible, leading to a greater enhancement of our knowledge of bee-plant interaction.

Conclusion

Further research into native bees and bee wildflower use will be useful to the long

term conservation of native bees, native plants and the ecological services they provide.

Understanding the links between bees and plants allows us as scientists to predict and

solve problems with pollination services that will occur as a result of habitat loss and

fragmentation, agricultural intensification and the globalization of bee pests and

diseases (National Research Council of the National Academies 2007). As in any

conservation effort, research alone is not enough. Extension and outreach efforts aimed

at convincing growers and homeowners that there is a problem and that they can help

are key steps in the conservation of native bees and other pollinators.

Ultimately native pollinators like the bees seen in my study help maintain a healthy

ecosystem and ensure food security. Our understanding of native bees and their importance is causing the fields of pollination ecology and pollinator conservation to blossom. New publications are beginning to bridge the gap between the science and its application (Kremen et al. 2007; Byrne & Fitzpatrick 2009; Tallamy 2009; Grissell 2010;

The Xerces Society 2011). It is my hope that the work I and others have done will help to preserve bees and the flowers that feed them for generations to come.

Table 4-1. Bee-plant associations from literature Bee Genus Known Floral Associations Augochlorella Asclepias, Coreopsis, Chamaecrista, Eryngium, Helianthus, Monarda, Rudbeckia, Salvia, Solidago, Trifolium, Vernonia Augochloropsis Asclepias, Chamaecrista, Coreopsis, Eryngium, Helianthus, Solidago, Trifolium Bombus Asclepias, Baptisia, Chamaecrista, Coreopsis, Eryngium, Gaillardia, Helianthus, Liatris, Monarda, Phlox, Rudbeckia, Solidago, Trifolium, Vernonia Melissodes Asclepias, Baptisia, Coreopsis, Helianthus, Monarda, Rudbeckia, Salvia, Solidago, Vernonia Svastra Gaillardia, Helianthus, Vernonia Agapostemon Chamaecreista, Coreopsis, Gaillardia, Helianthus, Monarda, Solidago, Vernonia Halictus Asclepias, Coreopsis, Eryngium, Helianthus, Liatris, Monarda, Rudbeckia, Solidago, Trifolium, Vernonia Lasioglossum Asclepias, Coreopsis, Eryngium, Helianthus, Liatris, Monarda, Rudbeckia, Salvia, Solidago, Trifolium, Vernonia Megachile Asclepias, Baptisia, Coreopsis, Eryngium, Gaillardia, Helianthus, Liatris, Monarda, Rudbeckia, Salvia, Solidago, Trifolium, Vernonia Xylocopa Chamaecrista, Solidago, Vernonia

LIST OF REFERENCES

Aizen, M. A., L. A. Garibaldi, S. A. Cunningham, and A. M. Klein. 2009. How much does agriculture depend on pollinators? Lessons from long-term trends in crop production. Annals of Botany 103(9):1579-1588.

Aizen, M. A., and L. D. Harder. 2009. The global stock of domesticated honey bees is growing slower than agricultural demand for pollination. Current Biology 19(11):915- 918.

Allen-Wardell, G., P. Bernhardt, R. Bitner, A. Burquez, S. Buchmann, J. Cane, P. A. Cox, V. Dalton, P. Feinsinger, M. Ingram, D. W. Inouye, C. E. Jones, K. Kennedy, P. Kevan, H. Koopowitz, R. Medellin, S. Medellin-Morales, and G. P. Nabhan. 1998. The potential consequences of pollinator declines on the conservation of biodiversity and stability of food crop yields. Conservation Biology 12:8-17.

Ascher, J. S., and J. Pickering. 2010. Bee species guide (Hymenoptera: Apoidea: Anthophila). Discover Life. Available from http://www.discoverlife.org/mp/20q?guide=Apoidea_species&flags=HAS: 2011.

Bäckman, J. C., and J. Tiainen. 2002. Habitat quality of field margins in a Finnish farmland area for bumblebees (Hymenoptera: Bombus and Psithyrus). Agriculture, Ecosystems & Environment 89(1-2):53-68.

Balduf, W. V. 1962. Life of the carpenter bee, Xylocopa virginica (Linn.) (Xylocopidae, Hymenoptera). Annals of the Entomological Society of America 55:263-271.

Barmaz, S., S. Potts, and M. Vighi. 2010. A novel method for assessing risks to pollinators from plant protection products using honeybees as a model species. Ecotoxicology 19(7):1347-1359.

Benedek, P. 1996. Structure and density of lucerne pollinating wild bee populations as affected by changing agriculture. Acta Horticulturae 437:353-357.

Biesmeijer, J. C., S. P. M. Roberts, M. Reemer, R. Ohlemüller, M. Edwards, T. Peeters, A. P. Schaffers, S. G. Potts, R. Kleukers, C. D. Thomas, J. Settele, and W. E. Kunin. 2006. Parallel declines in pollinators and insect-pollinated plants in Britain and the Netherlands. Science 313(5785):351-354.

Bosch, J., and W. P. Kemp. 2001. How to manage the blue orchard bee as an orchard pollinator. Sustainable Agriculture Network, Beltsville, MD.

Buchmann, S. L., and G. P. Nabhan. 1996. The Forgotten Pollinators. Island Press, Washington D.C.

Byrne, A., and Ã. Fitzpatrick. 2009. Bee conservation policy at the global, regional and national levels. Apidologie 40(3):194-210.

Cane, J. H., R. L. Minckley, and L. J. Kervin. 2000. Sampling bees (Hymenoptera: Apiformes) for pollinator community studies: Pitfalls of pan-trapping. Journal of the Kansas Entomological Society 73:225-231.

Canto-Aguilar, M. A., and V. Parra-Tabla. 2000. Importance of conserving alternative pollinators: Assessing the pollination efficiency of the squash bee, Peponapis limitaris in Cucurbita moschata (Cucurbitaceae). Journal of Insect Conservation 4(3):201-208.

Carreck, N. L., and I. H. Williams. 2002. Food for insect pollinators on farmland: Insect visits to flowers of annual seed mixtures. Journal of Insect Conservation 6:13-23.

Carvell, C., W. R. Meek, R. F. Pwell, D. Goulson, and M. Nowakowski. 2007. Comparing the efficacy of agri-environment schemes to enhance bumble bee abundance and diversity on arable field margins. Journal of Applied Ecology 44(1):29- 40.

Carvell, C., W. R. Meek, R. F. Pywell, and M. Nowakowski. 2004. The response of foraging bumblebees to successional change in newly created arable field margins. Biological Conservation 118(3):327-339.

Carvell, C., D. B. Roy, S. M. Smart, R. F. Pywell, C. D. Preston, and D. Goulson. 2006. Declines in forage availability for bumblebees at a national scale. Biological Conservation 132(4):481-489.

Carvell, C., P. Westrich, W. R. Meek, R. F. Pywell, and M. Nowakowski. 2006. Assessing the value of annual and perennial forage mixtures for bumblebees by direct observation and pollen analysis. Apidologie 37:326-340.

Corbet, S. A., I. H. Williams, and J. L. Osborne. 1991. Bees and the pollination of crops and wild flowers in the European community. Bee World 72:47-59.

Cresswell, J. E. 1997. Spatial heterogeneity, pollinator behaviour and pollinator- mediated gene flow: Bumblebee movements in variously aggregated rows of oil-seed rape. Oikos 78(3):546-556.

Decourtye, A., E. Mader, and N. Desneux. 2010. Landscape enhancement of floral resources for honey bees in agro-ecosystem. Apidologie 41(3):264-277.

Desneux, N., A. Decourtye, and J. Delpuech. 2007. The sublethal effects of pesticides on beneficial arthropods. Annual Review of Entomology 52(1):81-106.

Deyrup, M., J. Edirisinghe, and B. Norden. 2002. The diversity and floral hosts of bees at the Archbold Biological Station, Florida (Hymenoptera: Apoidea). Insecta Mundi 16(1- 3):86-120.

Drescher, W., and H. Geusen-Pfister. 1991. Comparative testing of the oral toxicity of acephate, dimethoate and methomyl to honeybees, bumblebees and Syrphidae. Acta Horticulturae 288:133-136.

Ellis, A., and K. S. Delaplane. 2008. Effects of nest invaders on honey bee (Apis mellifera) pollination efficacy. Agriculture, Ecosystems & Environment 127(3-4):201- 206.

Eltz, T., W. M. Whitten, D. W. Roubik, and K. E. Linsenmair. 1999. Fragrance collection, storage, and accumulation by individual male orchid bees. Journal of Chemical Ecology 25:157-176.

Elzinga, C. L., D. W. Salzer, and J. W. Willoughby. 1998. Measuring and monitoring plant populations. Bureau of Land Management National Business Center, Denver, Colorado.

Florida's Water Management Districts. 2009. Waterwise Florida Landscapes: Landscapes to promote water conservation using the principles of xeriscape.

Fontaine, C., I. Dajoz, J. Meriguet, and M. Loreau. 2006. Functional diversity of plant- pollinator interaction webs enhances the persistence of plant communities. Public Library of Science Biology 4(1):129-135.

Frankie, G., R. Thorp, J. Hernandez, M. Rizzardi, B. Ertter, J. Pawelek, S. Witt, M. Schindler, R. Coville, and V. Wojcik. 2009. Native bees are a rich natural resource in urban California gardens. California Agriculture 63(3):113-120.

Frankie, G. W., R. W. Thorp, L. E. Newstrom-Lloyd, M. A. Rizzardi, J. F. Barthell, T. L. Griswold, J. Kim, and S. Kappagoda. 1998. Monitoring solitary bees in modified wildland habitats: Implications for bee ecology and conservation. Environmental Entomology 27(5):1137-1148.

Frankie, G. W., R. W. Thorp, M. Schindler, J. Hernandez, B. Ertter, and M. Rizzardi. 2005. Ecological patterns of bees and their host ornamental flowers in two northern California cities. Journal of the Kansas Entomological Society 78(3):227-246.

Freitas, B., M., V. Imperatriz-Fonseca, L. Medina M., A. Kleinert de Matos Peixoto, L. Galetto, G. Nates-Parra, and J. J. Quezada-Euán G. 2009. Diversity, threats and conservation of native bees in the Neotropics. Apidologie 40(3):332-346.

Ghazoul, J. 2005a. Buzziness as usual? Questioning the global pollination crisis. Trends in Ecology & Evolution 20(7):367-373.

Ghazoul, J. 2005b. Response to Steffan-Dewenter et al.: Questioning the global pollination crisis. Trends in Ecology & Evolution 20(12):652-653.

Goulson, D. 2003. Conserving wild bees for crop pollination. Food, Agriculture & Environment 1(1):142-144.

Goulson, D., and B. Darvill. 2004. Niche overlap and diet breadth in bumblebees; are rare species more specialized in their choice of flowers? Apidologie 35(1):55-63.

Goulson, D., M. E. Hanley, B. Darvill, J. S. Ellis, and M. E. Knight. 2005. Causes of rarity in bumblebees. Biological Conservation 122(1):1-8.

Graenicher, S. 1930. Bee-fauna and vegetation of the Miami region of Florida. Annals of the Entomological Society of America 23(1):153-175.

Greenleaf, S. S., and C. Kremen. 2006. Wild bees enhance honey bees’ pollination of hybrid sunflower. Proceedings of the National Academy of Sciences 103(37):13890- 13895.

Grissell, E. 2010. Bees, wasps, and ants: The indispensable role of Hymenoptera in gardens. Timber Press, Portland, OR.

Grixti, J. C., L. T. Wong, S. A. Cameron, and C. Favret. 2009. Decline of bumble bees (Bombus) in the North American Midwest. Biological Conservation 142(1):75-84.

Grueber, C. E., J. M. Waters, and I. G. Jamieson. 2011. The imprecision of heterozygosity-fitness correlations hinders the detection of inbreeding and inbreeding depression in a threatened species. Molecular Ecology 20(1):67-79.

Hall, H. G., and J. S. Ascher. 2010. Surveys of bees (Hymenoptera: Apoidea: Anthophila) in natural areas of Alachua County in north-central Florida. Florida Entomologist 93(4):609-629.

Hardin, J. W., G. Doerksen, D. Herndon, M. Hobson, F. Thomas. 1972. Pollination ecology and floral biology of four weedy genera in southern Oklahoma. The Southwestern Naturalist 16(3/4):403-412.

Hurd, J., Paul D., W. E. LaBerge, and E. G. Linsley. 1980. Principal sunflower bees of North America with emphasis on the southwestern United States (Hymenoptera: Apoidea). Smithsonian Institution Press, Washington, D.C.

Jarzen, D. M. 2008. Guide to the operations and procedures adopted in the palynology laboratory.

Johansen, C. A. 1977. Pesticides and pollinators. Annual Review of Entomology 22(1):177-192.

Johansen, C. A., D. F. Mayer, and J. D. Eves. 1978. Biology and management of the alkali bee, Nomia melanderi Cockerell (Hymenoptera: Halictidae). Washington State Entomological Society, Pullman, WA.

Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61(1):65-71.

Kapp, R. O., O. K. Davis, and J. E. King. 2000. Pollen and Spores. 2nd edition. The American Association of Stratigraphic Palynologists Foundation.

Kearns, C. A., D. W. Inouye, and N. M. Waser. 1998. Endangered mutualisms: The conservation of plant-pollinator interactions. Annual Review of Ecology and Systematics 29(1):83-112.

Kells, A., J. Holland, and D. Goulson. 2001. The value of uncropped field margins for foraging bumblebees. Journal of Insect Conservation 5(4):283-291.

Kevan, P. G. 1999. Pollinators as bioindicators of the state of the environment: Species, activity and diversity. Agriculture, Ecosystems & Environment 74(1-3):373-393.

Kim, J., N. Williams, and C. Kremen. 2006. Effects of cultivation and proximity to natural habitat on ground-nesting native bees in California sunflower fields. Journal of the Kansas Entomological Society 79(4):309-320.

Kleijn, D., and I. Raemakers. 2008. A retrospective analysis of pollen host plant use by stable and declining bumble bee species. Ecology 89(7):1811-1823.

Klein, A., I. Steffan–Dewenter, and T. Tscharntke. 2003. Fruit set of highland coffee increases with the diversity of pollinating bees. Proceedings of the Royal Society of London B: Biological Sciences 270(1518):955-961.

Klein, A., B. E. Vaissière, J. H. Cane, I. Steffan-Dewenter, S. A. Cunningham, C. Kremen, and T. Tscharntke. 2007. Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society B: Biological Sciences 274(1608):303- 313.

Kluser, S., and P. Peduzzi. 2007. Global pollinator decline: A literature review. United Nations Environment Programme/Global and Regional Integrated Data, Geneva.

Kremen, C., and T. Ricketts. 2000. Global perspectives on pollination disruptions. Conservation Biology 14(5):1226-1228.

Kremen, C., N. M. Williams, M. A. Aizen, B. Gemmill-Herren, G. LeBuhn, R. Minckley, L. Packer, S. G. Potts, T. Roulston, I. Steffan-Dewenter, D. P. Vázquez, R. Winfree, L. Adams, E. E. Crone, S. S. Greenleaf, T. H. Keitt, A. 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. Ecology Letters 10(4):299- 314.

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. Ecology Letters 7(11):1109-1119.

Krombein, K. V. 1967. Trap-nesting wasps and bees: life histories, nests and associates. Smithsonian Press, Washington D.C.

LeBuhn, G., S. Droege, N. Williams, B. Minckley, T. Griswold, C. Kremen, O. Messinger, J. Cane, T. Roulston, F. Parker, V. Tepedino, and S. Buchmann. 2003. A standardized method for monitoring bee populations—the bee inventory plot. . Available from http://online.sfsu.edu/~beeplot/ 2011).

MacGregor, S. E. 1976. Insect pollination of cultivated crop plants. U.S.D.A. Handbook 496. U.S. Department of Agriculture, Agricultural Research Service, Washington.

Matteson, K. C., J. S. Ascher, and G. A. Langellotto. 2008. Bee richness and abundance in New York City urban gardens. Annals of the Entomological Society of America 101(1):140-150.

Mauseth, J. D. 2009. Botany: an introduction to plant biology. 4th edition. Jones and Bartlett Publishers, Sudbury, MA.

Meek, B., D. Loxton, T. Sparks, R. Pywell, H. Pickett, and M. Nowakowski. 2002. The effect of arable field margin composition on invertebrate biodiversity. Biological Conservation 106(2):259-271.

Michener, C. D. 2007. The bees of the world. 2nd edition. The Johns Hopkins University Press, Baltimore.

Michener, C.D. R.J. McGinley, B.N. Danforth. 1994. The bee genera of North and Central America (Hymenoptera: Apoidea). Smithsonian Institute Press, Washington D.C.

Mitchell, T. B. 1960. Bees of the eastern United States. I. Technical bulletin (North Carolina Experiment Station) 141:1-538.

Mitchell, T. B. 1960. Bees of the eastern United States. II. Technical bulletin (North Carolina Experiment Station) 152:1-557.

Murray, T. E., M. Kuhlmann, and S. Potts G. 2009. Conservation ecology of bees: Populations, species and communities. Apidologie 40(3):211-236.

Myers, R. L., and J. J. Ewel. 1990. Ecosystems of Florida. University of Central Florida Press, Gainesville, FL.

National Research Council of the National Academies. 2007. Status of pollinators in North America. The National Academies Press, Washington D.C.

Norcini, J. G., and J. H. Aldrich. 2004. Establishment of native wildflower plantings by seed. ENH 968. University of Florida IFAS Extension.

OECD (Organisation for Economic Co-operation and Development). 2010. OECD Survey of pollinator testing, research, mitigation and information management: Survey results. 52. Paris.

Parker, F. D. 1982. Efficiency of bees in pollinating onion flowers. Journal of the Kansas Entomological Society 55(1):171-176.

Pascarella, J. B. 2008. The bees of Florida. Available from http://www.bio.georgiasouthern.edu/Bio-home/Pascarella/Intro.htm 2011.

Pascarella, J. B., K. D. Waddington, and P. R. Neal. 1999. The bee fauna (Hymenoptera: Apoidea) of Everglades National Park, Florida and adjacent areas: Distribution, phenology, and biogeography. Journal of the Kansas Entomological Society 72(1): 32-45.

Patien, K. D., C. H. Shanks, and D. F. Mayer. 1993. Evaluation of herbaceous plants for attractiveness to bumble bees for use near cranberry farms. Journal of Apicultural Research 32(2):73-79.

Peterson, S. S., C. R. Baird, R. M. Bitner Parma, and C. Idaho. 1992. Current status of the alfalfa leafcutting bee, Megachile rotundata, as a pollinator of alfalfa seed. Bee Science 2:135-142.

Potts, S. G., B. A. Woodcock, S. P. M. Roberts, T. Tscheulin, E. S. Pilgrim, V. K. Brown, and J. R. Tallowin. 2009. Enhancing pollinator biodiversity in intensive grasslands. Journal of Applied Ecology 46(2):369-379.

Pywell, R. F., E. A. Warman, C. Carvell, T. H. Sparks, L. V. Dicks, D. Bennett, A. Wright, C. N. R. Critchley, and A. Sherwood. 2005. Providing foraging resources for bumblebees in intensively farmed landscapes. Biological Conservation 121(4):479-494.

Pywell, R. F., E. A. Warman, L. Hulmes, S. Hulmes, P. Nuttall, T. H. Sparks, C. N. R. Critchley, and A. Sherwood. 2006. Effectiveness of new agri-environment schemes in providing foraging resources for bumblebees in intensively farmed landscapes. Biological Conservation 129(2):192-206.

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, B. F. Viana. 2008. Landscape effects on crop pollination services: are there general patterns? Ecology Letters 11(5):499-515.

Rieff, S. K. 2011. Why native plants. The University of Texas at Austin. http://www.wildflower.org/whynatives/

Roulston, T. H., S. A. Smith, and A. L. Brewster. 2007. A comparison of pan trap and intensive net sampling techniques for documenting a bee (Hymenoptera: Apiformes) fauna. Journal of the Kansas Entomological Society 80(2):179-181.

Shipp, J. L., G. H. Whitfield, and A. P. Papadopoulos. 1994. Effectiveness of the bumble bee, Bombus impatiens Cr. (Hymenoptera: Apidae), as a pollinator of greenhouse sweet pepper. Scientia Horticulturae 57(1-2):29-39.

Steffan-Dewenter, I., S. G. Potts, and L. Packer. 2005. Pollinator diversity and crop pollination services are at risk. Trends in Ecology & Evolution 20(12):651-652.

Tallamy, D. W. 2009. Bringing nature home: How you can sustain wildlife with native plants. 2nd edition. Timber Press, Portland, OR.

Thompson, H. M., and L. V. Hunt. 1999. Extrapolating from honeybees to bumblebees in pesticide risk assessment. Ecotoxicology 8(3):147-166.

Triplehorn, C. A., and N. F. Johnson. 2005. Borror and DeLong’s Introduction to the Study of Insects. 7th edition. Brooks Cole, Belmont, CA.

Tuell, J. K., A. K. Fiedler, D. Landis, and R. Isaacs. 2008. Visitation by wild and managed bees (Hymenoptera: Apoidea) to eastern U.S. native plants for use in conservation programs. Environmental Entomology 37(3):707-718.

USDA, NRCS. 2011. The PLANTS Database. National Plant Data Center, Baton Rouge, LA http://plants.usda.gov

Van Dyck, H., A. J. Van Strien, D. Maes, and C. A. M. Van Swaay. 2009. Declines in common, widespread butterflies in a landscape under intense human use. Conservation Biology 23(4):957-965.

Vaughan, M., M. Shepherd, C. Kremen, and S. H. Black. 2004. Farming for bees: guidelines for providing native bee habitat on farms. The Xerces Society, Portland, OR.

Velthuis, H. H. W., and A. van Doorn. 2006. A century of advances in bumblebee domestication and the economic and environmental aspects of its commercialization for pollination. Apidologie 37(4):421-451.

Watanabe, M. E. 1994. Pollination worries rise as honey bees decline. Science 265(5176):1170.

Whitney, E. N., D. B. Means, and A. Rudloe. 2004. Priceless Florida: Natural ecosystems and native species. Pineapple Press Inc., Sarasota, FL.

Williams, N. M., D. Cariveau, R. Winfree, and C. Kremen. 2010. Bees in disturbed habitats use, but do not prefer, alien plants. Basic and Applied Ecology, In Press, Corrected Proof.

Williams, N. M., R. L. Minckley, and F. A. Silveira. 2001. Variation in native bee faunas and its implications for detecting community changes. Conservation Ecology 5(1):7.

Williams, N. M., R. Winfree, and E. McGlynn. 2009. Native bee benefits. Bryn Mawr College and Rutgers University.

Williams, P. 2005. Does specialization explain rarity and decline among British bumblebees? A response to Goulson et al. Biological Conservation 122(1):33-43.

Wilson, J. S., T. Griswold, and O. J. Messinger. 2008. Sampling bee communities (Hymenoptera: Apiformes) in a desert landscape: Are pan traps sufficient? Journal of the Kansas Entomological Society 81(3):288-300.

Winfree, R. 2010. The conservation and restoration of wild bees. Annals of the New York Academy of Sciences 1195(1):169-197.

Winfree, R., N. M. Williams, H. Gaines, J. S. Ascher, and C. Kremen. 2008. Wild bee pollinators provide the majority of crop visitation across land-use gradients in New Jersey and Pennsylvania, USA. Journal of Applied Ecology 45:793-802.

Xerces Society, The. 2011. Attracting native pollinators: Protecting North America's bees and butterflies. Storey Publishing, North Adams, MA.

Zayed, A. 2009. Bee genetics and conservation. Apidologie 40(3):237-262.

BIOGRAPHICAL SKETCH

Katharine Buckley graduated in 2011 with a master’s degree in entomology from

University of Florida. She received her bachelor of science in entomology from Purdue

University in August of 2009. As an undergraduate she did research on Florida mosquito population dynamics as affected by temperature, precipitation and hurricanes.

She plans on pursuing a PhD in entomology, working on native pollinator conservation and biological control in grapes at Washington State University where she is currently

enrolled.