Poster exhibited at GPSC Student Showcase, February 24th, 2016, University of Arizona. / We investigated whether bees have an innate preference for flowers with saturated pigments and whether experience altered any preference. Preference could be a result of reward quality varying by color morph and/or responses to the petals, anthers, or their combination. Consequently, we gave bees experience on one of four floral configurations created from two color morphs of Solanum tridynanum. We subsequently tested learned preference using an array of all four configurations. Changes in preference as a result of experience were not mediated by anthers, only by petals. Bees that first experienced configurations with purple petals subsequently preferred configurations with purple petals, relative to naïve bees. However, bees that first experienced white petals showed no subsequent change in preference relative to naïve bees. Surprisingly, naïve bees showed no preference for any particular floral configuration. Rather than an innate preference for flowers with more saturated colors, bees are less able to develop a preference for unsaturated types. Because individuals are more able to develop a preference for saturated flowers, these flowers experience greater visitation, and thus greater pollination success, over unsaturated types.

The risks posed by rapidly evolving RNA viruses to human and animal health are well recognized. Epidemics in managed and wildlife populations can lead to considerable economic and biodiversity losses. Yet, we lack understanding of the ecological and evolutionary factors that promote disease emergence. Host-switching viruses may be a particular threat to species important for human welfare, such as pollinating bees. Both honeybees and wild bumblebees have faced sharp declines in the last decades, with high winter mortality seen in honeybees. Infectious and emerging diseases are considered one of the key drivers of declines, acting in synergy with habitat loss and pesticide use. Here I focus on multihost viruses that pose a risk to wild bumblebees. I first identify the risk factors driving viral spillover and emergence from managed honeybees to wild bumblebees, by synthesising current data and literature. Biological factors (i.e. the nature of RNA viruses and ecology of social bees) play a clear role in increasing the risk of disease emergence, but anthropogenic factors (trade and transportation of commercial honeybees and bumblebees) creates the greatest risk of viral spillover to wild bees. Basic knowledge of the pathogenic effect of many common pollinator viruses on hosts other than A. mellifera is currently lacking, yet vital for understanding the wider impacts of infection at a population level. Here, I provide evidence that a common bumblebee virus, Slow bee paralysis virus (SBPV), reduces the longevity of Bombus terrestris under conditions of nutrition stress. The invasion of Varroa destructor as an ectoparasitic viral vector in European honeybees has dramatically altered viral dynamics in honeybees. I test how this specialist honeybee vector affects multi-host pathogens that can infect and be transmitted by both honeybees and wild bumblebees. I sampled across three host species (A. mellifera, B. terrestris and B. pascuorum) from Varroa-free and Varroa-present locations. Using a combination of molecular and phylogenetic techniques I find that this specialist honeybee vector increases the prevalence of four multi-host viruses (deformed wing virus (type A and B), SBPV and black queen cell virus) in sympatric wild bumblebees. Furthermore, wild bumblebees are currently experiencing a DWV epidemic driven by the presence of virus-vectoring Varroa in A. mellifera. Overall this thesis demonstrates that wild bumblebees are at high risk of viral disease emergence. My research adds to the ever-expanding body of evidence indicating that stronger disease controls on commercial bee operations are crucial to protect our wild bumblebees.

Changes in phenology due to climate warming could disrupt temporal overlap between interacting organisms when previously synchronized species respond to climate change at different rates. Phenologies of plants and insects are known to be sensitive to temperature and/or timing of snowmelt, with warmer temperatures and earlier snowmelt generally advancing spring flowering and emergence; however, some groups of pollinators, such as solitary bees, have been little explored in this context. One striking aspect of eastern hardwood forests is the emergence of understory wildflowers each spring, most of which rely, at least to some extent, on wild native pollinators for seed set. Without an understanding of the environmental drivers of phenology of these species, we have little ability to predict whether pollinators will continue to be well synchronized with flowering as the climate changes. In this study, I determined how spring temperatures and timing of snowmelt influence the phenology of spring wildflowers, activity of bees, and their temporal overlap in Gatineau Park, Québec. From 2013 to 2018, I characterized bee activity phenology and flowering phenology of understory plants in multiple study plots, focusing on early-flowering Anemone spp. and later-flowering Trillium grandiflorum. The sampled bee community was dominated by Andrena, Lasioglossum, and Nomada, all of which have similar activity periods. Degree-day accumulation was a better predictor of Anemone and Nomada phenology than were day of year or snowmelt date, whereas T. grandiflorum appeared to be more sensitive to photoperiodic cues; since day of year was the variable that best described its phenology. Activity periods of Andrena and Lasioglossum were equally well described by degree-day accumulation and by day of year. No taxon’s phenology was best predicted by snowmelt date. Despite these differences among taxa in the identities of the best predictors of phenology, bee activity and plant flowering phenologies responded at similar rates to interannual and among-site variation in snowmelt date and early spring temperature. Temporal overlap between flowering and bee activity was similar over the years of this study and was affected neither by snowmelt date nor by temperature. These results suggest that interacting plant and bee taxa may respond to different environmental variables but still maintain their synchrony under the conditions recorded so far.

No description available.

Temperate grasslands and the services they provide are threatened with severe degradation from human-driven land use changes. Among the worst affected services is pollinator support with grassland degradation contributing to the global decline in insect abundance due to habitat loss and a lack of floral resources. This has prompted conservation organizations to support pollinator conservation on working landscapes by increasing floral resources, but gaps remain in the consistent establishment of native forbs in intensively managed agricultural areas. We evaluated factors that influence native forb establishment during seed-based enrichment planting of tall fescue-dominated cattle pastures in two separate experiments: one testing a range of site preparation treatments with different seasons of sowing and comparing their effects on tall fescue suppression and seeded native plant establishment and the other investigating the effects of seed rate and pre-seeding cold stratification on native forb establishment on separate plots. For the site preparation experiment, we observed a mean target plant density of 0.12 target plants per m2 (SD = 0.247) The greatest target plant stem density (P < 0.0001) and species richness (P < 0.001) was in plots treated with a 2% glyphosate solution and sown with native seeds in early summer with the next five best treatments composed solely of fall sown replicates. For the seed and stratification experiment, we observed a mean target plant density of 88 target plants per m2 (SD = 73.9). Higher seeding resulted in greater target plant abundance in plots (P < 0.0001) with a seed rate of 56 kg/ha-1 resulting in almost three times as many target plants compared to 2.24 kg/ha-1. Pre-seeding stratification resulted in an increase in target plant abundance (P < 0.01). Target species richness was consistent between treatment levels. Results suggest that native forb establishment can be enhance by eliminating pasture grasses prior to seeding and the use of high seeding rates sown in the fall or using stratified seed. / Master of Science / Temperate grasslands and the services they provide are threatened with severe degradation from human-driven land use changes. Among the worst affected services is pollinator support with grassland degradation contributing to the global decline in insect abundance due to habitat loss and a lack of floral resources. To reverse this decline, conservation groups are encouraging the use of native plants throughout the landscape especially on farms and ranches to provide more resources for insect pollinators. One exciting opportunity exists in planting wildflowers into tall fescue-dominated cattle pastures that occupy millions of hectares of land in the Southeastern United States to provide food for pollinating insects. However little information exists on how to successfully establish wildflowers as much expertise is based on work done in the tallgrass prairie region of the Midwest. This study's goal was to investigate what control wildflower establishment by evaluating the success of an existing establishment experiment testing several site preparation techniques and different sowing seasons. A separate experiment was set up looking at the effect different seed rate and cold moist stratification had on establishment success of wildflowers. For the site preparation experiment, establishment was low for all treatments with a mean target plant density of 0.12 target plant per m2 (SD = 0.247). Summer sown 2% glyphosate had the highest wildflower richness and abundance at 0.35 target plants per m2 (SD = 0.247) and fall sown treatments were found to have higher sown wildflower abundance and richness than summer sown treatments. For the seed and stratification experiment, we observed a mean target plant density of 88 target plants per m2 (SD = 73.9). Target plant abundance did change between treatment levels with the highest and second highest seed levels yielding nearly three times and twice as many sown wildflowers as the lowest treatment respectively. Stratification resulted in an increase in sown wildflower abundance and sown wildflower richness did not differ significantly between treatment levels. Results suggest that native forb establishment can be enhance by eliminating pasture grasses prior to seeding and the use of high seeding rates sown in the fall using stratified seed.

As pollinator populations decline globally, public interest in creating pollinator gardens to help sustain bee abundance and diversity is rising, and there are many lists of recommended pollinator plantings and suggestions for bee-friendly flowers. However, these suggestions often lack grounding in empirical data. While anecdotal and expert suggestions are not without merit, it is crucial that we have clear, replicable, evidence-based planting recommendations to help boost the abundance and diversity of pollinators that will visit the plants. To fill this gap, we created a pollinator garden containing 25 pollinator plantings, some on previous recommendation lists, some not, and that were either native or non-native and perennial or annual. We surveyed the number and taxonomy of pollinator visitors to these plants through non-destructive methods. These data were analyzed to determine which plants are most effective at attracting abundant and diverse flower-visiting insects. We analyzed preference between native and non-native plants, perennials and annuals, and among different plant species. Our data revealed significant variation in the total abundance and diversity of flower-visiting attracted insects among plant species. Brown-eyed Susans attracted the highest abundance of insect visitors (average number of visitors per day = 53) and were 26-fold more attractive than Begonias, which attracted the lowest abundance of insect visitors (average number of visitors per day = 2). Lavender attracted the highest diversity of insect visitors (Simpson's Reciprocal Diversity Score: 3.5) compared to Pineapple Sage (Simpson's Reciprocal Diversity Score: 1.6), which attracted the lowest diversity of insect visitors. Additionally, we found that native perennials significantly attracted the greatest abundance of visitors compared to either non-native annuals or non-native perennials (p < 0.001). We conclude that ornamental landscape plants can support an abundance and diversity of pollinator visitors, and planting schemes should take into consideration the effects of plant species, plant lifespan, and plant origin. We can use these data to better inform the regional community how to attract and support abundant and diverse pollinator populations within urban and sub-urban ornamental landscape gardens. / Master of Science in Life Sciences / Many people want to build gardens with flowers that are attractive to pollinators, which can be bees, flies, butterflies, moths, and other insects. How, though, do we know what should be planted in that garden? Anyone with access to a computer can find lists of recommended plants for a pollinator garden, and they are provided by a variety of sources. However, the plants included on these lists are often not the same. Some lists may claim a handful of plants are the most attractive to pollinators, while another list will suggest different plants as the most attractive. In addition, these lists often are not based on scientific evidence. Our goal was to use these recommendations to build a pollinator garden so that we can collect data on the plants and determine which plants are best in our garden at attracting insect pollinators. We created a pollinator garden that includes 25 of popular landscape plants, many of which are found on lists of current plant recommendation lists. Then we let the plants grow and bloom, and during the summer, we counted and identified the different species of insects found on each plant. We then ranked plants, from best to worst, at their ability to attract insects and pollinators in numbers and in diversity. Additionally, we found that insects are generally more likely to visit plants that are native to the region. Using this information, we can share our results with the public so that anyone can build their own pollinator garden that will be the most attractive to pollinators.

Monopoloid potato (2n = 1x = 12) can be derived either paternally through anther/microspore culture or maternally through crossing with a haploid-inducing pollinator. Evidence from other genera suggests that androgenic and gynogenic haploid populations derived from the same parent may differ due to gametic selection and/or epigenetic factors. Our objectives were to derive androgenic and gynogenic monoploid populations from each of two diploid (2n = 2x = 24) Solanum phureja clones and compare their phenotypic characteristics in a greenhouse study. A haploid-inducing pollinator, S. phureja IVP101, was crossed to two selections of S. phureja (PP5 and BARD1-3). A total of 185 fruit was obtained from PP5 and 398 from BARD1-3, resulting from 85% and 65% fruit set, respectively. Seed lacking the dominant embryo spot marker carried by IVP101 were selected and germinated in vitro. From 29,300 PP5 x IVP101 seeds, 278 were spotless, resulting in 27 monoploids. Approximately 37 monoploids were obtained from the 99,500 BARD1-3 x IVP101 seeds of which 500 were spotless. In anther culture, PP5 and BARD1-3 yielded 0.16 and 1.67 embryos per anther respectively of which 51% and 44% of the regenerants were monoploid. A total of 32 anther-derived monoploids has been obtained from PP5 and 130 from BARD1-3. Rooted cuttings of 21 androgenic and 21 gynogenic PP5 monoploids were established in a greenhouse in each of three randomized complete blocks. Although the anther-derived monoploids appeared more vigorous, none of the differences between the two populations were significant. Flow cytometry revealed that anther derived progeny of PP5 and BARD1-3 had 3-8% greater nuclear DNA content than gynogenic monoploids derived from the same parental clones. / Master of Science

My dissertation and research focused on the effects of insecticides on squash bugs, its egg parasitoid, and pollinators in the production of cucurbits in Virginia. Plants in the cucumber family are dependent on insect pollination for successful fruit set, and are also susceptible to plant eating insects. Squash bugs are capable of transmitting cucurbit yellow vine decline, and their feeding can cause significant wilt and death in many varieties. To control for squash bug and other pests, growers commonly combine the application of broad-spectrum insecticides with the frequently applied prophylactic fungicides. Broad-spectrum insecticide applications are known to have negative effects on natural enemy populations, are capable of promoting insecticide resistance, and can have negative effects on pollinators if care in their use is not taken. Squash bugs have several natural enemies, but their predominant egg parasitoid is most effective at reducing damaging populations. The scelionid wasp Gryon pennsylvanicum Ashmead, is a prevalent egg parasitoid in Virginia and can be negatively affected by the application of broad-spectrum insecticides. Through survey efforts I found that G. pennsylvanicum is widely distributed throughout Virginia and is capable of high rates of egg parasitism (>90%). This is contrary to the 20% level previously assumed for the East Coast. I explored the effects of narrow-spectrum insecticides on the fate of the egg parasitoids, those developing in the host egg and emerged adults of G. pennsylvanicum. Contact assays showed that the insecticides λ-cyhalothrin and sulfoxaflor had caused high adult parasitoid mortality. As new insecticides get registered for use there is often concern about their effect on pollinators, specifically the European honey bee Apis mellifera L. I evaluated the use of large flight cages as a method to measure the sub-lethal effects of narrow-spectrum insecticides to honey bees, as a means to qualify risk. The method utilizes small colonies of honey bees (with stores of nectar and pollen) and their feeding at a treated sucrose solution after being trained to a feeder in an enclosed arena. This choice-test style behavioral experiment shows promise in qualifying the risks associated with insecticide exposure in the field. In the case of pyrifluquinazon, colonies repeatedly choose to avoid feeding at tainted feeders even after training with no other outside sources of food present. Further researching the sub-lethal behavioral effects that insecticides have on bees in a colony can help us better qualify their risk. / Ph. D.

Several aspects of plant-pollinator interactions are presented in the thesis. It contains a review on the open questions of plant-pollinator interactions from single species to complex networks. The following sections document novel results. Firstly, the conservation of complex pollination networks is addressed through the hierarchy of species' importance. In addition, the habitat requirements and interactions of a threatened rare pollinator species are explored. In the following chapters, the results from manipulative approaches applied in the field to plant-pollinator interactions are presented. The effect of pollinator's population decline on pollinators' foraging for pollen is investigated. Moreover, the way plant species loss impact several aspects of pollinator visitation is presented. Lastly, the impact of species removal on plant-pollinator network topology and on species ability of establishing new interactions is investigated.

Due to the growing recognition of the social and ecological consequences of the global decline in pollinator species, the need for more effective policies for the conservation of pollinator habitat is now more than ever. These trends call for research that provides a deeper understanding of farmers' decision-making processes. In this regard, this study tested a modified version of the Theory of Planned Behavior as a conceptual model for explaining farmers' perceptions and behavior regarding the adoption of pollinator conservation programs and practices. Specifically, the study tested how farmers' perceived behavioral control, attitudes, subjective norms, concern about herbicide resistance issues, and sociodemographic variables influence their intentions and actual adoption of pollinator conservation programs and practices. Quantitative survey data were gathered from 41 principal farm operators in the state of Illinois through the administration of a web-based survey. The resulting data were first explored using descriptive statistics and correlation analysis, following which multiple regression analysis was used to test four hypotheses on the predictors of farmers' intentions to adopt, as well as their adoption of pollinator conservation practices and programs. The results from the regression analysis showed that farmers' attitudes and their subjective norms had statistically significant positive effects on their adoption of pollinator conservation practices on their farms, as well as their intentions to adopt those practices in the near future. Perceived behavioral control also had a statistically significant positive effect on farmers' adoption of federal pollinator conservation programs, as well as their intentions to enroll in these programs in the future. Overall, these findings call for comprehensive pollinator conservation policies that facilitate the provision of information and incentives for farmers to voluntarily adopt pollinator conservation practices on their farms, as well as the provision of appropriate resources and opportunities for farmers to enroll in pollinator conservation programs over which they have minimal control.