Climate change is an important contributor to the modification of many bumblebee species’ range boundaries. It was linked to widespread decline at the southern edge of their distribution and to their inability to colonize new areas at the northern edge. Additionally, bumblebee decline is aggravated by other anthropogenic threats like land use change, agricultural practices and pathogen spillover. Predicted consequences are numerous, and could lead to severe economic and ecological impacts on human populations. A species-specific assessment of potential climate change impacts on North American bumblebees, based on the most recent global change scenarios as used in the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), was done for the first time. Using a massive dataset of georeferenced bumblebee observations and general circulation models, a series of species distribution models explore the impact of different climate change scenarios on climatically suitable areas of 30 bumblebee species. Northward range shifts occur in most bumblebee species’ projected climatic niches, revealing potential hotspots – places projected to be climatically suitable to multiple species – under future climate scenarios. Areas where species are likely to be lost in the absence of intervention are substantial, particularly in eastern parts of the continent. Models showed significant contractions of current ranges even under the very optimistic scenario in which all species disperse at 10 km/year. Results indicate that managed relocation as well as habitat management should be considered as a conservation strategy for some species. This research serves as a foundation for broader discussion and research in a nascent research area. It may assist in establishing localities where first conservation efforts could be directed for vulnerable bumblebee species.

Pedicularis sceptrum-carolinum is threatened in much of its geographic range. As it reproduces vegetatively only to a limited extent, the species relies heavily on pollinators for reproduction. One aspect often overlooked in restoration efforts is the pollination ecology of flowering plants and their pollinators. To get a better understanding of the pollination ecology of P. sceptrum-carolinum in northern Sweden, their flower visitors were identified, and the number of flower visits was counted for flowering individuals at three different sites. The nearby flowering plant community was inventoried around each investigated P. sceptrum-carolinum. It was tested whether the number of flower visits were similar among sites, between large and small populations, and between populations with a majority of fresh flowers and populations with a majority of senescent flowers. To investigate whether the nearby plant community is important for the pollination of the species, the correlation between flower visits of P. sceptrum-carolinum and nearby flower abundance and diversity was analysed. No difference was found between sites, between large and small populations, or between populations where the majority of flowers were fresh and populations with a majority of flowers which had senesced. A moderate positive correlation was found between number of flower visits and nearby flower abundance and diversity respectively, if one site with an exceptionally large abundance of Trifolium pratense was excluded. As abundance and diversity of the nearby flowering plants were themselves highly correlated, their individual effect on flower visits of the studied P. sceptrum-carolinum could not be told apart in this study.

Bumblebees have undergone dramatic declines both in Britain and further afield during the last century. Bumblebees provide a crucial pollination service to both crops and wild flowers. For these reasons, they have received a great deal of research attention over the years. However, the ecology of wild bumblebee nests and the interactions between nests and other species, particularly vertebrates has been somewhat understudied. This is largely due to the difficulty in finding sufficient nests for well replicated study and a lack of appropriate methods of observation. Here, methods for locating bumblebee nests were trialled. It was found that a specially trained bumblebee nest detection dog did not discover nests any faster than people who had received minimal instruction. Numbers of nest site searching queens provide a reliable indication of suitable nesting habitat (i.e. places where nests are more likely to be found). In order to investigate aspects of bumblebee nesting ecology wild nests were observed by filming or regular observations by either researchers or members of the public. Some buff-tailed bumblebee (Bombus terrestris) nests were collected and all the bumblebees were genotyped to identify any foreign individuals. A review of British mammalian dietary literature was conducted to identify those that predate bumblebees. Great tits (Parus major) were filmed predating bumblebees at nests and it was clear from the literature and observations that badger (Meles meles), pine martens (Martes martes) and hedgehogs (Erinaceus europaeus) predate bumblebee nests, as well as the wax moth (Aphomia sociella). No evidence for predation by any other vertebrate species was found. Behaviours recorded included parasitism by Psithyrus, apparent nectar theft and possible usurpation by true bumblebees, egg-dumping by foreign queens and drifting and drifter reproduction by foreign workers. These events may cause harm to colonies (for example, through horizontal transmission of pathogens, or exploitation of the host nest’s resources). Alternatively where for example, usurpation by true bumblebees, egg-dumping or drifting is successful, these alternative reproductive strategies may increase the effective population size by enabling a single nest to produce reproductives of more than one breeding female. These data found that wild buff-tailed bumblebees (B. terrestris) nests with a greater proportion of workers infected with Crithidia bombi were less likely to produce gynes than those with fewer infected workers. Gyne production also varied dramatically between years. There is a growing body of evidence that a class of frequently used insecticides called neonicotinoids are negatively impacting bumblebees. An experiment was conducted using commercial colonies of buff-tailed bumblebees (B. terrestris) which were fed pollen and nectar which had been treated with the neonicotinoid imidacloprid at field realistic, sub-lethal levels. Treated colonies, produced 85-90% fewer gynes than control colonies. If this trend is representative of natural nests feeding on treated crops, for example, oilseed rape and field beans or garden flowers, then this would be expected to cause dramatic population declines. In this thesis methods for locating bumblebee nests have been tested, new behaviours have been identified (for example, egg-dumping by queens and predation by great tits) and estimations for rates of fecundity and destruction by various factors have been provided. Doubt has been cast over the status of some mammals as predators of bumblebee nests and estimates for gyne production, nest longevity, etc, have been given. More work is needed, especially observations of incipient nests as this is when the greatest losses are thought to occur.

Social learning plays a valuable role in the lives of many animal taxa, sometimes allowing individuals to bypass the costs of personal exploration. The ubiquity of this behaviour may arise from the fact that learning from others is often underpinned by simple learning processes that also enable individuals to learn asocially. Insects have proven to be particularly valuable models for investigating parsimonious hypotheses with regards to social learning processes, due to their small brain sizes and the prevalence of social information use in their life histories. In this thesis, I use social insects to further investigate the mechanisms underlying more complex social learning behaviours and explore the circumstances under which social information use manifests. In the first chapter, I investigate the proximate mechanisms underlying social learning and demonstrate that even seemingly complex social learning behaviours can arise through simple associative learning processes. In Chapter two, I investigate whether bees are more predisposed to learning from conspecific cues and discover that social information is learnt to a greater extent than information originating from non-social sources. In Chapter four, I demonstrate that classical conditioning also underpins learning from evolved social signals in honeybees. Finally, I investigate whether social information is used adaptively by bumblebees: Chapter three demonstrates that joining behaviour in free-flying bees is contingent upon whether flowers are familiar or not, and in Chapter six, I show that when social information is costly to acquire, bees are more likely to rely on social information to make foraging decisions. Taken as a whole, my findings suggest that bees may be specially adapted for receiving social information, but the ability to learn from others arises through general associative learning mechanisms.

Bumblebees are important pollinators that are said to be in decline all over the world. Swedish bumblebees have now been inventoried through a nationwide monitoring program for which data for the first time have been analysed. The aim here was to give an indication of how 35 bumblebee species in Sweden have changed in occurrence over five years interval in the national inventory program. The results given in this report showed that 5 bumblebee species (B. lucorum, B. magnus, B. cryptarum, B. terrestris and B. pascuorum) had declined significantly. Furthermore bumblebees as a group had a significant decline while none of the species had increased. Moreover, declining Swedish bumblebees are species that are common in Europe while species that are declining in Europe do not seem to decline in Sweden. This result was somewhat unexpected, considering the European status of bumblebee species. Furthermore results showed that early emerging species and species living in ubiquitous and open areas had declined. This was also surprising comparing with previous research. Lastly the results showed that two bumblebee species (B. hortorum and B. terrestris) have shifted towards the south within Sweden while none shifted towards north or in a west-east direction.

Pollination, one of our ecosystem services, is considered to be in critical condition due to a worldwide reduction in pollinators and their biodiversity. As the agricultural landscape becomes more and more intense, the pollinators lose important food and living resources.     In temperate ecosystems, bumblebees (Bombus spp) are an important group of wild pollinators, and as with pollinators in general, they are declining in both abundance and richness, in Sweden as well as other countries.      The purpose of this study was to see if bumblebee species number of a location is linked to the location’s number of flowering angiosperm species in northwestern Scania when examining eight locations, and to see if gene flow existed between four chosen locations.        The result of this study suggests that it is not possible to tell from the flowering angiosperm species how many bumblebee species that will be abundant, but that it might be possible to tell the number of bumblebee individuals. With the number of bumblebee species, the abundant Fabaceae species was more important than the total number of flowering angiosperms of the location. The number of abundant Fabaceae species was strongly correlated to the bumblebee diversity index of the locations, indicating that it is a group of flowers closely linked to bumblebees.      To see if gene flow occurred between the chosen locations, mtDNA sequences were compared in neighbor joining trees. The result showed that though some tendencies of isolation existed, gene flow seemed to occur in general between the locations in that fragmented and human dominated landscape of northwestern Scania.

It has been argued that organic farming sustains a higher biodiversity than conventional farming. This might promote the ecosystem services that exist in agricultural landscapes such as pollination and pest control. Here, I examined the effect of farming system (organic vs. conventional) with respect to the time since farming system transition, landscape heterogeneity and plant richness on pollinating and predatory insects. In total, data from 30 farms were used, of which 20 were organic and 10 were conventional. The data were analyzed using general linear models and model averaging. The results show that insect groups responded differently to various factors. Pollinators were more sensitive to landscape complexity, showing an increase of abundance and species richness with an increased heterogeneity. Predators on the other hand reacted to farming system, where there was an increase in abundance and species richness on organic farms.

In cranberry (Vaccinium macrocarpon Aiton) cultivation, farmers typically rent colonies of honey bee (Apis mellifera) for pollination. However, the efficiency of this bee at pollinating cranberries in Oregon, as in other regions, is questionable. Bumble bees (Bombus sp.) are reportedly effective in other regions, but their impact in Oregon is unknown. My objectives were to: (i) Compare bumble bee and honey bee pollination efficiencies under caged conditions; (ii) Estimate the abundance of bumble bees, honey bees, and other pollinators on an Oregon cranberry farm; and iii) Analyze and compare sources of pollen collected by bumble bees and honey bees in Oregon cranberries. In comparing pollination efficiencies of bumble bees and honey bees under caged conditions, the analysis of variance of data from the cage study indicated that bumble bee and honey bee pollinated plants yielded statistically equivalent average numbers of cranberries (1421 ± 302.5 and 1405 ± 347.6 berries/m², respectively) and weight of berries (11.5 ± 2.42 and 11.5 ± 2.77g/m²). However, bumble bees may have increased fruit set in honey bee treatments. On one occasion, bumble bees were found in the honey bee treatment, and may have contributed to the pollination of flowers in these plots. To estimate the abundance of pollinators, visual observations and were blue vane traps were utilized. Thirty-four timed visual observations in transects of cranberry beds were performed over on four dates during cranberry bloom. Blue vane traps were set-up on five occasions during bloom for two day periods. In the visual observations, honey bees (3.5 ± 0.58/min) were observed more frequently than bumble bees (1.2 ± 0.20/min). Bumble bees of four species made up 69.1% of trapped bees while honey bees made up 16.6% of bees caught in blue vane traps. On an Oregon cranberry farm during bloom periods in 2009, 2010 and 2011, pollen was collected from honey bee colonies using pollen traps. In 2010 and 2011, pollen was hand collected from reared bumble bee (B. vosnesenkii) colonies at the same farm. A total of 2937 honey bee pollen loads and 171 bumble bee pollen samples (137 scopal pollen loads, and 34 samples from with the colony) were individually acetolyzed and compared to a reference collection using light microscopy. Each pollen load was homogenized and 100 pollen grains were identified and counted to determine the percentage of each pollen type. Pollen collected by honey bees consisted of 29.1 ± 1.4% (2009), 18.3 ± 2.4% (2010), and 23.0 ± 1.1% (2011) cranberry pollen. Cranberry pollen contributed a higher percentage (56.0 ± 6.1%, and 70.4 ± 4.3% in 2010, and 2011, respectively) in bumble bee collected pollen. Both bee species collected pollen from non-target plants including those in the following families: Asteraceae, Fabaceae, Ranunculaceae, and Roseaceae. Native bumble bees (B. vosnesenskii) collected more cranberry pollen than pollen from non-target plants, and consistently collected a higher proportion of cranberry pollen than honey bees. The results of these studies suggest that native bumble bees may be adequate for cranberry pollination in Curry County, Oregon. However, the size of bumble bee populations may vary from year to year due to climactic conditions, availability of nesting sites, and forage before and after cranberry bloom. Thus the dependability and consistency pollination services rendered to cranberry crops by bumble bees needs to be further investigated in relation to population fluxuations. / Graduation date: 2012

Society depends on bumblebees for the ecosystem service in the pollination of crops. Bumblebee declines have been documented, mostly due to intensification of agriculture and loss of species-rich semi-natural grasslands, an important bumblebee habitat. To conserve bumblebee diversity and the ecosystem service of pollination, it is important to do analyses on multiple spatial scales to see how the intensification of agriculture affects bumblebees at a landscape level. In this study, I related abundance of bumblebees in 476 sites in southern Sweden (total abundance and abundance of declining/non-declining, long-tongued/short-tongued, and species preferring open terrain vs. forest boundaries) to amount of land use types (semi-natural grassland, arable land, forest, water and “other land use”) at 34 spatial scales (radii 100 to 40,000 m). Arable land had a negative effect on total bumblebee abundance at scales from 464 to 10,000 m and forest had a negative effect at scales from 2929 to 5412 m. Semi-natural grassland showed no clear effects – however, the partial regression coefficients were consistently negative. Arable land had a negative effect on non-declining species, long- and short-tongued species and on species preferring forest boundaries at larger scales, e.g. regions dominated by agriculture. Forest had a positive effect at smaller scales on species preferring forest boundaries and a negative effect at larger scales on species preferring open terrain and on declining species. The results suggest that arable land is a non-habitat for bumblebees and that semi-natural grassland does not affect bumblebee abundance at a landscape level.

The interaction between pollinators and flowers affects the fitness of both pollinators and flowering plants, which can result in the co-evolution of pollinator and floral traits. Some pollination systems contain flower-dwelling ambush predators that attack visiting pollinators. This interaction could result in co-evolution of predator and pollinator traits as is typical in predator-prey systems. The presence of these predators could also have indirect negative fitness effects on flowering plants by killing or deterring pollinators. This raises the possibility that predator and floral traits also co-evolve. Furthermore, it is possible that this system is best described as a three-species game where predator, pollinator, and floral traits all co-evolve. The ultimate goal of my thesis is to explore this possibility. This is achieved in Chapter 5, which consists of a game theory model of the co-evolution of floral colour with predator and pollinator behavioural strategies. This model is novel, both within the pollination context described here and within a wider context. Furthermore this model is the first to propose that the evolution of floral colour might be affected by the presence of flower-dwelling ambush predators. This is particularly significant given that there has been little discussion about what floral traits might be adaptations to the presence of these predators. A secondary goal of this thesis is to explore how pollinators detect and respond to the presence of flower-dwelling ambush predators as an important subcomponent of predator-pollinator-flower co-evolutionary dynamic. Chapter 2 demonstrates that bumblebees avoid evidence of past predation events, and Chapter 3 demonstrates that the honeybee recruitment dance is affected by exposure to cues of predation risk in a way that should reduce the colony's exposure to predators. Chapter 4 is a model that suggests novel factors that might affect how a population of pollinators distributes itself between predator-free and predator-containing flowers. / Thesis / Doctor of Philosophy (PhD)