Changes in Composition and Structure of a Wild Bee Community and Plant- Pollinator Interactions in South-Central Ontario Over a Forty-Nine Year Period

Changes in composition and structure of a wild bee community and plant- pollinator interactions in South-Central Ontario over a forty-nine year period

Claire Rubens

A thesis presented to The University of Guelph

In partial fulfilment of requirements

for the degree of

Master of Science in Environmental Sciences

Guelph, Ontario, Canada

CHANGES IN COMPOSITION AND STRUCTURE OF A WILD BEE COMMUNITY AND

PLANT-POLLINATOR INTERACTIONS IN SOUTH-CENTRAL ONTARIO OVER A

FORTY-NINE YEAR PERIOD

Claire Rubens Advisor: University of Guelph, 2019 Professor Nigel E. Raine

Wild pollinators provide important ecosystem services for both agricultural and natural ecosystems. While there is evidence of global pollinator declines, more long-term studies are needed to assess population trends, and the potential impacts of environmental stress factors such as land-use intensification and climate change. This is the first study to examine long-term changes in a wild bee community in Canada. Wild bee abundance, species richness, diversity and evenness were compared across three sampling periods (1968-1969, 2002-03, and 2016-17) in Caledon,

Ontario over 49 years. Despite decreases in wild bee abundance since 2002-03, the diversity, evenness and richness increased over time. Extensive restructuring (including loss and frequency changes) of plant-pollinator interactions from 2002-03 to 2016-17 appeared not to affect network resilience. While local trends in land-use patterns did not predict changes in this wild bee community, climatic changes in temperature and snowfall correlated with wild bee abundance at the site.

iii

ACKNOWLEDGEMENTS

I would like to thank my advisor Dr. Nigel Raine for taking me on as an undergraduate work-study student and a master’s student, and for helping me to focus my interests on wild bees. His enthusiasm and experience have been very helpful throughout this study. Thank you so much to Dr. Alana Pindar for her help in every aspect of my thesis. I am so grateful for all the time and work she put in, even from across the country. Thank you also to the members of my advisory committee, Drs. Laurence Packer and Madhur Anand for their feedback on my project. I really appreciate the time they spent to help me make the best version of my thesis.

Thank you to all my field assistants, especially Emily Agar, Ellen Richard and Sage Handler, who made long days in the field so much more enjoyable. Thank you also to all the members of the Raine Lab for their support and input during this whole process.

I would also like to thank Drs. Cory Sheffield, Jason Gibbs and Thomas Onuferko for their identification expertise. Thank you to the Royal Ontario Museum for allowing me to visit their insect collection to review bee specimens.

Thank you to Jennifer Grixti for continuing to study this site and providing me with clarifications and extra information during the planning stages. A special thank you to Monika Gschwendtner and family for their continued interest in this project over the years and allowing us to explore and spend time on their beautiful property.

This work was supported by funds from the Natural Sciences and Engineering Research Council (NSERC Discovery grant 2015-06783), the Food from Thought: Agricultural Systems for a Healthy Planet Initiative, by the Canada First Research Excellent Fund (grant 000054) and the W. Garfield Weston Foundation awarded to Dr. Raine.

Finally, thank you to my parents Peter and Marion, my brother Calvin, my partner Alex and my cat Marcie for their support, patience and advice throughout the past three years.

TABLE OF CONTENTS

Acknowledgements ...... iii

Table of Contents ...... iv

List of Figures ...... viii

List of Appendices ...... xi

CHAPTER 1 Introduction...... 1

1.1 The importance of pollination ...... 1

1.2 Drivers of global pollinator declines ...... 3

1.2.1 Climate change...... 4

1.2.2 Invasive species ...... 6

1.2.3 Land use intensification ...... 6

1.2.4 Pests and pathogens ...... 11

1.2 Objectives ...... 11

CHAPTER 2 General data collection methods for analysis of wild bee community composition and structure, and plant-pollinator interactions in Caledon, Southern Ontario ...... 15

2.1 Study site ...... 15

2.2 Sampling procedures ...... 18

2.2.2 Sweep netting ...... 20

2.2.3 Comparing sampling methods between studies ...... 20

2.3 Specimen processing, curation and identification ...... 22

CHAPTER 3 Assessing wild bee community structure and composition changes over a forty- nine year period in a natural habitat in Caledon, Southern Ontario ...... 24

3.1 Introduction ...... 24

3.2 Analytical methods ...... 28 v

3.3 Results ...... 31

3.4 Discussion ...... 39

CHAPTER 4 Changes in plant-pollinator interactions over a 15-year period in Caledon, Southern Ontario ...... 47

4.1 Introduction ...... 47

4.2 Analytical methods ...... 49

4.3 Results ...... 53

4.4 Discussion ...... 61

CHAPTER 5 Impacts of climate and habitat change on wild bee diversity in Caledon, Southern Ontario ...... 64

5.1 Introduction ...... 64

5.1.1 Habitat change ...... 64

5.1.2 Climate change...... 66

5.2 Analytical methods ...... 67

5.3 Results ...... 69

5.4 Discussion ...... 75

CHAPTER 6 Conclusions...... 81

References ...... 83

Appendices ...... 108

Appendix I ...... 108

Appendix II ...... 133

LIST OF TABLES

Table 1: Differences in sampling methods and intensity between the years 1968-69 (MacKay, 1970), 2002-03 (Grixti, 2004) and 2016-17 at the site...... 21

Table 2: The number of families, genera and species found at the site in each year (1968, 1969, 2002, 2003, 2016 and 2017)...... 32

Table 4: Diversity, evenness, abundance, and richness of the bee community (excluding bumblebees) in all sampling years collected by sweep net at the site. Bumblebee and pan trap data were excluded from this chart to facilitate comparison of the calculations between years and year pairs...... 37

Table 5: Diversity, evenness, abundance, and richness of each bee guild (excluding bumblebees), collected by sweep net in each pair of sampling years (1968-69; 2002-03; 2016-17). Bumblebee and pan trap data were excluded from this chart to facilitate comparison of the calculations between year pairs...... 38

Table 6: Measures of plant-pollinator interaction network nestedness calculated using 50 runs of each null model (EE- equiprobable row and column totals, CE- proportional row and column totals, EF- equiprobable row totals and fixed column totals, FE- fixed row totals and equiprobable column totals) with the NeD software for the year pairs 2002-03 and 2016-17. NODF (Nestedness metric based on Overlap and Decreasing Fill) is the calculation of nestedness of the matrix input into the NeD program. The Z-score is a measure of how nested the focal matrix is compared to the null model. The RN (relative nestedness) is the degree of nestedness of the focal matrix compared to the null model. The P value corresponds to the Z-score, where the focal matrix is considered to be nested (compared to the null matrices) at a value of <0.05...... 59

Table 7: Climate variables near the site for each study year. Average minimum and maximum temperature are recorded for the study months (April, May, June, July, August, September), whereas rainfall and snowfall are annual totals. This information was obtained from the Government of Canada website, using a compilation of information from a variety of weather stations within 12.14-24.45 km away from the site...... 68

Table 8: R2 values, F statistics and p values for each multiple linear regression examining the relationship between climate (rainfall, snowfall, maximum temperature and minimum temperature) and evenness, diversity, richness and abundance for all functional guilds, except bumblebees. Significant values are represented by an asterisk...... 71

Table 9: Complete list of wild bee species, collected by pan traps and sweep net in the years 1968, 1969, 2002, 2003, 2016 and 2017 at the site. Functional guild classification for each vii

species is listed. The resources used for identification are also listed. The ‘*’ symbol represents specimens that are thought to have been misidentified. They appear to be missing from the collections and I was unable to verify their identification. The ‘**’ symbol represents specimens which need further verification through expert opinion and/or barcoding...... 108

Table 10: The “core species” at the site in each sampling year. Large numbers of these species appeared to drive trends in the overall wild bee community (e.g. abundance, guild proportions), including year to year variance...... 132

Table 11: Adjusted R2, Significance F and B values for each linear regression examining the relationship between individual climate variables (minimum temperature, maximum temperature, rainfall and snowfall) and evenness, diversity, richness and abundance for all functional guilds, except bumblebees. The ‘*’ symbol represents significant values (p<0.05)...... 133

viii

LIST OF FIGURES

Figure 1: An undated aerial photo of the site obtained from MacKay’s 1970 MSc thesis. The site of the house and pond is located within the red box (43°48'53.39"N 79°58'36.59"W). The location of the trees knocked down in a storm sometime between 2005 and 2016 is located in the blue box...... 17

Figure 2: A satellite image of the site in 2005, obtained from Google Earth (43°48'53.39"N 79°58'36.59"W). The site of the house and pond is located in the red box. The location of the trees knocked down in a storm sometime between 2005 and 2016 is located in the blue box. .... 17

Figure 3: A satellite image of the site in 2016, obtained from Google Earth (43°48'53.39"N 79°58'36.59"W). The site of the house and pond is located in the red box. The location of the trees knocked down in a storm sometime between 2005 and 2016 is located in the blue box. .... 18

Figure 4: Pan trap array composed of 30 plastic bowls (blue, white and yellow) located on the lawn between the house and man-made pond at the site...... 19

Figure 5: Venn diagram of the cavity nesting species collected in each pair of sampling years (1968-69: red circle; 2002-03: yellow circle; 2016-17: blue circle) at the site. Overlapping portions of each circle represent the total numbers of species in common between each pair of studies, with the middle section representing the species found in common between the three studies. See Table 9 for the specific species collected in each year...... 33

Figure 6: Venn diagram of the cleptoparasitic and social parasite species collected in each pair of sampling years (1968-69: red circle; 2002-03: yellow circle; 2016-17: blue circle) at the site. Overlapping portions of each circle represent the total numbers of species in common between each pair of studies, with the middle section representing the species found in common between the three studies. The genus Nomada is not included in this figure. See Table 9 for the specific species collected in each year...... 33

Figure 7: Venn diagram of the social ground nesting collected in each pair of sampling years (1968-69: red circle; 2002-03: yellow circle; 2016-17: blue circle) at the site. Overlapping portions of each circle represent the total numbers of species in common between each pair of studies, with the middle section representing the species found in common between the three studies. See Table 9 for the specific species collected in each year...... 34

Figure 8: Venn diagram of the solitary ground nesting collected in each pair of sampling years (1968-69: red circle; 2002-03: yellow circle; 2016-17: blue circle) at the site. Overlapping portions of each circle represent the total numbers of species in common between each pair of studies, with the middle section representing the species found in common between the three studies. Andrena species A and B are not included in this figure. See Table 9 for the specific species collected in each year...... 34 ix

Figure 9: The percentage of wild bee individuals collected by sweep net at the site per study year, belonging to each bee functional guild (excluding bumblebees)...... 35

Figure 10: Rarefaction curves for the wild bee communities in each sampling year in 1968, 1969, 2002, 2003, 2016 and 2017 at the site, obtained by sweep netting and pan traps. The shaded regions depict standard error...... 39

Figure 11: Adapted from Almeida-Neto et al., 2008. It is important to note that these matrices do not show the frequency of interactions, but rather the presence or absence of a certain interaction between two species. For the purposes of my thesis, the flowering plant species and the bee species would be represented by the x and y axis of each matrix. Each box represents an interaction between one species of plant, and one species of bee. The zeros and ones in each box represent the presence or absence of each interaction. The top four matrices show some level of nestedness among the rows and columns. A nested matrix would result from a network where generalist plants interact with specialist pollinators, and specialist plants interact with generalist pollinators. The bottom four matrices show no nestedness among rows and columns...... 51

Figure 12: Plant-pollinator interaction network of wild bees and flowering plants at the site, observed and collected by net in the years 2016 and 2017. Five functional guilds (bumblebees, cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 61 flowering plant groups were recorded, with 1616 interactions. The gray bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant...... 55

Figure 13: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants at the site, observed and collected by net in the years 2016 and 2017. Four functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 54 flowering plant groups were recorded in 1312 interactions. The gray bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant...... 56

Figure 14: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants at the site, observed and collected by net in the years 2002 and 2003. Four functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 75 flowering plant groups were recorded in 8580 interactions. The gray bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant...... 57

Figure 15: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants observed and collected by net in the years 2002 and 2003. Only flowering plants found in both the consecutive year pairs 2002-03, and 2016-17 are shown in this network. Four bee functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 30 flowering plant groups are common between the two studies. The gray bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant...... 58 x

Figure 16: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants observed and collected by net in the years 2016 and 2017. Only flowering plants found in both the consecutive year pairs 2002-03, and 2016-17 are shown in this network. Four bee functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 30 flowering plant groups are common between the two studies. The gray bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant...... 59

Figure 17: The positive relationship between snowfall in centimetres and abundance of the total bee population at the site per sampling year (B = 26.46, F1,4 = 0.03)...... 73

Figure 18: The positive relationship between snowfall in centimetres and the abundance of social ground nesters at the site per sampling year (B = 19.76, F1,4 = 0.04)...... 74

Figure 19: The negative relationship between minimum temperature in degrees Celsius and social ground nester abundance at the site per year (B = -1232.39, F1,4 = 0.03)...... 74

Figure 20: Phenology of the bee community, depicted as a percentage of the number of species in each functional guild (A: cavity nesters, B: social ground nesters, C: solitary ground nesters, C: cleptoparasites and social parasites) collected per week of the year in 2002-03 and 2016-17. The signal processing (binomial method) technique was used to create a trendline for 2002-03 and 2016-17 since no specific trend (e.g. exponential, linear etc.) was expected...... 75

LIST OF APPENDICES

Appendix I………………………………………………………………………………...……108

Appendix II…………………………………………………………………………………...... 133

CHAPTER 1 Introduction

1.1 The importance of pollination

Ecosystem services are provided by nature at no cost to humans, and are essential for our wellbeing (Costanza et al., 1997; Daily, 1997; Mace et al., 2012). Pollination is one such ecosystem service, which benefits both natural and agricultural ecosystems (Kremen et al., 2007). Approximately 87.5% of all flowering plant species rely on animals to some extent for pollen transfer (Ollerton et al., 2011; Ollerton, 2017). In temperate regions 78% of flowering plant species rely on pollination by animals (Ollerton et al., 2011). In Canada there are at least 855 bee species, with nearly 50% (420 species) found in Ontario. Because of the high proportion of species compared to the national total, Southern Ontario is considered to be one of the top bee biodiversity hotspots (Sheffield et al., 2011; Pindar et al., 2017) following British Columbia and Southern Alberta.

Pollinators play an important role in both natural ecosystems and agroecosystems, ensuring the reproduction of the majority of flowering plant species including many insect-dependent crops (Klein et al., 2007; Pindar et al., 2017). Animal-mediated pollination occurs when pollen grains become attached to a flower-visiting animal’s body during collection of nectar, pollen, oils and/or resin, and are deposited on the stigmas of subsequent conspecific flowers (IPBES, 2016). As specialist foragers on floral resources, bees are perhaps the most important group of pollinators. Additionally, bees have numerous morphological structures that enable the collection, transportation and manipulation of pollen (Danforth et al., 2006).

Pollinators are essential for the reproduction of many species of wild flowering plants (Senapathi et al., 2017). Aside from forming base of the food chain, plants are also important to soil structure and provide shelter for invertebrates that provide a number of ecosystem services (Tepedino, 1979; Potts et al., 2003; Meiners et al., 2019). The relationships between plants and pollinators are closely interconnected and can be represented in a network structure of mutualistic interactions, called plant-pollinator networks (e.g. Memmott et al., 2004). In some ways these networks can be

sensitive to changes, such as the loss of plant or insect species, creating a cascading negative effect as interactions are broken (Dune et al., 2002). Alternatively, plant-pollinator networks can be resilient to some degree of disturbance due to “built-in” functional redundancy, which can act as a safeguard.

With our rapidly growing global population, predicted to reach nine billion by 2043, comes an ever-increasing requirement for food (Albrectsen, 2013). To meet these needs agricultural production has increased enormously (average crop yields have increased by 1.5% per year from 1951-2006 (Aizen et al., 2008b; 2009; 2019)). Pollinators are a vital component of the agriculture industry for the pollination services that they provide to crops intended for both humans, as well as livestock consumption (Southwick & Southwick, 1992; Ollerton, 2017). The production of 35% of the global food supply relies on pollinators to some extent (Klein et al., 2007). These pollination services are valued at $235 to $577 billion USD annually (IPBES, 2016). Agriculture has become increasingly dependent on pollinators, though wild bees are likely to be a key factor in the sustainable long-term production of fruit and vegetable crops (Aizen et al., 2008b). It is difficult to determine precisely what value these wild pollinators contribute to agriculture (Pindar et al., 2017).

In Ontario honeybees are used for field crop pollination, while bumblebees pollinate greenhouse crops. It is estimated that $897 million is generated from crops pollinated by managed honeybees and bumblebees each year in Ontario (Government of Ontario, 2014). Across the province, it is thought that insect pollination is necessary to produce 34 major crops covering about 2.67 million hectares of land (Statistics Canada, 2014). A recent report on the status and trends of pollinators in Ontario suggests this may even be an underestimate considering the lack of information regarding Ontario crop pollination (Pindar et al., 2017).

Although it is common practice to bring managed pollinators into agricultural systems, honeybees are often less effective pollinators than wild insects for a variety of different crops (Westerkamp, 1991; Klein et al., 2007; Breeze et al., 2011; Garibaldi et al., 2013). The increase in area of insect pollinated crops and the decline of pollination services by honeybees in the UK further supports the idea that wild pollinators are necessary for production of many crops (Aizen et al., 2008b; 2

Breeze et al., 2011). Honeybees are thought to be less effective pollinators of some crops mainly because they are generalists, and do not have specific co-adaptations with plants (Westerkamp, 1991). Furthermore, honeybees have been observed avoiding pollen on flowers (which necessitates grooming) while searching for nectar (Westerkamp, 1991). For example, honeybees visiting apple flowers will often access nectaries from the side of the flower completely missing any contact with anthers or stigmas.

Crop yields do not typically reflect global pollinator declines, though it is likely only a matter of time before pollinator deficits become common and widespread for agricultural crops (Aizen et al., 2008b). A recent study shows that phylogenetic diversity of bees is reduced in agricultural landscapes, resulting in decreased pollination efficiency of two apple varieties (Grab et al., 2019). Agricultural land has been demonstrated to have negative impacts on the well-being of wild bees, especially at the early developmental stages (Nooten & Rehan, 2019). Specifically, the stem nesting bee Ceratina calcarata forages less effectively in agricultural landscapes resulting in lower reproductive success (Nooten & Rehan, 2019). Agricultural pesticides can also negatively impact the pollination services provided to certain crops, such as apples (Stanley et al., 2015). Together, these suggest that agricultural land can negatively affect the crop pollination services that wild bees provide.

1.2 Drivers of global pollinator declines

Biodiversity loss is a serious global issue which affects ecosystem function and services (Folke et al., 2004; Rockström et al., 2009). Hallmann et al. (2017) concluded that in just over 27 years, flying insect biomass has declined 76%. There is some indication that rates of species loss may have slowed recently, though this is not necessarily a positive indication (Carvalheiro et al., 2013). Such reduced rates of species loss could be driven by a decreased number of species, meaning there are simply fewer species to lose (Carvalheiro et al., 2013). Current estimates suggest that more than 40% of the world’s invertebrate pollinator species are at risk of extinction (IPBES, 2016). Land-use intensification, climate change, invasive species, pests and pathogens, and most

importantly their interactions are the key factors driving these declines (Vanbergen et al., 2013; Sanchez-Bayo & Wyckhuys, 2019).

1.2.1 Climate change

Anthropogenic climate change has been identified by many studies as the most important factor in pollinator declines (Williams & Osborne, 2009; Forrest & Miller-Rushing, 2010; Burkle & Alarcόn, 2011; Prather et al., 2013). Climate change is becoming increasingly problematic over time for organisms globally, due to the shifts in temperature and simultaneous shifts in broader weather patterns (Aguirre-Gutierrez et al., 2016). Climate change can impact pollinators with range shifts and reductions, changes in plant-pollinator interactions, and phenological mismatches. It is important to both monitor and predict how climate change will impact organisms to plan for the future (Parmesan & Yohe, 2003).

Range shifts

The potential of a species to shift their natural range seems like a promising response to climate change, if the population can be maintained in a more suitable location. These shifts have been observed in latitude and elevation (Parmesan et al., 1999; Devictor et al., 2008). Studies have predicted that many animals will exhibit northern displacement in response to global climate changes to cope with rising temperatures (Parmesan et al., 1999; Currie et al., 2004; Parmesan et al., 2011; Pellissier et al., 2013). Poleward range shifts in response to climate change have been observed in butterflies (Parmesan et al., 1999). However, looking at historical data, bumblebees do not appear to be tracking climate change in the same way. North American and European bumblebee species are not shifting further north but are losing ground at the southern edges of their ranges, suggesting a reduction in overall range (Kerr et al., 2015). Recent models suggest that North American bumblebees may not be able to shift their ranges at the rate required for future climate change projections (Sirois-Delisle & Kerr, 2018). It appears that many species may have already reached their “geographical and thermal safety limits” leaving no suitable environments (Devictor et al., 2008; Huey et al., 2009). Shifts in latitude and elevation could isolate habitat

patches, and/or reduce the area of habitat, which could ultimately lead to extinction if this global warming trend continues (Wilson et al., 2005).

Phenological Mismatches

Many phenological events such as reproduction and emergence, arise from complex interactions among the species and environmental factors. Climate change has the potential to shift the timing of environmental cues used by plants and pollinators for development and survival. Phenological shifts have been among the most obvious and thoroughly documented biological responses to the climate warming (Memmott et al., 2007; Bartomeus et al., 2011; Forrest & Thomson, 2011; Polce et al., 2014). Because of their close associations, it is predicted that declines in pollinators will result in similar declines in native plants (Potts et al., 2010).

It has been shown that over the past 130 years, the phenology of ten bee species from the northeastern USA have advanced approximately five days (Bartomeus et al., 2011). A shift in plant phenology has also be observed, where flowering plants are blooming earlier (Burkle et al., 2013). Since plants and bees in temperate environments both use temperature as a cue for timing their seasonal activity, it is possible that they could shift synchronously resulting in no or minimal mismatches (Hegland et al., 2009; Fründ et al., 2013; Visser, 2013). Conversely, the phenologies of bees and plants could shift at different rates if environmental cues other than temperature are important (Menzel & Fabian, 1999; Inouye et al., 2003; Hegland et al., 2009). There are a number of studies investigating the timing of these relationships, suggesting that plant and insect phenologies are shifting asynchronously. There is some evidence to suggest that insect phenology has shifted more rapidly than plant phenology over recent decades (e.g. Gordo & Sanz, 2005; Parmesan, 2007; Willmer, 2014). Other studies suggest the opposite pattern where plant phenology has advanced more rapidly than bee phenology (Forrest & Thomson, 2011; Kudo & Ida, 2013). Regardless of the shift direction, specialist bees are impacted more than generalist species because of their specific floral host requirements (Burkle et al., 2013).

1.2.2 Invasive species

Invasive species are non-native species that are present outside of their native range and often dominate a landscape if uncontrolled (Vanbergen et al., 2018). Wild pollinators can be affected by both invasive animals, as well as plants. Invasive plants can interrupt regular pre-existing plant- pollinator interactions through competition for pollination (Lopezaraiza-Mikel et al., 2007). This interruption could be beneficial to pollinators if high quality nectar and pollen is provided as a food resource, or it could exacerbate pollinator declines by competing with and replacing native plants (Vanbergen et al., 2018). Although research is lacking for Ontario, it is known that there are more than 500 invasive plant species in this province, providing many opportunities for the disruption of native plant-pollinator interactions (Pindar et al., 2017). Altering these interactions can lead to a number of other consequences including impediments in bee colony success, reduced bee population size and distribution, and changes in bee community structure (Aizen et al., 2008a; Stout & Morales, 2009).

Invasive insects have the potential to compete with native pollinators for two resources that are key to their survival: floral resources and nest sites. In fact, nearly half of pollinators nesting in bee hotels were found to be non-native, which demonstrates the possibility of competition between native and introduced species (MacIvor & Packer, 2015). Similar results were found in artificial nesting cavities where less than 25% of cavities were occupied by native species (Barthell et al., 1998). Invasive insects can also interfere with the reproduction of certain native insect species. Native and alien subspecies of Apis mellifera and Bombus terrestris have been known to reproduce (De la Rua et al., 2002, 2009; Jensen et al., 2005), and different species of bumblebees were mated in a lab setting (Kanbe et al., 2008). Additionally, negative consequences to offspring fitness have been recorded when reproductive interference occurs (Kanbe et al., 2008). This evidence suggests that repeated inter-species and inter-subspecies mating over time could negatively impact the survivability of entire wild bee populations.

1.2.3 Land use intensification

Overall, urbanization and agricultural intensification can change the landscape by fragmenting, destroying, and/or degrading the natural landscape. Habitat loss is thought to be the most important 6

factor contributing to pollinator declines, including bee declines (Biesmeijer et al., 2006; Brown & Paxton, 2009; Vanbergen et al., 2013; Ollerton et al., 2014; Goulson et al., 2015). Habitat destruction and degradation can reduce or eliminate opportunities for bees to forage and nest (Steffan-Dewenter & Tscharntke, 1999), while habitat fragmentation leaves islands of habitat surrounded by less hospitable regions. Outcomes of urbanization and agriculture can be positive or negative for different pollinator species, showing that there is no “one-size-fits-all” landscape (e.g. Ahrné et al., 2009; Matteson & Langellotto, 2010; Benjamin et al., 2014).

There is conflicting evidence about the effects of habitat fragmentation on wild pollinator health. A number of studies suggest that habitat fragmentation can negatively impact wild pollinators by decreasing abundance, diversity and species richness (Rathcke & Jules, 1993; Brosi et al., 2007; Krauss et al., 2009; Winfree et al., 2009; Bommarco et al., 2010). Decreased genetic diversity and inbreeding in these “islands” can negatively impact the long-term viability of populations (Zayed, 2009). Alternatively, there is evidence for increased pollinator diversity in fragmented habitats (Ricketts, 2004; Russell et al., 2005; Brosi et al., 2007; Williams & Kremen, 2007). The size of the habitat patch appears to determine whether effects of habitat fragmentation are beneficial or detrimental to wild pollinator populations (Cane, 2001).

Certain characteristics may enable bees to respond better to habitat fragmentation. Generalist pollinators which forage on many different plant types, may tolerate habitat fragmentation more effectively than specialist pollinators, which focus on one or a few types of floral resources. For example, many solitary bees are specialist foragers, whereas bumblebees are generalists and more likely to be able to find food in a landscape where wild flowers are less abundant (Michener, 2004). Additionally, larger bodied bees can fly longer distances than smaller bodied bees, having a larger range to find appropriate resources (Greenleaf et al., 2007; Wray et al., 2014). This is demonstrated in expansive cropping systems where larger bodied bumblebees are more successful than smaller solitary bees (Gathmann & Tscharntke, 2002; Benjamin et al., 2014; Kremen et al., 2015). In modified habitats such as urban areas, or agricultural land, wild bees are reliant on pockets of natural and semi-natural land for resources to survive (Morandin & Kremen, 2013a; Meiners et al., 2019). It appears that semi-natural land is especially important for wild bee nesting sites (Meiners et al., 2019). 7

Urbanization

By 2050, it is estimated that 68% of the global population will live in urban areas (United Nations, 2019). Wild bees can be affected by urbanization in both positive and negative ways (e.g. Matteson & Langellotto, 2010; Geslin et al., 2013; Carper et al., 2014). In fact, intermediate levels of urbanization generally show the greatest bee species richness (Pindar et al., 2017). Although somewhat counter-intuitive, urban areas can provide habitat for certain bee species. Often there are abundant semi-naturalized pockets such as parks and gardens, which contain flowering plants and nesting habitat, within urban landscapes. One such type of semi-naturalized area are green roofs which unlike typical urban parks and gardens, are not limited by sunlight (Colla et al., 2009; Matteson & Langellotto, 2010). Outside of these semi-naturalized areas, human made structures can also serve as nesting sites for cavity nesting species in particular (MacIvor & Packer, 2015).

Conversely, urbanization can be detrimental to bee fitness. Lack of flowering plant diversity can result in sporadic synchronized blooming which does not provide the necessary nutrition for pollinators throughout their flight period. Additionally, many plants in urban areas are hybrid varieties that can have reduced nutritional value or have a structure that is too complex for pollinators to access the pollen and nectar within. The replacement of natural areas with human- made structures can also be detrimental to wild bees. For example, urban areas are covered with impervious surfaces such as pavement, which can reduce ground or cavity nesting sites for bees which require these for reproduction (McKinney, 2008). Road mortality is also an often- overlooked hazard for pollinating insects (Baxter et al., 2015). Finally, an abundance of tall buildings in urban centers can impede dispersal, foraging and orientation by blocking sunlight (Matteson & Langellotto, 2010).

Agricultural intensification

Agricultural intensification has been linked to the loss of wild pollinator habitat (Kim et al., 2006; Grixti et al., 2009; Marini et al., 2012). Agricultural land generally shows a lower abundance and diversity of wild bees when compared to natural areas (Kremen et al., 2004; Morandin & Kremen, 2013b; Cutler et al., 2015). Expansive monoculture cropping systems can provide floral resources

for pollinators but lack the variety and asynchronous flowering periods that ensure the nutritional needs of a wide range of wild pollinator species (Pleasants, 1980). Floral plantings in agroecosystems, and/or proximity to natural areas is beneficial to the wellbeing of pollinators, and therefore the cropping system (Duelli & Obrist, 2003; Holzschuh et al., 2008; Ricketts et al., 2008; Garibaldi et al., 2011; Bartomeus et al., 2014). Aside from contributing to more effective pollination services, natural areas or floral plantings can provide food resources during times when the main crop is not in bloom, and provide nesting habitat (Meiners et al., 2019).

Agrochemicals are widely used around the world due to agricultural intensification, though their use is highly regulated in most countries, including Canada (Potts et al., 2010). Risks of exposure of pollinators to pesticides (insecticides, fungicides, herbicides and miticides) and antibiotics are well documented (Potts et al., 2010; Cutler et al., 2014a, 2015b; Raine & Gill, 2015). It is possible that the effects could be more detrimental if insects come into contact with multiple pesticides (Colin & Belzunces, 1992; Thompson & Wilkins, 2003; Gill et al., 2012; Johnson et al., 2013; Thompson et al., 2014). Research of agrochemical impacts on pollinators has primarily focused on honeybees, with more recent work on bumblebees (Potts et al., 2010; Pindar et al., 2017). This is problematic as it has been well established that agrochemicals can vary substantially in their toxicity to different pollinator taxa (Desneux et al., 2007; Blacquière et al., 2012; Arena & Sgolastra, 2014; Godfray et al., 2014, 2015). This is somewhat understandable due to the frequent use of honeybees for their pollination services, though it is important to shift focus if we are to understand declines in wild bees.

The impacts of fungicides, miticides and herbicides on wild bees are generally an understudied topic, especially in Ontario (Pindar et al., 2017). The fungicide Benomyl has been shown to alter bee community composition (Cahill et al., 2008). The focus on the impacts of miticides in Ontario has been on honeybees and the miticides used for controlling Varroa mites (Hillier et al., 2013). There is no evidence of assessments on the impacts of herbicides on wild bees in Ontario, which is concerning since they are the most commonly used pesticide on agricultural crops in the province (Statistics Canada, 2006; Pindar et al., 2017). There is, however, a study suggesting negative impacts of herbicides on butterflies (Russell & Schultz, 2009). Research in other countries suggests that herbicides mainly impact pollinators indirectly through the elimination of wild plants, 9

and therefore the availability of nectar and pollen (de Snoo & Van der Poll, 1999; Kleijn & Snoeiging, 1997).

Neonicotinoids include compounds like imidacloprid, thiamethoxam and clothianidin, are the most commonly used class of insecticide around the world. Research surrounding this insecticide class of pesticides and their effects on pollinators has been plentiful in recent years, but controversial (Walters, 2013; Godfray et al., 2014, 2015; Lundin et al., 2015; Tsvetkov et al., 2017). Of course, insecticides have lethal effects and can simply cause mortality by exposure (Alston et al., 2007). Sublethal impacts of insecticides are more difficult to study due to their more subtle effects on pollinators. Exposure to insecticides has been shown to have negative impacts on social bees (Godfray et al., 2014) such as impairments to foraging (Gill et al., 2012; Gill & Raine, 2014; Stanley et al., 2015), colony development (Gill et al., 2012), and queen production (Whitehorn et al., 2012; Rundlöf et al., 2015) at field realistic levels of exposure. Other sublethal impacts include decreased longevity of adult bees, effects on learning and memory, and decreased navigation ability (Pindar et al., 2017). The sublethal impacts of insecticide exposure on solitary bees are less studied, though it is likely that similar or potentially more severe effects could occur (Rundlöf et al., 2015; Raine, 2018).

1.2.4 Pests and pathogens

The impacts of pests and pathogens on pollinators are best documented in managed honeybees (e.g. Varroa mites, and the deformed wing virus they transmit to bees), probably due to their economic value and widespread use in agriculture. Their impacts are less well-studied in wild bee populations (Pindar et al., 2017), though it is known that certain pests and pathogens can be transferred between managed bees and wild bees (Colla et al., 2006; Cameron et al., 2011). This phenomenon, called pathogen spillover, can occur when managed or invasive pollinators encounter wild pollinators during foraging. Deformed wing virus and a number of different parasites are known to spread between managed honeybees and bumblebees (Fürst et al., 2014; Graystock et al., 2014; McMahon et al., 2015). Commercial bumblebee colonies can also spread pathogens to wild bumblebees through shared flower visitation, putting wild bumblebees foraging near greenhouses are at a higher risk of becoming infected (Colla et al., 2006). A recent study found that honeybee viruses could be transmitted to UK hoverflies, providing an interesting element to pollinator disease dynamics (Bailes et al., 2018). Pests and pathogens negatively impact the health of both wild and managed pollinators (Colla & Packer, 2008; Otterstatter & Thomson., 2008; Desai et al., 2016). The added stressor of additional pests and pathogens passing from managed bees to wild bees could contribute greatly to declines in wild pollinator populations when combined with climate change, the presence of invasive species and/or land use intensification.

1.2 Objectives

One of the oldest comprehensive studies of a bee community in Ontario was initiated in 1968 (MacKay, 1970; MacKay & Knerer, 1979) in which 105 species and 9784 individuals were collected over two consecutive seasons from a study site in Caledon, Ontario. This site was re- sampled in 2002 and 2003 (Grixti, 2004; Grixti & Packer, 2006). Since two sites were sampled by MacKay, she referred to this particular site as “Site A”, while Grixti referred to it as “the site.” Throughout this thesis, I will refer to the study location as “the site.” Grixti sampled the wild bee community and compared four measures between 1968-69 and 2002-03: abundance, species richness, diversity and evenness. A total of 150 species and 10437 individuals were collected, 11

showing increases in overall abundance, species richness, evenness and diversity since 1968-69 (Grixti, 2004). Between 1968-69 and 2002-03 there was a considerable change in the composition of the wild bee community, with a similarity of only 7.5% (Morisita-Horn similarity index) (Grixti, 2004). Abundance, species richness, diversity and evenness were also calculated for different bee groups defined by nesting preferences. Ground nesting and parasitic bee populations showed greater species richness, diversity and evenness in 2002-03 compared to 1968-69. The guild of cavity nesting species was more diverse and had higher evenness in 2002-03 compared to 1968- 69. Overall, the relative abundance of cavity nesters and parasitic bees increased between 1968- 69 and 2002-03, while the relative abundance of ground nesters decreased over time. These differences were attributed to succession and habitat fragmentation resulting from human driven modifications of the site (e.g. construction of a house). Climate change was not found to be a factor in wild bee community composition change over time, though male bees were recorded flying later in the 2002-03 compared to 1968-69. This observation could be an indication that climate change is impacting the bee community at this site without affecting its composition.

This study builds on the work of MacKay and Grixti, resampling the wild bee community at the site. My first objective was to examine changes in wild bee community composition and structure at the site over three time points spanning a 49-year period. Globally, there are very few long-term studies of wild bee populations (e.g. Marlin & LaBerge, 2001; Burkle et al., 2013; Gardner & Spivak, 2014; Meiners et al., 2019). This is the first time that wild bee communities in Canada have been compared over almost 50 years. Assessing changes in wild bee populations over time is becoming increasingly vital given the potential impacts of the four main environmental stressors: climate change, land-use intensification, pests and pathogens, and invasive species (Vanbergen et al., 2013) on these pollinator communities. These environmental stressors not only affect pollinators, but also their interactions with flowering plants due to their mutualistic, and highly interconnected relationships (e.g. Burkle et al., 2013). My null hypothesis is that the structure, composition and function of the wild bee community at the site would be maintained since 2002- 03. I predicted that the bee community in 2016-17 would show a similar abundance, diversity, evenness and species richness to the bee community in 2002-03. I expected this pattern to be consistent for the overall bee community, and for the cavity nester, bumblebee, solitary nester,

social ground nester, and cleptoparasite and social parasite guilds that make up the overall wild bee community. My alternative hypothesis is that the structure, composition and function of the bee community would be negatively impacted from 2002-03 to 2016-17. I predicted a decrease in abundance, diversity, evenness and richness, for both the overall bee community and for the cavity nester, bumblebee, solitary ground nester, social ground nester, and cleptoparasite and social parasite guilds within it. I expected these decreases because there was evidence of habitat loss and climate change effects on the pollinator community in 2002-03, and these stressors appear to be becoming increasingly important and impactful over time (Grixti, 2004; Vanbergen et al., 2013).

My second objective was to examine the interactions between flowering plants and wild bees at the site, to determine the presence and extent of changes between 2002-03 and 2016-17. My null hypothesis is that interactions between flowering plants and wild bees at the site would be maintained. I predicted that the strength and number of interactions in the corresponding plant- pollinator interaction network would remain similar over time. Additionally, the structure of the network would remain ‘nested’ (see Figure 12), a measure of resiliency to disturbance (Bascompte et al., 2003; Ollerton et al., 2003; Thompson, 2005; Guimarães et al., 2006; Lewinsohn et al., 2006; Santamaría & Rodríguez-Gironés, 2007). My alternative hypothesis is that the plant-pollinator network interactions would be negatively impacted over the past 15 years. I predicted that there would be fewer and weaker interactions between flowering plants and wild bees. I also expected the networks to show a more compartmentalized structure over time, corresponding with a reduction in overall resiliency to disturbance.

My third objective was to further investigate the impacts of climate change and land use change, focusing on their effects on wild bee species richness, diversity, evenness and abundance. The robust wild bee datasets from 1968-69 and 2002-03 at the site, provide opportunities to investigate the effects of the four main stress factors in Caledon, Ontario. Grixti focused on two particular stressors: land use change and climate change, and the impacts on the bee population at the site. She investigated land use change locally, determining that succession and habitat fragmentation affected the wild bee community composition. There was no evidence that climate change impacted the wild bee community structure, though flight period of males appeared to be altered. My aim was to continue to investigate the effects of these stressors adding additional bee 13

community data from 2016-17, using a more quantitative approach. Increased rainfall, rising minimum and maximum temperatures, and decreased snowfall is expected in Southern Ontario as a result of climate change (Fausto et al., 2015). I expected that climate data from the three study periods (1968-69, 2002-03 and 2016-17) near the site would align with these projections. There have been considerable increases in the degree of urbanization in and around the Greater Toronto Area since the late 1960s (Puric-Mladenovic & Strobi, 2006). Overall the amount of pollinator habitat surrounding the site was expected to decrease from 1968-69 through 2002-03 to 2016-17. My null hypothesis is that changes in climate and land-use would not affect the wild bee community at the site. I predicted that the species richness, diversity, evenness and abundance of the overall bee community and the guilds of cavity nesters, bumblebees, solitary nesters, social ground nesters, and cleptoparasites and social parasites would not be correlated with the climate and land-use change data at the site over time. My alternative hypothesis is that the climate and land-use changes over time would affect the wild bee community negatively. I predicted that both factors acting individually and in combination would correlate with a decline in wild bee species richness, diversity, evenness and abundance of the total bee community and the cavity nesters, bumblebees, solitary nesters, social ground nesters, and cleptoparasites and social parasites.

CHAPTER 2 General data collection methods for analysis of wild bee community composition and structure, and plant-pollinator interactions in Caledon, Southern Ontario

2.1 Study site

Wild bee specimens were collected from the study site over two seasons in 1968-69, (MacKay, 1970; MacKay & Knerer, 1979) and over two more seasons in 2002-03 (Grixti, 2004; Grixti & Packer, 2006). This site was originally chosen, and the study continued in 2016-17 due to the high abundance and species richness of wild bees, and general lack of long-term studies on wild bee communities. MacKay referred to this study location as ‘Site A’ since there were two sites examined in her thesis (MacKay, 1970; MacKay & Knerer, 1979), while Grixti changed the name to ‘the site’ (Grixti, 2004; Grixti & Packer, 2006). This same study site is examined in my thesis and I will refer to it as ‘the site.’

The site is a private property located in Peel county, in the city of Caledon, Ontario (43°48'53.39"N 79°58'36.59"W). The property is approximately seven hectares in area, with two sides perpendicularly bordered by Forks of the Credit Road and McLaren Road. The site is situated in the western section of the Oak Ridges Moraine, in the mixed wood plains ecozone. This zone is densely populated and agriculturally intensive, with great amounts of species loss demonstrated (Gibbs et al., 2009). The site is located in an ecologically important area near the Greater Toronto Area, with potential exposure to the effects of urbanization, agriculture and climate change (Puric- Mladenovic & Strobi, 2006).

Since the late 1960s, the site has changed substantially through both natural succession and anthropogenic influences. In 1968-69, the site was classified as an early successional field (MacKay, 1970), with mainly perennial plants, such as grasses (Figure 1). In 2002 and 2003 the site contained both early and mid-successional stages (Grixti, 2004) (Figure 2). Along with this change in successional stage, many flowering plants found at the site in 2002 and 2003 by Grixti, were not recorded by MacKay in 1968-1969 (Grixti, 2004; Grixti & Packer, 2006). Additionally,

a large stand of coniferous trees fell in a severe storm sometime between 2005 and 2015, providing abundant dead wood for cavity nesting bees (Figure 3).

Major anthropogenic influences at the site include the construction of a house and pond on the property in the late 1970s. A series of trails throughout the property, and a long gravel driveway off Forks of the Credit Road were also added, presumably at a similar time. The house is located at the top of a grassy hill, which seems to have consistently served as an important nesting site for ground nesting bees throughout 1968 to 2017.

~ 500 m N

Flower visitor and flower community data were collected at the site from May to October in the years 2016 and 2017. Previously, sweep nets were used to collect all specimens in 2002-03, and 1968-69 (MacKay, 1970; Grixti, 2004). For the 2016-17 seasons, the site was visited on one day every two weeks, when conditions were forecasted to be sunny with clear skies and temperatures above 15℃, as per the “Protocol to detect and monitor pollinator communities” (LeBuhn et al., 2016) sampling recommendations.

2.2.1 Pan trapping

I added pan trapping as a second sampling method, alongside sweep netting, to enhance the overall level of information that could be gained about the status of the pollinator community at this site in the 2016 and 2017 seasons. Pan trapping methods were informed by the “Protocol to detect and monitor pollinator communities” (LeBuhn et al., 2016). Three different coloured pan traps (blue, yellow and white) were made from 12-ounce Genpak Aristocrat solid plastic bowls. White bowls remained unpainted, while blue and yellow pans were spray painted with either Rust-oleum Ultracover Painter’s Touch in Oasis Blue or Sun Yellow, respectively. Pan trapping was conducted with one X- shaped array of 30 pan traps, 10 each of white, blue and yellow, each separated by approximately 1.5 metres (Figure 4). The order of the coloured pans in the array varied between each sampling event but was always set up such that adjacent pans were not the same colour. When sampling pollinator communities the entire array serves as the sampling unit, since different coloured bowls attract different species of pollinators. Each bowl was two thirds filled with ambient temperature tap water and one drop of clear, Nature Clean fragrance-free biodegradable dish washing soap to break the surface tension. The pan trap array was set out at approximately 09:00 and collected at approximately 16:00 on sampling days. The specimens collected in each coloured pan trap were separated from the liquid using individual strainers and placed in separated, labelled Whirl-Pak bags. The specimens were covered in 95% ethanol and placed in the refrigerator at the end of the day to minimize deterioration. 18

The set-up location of the array remained consistent throughout the 2016-17 sampling seasons. Since neither MacKay (1968-69) nor Grixti (2002-03) used pan traps at the site, the location was chosen based on accessibility and proximity to floral and nesting resources. The array was set on the lawn between the main house and the pond, bordered by various flower plantings and forest. The lawn was maintained by the property owner with frequent mowing, and the nearby garden plots containing perennials and wildflowers were minimally disturbed throughout the 2016 and 2017 sampling seasons.

Figure 4: Pan trap array composed of 30 plastic bowls (10 each of blue, white and yellow) located on the lawn between the house and man-made pond at the site.

2.2.2 Sweep netting

Sweep netting occurred on the same day as pan trapping, in four one-hour periods (10:00-11:00, 11:20-12:20, 12:40-13:40, 14:00-15:00) separated by 20-minute breaks for specimen labelling and plant identification. Netting was conducted by two people, across the entire property, following the protocol of Grixti and Packer (2006). Emphasis was placed on areas with flowering plants and nesting habitat, which changed throughout the season. Some areas were inaccessible during various times due to flooding in spring, or dense brush growth near the end of the sampling season. Bees were netted on plants, substrates, manmade objects, and in the air. Flowering plants were identified to species when possible using Plants of Southern Ontario (Dickinson & Royer, 2014) and Ontariowildflowers.com (Muma, 2012). Goldenrod (Solidago spp.) and Aster (Aster spp.) plants were only identified to genus due to time constraints in the field and the difficulty of their species level identification. Bee specimens were placed individually in vials with a label noting the date, time-period, collector and object (e.g. flower species) upon which they were collected. Specimens were stored in a shaded area until they could be brought back to the lab and placed in the freezer.

2.2.3 Comparing sampling methods between studies

Although there were some methodological differences among the three sampling periods (1968- 69 – MacKay; 2002-03 – Grixti; 2016-17 – Rubens: Table 1), data were ultimately adjusted for comparison. For example, in 1968-69 the timing of data collection was sporadic, which was difficult to replicate (MacKay, 1970). Furthermore, in 2002-03 sampling occurred once a week from 10:00 to 15:00 and only involved sweep-netting; whereas in 2016-17 sampling occurred once per fortnight for four hours (Grixti, 2004). In 2016-17, pan trapping was added to provide further information about the composition and dynamics of the bee community at the site. Some bees are more readily captured in pan traps (e.g. smaller bodied bees) and some better represented in sweep net surveys (e.g. bumblebees). I chose to employ both methods to minimize sampling biases and build the most complete representation of the overall bee community (Roulston et al., 2007; Popic et al., 2013; Rhodes et al., 2017).

In 2002-03, it was suggested that sampling occurred for five hours per week (Grixti, 2004). Further clarification with the author revealed that 30 minutes of the total sampling time per day in 2002 and 2003 was spent as a lunch break, and variable amounts of time were spent exploring the area for nesting sites and blooming flowers, and walking between these locations (Grixti, personal communication). This insight narrows the differences in sampling intensity between 2002-03 and 2016-17. Recent evidence suggests that sampling bee communities once every two weeks does not affect bee communities in the longer term, yet still provides enough information from which researchers can draw reasonable conclusions about the pollinator community (Gezon et al., 2015). The frequency and length of time spent sampling in 2016 and 2017 was deemed appropriate to gain an accurate representation of the bee community at the site, while remaining comparable to earlier sampling periods (MacKay, 1970; Grixti, 2004). In all three studies, honeybees were not collected. In 2016-17 they were present at the site in large numbers but collecting them would have taken time away from sampling wild bees.

Table 1: Differences in sampling methods and intensity between the years 1968-69 (MacKay, 1970), 2002-03 (Grixti, 2004) and 2016-17 at the site. Study year Sampling methods Sampling intensity Bee specimens collected Frequency Time

1968-69 “Intensive sweep once or twice unknown Bumblebees and netting” weekly honeybees omitted 2002-03 Sweep netting Once weekly 10:00-15:00 Queen bumblebees recorded but not collected, honeybees omitted 2016-17 Sweep netting Once biweekly Four one-hour Presence of periods from honeybees noted, 10:00-15:00 bumblebees collected and recorded Pan traps Once biweekly 9:00-16:00

2.3 Specimen processing, curation and identification

Bees collected in pan traps were sorted from other organisms in the same Whirl-pak and placed on paper towel to dry before pinning. Bees that were collected with sweep nets and then frozen, were thawed before pinning. Where needed, bees were placed in a relaxing chamber allowing for repositioning of body structures to aid identification. A label with the unique identification number, date collected, method of collection, location, GPS coordinates, name of collector and flowering plant on which the specimen was collected, was attached to each bee specimen with an appropriately sized insect pin. All specimens are housed and curated within the Raine Lab entomological collection at the University of Guelph.

Specimens collected by Grixti at this site in 2002-03 were kindly loaned to me for an extended time period by Dr. Laurence Packer, from the Packer collection at York University (PCYU; https://www.yorku.ca/bugsrus/PCYU/PackerCollection). These specimens were helpful as a reference collection when I started to identify specimens collected in 2016 and 2017. I reviewed most specimens sampled at this site in 1968-69 by visiting the Royal Ontario Museum (ROM). Existing databases for these two collections (PCYU and ROM) were used as a starting point to check for discrepancies in labelling or identification, and updates to taxonomy. A focus was placed on Lasioglossum (Dialictus), due to taxonomic updates in recent years. The number of individuals of some species recorded in the database was significantly higher than the number of physical specimens found in the museum collection. I recorded the higher number of individuals in these instances. The identification of a subset of specimens was verified for species with greater than approximately fifty individuals.

In addition to these reference collections, specimens were identified using a range of other taxonomic resources. Bees were first categorized to genus, using a combination of the Discover Life interactive key (Ascher & Pickering, 2018); The Bee Genera of Eastern Canada (Packer et al., 2007) and The Bees in Your Backyard (Wilson & Carril, 2016). Expert identification assistance for a variety of taxa was provided by Dr. Alana Pindar. Dr. Cory Sheffield verified the identification of a number of Bombus specimens, Dr. Jason Gibbs identified most Lasioglossum (Dialictus) specimens, and Dr. Thomas Onuferko identified Epeolus and Triepeolus specimens.

Keys from Mitchell (1960, 1962), and Discover Life were used in combination with a variety of other genera specific keys to identify bee specimens to species: Rehan and Sheffield (2011) was used to identify Ceratina specimens , Gibbs (2010, 2011) were used to identify a small number of Lasioglossum (Dialictus) specimens (those not directly identified by Dr. Jason Gibbs), and Romankova (2007) was used to identify Hylaeus specimens. Williams et al. (2014) was used to identify bumblebee specimens. Most bee specimens were identified to species, except for a small number that were substantially degraded during sampling, storage or processing. Nomada specimens were classified by morphological differences to morphospecies, rather than taxonomic description. This technique was also used by Grixti (2004) for two species of Andrena (Andrena species A and B).

All bees were classified by functional guild, defined as “a group of species that exploit the same class of environmental resources in a similar way” (Simberloff & Dayan, 1991). Species within the same genus do not necessarily have similar life history traits, such as nesting preference, sociality and host specialization (Hoehn et al., 2008). It has been suggested that functional diversity is a key factor in measuring diversity (Hulot et al., 2000; Lavorel & Garnier, 2002; Hoehn et al., 2008). As such, methods considering functional guilds can often provide more meaningful results since disturbances can affect each group differently (Williams et al., 2001; Grixti, 2004). Bees were classified by nesting biology into the following guilds: bumblebees, solitary ground nesters, social ground nesters, cleptoparasites and social parasites, and cavity nesters (see Table 9).

CHAPTER 3 Assessing wild bee community structure and composition changes over a forty-nine year period in a natural habitat in Caledon, Southern Ontario

3.1 Introduction

Long-term biodiversity studies are important to understand the impacts of environmental change, and to inform policy and management (Hughes et al., 2017). Only about 4.6% of global long-term studies include insects (Dornelas et al., 2018). There are few studies that have monitored long- term changes in wild bee biodiversity in North America (e.g. Marlin & LaBerge, 2001; Gardner & Spivak, 2014). The status and trends of wild pollinator diversity in Canada is a particularly large gap in current knowledge that urgently needs to be addressed (Pindar et al., 2017). This will allow us to better understand the extent and ramifications of pollinator declines. Gathering this information is useful in developing appropriate conservation strategies to ensure healthy wild bee populations and support high levels of species diversity across a range of habitats.

The limited research that currently exists documenting long-term trends in non-Bombus wild bee populations in Canada is concerning, especially considering the evidence of widespread pollinator declines around the world (Biesmeijer et al., 2006; Colla & Packer, 2008; Goulson, 2008; Grixti et al., 2009; Williams & Osbourne, 2009; Potts et al., 2010; Cameron et al., 2011; Kerr et al., 2015; Mathiasson & Rehan, 2019). Many studies amalgamate data from various research to gain a better understanding of a larger area, such as a country or continent (e.g. Cameron et al., 2011; Carvalheiro et al., 2013; Kerr et al., 2015). These studies are important to highlight general trends on a large spatial scale over the long term, however there are far fewer studies examining dynamics of pollinator communities at more specific locations. Although the results and outcomes of these local studies are less able to be generalized, they can provide much greater detail about pollinator community trends. My study addresses this research gap by focusing on the pollinator community in one specific area of southern Ontario and considering how this has changed over a forty-nine- year period.

A number of factors make long-term monitoring studies difficult to accomplish. To begin with, insect populations fluctuate from year to year, making it very difficult to fully understand the status of a species or community in an area without documenting trends over a number of consecutive years. In fact, to detect species declines of two percent per year, it could take up to 20 years of monitoring (LeBuhn et al., 2012). A 50% reduction in population could occur before a decline is even detected (LeBuhn et al., 2012). Despite requiring considerable investment of resources and time, longer-term studies are necessary to detect declines in bee populations. Furthermore, historical surveys or inventories may be lacking key information or consistent methodologies, which are important when attempting to make direct comparisons to subsequent follow-up studies. Finally, accessibility to sites can change over time potentially resulting in an inability to sample in the same locations inhibiting direct comparisons of populations over time. The attributes of a successful monitoring program are simplicity, replicability and low cost, though the most important feature is the ability to detect declines in biodiversity (LeBuhn et al., 2012).

To date comparatively few studies have tried to examine the trends in North American bees and these present somewhat variable conclusions. Indeed, Colla and colleagues suggested that most bee species recorded in North America have persisted from 1990 to 2009, with no clear evidence of larger scale extinctions in more recent history (Colla et al., 2012a). Although this study presents seemingly positive results about population persistence, other studies of North American bees suggest otherwise. For example, Cameron et al. (2011) assessed eight North American bumblebee species and reported that four of the species declined and their geographic ranges contracted. An even more recent study examined north eastern bee communities over 125 years and found composition changes with some species increasing and some declining (Mathiasson & Rehan, 2019). Most declining species were generalist solitary ground nesters, while most increasing species were generalist ground nesters (five solitary and three eusocial). Additionally, many of the species that were recorded as declining had appeared to shift their ranges in terms of elevation or latitude, suggesting effects of climate change (Kerr et al., 2015).

Studies examining the USA in general, recorded declines with a focus on bumblebees (Grixti et al., 2009; Bartomeus et al., 2013a; Koh et al., 2016). In the northeastern United States over a 140- year period, Bartomeus et al. (2013a) found that native bee species richness decreased weakly 25

overall, with significant species richness declines for bumblebees. Additionally, bee community composition seemed to be altered with significant changes in relative abundance in over half of the species examined. In Illinois, agricultural intensification was proposed as the driving factor behind the midcentury bumblebee species richness decline (Grixti et al., 2009).

Studies focusing on specific regions of the USA report a variety of trends in bee biodiversity over time. In Carlinville, Illinois there is bee community data from 1884, which has provided abundant opportunities for research on long-term changes in bee communities (Marlin & Laberge, 2001; Burkle et al., 2013). Surveys conducted at this site 75 years later found 14 new species to the area, but no evidence of 74 species which were initially recorded in the 1880s (Marlin & Laberge, 2001). A more recent study that built on these datasets assessed that 50% of bee species were extirpated over a period of 120 years, contributing to substantial degradation of the plant-pollinator network in the region (Burkle et al., 2013). In a restored tallgrass prairie habitat in Illinois, a bee community was examined over a 26-year period where bee abundance and species richness increased in the first three years, showing the potential for at least some recovery of bee diversity and community structure (Griffin et al., 2017).

Alternatively, a study examining the changes in megachilid bee communities in Itasca State Park in Minnesota found no evidence of significant declines in species richness and diversity across a 75-year period (Gardner & Spivak, 2014). Three new species to the area were found, though eleven species were recorded in 1937 and 1938 but not in the follow up survey (Gardner & Spivak, 2014). Similarly, the bee communities in Boulder county, Colorado grasslands seem to have remained mostly intact over the past century (Kearns & Oliveras, 2009). A recent study at Pinnacles National Park in California revealed that bee species richness remained consistently high over a number of extensive sampling events from 1996 to 2012 (Meiners et al., 2019). This area seems to be an important refuge area for native bees, supporting nearby agricultural lands and semi-natural areas with pollination services.

In Brazil and Costa Rica, urbanization appears to be a key factor driving population trends over the long term (Frankie et al., 2009; Martins et al., 2013; Nemesio et al., 2015). Bee species richness and abundance in a grassland site declined over a 40-year period (Martins et al., 2013). Similarly, 26

the diversity and abundance of bees found visiting a focal species of flowering tree declined significantly between 1972 and 2004 (Frankie et al., 2009). Conversely, orchid bee diversity, evenness and richness in an urban forest remnant appeared to remain stable over seven years (Nemesio et al., 2015). Another study focusing on orchid bees was conducted over seven years in Panama, where bee populations were found to be very stable over time at many of the study sites. Wet upland sites with higher bee diversity appeared to be the least stable (Roubik & Ackerman, 1987). In Britain, the evenness of wild bee and hoverfly species decreased from 1980 to 2013, with declines of over 55% in the upland species (Powney et al., 2019). This study supports the evidence of wild bee diversity declines previously found in Britain and the Netherlands (Biesmeijer et al., 2006). In contrast, there was an increase in bee species richness in Poland over a 30-year period (Banaszak & Ratynska, 2014).

Much of the evidence for pollinator declines has focused on honeybees and bumblebees (Goulson, 2008; Grixti et al., 2009; Williams & Osbourne, 2009; Potts et al., 2010; Cameron et al., 2011;), and similar patterns of research seem to be true for Ontario. Some studies suggest that responses to anthropogenic activities are drastically different between bumblebee species (Colla & Packer, 2008; Cameron et al., 2011; Colla et al., 2012a, 2012b). For example, Bombus impatiens populations seem to be increasing, and expanding their range (Colla et al., 2012a, 2012b, Goulson et al., 2008), while Bombus affinis was evaluated as critically endangered (Colla et al., 2012a, 2012b) and has not been seen in Canada since 2009 (Pindar et al., 2017). Conversely, North American and European bumblebee species have been shown to have similar responses to climate change over 110 years (Kerr et al., 2015).

Studies investigating wild bee biodiversity in specific locations and habitats are generally from the USA. Despite this interest in conducting inventories of wild bee biodiversity, these studies are rarely replicated (Meiners et al., 2019) and even fewer studies are conducted over the long term. Biodiversity is rarely evaluated on private land, though it has been demonstrated that these areas are important habitats (Hilty & Merenlender, 2003). My study contributes to the body of literature regarding the current status of wild bee biodiversity in Ontario and provides important information about the trends in community composition and structure on private property over a recent 49-year period. Comparisons of the entire bee community in 1968-69 relative to 2002-03 indicate that bee 27

species richness, evenness, diversity and abundance have all increased over a 34-year period, with many functional guilds exhibiting the same pattern (Grixti, 2004; Grixti & Packer, 2006). Ahead of collecting data from 2016-17, I predicted that the structure, composition and function of the wild bee community at the site would be maintained since 2002-03, showing a similar abundance, diversity, evenness and species richness. I expected this pattern to be consistent for the overall bee community, and for each functional guild (cavity nesters, bumblebees, solitary ground nesters, social ground nesters, cleptoparasites and social parasites). Alternatively, the structure, composition and function of the bee community at the site was hypothesized to be negatively impacted from 2002-03 to 2016-17, showing decreases in abundance, diversity, evenness and species richness in this time period, both overall and for each bee functional guild. I expected these decreases because there was evidence of habitat loss and climate change effects on the pollinator community in 2002-03 (Grixti, 2004). Both issues are becoming increasingly important globally.

3.2 Analytical methods

Wild bee biodiversity was estimated with four measures: abundance, species richness, diversity (H’) and evenness (J). These measures were calculated for the overall population and each functional guild in all sampling years (1968, 1969, 2002, 2003, 2016, 2017) and each study (1968- 69, 2002-03, 2016-17). An abundance estimate is simply the number of individuals collected over a given time, while estimates of species richness, diversity and evenness required further calculations (Gotelli & Ellison, 2012). Species richness, diversity and evenness were calculated for each year of the three studies using the program Species Diversity and Richness IV (Seaby & Henderson, 2006). These measures were recalculated for the 1968-69 and 2002-03 studies. Only samples obtained by sweep netting (the sampling method in common between the three study

periods) were compared over time. Specimens which were not identified to species (e.g. Nomada, Andrena species A and B) were not included in these calculations.

The Shannon Diversity Index (H’) (Equation 1) was used to calculate diversity estimates, facilitating comparisons to those calculated with data from 2002-03 by Grixti (2004). This widely- used index includes both the species richness and evenness to calculate a diversity estimate.

푠 ′ 퐻 = − ∑ 푝푖 log(푝푖) (1) 푖=1

The variable 푝푖 is the proportion of the abundance represented by species i. Generally, H’ values range between 1.5 and 3.5, with a higher value representing a higher diversity (Magurran, 1988; Gotelli & Ellison, 2012).

Evenness (J) (Equation 2) is a measure of the distribution of individuals within each species.

퐻′ 퐽 = (2) ln(푆)

The variable 푆 is the number of species in a given sample. Evenness values range from 0 to 1, with a value of 1 representing a community where all species are present in equal abundances (Heip et al., 1998).

Rarefaction curves are a more useful means of estimating and comparing true species richness than a simple count of the number of species in an area over a given time, for two main reasons. Firstly, species richness counts typically underestimate the actual number of species present, since rare species are difficult to detect (Colwell et al., 2012). Secondly, the number of individuals collected is positively correlated with the number of species collected (Colwell et al., 2012). Abundance varied greatly across sampling years at the site, so generating rarefaction curves was necessary to enable appropriate comparisons of species richness over time. The data collected were rarefied (randomized and repeatedly sampled without replacement) using the program EstimateS 9.1.0 (Colwell, 2013), and an individual-based rarefaction curve was generated. Extrapolation through 29

this method allows for comparison of smaller numbers to larger numbers of samples, without losing valuable data (Colwell et al., 2012). Rarefaction curves can also provide information about sampling intensity. Ideally, the resulting rarefaction curve would reach an asymptote, demonstrating that sampling effort was appropriate (Chiarucci et al., 2008). At such a point, species richness would not increase if sampling effort was intensified. Six rarefaction curves (one for each sampling year) were plotted together on one graph (Figure 11) using OriginPro, student version (OriginLab, 2019) allowing for the comparison of species richness over time, with varying sample sizes.

To visualize species composition across the three studies (temporal sampling events), I used the internet application BioVenn (Hulsen et al., 2008). Venn diagrams of total species richness of each sampling year pair were created, with additional diagrams for each functional guild. Jaccard’s similarity index (Equation 2) was calculated in the program Community Analysis Program V (Henderson & Seaby, 2014) to compare the species composition between 1968-69, 2002-03 and 2016-17. Jaccard’s similarity index is calculated with the following equation:

|퐴 ∩ 퐵| 퐽푎푐푐푎푟푑 퐼푛푑푒푥 = × 100% Equation (4) |퐴 ∪ 퐵|

In this instance, the variable A represents the species in one set, and B represents the species in the other set. The Jaccard similarity index indicates the percentage of similarity between the members of two or more datasets. 100% similarity indicates that the datasets include identical members, whereas datasets with 0% similarity would have no members in common.

Venn diagram comparisons between sampling year pairs, and calculations of Jaccard’s similarity index did not include the genera Bombus and Nomada, as Nomada were not identified to the same level, and bumblebees were not collected to the same extent in all three sampling periods. Andrena species A and B collected in 2002 and 2003 were not included in this analysis to facilitate comparisons.

3.3 Results

A total of 2853 bee specimens were collected using pan trapping and sweep netting in 2016 and 2017 field seasons respectively, representing 5 families, 27 genera, and 171 species (Tables 2 and 9). In the 2002-03 field seasons, a total of 10351 specimens were collected using sweep netting by Grixti and Packer (2006) representing 6 families, 27 genera and 155 species. In the 1968-69 field seasons, a total of 8201 specimens were collected using sweep netting by MacKay and Knerer (1979), representing 5 families, 24 genera and 98 species. In 2016-17, 47 species were singletons; 27 species were singletons in both 2002-03 and 1968-69.

The Jaccard similarity index revealed a 42.86% similarity between the wild bee species in 1968- 69 and 2002-03. The species in the 2016-17 community and 2002-03 community were 42.57% similar, followed by a 34.43 % similarity between 2016-17 and 1968-69. The Jaccard similarity indices were calculated using only sweep net samples to facilitate comparisons between the three studies. There were 56 species in common between the three studies. Differences in the species composition of the three studies was further analysed by dividing the species into functional guilds. Specimens which could not yet be identified to species were excluded from Venn diagram comparisons (Figures 5, 6, 7, 8). Twelve cavity nesting species were found throughout all three year-pairs, with 2016-17 species richness most similar to 2002-03 (20 species shared, 57.1% of 2016-17 species richness: Figure 5). Six cleptoparasitic and social parasite species were collected in all three studies. Similarly, in terms of species composition, 2016-17 are most similar to 2002- 03, with 11 species in common (73.3% of 2016-17 species richness: Figure 6). Bee species composition overlapped the most in the social ground nester guild, with nearly identical species found across all three studies (Figure 7). There were 17 social ground nesting species in common between all three year-pairs, with 20 species found in 2016-17, and 2002-03 (100% of 2016-17 species richness). Twenty-one solitary ground nesting species were found throughout the three studies, with species composition in 2016-17 most similar to 2002-03 (38 species in common, 46.3% of 2016-17 species richness: Figure 8).

In addition to these changes in species composition, there are significant differences in the percentage of individuals belonging to each functional guild across time at the site (Figure 9). 31

Bumblebees were excluded from this analysis due to inconsistent sampling methods for this genus across these three studies. The bee communities in both 1968-69 were dominated by social ground nesters (making up 85.6% and 88.8% of the individuals sampled in each year, respectively). After 1968 and 1969, the functional guild evenness increases appreciably, though the cleptoparasites and social parasite functional guild remained the smallest proportion throughout all six sampling years (ranging from a minimum 0.53% in 1968, to a maximum of 5.1% in 2003). The percentage of cavity nesters shows a steady increase over time from 1.4% in 1969 to 42.9% in 2017. The relative abundance of solitary ground nesters also increased over time, from an average of 8.8% in 1968 and 1969 to an average of 43% across the years 2002, 2003, 2016 and 2017.

Table 2: The number of wild bee families, genera and species found at the site in each year (1968, 1969, 2002, 2003, 2016 and 2017). Sampling Year Families Genera Species 1968 5 24 90 1969 5 16 51 1968 and 1969 5 24 98 2002 5 24 101 2003 6 26 146 2002 and 2003 6 27 155 2016 5 23 123

2017 5 24 122 2016 and 2017 5 27 171

the species found in common between the three studies. The genus Nomada is not included in this figure. See Table 9 for the specific species collected in each year.

Figure 7:Venn diagram of the social ground nesting collected in each pair of sampling years (1968-69: red circle; 2002-03: yellow circle; 2016-17: blue circle) at the site. Overlapping portions of each circle represent the total numbers of species in common between each pair of studies, with the middle section representing the species found in common between the three studies. See Table 9 for the specific species collected in each year.

Figure 8:Venn diagram of the solitary ground nesting collected in each pair of sampling years (1968-69: red circle; 2002-03: yellow circle; 2016-17: blue circle) at the site. Overlapping portions of each circle represent the total numbers of species in common between each pair of studies, with the middle section representing the species found in common between the three studies. Andrena species A and B are not included in this figure. See Table 9 for the specific species collected in each year.

Figure 9:The percentage of wild bee individuals collected by sweep net at the site per study year, belonging to each bee functional guild (excluding bumblebees).

Diversity, evenness, abundance and species richness were calculated for the complete bee community in 2016-17 (Table 3), for both pan trap and sweep net samples. Examining the years separately and combined, the estimates of diversity, abundance and species richness were higher for the sweep net samples, while pan trap samples showed higher evenness estimates. Examining the samples collected from both active and passive sampling methods in combination, the diversity, species richness and evenness estimates were higher in 2016, while the abundance estimate was higher in 2017.

Estimates of diversity, evenness, abundance and species richness of sweep net samples were also calculated for each sampling year and year-pairing for the two historical datasets, reported with 2016-17 values in Table 4. For individual collecting years the estimates of diversity, abundance, 35

and species richness were highest in 2003, and the evenness estimate was highest in 2016. For consecutive year-pairs, diversity and evenness estimates were highest in 2016-17, while the 2002- 03 year-pair showed the highest abundance and species richness estimates (Table 4).

Further categorizing the sweep net specimens into functional guilds (solitary ground nesters, social ground nesters, cavity nesters and cleptoparasites and social parasites) provides greater insights over time (Table 5). Social ground nester diversity and evenness estimates were highest in the 2016-17 year-pair, while the abundance and species richness estimates were highest in 1968-69 (Table 5). Estimates of solitary ground nester diversity, evenness and species richness were highest in 2016-17, while the abundance estimate was highest in 2002-03 (Table 5). Cleptoparasite and social parasite diversity, abundance and richness estimates were highest in 2002-03, while the evenness estimate was highest in 2016-17. Cavity nester diversity and species richness estimates were highest in 2016-17, while the evenness and abundance estimates were highest in 2002-03 (Table 5).

Rarefaction curves accurately show how species richness estimates have changed across sampling years at the site (Figure 10). Using the method of extrapolation discussed earlier in this chapter, an estimate of the true species richness was obtained. Since this method accounts for abundance, the estimates of species richness obtained by rarefaction can be compared across the six sampling years. Extrapolation of the 2016 and 2017 curves revealed a higher estimate of the true species richness when compared to extrapolated species richness curves of previous sampling years (1968, 1969, 2002, 2003). The shape of the 2017 and 1969 rarefaction curves closely approaches an asymptote, demonstrating that the number of bee species present at the site in these years is accurately represented by the sampling. The slopes of the remaining curves (2016, 1968, 2002 and 2003) at 6000 individuals are steeper in comparison, demonstrating that the number of individuals sampled does not fully represent the true bee species richness at the site (Figure 10).

Table 3: Diversity, evenness, abundance, and richness of the complete bee community collected by both pan trap and sweep net in the years 2016 and 2017 at the site, calculated for each year separately and combined. 2016 2017 2016 & 2017 combined Pan Net Total Pan Net Total Pan Net Total Diversity (H’) 3.49 3.50 3.81 3.43 3.51 3.71 3.69 3.70 3.93 Evenness (J) 0.81 0.78 0.79 0.82 0.75 0.77 0.80 0.74 0.76 Abundance 419 927 1346 370 1137 1507 789 2064 2853 Richness 74 90 123 65 108 122 104 145 171

1968 1.74 0.39 5102 90 1969 1.63 0.41 3058 51 1968 & 1969 1.74 0.38 8160 98 2002 3.24 0.70 4208 101 2003 3.78 0.76 6143 146 2002 & 2003 3.65 0.72 10351 155 2016 3.46 0.79 626 81 2017 3.49 0.76 746 101 2016 & 2017 3.70 0.75 1372 135

Diversity (H’) Evenness (J) Abundance Richness Social ground nesters 1968 & 1969 1.10 0.35 7096 22 2002 & 2003 2.16 0.71 3163 21 2016 & 2017 2.27 0.79 303 18 Solitary ground nesters 1968 & 1969 2.69 0.71 723 45 2002 & 2003 2.78 0.66 4604 69 2016 & 2017 3.23 0.76 472 71 Cleptoparasites and social parasites 1968 & 1969 2.02 0.73 66 16 2002 & 2003 2.90 0.80 465 37 2016 & 2017 2.45 0.93 29 14 Cavity nesters 1968 & 1969 1.28 0.47 307 15 2002 & 2003 2.16 0.65 2119 28 2016 & 2017 2.21 0.64 592 32

3.4 Discussion

The estimates of diversity, evenness, abundance and species richness of the 2016 and 2017 bee communities were compared with historical datasets from 1968, 1969, 2002 and 2003 (Table 5). Diversity and evenness estimates increased over the three studies (1968-69, 2002-03 and 2016- 17). Abundance estimates increased from 1968-69 to 2002-03 but decreased dramatically in 2016- 17. Similarly, there was no clear trend in species richness estimates, as the number of species increased from 1968-69 to 2002-03 but decreased slightly to 2016-17. Further analysis involving rarefaction curves, revealed that estimates of true species richness are highest for 2016-17, followed by 2002-03 and 1968-69, respectively (Figure 10). This is the most useful method to compare the species richness of the bee community over six years. Therefore, there is evidence to suggest that estimates of true species richness of wild bees are increasing over time at the site. In 39

a long-term study of a wild bee community in Illinois, about 50% of bee species were thought to be extirpated from the study site from the late 1800s to 2010 (Burkle et al., 2013). A similar pattern of species turnover was evident in this study, although over a much shorter period over time. The species that compose the wild bee communities at three different time points (1968-69, 2002-03 and 2016-17) were most similar between 1968-69 and 2002-03 (42.86% similarity), followed closely by the communities in 2002-03 and 2016-17 (42.57% similarity). The communities in 1969-69 and 2016-17 were the least similar (34.43%). Overall, the wild bee diversity, evenness and richness estimates increased over time, while abundance estimates decreased, presenting very interesting and seemingly contradictory outcomes. These results suggest that the evenness, richness and diversity can potentially be maintained on a small spatial scale, even if there is a dramatic loss in the number of individuals. Though the pollination services provided by bees may be dramatically reduced in this area (due to declines in abundance), the diversity of pollinators does not seem to be declining concomitantly.

Functional guild trends

Separate analysis of the estimates of abundance, evenness, diversity and richness for each wild bee functional guild shows varied trends. There is evidence to suggest that guilds are differently affected by disturbance (Cane et al., 2006; Williams et al., 2010; Wray et al., 2014; Roberts et al., 2017). Delving further into these functional guilds, certain species within are more abundant than others. These “core species” are likely the drivers of trends in the community, such as guild proportions (Hanski, 1982). On the other hand, “satellite species” are those which are uncommon or rare and are less likely to be responsible for significant trends, though they may still provide some interesting insight into the overall bee community (Hanski, 1982).

Four functional guilds were examined in this study: social ground nesters, solitary ground nesters, cleptoparasites and social parasites, and cavity nesters. Out of each of these guilds, social ground nesters showed the lowest amount of species turnover across the three studies, with the vast majority of species collected in at least one year of each study (Figure 7). Social ground nesters showed decreases in estimates of both abundance and species richness, with increases in estimates of diversity and evenness. This decrease in abundance aligns with overall trends of the entire wild 40

bee community. Reduced richness could have resulted from a decreased number of social ground nester specimens collected in each study over time. However, the richness did not decrease proportionally to the abundance, expressed as an increase in diversity and evenness. The significant decrease in the proportion of social ground nesters between 1969 and 2002 correlates with some habitat changes at the site, identified later in this chapter. Four main social ground nester species appeared to drive trends within this functional guild. In 1968-69, there were thousands of Lasioglossum imitatum individuals, which were observed much less frequently in consequent studies. Across all three studies Lasioglossum lineatulum decreased steadily (Table 10). From 2002-03 to 2016-17, there was a large decrease in both Halictus ligatus and Lasioglossum cinctipes (Table 10). These species are all common and widespread in southern Ontario, so the trends in these species at the site are likely driven by local factors.

Interestingly, the relative abundance of solitary ground nesters dramatically increased between 1969 and 2002. This suggests that the driving factor behind these changes in both the social and solitary ground nester guilds may not be a result of limited nesting sites, since both guilds nest in the ground. This switch in the percentage of social ground nesters and solitary ground nesters between the first two studies may suggest that solitary ground nesters filled a niche previously occupied by social ground nesters. Other studies have found that social bees are generally more sensitive to disturbance than solitary bees (Klein et al., 2003a, 2003b; Ricketts et al., 2008; Winfree et al., 2009). The social ground nesters at the site could possibly have been more affected by a disturbance, such as climate or habitat change (further explored in Chapter 5), a pattern observed in tropical forest ecosystems (Aizen & Feinsinger, 1994; Klein et al., 2003a, 2003b). In temperate grasslands, solitary bees have been found to be more sensitive to disturbance (Steffan-Dewenter et al., 2002). The link between sociality and response to land use change appears to differ between habitats, and the biology of bees native to the region (Ricketts et al., 2008). Social bees appeared to be more sensitive to disturbance in tropical regions because many are stingless bees, which forage in relatively small areas and nest in tree cavities (normally more plentiful in a natural habitat) (Ricketts et al., 2008). In temperate areas, many of the social bees are bumblebees which have a comparatively large foraging range and are able to find nesting sites in more intensified landscapes (Ricketts et al., 2008).

Solitary ground nesters show an increase in diversity and richness estimates over time. Additionally, a number of specific species show peaks in abundance in 2002-03. Among these are Colletes inaequalis, Andrena fragilis, Andrena integra, Andrena wilkella and Lasioglossum zonulum. Interestingly, no bees in the genus Melissodes were collected in 1968-69, but a variety of species were found in the two latter studies. A peak in abundance in 2002-03 was also observed in the species Melissodes druriellus and Melissodes illatus. Some other notable changes in the solitary ground nester guild include the presence of particular species in early studies with no evidence in later studies, or the apparent extirpation of these species at the site. Calliopsis andreniformis and Perdita octomaculata were abundant in large numbers in 1968-69, with no records in the following two studies. Similarly, Pseudopanurgus rudbeckiae was observed by MacKay in 1968-69 and Grixti in 2002-03, but not in 2016-17. Additionally, the only observations of Macropis nuda were in 2003.

A steady increase in the relative abundance of cavity nesters collected in each sampling year was observed since 1969. Cavity nester diversity and richness have also increased over time. Ceratina calcarata and Ceratina dupla have greatly increased substantially in abundance from 1968-69 to 2002-03. Conversely, fewer individuals belonging to the genus Hylaeus were collected in 2016- 17 compared to the previous two studies. Ontario’s only species of Xylocopa (X. virginica) was not found by both MacKay or Grixti but was recorded at the site in 2016-17 (only one individual each year). A potential explanation is that this bee resembles members of the genus Bombus, and therefore may have been avoided. As previously mentioned, MacKay and Grixti did not sample bumblebees consistently, and therefore may have overlooked this species. Another reason for this new observation at the site is that this species is at its northern range in Ontario (Skandalis et al., 2011). It is possible that the range has extended further over time to include the site a result of variation in precipitation and temperature due to climate change (Skandalis et al., 2011).

At the site, the cleptoparasite and social parasite guild represented the smallest percentage of bees in each sampling year, aligning with the trends generally seen in bee communities (Wcislo, 1981). The proportion of cleptoparasites and social parasites increased from 1968-69 to 2002-03, then stayed at a similar level in 2016-17, and showed a substantial amount of species turnover over time. As the apex of the bee community, cleptoparasite diversity and relative abundance could 42

indicate the status of the total bee community (Sheffield et al., 2013). The small increase in relative abundance of this guild provides evidence that the overall bee community at the site could have become slightly more robust, or at least maintained its health over time. Other trends in the cleptoparasite and social parasite functional guild include an increase in evenness over time. This could potentially be linked with the evenness of hosts in the social ground nester and cavity nester guild that also increased over time. Cleptoparasite and social parasite diversity and richness are highest in 2002-03, which could be influenced by the high number of specimens collected in 2002- 03.

One genus of cleptoparasite found at the site is Epeolus, which was collected in low numbers throughout all three studies. The suspected hosts of the Epeolus species found at the site are Colletes americanus, C. compactus, C. kincaidii, and C. simulans (T. Onuferko, personal communication, January 22, 2018). Three of these species were observed in each study, while C. americanus, which was not collected in 2016 or 2017. Triepeolus pectoralis is a cleptoparasite of Melissodes druriellus, both were observed in 2002-03 and 2016-17 (T. Onuferko, personal communication, January 22, 2018). Stelis lateralis was only seen in low numbers in 2003, and is a parasitic species of Hoplitis pilosifrons, Hoplitis producta and Hoplitis simplex (Onuferko et al., 2015). All three host species were observed throughout the three studies (except H. producta in 2016-17), demonstrating that their parasite either evaded collection in some years, or was not collected for another reason other than host availability. Calliopsis andreniformis was recorded in 1968 but was not observed at the site in the following years. This species’ parasite, Holcopasites calliopsidis was also seen in 1968 and 1969, but unrecorded after this time (Mitchell, 1962). It is not known why these species were not collected at the site, but it was not likely related to the availability of floral resources, as C. andreniformis is a generalist species (Mitchell, 1960). Another parasitic genus, Coelioxys was collected throughout all three studies at the site. Megachile centuncularis and M. campanulae are hosts of Coelioxys modesta (Mitchell, 1962). C. modesta was only observed in 2003, though M. campanulae was present in 2002 and 2003, and M. centuncularis was present in 2017. It is possible that C. modesta may not have been able to transition hosts over time at the site. Coelioxys sayi is a parasite of Megachile mendica and both species were observed simultaneously over time (Mitchell, 1962). Coelioxys rufitarsus parasitizes

Megachile latimanus and Megachile melanophoea, which were also observed simultaneously (Mitchell, 1962).

Although the genus Bombus was not included in the analyses over time, it should be noted that Bombus terricola and Bombus affinis were recorded at the site in 1968-69. B. terricola is currently listed as vulnerable and B. affinis is critically endangered (Hatfield et al., 2015a, 2015b). There are no bumblebee data for 2002-03, so it is unknown if these important species were present during these years. Bombus terricola was once again recorded in 2016-17 at the site, which is of particular significance as it is a species of special concern in Ontario (Colla & Packer, 2008). The last sighting of Bombus affinis in Ontario was in 2009 (Colla & Taylor, 2011), so this species was not expected to be seen at the site in 2016-17.

In this study, a combination of pan trapping and sweep netting was used to sample the wild bee community at the site. Sweep nets were used exclusively in 1968-69 and 2002-03. Although sampling methods varied between the three studies, only sweep net samples were used when comparing data across the three sampling periods (1986-69, 2002-03, 2016-17). Pan trapping, a passive sampling method, was added in 2016-17 to enable the collection of additional information about the bee community while simultaneously collecting data with sweep nets.

Both sweep netting and pan trapping sampling methods have their biases which may have had an impact on the bee species collected in each study year (Roulston et al., 2007; Popic et al., 2013; Rhodes et al., 2017). Pan traps specifically target flower visiting insects (including many pollinators), reducing the amount of bycatch, and also avoiding collector bias. They are simple and cost effective to construct, and are easy to set up, requiring no attention until pickup. Pan traps typically catch relatively few large-bodied bees, such as bumblebees, since they are more easily able to crawl out of the trap. Cleptoparasites are also fairly under-represented in pan traps, likely because they spend more time searching for their host nests rather than flowers (Marlin & LaBerge, 2001). In contrast, Lasioglossum (Dialictus) seem to be over-represented in pan trap specimens (Onuferko, 2014). Considering the number of Lasioglossum (Dialictus) individuals, this could be a factor affecting samples collected at the site. It has also been suggested that pan traps could be less attractive to bees foraging in tall plants, since they are most often placed on the ground and 44

may be overlooked by bees amongst higher vegetation (Cane et al., 2000). This was not likely a concern at the site, since the pan traps were placed on an open lawn, that was mowed frequently throughout both the 2016 and 2017 sampling seasons.

As with any sampling method, there are some biases to consider with sweep netting. The number of bee samples collected in a given time can vary from person-to-person depending on their sweep netting experience, skill and motivation. Initially, it can be difficult for a beginner to determine if the target insect is a bee or other flying insect, which can affect the number of bees collected within a sampling period. This difference in experience and skill level will affect long term studies such as this one, since there are many different samplers over time. Additionally, smaller bodied bees are often more difficult to detect, and therefore less likely to be collected by sweep net than larger bodied bees.

Samples from both sweep netting and pan traps have been shown to well-represent a bee community, with a correlation between the abundance of bees collected using each method (Richards et al., 2011). This study does not align with Richards et al. (2011) findings, showing differences estimates of abundance, evenness, richness and diversity in each method. Pan trap and sweep net samples showed a similar diversity estimate. However, the evenness estimate was greater for pan trap samples, supporting the idea that pan traps can be subject to less collector bias than sweep netting. Sweep netting resulted in much higher estimate of abundance, which was expected since samplers were actively searching for bees. Sweep net samples also had a higher richness estimate This result was expected considering that the species richness is related to the number of individuals. It has been suggested that social bees may be more readily collected than solitary bees, since they nest in aggregations and exist in higher numbers (Goldstein & Scott, 2015). This was true for the samples in 1968 and 1969 but did not hold true for the subsequent studies in which greater numbers of solitary bees were collected.

As outlined in Table 1, the sampling frequency was variable between the three studies. In 1968- 69 sampling was sporadic, while in 2002-03 sampling occurred once per week. In 2016-17, sampling was conducted once every other week, while increasing the number of sweep netting hours per sampling day to make the sampling intensity comparable to 2002-03. Sampling 45

frequency was altered for a number of reasons and the potential implications of this are investigated further in this chapter. Another factor that may have impacted the representation of bee communities at this site over time are differences in the sampling start dates. Sampling began at the start of May in 2016-17, which was later than the previous two studies. Because of this shorter sampling season, it is possible that the presence of early season bees is not accurately documented in the 2016-17 data.

Although taxon concepts are fluid and change over time (Packer et al., 2018), it is unlikely that these changes significantly contributed to the patterns seen in bee community composition and structure, and their interactions with flowering plants. In terms of plant identification, the taxonomy of common Ontario wildflowers is well established and has not changed drastically since the start of this study in 1968 (except in terms of naming). In cases where plants were not identified to the same level, the highest level of classification was used to facilitate comparisons across years. Conversely, wild bee taxonomy has changed dramatically over the years, especially in select groups like Lasioglossum (Dialictus). Using fully updated taxonomic information, the 1968-69 and 2002-03 bee collections were re-examined to revise species identifications, identify missed specimens and correct mistakes.

Using data from three time points over three pairs of sampling years (6 years of sampling in total), the presence and extent of changes in the bee community at the site over 49 years was examined. Certainly, the results of this study would benefit from a greater number of sampling years more evenly spread across the full period. However, adopting the paired-year approach gives some indication of year-to-year variation in the bee community at each of these discrete sampling times across this almost five-decade long timeline. Continuing this study would be beneficial to determine whether the changing patterns in bee community composition and structure described in this study hold true over the longer-term.

CHAPTER 4 Changes in plant-pollinator interactions over a 15- year period in Caledon, Southern Ontario

4.1 Introduction

Plant-pollinator interactions refer to the strong mutualistic relationships between flowering plants and pollinators. Pollinators ensure the reproduction of the majority of wild flowering plants (Ollerton et al., 2011), which support general biodiversity in most terrestrial ecosystems (Loreau et al., 2001). Plant-pollinator interactions are often depicted in a network structure, due to their interconnected relationships. Pollinator networks face many anthropogenic disturbances and 47

functional redundancy can provide some resilience to these networks. Functional redundancy in plant-pollinator community networks means that some species and interactions can be lost before ecosystem function is significantly impaired (Burkle et al., 2013). Functional redundancy in a plant-pollinator network is directly related to the diversity of pollinators and flowering plants in a community. If there are multiple generalist species performing similar roles, this can diminish the negative effects if any one species becomes locally extirpated or extinct. Maintaining high levels of biodiversity is important for such networks to function (Bartomeus et al., 2013b). With additional pressures (such as climate change, pests and pathogens, land use intensification and invasive species) on the network, interactions between plants and pollinators can be weakened, which can disrupt overall ecosystem function (Burkle et al., 2013).

Plant-pollinator networks have been found to be quite resilient to disturbance. However, these networks are only resilient up to a point, with further species losses creating a snowballing effect (Burkle et al., 2013). The very feature that makes these networks resilient, can also put both plants and pollinators at risk. Highly connected species can protect the network from further loss by fulfilling similar roles. However, the removal of a highly connected species from the network has the most serious consequences. Bumblebees and some solitary bees are the most highly connected members of plant-pollinator networks and therefore are the most important to conserve (Memmott et al., 2004). These highly connected species are generalists, which forage on many different floral resources.

Loss of interactions between plants and pollinators can also occur through phenological mismatches due to changing patterns of land-use or climate change. There is some competing evidence from the USA and Europe indicating that climate change is impacting the synchronicity of plant-pollinator interactions (Memmott et al., 2010; Bartomeus et al., 2011; Forrest & Thomson, 2011; Bartomeus et al., 2013a; Burkle et al., 2013; Petanidou et al., 2014; Polce et al., 2014). These mismatches are detrimental to solitary bees when they emerge before plants, causing reductions in survival, activity, the number of brood cells and nests (Schenk et al., 2017). The effects of these mismatches between food source and insect could contribute to pollinator declines, and consequently influence pollination services in these locations (Potts et al., 2010; Schenk et al., 2017). 48

The concept of ‘nestedness’ can be used as a measure of resilience of a plant-pollinator network to disturbance (Burgos et al., 2007). A network is described as ‘nested’ when generalist plants interact with specialist pollinators, and specialist plants interact with generalist pollinators (Memmott, 2004). This feature of a network protects species from extinction by having another species that performs the same role in the ecosystem (pollinates a specific plant or provides resources for a specific pollinator). Alternatively, a compartmentalized structure exists when generalists interact with generalists, and specialists interact with specialists (i.e. if a single pollinator species interacts with one species of plant, extinction of either would ensure the loss of both species).

Here I examine the structure of interactions between flower visiting bees and wild flowering plants using data collected by Grixti in 2002-03 and data collected in 2016-17. Comparing the structure, topology and nestedness of these interaction networks allows me to examine how they have changed over this 15-year period (2002-2017). One prediction is that interactions between flowering plants and wild bees at the site would be maintained, with the strength and number of interactions in the corresponding plant-pollinator interaction network remaining similar over time. Additionally, the structure of the network would remain ‘nested’ (see Figure 12), a measure of resiliency to disturbance (Bascompte et al., 2003; Ollerton et al., 2003; Thompson, 2005; Guimarães et al., 2006; Lewinsohn et al., 2006; Santamaría & Rodríguez-Gironés, 2007). Alternatively, the plant-pollinator network interactions could be negatively impacted over the past 15 years, with fewer and weaker interactions between flowering plants and wild bees. If so, I would expect networks to show a more compartmentalized structure over time, corresponding with a reduction in overall resiliency to disturbance.

4.2 Analytical methods

To consider the mutualistic relationship between flowering plants and bees, it was necessary to evaluate how these interactions have changed over time at the site. Both the 2002-03 and 2016-17 datasets include information regarding the flowering plant upon which each bee specimen was collected, whereas these interaction data were not available for collections made in 1968-69. 49

Interactions between bees and flowering plants were defined as flower visits and did not measure pollination. Previous studies did not include a full inventory of plant species, so changes in vegetation at the site were not able to be compared over time. Data were entered into Food Web Designer 3.0 (Sint & Traugott, 2015), generating five plant-pollinator interaction networks. Complete networks for 2002-03 and 2016-17, and a network for 2016-17 without bumblebee species were constructed. Additionally, two networks including only the plant species in common between both year-pairs were constructed.

Along with simple calculations of the number of interactions lost, gained and maintained over time, nestedness was evaluated and compared for each network. In an interaction network with perfect nestedness, specialist pollinators which focus on specific flowering plants, interact with only a subset of the plant species visited by more generalist pollinator species (Bascompte et al., 2003). Examples of this matrix quality are depicted in Figure 11. There is abundant evidence suggesting that species interaction networks with high levels of nestedness are more resilient to disturbances (Ollerton et al., 2003; Thompson, 2005; Guimarães et al., 2006; Lewinsohn et al., 2006; Santamaría & Rodríguez-Gironés, 2007). Nestedness was evaluated with the Nestedness for Dummies (NeD) (Strona et al., 2014a, 2014b) software. NODF (Nestedness metric based on Overlap and Decreasing Fill) was used as the nestedness metric. This metric was proposed by Almeida-Neto et al. (2008) to more accurately identify nested matrices, compared to previously used metrics. NODF is “the percentage of presences in inferior rows and in right columns that are in the same position (column or row) of the presences in, respectively, upper rows and left columns with higher marginal totals for all pairs of columns and of rows” (Almeida-Neto et al., 2008). The NODF value can range from zero (non-nested structure) to 100 (perfectly nested).

Figure 11: Matrices illustrating varying degrees of nestedness in plant-pollinator communities adapted from Almeida- Neto et al. (2008). It is important to note that these matrices do not show the frequency of interactions, but rather the presence or absence of a certain interaction between two species. For the purposes of my thesis, the flowering plant species and the bee species would be represented by the x and y axis of each matrix. Each box represents an interaction between one species of plant, and one species of bee. The zeros and ones in each box represent the presence or absence of each interaction. The top four matrices show some level of nestedness among the rows and columns. A nested matrix would result from a network where generalist plants interact with specialist pollinators, and specialist plants interact with generalist pollinators. The bottom four matrices show no nestedness among rows and columns.

The NeD program was run, randomizing the data 50 times, with each of the available null models, resulting in NODF, Z-value, relative nestedness and the nestedness probability level outputs. The dataset was randomized 50 times based on the standard number of runs in the NeD program and a study using a similar procedure (Bascompte et al., 2003). Nestedness was evaluated with a presence-absence matrix, so the strength of interactions between species was not considered in this calculation. The Z value (Equation 5) is measured in standard deviations of the degree of nestedness in the focal matrix from the mean of the distribution of the null matrix (Kaiser, 2015). This value determines whether the degree of nestedness in the focal matrix is significant compared to the nestedness of the null matrix.

푁푂퐷퐹 − 푚푒푎푛(푁푂퐷퐹 ) 푍 = 푚푎푡푟푖푥 푛푢푙푙 (5) 푠푡푑(푁푂퐷퐹푛푢푙푙)

NODFmatrix is the NODF value for the matrix under evaluation. NODFnull is the set of NODF values calculated for the null model and std(NODFnull) is their standard deviation. When using NODF, significance at p < 0.05 occurs when the Z values exceed 1.64. These NODF variables are also used when calculating relative nestedness (RN) (Equation 6) of the matrix under evaluation to the null model matrix.

푁푂퐷퐹 − 푚푒푎푛(푁푂퐷퐹 ) 푅푁 = 푚푎푡푟푖푥 푛푢푙푙 (6) 푚푒푎푛(푁푂퐷퐹푛푢푙푙)

High relative nestedness values corresponds to a matrix that is more nested than the null model matrix. A relative nestedness value of zero or below means that the focal matrix is not as nested as the null model matrix, whereas a value of one corresponds to complete nestedness (Bascompte et al., 2003; Nielsen & Bascompte, 2007).

Considering the lack of consensus around choosing null models for nestedness evaluation, so multiple null models available in the NeD program were used. The EE model (equiprobable row totals, equiprobable column totals) varies the number of species occurrences in rows and columns, while maintaining the total number in the overall matrix. The CE model (proportional row totals, proportional column totals) uses the row and column total proportions to assign an occupational probability (Equation 7) to each cell in the matrix.

푇표푡푅 푇표푡퐶 ( 푖 + 푗) 푃푟표푏푎푏푖푙푖푡푦 표푓 푐푒푙푙 표푐푐푢푝푎푡푖표푛 = 퐶 푅 (7) 2

TotRi is the number of presences in the row i, TotCj is the number of presences in the column j, C is the number of matrix columns and R is the number of matrix rows. The FE model (fixed row totals, equiprobable column totals), maintains the species occurrence frequencies (row totals),

while randomly varying the column totals. The EF (equiprobably row totals, fixed column totals) model functions oppositely, with the column totals maintained and the row totals varied.

4.3 Results

The plant-pollinator network included 1616 interactions between 61 flowering plants and five functional groups of wild bees at the site in 2016-17 (Figure 12). The highest number of interactions occurred with the cavity nester functional group. The three plants involved in the most interactions were Sorbaria sorbarifolia (False spirea), Taraxacum (Dandelion) and Cirsium vulgare (Bull thistle). The top three strongest interactions were between cavity nesters and Leucanthemum vulgare (Oxeye daisy), social ground nesters and Sorbaria sorbarifolia (False spirea), and solitary ground nesters and Taraxacum (Dandelion) (Figure 12).

Figures 13 and 14 show the interactions recorded for 2016-17 and 2002-03 without bumblebees, to facilitate comparison between studies. The remaining number of interactions without bumblebees in 2016-17 was 1312 (indicating that 18.8% of plant interactions in 2016-17 involved bumblebees), while the number of interactions in 2002-03 was 8580. These figures show a greater than 6-fold decrease in the number of interactions between plants and pollinators over this time period, while the frequency of specific interactions increases. In 2002 and 2003, 75 plant groups were recorded in interaction with four functional groups of bees. The three most frequently recorded interactions are those between Securigaria varia (Crown vetch) and solitary ground nesters, Chrysanthemum leucanthemum (Oxeye daisy) and social ground nesters, and Eupatorum fistulosum (Joe-pye weed) and social ground nesters. In 2016-17, 54 plant groups were observed interacting with four functional groups of bees. The three strongest interactions are those between social ground nesters and Sorbaria sorbarifolia (False spirea), Taraxacum (Dandelion) and solitary ground nesters, and Leucanthemum vulgare (Oxeye daisy) and cavity nesters.

Considering the high degree of flowering plant species turnover at the site between 2002-03 and 2016-17, it is important to focus on the interactions involving the flowering plant species in common between these studies (Figure 15 and Figure 16). There were only 30 plant groups in

common between the two studies. Forty-five plant groups (74%) involved in bee interactions in 2002-03 were not observed in 2016-17 interactions. In 2016-17, 24 plant groups served as resources for wild bees: interactions that were not observed in the previous study. From a total of 1381 interactions from both datasets, only six percent of the interactions persisted. Seventy-one percent of the interactions were lost from 2002-03 to 2016-17, while only 23 percent were gained. Figures 13 and 14 present a similar pattern to those in Figures 15 and 16, showing reduced interconnectedness between plants and pollinators in the network.

Calculations of nestedness revealed that both the 2002-03 and 2016-17 networks are nested (p<0.05) irrespective of which (CE, EE, EF and FE) null models are used (Table 6). Comparing nestedness between 2002-03 and 2016-17 is not possible due to the lack of replicates. The different null models are used to calculate relative nestedness in different ways; but these are not considered replicates since calculations are based on the same plant-pollinator interaction matrix. Since there are only two values of relative nestedness (one each for 2002-03 and 2016-17 respectively) it was not possible to statistically test if they are different.

Figure 12: Plant-pollinator interaction network of wild bees and flowering plants at the site, observed and collected by sweep netting in the years 2016 and 2017. Five functional guilds (bumblebees, cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 61 flowering plant groups were recorded, with 1616 interactions. The grey bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant.

Figure 13: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants at the site, observed and collected by sweep netting in the years 2016 and 2017. Four functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 54 flowering plant groups were recorded in 1312 interactions. The grey bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant.

Figure 14: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants at the site, observed and collected by sweep netting in the years 2002 and 2003. Four functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 75 flowering plant groups were recorded in 8580 interactions. The grey bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant.

Figure 15: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants observed and collected by sweep netting in the years 2002 and 2003. Only flowering plants found in both the consecutive year pairs, 2002-03 and 2016-17, are shown in this network. Four bee functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 30 flowering plant groups are common between the two studies. The grey bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant.

Figure 16: Plant-pollinator interaction network of wild bees (excluding bumblebees) and flowering plants observed and collected by sweep netting in the years 2016 and 2017. Only flowering plants found in both the consecutive year pairs, 2002-03 and 2016-17, are shown in this network. Four bee functional guilds (cavity nesters, cleptoparasites and social parasites, social ground nesters, and solitary ground nesters) and 30 flowering plant groups are common between the two studies. The grey bars represent the relative frequency of interactions with each plant and the coloured arrows represent the frequency of interactions between each bee group and plant. Table 6: Measures of plant-pollinator interaction network nestedness calculated using 50 runs of each null model (EE- equiprobable row and column totals, CE- proportional row and column totals, EF- equiprobable row totals and 59

fixed column totals, FE- fixed row totals and equiprobable column totals) with the NeD software for the year pairs 2002-03 and 2016-17. NODF (Nestedness metric based on Overlap and Decreasing Fill) is the calculation of nestedness of the matrix input into the NeD program. The Z-score is a measure of how nested the focal matrix is compared to the null model. The RN (relative nestedness) is the degree of nestedness of the focal matrix compared to the null model. The P value corresponds to the Z-score, where the focal matrix is considered to be nested (compared to the null matrices) at a value of <0.05.

Null models CE EE EF FE 2002/2003 NODF 68.98 68.98 68.98 68.98 Z-score 4.21 3.15 14.30 2.33 RN 0.13 0.15 0.07 0.12 P value < 0.001 < 0.001 < 0.001 < 0.05 Nested? Yes Yes Yes Yes 2016/2017 NODF 67.46 67.46 67.46 67.46 Z-score 6.17 8.33 15.27 2.88 RN 0.34 0.38 0.26 0.19 P value < 0.001 < 0.001 < 0.001 < 0.01 Nested? Yes Yes Yes Yes

4.4 Discussion

Between the two studies, the number of interactions decreased dramatically from 8580 to 1312. Out of the common interactions between 2002-03 and 2016-17, only 6% of these persisted. Considering that 71% of the interactions were lost, while only 23% new interactions were gained, this shows there was a significant overall loss of interactions between plants and pollinators between these time points. This disparity could be a result of the extirpation of bee and/or plant species from the site. Plant turnover appears to be a significant factor in the network changes seen over time. There was a high degree of plant turnover between the years 2002-03 and 2016-17. Forty-five plant groups (71%) were lost in this time and only 24 new plant groups were gained. The previous studies in 1968-69 and 2002-03 did not include a comprehensive inventory of plant species so it is unknown if the plants involved in “lost” interactions actually persisted at the site. For example, Cannabis was observed at the site in the most recent two studies. Interactions between this plant and wild bees were recorded in 2002-03, but not in 2016-17. Further studies should include a comprehensive plant inventory to potentially account for the loss of certain interactions between plants and flower visitors. In another long-term study of a bee community in Illinois, it was determined that the loss of bee species from the site was one of the key factors in plant-pollinator interaction loss and network degradation from the late 1800s to 2010 (Burkle et al., 2013). In contrast to the findings in this thesis, all plant species that were observed in interactions with pollinators remained present at the site over time (Burkle et al., 2013).

Comparisons of plant-pollinator networks between 2002-03 and 2016-17 at the site showed evidence of decreased resiliency (measured by nestedness) over time, though this cannot be statistically tested. It has been suggested that plant-pollinator networks are somewhat resilient to disturbances over time (Burkle et al., 2013). This is thought to be due to functional redundancy in the interactions. Functional redundancy is visually represented by a highly connected, and nested structure, where specialist pollinators interact with a subset of the flowering plants with which generalist pollinators also interact (Petanidou & Ellis, 1996; Bascompte et al., 2003; Ollerton et al., 2003; Waser & Ollerton, 2006; Alarcόn et al., 2008). Fewer interactions in a mutualistic network puts increased pressure on each individual species. Increased pressures can be detrimental to a network, resulting in a domino effect of further species loss (Gilbert, 2009). 61

The NODF values are similar between the two studies, though slightly higher in 2002-03 compared to 2016-17. This is somewhat expected since there were many more interactions recorded in 2002- 03. This value is not typically compared between matrices since differences in the number of interactions are not accounted for (Song et al., 2017). The NODF values for both the 2002-03 (NODF=68.98) and 2016-17 (NODF=67.46) plant-pollinator interaction networks are closer to 100 than zero, indicating that both networks could be considered quite nested.

On average, relative nestedness scores from all the null models used in the 2016-17 matrix are higher compared to the relative nestedness values from the 2002-03 matrix. This calculation accounts for differences in the number of interactions, and can be compared between studies, but these differences cannot be statistically tested due to the fact that there are only two values (Song et al., 2017; Strona et al., 2014a, 2014b). Although both networks are considered to be nested, the 2016-17 matrix is more nested than the 2002-03 matrix. These values are somewhat comparable to other studies of mutualistic networks, with values ranging from approximately 0.19 to 0.33.

This network is similar to others in that the number of bee species was greater than the number of plant species (Petanidou & Ellis, 1993; Alarcόn et al., 2008). Bee abundance is considered to be a major influence on the nestedness of a plant-pollinator network (Vázquez & Aizen, 2004; Stang et al., 2006; Alarcόn et al., 2008). The number of species sampled affects the probability of finding a particular interaction between plant and pollinator. The substantial decrease in the estimate of wild bee abundance from 2002-03 to 2016-17 could have limited the number of rare interactions found at the site. However, it is not likely that further sampling would influence the plant-pollinator network structure significantly, since only rare plant and bee species, and rare interactions would be added (Alarcόn et al., 2008).

From this analysis, it appears that the changes in the plant community at the site are important factors in the structure of the plant-pollinator networks, considering the decrease in the number of plant species documented in interaction with bees between studies. If these plants are present, this could suggest that phenological mismatch between bees and flowering plants. Specialist bees would be impacted by phenological mismatch more than generalist species because of their specific floral resource requirements (Burkle et al., 2013). Further sampling of the bee community 62

could help to determine the reason for these extensive structural changes in the plant-pollinator network between 2002-03 to 2016-17. Future work at this site should focus on conducting a plant inventory to determine if the plants found in interactions at the site in 2002-03 that were not found in interaction in 2016-17 are still present at the site.

CHAPTER 5 Impacts of climate and habitat change on wild bee diversity in Caledon, Southern Ontario

5.1 Introduction

5.1.1 Habitat change

Habitat loss is often cited as one of the most important factors in pollinator decline (Biesmeijer et al., 2006; Brown & Paxton, 2009; Vanbergen et al., 2013; Ollerton et al., 2014; Goulson, 2015). Both urbanization and agricultural intensification are significant contributors to pollinator habitat loss and fragmentation. This is also likely to be true for Southern Ontario, a region that has undergone recent significant urbanization and agricultural intensification and is a bee biodiversity hotspot in Canada (Pindar et al., 2017).

Recent data indicated that 86% of the human population of Ontario are now living in urban areas (Statistics Canada, 2011), yet there is little research on the effects of urbanization on wild bee populations in southern Ontario (Horn, 2010; MacIvor 2015; Vickruck & Richards 2017). Comparisons of a variety of land uses within urban settings in a number of Southern Ontario cities indicated that the existence of naturalized areas was correlated with both higher wild bee abundance and diversity (Horn, 2010). This study emphasized that bee habitat could be added to any location within a city, since land use did not seem to affect bee abundance. Additionally, a relatively small naturalized area was found to have significant positive impacts on bee biodiversity (Horn, 2010). One way to incorporate pollinator habitat in urban settings is the addition of green infrastructure (Colla et al., 2009; MacIvor & Packer, 2015; Aronson et al., 2017). Green rooves can provide locations for bee nesting and foraging within cities for bees to nest and forage (Colla et al., 2009). Biodiversity surveys of green rooves in Toronto, Ontario, determined that there were no substantial differences in the bee community compared to the ground level sampling (Colla et al., 2009). However, bee species richness on green rooves was much lower than that of the countryside just outside Toronto. MacIvor & Packer (2015) suggested that the design of green infrastructure should consider building height and the proximity to ground level semi-naturalized areas. On Toronto green rooves, the height of the building negatively impacted bee populations, 64

while the presence of green space nearby was beneficial (MacIvor, 2011). In St. Catharines, Ontario, the Eastern Carpenter Bee (Xylocopa virginica) is thought to be one of the few species that thrive in the face of anthropogenic change, specifically urbanization (Vickruck & Richards, 2017). Nesting sites are important resources for wild bees and are often scarce in urban settings. Eastern Carpenter Bees nest almost exclusively in man-made wood structures, so benefit from living in urban habitats (Vickruck & Richards, 2017).

Approximately one quarter of Canada’s farmers live in Ontario, covering 8% of the total Canadian farmland (Statistics Canada, 2012). Agricultural practices have changed significantly in Ontario over the last century with fewer farmers and farms, but increased farm size (Statistics Canada, 2012). Over time farmland area has decreased, but Southern Ontario is still considered to be mainly “dependable agricultural land with pockets of settlement areas scattered throughout” (Pindar et al., 2017). Studies on pollinator habitat loss due to agriculture in Southern Ontario are minimal. There is evidence to suggest that habitat loss (and pesticides) are not resulting in declines of certain bumblebee species (Szabo et al., 2012). Additionally, it is known that squash bees live in agricultural land, requiring cucurbit crops to survive (Willis & Kevan, 1995).

Habitat restoration refers to efforts to return degraded areas to their natural state, with the goal of restoring biodiversity. Pollinator habitat is considered to be restored when “all functional pollinator groups are retained, and seed set is occurring” (Forup & Memmott, 2005). Nesting sites and floral resources are the most important considerations when restoring a habitat for pollinators (Murray et al., 2012; Sydenham et al., 2014). Specifically, floral resources might be the key determination of wild bee community composition after restoration (Cusser & Goodell, 2013; Morandin & Kremen, 2013b; Havens & Vitt, 2016). In Ontario, there is well established evidence that restoration practices have positive effects on bee communities (Pindar et al., 2017). Wild bee community composition appears to vary over time, dependent on the length of time since restoration (McLeod, 2013; Rutgers-Kelly & Richards, 2013). In Southern Ontario, comparisons of bee communities at a wet meadow site both four to six years after restoration suggested that both diversity and evenness of wood nesters and cleptoparasites was lower in more recently restored habitats (McLeod, 2013). Relative abundance of different bee nesting guilds appeared to be affected by different restoration methods. Comparisons of bee communities at restored, former 65

landfill sites reported that bee abundance was highest at the recent landfill site, compared to the degraded and control sites (Richards et al., 2011; Rutgers-Kelly & Richards, 2013; Onuferko et al., 2017). In the three to four years after restoration, the abundance and species richness of bees increased rapidly, followed by a decline (Onuferko et al., 2017). Ground nesting species appeared to move into the restored area faster than cavity nesters, probably due to the availability of nesting sites. Other factors including severe droughts, succession, and suburbanization appeared to affect the bee community both at the control and landfill sites (Onuferko et al., 2017).

5.1.2 Climate change

In Ontario, the distributions of many wild pollinator species are understudied, which makes it difficult to evaluate how well they are coping with climate change. Shifts in range and/or phenology have been suggested as possible responses to climate change. A number of studies, mostly focusing on bumblebees, have been conducted to investigate the effect of climate change. North American bumblebees have not been tracking warming trends northward, while their southern range boundaries are shrinking (Kerr et al., 2015). Furthermore, bumblebees are not moving to higher elevations as the climate warms across North America (Kerr et al., 2015; Sirois- Delisle & Kerr, 2018). Climate change projections indicate that temperatures may rise faster than North American bumblebees are able to shift their ranges (Sirois-Delisle & Kerr., 2018). Those bumblebee species able to tolerate a wider variety of climates may be at an advantage during this period of global warming (Williams & Osbourne., 2009). Similar to bumblebees, it is thought that Xylocopa virginica populations in Ontario will not be able to shift their range northward in response to climate change (Skandalis et al., 2011).

Climate change can also prompt shifts in wild bee phenology (Bartomeus et al., 2011). There are few studies that focus on this topic in Ontario specifically, though pollinator phenological shifts have been well documented in other regions (Menzel et al., 2006; Burkle et al., 2013). These shifts are particularly concerning due to the close interactions between plants and pollinators. Some studies suggest that plants and pollinators may shift their phenologies synchronously, while others provide evidence that phenologies of either plants or pollinators have shifted, and/or are changing more rapidly (e.g. Forrest & Thomson, 2011; Willmer, 2014). 66

Increased rainfall, rising minimum and maximum temperatures, and decreased snowfall is expected in Southern Ontario as a result of climate change (Fausto et al., 2015). Climate data from the three study periods (1968-69, 2002-03 and 2016-17) near the site were expected to align with these projections. There have been considerable increases in the degree of urbanization in and around the Greater Toronto Area since the late 1960s (Puric-Mladenovic & Strobi, 2006). Overall the amount of pollinator habitat surrounding the site was expected to decrease from 1968-69 through 2002-03 to 2016-17. If changes in climate and land-use did not affect the wild bee community at the site, the species richness, diversity, evenness and abundance for both the overall bee community and within each of the functional guilds would not be correlated with the climate and land-use change data at the site over time. However, if the climate and land-use changes over time affected the wild bee community negatively, both factors acting individually or in combination were predicted to correlate with a decline in wild bee species richness, diversity, evenness and abundance for the total bee community, and within each functional guild.

5.2 Analytical methods

Land-use data were obtained from the Southern Ontario Land Resource Information System (SOLRIS) (Ministry of Natural Resources and Forestry, 2016). Sixteen buffer zones, ranging from 50 meters to 2000 metres in radius were used for the analysis. These buffer zones were chosen based on the variety of distances wild bees are able fly during foraging bouts, depending on their size (Gathmann & Tscharntke, 2002; Greenleaf et al., 2007; Zurbuchen et al., 2009). SOLRIS data ranges from 1999-2002 and 2009-2011, beginning later than the first wild bee sampling event at the site in 1968, and stopping short of the last bee sampling event in 2017. Data from the year 1999 were used to represent the years 1968-69, and 2002 data were used to represent Grixti’s study conducted in 2002-03. 2011 data were used to represent the land use in 2016-17.

Historical rainfall, snowfall, minimum and maximum temperature climate data were obtained from the Government of Canada website (2019) and are shown in Table 7. These four measures of climatic data were not available from the same weather station for the years 1968-69, 2002-03 or 2016-17. As such, data from multiple weather stations (Glen Haffy Mono Mills, Orangeville, 67

Orangeville MOE, Sandhill, Mono Centre, and Georgetown WWTP) ranging from 12.14-24.45 kilometres away from the site, were used in combination to fill these information gaps. Rainfall and snowfall data were obtained for the entire year, while minimum and maximum temperature data were obtained per sampling month and averaged throughout the entire sampling season. Multiple linear regressions and simple linear regressions were conducted (using the statistical program R 3.5.1; R Core Team, 2013) to examine the effects of habitat and climate, acting together and individually, on the species richness, evenness, diversity and abundance of the total wild bee community and for each functional guild. The method of wild bee data collection is described in Chapter 2. These regressions were evaluated with a significance of p < 0.05 for the overall model, and each of the individual predictors. The R pwr package was used to evaluate the power of the regressions (Champely, 2015).

To visualize bee abundance, evenness, diversity and richness, changes in bee phenology through time, and impacts of climate, the data were plotted using OriginPro, student version (OriginLab, 2019). Phenology graphs were created for the complete 2002-03 and 2016-17 datasets, and the bee species in common between these two consecutive year pairs. The 1968-69 data were not used in comparisons of phenology due to insufficient information regarding dates of specimen collection. The signal processing technique, smoothing (binomial method) was used to produce the trendlines in phenology over the course of the sampling seasons since no particular trend was expected (e.g. exponential, linear etc.).

1968 0.5 10.6 714.5 197.1

1969 0.6 10.5 658.2 193.5

2002 2.4 12.0 543.7 170.8

2003 0.7 11.0 741.4 177.0

2016 3.1 12.7 305.8 62.0

2017 2.3 11.4 563.5 49.0

5.3 Results

Climate and land-use data between 1968 and 2017 were assessed for correlation with abundance evenness, richness and diversity of the wild bee community. The total pollinator habitat within a two-kilometre radius from the site was 992.4 hectares, composed of treed swamp (30.62%), mixed forest (21.79%), deciduous forest (19.32%), pasture (16.68%), forest (1.23%), coniferous forest (4.98%), thicket swamp (0.89%), marsh (1.07%), plantation (2.59%) and hedge row (0.17%). In 1999, 2002 and 2011, no changes in habitat classification were detected using SOLRIS data in ArcGIS at 16 buffer radii from 50 to 2000 metres from the site. Since the abundance and distribution of habitat types surrounding the site did not change significantly over time, this suggests that land-use patterns are unlikely to explain patterns in abundance, diversity, evenness and richness of the wild bee communities during the three studies.

Conversely, minimum temperature, maximum temperature, rainfall and snowfall data relevant to the site did vary over the six years of data collection. Multiple linear regressions showed some evidence that the four climate variables affected species richness (Tables 8, 11). The total species 2 richness (R = 0.9999, F1,4 = 12520.00, p = 0.01) was predicted by the four climate variables: minimum temperature (B= -33.43, p = 0.01), maximum temperature (B = 81.85, p < 0.01), rainfall (B = 0.31, p < 0.01) and snowfall (B = -0.50, p < 0.01). This model predicts that for every degree increase in minimum temperature or additional centimetre of snowfall, then total species richness at the site would either increase by 33.43 or fall by 0.50 units, respectively. Additionally, for each increase by one degree Celsius in maximum temperature and millimetre of rainfall over time, 81.85% and 0.31% more species would be observed at the site. There was no evidence of a

relationship between climate and wild bee species evenness, diversity or abundance overall and 2 for each functional guild (Table 8). The richness of solitary ground nesters (R = 0.9955, F1,4 = 2 280.1, p = 0.04) and cavity nesters (R = 0.9973, F1,4 = 456, p = 0.04), specifically, appeared to rely on climate. Decreased snowfall over time (B = -0.31, p = 0.03), higher rainfall (B = 0.15, p = 0.03), increased maximum temperatures (B = 39.46, p = 0.03), and decreased minimum temperatures (B = -18.24, p = 0.06) correlated to a greater species richness in the solitary ground nester guild. Similarly, decreased snowfall (B = -0.10, p = 0.03), higher maximum temperatures (B = 16.22, p = 0.03), increased rainfall (B = 0.11, p = 0.02), and increased minimum temperatures (B = 0.68, p = 0.49) positively affected the species richness within the cavity nester guild.

A power test for the multiple regressions was conducted in the statistical program R, using the ‘pwr’ package (power = 0.051). Considering the commonly used power threshold of 0.8, the power of the regressions conducted to test for effects of climate variables on evenness, abundance, diversity richness is extremely low. This suggests that the likelihood of a statistically significant result in the linear regressions is about 5%, meaning that the probability of a Type II error (a “false negative”) is very high.

Simple linear regressions were also used to analyze the effects of individual climate variables on the abundance, richness, diversity and evenness of the total bee community and for each functional guild (Table 11). The abundance of the total bee community was predicted by the snowfall, where increased snowfall is correlated with a higher abundance (B = 26.46, F1,4 = 0.03) (Figure 17). For every one-centimetre increase in snowfall, total wild bee abundance will increase by 26.46 individuals. Increased snowfall (B = 19.76, F1,4 = 0.04) and increased minimum temperature (B =

-1232.39, F1,4 = 0.03) is correlated with increases in social ground nester abundance and decreases in social ground nester abundance respectively (Figures 18 and 19). For every one-centimetre increase in snowfall, social ground nester abundance will increase by 19.76 individuals. For every one-degree Celsius increase in minimum temperature, social ground nester abundance will decrease by 1232.39 individuals.

2 Adjusted R F4,1 p

Evenness (J)

Solitary ground nesters 0.932 18.190 0.174

Social ground nesters 0.545 2.495 0.439

Cavity nesters -2.235 0.136 0.946

Cleptoparasites and social parasites -1.543 0.242 0.888

Total 0.480 2.156 0.467

Diversity (H’)

Solitary ground nesters 0.993 169.700 0.058

Social ground nesters 0.602 2.893 0.412

Cavity nesters 0.131 1.188 0.589

Cleptoparasites and social parasites 0.331 1.620 0.524

Total 0.633 3.155 0.397

2 Adjusted R F4,1 p

Richness

Solitary ground nesters 0.996 280.100 0.045*

Social ground nesters -0.048 0.942 0.639

Cavity nesters 0.997 456 0.035*

Cleptoparasites and social parasites 0.779 5.396 0.311

Total 0.999 12520.000 0.007*

Abundance

Solitary ground nesters 0.5198 2.353 0.450

Social ground nesters 0.1368 1.198 0.587

Cavity nesters 0.5017 2.259 0.4578

Cleptoparasites and social parasites 0.01282 1.016 0.6226

Total 0.9747 49.19 0.1065

Phenological patterns of bee species activity provides some more information about the potential effects of climate variability on the bee community at the site (Figure 20). For each of the functional guilds examined, it appears that a higher percentage of the total number of species were collected in most weeks in 2002-03 compared to 2016-17.

Comparing the trends in species richness percentage over the entire sampling season between the years 2002-03 and 2016-17 can provide some further insights, as different species are more active during different times. Examining differences in peaks (weeks with comparatively high proportions of species) and valleys (weeks with comparatively low proportions of species) could suggest phenological shifts of wild bees at the site over time. Cavity nester species richness percentages first peak aligns between the two time periods (around week 27 and 28). The next trough (2002-03: week 31, 2016-17: week 30) and peak (2002-03: week 33, 2016-17: week 32) are slightly misaligned (Figure 20), occurring earlier in 2016-17. Solitary ground nester species richness patterns appear similar throughout the sampling periods in the two studies (Figure 20). The overall patterns of the social ground nesting species richness over the course of the sampling periods in the two studies are challenging to determine, showing much more variability (Figure 20). Overall, cleptoparasite and social parasite species richness appears to be more stable in 2016- 17 than 2002-03, though the pattern is not very clear (Figure 20).

Figure 17: The positive relationship between snowfall in centimetres and abundance of the total bee population at the site per sampling year (B = 26.46, F1,4 = 0.03).

Figure 18: The positive relationship between snowfall in centimetres and the abundance of social ground nesters at the site per sampling year (B = 19.76, F1,4 = 0.04).

Figure 19: The negative relationship between minimum temperature in degrees Celsius and social ground nester abundance at the site per year (B = -1232.39, F1,4 = 0.03). 74

5.4 Discussion

Various climatic events affected the weather in Southern Ontario over the duration of the studies. El Niño and La Niña events coincided with some sampling years, potentially influencing the bee community. Combined with global warming, the weather in these years was considered to be abnormal (McDermid et al., 2015). El Niño events occurred in 2002-03 and 2015-16, overlapping with sampling in the spring and summer of 2002, 2003 and 2016. The El Niño event in 2002-03 was considered to be moderate, while the event in 2015-16 was considered to be very strong and followed by a weak La Niña event in 2016-17 (Null, 2019). The year 2015 was the warmest on record, which almost certainly impacted the bees at the site (McDermid et al., 2015). El Niño and

La Niña events affect the weather most in the winter and spring months, during the overwintering period and perhaps affecting emergence times in the bee lifecycle.

The decreasing trend in snowfall and rainfall, coupled with the rise in average minimum temperature at the site, is somewhat consistent with climate change predictions for Southern Ontario (Fausto et al., 2015). It was reported that average temperatures are expected to rise, with more precipitation falling as rain rather than snow, and an increase in rainfall overall. Higher energy expenditure as a result of winter warming has been found to negatively impact certain bee species through higher levels of weight loss (e.g. Osmia cornuta) (Frund et al., 2013). Both advanced and delayed emergence was also recorded as a response to high winter temperatures, varying between species and sexes, which could lead to phenological mismatches with floral resource blooming (Frund et al., 2013). Other bees appear to be less sensitive to these climate changes, likely due to differing physiology between species (Frund et al., 2013). Reduced snowfall can impact the survival of wild bees in the overwintering stage through reduced insulation (Kudo, 2014). Additionally, there is evidence to suggest that snowmelt timing can affect the emergence times of overwintering queens (Kudo & Ida, 2013). Although rainfall is expected to increase in Southern Ontario, the reduced rainfall in 2016 and 2017 could have impacted the growth and reproduction of flowering plants, which has been shown to also limit bee activity (Minckley et al., 2013). Bivoltine bees might be especially affected by reduced floral resources, due to continuous flowering required throughout the reproductive season (Onuferko, 2014). Bee life cycles can also be affected by temperature in other ways, dependent on location and habitat. In the Rocky Mountains in Colorado, warmer winters can trigger a 1-year life cycle (rather than 2 years) in species with the ability to vary their life cycle like Osmia iridis (Forrest et al., 2019). A one-year life cycle is a risky strategy because individuals may not be able to pupate before winter, whereas there is more time to complete this stage over two years (Forrest et al., 2019).

Extreme weather events, which are predicted to increase in frequency due to climate change, can also cause direct mortality to wild bees (Oyen et al., 2016). Some species may be able to tolerate climate change better than others. It has been suggested that nesting in cavities may provide a buffer for more extreme temperatures and weather events (Villalobos & Vamosi, 2018). Additionally, some insects are able to shift their ranges in response to increased temperatures (Roy 76

& Sparks, 2000; Walther et al., 2002; Parmesan & Yohe, 2003); however, this may not be possible if the species is at its northern range limits, resulting in habitat loss in the southern portion of the range (Kerr et al., 2015). Some species may even benefit from rising temperature, which could extend the nesting season, providing greater opportunities to reproduce (Richards et al., 2015).

Some evidence of the impacts of climate change was observed in 2002-03, with male bees recorded flying later in the season (Grixti, 2004). I found some evidence that climate change was impacting the bee community at the site over three points in spanning 49 years. Specifically, the species richness of the total bee community was predicted by snowfall, rainfall minimum temperature and maximum temperature. Solitary ground nester and cavity nester species richness was also predicted by these climate variables. Considering the low power (see Chapter 5: Results) of these multiple regressions, it is possible that there could be an effect of certain climatic variables on species richness, diversity, abundance and evenness when none were detected by the regressions. This could have resulted from an insufficient number of data points (sampling years), or lack of replicate datasets. Multicollinearity is also a factor to consider in these multiple regressions. The four predictors (minimum temperature, maximum temperature, snowfall and rainfall) are all climate measures, and are highly related to each other. Considering the high R2 values, it is possible that the variation in these climate variables are being accounted for multiple times, skewing the regressions. To eliminate this issue, simple linear regressions were conducted to analyze the relationship between each climate variable individually, and the abundance, richness, evenness and diversity overall and for each functional guild. The total wild bee abundance was predicted by snowfall. The social ground nester abundance was affected by the minimum temperature and snowfall at the site. Snowfall levels impact the insulation of overwintering bees, which is an important factor for ground nesters (Kudo, 2014). It seems that increased snowfall may have impacted the ability of social ground nesters to survive the winter months. Social ground nesters composed an extremely large proportion of the total bee community in 1968-69, which may have had an effect on the relationship between the overall bee population and snowfall over the duration of the three studies. Increasing minimum temperatures were shown to negatively impact social ground nesters which could have impacted the phenology of plants and bees at the site leading to

a phenological mismatch. It is important to continue this historical study to determine with more certainty if the changing climate is influencing the wild bee community at the site.

There is some potential evidence of phenological change in the forward direction in one functional guild (cavity nesters) between the years 2002-03 and 2016-17. With the results of the multiple regressions, it appears that cavity nesters are one of the groups most affected by climate change at this site. This phenological change demonstrates adaptability to climate change effects. This phenological change could have contributed to the structural differences in the plant-pollinator interaction network between 2002-03 and 2016-17, where many interactions were lost over time. In 2002-03 there were 1492 more interactions between cavity nesters and flowering plants compared to 2016-17. Overall, solitary ground nester species richness patterns appear quite similar between 2002-03 and 2016-17 suggesting that their phenology is not being affected by climate change currently. Social ground nesting species richness is less stable throughout the sampling seasons in 2002-03 than 2016-17, with more extreme variability. There may be some other factors influencing these patterns since there appears to be shifts in both directions. For example, the valley at week 31 in 2002-03 occurs earlier than the valley in 2016-17 (week 33). Notably, there is also a large peak early in the season in 2002-03 which is not seen in 2016-17, and a large trough at the end of the season in 2016-17 that is not seen in 2002-03. Despite the evidence suggesting that social ground nesters are impacted by climate, there is no evidence that climate is affecting the social ground nester phenology. Cleptoparasites and social parasite species richness also appears more stable in 2016-17 compared to 2002-03, showing no clear phenological shifts over time.

In 2002-03, a higher percentage of the total species were recorded each week compared in 2016- 17. Overall, more individuals were collected in 2002-03 than 2016-17, which is likely the reason for this trend. Substantially more individuals collected overall in 2002-03 would likely mean that more individuals were collected during most weeks when compared to 2016-17. Abundance is linked with species richness, as discussed in Chapter 3, so collecting a higher percentage of species in each guild would be expected in 2002-03. This is especially apparent in the cavity nester and social ground nester functional guilds, perhaps because they are more abundant in general compared to the other guilds. 78

It is not possible to determine if the phenology of cavity nesters has actually shifted given the relatively short period of time separating the two studies (15 years), compared with other studies. Phenological shifts in bees were observed over a period of 120 years (Burkle et al., 2013). The phenologies of ten northeastern North American bee species representing a range of socialities and nesting preferences (including cavity nesters) were also found to have advanced by a mean time of 10 days over 130 years (Bartomeus et al., 2011). Eight of the species analyzed by Burkle and colleagues (2013) were recorded at the site in 1968-69, 2002-03 and/or 2016-17. Since data were collected every week in 2002-03 and every two weeks in 2016-17, the sampling events were also likely not frequent enough to determine the presence of phenological change. It is possible that starting data collection earlier in the year could have revealed some insights into changes phenology, since sampling in 2016-17 began in early May, after the emergence of some early emerging bee species (Rutgers-Kelly & Richards, 2013). The greatest phenology shifts occur in bees that are active in early spring (Burkle et al., 2013; Frund et al., 2013). Continuing this historical study, with more frequent sampling over the entire active season of wild bees would provide more conclusive evidence to determine if the wild bee phenology is changing at the site over time.

There is no numerical evidence to suggest that land use change up to two kilometers away from the site is contributing to changes in richness, abundance, diversity or evenness. Though no changes were seen in the SOLRIS habitat data near the site, comparisons of satellite and aerial images, and written descriptions in MacKay and Grixti’s theses revealed some differences. It is possible that some of these habitat changes were not represented in the SOLRIS data, firstly because these changes occurred well before the habitat data were collected (no data exists for 1968- 69 so the earliest data were used), and secondly, because the habitat changes were too fine-scale to be detected at the resolution of the SOLRIS datasets. Overall, the observed changes in habitat at the site over the duration of the study have resulted from both the land owners and likely succession, as described by Grixti (2004). Major differences in habitat include the construction of a house, pond, and long gravel driveway sometime between 1969 and 2002. During this time, a large stand of coniferous trees was also planted. After 2005, but before 2016, a portion of these trees fell in a storm, creating an open area in the forest.

In 1968-69, the land was classified as an early successional field, which has been demonstrated to very good wild bee habitat (Carvell, 2002; Potts et al., 2003; Taki et al., 2013). The man-made structures reduced the overall amount of habitat that could be used by bees. The lawn presumably started being maintained near the time of house construction, and potentially reduced the abundance of flowering plants. In the years 2016-17, the garden areas mostly contained native wild flowers alongside Hemerocallis fulva (Orange daylily), Sorbaria sorbifolia (False spirea) and various species of ornamental sunflowers. Some bees were collected on sunflowers, but a very large number were collected on Sorbaria sorbifolia bushes bordering the house, suggesting that these garden areas are important resources for the bee community. This aligns with other findings that suggest gardens can be important resources for bees in both urban and agricultural landscapes (e.g. Samnegard et al., 2011; Plascencia and Philpott, 2017).

The area of fallen trees within the forest was a key section of the property where specimens were collected in 2016-17. The fallen trees served as nesting sites for cavity nesting bees, and presumably this, along with natural ecological succession, was the driving force behind the increase in cavity nesters at the site (Grixti, 2004; Grixti & Packer, 2006). This location also contained the greatest wildflower diversity and appeared to provide one of best foraging locations for wild bees at the site. The formation of this clearing (which resulted from a storm) is thought to have drastically altered the overall flowering plant community structure, and consequently the bee community over time (Banaszak & Ratynska, 2014). Forest openings contain more abundant and diverse bee communities than mature forests, which seems to align with what was observed during sampling in this particular area (Roberts et al., 2017). The significant reduction in bee abundance over time at the site is very interesting when compared with the findings of Winfree et al. (2009). Significant reduction in bee abundance and species richness has been demonstrated in areas experiencing extreme habitat loss (Winfree et al., 2009). The drastic reduction in abundance between the three time periods at the site would suggest that great habitat loss has occurred, though this was not represented in the numerical habitat data.

The clades least affected by land-use change seem to be Bombus and Lasioglossum (Dialictus) (Grab et al., 2019). The sociality of these groups, and their longer flight period may play a role in their resilience to habitat degradation and loss. Alternatively, Andrena is thought to be more 80

sensitive as the species belonging to this genus are generally solitary and have a short flight period. Overall, changes observed in habitat at the site appeared to affect species richness wild bee community at the site, but the lack of numerical evidence of habitat change (over a large spatial scale) made these impacts hard to quantify. Since further changes in habitat are predicted in the surrounding area, a special focus on the genus Andrena may be of particular importance in future conservation considerations.

CHAPTER 6 Conclusions

This is one of very few long-term studies on wild bee diversity in Canada, with detailed information about a bee community on a private property in Caledon, Ontario at three timepoints spanning almost 50 years. The overall diversity, evenness and richness of the wild bee community actually increased, while the abundance decreased dramatically from 2002-03 to 2016-17. Interestingly, these results suggest that wild bee diversity and ecological function can mostly be conserved despite significant reductions in the number of individuals.

This study shows that the plant-pollinator community has remained quite resilient despite prevalent plant species loss and bee abundance declines at the site. Plant-pollinator interactions were compared for the years 2002-03 and 2016-17, revealing that both networks were considered to be nested, a measure of resiliency. The 2016-17 network showed a more nested structure, after 81

accounting for the differences in plants at the site between the two time periods. This evidence suggests that plant-pollinator networks can maintain a highly nested pattern despite drastic restructuring over time. Some evidence of bee phenology change was observed in the cavity nester guild, which could have contributed to interaction loss within the plant-pollinator network, though it is unlikely phenological change could be detected in less than 50 years, sampling once a week and once every two weeks. If this study is continued, I would recommend completion of a full inventory of flowering plants be conducted to determine the causes of interaction losses over time. More frequent sampling would also help to detect any bee phenology changes.

Many habitat changes were observed throughout the duration of the three studies, but their effects on the wild bee population can only be speculated upon since they were not represented numerically in the SOLRIS dataset. It is likely that future SOLRIS data would detect loss in pollinator habitat considering the current rate of urbanization near the site. Climatic variation in snowfall, rainfall, minimum temperature and maximum temperature appears to be correlated with total species richness, and species richness of cavity nesters and solitary ground nesters. A power test revealed that the number of study years may have been insufficient to determine the effects of climate and habitat change on species richness, diversity, evenness and abundance, resulting in a false negative. Additionally, collinearity may be a factor affecting the multiple regressions. When tested individually, snowfall appeared to have a positive effect on the total bee abundance and social ground nester abundance at the site, possibly due to increased insulation during overwintering. Minimum temperature appeared to have a negative impact on the social ground nester abundance, potentially leading to phenological mismatches between bees and flowering plants. The effects of the two important stressors: climate change and land use change could be further investigated if this study continues in the future. Since no significant changes in land use have been recorded in the area, it is possible that the site is acting as a pollinator sanctuary. Based on the observed changes in habitat it appears that microhabitat provisions are important, especially when landscape scale changes cannot be detected.

This study demonstrates the need to continue long-term studies of wild bee populations, especially to determine the driving factors of changes in community composition and structure. Future studies at this site in particular, could provide further information regarding the effects of climate change 82

and habitat change on the wild bee community structure and composition. These two stressors are expected to become increasingly important over time, and additional years of data could contribute to a more robust analysis of their effects on wild bee communities. Another strategy could be to include replicate datasets (sample bee communities in sites with similar characteristics nearby). Overall, the findings of this study show somewhat positive outcomes in terms of the functionality of the bee community at the site, and its interactions with flowering plants, despite significant reductions in bee abundance and the overall number of plant-pollinator interactions. This study contributes important information about the status of wild bees in Southern Ontario, and the potential long-term effects of climate change and habitat change.

REFERENCES

Aguirre-Gutiérrez, J., Kissling, W. D., Carvalheiro, L. G., Wallis DeVries, M. F., Franzén, M., & Biesmeijer, J. C. (2016). Functional traits help to explain half-century long shifts in pollinator distributions. Scientific Reports, 6, 24451, doi:10.1038/srep24451

Ahrné, K., Bengtsson, J., & Elmqvist, T. (2009). Bumble bees (Bombus spp) along a gradient of increasing urbanization. PLoS One, 4, e5574, doi:10.1371/journal.pone.0005574

Aizen, M. A. & Feinsinger, P. (1994). Habitat fragmentation, native insect pollinators, and feral honeybees in Argentine ‘Chaco Serrano’. Ecological Applications, 4, 378-392, doi:10.2307/1941941

Aizen, M. A., Aguiar, A., Biesmeijer, J. C., Garibaldi, L. A., Inouye, D. W., Jung, C. et al. (2019). Global agricultural productivity is threatened by increasing pollinator dependence without a parallel increase in crop diversification. Global Change Biology, doi:10.1111/gcb.14736

Aizen, M. A., Garibaldi, L. A., Cunningham, S. A. & Klein, A. M. (2009). How much does agriculture depend on pollinators? Lessons from long-term trends in crop production. Annals of Botany, 103, 1579-1588, doi:10.1093/aob/mcp076

Aizen, M. A., Garibaldi, L.A., Cunningham, S. A. & Klein, A. M. (2008b). Long-term global trends in crop yield and production reveal no current pollination shortage but increasing pollinator dependency. Current Biology, 18, 1572-1575, doi:10.1016/j.cub.2008.08.066

Aizen, M. A., Morales, C. L., & Morales, J. M. (2008a). Invasive mutualists erode native pollination webs. PLoS Biology, 6, e31, doi:10.1371/journal.pbio.0060031

Alarcόn, R., Waser, N. M., & Ollerton, J. (2008). Year-to-year variation in the topology of a plant- pollinator interaction network. Oikos, 117, 1796-1807, doi:10.1111/j.0030-1299.2008.16987.x

Albrectsen, A. (2013). Sustainable development and population dynamics: placing people at the centre. Retrieved from https://www.unfpa.org/press/sustainable-development-and-population- dynamics-placing-people-centre

Almeida-Neto, M., Guimaraes, P., Guimaraes, P. R., Loyola, R. D. & Ulrich, W. (2008). A consistent metric for nestedness analysis in ecological systems: reconciling concept and measurement. Oikos, 117, 1239-2008, doi:10.1111/j.0030-1299.2008.16644.x

Alston, D. G., Tepedino, V. J., Bradley, B. A., Toler, T. R., Griswold, T. L., & Messinger, S. M. (2007). Effects of the insecticide phosmet on solitary bee foraging and nesting in orchards of Capitol Reef National Park, Utah. Environmental Entomology, 36, 811-816, doi:10.1603/0046- 225x(2007)36[811:eotipo]2.0.co;2

Arena, M., & Sgolastra, F. (2014). A meta-analysis comparing the sensitivity of bees to pesticides. Ecotoxicology, 23, 324-334, doi:10.1007/s10646-014-1190-1

Aronson, M. F. J., Lepczyk, C. A., Evans, K. L., Goddard, M. A., Lerman, S. B., MacIvor, J. S. et al. (2017). Biodiversity in the city: key challenges for urban green space management. Frontiers in Ecology and the Environment, 15, 189-196, doi:10.1002/fee.1480

Ascher, J. S., & Pickering, J. (2018). Discover Life bee species guide and world checklist (Hymenoptera: Apoidea: Anthophila). Retrieved from http://www.discoverlife.org/mp/20q?guide=Apoidea_species

Bailes, E. J., Deutsch, K. R., Bagi, J., Rondissone. L., Brown, M. J. F., & Lewis, O. T. (2018). First detection of bee viruses in hoverfly (syrphid) pollinators. Biology Letters, 14, 20180001, doi:10.1098/rsbl.2018.0001

Banaszak, J. & Ratynska, H. (2014). Local changes in communities of wild bees (Hymenoptera: Apoidea: Apiformes): 30 years later. Polish Journal of Entomology, 83, 325-351, doi:10.2478/pjen-2014-0025

Barthell, J. F., Frankie, G. W., & Thorp, R. W. (1998). Invader effects in a community of cavity nesting megachilid bees (Hymenoptera: Megachilidae). Environmental Entomology, 27, 240-247, doi:10.1093/ee/27.2.240

Bartomeus, I., Ascher, J. S., Gibbs, J., Danforth, B. N., Wagner, D. L., Hedtke, S. M. et al. (2013a). Historical changes in northeastern US bee pollinators related to shared ecological traits. Proceedings of the National Academy of Sciences USA, 110, 4656-4660, doi:10.1073/pnas.1218503110 84

Bartomeus, I., Ascher, J. S., Wagner, D., Danforth, B. N., Colla, S., Kornbluth, S. et al. (2011). Climate-associated phenological advances in bee pollinators and bee-pollinated plants. Proceedings of the National Academy of Sciences USA, 108, 20645-20649, doi:10.1073/pnas.1115559108

Bartomeus, I., Park, M. G., Gibbs, J., Danforth, B. N., Lakso, A. N., & Winfree, R. (2013b). Biodiversity ensures plant-pollinator phenological synchrony against climate change. Ecology Letters, 16, 1331-1338, doi:10.1111/ele.12170

Bartomeus, I., Potts, S. G., Steffan-Dewenter, I., Vaissière, B. E., Woyciechowski, M., Krewenka, K. M. et al. (2014). Contribution of insect pollinators to crop yield and quality varies with agricultural intensification. PeerJ, 2, e328, doi:10.7717/peerj.328

Bascompte, J., Jordano, P., Melian, C. J., & Olesen, J. M. (2003). The nested assembly of plant- animal mutualistic networks. Proceedings of the National Academy of Sciences USA, 100, 9383- 9387, doi:10.1073/pnas.1633576100

Baxter-Gilbert, J. H., Riley, J. L., Neufeld, C. J. H., Litzgus, J. D., & Lesbarrères, D. (2015). Road mortality potentially responsible for billions of pollinating insect deaths annually. Journal of Insect Conservation, 19, 1029–1035, doi:10.1007/s10841-015-9808-z

Benjamin, F. E., Reilly, J. R., & Winfree, R. (2014). Pollinator body size mediates the scale at which land use drives crop pollination services. Journal of Applied Ecology, 51, 440-449, doi:10.1111/1365-2664.12198

Biesmeijer, J., C., Roberts, S. P. M., Reemer, M., Ohlemueller, R., Edwards, M., Peeters, T. et al. (2006). Parallel declines in pollinators and insect-pollinated plants in Britain and the Netherlands. Science, 313, 351-354, doi:10.1126/science.1127863

Blacquière, T., Smagghe, G., van Gestel, C. A. M., & Mommaerts, V. (2012). Neonicotinoids in bees: a review on concentrations, side-effects and risk assessment. Ecotoxicology, 21, 973-992, doi:10.1007/s10646-012-0863-x

Bommarco, R., Biesmeijer, J. C., Meyer, B., Potts, S. G., Poyry, J., Roberts, S. P. M. et al. (2010). Dispersal capacity and diet breadth modify the response of wild bees to habitat loss. Proceedings of the Royal Society B: Biological Sciences, 277, 2075-2082, doi:10.1098/rspb.2009.2221

Breeze, T. D., Bailey, A. P., Balcombe, K. G., & Potts, S. G. (2011). Pollination services in the UK: how important are honeybees? Agriculture Ecosystems & Environment, 142, 137-143, doi:10.1016/j.agee.2011.03.020

Brosi, B. J., Daily, G. C., & Ehrlich, P. R. (2007). Bee community shifts with landscape context in a tropical countryside. Ecological Applications, 17, 418-430, doi:10.1890/06-0029

Brown, M. J. F., & Paxton, R. J. (2009). The conservation of bees: a global perspective. Apidologie, 40, 410-416, doi:10.1051/apido/2009019 85

Burgos, A., Ceva, H., Perazzo, R. P. J., Devoto, M., Medan, D., Zimmerman, M. et al. (2007). Why nestedness in mutualistic networks? Journal of Theoretical Biology, 249, 307-313, doi:10.1016/j.jtbi.2007.07.030

Burkle, L. A., & Alarcόn, R. (2011). The future of plant-pollinator diversity: understanding interaction networks across time, space, and global change. American Journal of Botany, 98, 528- 538, doi:10.3732/ajb.1000391

Burkle, L. A., Marlin, J. C., & Knight, T. M. (2013). Plant-pollinator interactions over 120 years: loss of species, co-occurrence, and function. Science, 339, 1611-1615, doi:10.1126/science.1232728

Cahill, J. F. Jr, Elle, E., Smith, G. R., & Shore, B. H. (2008). Disruption of a belowground mutualism alters interactions between plants and their floral visitors. Ecology, 89, 1791-1801, doi:10.1890/07-0719.1

Cameron, S. A., Lozier, J. D., Strange, J. P., Kock, J. B., Cordes, N., Solter, L. F. et al. (2011). Patterns of Widespread decline in North American bumblebees. Proceedings of the National Academy of Sciences USA, 108, 662-667, doi:10.1073/pnas.1014743108

Cane, J. H. (2001). Habitat fragmentation and native bees: a premature verdict? Conservation Ecology, 5, 3, doi:10.5751/ES-00265-050103

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

Cane, J. H., Minckley, R. L., Kervin, L., Roulston, T. H., Williams, N. (2006). Complex responses within a desert bee guild (Hymenoptera: Apiformes) to urban habitat fragmentation. Ecological Applications, 16, 632–644, doi:10.1890/10510761(2006)016[0632:CRWADB]2.0.CO;2

Carper, A. L., Adler, L. S., Warren, P. S., & Irwin, R. E. (2014). Effects of suburbanization on forest bee communities. Environmental Entomology, 43, 253-262, doi:10.1603/EN13078

Carvalheiro, L. G., Kunin, W. E., Keil, P., Aguirre-Gutiérrez, J., Ellis, W. N., Fox, R. et al. (2013). Species richness declines and biotic homogenization have slowed down for NW-European pollinators and plants. Ecology Letters, 16, 870-808, doi:10.1111/ele.12121

Carvell, C. (2002). Habitat use and conservation of bumble bees (Bombus spp.) under different grazing regimes. Biological Conservation, 103, 33-49, doi:10.1016/S0006-3207(01)00114-8

Champely, S. (2018). pwr: Basic functions for power analysis. Retrieved from https://cran.r- project.org/web/packages/pwr/index.html

Chiarucci, A., Bacaro, G., Rocchini, D., & Fattorini, L. (2008). Discovering and rediscovering the sample-based rarefaction formula in the ecological literature. Community Ecology, 9, 121-123, doi:10.1556/ComEc.9.2008.1.14

Colin, M. E., & Belzunces, L. P. (1992). Evidence of synergy between prochloraz and deltamethrin in Apis mellifera L.: a convenient biological approach. Pesticide Science, 36, 115–119, doi:10.1002/ps.2780360206

Colla, S. R., & Packer, L. (2008). Evidence for decline in eastern North American bumblebees (Hymenoptera: Apidae), with special focus on Bombus affinis Cresson. Biodiversity and Conservation, 17, 1379-1391, doi:10.1007/s10531-008-9340-5

Colla, S. R., & Taylor-Pindar, A. (2011). Recovery Strategy for the Rusty-patched Bumble Bee (Bombus affinis) in Ontario. Retrieved from http://files.ontario.ca/environment-and- energy/species-at-risk/stdprod_086037.pdf

Colla, S. R., Ascher, J. S., Arduser, M., Cane, J., Deyrup, M., Droege, S. et al. (2012a). Documenting Persistence of Most Eastern North American Bee Species (Hymenoptera: Apoidea: Anthophila) to 1990–2009. Journal of the Kansas Entomological Society, 85, 14-22, doi:10.2317/JKES110726.1

Colla, S. R., Gadallah, F., Richardson, L., Wagner, D., & Gall, L. (2012b). Assessing declines of North American bumble bees (Bombus spp.) using museum specimens. Biodiversity and Conservation, 21, 3585-3595, doi:10.1007/s10531-012-0383-2

Colla, S. R., Otterstatter, M. C., Gegear, R. J. & Thomson, J. D. (2006). Plight of the bumble bee: pathogen spillover from commercial to wild populations. Biological Conservation, 129, 461-467, doi:10.1016/j.biocon.2005.11.013

Colla, S. R., Willis, E., & Packer, L. (2009). Can green roofs provide habitat for urban bees (Hymenoptera: Apidae)? Cities and the Environment, 2, 1-12.

Colwell, R. K. (2013). EstimateS: Statistical estimation of species richness and shared species from samples, version 9. Retrieved from http://purl.oclc.org/estimates

Colwell, R. K., Chao, A., Gotelli, N. J., Lin, S., Mao, C. X., Chazdon, R. L. et al. (2012). Models and estimates linking individual-based and sample-based rarefaction, extrapolation and comparison of assemblages. Journal of Plant Ecology, 5, 3-21, doi:10.1093/jpe/rtr044

Costanza, R., d'Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B. et al. (1997). The value of the world's ecosystem services and natural capital. Nature, 387, 253-260, doi:10.1038/387253a0

Currie, D. J., Mittelbach, G. G., Cornell, H. V., Field, R., Guegan, J. F., Hawkins, B. A. et al. (2004). Predictions and tests of climate-based hypotheses of broad-scale variation in taxonomic richness. Ecology Letters, 7, 1121-1134, doi:10.1111/j.1461-0248.2004.00671.x 87

Cusser, S., & Goodell, K. (2013). Diversity and distribution of floral resources influence the restoration of plant-pollinator networks on a reclaimed strip mine. Restoration Ecology, 21, 713- 721, doi:10.1111/rec.12003

Cutler, G. C., Nams, V. O., Craig, P., Sproule, J. M., Sheffield, & C. S. (2015). Wild bee pollinator communities of lowbush blueberry fields: spatial and temporal trends. Basic and Applied Ecology, 16, 73-85, doi:10.1016/j.baae.2014.11.005

Cutler, G. C., Scott-Dupree, C. D., & Drexler, D. M. (2014a). Honey bees, neonicotinoids and bee incident reports: the Canadian situation. Pest Management Science, 70, 779-783, doi:10.1002/ps.3613

Cutler, G. C., Scott-Dupree, C. D., Sultan, M., McFarlane, A. D., & Brewer, L. (2014b). A large- scale field study examining effects of exposure to clothianidin seed-treated canola on honey bee colony health, development, and overwintering success. PeerJ, 2, e652, doi:10.7717/peerj.652

Daily, G. C. (1997). Nature's Services: Societal dependence on natural ecosystems. Washington DC: Island Press.

Danforth, B. N., Sipes, S., Fang, J., & Brady, S. G. (2006). The history of early bee diversification based on five genes plus morphology. Proceedings of the National Academy of Sciences USA, 103, 15118–15123, doi:10.1073/pnas.0604033103

De la Rúa, P., Jaffe, R., Dall’Olio, R., Munoz, I., & Serrano, J. (2009). Biodiversity, conservation and current threats to European honeybees. Apidologie, 40, 263-284, doi:10.1051/apido/2009027

De la Rúa, P., Serrano, J., & Galian, J. (2002). Biodiversity of Apis mellifera populations from Tenerife (Canary Islands) and hybridisation with east European races. Biodiversity Conservation, 11, 59-67, doi:10.1023/A:1014066407307 de Snoo, G. R., & Van der Poll, R. J. (1999). Effect of herbicide drift on adjacent boundary vegetation. Agriculture, Ecosystems & Environment, 73, 1-6, doi:10.1016/S0167-8809(99)00008- 0

Desai, S. D., Kumar, S., & Currie, R. W. (2016). Occurrence, detection, and quantification of economically important viruses in healthy and unhealthy honey bee (Hymenoptera: Apidae) colonies in Canada. Canadian Entomologist, 148, 22–35, doi:10.4039/tce.2015.23

Desneux, N., Decourtye, A., & Delpuech, J. M. (2007). The sublethal effects of pesticides on beneficial arthropods. Annual Review of Entomology, 52, 81-106, doi:10.1146/annurev.ento.52.110405.091440

Devictor, V., Julliard, R., Couvet, D., & Jiguet, F. (2008). Birds are tracking climate warming, but not fast enough. Proceedings of the Royal Society B: Biological Sciences, 275, 2743-2748, doi:10.1098/rspb.2008.0878

Dickinson, R., & Royer, F. (2014). Plants of southern Ontario: trees, shrubs, wildflowers, grasses, ferns and aquatic plants. Vancouver BC: Lone Pine Publishing.

Dornelas, M., Antão, L. H., Moyes, F., Bates, A. E., Magurran, A. E., Dusan, A. (2018). BioTIME: a database of biodiversity time series for the Anthropocene. Global Ecology and Biogeography, 27, 760-786, doi:10.1111/geb.12729

Duelli, P. & Obrist, M. K. (2003). Regional biodiversity in an agricultural landscape: the contribution of seminatural habitat islands. Basic and Applied Ecology, 4, 129-138, doi:10.1078/1439-1791-00140

Dune, A., Williams, R. J. & Martinez, N. D. (2002). Network structure and biodiversity loss in food webs: robustness increases with connectance. Ecology Letters, 5, 558-567, doi:10.1046/j.1461-0248.2002.00354.x

Fausto, E., Milner, G., Nikolic, V., Briley, L., Basile, S., Behan, K. et al. (2015). Historical and Future Climate Trends in York Region. Retrieved from https://climateconnections.ca/app/uploads/2015/02/Historical-and-Future-Climate-Trends-in- York-Region_Report-1.pdf

Folke, C., Carpenter, S. R., Walker, B., Scheffer, M., Elmqvist, T., Gunderson, L. et al. (2004). Regime shifts, resilience and biodiversity in ecosystem management. Annual Review in Ecology, Evolution and Systematics, 35, 557–581, doi:10.1146/annurev.ecolsys.35.021103.105711

Forrest, J. R. K., & Thomson, J. D. (2011). An examination of synchrony between insect emergence and flowering in Rocky Mountain meadows. Ecological Monographs, 81, 469-491, doi:10.1890/10-1885.1

Forrest, J. R. K., Cross, R. & CaraDonna, P. J. (2019). Two-year bee, or not two-year bee? How voltinism is affected by temperature and season length in a high-elevation solitary bee. The American Naturalist, 193, 560–574, doi:10.1086/701826

Forrest, J., & Miller-Rushing, A. J. (2010). Toward a synthetic understanding of the role of phenology in ecology and evolution. Philosophical Transactions of the Royal Society B: Biological Sciences, 365, 3101-3112, doi:10.1098/rstb.2010.0145

Forup, M. L., & Memmott, J. (2005). The relationship between the abundances of bumblebees and honeybees in a native habitat. Ecological Entomology, 30, 47-57, doi:10.1111/j.0307- 6946.2005.00660.x

Frankie, G. M., Rizzard, M., Vinson, B., & Griswold, T. L. (2009). Decline in bee diversity and abundance from 1972-2004 on a flowering leguminous tree, Andira inermis in Costa Rica at the interface of disturbed dry forest and the urban environment. Journal of the Kansas Entomological Society, 82, 1-20, doi:10.2317/JKES708.23.1

Frund, J., Zieger, S. L., & Tscharntke, T. (2013). Response diversity of wild bees to overwintering temperatures. Oecologia, 173, 1639-1648, doi:10.1007/s00442-013-2729-1

Fürst, M. A., McMahon, D. P., Osborne, J. L., Paxton, R. J., & Brown, M. J. F. (2014). Disease associations between honeybees and bumblebees as a threat to wild pollinators. Nature, 506, 364- 366, doi:10.1038/nature12977

Gardner, J. D., & Spivak, M. (2014). A survey and historical comparison of the Megachilidae (Hymenoptera: Apoidea) of Itasca State Park, Minnesota. Annals of the Entomological Society of America, 107, 983–993, doi:10.1603/AN14023

Garibaldi, L. A., Steffan-Dewenter, I., Winfree, R., Aizen, M. A., Bommarco, R., Cunningham, S. A. et al. (2013). Wild pollinators enhance fruit set of crops regardless of honey bee abundance. Science, 339, 1608-1611, doi:10.1126/science.1230200

Garibaldi, L. A., Steffan-Dewenter, I., Kremen, C., Morales, J. M., Bommarco, R., Cunningham, S. A. et al. (2011). Stability of pollination services decreases with isolation from natural areas despite honey bee visits. Ecology Letters, 14, 1062–1072, doi:10.1111/j.1461-0248.2011.01669.x.

Gathmann, A., & Tscharntke, T. (2002). Foraging ranges of solitary bees. Journal of Animal Ecology, 71, 757-764, doi:10.1046/j.1365-2656.2002.00641.x

Geslin, B., Gauzens, B., Thebault, E., & Dajoz, I. (2013). Plant pollinator networks along a gradient of urbanisation. PLoS One, 8, e63421, doi:10.1371/journal.pone.0063421

Gezon, Z. J., Wyman, E. S., Ascher, J. S., Inouye, D. W., & Irwin, R. E. (2015). The effect of repeated, lethal sampling on wild bee abundance and diversity. Methods in Ecology and Evolution, 6, 1044-1054, doi:10.1111/2041-210X.12375

Gibbs, J. (2010). Revision of the metallic species of Lasioglossum (Dialictus) in Canada (Hymenoptera, Halictidae, Halictini). Zootaxa, 2591, 1-382, doi:10.11646/zootaxa.2591.1.1

Gibbs, J. (2011). Revision of the metallic Lasioglossum (Dialictus) of eastern North America (Hymenoptera: Halictidae: Halictini). Zootaxa, 3073, 1-216, doi:10.11646/zootaxa.2591.1.1

Gibbs, K. E., Mackey, R. L., & Currie, D. J. (2009). Human land use, agriculture, pesticides and losses of imperiled species. Diversity and Distributions, 15, 242-253, doi:10.1111/j.1472- 4642.2008.00543.x

Gilbert, A. (2009). Connectance indicates the robustness of food webs when subjected to species loss. Ecological Indicators, 9, 72-80, doi:10.1016/j.ecolind.2008.01.010

Gill, R. J. & Raine, N. E. (2014). Chronic impairment of bumblebee natural foraging behaviour induced by sublethal pesticide exposure. Functional Ecology, 28, 1459-1471, doi:10.1111/1365- 2435.12292

Gill, R. J., Ramos-Rodriguez, O., Raine, N. E. (2012). Combined pesticide exposure severely affects individual- and colony-level traits in bees. Nature, 491, 105-108, doi:10.1038/nature11585

Godfray, H. C. J., Blacquière, T., Field, L. M., Hails, R. S., Petrokofsky, G., Potts, S. G. et al. (2014). A restatement of the natural science evidence base concerning neonicotinoid insecticides and insect pollinators. Proceedings of the Royal Society B: Biological Sciences, 281, 20140558, doi:10.1098/rspb.2014.0558.

Godfray, H. C. J., Blacquière, T., Field, L. M., Hails, R. S., Potts, S. G., Raine, N. E. et al. (2015). A restatement of recent advances in the natural science evidence base concerning neonicotinoid insecticides and insect pollinators. Proceedings of the Royal Society B: Biological Sciences, 282, 20151821, doi:10.1098/rspb.2015.1821

Goldstein, P. Z., & Scott, V. L. (2015). Taxonomic and behavioral components of faunal comparisons over time: The bees of Boulder County past and present (Colorado, USA) (Hymenoptera: Anthophila). Proceedings of the Entomological Society of Washington, 117, 290- 346, doi:10.4289/0013-8797.117.3.290

Gordo, O., & Sanz, J. J. (2005). Phenology and climate change: a long-term study in a Mediterranean locality. Oecologia, 146, 484-495, doi:10.1007/s00442-005-0240-z

Gotelli, N. J., & Ellison, A. M. (2012). A Primer of Ecological Statistics (2nd edition). Sunderland, MA: Sinaur Associates.

Goulson, D. (2015). Neonicotinoids impact bumblebee colony fitness in the field; a reanalysis of the UK’s food & environment research agency 2012 experiment. Peer J, 3, e854, doi:10.7717/peerj.854

Goulson, D., Lye, G. C., & Darvill, B. (2008). Decline and conservation of bumble bees. Annual Review of Entomology, 53, 191-208, doi:10.1146/annurev.ento.53.103106.093454

Government of Canada. (2019). Historical Data. Retrieved from:http://climate.weather.gc.ca/historical_data/search_historic_data_e.html

Government of Ontario. (2014). Pollinator Health. Retrieved from https://www.ontario.ca/page/pollinator-health

Grab, H., Branstetter, M. G., Amon, N., Urban-Mead, K. R., Park, M. G., Gibbs, J. et al. (2019). Agriculturally dominated landscapes reduce bee phylogenetic diversity and pollination services. Science, 363, 282-284, doi:10.1126/science.aat6016

Graystock, P., Goulson, D., & Hughes, W. O. H. (2014). The relationship between managed bees and the prevalence of parasites in bumblebees. PeerJ, 2, e522, doi:10.7717/peerj.522m

Greenleaf, S. S., Williams, N. M., Winfree, R., & Kremen, C. (2007). Bee foraging ranges and their relationship to body size. Oecologia, 153, 589–596, doi:10.1007/s00442-007-0752-9 91

Griffin, S. R., Bruninga-Socolar, B., Kerr, M. A., Gibbs, J., & Winfree, R. (2017). Wild bee community change over a 26-year chronosequence of restored tallgrass prairie. Restoration Ecology, 25, 650-660, doi:10.1111/rec.12481

Grixti, J. C. (2004). The bee fauna of an old field site in southern Ontario, revisited after 34 years: a case for change (Master’s thesis). Retrieved from Dissertations and Theses @ York University.

Grixti, J. C., & Packer, L. (2006). Changes in the bee fauna (Hymenoptera: Apoidea) of an old field site in southern Ontario, revisited after 34 years. Canadian Entomologist, 138, 147-164, doi:10.4039/n05-034

Grixti, J. C., Wong, L. T., Cameron, S. A., & Favret, C. (2009). Decline of bumble bees (Bombus) in the North American Midwest. Biological Conservation, 142, 75-84, doi:10.1016/j.biocon.2008.09.027

Guimaraes, P. R. Jr., & Guimaraes, P. (2006). Improving the analyses of nestedness for large sets of matrices. Environmental Modelling and Software, 21, 1512-1513, doi:10.1016/j.envsoft.2006.04.002

Hallmann, C. A., Sorg, M., Jongejans, E., Siepel, H., Hofland, N., Schwan, H. et al. (2017). More than 75 percent decline over 27 years in total flying insect biomass in protected areas. PLoS One, 12, e0185809, doi:10.1371/journal.pone.0185809

Hanski, I. (1982). Dynamics of regional distribution: the core and satellite species hypothesis. Oikos, 38, 210-221, doi:10.2307/3544021

Hatfield, R., Jepsen, S., Thorp, R., Richardson, L., Colla, S., Folts Jordan, S. et al. (2015a). Bombus affinis. The IUCN Red List of Threatened Species 2015. Retrieved from http://dx.doi.org/10.2305/IUCN.UK.2015-2.RLTS.T44937399A46440196.en.

Hatfield, R., Jepsen, S., Thorp, R., Richardson, L., Colla, S., Folts Jordan, S. et al. (2015b). Bombus terricola. The IUCN Red List of Threatened Species 2015. Retrieved from http://dx.doi.org/10.2305/IUCN.UK.2015-2.RLTS.T44937505A46440206.en.

Havens, K., & Vitt, P. (2016). The importance of phenological diversity in seed mixes for pollinator restoration. Natural Areas Journal, 36, 531-537, doi:10.3375/043.036.0419

Hegland, S. J., Nielsen, A., Lázaro, A., Bjerknes, A. L., & Totland, O. (2009). How does climate warming affect plant-pollinator interactions? Ecology Letters, 12, 184-195, doi:10.1111/j.1461- 0248.2008.01269.x

Heip, C. H. R., Peter, M. J., Soetaert, H., & Soetaert, K. (1998). Indices of diversity an evenness. Océanis, 24, 41-87.

Henderson, P.A. & Seaby, R. M. H. (2014). Community Analysis Package Version 5, Pisces Conservation Ltd, Lymington, UK. 92

Hillier, K., Frost, E., & Shutler, D. (2013). Fate of dermally applied miticides fluvalinate and amitraz within honey bee bodies. Journal of Economic Entomology, 106, 558-565, doi:10.1603/EC12300

Hilty, J., & Merenlender, A. M. (2003). Studying biodiversity of private lands. Conservation Biology, 17, 132-137, doi:10.1111/brv.12002

Hoehn, P., Tscharntke, T., Tylianakis, J. M. & Steffan-Dewenter, I. (2008). Functional group diversity of bee pollinators increases crop yield. Proceedings of the Royal Society B: Biological Sciences, 275, 2283-2291, doi:10.1098/rspb.2008.0405

Holzschuh, A., Steffan-Dewenter, I., & Tscharntke, T. (2008). Agricultural landscapes with organic crops support higher pollinator diversity. Oikos, 117, 354-361, doi:10.1111/j.2007.0030- 1299.16303.x

Horn, M. (2010). A comparison of pollinator biodiversity between green spaces, industrial areas and residential land-use zones in urban, southern Ontario (Master’s thesis). Retrieved from The Atrium at the University of Guelph.

Huey, R. B., Deutsch, C. A., Tewksbury, J. J., Vitt, L. J., Hertz, P. E., Alvarez Perez, H. J. et al. (2009). Why tropical forest lizards are vulnerable to climate warming. Proceedings of the Royal Society B: Biological Sciences, 276, 1939-1948, doi:10.1098/rspb.2008.1957

Hughes, B. B., Beas-Luna, R., Barner A. K., Brewitt, K., Brumbaugh, D. R., Cerny-Chipman, E. B. et al. (2017). Long-term studies contribute disproportionately to ecology and policy. BioScience, 67, 271-281, doi:10.1093/biosci/biw185

Hulot, F. D., Lacroix, G., Lescher-Moutoue, F. O. & Loreau, M. (2000). Functional diversity governs ecosystem response to nutrient enrichment. Nature, 405, 340-344, doi:10.1038/35012591

Hulsen, T., de Vlieg, J & Wynand, A. (2008). BioVenn – a web application for the comparison and visualization of biological lists using area-proportional Venn diagrams. BioMed Central: The Open Access Publisher. Retrieved from http://temoa.tec.mx/node/453436

Inouye, D. W., Saavedra, F., & Lee-Yang, W. (2003). Environmental influences on the phenology and abundance of flowering by Androsace septentrionalis (Primulaceae). American Journal of Botany, 90, 905-110, doi:10.3732/ajb.90.6.905

IPBES. (2016). Summary for policymakers of the assessment report of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services on pollinators, pollination and food production. Retrieved from https://www.actu-environnement.com/media/pdf/news-26331- synthese-ipbes-decideurs-pollinisateurs.pdf

Jensen, A. B., Palmer, K. A., Boomsma, J. J., & Pedersen, B. V. (2005). Varying degrees of Apis mellifera ligustica introgression in protected populations of the black honeybee, Apis mellifera

mellifera, in northwest Europe. Molecular Ecology, 14, 93-106, doi:10.1111/j.1365- 294X.2004.02399.x

Johnson, R. M., Dahlgren, L., Siegfried, B. D., & Ellis, M. D. (2013). Acaricide, fungicide and drug interactions in honey bees (Apis mellifera). PLoS One, 8, e54092, doi:10.1371/journal.pone.0054092

Kaiser, L. (2015). Nestedness and null models in ecology (Bachelor’s thesis). Retrieved from the University of Freiburg

Kanbe, Y., Okada, I., Yoneda, M., Goka, K., & Tsuchida, K. (2008). Interspecific mating of the introduced bumblebee Bombus terrestris and the native Japanese bumble Bombus hypocrita sapporoensis results in inviable hybrids. Naturwissenschaften, 95, 1003-1008, doi:10.1007/s00114-008-0415-7

Kearns, C. A., & Oliveras, D. M. (2009). Boulder County bees revisited: a resampling of Boulder Colorado bees a century later. Journal of Insect Conservation, 13, 603-613, doi:10.1007/s10841- 009-9211-8

Kerr, J. T., Pindar, A., Galpern, P., Packer, L., Potts, A., Roberts, S. M. et al. (2015) Climate change impacts on bumblebees converge across continents. Science, 349, 177-180, doi:10.1126/science.aaa7031

Kim, J., Williams, N., & Kremen, C. (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, 309-320, doi:10.2317/0507.11.1

Kleijn, D., & Snoeijing, G. (1997) Field boundary vegetation and the effects of agrochemical drift: botanical change caused by low levels of herbicide and fertilizer. Journal of Applied Ecology, 34, 1413-1425, doi:10.2307/2405258

Klein, A. M., Vaissière, B. E., Cane, J. H., Steffan-Dewenter, I., Cunningham, S. A., Kremen, C. et al. (2007). Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society of London Series B: Biological Sciences, 274, 303-313, doi:10.1098/rspb.2006.3721

Klein, A., Steffan-Dewenter, I. & Tscharntke, T. (2003a). Fruit set of highland coffee increases with the diversity of pollinating bees. Proceedings of the Royal Society of London Series B: Biological Sciences, 270, 955-961, doi:10.1098/rspb.2002.2306

Klein, A.-M., Steffan-Dewenter, I. & Tscharntke, T. (2003b). Pollination of Coffea canephora in relation to local and regional agroforest management. Journal of Applied Ecology, 40, 837-845, doi:10.1046/j.1365-2664.2003.00847.x

Koh, I., Lonsdorf, E. V., Williams, N. M., Brittain, C., Isaacs, R., Gibbs, J. et al. (2016). Modeling the status, trends, and impacts of wild bee abundance in the United States. Proceedings of the National Academy of Sciences USA, 113, 140-145, doi:10.1073/pnas.1517685113 94

Krauss, J., Alfert, T., & Steffan-Dewenter, I. (2009). Habitat area but not habitat age determines wild bee richness in limestone quarries. Journal of Applied Ecology, 46, 194-202, doi:10.1111/j.1365-2664.2008.01582.x

Kremen, C., & M'Gonigle, L.K. (2015). Small-scale restoration in intensive agricultural landscapes supports more specialized and less mobile pollinator species. Journal of Applied Ecology, 52, 602-610, doi:10.1111/1365-2664.12418

Kremen, C., Williams, N. M., Aizen, M. A., Gemmill-Herren, B., LeBuhn, G., Minckley, R. et al. (2007). Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecology Letters, 10, 299–314, doi:10.1111/j.1461- 0248.2007.01018.x

Kremen, C., Williams, N. M., Bugg, R. L., Fay, J. P., & Thorp, R. W. (2004). The area requirements of an ecosystem service: crop pollination by native bee communities in California. Ecology Letters, 7, 1109-1119, doi:10.1111/j.1461-0248.2004.00662.x

Kudo, G. (2014). Vulnerability of phenological synchrony between plants and pollinators in an alpine ecosystem. Ecological Research, 29, 571-581, doi:10.1007/s11284-013-1108-z

Kudo, G., & Ida, T. Y. (2013). Early onset of spring increases the phenological mismatch between plants and pollinators. Ecology, 94, 2311–2320, doi:10.1890/12-2003.1

Lavorel, S. & Garnier, E. (2002). Predicting changes in community composition and ecosystem functioning from plant traits:revisiting the Holy Grail. Functional Ecology, 16, 545-556, doi:10.1046/j.1365-2435.2002.00664.x

LeBuhn, G., Droege, S., Connor, E. F., Gemmill-Herren, B., Potts, S., Minckley, R. L. et al. (2012). Detecting insect pollinator declines on regional and global scales. Conservation Biology, 27, 113- 120, doi:10.1111/j.1523-1739.2012.01962.x.

LeBuhn, G., Droege, S., Connor, E., Gemmill-Herren, B., & Azzu, N. (2016). Protocol to detect and monitor pollinator communities. Retrieved from http://www.fao.org/3/a-i5367e.pdf

Lewinsohn, T.M., Prado, P. I., Jordano, P., Bascompte, J., & Olesen, J.M. (2006). Structure in plant‐animal interaction assemblages. Oikos, 113, 174–184, doi:10.1111/j.0030- 1299.2006.14583.x

Lopezaraiza‐Mikel, M. E., Hayes, R., Whalley, M., & Memmott, J. (2007). The impact of an alien plant on a native plant–pollinator network: an experimental approach. Ecology Letters, 10, 539– 550, doi:10.1111/j.1461-0248.2007.01055.x

Loreau, M., Naeem, S., Inchausti, P., Bengtsson, J., Grime, J. P., Hector, A. et al. (2011). Biodiversity and ecosystem functioning: current knowledge and future challenges. Science, 294, 804-808, doi:10.1126/science.1064088

Lundin, O., Rundlöf, M., Smith, H. G., Fries, I., & Bommarco, R. (2015). Neonicotinoid insecticides and their impacts on bees: a systematic review of research approaches and identification of knowledge gaps. PLoS One, 10, e0136928, doi:10.1371/journal.pone.0136928

Mace, G. M., Norris, K. & Fitter, A. H. (2012). Biodiversity and ecosystems services: a multilayered relationship. Trends in Ecology & Evolution, 27, 19-26, doi:10.1016/j.tree.2011.08.006

MacIvor, J. S., & Packer, L. (2015). 'Bee hotels' as tools for native pollinator conservation: a premature verdict? PLoS One, 10, e0122126, doi:10.1371/journal.pone.0122126

MacIvor, S. J. & Lundholm, J. (2011). Insect species composition and diversity on intensive green roofs and adjacent level-ground habitats. Urban Ecosystems, 14, 225-241, doi:10.1007/s11252- 010-0149-0

MacKay, P. A. (1970). Observations on periodicity and ecological interactions of the apoidea and their parasite complexes in an Ontario biotope (Master’s thesis). Retrieved from the University of Toronto Archives.

MacKay, P. A., & Knerer, G. (1979). Seasonal occurrence and abundance in a community of wild bees from an old field habitat in Southern Ontario. Canadian Entomologist, 111, 367-376, doi:10.4039/Ent111367-3

Magurran, A. E. (1988). Ecological diversity and its measurement. London: Chapman and Hall.

Marini, L., Quaranta, M., Fontana, P., Biesmeijer, J. C., & Bommarco, R. (2012). Landscape context and elevation affect pollinator communities in intensive apple orchards. Basic and Applied Ecology, 13, 681-689, doi:10.1016/j.baae.2012.09.003

Marlin, J. C., & LaBerge, W. E. (2001). The native bee fauna of Carlinville, Illinois, revisited after 75 years: a case for persistence. Conservation Ecology, 5, 9.

Martins, A. C., Goncalves, R. B., & Melo, G. A. R. (2013). Changes in wild bee fauna of a grassland in Brazil reveal negative effects associated with growing urbanization during the last 40 years. Zoologie (Curitiba), 30, doi:10.1590/S1984-46702013000200006

Mathiasson, M. E., & Rehan, S. M. (2019). Status changes in the wild bees of north-eastern North America over 125 years revealed through museum specimens. Insect Conservation and Diversity, 12, 278-288, doi:10.1111/icad.12347

Matteson, K. C., & Langellotto, G. A. (2010). Determinates of inner city butterfly and bee species richness. Urban Ecosystems, 13, 333-347, doi:10.1007/s11252-010-0122-y

McDermid, J., Fera, S., & Hogg, A. (2015). Climate change projections for Ontario: An updated synthesis for policymakers and planners. Retrieved from http://www.climateontario.ca/MNR_Publications/CCRR-44.pdf 96

McKinney, M. L. (2008). Effects of urbanization on species richness: a review of plants and animals. Urban Ecosystems, 11, 161–176, doi:10.1007/s11252-007-0045-4

McLeod, K. (2013). Native bee (Hymenoptera: Apiformes) response to ecological restoration in southern Ontario (Master’s thesis). Retrieved from UWSpace Electronic Theses and Dissertations Collection at the University of Waterloo.

McMahon, D. P., Fürst, M. A., Caspar, J., Theodorou, P., Brown, M. J. F., & Paxton, R. J. (2015). A sting in the spit: widespread cross-infection of multiple RNA viruses across wild and managed bees. Journal of Animal Ecology, 84, 615-624, doi:10.1111/1365-2656.12345

Meiners, J. M., Griswold, T. L. & Carril, O. M. (2019). Decades of native bee biodiversity surveys at Pinnacles National Park highlight the importance of monitoring natural areas over time. PLoS One, 14, e0207566, doi:10.1371/journal.pone.0207566

Memmott, J., Carvell, C., Pywell, R. F., & Craze, P. G. (2010). The potential impact of global warming on the efficacy of field margins sown for the conservation of bumble bees. Philosophical Transactions of the Royal Society B: Biological Sciences, 365, 2071-2079, doi:10.1098/rstb.2010.0015

Memmott, J., Craze, P. G., Waser, N. M., & Price, M. V. (2007). Global warming and the disruption of plant-pollinator interactions. Ecology Letters, 10, 710-717, doi:10.1111/j.1461- 0248.2007.01061.x

Memmott, J., Waser, N. M., & Price, M. V. (2004). Tolerance of pollination networks to species extinctions. Proceedings of the Royal Society B: Biological Sciences, 271, 2605-2611, doi:10.1098/rspb.2004.2909

Menzel, A., Sparks, T. H., Estrella, N., Koch, E., Aasa, A., Ahas, R. et al. (2006). European phenological response to climate change matches the warming pattern. Global Change Biology, 12, 1969-1976, doi:10.1111/j.1365-2486.2006.01193.x

Menzel, P., & Fabian, P. (1999). Growing season extended in Europe. Nature, 397, 659, doi:10.1038/17709

Michener, C. D. (2004). Bees of the World. Baltimore, MD: The John Hopkins University Press.

Minckley, R. L., Roulston, T. H., & Williams, N. M. (2013). Resource assurance predicts specialist and generalist bee activity in drought. Proceedings of the Royal Society B: Biological Sciences, 280, 20122703. doi:10.1098/rspb.2012.2703

Ministry of Natural Resources and Forestry. (2016). Southern Ontario Land Resource Information System (SOLRIS) 2.0. Retrieved from https://www.javacoeapp.lrc.gov.on.ca/geonetwork/srv/en/main.home?uuid=635529ce-2639- 46f8-9fc2-43fbdd68aad1

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

Mitchell, T.B. (1962). Bees of the eastern United States II. Technical bulletin (North Carolina Agricultural Experiment Station), 152, 1-557.

Morandin, L. A., & Kremen, C. (2013a). Hedgerow restoration promotes pollinator populations and exports native bees to adjacent fields. Ecological Applications, 23, 829-839, doi:10.1890/12- 1051.1

Morandin, L. A., & Kremen, C. (2013b). Bee preference for native versus exotic plants in restored agricultural hedgerows. Restoration Ecology, 21, 26-32, doi:10.1111/j.1526-100X.2012.00876.x

Muma, W. (2012). Ontario Wildflowers. http://www.ontariowildflowers.com

Murray, T.E., Fitzpatrick, U., Byrne, A., Fealy, R., Brown, M. J. F., & Paxton, R. J. (2012). Local- scale factors structure wild bee communities in protected areas. Journal of Applied Ecology, 49, 998-1008, doi:10.1111/j.1365-2664.2012.02175.x

Nemesio, A., Santos, L. M., & Vasconcelos, H. L. (2015). Long-term ecology of orchid bees in an urban forest remnant. Apidologie, 46, 359-368, doi:10.1007/s13592-014-0328-8

Nielsen, A., & Bascompte, J. (2007). Ecological networks, nestedness, and sampling effort. Journal of Ecology, 95, 1134-1141, doi:10.1111/j.1365-2745.2007.01271.x

Nooten, S. S. & Rehan, S. M. (2019). Agricultural land use yields reduced foraging efficiency and unviable offspring in the wild bee Ceratina calcarata. Ecological Entomology, 44, 534-542, doi:10.1111/een.12730

Null, J. (2019). El Nino and La Nina years and intensities. Retrieved from https://ggweather.com/enso/oni.htm

Ollerton, J. (2017). Pollinator diversity: distribution, ecological function, and conservation. Annual Review of Ecology, Evolution and Systematics, 48, 353-376, doi:10.1146/annurev-ecolsys- 110316-022919

Ollerton, J., Erenler, H., Edwards, M., & Crockett, R. (2014). Extinctions of aculeate pollinators in Britain and the role of large-scale agricultural changes. Science, 346, 1360-1362, doi:10.1126/science.1257259

Ollerton, J., Johnson, S. D., Crammer, L., & Kellie, S. (2003). The pollination ecology of an assemblage of grassland asclepiads in South Africa. Annals of Botany, 92, 807–834, doi:10.1093/aob/mcg206

Ollerton, J., Winfree, R., & Tarrant, S. (2011). How many flowering plants are pollinated by animals? Oikos, 120, 321-326, doi:10.1111/j.1600-0706.2010.18644.x 98

Onuferko, T. (2014). Restoration and succession of a bee community in southern St. Catharines, Ontario, within a ten-year study period (Master’s thesis). Retrieved from Brock University Digital Repository.

Onuferko, T. M., Kutby, R., & Richards M. H. (2015). A list of bee species (Hymenoptera: Apoidea) recorded from three municipalities in the Niagara region of Ontario, including a new record of Lasioglossum furunculum Gibbs (Halictidae) in Canada. Journal of the Entomological Society of Ontario, 146, 3–22.

Onuferko, T. M., Skandalis, D. A., Cordero, R. L. & Richards, M. (2017). Rapid initial recovery and long-term persistence of a bee community in a former landfill. Insect Conservation and Diversity, 11, 88-99, doi:10.1111/icad.12261

OriginLab Corporation. (2019). Origin Pro Student Version. Retrieved from https://www.originlab.com/

Otterstatter, M. C., & Thomson, J. D. (2008). Does pathogen spillover from commercially reared bumble bees threaten wild pollinators? PLoS One, 3, e2771, doi:10.1371/journal.pone.0002771

Oyen, K. J., Giri, S., & Dillon, M. E. (2016). Altitudinal variation in bumble bee (Bombus) critical thermal limits. Journal of Thermal Biology, 59, 52-57, doi:10.1016/j.jtherbio.2016.04.015

Packer, L., Genaro, J. A., & Sheffield, C. S. (2007). The Bee Genera of Eastern Canada. Canadian Journal of Arthropod Identification, 3, doi:10.3752/cjai.2007.03

Packer, L., Monckton, S. K., Ferrari, R. R., & Onuferko, T. (2018). Validating taxonomic identifications in entomological research. Insect Conservation and Diversity, 11, 1-12, doi:10.1111/icad.12284

Parmesan, C. (2007). Influences of species, latitudes and methodologies on estimates of phenological response to global warming. Global Change Biology, 13, 1860-1872, doi:10.1111/j.1365-2486.2007.01404.x

Parmesan, C., & Yohe, G. (2003). A globally coherent fingerprint of climate change impacts across natural systems. Nature, 421, 37-42, doi:10.1038/nature01286

Parmesan, C., Duarte, C., Poloczanska, E., Richardson, A. J., & Singer, M. C. (2011). Overstretching attribution. Nature Climate Change, 1, 2-4, doi:10.1038/nclimate1056

Parmesan, C., Ryrholm, N., Stefanescu, C., Hill, J. K., Thomas, C. D., Descimon, H. et al. (1999). Poleward shifts in geographical ranges of butterfly species associated with regional warming. Nature, 399, 579-583, doi:10.1038/21181

Pellissier, L., Bråthen, K. A., Vittoz, P., Yoccoz, N. G., Dubuis, A., Meier, E. S. et al. (2013). Thermal niches are more conserved at cold than warm limits in arctic-alpine plant species. Global Ecology and Biogeography, 22, 933-941, doi:10.1111/geb.12057 99

Petanidou, T., & Ellis, W. N. (1993), Pollinating fauna of a Phryganic ecosystem: composition and diversity. Biodiversity Letters, 1, 9-22, doi:10.2307/2999643

Petanidou, T., & Ellis, W. N. (1996). Interdependence of native bee faunas and floras in changing Mediterranean communitites. Linnean Society Symposia Series, 18, 201-226.

Petanidou, T., Kallimanis, A. S., Sgardelis, S. P., Mazaris, A. D., Pantis, J. D., & Waser, N. M. (2014). Variable flowering phenology and pollinator use in a community suggest future phenological mismatch. Acta Oecologica, 59, 104-111, doi:10.1016/j.actao.2014.06.001

Pindar, A., Mullen, E. K., Tonge, M. B., Guzman-Novoa, E., & Raine, N. E. (2017). Status and Trends of Pollinator Health in Ontario. University of Guelph report prepared for Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA). 238 pages. Retrieved from https://rainelab.files.wordpress.com/2015/12/status-and-trends-of-pollinator-health-in-ontario- march-8-2017-tagged.pdf

Plascencia, M. & Philpott, S. M. (2017). Floral abundance, richness, and spatial distribution drive urban garden bee communities. Entomological Research, 107, 658-667, doi:10.1017/S0007485317000153

Pleasants, J. M. (1980). Competition for bumblebee pollinators in Rocky Mountain plant communities. Ecology, 61, 1446-1459, doi:10.2307/1939053

Polce, C., Garratt, M. P., Termansen, M., Ramirez-Villegas, J., Challinor, A. J., Lappage, M. G. et al. (2014). Climate-driven spatial mismatches between British orchards and their pollinators: increased risks of pollination deficits. Global Change Biology, 20, 2815-2828, doi:10.1111/gcb.12577

Popic, T. J., Davila, Y. C., & Wardle, G. M. (2013). Evaluation of common methods for sampling invertebrate pollinator assemblages: net sampling out-perform pan traps. PLoS One, 8, e66665, doi:10.1371/journal.pone.0066665

Potts, S. G., Biesmeijer, J. C., Kremen, C., Neumann, P., Schweiger, O., & Kunin, W. E. (2010). Global pollinator declines: trends, impacts and drivers. Trends in Ecology & Evolution, 25, 345- 353, doi:10.1016/j.tree.2010.01.007

Potts, S. G., Vulliamy, B., Dafni, A., Ne’eman, G., & Willmer, P. (2003). Linking bees and flowers: how do floral communities structure bee communities. Ecology, 84, 2628-2642, doi:10.1890/02-0136

Powney, G. D., Carvell, C., Edwards, M., Morris, R. K. A., Roy, H. E, Woodcock, B. A. et al. (2019). Widespread losses of pollinating insects in Britain. Nature Communications, 10, 1018, doi:10.1038/s41467-019-08974-9

100

Prather, C. M., Pelini, S. L., Laws, A., Rivest, E., Woltz, M., Bloch, C.P. et al. (2013). Invertebrates, ecosystem services and climate change. Biological Reviews, 88, 327-348, doi: 10.1111/brv.12002

Puric-Mladenovic, D., & Strobi, S. (2006). Delineating conservation areas on the Oak Ridges Moraine using a systematic conservation planning approach. The Forestry Chronicle, 82, 395-402, doi:10.5558/tfc82395-3

R Core Team. (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Raine, N. E. (2018). Pesticide affects social behaviour of bees. Science, 362, 643-644, doi:10.1126/science.aav5273

Raine, N. E., & Gill, R. J. (2015). Tasteless pesticides affect bees in the field. Nature, 521, 38-40, doi:10.1038/nature14391

Rathcke, B. J., & Jules, E. S. (1993). Habitat fragmentation and plant-pollinator interactions. Current Science, 65, 273-277.

Rehan, S. M. & Sheffield, C. S. (2011). Morphological and molecular delineation of a new species in the Ceratina dupla species-group (Hymenoptera: Apidae: Xylocopinae) of eastern North America. Zootaxa, 2873, 35-50, doi:10.11646/zootaxa.2873.1.3

Rhodes, M., Fant, J., & Skogen, K. (2017). Pollinator identity and spatial isolation influence multiple paternity in an annual plant. Molecular Ecology, 26, 4296–4308, doi:10.1111/mec.14115

Richards, M. H., Onuferko, T. M., & Rehan, S. M. (2015). Phenological, but not social, variation associated with climate differences in a eusocial sweat bee, Halictus ligatus, nesting in southern Ontario. Journal of Hymenoptera Research, 43, 19-44, doi:10.3897/JHR.43.8756

Richards, M. H., Rutgers-Kelly, A., Gibbs, J., Vickruck, J. L., Rehan, S. M, & Sheffield, C. S. (2011). Bee diversity in naturalizing patches of Carolinian grasslands in southern Ontario, Canada. Canadian Entomologist, 143, 279-299, doi:10.4039/n11-010

Ricketts, T. H. (2004). Tropical forest fragments enhance pollinator activity in nearby coffee crops. Conservation Biology, 18, 1262-1271, doi:10.1111/j.1523-1739.2004.00227.x

Ricketts, T. H., Regetz, J., Steffan‐Dewenter, I., Cunningham, S. A., Kremen, C., Bogdanski, A. et al. (2008). Landscape effects on crop pollination services: are there general patterns? Ecology Letters, 11, 499-515, doi:10.1111/j.1461-0248.2008.01157.x.

Roberts, H. P., King, D. I. & Milam, J. (2017). Factors affecting bee communities in forest openings and adjacent mature forest. Forest Ecology and Management, 394, 111-122, doi:10.1016/j.foreco.2017.03.027

101

Rockström, J., Steffen, W., Noone, K., Persson, Å., Chapin, F. S., Lambin, E. et al. (2009). Planetary boundaries: exploring the safe operating space for humanity. Ecology and Society, 14, 32, doi:10.5751/ES-03180-140232

Romankova, T. G. (2007). Bees of the genus Hylaeus of Ontario (Hymenoptera: Apoidea: Colletidae). Journal of the Entomological Society of Ontario, 138, 137-154.

Roubik, D. W. & Ackerman, J. D. (1987). Long-term ecology of euglossine orchid-bees (Apidae: Euglossini) in Panama. Oecologia, 73, 321-333, doi:10.1007/BF00385247.

Roulston, T. H., Smith, A. S., & Brewster, A. L. (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, 179-181, doi:10.2317/0022- 8567(2007)80[179:ACOPTA]2.0.CO;2

Roy, D. B., & Sparks, T. H. (2000). Phenology of British butterflies and climate change. Global Change Biology, 6, 407-416, doi:10.1046/j.1365-2486.2000.00322.x

Rundlöf, M., Andersson, G. K. S., Bommarco, R., Fries, I., Hederström, V., Herbertsson, L. et al. (2015). Seed coating with a neonicotinoid insecticide negatively affects wild bees. Nature, 521, 77–80, doi:10.1038/nature14420

Russell, C., & Schultz, C. B. (2009). Effects of grass-specific herbicides on butterflies: an experimental investigation to advance conservation efforts. Journal of Insect Conservation, 14, 53-63, doi:10.1007/s10841-009-9224-3

Russell, K. N., Ikerd, H., & Droege, S. (2005). The potential conservation value of unmowed powerline strips for native bees. Biological Conservation, 124, 133-148, doi:10.1016/j.biocon.2005.01.022

Rutgers-Kelly, A. C., & Richards, M. H. (2013). Effect of meadow regeneration on bee (Hymenoptera: Apoidea) abundance and diversity in southern Ontario, Canada. Canadian Entomologist, 145, 655-667, doi:10.4039/tce.2013.42

Samnegard, U., Persson, A. S., & Smith, H. G. (2011). Garden benefit bees and enhance pollination in intensively managed farmland. Biological Conservation, 144, 2602-2606, doi:10.1016/j.biocon.2011.07.008

Sanchez-Bayo, F. & Wyckhuys, K. (2019). Worldwide decline of the entomofauna: a review of its drivers. Biological Conservation, 232, 8-27, doi:10.1016/j.biocon.2019.01.020

Santamaria, L., & Rodriguez-Gironés, M. A. (2007). Linkage rules for plant-pollinator networks: trait complementarity or exploitation barriers. PLoS Biology, 5, e31, doi:10.1371/journal.pbio.0050031

102

Schenk, M., Kraus, J., & Holzschuh, A. (2017). Desynchronizations in bee-plant interactions cause severe fitness losses in solitary bees. Journal of Animal Ecology, 87, 139-149, doi:10.1111/1365- 2656.12694

Seaby, R. M., & Henderson, P. A. (2006). Species Diversity and Richness Version 4. Pisces Conservation Ltd., Lymington, England.

Senapathi, D., Goddard, M. A., Kunin, W. E. & Baldock, K. C. R. (2017). Landscape impacts on pollinator communities in temperate systems: evidence and knowledge gaps. Functional Ecology, 31, 26–37, doi:10.1111/1365-2435.12809

Sheffield, C. S., Dumesh, S., & Cheryomina, M. (2011). Hylaeus punctatus (Hymenoptera: Colletidae), a bee species new to Canada, with notes on other non-native species. Journal of Entomological Society of Ontario, 142, 29-43.

Sheffield, C., Pindar, A., Packer, L., & Kevan, P. G. (2013). The potential of cleptoparasitic bees as indicator taxa for assessing bee communities. Apidologie, 44, 501-510, doi:10.1007/s13592- 013-0200-2

Simberloff, D., & Dayan, T. (1991). The guild concept and the structure of ecological communities. Annual Review of Ecology and Systematics, 22, 115-143, doi:10.1146/annurev.es.22.110191.000555

Sint, D., & Traugott, M. (2015). Food Web Designer: a flexible tool to visualize interaction networks. Journal of Pest Science, 89, 1-5, doi:10.1007/s10340-015-0686-7

Sirois-Delisle, C., & Kerr, J. T. (2018). Climate change-driven range losses among bumblebee species are poised to accelerate. Scientific Reports, 8, 14464, doi:10.1038/s41598-018-32665-y

Skandalis, D. A., Richards, M. H., Sformo, T. S., & Tattersall, G. J. (2011). Climate limitations on the distribution and phenology of a large carpenter bee, Xylocopa virginica (Hymenoptera: Apidae). Canadian Journal of Zoology, 89, 785-795, doi:10.1139/z11-051

Song, C., Rohr, R. P., & Saavedra, S. (2017). Why are some plant-pollinator networks more nested than others? Journal of Animal Ecology, 86, 1417–1424, doi:10.1111/1365-2656.12749

Southwick, E. E., & Southwick, L. (1992). Estimating the economic value of honey bees (Hymenoptera, Apidae) as agricultural pollinators in the United States. Journal of Economic Entomology, 85, 621-633, doi:10.1093/jee/85.3.621

Stang, M., Klinkhamer, P. G. L., Meijden, E. V. D. (2006). Size constraints and flower abundance determine the number of interactions in a plant-flower visitor web. Oikos, 112, 111-121, doi:10.1111/j.0030-1299.2006.14199.x

103

Stanley, D. A., Garratt, M. P. D., Wickens, J. B., Wickens, V. J., Potts S.G.& Raine, N. E. (2015). Neonicotinoid pesticide exposure impairs crop pollination services provided by bumblebees. Nature, 528, 548-550, doi:10.1038/nature16167

Statistics Canada. (2006). Snapshot of Canadian Agriculture. Retrieved from http://www.statcan.gv.ca/ca-ra2006/articles/snapshot-portrait-eng.htm.

Statistics Canada. (2011). The Canadian population in 2011: population counts and growth. Retrieved from http://www12.statcan.ca/census-recensement/2011/as-sa/98-310-x/98-310- x2011001-eng.cfm

Statistics Canada. (2012). Number and area of census farms, Canada and the provinces, 1996, 2001, 2006 and 2011. Retrieved from http://www.omafra.gov.on.ca/english/stats/census/number.htm

Statistics Canada. (2014). Total farm cash receipts by province. Retrieved from http://www.omafra.gov.on.ca/english/stats/finance/crprov.htm

Steffan-Dewenter, I., & Tscharntke, T. (1999). Effects of habitat isolation on pollinator communities and seed set. Oecologia, 121, 432-440.

Steffan-Dewenter, I., Muzenberg, U., Burger, C., Thies, C., & Tscharntke, T. (2002). Scale dependent effects of landscape context on three pollinator guilds. Ecology, 83, 1421-1432, doi:10.1890/0012-9658(2002)083[1421:SDEOLC]2.0.CO;2

Stout, J. C., & Morales, C. L. (2009). Ecological impacts of invasive alien species on bees. Apidologie, 40, 388-409, doi:10.1051/apido/2009023

Strona, G., Galli, P., Seveso, D., Montano, S., & Fattorini, S. (2014a). Nestedness for Dummies (NeD): A User-Friendly Web Interface for Exploratory Nestedness Analysis. Journal of Statistical Software, 59, 1-9, doi:10.18637/jss.v059.c03

Strona, G., Nappo, D., Boccacci, F., Fattorini, S., & San-Miguel-Ayanz, J. (2014b). A fast and unbiased procedure to randomize ecological binary matrices with fixed row and column totals. Nature Communications, 5, 4114, doi:10.1038/ncomms5114

Sydenham, M. A. K., Eldegard, K., & Totland, O. (2014). Spatio-temporal variation in species assemblages in field edges: seasonally distinct responses of solitary bees to local habitat characteristics and landscape conditions. Biodiversity and Conservation, 37, 1-14, doi:10.1007/s10531-014-0729-z

Szabo, N. D., Colla, S. R., Wagner, D. L., Gall, L. F., & Kerr, J. T. (2012). Do pathogen spillover, pesticide use, or habitat loss explain recent North American bumblebee declines? Conservation Letters, 5, 232-239, doi:10.1111/j.1755-263X.2012.00234.x

104

Taki. H., Okochi, I., Okabe, K., Inoue, T., Goto, H., Matsumura, T. et al. (2013). Succession influences wild bees in a temperate forest landscape: the value of early successional stages in naturally regenerated and planted forests. PLoS One, 8, e56678, doi:10.1371/journal.pone.0056678

Tepedino, V. J. (1979). The importance of bees and other insect pollinators in maintaining floral species composition. Great Basin Naturalist Memoirs, 3, 139–150.

Thompson, H. M., Levine, S. L., Doering, J., Norman, S., Manson, P., Sutton, P. et al. (2014). Evaluating exposure and potential effects on honeybee brood (Apis mellifera) development using glyphosate as an example. Integrated Environmental Assessment and Management, 10, 463-470, doi:10.1002/ieam.1529.

Thompson, H., & Wilkins, S. (2003). Assessment of the synergy and repellency of pyrethroid/fungicide mixtures. Bulletin of Insectology, 56, 131-134.

Thompson, J. N. (2005). The geographic mosaic of coevolution. Chicago, IL: University of Chicago Press.

Tsvetkov, N., Samson-Robert, O., Sood, K., Patel, H. S., Malena, D. A., Gajiwala, P. H. et al. (2017). Chronic exposure to neonicotinoids reduces honeybee health near corn crops. Science, 356, 1395-1397, doi:10.1126/science.aam7470.

United Nations, Department of Economic and Social Affairs, Population Division (2019). World Urbanization Prospects: The 2018 Revision (ST/ESA/SER.A/420). New York: United Nations.

Vanbergen, A. J., Baude, M., Biesmeijer, J. C., Britton, N. F., Brown, M. J. F., Brown, M. et al. (2013). Threats to an ecosystem service: pressures on pollinators. Frontiers in Ecology and the Environment, 11, 251-259, doi:10.1890/120126

Vanbergen, A. J., Espíndola, A., & Aizen, M. A. (2018). Risks to pollinators and pollination from invasive alien species. Nature, Ecology & Evolution, 2, 16-25, doi:10.1038/s41559-017-0412-3

Vázquez, D.P. and Aizen, M.A. (2004) Asymmetric specialisation: a pervasive feature of plant- pollinator interactions. Ecology, 85, 1251-1257, doi:10.1890/03-3112

Vickruck, J. L. & Richards, M. H. (2017). Nestmate discrimination based on familiarity but not relatedness in eastern carpenter bees. Behavioural Processes, 145, 73-80, doi:10.1016/j.beproc.2017.10.005

Villalobos, S., & Vamosi, J. C. (2018). Climate and habitat influences on bee community structure in Western Canada. Canadian Journal of Zoology, 96, 1002-1009, doi:10.1139/cjz-2017-0226

Visser, Z. (2013). Farming bees in a dynamic social-ecology: An ethnographic exploration of knowledge practices among commercial bee farmers in the Western Cape (Master’s thesis). Retrieved from OpenUCT at the University of Cape Town. 105

Walters, K. (2013). Data, data everywhere but we don't know what to think? Neonicotinoid insecticides and pollinators. Outlooks on Pest Management, 24, 151-155, doi:10.1564/v24_aug_02

Walther, G., Post, E., Convey, P., Menzels, A., Parmesan, C., & Beebee, T. J. (2002). Ecological responses to recent climate change. Nature, 416, 389-395, doi:10.1038/416389a

Waser, N. M., & Ollerton, J. (2006). From specialization to generalization. Chicago: The University of Chicago Press.

Wcislo, W. T. (1981). The roles of seasonality, host synchrony, and behaviour in the evolutions and distributions of nest parasites in Hymenoptera (Insecta), with special reference to bees (Apoidea). Biological Reviews, 62, 515-543.

Westerkamp, C. (1991). Honeybees are poor pollinators: why? Plant Systematics and Evolution, 177, 71-75, doi:10.1007/BF00937827

Whitehorn, P. R., O’Connor, A., Wackers, F. L., & Goulson, D. (2012). Neonicotinoid pesticide reduced bumblebee colony growth and queen production. Science, 336, 351-352, doi:10.1126/science.1215025.

Williams, G. R., Shutler, D., & Rogers, R. E. L. (2010). Effects at nearctic north-temperate latitudes of indoor versus outdoor overwintering on the microsporidium Nosema ceranae and western honey bees (Apis mellifera). Journal of Invertebrate Pathology, 104, 4-7, doi:10.1016/j.jip.2010.01.009

Williams, N. M., & Kremen, C. (2007). Resource distribution among habitats determines solitary bee offspring production in a mosaic landscape. Ecological Applications, 17, 910–921, doi:10.1890/06-0269

Williams, P. H., & Osborne, J. L. (2009). Bumblebee vulnerability and conservation world-wide. Apidologie, 40, 367-387, doi:10.1051/apido/2009025

Williams, P. H., Thorp, R. W., Richardson L. & Colla, S. (2014). Bumble Bees of North America: an Identification Guide. Princeton University Press.

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

Willis Chan, D. S. & Kevan, P. (1995). Foraging dynamics of Peponapis pruinosa (Hymenoptera: Anthophoridae) on pumpkin (Cucurbita pepo) in southern Ontario. Canadian Entomologist, 127, 167-175, doi:10.4039/Ent127167-2

Willmer, P. (2014). Climate change: bees and orchids lose touch. Current Biology, 24, R1133- R1135, doi:10.1016/j.cub.2014.10.061

106

Wilson, J. S., & Carril, O. M. (2016). The bees in your backyard: a guide to North America’s bees. Princeton, NJ: Princeton University Press.

Wilson, R. J., Gutierrez, D., Gutierrez, J., Martinez, D., Agudo, R., & Monserrat, V. J. (2005). Changes to the elevational limits and extent of species ranges associated with climate change. Ecology Letters, 8, 1138-1146, doi:10.1111/j.1461-0248.2005.00824.x.

Winfree, R., Aguilar, R., Vázquez, D. P., LeBuhn, G., & Aizen, M. A. (2009). A meta-analysis of bees' responses to anthropogenic disturbance. Ecology, 90, 2068-2076, doi:10.1890/08-1245.1

Wray, J. C., Neame, L. A., & Elle, E. (2014). Floral resources, body size, and surrounding landscape influence bee community assemblages in oak-savannah fragments. Ecological Entomology, 39, 83-93, doi:10.1111/een.12070

Zayed, A. (2009). Bee genetics and conservation. Apidologie, 40, 237-262, doi:10.1051/apido/2009026

Zurbuchen, A., Landert, L., Klaiber, J., Müller, A., Hein, S. & Dorn, S. (2009). Maximum foraging ranges in solitary bees: only a few individuals have the capability to cover long foraging distances. Biological Conservation, 143, 669–-676, doi:10.1016/j.biocon.2009.12.003

107

APPENDICES

Appendix I

Table 9: Complete list of wild bee species, collected by pan traps and sweep net in the years 1968, 1969, 2002, 2003, 2016 and 2017 at the site. Functional guild classification for each species is listed. The resources used for identification are also listed. The ‘*’ symbol represents specimens that are thought to have been misidentified. They appear to be missing from the collections and I was unable to verify their identification. The ‘**’ symbol represents specimens which need further verification through expert opinion and/or barcoding. Family / Genus Subgenus / Species Functional Number of specimens collected per year Species level Guild identification Classification 1968 1969 2002 2003 2016 2017 Total

Andrenidae Ascher & Pickering, 2018 Mitchell 1960, Andrena Fabricius, algida Smith, 1853 solitary 2 2 1962 1775 ground nester

alleghaniensis Viereck, 1907 solitary 14 14 ground nester

arabis Robertson, 1897 solitary 2 2 ground nester

asteris Robertson, 1891 solitary 15 8 1 24 ground nester

barbilabris Kirby, 1802 solitary 2 2 ground nester

108

bisalicis Viereck, 1908 solitary 1 1 1 3 ground nester bradleyi Viereck, 1907 solitary 1 4 5 ground nester brevipalpis Cockerell, 1930 solitary 3 3 ground nester canadensis Dalla Torre, 1896 solitary 25 5 30 ground nester carlini Cockerell, 1901 solitary 2 15 25 6 8 56 ground nester ceanothi Viereck, 1917 solitary 4 4 ground nester commoda Smith, 1879 solitary 1 14 15 ground nester crataegi Robertson, 1893 solitary 8 1 8 41 18 27 103 ground nester cressonii Robertson, 1891 solitary 1 2 10 12 2 27 ground nester distans Provancher, 1888 solitary 1 1 ground nester dunningi Cockerell, 1898 solitary 1 1 3 5 ground nester erigeniae Robertson, 1891 solitary 1 1 ground nester erythrogaster Ashmead, 1890 solitary 5 20 29 54 ground nester 109

erythronii Robertson, 1891 solitary 2 23 25 ground nester

forbesii Robertson, 1891 solitary 2 2 ground nester

fragilis Smith, 1853 solitary 70 101 1 172 ground nester

fulgida LaBerge, 1980 solitary 4 4 ground nester

geranii Robertson, 1891 solitary 1 1 ground nester

hippotes Robertson, 1895 solitary 14 14 ground nester

hirticincta Provancher, 1888 solitary 1 19 20 3 3 46 ground nester

imitatrix Cresson, 1872 solitary 1 1 1 12 1 1 17 ground nester

integra Smith, 1853 solitary 9 2 257 35 2 305 ground nester

krigiana Robertson, 1901 solitary 3 3 ground nester macoupinensis ** Robertson, solitary 1 1 1900 ground nester

110

mandibularis Robertson, 1892 solitary 1 1 ground nester

mariae Robertson, 1891 solitary 11 11 ground nester

melanochroa Cockerell, 1898 solitary 3 1 10 1 3 18 ground nester

milwaukeensis Graenicher, 1903 solitary 5 5 ground nester

miranda Smith, 1879 solitary 1 2 1 1 5 ground nester miserabilis Cresson, 1872 solitary 3 34 8 37 82 ground nester morrisonella Viereck, 1917 solitary 3 3 ground nester morrisonella ** Viereck, 1917/ solitary 1 1 wheeleri ** Graenicher 1904 ground nester

nasonii Robertson, 1895 solitary 1 54 139 58 75 327 ground nester nasonii ** Robertson, 1895/ solitary 13 13 wheeleri ** Graenicher 1904 ground nester nivalis Smith, 1853 solitary 3 3 ground nester nubecula Smith, 1853 solitary 5 1 6 ground nester peckhami Cockerell, 1902 solitary 1 1 ground nester 111

persimulata Viereck, 1917 solitary 1 1 2 ground nester placata Mitchell, 1960 solitary 1 1 ground nester robertsonii Dalla Torre, 1896 solitary 7 2 1 1 2 13 ground nester runcinatae Cockerell, 1906 solitary 2 2 ground nester rufosignata Cockerell, 1902 solitary 3 2 5 ground nester rugosa Robertson, 1891 solitary 90 6 3 99 ground nester salictaria Robertson, 1905 solitary 6 6 12 ground nester sigmundi Cockerell, 1902 solitary 20 20 ground nester simplex Smith, 1853 solitary 3 1 1 5 ground nester species A ** solitary 1 1 Identified by ground nester Grixti species C ** solitary 3 3 Identified by ground nester Grixti spiraeana Robertson, 1895 solitary 3 1 4 Ascher & ground nester Pickering, 2018

112

thaspii Graenicher, 1903 solitary 3 3 Mitchell 1960, ground nester 1962

vicina Smith, 1853 solitary 1 6 1 8 ground nester

wheeleri Graenicher, 1904 solitary 9 1 8 18 ground nester

wheeleri ** Graenicher, 1904/ solitary 7 2 9 fulgida ** LaBerge, 1980 ground nester

wheeleri * Graenicher, 1904/ solitary 1 1 fulgida * LaBerge, 1980/ ground nester salictaria * Robertson, 1905 wheeleri ** Graeniher, 1904/ solitary 1 ziziae ** Robertson, 1891 ground nester

wilkella Kirby, 1802 solitary 13 765 396 7 1181 ground nester

w-scripta Viereck, 1904 solitary 7 4 11 ground nester

ziziae Robertson, 1891 solitary 1 2 3 ground nester

spp.** solitary 5 5 ground nester Calliopsis Smith, andreniformis Smith, 1853 solitary 107 107 Ascher & 1853 ground nester Pickering, 2018 Packer et al., 2007 Perdita Smith, 1853 halictoides Smith, 1853 solitary 1 3 4 Ascher & ground nester Pickering, 2018

octomaculata Say, 1824 solitary 121 95 216 ground nester 113

Pseudopanurgus rudbeckiae Robertson, 1895 solitary 2 2 6 25 35 Ascher & Cockerell, 1897 ground nester Pickering, 2018

Andrenidae TOTAL 319 156 1249 1081 171 210 3186

Apidae

Anthophora Latreille, bomboides Kirby, 1837 solitary 1 1 Ascher & 1803 ground nester Pickering, 2018

furcata Panzer, 1798 cavity nester 1 1 2

terminalis Cresson, 1869 cavity nester 9 11 7 27

Bombus Latreille, affinis Cresson, 1863 bumblebee Cory Sheffield 1802 Williams et al., 2014 bimaculatus Cresson, 1863 bumblebee 29 19 48

borealis Kirby, 1837 bumblebee 13 1 14

fervidus Fabricius, 1798 bumblebee

griseocollis DeGeer, 1773 bumblebee 12 11 23

impatiens Cresson, 1863 bumblebee 120 115 235

114

perplexus Cresson, 1863 bumblebee 2 1 3

rufocinctus Cresson, 1863 bumblebee 3 3

sandersoni Franklin, 1913 bumblebee 1 1

ternarius Say, 1837 bumblebee 5 5

terricola Kirby, 1837 bumblebee 2 2

vagans Smith, 1854 bumblebee 16 16

Ceratina Latreille, calcarata Robertson, 1900 cavity nester 3 283 511 62 185 1044 Rehan & 1802 Sheffield, 2011

dupla ** Say, 1837 cavity nester 11 13 36 305 12 35 412

mikmaqi ** Rehan and cavity nester 9 37 46 Sheffield, 2011

strenua Smith, 1879 cavity nester 1 1

Epeolus Latreille, autumnalis Robertson, 1902 cleptoparasite/ 1 1 Identified by 1802 social parasite Onuferko

bifasciatus Cresson, 1864 cleptoparasite/ 1 1 social parasite lectoides Robertson, 1901 cleptoparasite/ 5 5 social parasite 115

minimus Robertson, 1902 cleptoparasite/ 1 1 social parasite

pusillus Cresson, 1864 cleptoparasite/ 1 1 social parasite

scutellaris Say, 1824 cleptoparasite/ 1 2 8 4 15 social parasite

Holcopasites calliopsidis Linsley, 1943 cleptoparasite/ 1 4 5 Ascher & Ashmead, 1899 social parasite Pickering, 2018 Packer et al., 2007

Melissodes Latreille, agilis Cresson, 1878 solitary 2 2 Ascher & 1829 ground nester Pickering, 2018 Mitchell, 1960, apicata Lovell and Cockerell, solitary 4 4 1962 1906 ground nester

bidentis Cockerell, 1914 solitary 1 2 3 ground nester

denticulatus Smith, 1854 solitary 2 2 ground nester

desponsus Smith, 1854 solitary 22 5 44 21 92 ground nester

druriellus Kirby, 1802 solitary 42 112 6 4 164 ground nester

illatus Lovell and Cockerell, solitary 115 83 22 220 1906 ground nester

116

subillatus LaBerge, 1961 solitary 1 1 2 ground nester

trinodis Robertson, 1901 solitary 7 7 ground nester

wheeleri Cockerell, 1906 solitary 6 6 ground nester

spp. ** solitary 2 2 ground nester

Nomada Scopoli, articulata Smith, 1854 cleptoparasite/ 1 1 Ascher & 1770 social parasite Pickering, 2018 Mitchell 1960, bella Cresson, 1863 cleptoparasite/ 4 4 1962 social parasite

ceanothi Cockerell, 1907 cleptoparasite/ 7 7 social parasite

cressonii Robertson, 1893 cleptoparasite/ 4 3 6 12 25 social parasite

cuneata Robertson, 1903 cleptoparasite/ 2 3 5 social parasite

denticulata Robertson, 1902 cleptoparasite/ 1 1 social parasite

illinoensis Robertson, 1900 cleptoparasite/ 14 14 social parasite

imbricata Smith, 1854 cleptoparasite/ 2 2 social parasite inepta Mitchell, 1962 cleptoparasite/ 1 1 social parasite 117

lepida Cresson, 1863 cleptoparasite/ 6 6 social parasite luteoloides Robertson, 1895 cleptoparasite/ 4 8 12 social parasite maculata Cresson, 1863 cleptoparasite/ 1 4 14 12 31 social parasite obliterata Cresson, 1863 cleptoparasite/ 2 6 4 12 social parasite pygmaea Cresson, 1863 cleptoparasite/ 26 29 55 social parasite sayi Robertson, 1893 cleptoparasite/ 4 26 33 18 81 social parasite sobrina Mitchell, 1962 cleptoparasite/ 1 1 social parasite vicina Cresson, 1863 cleptoparasite/ 5 11 16 social parasite species A ** cleptoparasite/ 2 2 social parasite species B ** cleptoparasite/ 4 4 social parasite species C ** cleptoparasite/ 1 1 social parasite species D ** cleptoparasite/ 2 2 social parasite

118

species E ** cleptoparasite/ 3 3 social parasite

species F ** cleptoparasite/ 2 2 social parasite

species G ** cleptoparasite/ 2 2 social parasite

species H ** cleptoparasite/ 1 1 social parasite

spp. ** cleptoparasite/ 1 1 social parasite

Triepeolus pectoralis Robertson, 1897 cleptoparasite/ 20 4 4 28 Identified by Robertson, 1901 social parasite Onuferko

Xylocopa Latreille, virginica Linnaeus, 1771 cavity nester 1 1 2 Ascher & 1802 Pickering, 2018 Packer et al., 2007 Apidae TOTAL 27 53 609 1198 362 489 2738

Colletidae

Colletes Latreille, americanus Cresson, 1868 solitary 10 15 25 36 86 Ascher & 1802 ground nester Pickering, 2018 Mitchell 1960, compactus Cresson, 1868 solitary 8 15 52 67 6 148 1962 ground nester

impunctatus Nylander, 1852 solitary 1 1 ground nester

inaequalis Say, 1837 solitary 14 401 328 13 4 760 ground nester 119

kincaidii Cockerell, 1898 solitary 18 1 1 20 ground nester

latitarsus Robertson, 1891 solitary 1 1 ground nester

nudus Robertson, 1898 solitary 1 1 ground nester

simulans Swenk, 1868 solitary 1 1 12 25 2 41 ground nester

thoracicus Smith, 1853 solitary 3 8 11 ground nester

spp. ** solitary 1 1 ground nester

Hylaeus Fabricius, affinis Smith, 1853 cavity nester 185 4 1 31 1 222 Romankova, 2007 1793 Ascher & Pickering, 2018 annulatus Linnaeus, 1758 cavity nester 2 17 19 Mitchell 1960, 1962

basalis Smith, 1853 cavity nester 1 1

mesillae Cockerell, 1896 cavity nester 1 15 16

leptocephalus Morawitz, 1871 cavity nester 1 1

modestus Say, 1837 cavity nester 22 21 14 135 58 250

120

verticalis Cresson, 1869 cavity nester 1 1

Colletidae TOTAL 244 70 496 644 104 22 1580

Halictidae

Agapostemon Guérin- sericeus Forster, 1771 solitary 10 38 45 2 4 99 Ascher & Méneville, 1844 ground nester Pickering, 2018

texanus Cresson, 1872 solitary 3 3 ground nester

virescens Fabricius, 1775 solitary 1 4 5 ground nester

Augochlora Smith, pura Say, 1837 cavity nester 6 2 75 121 39 63 306 Ascher & 1853 Pickering, 2018 Packer et al., 2007 Augochlorella aurata Smith, 1853 social ground 14 14 53 212 10 7 310 Ascher & Sandhouse, 1937 nester Pickering, 2018 Packer et al., 2007 Augochloropsis metallica Fabricius, 1793 social ground 26 28 21 4 13 92 Ascher & Cockerell, 1897 nester Pickering, 2018 Packer et al., 2007 Dufourea Lepeletier, harveyi Cockerell, 1906 solitary 1 1 Ascher & 1841 ground nester Pickering, 2018 Packer et al., 2007 Halictus Latreille, confusus Smith, 1853 social ground 350 226 120 143 42 9 890 Ascher & 1804 nester Pickering, 2018

ligatus Say, 1837 social ground 64 88 493 450 58 8 1161 nester

121

rubicundus Christ, 1791 social ground 25 19 18 50 12 4 128 nester

Lasioglossum Curtis, abanci Crawford, 1932 solitary 2 3 5 Identified by 1833 (Dialictus ground nester Gibbs Robertson, 1902) Gibbs, 2010, 2011 admirandum Sandhouse, 1924 social ground 1 3 13 17 nester

anomalum Robertson, 1892 solitary 12 7 19 ground nester

atwoodi Gibbs, 2010 solitary 1 1 ground nester

cattellae Ellis, 1913 solitary 3 3 ground nester

cephalotes Dalla Torre, 1896 cleptoparasite/ 4 4 social parasite

coeruleum Robertson, 1893 cavity nester 2 3 9 4 1 19

creberrimum Smith, 1853 solitary 1 1 ground nester

cressonii Robertson, 1890 cavity nester 2 51 69 21 56 199

ephialtum Gibbs, 2010 solitary 1 1 2 ground nester

fattigi Mitchell, 1960 solitary 2 2 4 ground nester

122

foveolatum Robertson, 1902 solitary 3 2 5 ground nester heterognathum Mitchell, 1960 social ground 15 36 19 5 3 3 81 nester hitchensi Gibbs, 2012 solitary 1 1 2 2 6 ground nester imitatum Smith, 1853 social ground 3162 1894 15 81 9 1 5162 nester laevissimum Smith, 1853 social ground 4 4 27 131 53 23 242 nester leucocomum Lovell, 1908 solitary 1 1 ground nester lineatulum Crawford, 1906 social ground 488 372 157 304 47 18 1386 nester nigroviride Graenicher, 1911 social ground 1 1 14 7 7 30 nester oblongum Lovell, 1905 cavity nester 1 21 27 2 51

obscurum Robertson, 1892 solitary 1 1 ground nester oceanicum Cockerell, 1916 social ground 62 12 20 43 14 3 154 nester paradmirandum Knerer and social ground 1 1 Atwood, 1966 nester

123

perpunctatum Ellis, 1913 social ground 1 8 2 3 1 15 nester pilosum Smith, 1853 social ground 55 18 23 4 1 101 nester rohweri Ellis, 1915 social ground 1 1 54 56 nester smilacinae Robertson, 1897 social ground 11 11 nester subversans Mitchell, 1960 social ground 5 5 nester subviridatum Cockerell, 1938 cavity nester 1 1 2

taylorae Gibbs, 2010 solitary 1 1 ground nester tegulare Robertson, 1890 social ground 3 6 5 4 2 20 nester versans Lovell, 1905 social ground 5 8 45 5 4 67 nester versatum Robertson, 1902 social ground 1 35 34 70 nester viridatum Lovell, 1905 social ground 3 2 5 nester weemsi Mitchell, 1960 social ground 1 1 2 nester

124

zephyrum Smith, 1853 social ground 3 2 1 1 7 nester

L. (Evylaeus cinctipes Provancher, 1888 social ground 93 23 152 436 1 705 Identified by Robertson, 1902) nester Gibbs Gibbs et al., 2013 L. (Hemihalictus birkmanii Crawford, 1906 solitary 1 1 Identified by Cockerell, 1897) ground nester Gibbs Gibbs et al., 2013 foxii Robertson, 1895 solitary 7 4 3 8 4 3 29 ground nester

inconditum Cockerell, 1916 solitary 1 1 ground nester

macoupinense Robertson, 1895 solitary 2 3 5 ground nester

pectinatum Robertson, 1890 solitary 1 1 ground nester

pectorale Smith, 1853 solitary 29 23 8 8 1 69 ground nester

L. (Lasioglossum) coriaceum Smith, 1853 solitary 14 7 16 3 5 45 Identified by ground nester Gibbs Gibbs, 2010, 2011 forbesii Robertson, 1890 solitary 2 1 3 ground nester

paraforbesii McGinley, 1986 solitary 2 1 3 ground nester

titusi Crawford, 1902 solitary 1 1 ground nester

125

L. (Leuchalictus divergens Lovell, 1905 solitary 2 2 Identified by Warncke, 1975) ground nester Gibbs Gibbs, 2010, 2011 leucozonium Schrank, 1781 solitary 21 17 107 196 81 59 481 ground nester

zonulum Smith, 1848 solitary 2 179 201 382 ground nester

L. (Sphecodogastra) quebecense Crawford, 1907 solitary 1 55 1 57 ground nester

Lasioglossum Curtis, pusillus * solitary 1 1 1833 ground nester

rufitarsum * solitary 1 3 4 14 22 ground nester

scutellaris * solitary 1 1 ground nester

spp. ** unknown 7 7 20 23 57

Sphecodes Latreille, aroniae Mitchell, 1960 cleptoparasite/ 1 1 Ascher & 1804 social parasite Pickering, 2018 Mitchell, 1960, atlantis Mitchell, 1956 cleptoparasite/ 3 3 1962 social parasite

brachycephalus Mitchell, 1956 cleptoparasite/ 3 2 5 social parasite

clematidis Robertson, 1897 cleptoparasite/ 1 26 27 social parasite

126

coronus Mitchell, 1956 cleptoparasite/ 6 6 social parasite

cressonii Robertson, 1903 cleptoparasite/ 2 2 social parasite

davisii Robertson, 1897 cleptoparasite/ 2 21 16 5 8 52 social parasite

dichrous Smith, 1853 cleptoparasite/ 1 3 12 1 17 social parasite

galerus Lovell and Cockerell, cleptoparasite/ 1 2 1 4 1907 social parasite

heraclei Robertson, 1897 cleptoparasite/ 2 5 1 8 social parasite illinoensis Robertson, 1903 cleptoparasite/ 1 1 social parasite

levis Lovell and Cockerell, 1907 cleptoparasite/ 1 1 1 3 social parasite

minor Robertson, 1898 cleptoparasite/ 18 68 1 87 social parasite

prosphorus Lovell and cleptoparasite/ 1 1 Cockerell, 1907 social parasite

ranunculi Robertson, 1897 cleptoparasite/ 1 1 2 4 social parasite

stygius Robertson, 1893 cleptoparasite/ 1 4 1 3 9 social parasite

Halictidae TOTAL 4505 2784 1671 2962 536 389 1284 7 127

Megachilidae

Anthidiellum notatum Latreille, 1809 solitary 1 1 Packer et al., 2007 Cockerell, 1904 ground nester

Anthidium Fabricius, manicatum Linnaeus, 1758 solitary 5 3 1 9 Ascher & 1804 ground nester Pickering, 2018

oblongatum Illiger, 1806 solitary 1 11 12 ground nester

Chelostoma Latreille, campanularum Kirkby, 1802 cavity nester 1 1 Ascher & 1809 Pickering, 2018

rapunculi Lepeletier, 1841 cavity nester 10 10

Coelioxys Latreille, modesta Smith, 1854 cleptoparasite/ 3 3 Ascher & 1809 social parasite Pickering, 2018

rufitarsis Smith, 1854 cleptoparasite/ 1 1 1 3 6 social parasite

sayi Robertson, 1897 cleptoparasite/ 1 2 3 social parasite

Heriades Spinola, carinata Cresson, 1864 cavity nester 2 1 2 7 2 5 19 Ascher & 1808 Pickering, 2018

leavitti Crawford, 1913 cavity nester 20 49 2 5 76

variolosa Cresson, 1872 cavity nester 18 15 33

128

Hoplitis Klug, 1807 truncata Cresson, 1878 cavity nester 1 1 Ascher & Pickering, 2018

pilosifrons Cresson, 1864 cavity nester 3 1 1 5

producta Cresson, 1864 cavity nester 4 1 1 4 10

spoliata Provancher, 1888 cavity nester 1 2 2 5

Megachile Latreille, addenda Cresson, 1878 cavity nester 1 1 Ascher & 1802 Pickering, 2018 Mitchell, 1960, brevis Say, 1837 cavity nester 2 2 1962

campanulae Robertson, 1903 cavity nester 1 2 3

centuncularis Linnaeus, 1758 cavity nester 3 3

frigida Smith, 1853 cavity nester 1 7 1 2 11

inermis Provancher, 1888 cavity nester 1 44 77 8 1 131

ingenua Cresson, 1878 cavity nester 2 2

lapponica Thompson, 1872 cavity nester 1 1

latimanus Say, 1823 solitary 15 29 1 1 46 ground nester

129

melanophaea Smith, 1853 solitary 1 1 ground nester

mendica Cresson, 1878 cavity nester 31 13 5 4 53

parallela Smith, 1853 solitary 1 1 ground nester

pugnata Say, 1837 cavity nester 1 2 2 5

relativa Cresson, 1878 cavity nester 8 6 3 17

rotundata Fabricius, 1793 cavity nester 42 45 9 3 99

texana Cresson, 1878 solitary 1 2 10 13 ground nester

gemula Cresson, 1878 solitary 1 1 ground nester

Osmia Panzer, 1806 atriventris Cresson, 1864 cavity nester 2 3 5 Ascher & Pickering, 2018

lignaria Say, 1837 cavity nester 1 1

pumila Cresson, 1864 cavity nester 1 1 8 10

tersula Cockerell, 1912 cavity nester 1 1

Stelis Panzer, 1806 lateralis Cresson, 1864 cleptoparasite/ 2 2 Ascher & social parasite Pickering, 2018 130

Megachilidae 14 2 185 251 54 97 603 TOTAL Melittidae

Macropis Panzer, nuda Provancher, 1882 solitary 6 6 Ascher & 1809 ground nester Pickering, 2018

Melittidae TOTAL 6 6

Grand TOTAL 5102 3058 4208 6143 626 746 1988 3

131

Table 10: The most abundant, or “core species”, at the site in each sampling year. Large numbers of individuals for each of these species appeared to drive trends in the overall wild bee community (e.g. abundance, guild proportions), including year-to-year variance. Functional Guild Species Number of specimens collected per year Classification 1968 1969 2002 2003 2016 2017

Social ground Halictus ligatus 64 88 493 450 58 8 nesters Lasioglossum cinctipes 93 23 152 436 1

Lasioglossum imitatum 3162 1894 15 81 9 1

Lasioglossum lineatulum 488 372 157 304 47 18

Solitary ground Andrena fragilis 70 101 1 nesters Andrena integra 9 2 257 35 2

Andrena wilkella 13 765 396 7

Calliopsis andreniformis 107

Colletes inaequalis 14 401 328 13 4

Lasioglossum zonulum 2 179 201

Perdita octomaculata 121 95

Cavity nesters Ceratina calcarata 3 283 511 62 185

Ceratina dupla 11 13 36 305 12 35

132

Appendix II

Minimum temperature Maximum temperature Rainfall Snowfall

2 2 Adj. R F1,4 B Adj. R F1,4 B Adj. F1,4 B Adj. F1,4 B R2 R2 Abundance

Solitary -0.25 0.94 -34.24 -0.23 0.80 156.80 -0.16 0.61 1.64 -0.10 0.50 5.09 ground nesters Social ground 0.66 0.03* -1232.39 0.47 0.08 -1404.53 0.39 4.18 7.22 0.61 0.04* 19.76 nesters Cavity nesters -0.22 0.76 -73.14 -0.25 0.97 13.68 -0.09 0.48 1.14 -0.23 0.79 1.05

Clepto- & -0.17 0.63 -25.93 <0.01 0.96 -3.49 -0.06 0.45 0.28 -0.11 0.52 0.57 social parasites Total 0.40 0.11 -1365.70 0.07 0.30 -1237.54 0.50 0.07 10.28 0.65 0.03* 26.46

133

Evenness

Solitary -0.23 0.79 -0.01 -0.16 0.61 -0.02 -0.23 0.81 <0.01 0.04 0.33 <-0.01 ground nesters Social ground 0.36 0.12 0.13 0.35 0.13 0.17 0.11 0.28 <-0.01 0.55 0.06 <-0.01 nesters Cavity nesters 0.12 0.27 0.06 0.13 0.26 0.08 0.19 0.22 <-0.01 -0.02 0.40 <-0.01

Clepto- -0.03 0.41 0.04 <-0.01 0.38 0.05 -0.09 0.49 <-0.01 0.18 0.23 <-0.01 parasites & social parasites Total 0.38 0.12 0.12 0.44 0.09 0.16 0.10 0.28 <-0.01 0.37 0.12 <-0.01

Diversity

Solitary -0.19 0.67 0.08 -0.19 0.69 0.10 -0.24 0.85 <-0.01 0.30 0.15 <-0.01 ground nesters Social ground 0.32 0.14 0.37 0.38 0.11 0.50 0.09 0.29 <-0.01 0.47 0.08 <-0.01 nesters Cavity nesters 0.43 0.09 0.33 0.52 0.06 0.45 0.18 0.22 <-0.01 0.30 0.15 <-0.01

Clepto- & -0.18 0.65 0.14 <-0.01 0.38 0.33 -0.24 0.84 <-0.01 -0.20 0.70 <-0.01 social parasites Total 0.28 0.16 0.60 0.38 0.12 0.83 0.02 0.35 <-0.01 0.31 0.15 -0.01

Richness

Solitary -0.06 0.45 5.34 0.02 0.36 8.17 -0.19 0.67 -0.02 0.22 0.20 -0.14 ground nesters Social ground -0.21 0.73 -0.30 -0.23 0.80 0.28 -0.23 0.81 <0.01 -0.14 0.57 <0.01 nesters Cavity nesters 0.06 0.32 3.56 0.08 0.29 4.77 -0.19 0.68 -0.01 0.15 0.25 <-0.07

134

Clepto- & -0.16 0.62 1.38 0.06 0.32 3.36 -0.20 0.70 <-0.01 -0.14 0.56 -0.03 social parasites Total -0.06 0.45 9.99 0.07 0.31 16.58 -0.19 0.69 -0.04 0.10 0.28 -0.22

135