Flight phenology of oligolectic solitary bees are affected by flowering phenology

Linköping University | Department of Physics, Chemistry and Biology Bachelor’s Thesis, 16 hp | Educational Program: Physics, Chemistry and Biology Spring term 2021 | LITH-IFM-G-EX—21/4000--SE

Flight phenology of oligolectic solitary

bees are affected by flowering phenology

Anna Palm

Examinator, György Barabas Supervisor, Per Millberg

Table of Content

1 Abstract ... 1

2 Introduction ... 1

3 Material and methods ... 3

3.1 Study species ... 3 3.2 Flight data ... 4 3.3 Temperature data ... 4 3.4 Flowering data ... 4 3.5 Combining data ... 5 3.6 Statistical Analysis ... 5 4 Result ... 5 4.1 Number of observations ... 5 4.2 Salix caprea ... 6

4.3 Succisa pratensis and Knautia arvensis ... 10

5 Discussion ... 14

6 Societal & ethical considerations ... 18

7 Acknowledgements ... 18

1 Abstract

Understanding the relationships between solitary bees’ flight phenology and flowering

phenology is important in the context of global warming. Using Swedish citizen science data, observations of oligolectic solitary bees and flowering phenology were used together with temperature data. All five bees studied had flight period that overlapped with the flowering period their corresponding host plant. None of the species were affected by the temperature, although there was a correlation between earliest observations of flowering phenology and flight phenology. The later the flowering observation was made, the later the flight

observation was made. No correlation was found between the length of flight period and length of the flowering period. Increasing temperature is not the only factor that effects flight phenology and flowering phenology.

Key words: Flight phenology, Flowering phenology, Global warming, Oligolectic, Solitary bee

2 Introduction

Around the world 78 % to 94 % of all flowering plants require animals to pollinate in some way for the plant to reproduce with success (Ayers et al., 2021; Ollerton, 2011). Of these animals a significant portion is bees. Not only are the plants and bees effected by this

pollination relationship but also other animals and plants who rely on these services for food or other resources (Ayers et al., 2021). There are approximately 250 species of solitary bees in Sweden (Stenmark, 2016). Solitary bees are so called because they do not live in colonies as e.g. honeybees. Different species of solitary bees are active at different times of the year. Depending on the species their flight period can occur anytime between spring to early autumn (Stenmark, 2016). Phenology describes the specific date of natural occurring events that happens every year (SMHI, 2021a). These events can be both biotic and abiotic for example when the first leaves changes colour in autumn, when migrating birds return in spring or when the first snow falls in winter (SMHI, 2021a). In the current report focus is on flowering phenology and flight phenology. Flowering phenology indicate when flowering was observed. Flight phenology indicate when flying solitary bee were observed.

Solitary bees can be divided in to two different groups, the oligolectic species and the polylectic species (Pekkarinen, 1997). The oligolectic bees have specialized in collecting

pollen and nectar from a specific plant family, or a small number of species. If a bee only collects pollen and or nectar from one species, it is called narrow oligolectic or strictly oligolectic. (Pekkarinen, 1997; Schlindwein, 2004). The polylectic bees are more general when it comes to pollen and nectar preference, this means that these bees can collect pollen and nectar from a range of different plant families. The polylectic bees can also collect pollen and nectar only within a single plant family but from more species than an oligolectic bee would collect (Pekkarinen, 1997; Schlindwein, 2004).

The Intergovernmental Panel on Climate Change (IPCC, 2014) define global warming as a gradual increase of global surface temperature. This has a substantial effect on our

environment. Glaciers are melting, more countries experience more intense heat waves and sea levels are rising (IPCC, 2014). Phenology is also affected by global warming (Hegland et al., 2009). The warmer temperatures can cause both plants and animals to flower or to become active earlier than they used to (Hegland et al., 2009; Kharouba et al., 2018). Because of this, with the help of phenology observations through several years, changes in climate and global warming can be observed (SMHI, 2021a).

Understanding the relationships between solitary bees’ flight phenology and flowering phenology is important when climate change affects our world (Gallagher et al., 2020). Because the oligolectic solitary bees only collect pollen from specific species it is interesting to know when these bees are active and search for pollen and nectar and when the plants are flowering. The aim of this report is therefore to compare data for the oligolectic solitary bees’ flight period and the flowering of the flowers sought after. More specifically, (1) what is the relationship between solitary bees’ first flight and first flowering over different years?, (2) how does the length of the flowering period influence the flight period?

I expected that both flowering phenology and flight phenology would start earlier warmer years. I also expected that the flowering period would be longer during the warmer years, which would imply an extended flight period for the bees studied. To test these assumptions, I compiled citizen science data for five oligolectic bee species and the flowering of their host plants from an 11-year period for southern Sweden.

3 Material and methods

Three types of data were used in the present study: weather data, insect data, and flowering data. The solitary bee data and flowering data is based on Swedish citizen data while the weather data is from historical temperature data. The regions Götaland and Svealand in southern Sweden were in focus of in the present study.

3.1 Study species

This report takes a closer look at five different solitary bee species Andrena vaga, Andrena praecox, Colletes cunicularius, Andrena marginata and Andrena hattorfiana. Andrena vaga, Andrena praecox and Colletes cunicularius collect pollen from Salix caprea and other Salix species which means these solitary bees are flying early in the year when most Salix flower. Andrena vaga digs holes in sandy grounds to make nests (Holmström et al., 2018a). Andrena praecox also make nests in sandy grounds and like to live close to Salix (Holmström et al., 2018b). Only the females are oligolectic for Salix caprea while the males have a more diverse diet. Flowers that males collect nectar from include Tussilago farfara and Crocus vernus, which are both also flowering in early spring (Holmström et al., 2018b). Colletes cunicularius create nests in fine sandy grounds often close to water (Cederberg et al., 2021). Andrena marginata collect pollen from several different Dipsacaceae species mainly Succisa pratensis and Knautia arvensis, though there are a few other species that are rare and with limited geographic distribution for example Scabiosa canescens, and Scabiosa columbaria (Holmström et al., 2018c). Therefore, these two plant species were not considered in the current study. Andrena marginata depends on having a good nest in close proximity to flowers like Succisa pratensis and Knautia arvensis (Holmström et al., 2018c). Andrena hattorfiana is strongly specialized in what flowers they collect pollen from (Holmström et al., 2018d). Because Andrena hattorfiana is strictly oligolectic, it only collect pollen from

Knautia arvensis. Pollen collecting from Scabiosa columbaria may occur in the most southern parts of Sweden and in Europe. Andrena hattorfiana live in sandy dry grounds that are more open but still close to where Knautia arvensis grows (Holmström et al., 2018d). In the summer the nests house the development from larva to adult bee, this applies to all the studied species. Even though the bees become adults before the summer ends, they stay in their nest throughout the winter until Salix caprea flower in early spring and in summer when Succisa pratensis and Knautia arvensis flower (Cederberg et al., 2021; Holmström et al., 2018c).

3.2 Flight data

Data over observations of the solitary bees Andrena vaga, Andrena praecox, Colletes cunicularius, Andrena marginata and Andrena hattorfiana was downloaded. These five species were chosen because of their oligolectic preference and because of the amount of observations reported. These data came from the open register on Artportalen which is

harbouring Swedish citizen science data (SLU Artdatabanken, 2021). Swedish citizen science data means that both professionals and amateurs can upload data of observations made. Of the downloaded data, I selected the years 2010-2020, and only Imago/adult individuals, and excluded dead individuals. All graphs and analysis were made in Rstudio (RStudio Team, 2020). The day of the year an observation was made was calculated. The provinces were grouped geographically after the regions Götaland, Svealand, and Norrland. Because observations turned out to be low in Norrland, this region was excluded. For Andrena marginata and Andrena hattorfiana there were too few observations in Svealand for a meaningful analysis, so only Götaland was included. The data were also formatted so every row in the data represented one individual solitary bee instead of one observation time where multiple bees could have been observed. The relevant observation months were noted for all species so that the weather data could be filtered accordingly.

3.3 Temperature data

Monthly data of daily averages temperature from all Swedish weather stations was

downloaded from SMHI (SMHI, 2021b). I compiled such monthly data for March to October for each year and region between 2010 to 2020. In addition, corresponding values for the climatic reference period 1961-1990 was also compiled. The data actually used in the present study was deviation of the measured temperature from this 30-year average. The bee species have different flight period. Hence, species-wise temperature data was compiled: Andrena vaga March to July; Andrena praecox March to June; Colletes cunicularius March to September; Andrena marginata; June to September; Andrena hattorfiana May to August.

3.4 Flowering data

Flowering phenology data on Salix caprea were received from Ola Langvall his work with Åslög Dahl and Kjell Blomgren on phenology (Blomgren et al., 2015). These data were used in reference for the bees Andrena vaga, Andrena praecox and Colletes cunicularius. ArcMap (ESRI, 2011) was used to assign a region to each report, based on its coordinates. For each

year and every region, the first and last flowering date was defined after excluding the first 5 % and the last 5 % of the yearly and regional data (to limit the influence of outliers in data).

For Andrena marginata and Andrena hattorfiana flowering data on Knautia arvensis and Succisa pratensis were downloaded from Artportalen (SLU Artdatabanken, 2021). For Andrena marginata both Knautia arvensis and Succisa pratensis was used but for Andrena hattorfiana only Knautia arvensis was used. When the species was identified it was also noted (for most observations) if it was flowering or not.

3.5 Combining data

Flight data and flowering data were combined in violin-graphs. Flowering data on Salix caprea were not suited to be presented in a violin-graph, and were instead represented as first and last flowering.

3.6 Statistical Analysis

To evaluate whether flowering phenology (start and length of flowering) affected the flight period, a generalized liner model was done for each bee species. Where relevant, “region” was included in the model. The first 5 % and last 5 % of observations (per year and region) were excluded to exclude any extreme values. Both significant relationships (P < 0.05) and non-significant relationships (P > 0.05) were plotted in scatter-graphs.

4 Result

4.1 Number of observations

The bee species with the most individuals was Andrena vaga (median and minimum number of individuals per year were 704 and 171; Table 1). Andrena praecox had the least individuals observed (median and minimum number of individuals per year were 33.5 and 11; Table 1).

The flowering phenology of Salix caprea, Knautia arvensis, and Succisa pratensis was based on 305.5, 624, and 1353 observations per year (median), respectively, Table 2.

Tabell 1. Observations of Solitary bee species between 2010-2020 in two regions: Götaland and Svealand.

Tabell 2. Observations of Salix caprea, Knautia arvensis, and Succisa pratensis between the years of 2010-2020. The observations are divided into the regions Götaland and Svealand.

4.2 Salix caprea

Andrena vaga and Salix caprea had a similar length for flight and flowering which was to be expected (Figure 1). For most years, the majority of Andrena vaga observations were made within the flowering period for both regions. Some years, however, there was a mismatch for example 2012 in Svealand where flowering started before any observations of Andrena vaga were made.

Species Andrena vaga Andrena praecox Colletes cunicularius Andrena marginata

Andrena hattorfiana

Region Götaland Svealand Götaland Svealand Götaland Svealand Götaland Götaland 2010 270 427 158 43 841 1004 513 104 2011 1022 2781 78 21 1035 632 10 376 2012 688 666 25 22 381 1325 95 324 2013 2042 1916 26 15 840 752 248 1141 2014 2621 1053 59 107 387 133 92 183 2015 720 1137 50 14 329 629 344 339 2016 2551 1669 24 26 130 437 236 400 2017 171 549 39 20 292 281 297 349 2018 451 463 62 11 249 978 151 119 2019 523 576 66 39 95 106 58 195 2020 412 1409 103 28 430 1919 188 101

Species Salix caprea Knautia arvensis Succisa pratensis

Region Götaland Svealand Götaland Götaland

2010 94 48 91 1353 2011 78 58 52 1578 2012 84 60 1064 587 2013 109 48 84 1328 2014 117 53 861 316 2015 336 275 702 5956 2016 492 374 291 4561 2017 697 588 624 42 726 2018 639 683 1117 150 2019 549 742 2086 378 2020 496 1037 612 5149

Figure 1. Observations Andrena vaga and flowering of Salix caprea in Götaland (A) and Svealand (B). Violin-graph over observations of Andrena vaga every year sorted from coldest year to the warmest year (n.b. the different order in the two regions). The box-plot inside the violins represent the quartiles of observations. The two points every year represent the first and last flowering observation of Salix caprea, after excluding the first 5 % and last 5 % of observations.

Andrena praecox also had a similar flight period compared to the flowering period of Salix caprea (Figure 2). The number of observations for Andrena praecox had a minimum of 10 observations per year/region.

Figure 2. Observations Andrena praecox and flowering of Salix caprea in Götaland (A) and Svealand (B). Violin-graph over observations of Andrena praecox every year sorted from coldest year to the warmest year (n.b. the different order in the two regions). The box-plot inside the violins represent the quartiles of observations. The two points every year represent the first and last flowering observation of Salix caprea, after excluding the first 5 % and last 5 % of observations.

The flight period of Colletes cunicularius, a relatively common species (minimum 90 observation per year/region), overlapped well with the flowering period of Salix caprea (Figure 3).

Figure 3. Observations Colletes cunicularius and flowering of Salix caprea in Götaland (A) and Svealand (B). Violin-graph over observations of Colletes cunicularius every year sorted from coldest year to the warmest year (n.b. the different order in the two regions). The box-plot inside the violins represent the quartiles of observations. The two points every year represent the first and last flowering observation of Salix caprea, after excluding the first 5 % and last 5 % of observations.

The first flight of the Salix-pollinating bees started around day 75 to 110 depending on the region, species, and year while the first flowering was observed around day 75 to 120 (Figure 4A). Generally, the first flight started at a later date when the flowering was late (Figure 4A). First observations of both Andrena vaga and Colletes cunicularius had a significant positive correlation with first observations of Salix caprea according to the generalized linear models (Table 3). Colletes cunicularius had a significantly earlier flight phenology in Svealand then in Götaland (Table 3).

The flight period of the Salix-visiting bees ranged from around 15 to 75 days depending on the region, species, and year while the range of flowering was observed around day 10 to 45 (Figure 4B). The flowering period of Salix caprea was, in most cases, around 30 days long.

The generalized liner model showed that there was no significant correlation with any of the three species and the flight period length or the region (Table 3).

Figure 4. Observations of Andrena vaga (yellow), Andrena praecox (purple), and Colletes

cunicularius (cyan). The first observation day of year compared to the first observation day of the year for Salix caprea (A). The flight period compared to the flowering period for Salix caprea in the same year (B). The circular points represent observations in Götaland and triangular points represent observations in Svealand.

Table 3. Results of generalized linear model of first flight and length of flight period for Andrena vaga, Andrena praecox, and Colletes cunicularius in southern Sweden. Statistical Z-values for first flowering and flowering period of Salix caprea and two geographic regions.

1 _{Götaland compared to Svealand: earlier flight starts if negative, longer flight period if positive.}

Significant at * P < 0.05, ** P < 0.005, *** P < 0.000001.

4.3 Succisa pratensis and Knautia arvensis

The flight period of Andrena marginata coincided well with the flowering of Succisa pratensis (Figure 5B) but in most years Knautia arvensis had ceased flowering when flight commenced (Figure 5A).

Flight start: Z-value Flight period: Z-value Species Salix caprea Region1 _{Salix caprea Region}1

Andrena vaga 2.817** 1.618 -0.702 -0.406

Andrena praecox 1.465 -0.615 0.580 -0.379

Figure 5. Observations of Andrena marginata and flowering of Knautia arvensis (A) and flowering of Succisa pratensis (B). Violin-graph over observations of Andrena marginata every year sorted from coldest year to the warmest year. The box-plot inside the violins represent the quartiles of

observations. The lighter violins every year represent the flowering observation of Knautia arvensis (A) and Succisa pratensis (B).

Andrena hattorfiana had a high overlap with flowering observations of Knautia arvensis (Figure 6).

Figure 6. Observations Andrena hattorfiana and flowering of Knautia arvensis. Violin-graph over observations of Andrena hattorfiana every year sorted from coldest year to the warmest year. The box-plot inside the violins represent the quartiles of observations. The lighter violins every year represent the flowering observation of Knautia arvensis.

The first flight of Andrena hattorfiana consistently started earlier then the first flight of Andrena marginata when compared to the first flowering observation (Figure 7A). Later start of flowering consistently lead to later flight, for both bees. When comparing the first flight observations of both bees to first flowering observation of Knautia arvensis significant results were found (Table 4).

The flight period of Andrena marginata and Andrena hattorfiana compared to the flowering period of Knautia arvensis were mixed (Figure 7B). The flight period was between around 10 to 55 days depending on the species and year. The flowering periods length was between around 5 to 115 days depending on the year the observations were made. The flight period of Andrena hattorfiana tended to be 30 to 45 days with only small variation between years

(Figure 7B). Andrena marginata flight period varied more between years (Figure 7B). There was, however, no correlation between Andrena hattorfiana and Andrena marginata (Table 4).

Figure 7. Observations of Andrena marginata (yellow) and Andrena hattorfiana (purple). The first observation day of year compared to the first observation day of the year for Knautia arvensis (A). The flight period compared to the flowering period for Knautia arvensis the same year (B).

Table 4. Results of generalized linear model of first flight and length of flight period for Andrena marginata and Andrena hattorfiana in southern Sweden. Statistical Z-values for first flowering and flowering period of Knautia arvensis and Succisa pratensis.

Significant at * P < 0.05, ** P < 0.005

The first flight observations of Andrena marginata were spared out (Figure 8A). There was no correlation between first flight observations of Andrena marginata and first flowering observation of Succisa pratensis (Table 4). Flight period and flowering period did not have a correlation with each other either (Table 4). The flight period was around 10 to 55 days long depending on the year. The length of flowering Succisa pratensis was between 5 to 120 days long.

First Flight: Z-value Flight Period: Z-value

Species Knautia arvensis Succisa pratensis Knautia arvensis Succisa pratensis

Andrena marginata 2.128* -0.193 0.448 -0.042

Figure 8. Observations of Andrena marginata. The first observation day of year compared to the first observation day of the year for Succisa pratensis (A). The flight period compared to the flowering period for Succisa pratensis the same year (B).

5 Discussion

The observations of Andrena vaga and Colletes cunicularius compared to observations of Salix caprea show similar overlapping periods. Andrena vaga and Colletes cunicularius had an overlap of nine years of eleven and ten of eleven years of observations in Götaland. In Svealand both bees had an overlap of seven of eleven years of observations. Both of these species collect pollen from only Salix plants where the main species is Salix caprea

(Holmström et al., 2018a; Cederberg et al., 2021). The results show that it does not matter if the years are sorted by temperature. The temperature in the species-specific interval does not have an effect on when the bees are active and when the plants are flowering. The solitary bees are not flying during a longer period in warm years of 2018 or 2020 compared to the cool year 2013 (Figure 1, Figure 3). The temperature period that does have an effect on the length and start for both flight period and flowering period is the temperature in winter and spring. However, in the current study only the months when the flight period occurred were included in the temperature interval for each species, and it is possible that considering temperature in the preceding month(s) would have been more useful.

The data on Andrena praecox can be somewhat misguiding because of the low amount of observations per in some combinations of year and region. The worst data point for Andrena

praecox was 2018 in Svealand where only 11 observations were made. Because of the low number of observations some violins are more uncertain then others. Despite the low observations of Andrena praecox the overlap with Salix caprea was high. In Götaland the overlap was ten of eleven years of observations and in Svealand observations for nine of eleven years overlapped with Salix caprea observations. The flowering and flight period of Salix caprea and Andrena praecox were very similar even though only the females are oligolectic for Salix. (Holmström et al., 2018b).

Two of the Salix-visiting bees was affected by the onset of Salix flowering: the later the flowering the later the first bee will be recorded. Andrena vaga’s and Colletes cunicularius’ first flight appears at a significant earlier date when first flowering of Salix caprea is earlier. Colletes cunicularius’ first flight was also significantly affected by what region the

observations were made in. The observations of Colletes cunicularius were made significantly earlier when they were observed in Svealand compared with Götaland. This is surprising considering Götaland is located further south than Svealand, which would indicate that higher temperatures would start in Götaland before in Svealand.

According to faunistic literature, the oligolectic Andrena marginata uses both Knautia arvensis and Succisa pratensis. The current study, however, show that in total only five of eleven years of observations had a clear overlap between Andrena marginata and Knautia arvensis. The flight periods overlap with the flowering period of Succisa pratensis was clearer with nine of eleven years of observations overlapping. This indicates that Andrena marginata have a much closer relationship with Succisa pratensis then with Knautia arvensis. However, Andrena marginata had significantly earlier first flight observations when compared to first flowering observations of Knautia arvensis not Succisa pratensis. Succisa pratensis flowering starts later then flowering of Knautia arvensis which would indicate why Knautia arvensis have an effect on first flight observations instead of Succisa pratensis.

Observations of Andrena hattorfiana and Knautia arvensis had a great overlap, this is not surprising considering Andrena hattorfiana being strictly oligolectic. In total eleven of eleven years of observations overlapped. However, nature is not perfect so the phenology periods can be shifted depending on many factors. To maintain a close relationship between these two species a consistent temperature is important. If one of their periods get askew because of global warming the other will have to try to adjust their flight period or flowering period also. Those types of desynchronization between the bee and flower phenology cause fitness losses

in one of them or in both species (Villagomez et al., 2021). Andrena hattorfiana had

significantly earlier first flight observations that depend on when flowering phenology starts for Knautia arvensis.

For Andrena vaga, Andrena praecox, and Colletes cunicularius the majority of observations, regardless of year and flowering period, had a flight period that was between 15 and 60 days. The flowering period of Salix caprea do not reach 60 days. Generally, the species with the longest flight period was Andrena praecox which also happens to be the least oligolectic species of these three bees. That means that Andrena praecox flight phenology is not limited to when Salix caprea is flowering (Holmström et al., 2018b). Flight period of Andrena

marginata and Andrena hattorfiana was also unaffected by the length of flowering of Knautia arvensis and Succisa pratensis. Andrena marginata had a more extended flight period

compared with Andrena hattorfiana. Andrena hattorfiana had a concentrated flight period around 30 to 45 days, regardless of year and flowering period. It is understandable that Andrena hattorfiana would have a concentrated flight period as it is considered strictly oligolectic towards Knautia arvensis (Holmström et al., 2018d).

Not many studies have covered the relationship between pollinators and flowering plants within the context of global warming. Villagomez et al. (2021) have conducted a study that explores the affect global warming can have on the relationship between the overwintering colony bee Apis mellifera and the early spring flowering Crocus sierebi. They suggested that the increasing temperatures that follow global warming would have the largest effect on the species when changing the temperatures in winter and spring. When doing this they thought the flowering phenology would advance but also that the honey bees brood rearing activity would increase (Villagomez et al., 2021). The results of their study show that if the

temperature increase in the winter and in early spring, flowering phenology will advance. Even though Villagomez et al. (2021) study is focused on different species than the current study, the same conclusions can be drawn. A good synchronization between the species is critical for both plant and bee fitness. Temperature increase in the winter and spring have the largest impact on the relationship of overwintering bees and early flowering plants

(Villagomez et al., 2021).

Hegland et al. (2009) conducted a study about how global warming affect the interaction between a plant and its pollinator. The study resulted in that both plants and pollinators phenology were affected by global warming (Hegland et al., 2009). Other studies show that

different plants are affected by global warming in different ways (SMHI, 2021a). For example, Tussilago farfara and Hepatica nobilis both flower around March and April in Sweden. During the last few decades Hepatica nobilis flowering phenology have shifted to flowering earlier than Tussilago farfara (SMHI, 2021a).

Even within the same species phenology shifts can occur (Kehrberger et al., 2019). The study conducted by Kehrberger et al. (2019) focuses on the phenology of the solitary bees Osmia cornuta and Osmia bicornis. Cocoons of both species were placed on seven different

meadows with different mean temperatures to study if the warmer temperatures would cause the solitary bees to emerge faster. The results showed that Osmia cornuta males did have an emergence shift when the temperatures increased. However, the females did not experience a shift (Kehrberger et al., 2019). Plants are not only affected by temperature but also amount of rain and the length of the photoperiod play a part in then plants flower (SMHI 2021a,

Villagomez. et al 2021).

Global warming is an important factor to consider when looking at the relationship between flight phenology for solitary bees and flowering phenology. However, rising temperatures is not the only factor to take into account when looking at flowering phenology. To get a more accurate model over how flowering phenology have changed things like the amount of light plants is exposed to and the amount of rain a plant gets must be considered. In the current study these factors were not in mind when planning the study and handling the data. In Kehrberger et al. (2019) study there were differences between the sexes within the same species of solitary bee. My solitary bee data was not divided into males and females. This was done because it was not the focus of this study and also because the observation data from Artportalen (SLU Artdatabanken, 2021) in most cases did not indicate the sex of the solitary bee. In future studies, it would be interesting to consider more detailed weather data, e. g. winter temperatures and rainfall.

In conclusion, most species had very similar overlap between solitary bee and flower

observations. The ones that had less overlap were not exclusively collecting pollen and nectar from the flower I compared it with, for example Andrena praecox. Andrena hattorfiana on the other hand had the best overlap because of it is strictly oligolectic for Knautia arvensis. There were multiple significant correlations between bees’ first flight observations and first

flowering observations. However, unexpectedly the flight periods length was not affected by the flowering period for any species.

6 Societal & ethical considerations

Pollinators like the solitary bee contributes to ecosystem services in the form of pollination, which is crucial for our food production systems. The importance of ecosystem services in the form of pollination can affect the general public to understand how important solitary bees are and how supporting them are important. Without pollinators we would not have most of the food products we eat every day. This study has a connection to goal 15 about supporting life on land and biodiversity (UNDESA, 2021). Temperature changes affect the phenology of both solitary bees and flowering plants. If global warming continues to increase the relationship between solitary bee and flowering plant may be lost.

All insect observations data that comes from Artportalen (SLU Artdatabanken, 2021), and have been reported by the general public. It is unknown to what extent reports on bees are based on destructive sampling. Either way, as this study re-uses already existing data, no further harm is inflicted on the animal populations.

7 Acknowledgements

I would like to thank my supervisor Per Milberg for guidance during the making of this thesis. I would also like to thank Ola Langvall who provided phenology data of Salix caprea. I would also like to thank Malin Magnusson Rundqvist for always being there to discuss problems with during the making of this thesis.

8 References

Ayers, A. C., Rehan, S. M. (2021) Supporting Bees in Cities: How Bees Are Influenced by Local and Landscape Features. Insects 12(2): 128. https://doi.org/10.3390/insects12020128

Blomgren, K., Langvall, O., Dahl, Å., Göthlin, E., Hassel, L., and Grundström, S. (2015). Vårkollen, fenologi-väkteri och en ny miljömålsindikator. Svensk Botanisk Tidskrift, 109, 112-119

Cederberg, B. et al., 2021. Svenska bin. https://artfakta.se/naturvard/taxon/103078 (accessed 2021-05-12)

ESRI 2011. ArcGIS Desktop: Release 10. Redlands, CA: Environmental Systems Research Institute.

Gallagher, M. K. and Campbell, D. R. 2020. Pollinator visitation rate and effectiveness vary with flowering phenology. American Journal of Botany 107: 445– 455.

https://doi-org.e.bibl.liu.se/10.1002/ajb2.1439

Hegland, S. J., Nielsen, A., Lázaro, A., Bjerknes, A.‐L., & Totland, Ø. (2009). How does climate warming affect plant‐pollinator interactions?. Ecology Letters 12: 184-195.

https://doi-org.e.bibl.liu.se/10.1111/j.1461-0248.2008.01269.x

Holmström, G. et al., 2018a. Andrena – sandbin. https://artfakta.se/naturvard/taxon/andrena-vaga-103114 (accessed 2021-05-11)

Holmström, G. et al., 2018b. Andrena – sandbin. https://artfakta.se/naturvard/taxon/andrena-praecox-103102 (accessed 2021-05-11)

Holmström, G. et al., 2018c. Andrena – sandbin. https://artfakta.se/naturvard/taxon/andrena-marginata-102671 (accessed 2021-05-12)

Holmström, G. et al., 2018d. Andrena – sandbin. https://artfakta.se/naturvard/taxon/andrena-hattorfiana-102668 (accessed 2021-05-12)

IPCC. (2014). Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (Eds.)]. IPCC, Geneva, Switzerland.

Kehrberger, S., & Holzschuh, A. (2019). Warmer temperatures advance flowering in a spring plant more strongly than emergence of two solitary spring bee species. PLoS ONE 14(6): e0218824. https://doi.org/10.1371/journal.pone.0218824

Kharouba, H.M., Ehrlén, J., Gelman, A., Blomgren, K., Allen, J.M., Travers, S.E., and Wolkovich, E.M. (2018). Global shifts in the phonological synchrony of species interactions over recent decades. PNAS. 115: 5211-5216. https://doi.org/10.1073/pnas.1714511115

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

Pekkarinen, A. (1997). Oligolectic bee species in Northern Europe (Hymenoptera, Apoidea). Entomologica Fennica, 8(4), 205-214. https://doi.org/10.33338/ef.83945

RStudio Team. (2020). RStudio: Integrated Development Environment for R. RStudio, PBC, Boston, MA URL http://www.rstudio.com/

Schlindwein, C. (2004). Are oligolectic bees always the most effective pollinators. Solitary bees: conservation, rearing and management for pollination. Imprensa Universitária, Fortaleza 231-240. http://www.webbee.org.br/bpi/solitary/livro_04.pdf#page=228

SLU Artdatabanken. (2021). Artportalen. https://www.artportalen.se/ (accessed 2021-04-13)

SMHI. (2021a). Fenologi – naturens återkommande tidsmönster.

https://www.smhi.se/kunskapsbanken/klimat/fenologi/fenologi-naturens-aterkommande-tidsmonster-1.5189 (accessed 2021-05-18)

SMHI. (2021b). Års- och månadsstatistik. https://www.smhi.se/klimat/klimatet-da-och-nu/manadens-vader-och-vatten-sverige/manadens-vader-i-sverige/ars-och-manadsstatistik

(accessed 2021-04-16)

Stenmark, M. (2016). Solitärbin.

https://www2.jordbruksverket.se/download/18.8e04a5f15891f622e3549c0/1479989001295/o vr265_15v4.pdf (accessed 2021-03-16)

UNDESA. (2021). Goal 15. https://sdgs.un.org/goals/goal15. (accessed 26 April 2021).

Villagomez, GN, Nürnberger, F, Requier, F, Schiele, S, Steffan‐Dewenter, I. (2021). Effects of temperature and photoperiod on the seasonal timing of Western honey bee colonies and an early spring flowering plant. Ecology & Evolution; in press. https://doi.org/10.1002/ece3.7616