This dissertation has been submitted by a student. This is not an example of the work written by our professional dissertation writers.

Local Environment Europe

The effect of the local environment on bee abundance and diversity in regions throughout Europe.

Bees have an important ecological role; they are insect pollinators providing a crucial service. Without insect pollination human diet would be very different to how we know it now. Declines in pollinators have been reported and by attempting to understand the how the local environment affects bee abundance and diversity it may be possible to prevent any further decline. Samples were collected at six sites across Europe in each site there was a disturbed landscape and a natural landscape and within each of these a hot and a cold area.

Samples were collected, pinned and identified to genera and the Halictus measured. Analysis found that there was no significant difference in size between different countries, landscape and temperature. The number of individuals and the number of genera did not significantly differ between temperature, landscape and country however on a smaller country level there were differences in the numbers of individuals and genera at different landscapes.

The number of bumblebees was affected by the landscape with more individuals found in the disturbed landscapes; it is possible that due to the foraging ability and feeding preferences that bumblebees are able to gain an advantage in a disturbed landscape. Different genera were found in different regions with high numbers of Panurgus and Panurginus found in Spain and Catalunya. The number of bumblebees was also found to be significantly related to latitude. These differences in composition in different areas could be seriously affected in the face of climate change. The effect of the local environment on bee abundance and diversity in regions throughout Europe.

1: Background and importance

“If the bee disappeared off the surface of the globe then man would only have four years of life left. No more bees, no more pollination, no more plants, no more animals, no more man.” Albert Einstein

1.1: The importance of bees

Bees provide the critical ecosystem service of pollination (Kearns et al 1998). Insect pollination is essential for our life as we know it. 84% of crops in the EU depend on insect pollination (Williams 1994) and one third of our diet can be attributed to insect pollination, either directly or indirectly (McGregor 1976). Of the insect pollinators it is bees which provide the most pollination, bees which are highly adapted to flower visitation, have been confirmed to be pollinators for 72.7% of crop species and it is thought they could be responsible for the pollination of another 10.2% (Williams 1994, Roubik 1995).

Declines in bees point towards serious consequences for natural ecosystem process and agricultural processes (Biesmeijer et al 2006). The declines being experienced on local and regional scales present a worrying situation with habitat loss, fragmentation, agricultural intensification and pesticide use causing declines in honey bees, solitary bees, wild bees and bumble bees (Steffan-Dewenter et al 2005).

The greatest diversity of bees in the world is experienced in arid and semi-arid regions of the world including the Mediterranean regions of Southern Europe (Danforth 2007). Most of the bees in the world are solitary bees (National Research Council of the National Academies 2007) and of the solitary bees the majority of them are resource specialists, oligolectic (Wcislo and Cane 1996). Oligolectics are bee species which collect pollen from one genus or species but can collect nectar from a variety of plants, they are often referred to as specialists.

Polylectic bee species are generalists; they can collect pollen from a variety of flowering plants and include the honeybee (Apis) and the bumblebee (Bombus). In theory the risk of extinction is elevated in oligolectic bees as their presence and distribution is limited by just one floral host (Zayed and Packer 2007). Work by Cane et al (2006) into urban habitat fragmentation showed the abundance and richness of oligolectic bees to have declined but not to have declined in the polylectic bees.

Due to the important role of bees it is essential to understand the abundance and diversity of bees across the landscape and the local factors that affect them. By understanding the local factors affecting the diversity and abundance of bees it may be possible to effectively manage and conserve bees and help to prevent any further declines in diversity and abundance.

1.2: Landscape

Much of the natural habitat in Europe has been lost and the habitats with the highest species richness are the remaining semi-natural areas (Pimentel et al 1992).

The impact of disturbance on insect communities is not so extensively studied as the impact on vegetation, on the studies that are available results show that different insect groups respond differently to disturbance (Steffan-Dewenter and Leschke 2003).

Study by Steffan-Dewenter and Leschke (2003) on the effect of habitat management and landscape on bees and wasps in orchards in Europe showed that the vegetation was more significantly affected by the management practices than the insects.

Bee species richness is correlated with the percentage of grassland in the surrounding landscape (Dauber et al 2003, Hendrickx et al 2007, Steffan-Dewenter et al 2002).

The bees in the study by Hendrickx et al (2007) showed not only a decline with distance from semi-natural patches but also a decline with increasing management practices. The other groups in the study experienced increased numbers with proximity to semi-natural habitat but no significant declines with increasing agricultural management. The results for bees can be attributed due to bees having such a strong dependence on floral resources (Tscharntke et al 1998). Low plant diversity with limited floral resources may not to be able to support a high diversity of insects thus resulting in lower insect diversity and the ability to support only the generalist species (Westphal et al 2003).

Proximity to floral resources and nesting sites is important as foraging distances can be fairly small. Large bumblebees such as Bombus terrestris can forage distances up to 3000m, as foraging distances are related to body size, smaller bees may only be able to forage a few metres (Westphal et al 2006). In the tropical forests of Costa Rica pasture management and the floral resources showed to have no significant impact on the diversity or abundance of bees, however deforested countryside just metres away from the forest contained a different community composition (Brosi et al 2006).

The complexity of landscapes means that the impact of disturbance can vary depending on the frequency, intensity and extent of the disturbance (Samways 2005).

Moderate disturbance can actually increase the diversity of the area by opening up areas for colonisation by providing ecological niches and opportunities for rarer species (Caswell 1976, Connell 1978, Petraitis et al 1989). Alternately diversity could be lowered as the dominance of opportunistic species is increased (Margalef 1968).

Bees depend on floral resources for nectar and pollen and can only travel certain distances from their nesting site to reach it, both flowers and nests need to be close by. Therefore declining floral resources, and declining suitable nest sites, as experienced in large scale disturbed areas, may result in the declining numbers and diversity of bees.

1.3: Microclimate- temperature

The microclimate, the lowest two metres of the atmosphere (Stoutjesdijk and Barkman 1992), is the layer of the atmosphere where the majority of plants and animals live (Unwin and Corbet 1991). The soil surface (or other substance, for example forest or concrete) influences the heat and moisture budget of the surrounding atmosphere producing localised variations in the climatic conditions, for example temperature, windspeed and humidity.

The relationship between plant and microclimate is a close one with plants affecting the climatic conditions around them and the microclimate affects the factors controlling the functioning of the plant including the availability of the products required for photosynthesis. Insects benefit from this interaction and due to the close mutalistic relationship between some plant and insect species, for example plants and pollinators, are dependent on a healthy relationship between microclimate and plants. An unsuitable microclimate will lead to the deterioration of plant life and eventual death of the plant and insects dependent on it.

1.4: Insects, temperature and body size

In many insects body temperature is essential in order to gain flight. An insect needs to gain enough energy to fly; it needs to raise the thoracic temperature above the temperature of the environment (Bishop and Armbruster 1999) this can be achieved by basking and endothermy (producing heat in the muscles) (Unwin and Corbet 1991).

The size of the insect plays a vital role in the ability to heat up and subsequently fly and forage. A study by Casey and Joos (1983) found that the proportion of heat lost from the thorax per time unit decreases as the body mass of the insect increases, therefore larger insects are slower at gaining and loosing heat. Bishop and Armbruster (1999) also concluded that the ability to raise temperature in order to fly increases with body size making bumblebees better thermoregulators than solitary bees. Even when looking at solitary bees larger solitary bees will be better at thermoregulation than smaller solitary bees.

Foraging activity can be restricted by thermoregulation factors (Heinrich 1974) and not just over winter. In the summer months foraging at high latitudes and higher temperatures may prove to be difficult for larger insects with solitary bees reaping the floral rewards. Whilst in cooler areas at lower latitudes larger bees, such as bumblebees will have the advantage (Bishop and Armbruster 1999).

Tropical bumblebees have been found to be the largest bees, an exception to the rest of the findings by Peat et al (2205). They found that the mean size of bumblebees varies between different climates with colder climates having a larger mean size than those of warmer climates. Size variation of bumblebees within a region was found not to be related to temperature but other factors, possibly to improve colony foraging with different sizes able to visit different flowers (Peat et al 2005).

It is not just at different temperatures, different latitudes and different elevations that there are heat constraints on the species present but also a daily sequence. Heinrich (1976) observed bees visiting flower patches and noted the day sequence process. Large insects, such as large bumblebees, are able to achieve a body temperature high enough to fly at a lower temperature than a smaller insect, for example a small solitary bee. This then means that earlier in the day the bumblebee can begin to forage and last longer into the evening when the temperature of the environment begins to fall. However in the midday heat the bumblebee may become overheated and need to retreat and cease flying for a few hours. The small solitary bee although not able to start until later and unable to continue into the evening will be able to cope in the midday heat and continue to forage (Unwin and Corbet 1991).

The temperature of the area determines the foraging activity of bees and will influence the bees present in the area. What is under-researched is the effect of very localised temperature has on the bees and size of bees present.

1.5: Climate

The temperature of the environment does not only determine the body temperature of the insect but also the geographical range (Gates 1993). Over the past 30years shifts in the abundance and distribution of a variety of species have been witnessed due to climate change (Parmesan and Yohe 2003). Hickling et al (2006) studied the distributions of different taxonomic groups in Britain over the last 25years to examine any shifts in range that may have occurred. A shift in distribution upwards and northwards was found in most taxonomic groups with the latitude being a more significant factor than elevation.

Alterations to geographic ranges will impact different organisms in different ways and at different times in their lifecycle. It is possible that the interactions between organisms could be severely affected and possibly even destroyed, in some instances resulting in the extinction of one or both of the species. With these shifts in distributions comes the increased possibility of species extinctions, one prediction for 2050 using a mid-range climate scenario showed 15-37% of species committed to extinction (Thomas et al 2004). In order to avoid the risk of extinction species will have to be able to keep up with the changing climate by migrating at fast enough rates however barriers such as mountains and fragmented, disturbed landscapes may hinder this migration (Pearson and Dawson 2003).

General climate models which observe the possible consequences of climate change show a general pattern of the increasing of the Mediterranean summer drought (Gates 1993). As a result it is expected that a shift in species composition will occur and drought conditions will lead to reduced plant cover. This will inevitably impact many insect species including pollinators, such as bees, that will lose their source of nectar and pollen. Research suggests that resource specialists are likely to be the first to suffer declines as they rely on just one plant for their pollen (Cane et al 2006).

Looking at the effect of local temperatures on abundance and richness may be able to give an indication of what will follow with global climate change and thus be an aid for planning and conservation measures.

2: Aims and Objectives

Bees are essential for pollination and are the key to maintaining life as we know it. Reaching and maintaining the right temperature is essential for an insect’s flight, there is evidence that reaching this temperature is related to body size but does it vary with temperature within a microclimate? Does the local temperature affect the bee diversity and abundance and will this provide any insights into what may happen in the face of global climate change? Within Europe it has been reported that it is the remaining semi-natural habitats that contain the most species richness. If this is the case it would be expected that areas of human disturbance would experience a much lower diversity and abundance.

In this project the aim is to examine the effect that the local conditions, temperature and landscape, have on the abundance and genera of bees in a selection of regions across Europe. Within this there are three main objectives to be examined:

  • To determine if the local temperature affects the abundance and diversity of bees.
  • To determine if the surrounding landscape, disturbed or natural, affects the diversity and abundance of bees.
  • To establish whether the size of certain genera are significantly affected by the local environment.

3: Methodology

3.1: Site selection

Samples of pollinators were collected in field sites throughout Europe in the summer of 2007 as part of the CITIRAT (Climate Interactions with Terrestrial plant Interactions a Risk Assessment Tool) project. The CITIRAT project is part of the wider EU ALARM (Assessing LArge scale Risks for biodiversity with tested Methods) project (http://www.alarmproject.net/alarm/). The field sites for the CITIRAT project were pre-determined by ALARM, with the core sites situated in different regions throughout Europe allowing the study of most of the climatic regions in Europe.

For each of the core sites there are two sites measuring 4km by 4km within 50km of each other. One of the two sites being predominantly natural or semi-natural and the other site a disturbed landscape. The two focal sites have being selected so that the geological and environmental parameters are as similar as possible allowing the human disturbance to be the most distinguishing features between the sites. Figure 3.1.1 shows examples of the land cover in each category.

Table 3.1.1: An example of the classification of disturbed and natural sites, categories taken from the level 3 CORINE 2000 land cover classification.

Disturbed

Non-irrigated arable land, pastures, discontinuous urban fabric, complex cultivation procedures, fruit trees and berry plantations, agro-forestry areas, olive groves, permanently irrigated land.

Natural/semi-natural

Mixed forest, coniferous forest, broadleaved forest, transitional woodland-scrub, sclerophyllous vegetation, natural grasslands.

Using GIS analysis the temperature for each of the disturbed and natural areas was calculated using a model which combined the elevation, slope, aspect, average daytime temperature, clear sky solar radiation maps. This model then gave the temperatures for points throughout the landscape, the hottest 10% and coldest 10% of points were selected and ranked, the top two temperature points for both hot and cold were then determined and ready for fieldwork to begin.

3.2: Sampling method

Each of the two landscapes (disturbed and natural) had two sampling rounds approximately 2 weeks apart. Within each sampling round two hot and two cold temperature sites were used (as predetermined by the GIS analysis). Each temperature spot had three cluster sets of pan traps, one white, one yellow and one blue. Each cluster contained five pan traps of a single colour. Each cluster was situated five metres apart in open, low vegetation at ground level. The pan traps were left out over a two day period in dry conditions with low wind and a temperature of greater than 15ºC. Leaving pan traps out over a two day period eradicated any daily variation in bee species present due to daily temperature fluctuations. By using all three coloured pan traps bias was reduced as a range of colour preferences could be catered for (Leong and Thorp 1999).

When the samples, preserved in alcohol, were returned to Leeds the samples were sorted taking note of the number of honeybees, number of bumble bees, number of other bees, number of hoverflies and the number of butterflies. Anything else that was collected in traps was discarded.

The bumblebees and other bees were removed from the sample tubes, and were dried, pinned and labelled. The bees were then identified to genus level and the results recorded.

Figure 3.3.1: Map of Europe showing the ALARM core sites. The yellow dots indicate the sites used in this analysis and their ‘country’ label. Adapted from an image available at: http://www.alarmproject.net.

3.3: Analytical method

Samples were collected at sites all across Europe. Time and resource restraints meant that not all of the sites sampled could be pinned and identified for use in this study. The sites used were carefully selected with sites showing high variation in elevation and therefore temperature differences chosen. Figure 3.3.1 shows the European sites used in this project and table 3.3.1 shows the latitude and longitude of the sites. From here on these ‘sites’ will be called countries to avoid confusion.

Table 3.3.1: Sites used with the latitude and Longitude

Country

Landscape

Latitude

Longitude

Austria

Disturbed

47.5205

14.1432

Austria

Natural

48.0125

15.1620

Catalunya

Disturbed

41.2620

1.7714

Catalunya

Natural

41.2526

1.9006

Germany

Disturbed

51.5491

9.7754

Germany

Natural

51.4540

12.9410

Italy

Disturbed

45.6202

12.4526

Italy

Natural

45.7775

12.6088

Spain

Disturbed

39.3153

-4.0661

Spain

Natural

39.4133

-4.0650

UK

Disturbed

51.5082

-1.5310

UK

Natural

51.7650

-0.4585

To calculate the diversity for each of the conditions at each of the sites the Simpson’s diversity index, which is “one of the most meaningful and robust diversity measures”(Magurran 2004) was used. The index works by calculating the probability, that from a community of infinite size, two individuals will belong to the same species. The Simpson diversity index was expressed as 1-D therefore meaning that as the Simpson’s diversity decreases as does the diversity, this logical adaptation of the index mean that the diversity of the samples could easily be calculated and compared.

In order to determine if the size of bees are affected by the local conditions one genus, Halictus was chosen due to them making up a large proportion of total individuals present. To measure the Halictus samples a random number table was used to determine which specimens should be measured. All of the specimens were females and from two countries, Spain and Germany. Digital callipers were used under a microscope to measure the width of the thorax, in-between the base of the wings.

The numbers of Bombus’ were looked at as well as the size of the Halictus. Bombus’ are known to be (generally) a larger body size and better thermoregulators so provide a good genus to use as an indication of distribution related to the local environment factors.

The information available for use in the statistical analysis was the number of individuals, the number of genera, the temperature (hot or cold), the landscape (disturbed or natural), the country, the sample round (1 or 2), the site (either 1 or 2), the diversity (Simpson’s 1-D), the number of bumblebees, the number of solitary bees and for a selection of sites the size of Halictus.

The statistical analysis was carried out using R and Minitab for the principal component analysis. Excel was used for the production of some of the graphics. Not all the data was normally distributed, distributions were checked using the Shapiro-Wilk test. The analysis used was a mixed effects model but not all data meet the assumptions so where unavoidable non-parametric tests were used, a generalised mixed effects model (glmmPQL).

4: Results

Nineteen Genera were identified; a list of these genera and authorities can be viewed in the appendix A. One genus could not be confirmed despite various opinions but is suspected that it might be Panurginus.

4.1: Individuals and genera

Figure 4.1.1: The mean number of individuals per sample round, error bars indicate ±1 SE.

(t66= -5.804, p=<0.0001, 95%CL)

A mixed effects model was used for the analysis of the individuals. The random effects were site, landscape and country. The standard deviation estimate for country was 0.528 showing that for the countries there was a variation from the mean, this may affect the outcome of the model. The only significant factor was sample round (t66=-6.456, p=<0.001, 95%CL).

There were no significant differences in individuals within temperature, landscape, sample round or any of the interaction terms. To alleviate the problem of countries having a great variation in the numbers of individuals the model was rerun with countries as a fixed factor. This reduced the variation of the random effects and recalculated the fixed effects. Sample round remained the only significant factor (t66= 5.804, p=<0.001, 95%CL) (figure 4.4.1).

The dataset for genera was non-parametric so the model used was the glmmPQL. A very low standard deviation estimate was given for each of the random factors (country=<0.060, site=<0.001) therefore for each of the random effects there is little variation meaning they have little effect on the overall model. There were no significant fixed factors in the model.

4.2: Diversity

The generalised mix effects model for diversity used Simpson’s 1-D values. The estimates of standard deviation for the random effect of country were extremely low, <0.001, showing that for diversity there was almost no variation from the mean (figure 4.1.2). Of the fixed factors for diversity none were significant (95%CL).

Figure 4.1.2: The mean diversity (Simpson’s 1-D) for each country. Error bars indicate ± 1 SE. The diversity was not significantly for any of the factors, Standard deviation between countries was low at <0.001.

4.3: Bumblebees and other bees

The numbers of bumblebees (Bombus spp) were used in a generalised mixed effects model (glmmPQL) in order to determine if there were significant differences in the variation between temperature, landscape and sample round. The standard deviation of country was high at 1.376 showing that within the effect of country there was a lot of variation from the mean, thus contributing to the variation in bumblebees and possibly influencing the overall model. Of the fixed factors sample round and landscape were shown to be significant.

Bumblebee numbers were significantly different for sample round (t79=-3.59, p=0.001, 95%CL) and landscape (t76= -3.314, p=0.001, 95%CL). Rerunning the model with country as a fixed factor changed the results. The standard deviation of the site was low at <0.001 therefore not likely to affect the fixed factors. There were significant differences in landscape (p=0.002, t 81= -3.153, 95%CL), sample round (p=0.001, t81 = -3.394,) and also several countries were significantly different from the control country which was Austria. Catalunya (p=0.001, t81=-3.488, 95% CL), Italy (p=0.043, t81=-2.060, 95%CL), Spain (p=0.014, t81=-2.513, 95%CL) and the UK (p=0.002, t81=3.266, 95% CL). Germany was proven to not be significantly different from Austria (P=0.392, t81=-0.861, 95%CL) (figure 4.3.3).

Figure 4.3.2: The number of bumblebees per landscape. Error bars indicate ± 1SE (t81=-3.153, p=0.002, 95%CL).

Figure 4.3.1: The number of bumblebees per sample round. Error bars indicate ± 1SE (t81=-3.394, p=0.001, 95%CL).

Figure 4.3.3: The mean number of bumblebees per country, error bars represents ± 1 SE.

The number of other bees (bees that were not honeybee or bumblebees) were taken and used in a mixed effect generalised linear model (glmmPQL). The model was initially run with the random factors of country and site. The standard deviation for country was 0.968 showing a large variance that could potentially affect the model output. The model was then rerun with country as a fixed factor. This second model had site as a random factor with a low standard deviation of (<0.001), the fixed factors that were significant were sample (p=0.003, t81=-3.069, 95%CL), Catalunya (p=0.002, t81=-3.29, 95%CL), Italy (p=0.001, t81=-3.502, 95%CL), Spain (p=0.001, t81=-3.464, 95%CL) and UK (p=0.007, t81=-2.756, 95%CL). As with the number of bumblebees the only country that is not significantly different from Austria is Germany (figure 4.3.4).

Figure 4.3.4: The mean number of other bees per country. Error bars represent ±1 SE.

Regression analysis was carried out on the number of bumblebees and latitude. In order to meet the assumptions the number of bumblebees had to be transformed using a log+1 transformation. There was a significant positive relationship between the number of bumblebees and the latitude (F1 90=60.79, p<0.001, r2=0.403, 95%CL).

Regression analysis was also carried out on the number of other bees and the latitude, the same transformation as on the bumblebee regression had to be carried out (log+1). There was no significant relationship between the number of other bees and latitude (F1 90=0.187, p=<0.667, r2=0.002, 95%CL). Figures 4.3.5 and 4.3.6 show the relationships.

Figure 4.3.5: The positive relationship between the number of bumblebees and latitude (F1 90=60.79, p<0.001, r2=0.403, 95%CL).

Figure 4.3.6: The positive relationship between the number of other bees and latitude (F1 90=0.187, p=<0.667, r2=0.002, 95%CL)

4.4: Size

Table 4.4.1:The mean size (mm) of Halictus for each of the variables

Size-Landscape

Natural

Disturbed

1.601143

1.49175

Size-Country

Germany

Spain

1.61625

1.458857

Size-Temperature

Cold

Hot

1.50275

1.588571

For size the model showed that the standard deviation within the random factor of country to be small at <0.0001. None of the fixed factors were significant; landscape and temperature experience no significant differences in the size of Halictus present (95% CL). Table 4.4.1 shows the mean size (mm) for each variable.

4.5: Individual country analysis

In order to determine in the local environment was having an impact within the separate countries statistical analysis was performed for each of the countries to determine the effect. The assumptions of homoscedasticity, normality of within group error and normality of random effect were all checked to ensure the model was appropriate. All of the models contained landscape, temperature and sample round as fixed factors and site for the random effect. A generalised linear mixed effect model (glmm.PQL) was run for genera, individuals and diversity for each of the countries. Figures of significant results can be seen in the appendix B.

For Spain the standard deviation estimate of the site was moderately low at 0.318 for the number of individuals showing that there is a small variance from the mean for individuals in relation to site. For genera the standard deviation of site was low at 0.192 with little variation. There was very little variance from the mean for the diversity scores and site with the standard deviation estimation for diversity very low at 0.083.

With regard to the fixed factors for the number of individuals and diversity none were significant. For the number of genera in Spain landscape was a significant factor with there being significant variations in the number of genera between disturbed and natural sites (t10= 3.295, p=0.008, 95%CL).

In Italy the standard deviation estimate for sites were very small (individuals, <0.001, genera <0.001 and diversity <0.001) with almost no variation in the sites. The only significant factors were related to the number of individuals. There was not a significant difference in the number of individuals in the different temperatures (t10 = -1.517, p=0.160, 95%CL). There were significant differences in the number of individuals between landscape (t10= 3.333, p=0.008, 95%CL) and in the variation of number of individuals with sample round (t10= -2.467, p=0.033, 95%CL).

For Catalunya the estimates for the standard deviation of the fixed factor site were low in all of the models (individuals= 0.183, genera =<0.001, diversity =<0.001). For diversity there was no significant difference for temperature, landscape or sample size. The number of individuals did significantly vary between sample rounds (t10 =2.525, p=0.030, 95%CL) but not between temperatures and landscape. The number of genera also varied just between sample rounds (t10=-4.847, p=0.007, 95%CL).

In Austria for the random effect of site there is very little variation from the mean with individuals (=<0.001), genera (=<0.001) and diversity (=0.001). There were no significant fixed factors for any of the four models.

For Germany the estimate of standard deviation of site for the individuals was low at 0.193, for genera it was also low at 0.153 showing little variation from the mean. The standard deviation of site for diversity was extremely low at <0.001 for both models. Of all the models only the model for individuals showed any significant factors, a significant difference in individuals between sample round (t10=-2.966, p=0.014, 95%CL).

In the United Kingdom the estimate of the standard deviation for site is low with individuals (<0.001), this is also the case for genera (<0.0001), diversity (<0.0001). Temperature, landscape and sample round are not significant in any of the models for the UK.

4.6: Principal component analysis

Figure 4.6.1: Loading plot from the principal component analysis

Figure 4.6.2: Score plot from the principal component analysis

Principal component analysis was carried out on the data in an attempt to understand the underlying data structure. The principal components do not explain a lot of the variation with the first component explaining just 9.3% of the variance (eigen value 2.232) and the second component explaining 6.5% (eigen value 1.551).

Principal component one shows a positive correlation with latitude, Lasioglossum, Halictus, Andrena, Sphecodes and Hoplosmia. These are the genera which are most abundant out of the whole dataset. Principal component two has a positive correlation with latitude and sample round and a negative correlation with Panurginus, Panurgus and Osmia. These genera are less frequently abundant in the data and are found in the sites with lower latitudes such as Spain and Catalunya.

Figure 4.6.1 Shows the loading plot produced from the principal component analysis. Studying the variables and their ‘loading’ on the components can reveal patterns in the data with variables close together having similarities. Sample is the only variable that appears to have a strong loading on the second component. Lasioglossum and Halictus are isolated from the rest of the variables and show a very strong loading on the first principal component and a neutral component on the second component, honeybees, Sphecodes, Andrena, Hoplosmia, Hylaeus and temperature also have a loading on the first component but not as strongly as Lasioglossum and Halictus.

Latitude has a strong loading on the first and second components. Panurgus and Panurginus have negative loadings on both components.

Figure 4.6.2 shows the scores of the principal component analysis separated into countries. The outliers were identified in an attempt to further understand the data.

The outliers that have a strong negative score on both the first and second components were from the Spanish sites. Analyses of the Spanish outliers show that they all have the genera Panurgus and Panurginus present. All of these outliers were found in the natural landscape.

The outliers from Germany all contained a number of genera that were not commonly abundant, for example Hoplosmia, Hylaeus, Ceratina and Andrena were found to be present at these outlier points. The most extreme of these outliers had a number of different genera, for example the furthest outlier was in the disturbed landscape in a hot area and was found to have four different genera, Lasioglossum (13 individuals), Halictus (8), Sphecodes (5) and Hylaeus (1).

All of the Austrian outliers were found in the natural sites three out of the four were sited in the hot areas and they all had high numbers of Halictus and Lasioglossum (furtherest outlier, 23 Halictus and 3 Lasioglossum). The outlier that was the cold areas of the landscape contained 11 Lasioglossums, 5 Halictus’ and 1 Andrena.

The Catalan outliers were not greatly separated and were very close to the Spanish outliers, like the Spanish outliers, they all had Panurgus or Panurginus present.

There was one UK sample that was slightly detached from the rest; this point contained 9 Bombus and no other individuals.

5: Discussion

5.1: Sample round

Sample round was found to be significant in determining the total number of individuals, the number of bumblebees and the number of other bees. On an individual country level sample round was also found to affect the numbers of individuals found in Italy, Catalunya and Germany.

There were gaps of two to three weeks between the sampling efforts; this difference in sampling round could have several different causes. Bee fauna is very diverse with a high proportion of rare species making up the communities (Williams et al 2001) one factor resulting from the high and diverse number of individuals is that the life history and foraging activities are not well known (Kearns et al 1998). As little is known about bee activity and it is not fully researched as to what individual species are particularly more abundant when and for what reasons.

Due to the mutualistic relationship between plants and pollinators oligolectic species are going to be abundant when their host plant is in bloom (McIntyre and Hostetler 2001) therefore the species present in the traps will depend on whether the sampling time coincides with the plant bloom. However if the plant bloom is particularly large this may influence the specimens collected as ‘real’ flowers will attract a number of bees reducing the number of bees collected in the pan traps (Minkley et al 1999). The climate will also affect the species caught in the pan traps. Different genera have different active seasons and dramatic changes can be experienced within short periods of time (Minckley et al 1994, Williams et al 2001).

Different coloured pan traps can attract different species/genera and even different sexes (Leong and Thorp 1999), as this study used all three pan trap colours the traps used should not greatly affect the outcome of the study.

5.2: Landscape

Of the variables tested it is landscape that was shown to be the most significant factor. For the overall models landscape was shown to cause variation only in the number of bumblebees with more bumblebees found in disturbed landscapes than in natural landscapes. Mean number of bumblebees in disturbed landscapes was 4 compared to 2 for the natural sites.

The species richness of bee communities tends to be higher in natural areas with diversity increasing with the proportion of natural habitat in the proximity (Dauber et al 2003, Hendrickx et al 2007, Steffan-Dewenter et al 2002). However if has been noted that although this is the pattern on average it is not always the case and pollinator diversity as a reaction to landscape can vary (Ricketts et al 2008). The higher level of diversity can be attributed to the higher numbers of floral resources available in the natural areas (Tscharntke et al 1998). A 4km by 4km ‘disturbed’ site will still contain some floral resources although not as many as would be expected in the natural areas.

As bumblebees are polylectic, generalists, they may be able to capitalise on the floral resources available in the disturbed sites. Bumblebees are generally large sized bees and are able to forage at further distances away from their nesting site (Westphal et al 2006) a study by Osborne et al (2001) showed that bumblebees do not necessary forage close to their nest often flying beyond the nearest available forage. Therefore it is possible that bumblebees in a disturbed landscape can capitalise on their foraging distance and generalist capabilities.

The numbers of genera were significantly different between landscapes in Spain and the numbers of individuals were significantly different between landscapes in Italy. As these observations were not observed in the overall model or in any of the other countries it is likely that factors specific to the location of the site are influencing the differences.

In Spain the mean number of genera in the disturbed site was 2 compared to the natural site where the mean number of genera was 4. The Spanish natural site contains a higher mean number of genera, the land use in the Spanish natural site consists of a large area of agro-forestry alongside areas of coniferous and sclerophyllous vegetation.

The mean number of individuals in the Italian disturbed site was 11 compared to the natural site which had 21. The disturbed site contained a large proportion of non-irrigated arable land, complex cultivation patterns, discontinuous urban fabric and vineyards. The surrounding landscape was a matrix consisting of the same land use types with no semi-natural areas nearby.

The pattern of genera in landscapes in Spain and individuals in Italy follows the findings of Dauber et al (2003), Hendrickx et al (2007) and Steffan-Dewenter et al (2002) that bee diversity is related to the proportion of natural habitat in the surrounding areas.

5.3: Temperature and size

In order to fly bees need to raise their body temperature above the critical temperature (Bishop and Armbruster 1999), the ability to reach and exceed this critical temperature is directly related to body size (Casey and Joos 1983). Bumblebees are better thermoregulators than smaller bees, this thermoregulation ability can help or hinder their foraging activity (Heinrich 1974).

Temperature did not significantly affect any of the factors tested. The local temperature does not significantly affect the number of all individuals, bumblebees, other bees or the number of genera. The temperature differences in the landscape are slight enough to not be significant in affecting the ability of bees to fly. Bumblebees are better thermoregulators (Bishop and Armbruster 1999) so are able to achieve flight easier than smaller bees so in colder sites it would be likely to see higher proportions of bumblebees than smaller other bees.

For the Halictus measured there were no factors that caused significant differences in size between them. Country differences, landscape differences and size differences were not found. As larger individuals are better at thermoregulation it would be expected that larger individuals may be found in colder areas and at higher latitudes with a cooler climate, this was not the case for the size of Halictus.

Overall the number of bumblebees and the number of other bees differed between countries with the most bumblebees found in the most northerly country (UK). There is a positive relationship between the number of bumblebees and latitude supporting that bumblebee numbers increase with cooler climates (Peat et al 2005) due to better thermoregulation. Although there is no relationship between the number of individuals, the number of genera, the number of bumble and the number of other bees and the local temperature there is a significant relationship between bumblebees and latitudes.

5.4: Diversity

Diversity was found to not be significantly different between landscape, temperature or country. Simpson’s 1-D was used to describe the diversity, it calculates the probability that two samples will belong to the same species but does not give any account of the differences in composition.

Principal component analysis can give some insight into the underlying patterns and indications of the composition. The first principal component indicates high latitudes are correlated with high numbers of Lasioglossum and high numbers of Halictus. The second component suggests that high latitudes are correlated with low numbers of Panurgus and Panurginus.

Panurgus and Panurginus belong to the subfamily Panurginae. Panurgus is a typical Mediterranean genus (Patiny et al 2005) and therefore would be associated with the lower latitudes and found in sites in Spain and Catalunya, as suggested by the principal component analysis. Although Halictus and Lasioglossum were found in abundance throughout the different countries large numbers of Lasioglossums and Halictus were found in Germany and Austria which is high latitude.

5.5: Climate change

Climate change is resulting in the change of geographical range for many species (Gates 1993) with distributions shifting upwards and northwards (Hickling et al 2006). With bee distributions being forced to move northwards in the face change barriers could be experienced that increase the risk of extinction (Thomas et al 2004). Landscape significantly affected the number of individuals in Italy and the number of genera in Spain. As species distributions shift northwards they may have to leave an area high in semi-natural and natural habitats and move to a disturbed area which if contains less floral resources will affect the pollinators (Steffan-Dewenter et al 2002).

Barriers such as the landscape structure and artificial fragmentation (i.e. disturbed landscapes) will affect species migration and thus make it impossible for species to keep up with the changing climate (Pearson and Dawson 2003). Different species were associated with different latitudes, Panurgus and Panurginus were shown to be associated with lower latitudes and the number of bumblebees positively related with the latitude, more bumblebees the higher the latitude, as the latitudes become warmer smaller bees such as solitary may be able to compete and gain a foraging advantage (Heinrich 1974) leading to a changing composition and declining bumblebees in the higher latitudes.

The Panurgus and Panurginus which are associated with warmer latitudes may extend their range northwards however as the majority of the two genera are found in Spain and Catalunya the Pyrenees may act as a barrier preventing the migration (Pearson and Dawson 2003). Cane et al (2006) noted that the specialists will be greatest at risk from extinction. Oligolectic bees are therefore at a higher risk of extinction, the survival of the bees will also depend on the survival of the host plant. If the plant and bees are not able to respond at the same rate it could provide fatal for one, if not both of the species.

5.6: Limitations and considerations:

When examining the results of this study there are some considerations to be aware of. Due to time restrictions only six sites across Europe were able to be analysed, the countries were distributed throughout Western Europe. It is important to note when making comparisons that there will be inevitable variations in climate/weather conditions, the vegetation and the surrounding landscape. Although great care was taken when picking the sites variations between them will be inevitable.

This study looked at bee genera rather than individual species, some genera are much larger than others containing a large number of individual species. Responses to temperature and landscape therefore may masked by observing genera rather than species. Due to time and identification skills, most genera contain both oligolectic and polylectic species (Wcislo and Cane 1996).

An extension of this project could look at the individual species and also other features of the local environment that could affect the diversity, for example rainfall, slope, aspect and humidity.

6: Conclusion

The local environment appears to have a relatively small impact on the bee diversity of an area. The local temperature does not impact bee numbers or the number of genera present with the temperature not being low enough to impact flight availability. The body size of the specimens measured showed no difference between country, landscape or temperature indicating that for Halictus the body size is not significantly affected by the local environment.

The local landscape does not affect the number of individuals, genera of other bees but does affect the number of bumblebees. More bumblebees were found in the disturbed site that in the natural sites, against what was expected. This may be down to bumblebees being able to maximise on their generalist feeding habits and their ability to forage further distances. Individual countries had different patterns for the number of genera and number of individuals which fell in line with the existing research that the diversity of bees’ increases with the proportion of semi-natural habitats in the surrounding areas.

Climate change could have serious implications for community composition. Oligolectic species are more vulnerable to extinction as they depend on one host plant; however bumblebees, which are generalists, are associated with higher latitudes so warmer temperatures could see their ranges diminish. Survival for some genera will mean having to extend their range past abiotic and biotic barriers.

Acknowledgments

Many thanks go out to Koos Biesmeijer, Stuart Roberts for the identification help, Elaine and Terry Hawes and Andrew Lawrence. Thanks to Terry Hawes for the photos.

References:

Biesmeijer JC, SPM Roberts, M Reemer, R Ohlemüller, M Edwards, T Peeters, AP Schaffers, SG Potts, R Kleukers, CD Thomas, J Settele and WE Kunin (2006) Parallel Declines in Pollinators and Insect-Pollinated Plants in Britain and the Netherlands: Science: 313: 351-354.

Brosi BJ, Daily GC and PR Ehrlich (2006) Bee community shifts with landscape context in a tropical countryside: Ecological Applications: 17: 418-430

Cane JH, Minckley RL, Kervin LJ, Roulston TH, NM Williams (2006) Complex responses within a desert bee guild (Hymenoptera: Apiformes) to urban habitat fragmentation: Ecological Applications 16:632-644

Casey TM and BA Joos (1983) Morphometrics, conductance, thoracic temperature and flight energetics of noctuid and geometrid moths: Physiological Zoology: 56:160-173

Caswell H (1976) Community structure: a neutral model analysis. Ecological Monographs: 46: 327-354

Connell JM (1978) Diversity in tropical rainforests and reefs: Science: 199: 1302-1310.

Danforth B (2007) Bees: Current Biology: 17: 156-161

Dauber J, Hirsch M, Simmering D, Waldhardt R, Otte A and V Walters (2003) Landscape structure as an indicator of biodiversity: Matrix effects on species richness. Agriculture, Ecosystems and Environment: 98:329

Delaplane KS & DF Mayer (2000) Crop pollination by bees. CABI publishing, Oxon.

Gates DM(1993) Climate change and its biological consequences. Sinauer Associates Inc. Sunderland, Massachusetts

Heinrich (1974) Thermoregulation in endothermic insects. Science: 185: 747-756

Heinrich (1976) Resource partitioning among some eusocial insects: Bumble bees. Ecology: 57: 874-889

Hendrickx F, Maelfait JP, Van Wingerden W, Schweiger O, Speelmans M, Aviron S, Augenstein I, Billeter R, Bailey D, Bukacek R, Burel F, Diekӧtter T, Dirksen J, Herzog F, Liira J, Roubalova M, Vandomme V and R Bugter (2007) How landscape structure, landuse intensity and habitat diversity affect components of total arthropod diversity in agricultural landscapes. Journal of Applies Ecology: 44: 340-351.

Hickling R, Roy DB, Hill JK, Fox R and CD Thomas (2006) The distributions of a wide range of taxonomic groups are expanding polewards: Global Change Biology 12: 450-455

Kearns CA, Inouye DW and NM Waser (1998) Endangered mutualisms: The conservation of plant-pollinator interactions: Annual Review Ecology Systematics: 29:83-112

Kremen C, Williams NM, Aizen MA, Gemmill-Herren B, LeBuhn G, Minckley R, Packer L, Potts SG, Roulston T, Steffan-Dewenter I, Vazquez DP, Winfree R, Adams L, Crone EE, Greenleaf SS, Keitt TH, Klein AM, Regetz J and TH Ricketts (2007) Pollination and other ecosystem services provided by mobile organisms: A conceptual framework for the effects of land use change: Ecology letters: 10: 299-314

Leong JM and RW Thorp (1999) Colour-coded sampling: The pan trap colour preferences of oligolectic and non-oligolectic bees associated with vernal pool plant: Ecological Entomology: 24: 329-335

Magurran AE (2004) Measuring biological diversity. Blackwell, Malden USA.

Margalef R (1968) Perspectives in Ecological Theory. University of Chicago Press, Chicago p111.

McGregor SE (1976) Insect pollination of cultivated crop plants. USDA Agricultural handbook 496. Washington DC: US Department of Agriculture: Available online at http://gears.tucson.ars.ag.gov/book/index.html accessed 04/03/2008

McIntyre NE and ME Hostetler (2001) Effects of urban landuse on pollinator (Hymenoptera: Apoidea) communities in a desert metropolis: Basic and Applied Ecology: 2: 204-218

Minckley RL, Wcislo WT, Yanega DA and SL Buchmann (1994) Behaviour and phenology of a specialist bee (Dieunomia) and sunflower (Helianthus) pollen availability. Ecology: 75: 1406-1419

Minckley RL, Cane JH, Kervin L and TH Roulston (1999) Spatial predictability and resource specification of bees (Hymenoptera:Apoidea) at a superabundant wide spread resource: Biological Journal of the Linnean Society: 67:119-147

National Research Council of the National Academies (2007) The status of pollinators in North America: National Academies Press, Washington DC.

Patiny S, Oritiz-Sánchez FJ and D Michez (2005) A review of Panurgus (Pachcephalepanurgus), with the description of a new species from Spain: Zootaxa: 1037: 37-48

Parmesan C and GA Yohe (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature: 421: 37-42

Pearson RG and Dawson TP (2003) Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Global Ecology and Biogeography:12 :361-371

Peat J, Darvill B, Ellis J and D Goulson (2005) Effects of climate on intra- and interspecific size variation in bumble-bees: Functional Ecology: 19: 145-151

Petraitis PS, Latham RE and RP Niesenbaum (1989) The maintenance of species diversity by disturbance. Quarterly review of biology: 64:393-418

Pimentel D, Stachow U, Takars DA, Brubaker HW, Dunmas AR and JJ Meaney (1992) Conserving biological diversity in agricultural/forestry systems. Most biological diversity exists in human managed ecosystems. Bioscience: 42:354-362.

Ricketts TH, Regetz J, Steffan-Dewenter I, Cunningham SA, Kremen C, Bogdanski A, Gemmill-Herren B, Greenleaf SS, Klein AM, Mayfield MA, Morandin LA, Ochieng A and BF Viana (2008) Landscape effects on crop pollination services: are there general patterns? Ecology letters: 11: 499-515

Roubik DW(1995) Pollination of cultivated crops in the tropics: Food and Agricultural Organisation. Rome

Samways MJ (2005) Insect diversity and Conservation: Cambridge university press, Cambridge,

Steffan-Dewenter I, Munzenberg U, Burger C, Thies C and T Tscharntke (2002) Scale-dependent effects of landscape context on three pollinator guilds: Ecology: 83: 1421-1432

Steffan-Dewenter I and K Leschke (2003) Effects of habitat management on vegetation and above ground nesting bees and wasps of orchard meadows in central Europe: Biodiversity and Conservation: 12: 1953-1968

Steffan-Dewenter I, Potts SG and L Packer (2005) Pollinator diversity and crop pollination services are at risk: Trends in ecology and evolution: 20: 651-652

Stoutjesdijk PH and JJ Barkman (1992) Microclimate, vegetation and fauna, OPULUS press AB, Sweeden.

Tscharntke T, Gathmann A and I Steffan-Dewenter (1998) Bioindication using trap nesting bees and wasps and their natural enemies; community structure interactions: Journal of Applied Ecology: 35: 708-719

Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YJ, Erasmus BFN, Ferreira de Siqueira M, Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA, Townsend Peterson A, Phillips OL and SE Williams (2004) Extinction risk from climate change: Nature: 427: 145-148

Unwin DM and SA Corbet (1991) Insects, plants and microclimate. Naturalists handbooks 15. Richmond Publishing group, Slough.

Wcislo WT and JH Cane (1996) Floral resource utilisation by solitary bees (Hymenoptera: Apoidea) and exploitation of their stored foods by natural enemies: Annual Review of Entomology: 41:257-286

Westphal C, Steffan-Dewenter I and T Tscharntke (2003) Mass flowering crops enhance pollinator densities at a landscape scale. Ecology letters: 6: 961-965

Williams IH (1994) The dependence of crop production within the European Union on pollination by honey bees: Agricultural Zoology Reviews: 6: 229-257

Westphal C, Steffan-Dewenter I and T Tscharntke (2006) Bumblebees experience landscapes at different spatial scales: possible implications for coexistence: Oecologia: 149: 289-300

Williams NM, Minckley RL, and FA Silveira (2001) Variation in native bee faunas and its implications for detecting community changes. Conservation Ecology 5:7. [online] URL: http://www.consecol.org/vol5/iss1/art7/

Zayed A and L Packer (2007) The population genetics of solitary oligolectic sweat bees Lasioglossum (Sphecodogastra) oeriotherae (Hymenoptera:Halictidae): Heredity: 99: 397-405

Zhang W, Ricketts TH, Kremen C, Carney K and SM Swinton (2007) Ecosystem services and dis-services to agriculture. Ecological Economics: doi:10.1016/j.ecolecon.2007.02.024

Appendix A: Bee Genera and authorities identified.

The genera identified from the samples

Order

Family

Subfamily

Genus

Authority

Hymenoptera

Andrenidae

Andrenidae

Andrena

Fabricius 1775

Hymenoptera

Megachilidae

Megachilinae

Anthidium

Spinola 1838

Hymenoptera

Apidae

Apinae

Anthophora

Latreille 1803

Hymenoptera

Apidae

Apinae

Apis

Linnaeus 1758

Hymenoptera

Apidae

Apinae

Bombus

Latreille 1802

Hymenoptera

Apidae

Xylocopinae

Ceratina

Latreille 1802

Hymenoptera

Apidae

Apinae

Eucera

Latreille 1810

Hymenoptera

Halictidae

Halictinae

Halictus

Latreille 1804

Hymenoptera

Megachilidae

Megachilinae

Heriades

Linnaeus 1758

Hymenoptera

Megachilidae

Megachilinae

Hoplosmia

Kirby 1802

Hymenoptera

Colletidae

Hylaeinae

Hylaeus

Fabricius 1793

Hymenoptera

Halictidae

Halictinae

Lasioglossum

Curtis 1833

Hymenoptera

Megachilidae

Megachilinae

Megachile

Latreille 1802

Hymenoptera

Megachilidae

Megachilinae

Osmia

Panzer 1806

Hymenoptera

Andrenidae

Panurginae

?Panurginus

Giraud 1861

Hymenoptera

Andrenidae

Panurginae

Panurgus

Kirby 1802

Hymenoptera

Halictidae

Anthophorinae

Sphecodes

Latreille 1804

Hymenoptera

Megachilidae

Megachilinae

Stelis

Nylander 1848

Hymenoptera

Halictidae

Rophitinae

Systropha

Scopoli 1770

Mean number of individuals per sample round in Catalunya.( t10= 2.525, p=0.030, 95%CL)

Appendix B: Figures from the significant country results.

Mean number of individuals per sample round in Germany

Mean number of individuals per sample round in Germany.( t10=-2.966, p=0.014, 95%CL)

Mean number of genera per sample round in Catalunya.( t10=-4.847, p=0.007, 95%CL)

Mean number of individuals by sample round in Italy.( t10=-2.467, p=0.033, 95%CL)

Mean number of genera by landscape in Spain.( t10=3.295, p=0.008, 95%CL)

Mean number of individuals by landscape in Italy.( t10=3.333, p=0.008, 95%CL)

Mean number of genera by landscape in Spain

Mean number of individuals by landscape in Italy.

Writing Services

Essay Writing
Service

Find out how the very best essay writing service can help you accomplish more and achieve higher marks today.

Assignment Writing Service

From complicated assignments to tricky tasks, our experts can tackle virtually any question thrown at them.

Dissertation Writing Service

A dissertation (also known as a thesis or research project) is probably the most important piece of work for any student! From full dissertations to individual chapters, we’re on hand to support you.

Coursework Writing Service

Our expert qualified writers can help you get your coursework right first time, every time.

Dissertation Proposal Service

The first step to completing a dissertation is to create a proposal that talks about what you wish to do. Our experts can design suitable methodologies - perfect to help you get started with a dissertation.

Report Writing
Service

Reports for any audience. Perfectly structured, professionally written, and tailored to suit your exact requirements.

Essay Skeleton Answer Service

If you’re just looking for some help to get started on an essay, our outline service provides you with a perfect essay plan.

Marking & Proofreading Service

Not sure if your work is hitting the mark? Struggling to get feedback from your lecturer? Our premium marking service was created just for you - get the feedback you deserve now.

Exam Revision
Service

Exams can be one of the most stressful experiences you’ll ever have! Revision is key, and we’re here to help. With custom created revision notes and exam answers, you’ll never feel underprepared again.