Institutionen för fysik, kemi och biologi
Examensarbete 16 hp
Management effects on butterfly and bumblebee abundance in Swedish semi-natural grasslands
Kristoffer Dahl
LiTH-IFM- Ex--15/3035--SE
Handledare: Per Milberg and Malin Tälle, Linköpings universitet Examinator: Anders Hargeby, Linköpings universitet
Institutionen för fysik, kemi och biologi
Linköpings universitet
Rapporttyp Report category Examensarbete C-uppsats Language English Titel/Title:
Management effects on butterfly and bumblebee abundance in Swedish semi-natural grasslands
Författare/Author: Kristoffer Dahl Sammanfattning/Abstract:
The amount of semi-natural grasslands has decreased in the agricultural landscape and because of this many grassland species are declining. Semi-natural grasslands are dependent on management, but different studies suggest that different management methods are most favorable to species richness and abundance. The aim of this study was to compare the effects of mowing and grazing on butterfly and bumblebee abundance in Swedish semi-natural grasslands. We used data collected through the Swedish environmental monitoring program NILS. We identified 31 comparable meadows and chose 1-10 pastures in the area surrounding each meadow. For six different species groups we calculated the number of individuals found per 100 m. By standardizing the values of the difference between the means for meadows and pastures we were able to compare the effects of the two methods on abundance in each species group. No species group showed a preference for pastures. A significant preference for meadows was found for two butterfly species groups. The four other groups showed a trend to prefer meadows before pastures. According to our study the recommended management method should be mowing, considering protection of butterflies and bumblebees.
ISBN
LITH-IFM-G-EX—15/3035—SE
__________________________________________________ ISRN
__________________________________________________ Serietitel och serienummer ISSN
Title of series, numbering
Handledare/Supervisor Per Milberg, Malin Tälle
Ort/Location: Linköping
Nyckelord/Keyword:
Bumblebees; butterflies; grazing; management; mowing; semi-natural grasslands
Datum/Date
2015-06-10 Department of Physics, Chemistry and
Biology
Avdelningen för biologi
Table of content
1 Abstract ... 2
2 Introduction ... 2
3 Material & Methods ... 4
3.1 Study sites and sampling ... 4
3.2 Identification of grassland objects ... 5
3.3 Classification ... 5
3.4 Statistical analysis ... 6
4 Results ... 6
5 Discussion ... 7
5.1 Abundance in butterfly and bumblebee species ... 7
5.2 Mowing versus grazing ... 8
5.3 Other factors to consider ... 8
5.4 Social and ethical aspects ... 9
6 Conclusions ... 9
7 Acknowledgements ... 9
1 Abstract
The amount of semi-natural grasslands has decreased in the agricultural landscape and because of this many grassland species are declining. Semi-natural grasslands are dependent on management, but different studies suggest that different management methods are most favorable to species richness and abundance. The aim of this study was to compare the effects of mowing and grazing on butterfly and bumblebee abundance in Swedish semi-natural grasslands. We used data collected through the Swedish environmental monitoring program NILS. We identified 31 comparable meadows and chose 1-10 pastures in the area surrounding each meadow. For six different species groups we calculated the number of individuals found per 100 m. By standardizing the values of the
difference between the means for meadows and pastures we were able to compare the effects of the two methods on abundance in each species group. No species group showed a preference for pastures. A significant preference for meadows was found for two butterfly species groups. The four other groups showed a trend to prefer meadows before pastures. According to our study the recommended management method should be mowing, considering protection of butterflies and bumblebees.
2 Introduction
In Åbo archipelago, Finland, the management of semi-natural grasslands changed during the 1950’s. Due to intensification of agriculture,
fertilization or abandonment semi-natural grasslands were turned into arable land or less diverse shrub land and forest. It was no longer
economically feasible to manage the grasslands using grazing or mowing (Kotiluoto 1998). This is an example of what happened in all of Europe at the time (Benton et al. 2002, Robinson & Sutherland 2002, Kadlec et al. 2010). As a consequence of this, semi-natural grasslands is a habitat type declining in both area and amount, which can have serious effects on the biota of all trophic levels (Pettersson et al. 2013, Simons et al. 2014). Furthermore, the ecosystem services (e.g. pollination) provided by the semi-natural grasslands and the organisms in them, are being
jeopardized (Kremen et al. 2007).
It is commonly considered that semi-natural grassland habitats were created and maintained by the megaherbivores. After the extinction of megaherbivores, approximately 12 000 years ago, humans have taken their ecological role in grasslands, through cattle grazing and/or mowing for hay in order to provide the farm animals with food during the winter (Pykälä 2000).
The two most commonly used management methods are mowing and grazing. However, few studies have examined their effect on biodiversity and abundance of semi-natural grasslands, and the ones that exist have contradicting results. One of the most studied species groups linked to semi-natural grasslands are butterflies and moths (Saarinen & Jantunen 2005, Bergman et al. 2008, D’Aniello et al. 2011, Skórka & Lenda 2011). Loos et al. (2014) have for instance shown that low-intensity grasslands in Transylvania, Eastern Europe, contain a high butterfly diversity. However, this study concluded that an intensification and increase in modern agriculture can seriously threaten the butterfly fauna. More industrialized countries have already lost many valuable semi-natural grasslands, leading to declining butterfly populations (Robinson & Sutherland 2002, Van Dyck et al. 2009, Pettersson et al. 2013), which could affect the pollination of grain and their function as pest-controllers (Skórka & Lenda 2011). Another ecologically important species group due to e.g. their pollination of grain is bumblebees (Goulson et al. 2005). Like the butterflies, the bumblebee populations are declining in Europe (Kosior et al. 2007) and the rest of the world (Williams & Osborne 2009). For instance Bommarco et al. (2012) showed a severe shift in the
composition of bumblebee species in Swedish red clover fields, and several other studies have showed a dramatic decline in bumblebees worldwide (Goulson et al. 2005, Kosior et al. 2007, Goulson et al. 2008, Cameron et al. 2011).
Previous studies have examined the effect of mowing and grazing on butterflies and bumblebees.Saarinen & Jantunen (2005), for example, found no significant difference in the effect of mowing and on the
butterfly diversity in eight mown grasslands (meadows) and eight grazed grasslands (pastures) in north-western Russia/north-eastern Finland. Smith & Cherry (2014) found a significantly higher diversity and species richness in mowed compared to grazed grasslands. In addition to species richness and diversity they also found a significant higher abundance in fields mowed only every third year compared to fields mowed or grazed every year. In contrast, D’Aniello et al. (2011) found a significantly higher butterfly abundance and diversity in the grazed grasslands in southern Italy. In addition, WallisDeVries & Raemakers (2001) showed that butterfly abundance in eleven low-intensity grazed grasslands in the Netherlands were higher in comparison to four mowed objects. They also showed that several threatened butterfly species benefited from grazing and not from mowing. Carvell (2002) showed that cattle grazing were essential in maintaining a species rich vegetation, and crucial for several bumblebee species, although only in comparison with no management or
sheep grazing. The effects of mowing on bumblebees have not been studied.
In the balance between profitable and biologically sustainable agriculture the political control and available resources must be directed as
effectively as possible. The capital invested in environmental protection is often very limited which makes it very important that the money is spent in the best way to ensure the species richness and abundance remain high and stable. Today, in Sweden, more money is received by the farmers who manage using mowing instead of grazing. However, if studies show that grazing is more favorable as management method, then grazed grassland should receive more economic subsidies.
Because of the great need of protection of the traditional grassland habitats where many threatened species exist and because of the lack of studies in the subject, the aim of this study was to investigate effects of different management methods on butterfly and bumblebee fauna in Swedish semi-natural grasslands. Even if there is a lack of studies in the subject, there is a lot of potential data. From the Swedish environmental monitoring program NILS we studied if the management method could contribute to different effects on butterfly and bumblebee abundance, and which management method is most favorable.
3 Material & Methods 3.1 Study sites and sampling
The National Inventory of Landscapes in Sweden (NILS) is an environmental monitoring program started in 2003 by the Swedish Environmental Protection Agency, with the goal of monitoring of the landscapes in Sweden. For this program 600 squares (5×5 km2 each) have been evenly allocated across Sweden. In these squares, 696 grassland objects have been randomly selected.
The grassland objects were visited once every fifth year, and during this year each object is visited four times. During visit one, two and three the butterfly fauna was surveyed. During visit four the bumblebee fauna was surveyed. Each object is divided into several transects. During the survey every butterfly species within 10 m and every bumblebee species within 4 m of the transect is noted. In addition to species data, information on the grass sward height, management and temperature were noted. (Swedish University of Agricultural Sciences 2014, Swedish University of
3.2 Identification of grassland objects
For the present study data from 2006 to 2014 was used. We identified 69 meadows (i.e. grasslands managed by mowing). NILS classifies
grasslands into six categories according to the management status. We classified grasslands as meadows if they were mowed this year (category 4) or mowed earlier years (category 5) If the management status was classified as ongoing grazing (category 1) or grazing unsure (category 2) in any of the visits that year, all data from that point forward was
disqualified as meadow data. If the management status was unclear, for example if the management was classified as no grazing (category 3) or un-continued management (category 6), complementary data from the database TUVA was used to determine an objects management status. The TUVA database comes from a national survey executed in 2002-2004 by the Swedish Board of Agriculture (Persson 2005) and contains information on meadows and pastures in Sweden. Data from the latest year of visit were always used except if the management changed from the first year of visit to category 1 or 2 the second year of visit, as mentioned above. Then data from the first year of visit were used. For each meadow object all pasture objects in the same and in the eight surrounding large NILS-squares were identified. To qualify as a pasture object the management status had to be classified as 1 or 2 and the grazing animal had to be cattle. Cattle are the most common grazing animal used, and also the most interesting in an agricultural context. For the pasture objects, data from the year of visit closest in time to the corresponding meadow were used.
Data for all visits (during one year of visit) and for all transects were combined and used in the analysis. One meadow object was compared to 1-10 pasture objects (depending on the number of pasture objects in the eight surrounding squares). The meadows without any nearby pasture objects were disqualified from further analysis. That left us with 31 comparable meadow objects.
3.3 Classification
The species were divided into eight groups: 1) European butterfly grassland indicator (ButterflyEU) (Biała 2013), 2) Grassland specialist butterflies according to Bergman1 (ButterflySpec.) 3) Red-listed
1
Bergman, K-O. Landscape Mediated Patterns of Species Richness for Butterflies in Southern Sweden. In prep.
butterflies (ArtDatabanken 2015), 4) All butterfly species
(ButterflyTotal), 5) Declining bumblebees (BumblebeeDecline) (Botham et al.2013), 6) Bumblebees bound to open fields (BumblebeeOpen) (Svensson et al.2000), 7) Red-listed bumblebees (ArtDatabanken 2015) and 8) All bumblebee species (BumblebeeTotal). The occurrence of red-listed species (both butterflies and bumblebees) was too low to enable analysis.
3.4 Statistical analysis
The mean number of individuals per 100 m per object was calculated for each species group. The two means (i.e. one for the meadow object and one for the pasture object(s)) were compared by subtraction, resulting a difference value for each meadow and species group (a total of 6 species groups × 31 meadows = 186 values). The values for the differences were then standardized using the mean and standard deviation of all values. Standardizing the values of the difference between the means make comparison of different species groups easier to interpret. A positive value indicates a higher abundance in the meadows whilst a negative value indicates a higher abundance in the pastures.
4 Results
The standardized difference between the means was positive for every species group. Consequently, there were more individuals per 100 m in the meadows compared to the pastures for all species groups. For the European butterfly grassland indicators and for all butterflies we found a significantly positive standardized difference between the means (Figure 1). Note that by standardizing the values, the standard deviation and the confidence interval were the same for all species groups: StD = 1 and CI = 0,352.
Figure 1. Confidence interval shown for the standardized values of six species groups: declining bumblebees, bumblebees bound to open fields, all
bumblebee species, European butterfly grassland indicators, grassland specialist butterflies and all butterfly species. If the value is positive the abundance is higher in the meadows whilst a negative value indicates a higher abundance in the pastures. A significant positive difference between the means was found for the European butterfly grassland indicators and for all butterfly species (marked with asterisks).
5 Discussion
5.1 Abundance in butterfly and bumblebee species
From our results we can see that no species group had higher abundance in the grazed grasslands. We can also see a significant positive difference between means in the groups European butterfly grassland indicators and all butterfly species. Butterflies seem to prefer meadows to a higher extent than bumblebees. Bumblebees although shows a trend to also prefer meadows before pastures. The results suggest an overall preference for meadows compared with pastures. This is consistent with results from Smith & Cherry (2014) who found low intensity mowing (every third year) to be the most favorable to butterfly abundance, and that annual mowing was more favorable for species richness and diversity than grazing. Any previous studies on the effects of mowing on bumblebees are not to be found. However, our results are not in line with several other studies: D’Aniello et al (2011) and WallisDeVries & Raemakers (2001) both found grazing to be more favorable for butterfly abundance and diversity. The difference in effects on bumblebees are not well studied but Carvell (2002) showed a preference for cattle grazed grasslands over sheep grazed grasslands and grasslands without management.
0 1 2 3 4 5 6 7 -1,5 0 1,5 Standard deviation BumblebeeDecline BumblebeeOpen BumblebeeTotal ButterflyEU* ButterflySpec. ButterflyTotal *
However, our data have limitations. We can, for example, not say
anything about the grazing pressure in the pastures or the time of mowing in the meadows, which might also affect the results. This must be taken in to account when interpreting the results.
5.2 Mowing versus grazing
Because of the scarce and contradictory results from studies in this subject it is difficult to know the reason behind our results. However, there are some theories. One theory is that grazing reduces the amount of nectar and pollen throughout the whole summer while mowing, if
executed in the late summer, reduces the amount of nectar and pollen late in the season, and therefore is more favorable for nectar-seeking
organisms like butterflies and bumblebees (Saarinen & Jantunen 2005). This could be one of the reasons for the observed higher abundance of butterflies in mowed grasslands. Intensive grazing also has a negative impact on both abundance and biodiversity in several organism groups (Robinson & Sutherland 2002, Cousins 2009, Jugovic et al. 2013). For both butterflies and bumblebees the flower abundance is important, which is dependent on, among other factors, the seed dispersal. Some studies have shown that grazing is more favorable for seedling and seed dispersal (Rico et al.2011, Auffret et al. 2012, Marteinsdóttir 2014), but mowing could also benefit seed dispersal (Bakker et al. 1996). Low-intensity grazing can result in a mosaic grassland, due to the selective way cattle grazes the land (Jugovic et al. 2013). This will create patches with higher grass sward height, which can benefit many butterfly species (Bergman et al. 2008, Skórka & Lenda 2011), contradicting the result in our study.
5.3 Other factors to consider
In the decision of which management method to use in a grassland more factors than the ones our study looked at must be considered. The
butterfly and bumblebee species richness could be a good indicator of the total species richness in a grassland, but other species dependent on semi-natural grasslands that are not included in this study e.g. other insects, birds and plants, needs to be considered. For this study we compared mowing and cattle grazing. However, we have no information on the grazing pressure nor time of mowing. These factors are important to consider, as mowing during the butterflies flight period could be
catastrophic (D’Aniello et al. 2011) and high-intensity grazing could lead to very species-poor grasslands (Hahn et al. 2015). Furthermore, mowing and grazing are two quite different and demanding management methods. Keeping animals for grazing is both time-consuming and expensive for the farmer whilst mowing could be done in a few days (depending on the
area) each year, or even as rarely as every second or third year, according to Smith & Cherry (2014). Mowing could also have great costs, though. Depending on mowing method, area and location you have to be able to afford different types of tools, machines and man-power. It is also important to consider the different conditions in different parts of the world. For example, the optimal management method in Sweden might not be the optimal in Italy. The conditions can also differ within a country, for example in a wet versus a dry grassland. In addition, the management history of the grassland could affect the optimal
management today2.
5.4 Social and ethical aspects
With an increased level of knowledge in this subject we will be able to direct the political control and economic resources to get a more effective protection of semi-natural grassland habitats and the species in them. This study will therefore be of great relevance to farmers and policymakers. We are also increasing the knowledge in agricultural ecology which could lead to an increase in economical funds available for
environmental protection. For example, as our results reveal that
meadows are expected to contain more butterflies and bumblebees than pastures, the higher subsidies to mowed grasslands are justified.
Furthermore, preservation of ecosystem services is very important, and sustainable agriculture is very important for social, economic and environmental aspects.
6 Conclusions
The optimal management method depends on several different factors. In this study we looked at the difference in abundance of butterflies and bumblebees between mowed and grazed grasslands, and found that the abundance of both species groups are higher in mowed grasslands. When deciding the optimal management method in a grassland object it is however important to consider other organisms, geographical conditions, historical management as well as economic factors.
7 Acknowledgements
I am grateful to my supervisors Per Milberg, and Malin Tälle for support during the data processing, analysis and writing phase. I also want to
2
Tälle M., Deák B., Poschlod P., Valkó O., Westerberg L., Milberg P. Grazing vs. mowing: a meta-analysis of biodiversity benefits for grassland management. In prep.
thank Erik Cronvall and Åsa Eriksson at the Swedish University of Agricultural Sciences for data providing and information on the NILS-project.
8 References
ArtDatabanken 2015. Rödlistade arter i Sverige 2015. ArtDatabanken SLU, Uppsala
Auffret G., Schmucki R., Reimark J., Cousins S.A.O. (2012) Grazing Networks Provide Useful Functional Connectivity for Plants in
Fragmented Systems. Journal of Vegetation Science: Official Organ of the International Association for Vegetation Science 23: 970-977. Bakker J.P., Poschlod P., Strykstra R.J., Bekker R.M., Thompson K. (1996) Seed Banks and Seed Dispersal: Important Topics in Restoration Ecology \S. Acta Botanica Neerlandica 45: 461-490.
Benton T.G., Bryant D.M., Cole L., and Crick H.Q.P. (2002) Linking Agricultural Practice to Insect and Bird Populations: A Historical Study over Three Decades. The Journal of Applied Ecology 39: 673–87. Bergman K-O, Ask L., Askling J., Ignell H., Wahlman H., Milberg P. (2008) Importance of Boreal Grasslands in Sweden for Butterfly Diversity and Effects of Local and Landscape Habitat Factors. Biodiversity and Conservation 17: 139–53.
Biała K. (2013) The European Grassland Butterfly Indicator: 1990–2011. European Environment Agency.
Bommarco R., Lundin O., Smith H.G., Rundlöf M. (2012) Drastic Historic Shifts in Bumble-Bee Community Composition in Sweden. Proceedings. Biological Sciences / The Royal Society 279: 309-315. Botham MS, Brereton TM, Middlebrook I., Randle Z, Roy D.B. (2013) UKBMS Annual Report 2012.
Cameron S.A., Lozier J.D., Strange J.P., Koch J.B., Cordes N., Solter L.F., Griswold T.L. (2011) Patterns of Widespread Decline in North American Bumble Bees. Proceedings of the National Academy of Sciences of the United States of America 108: 662-667.
Carvell C. (2002) Habitat use and conservation of bumblebees (Bombus spp.) under different grassland management regimes. Biological
Conservation 103: 33-49
Cousins, S.A.O. (2009) Landscape History and Soil Properties Affect Grassland Decline and Plant Species Richness in Rural Landscapes. Biological Conservation 142: 2752–58.
D’Aniello B., Stanislao I., Bonelli S., Balletto E. (2011) Haying and Grazing Effects on the Butterfly Communities of Two Mediterranean-Area Grasslands. Biodiversity and Conservation 20: 1731-1744.
Goulson D., Hanley M.E., Ellis J.S., Knight M.E. (2005) Causes of Rarity in Bumblebees. Biological Conservation 122: 1-8.
Goulson D., Lye G.C., Darvill B. (2008) Decline and Conservation of Bumble Bees. Annual Review of Entomology 53: *****.
Hahn M., Schotthöfer A., Schmitz J., Franke L.A., Brühl C.A. (2015) The Effects of Agrochemicals on Lepidoptera, with a Focus on Moths, and Their Pollination Service in Field Margin Habitats. Agriculture,
Ecosystems and Environment 207: 153-162.
Jugovic J., Črne M., Pečnikar Z.F. (2013) The Impact of Grazing,
Overgrowth and Mowing on Spring Butterfly (Lepidoptera: Rhopalocera) Assemblages on Dry Karst Meadows and Pastures. Natura Croatica: Periodicum Musei Historiae Naturalis Croatici = Casopis Hrvatskoga Prirodoslovnog Muzeja 22: 157–69.
Kadlec T., Vrba P., Kepka P., Schmitt T., Konvicka M. (2010) Tracking the Decline of the Once-Common Butterfly: Delayed Oviposition,
Demography and Population Genetics in the Hermit Chazara Briseis. Animal Conservation 13: 172–83.
Kosior A., Celary W., Olejniczak P., Fijal J., Kró, Wieslaw L., Wojciech S., Plonka P. (2007) The Decline of the Bumble Bees and Cuckoo Bees (Hymenoptera: Apidae: Bombini) of Western and Central Europe. Oryx: The Journal of the Fauna Preservation Society 41. Cambridge University Press: 79–88.
Kotiluoto R. (1998) Vegetation Changes in Restored Semi-Natural Meadows in the Turku Archipelago of SW Finland. Plant Ecology 136 53-67.
Kremen C., Williams N.M., Aizen M.A., Gemmill-Herren B., LeBuhn G., Minckley R., Packer L. (2007) Pollination and Other Ecosystem Services Produced by Mobile Organisms: A Conceptual Framework for the Effects of Land-Use Change. Ecology Letters 10: 299-314.
Loos J., Dorresteijn I., Hanspach J., Fust P., Rakosy L., Fischer J. (2014) Low-Intensity Agricultural Landscapes in Transylvania Support High Butterfly Diversity: Implications for Conservation. PloS One 9 (7): e103256.
Marteinsdóttir B. (2014) Seed Rain and Seed Bank Reveal That Seed Limitation Strongly Influences Plant Community Assembly in
Grasslands. PloS One 9 (7): e103352.
Persson K. (2005) Ängs- Och Betesmarksinventeringen 2002-2004: Rapport 2005:1. Swedish Board of Agriculture.
Pettersson L., Nilsson S.G., Franzén M. (2013) Land-Use Changes, Farm Management and the Decline of Butterflies Associated with Semi-Natural Grasslands in Southern Sweden. Nature Conservation 6: 31-48.
Pykälä J. (2000) Mitigating Human Effects on European Biodiversity through Traditional Animal Husbandry. Conservation Biology: 14: 705-712.
Rico Y., Boehmer H.J., Wagner H.H. (2011) Determinants of Actual Functional Connectivity for Calcareous Grassland Communities Linked by Rotational Sheep Grazing. Landscape Ecology 27: 199-209.
Robinson R. A., Sutherland W.J. (2002) Post-War Changes in Arable Farming and Biodiversity in Great Britain. The Journal of Applied Ecology 39: 157–76.
Saarinen K., Jantunen J. (2005) Grassland Butterfly Fauna under Traditional Animal Husbandry: Contrasts in Diversity in Mown
Meadows and Grazed Pastures. Biodiversity and Conservation 14: 3201-3213.
Simons N.K., Gossner M.M., Lewinsohn T.M., Boch S., Lange M., Müller J., Pašalić E. (2014) Resource-Mediated Indirect Effects of Grassland Management on Arthropod Diversity. PloS One 9: e107033.
Skórka P, Lenda M. (2011) Abandoned Fields as Refuges for Butterflies in the Agricultural Landscapes of Eastern Europe. In Insect Habitats: Characteristics, Diversity and Management, 83–103. Nova Science Publishers, Inc.
SLU (2014) Fältinstruktion För Fjärilar, Humlor, Grova Träd Och Lavar I Ängs- Och Betesmarker NILS. Sveriges lantbruksuniversitet,
Institutionen för skoglig resurshushållning.
SLU (2015) Nationell Inventering Av Landskapet I Sverige. Sveriges Lantbruksuniversitet. May 6. http://www.slu.se/sv/centrumbildningar-och-projekt/nils/mer-om-nils1/.
Smith L.M., Cherry R. (2014) Effects of Management Techniques on Grassland Butterfly Species Composition and Community Structure. The American Midland Naturalist 172: 227–35.
Svensson B. Lagerlöf J., Svensson B.G. (2000) Habitat Preferences of Nest-Seeking Bumble Bees (Hymenoptera: Apidae) in an Agricultural Landscape. Agriculture, Ecosystems and Environment 77: 247–55. van Dyck H., van Strien A.J., Maes D., van Swaay C.A.M. (2009) Declines in Common, Widespread Butterflies in a Landscape under Intense Human Use. Conservation Biology: The Journal of the Society for Conservation Biology 23: 957–65.
WallisDeVries M.F., Raemakers I. (2001) Does Extensive Grazing Benefit Butterflies in Coastal Dunes? Restoration Ecology 9: 179-188. Williams P.H., Osborne J.L. (2009) Bumblebee Vulnerability and Conservation World-Wide. Apidologie 40: 367-387.
