Climatic Risk and 
Distribution Atlas of 
European Bumblebees 

PIERRE RASMONT, MARKUS FRANZEN, THOMAS LECOCQ, ALEXANDER HARPKE, 
STUART P.M. ROBERTS, JACOBUS C. BIESMEIJER, LEOPOLDO CASTRO, 
ByJORN CEDERBERG, LIBOR DvorAK, UNA FITZPATRICK, YVES GONSETH, 
Er1c HAUBRUGE, GILLES MAHE, AULO MANINO, DENIS MICHEZ, 
JOHANN NEUMAYER, FRODE @DEGAARD, JUHO PAUKKUNEN, 
TADEUSZ PAWLIKOWSKI, SIMON G. Potts, MENNO REEMER, 

JOSEF SETTELE, JAKUB STRAKA & OLIVER SCHWEIGER 

Dedicated to 

Astrid Loken, Bruno Pittioni, William FE Reinig & Anton S. Skorikov 
who pioneered bumblebee biogeography 

John Heath & Jean Leclercq 

founders of the European Invertebrate Survey - 
Cartographie des Invertébrés Européens - Erfassung der Europaischen Wirbellosen 



Climatic Risk 

and Distribution 
Atlas of European 
Bumblebees 

by 

PIERRE RASMONT, MARKUS FRANZEN, THOMAS LECOCQ, 
ALEXANDER HARPKE, STUART P.M. ROBERTS, JACOBUS C. 
BIESMEIJER, LEOPOLDO CASTRO, BJORN CEDERBERG, 
Lrpor DvoRAk, UNA FITZPATRICK, YVES GONSETH, 
Er1Ic HAUBRUGE, GILLES MAHE, AULO MANINO, 
DENIS MICHEZ, JOHANN NEUMAYER, FRODE @DEGAARD, 
JUHO PAUKKUNEN, TADEUSZ PAWLIKOWSKI, 
SIMON G. POTTS, MENNO REEMER, JOSEF SETTELE, 
JAKUB STRAKA & OLIVER SCHWEIGER 

Biorisk 10 (Special Issue) 

> PENSOFT. 

2015 


ep The research leading to these results has received funding from the 

@ a European Community's Seventh Framework Programme (FP7/2007- 

4: 2013) under grant agreement no 244090, STEP Project (Status and 
Trends of European Pollinators, www.step-project.net) 

CLIMATIC RISK AND DISTRIBUTION ATLAS OF EUROPEAN BUMBLEBEES 

Citation: Rasmont P., Franzén M., Lecocq T., Harpke A., Roberts S.PM., Biesmeijer J.C., Castro L., 
Cederberg B., Dvorak L., Fitzpatrick U., Gonseth Y., Haubruge E., Mahé G., Manino A., Michez D., 
Neumayer J., Odegaard FE, Paukkunen J., Pawlikowski T., Potts S.G., Reemer M.., J. Settele, J. Straka, 
Schweiger O. (2015) Climatic Risk and Distribution Atlas of European Bumblebees. Biorisk 10 (Special 
Issue), 246 pp. 

Front cover: Bombus hyperboreus, an Arctic bumblebee species that is threatened by global warming. 
© Photo: Goran Holmstrém. 

Disclaimer: The views expressed in this publication are those of the authors and do not necessarily reflect 

the views or opinions of the funders or reviewers. 

Biorisk 10 (Special Issue) 
ISSN 1313-2644 (print) 
ISSN: 1313-2652 (online) 
doi: 10.3897/biorisk. 10.4749 

First published 2015 
ISBN: 978-954-642-768-7 (hardback) 
ISBN: 978-954-642-769-4 (e-book) 

Pensoft Publishers 

12, Prof. Georgi Zlatarski St. 
1700 Sofia, Bulgaria 

e-mail: info.pensoft.net 

www.pensoft.net 

All content is Open Access, distributed under the terms of the Creative Commons Attribution License 
(CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided 

that the original author and source are credited. 

Design by WPENSUFT. 
Printed in Bulgaria, January 2015 


Climatic Risk and Distribution Atlas of European Bumblebees 

TABLE OF CONTENTS 

fh Alpe. Ofc O TiC UES a 98 oop run he ne ress ee pal acne Aosta eae od 5 
2 ANC tT OW LER COIMETIES, 0 he neheces pecan edie Berta an tt sete oatnen Drees ier ca dthcotath lataientderheetoutedy 6 
BUR GEE WOE te svc coteceesve denseaanevonneresn value meen ah oragemeon cap tasandnvs cans exter dent raven ce esaroe oat 7 
EG Elna) <4 Bel en ey Sat RA WAR OUR. Once aL faith nA Ark aie Ree ean. kOe haa RRR en! 8 
pve Unt aya thts, ako) «an amit tia ORME etn ES ROR REI A OP RERE 11 
6. Methodology set betes tenswchnestetle Mea Medel bth erha leet Atandaleah eet rehtshsvreen Aaah geetee the 
7+ Ghecklistot the: European. bumblebee species vescevercevssazea sanvederecgiancessoaonrse 27 
8. Climatic risks of Bore pearl burmnblebets:).1...5..8 ace at aceayetearrertldeieonenotay 30 
SaNonzmedelled*urdpean, bumblebees peciesien. sep: ansaendnonnteraion ea 144 
NO ctaenietal Pacterinshok Leite stishey ew os cacs te cburhcetneteatiosneente tatu teen cbaeietine’ Le 
lala Methodoldgical lmaitatt ons cits tired teeter ecteiala hesicde tos ten Becereds ban sineees 159 
|L§2208 BS Tey ches col otis sth) geen er Mee, al on en AUR See ee ER een nya ot J oeebatn 169 
13 G@limate.chaneeand bumblebeeiconsetvation,..2.0b ecw Reed diated: 173 
VAP COSTICIUS TONS ROAR AY alee ROR 5 MERA #8 OLRE GMO Red EDS ae, ot bal Sree yiavan oct 179 
VD in RETEREINCES exci vite een veptres wesivsaise Sas oes Vow beh baal eo edad wgromessnrs teas eoncear bea te rota 181 
NGAP DOMCICeS M72) t aim! ka REAR ane ie Rind ot Behn e nl oe Re Eat canta 199 
17. Distribution maps of West-Palavearctic bumblebees..............:::eeeeeeeees 212 
(Ronee yOl nai 1e ig GRR e Mee Re Rae Retr en Caney SEAL RIA dy Oe Va ata Re de hid 234 

TOTES OL AIL LOTS ere sa Patra Meter idee ttre stall, onda tts berncdt tern Wt te tegates 236 


Climatic Risk and Distribution Atlas of European Bumblebees 

2. Acknowledgements 

We especially thank Y. Barbier and S. Iserbyt (Belgium) for their help in data 
management, to P. Stoev and L. Penev (Bulgaria) for their help in the layout and 
R. De Jonghe (Belgium) for his numerous pieces of advice and his great experience of 
bumblebee behaviour and ecology. 

We thank a lot all the people who kindly and promptly provided their data: 

J. Gokcezade, A. Aichhorn. T. Kiipper and numerous entomologists from Austria; J. 
D’Haeseleer, J. Devalez, K. Janssens, D. Laget (Belgium); V.L. Kazenas (Bulgaria); 
P. Bogusch, (Czech Republic); I. Calabuig (Denmark); G. Sdderman, I. Teras 
(Finland); S. Gadoum, X. Lair, D. Roustide, P. Stallegger (France); H.J. Fliigel, C. 
Schmid-Egger, N. Schneider (Germany); Y.J. Anagnostopoulos, T. Petanidou (Greece); 
M. Cornalba, E. Gabriele, E Intoppa, G. Pagliano; M.G. Piazza, M. Quaranta (Italy); 
E. Budrys (Lithuania); E Feitz, N. Schneider (Luxembourg); I. Peeters, F van der 
Meer, J. Verhulst (and numerous entomologists from EISN) (The Netherlands); 
W. Celary, W.E. Dramstad, A. Kosior, W. Solarz (Poland); D. Baldock (Portugal); 
M.V. Berezin, N. Filipov, T.V. Levchenko, Z. Yefremova (Russia); A. Gogala, A. Jeni¢ 
(Slovenia); P. Sima, V. Smetana (Slovakia); J.R. Obeso, EJ. Ortiz (Spain); M. Larsson, 
Bo Séderstrém, M. Stenmark (and numerous entomologists from SSIC) (Sweden); 
A. Miller, M.K. Obrist (Switzerland); A.M. Aytekin (Turkey); D.W. Baldock, M. 
Edwards, G.R. Else (and numerous entomologists from BWARS), D. Goulson (UK); 
I. Konovalova, V. Radchenko (Ukraine). 

The following persons kindly authorised the reproduction of their pictures: 

J.-S. Carteron (France), G. Holmstrém (Sweden), J.-C. Kornmilch (Germany), O. 
Korsun (Russia), G. Millet (France), A.G. Maldonado (Spain), M. McGlinchey (UK), 
J. Michailowski (Belgium), A. Pauly (Belgium), G. Pisanti (Israel). 

The research leading to this book has received funding from the European Commission's 
Seventh Framework Programme (FP7/2007-2013) under grant agreement no 244090, 
STEP Project (Status and Trends of European Pollinators, www.step-project.net). It also 
received funds from the BELSPO - Belgian Research Action through Interdisciplinary 
Networks (BRAIN) BELBEES project (www.belbees.be). 

Most of the pictures by P. Rasmont have been made possible thanks to Fonds de la 
Recherche Scientifique (Belgium) travel grants, to the Eyne Municipality (France, 
A. Bousquet, R. Staats), to a funding from the European Commission's Seventh 
Framework Programme (FP7/2007-2013) project INTERACT, and to efforts and 
personal contributions of R. De Jonghe (Belgium). 


Climatic Risk and Distribution Atlas of European Bumblebees 

3. Foreword 

Pollinators are increasingly recognised as providing a vital ecosystem service, not least 
for feeding people, and bumblebees are among the most important pollinators in 
north temperate regions like Europe. Like most animals, bumblebees are sensitive to 
climate, in part through their geographically varying interactions with other pressures, 
such as land use and pesticide use. Climate change, for which the evidence is now 
unequivocal, is therefore expected to affect bumblebee distributions across Europe. For 
relatively cool-adapted animals like bumblebees, many of the already evident and likely 
future climate changes are unlikely to be good news. This may be especially challenging 
if constraints on the ability of bumblebees to spread to keep up with climate changes 
will make it difficult for them to compensate in terms of distribution extent by moving 
into new areas of Europe. This atlas, considering the likely effects of climate change 
on bumblebees in Europe, is therefore a timely and vital work. It is an important 
complement to the earlier Red List of Threatened Species for the IUCN Bumblebee 
Specialist Group, in which the BBSG Regional Coordinators Pierre Rasmont and 
Stuart Roberts also took leading roles. 

The challenge is one of dealing with very complex systems. Even if we know which 
service or function is needed in a changing world, we may not always be able to 
predict precisely which species will best be able to carry out that role as the system 
changes. Therefore while we can try to target efforts on the currently most critical 
species, it is also important to conserve the diversity of species, as an insurance against 
unpredictable outcomes from complex systems in which unexpected species prove to 
become the most important in the future. 

This atlas breaks new ground in assessing the most likely consequences of climate 
change for these important pollinators in Europe. The prognosis is shown to be dire. 
But it should be an inspiration and a stimulus to encourage people to look urgently 
at similar model projections in other, often less well studied, parts of the world. ‘The 
clock is ticking and we need to see progress in all regions, before it becomes too late 
for some species. 

Paul Williams 
Chair, IUCN Bumblebee Specialist Group 


Climatic Risk and Distribution Atlas of European Bumblebees 

4. Context 
4.1. An overview of the bumblebees 

Bumblebees are amongst the most familiar insects inhabiting meadows, gardens, 
and grasslands of the temperate regions of the World. They have long been popular 
with field biologists and naturalists thanks to their bright colours, large body size, 
and abundance. Bumblebees (genus Bombus) are insects closely related to honey 
bees, stingless bees, cuckoo bees, carpenter bees and orchid bees which together 
constitute the family Apidae within the order Hymenoptera (Michener, 2000). 
Today, approximately 250 species assigned to 15 subgenera are recognized world- 
wide (Williams, 1998; Williams et a/., 2008). Most of the bumblebees are eusocial 
species while few of them are socially parasitic bees (i.e. inquiline species; the 27 
species included in the subgenus Psithyrus and two other species). Like other Apidae, 
the bumblebees are recognised as pollinator species (Neff & Simpson, 1993). Among 
animal pollinators of the Northern Hemisphere, only few achieve such a numerical 
dominance as flower visitors as bumblebees. This makes bumblebees a critically 
important functional group providing ecosystem services for natural environments 
and for agricultural crops (e.g. Free, 1993; Klein et al., 2007). 

Although the distribution of bumblebees encompasses a wide geographic range from 
Arctic tundra to lowland tropical forest, they are clearly most abundant in mountain 
habitats and cold and temperate regions of the Northern Hemisphere (Williams, 
1998). Indeed, these robust hairy bees have thermoregulatory adaptations involving 
facultative endothermy (Heinrich, 1979), that enable them to live in the coldest areas 
inhabited by insects. Thanks to these adaptations, bumblebees have been able to 
recolonise areas depopulated by Ice Ages in the last three million years (Hines, 2008). 
However these adaptations to cold climate raise the question of what will be the fate of 
bumblebees under current global warming. Investigating this question requires large 
biogeographic databases which, until recently, have been unavailable. 

4.2. Advances in the study of bumblebee biogeography 

The late 19th century and the first half of the 20th century brought the first modern 
biogeographical studies based on species mapping (for bumblebees see e.g. Reinig, 1937, 
1939; Pittioni, 1938, 1942, 1943). However, these first studies were based on highly 
uncertain geographic locations which strongly limited the potential usefulness of these 
first biogeographic data. Advances in geographic localisation methods and instruments 
in the second half of the 20th century have tremendously increased the quality of 
distribution data and led to the development of the first biogeographic databases. In 
the early 1970's, the foundation of the European Invertebrate Survey - Cartographie des 
Invertébrés Européens - Erfassung der Europdischen Wirbellosen launched the mapping of 


Context 

insects from Europe (Heath & Leclercq, 1969). However, the technological limitations 
of the 1970's allowed only few pioneer results at the continental level (Heath & Leclercq, 
1981). Once again, technological advances, especially in micro-computing, database 
management and geographical information systems allowed further progress by gathering 
huge amounts of data. The rapid increase in the accessibility of the modern technology 
has allowed collecting an unprecedented number of biogeographical data by numerous 
professional and citizen scientists in many countries. 

This has led to many high level studies in several groups of organisms (e.g. Tutin et al., 
2001; Settele et a/., 2008; Kudrna et a/., 2011). In contrast, for bees, despite the great 
interest of biologists and the wider society, the complexity of the taxonomy considerably 
delayed the establishment of a database. In this context, the European Union FP7 
“STEP” project (“Status and Trends of European Pollinators”; www.step-project. 
net; Potts et al., 2011) appeared as one of the first occasions (but see also ALARM 
project; www.alarmproject.net) to make a significant advance in the knowledge of bees 
from the whole European region. The outcome of this European collaboration has 
exceeded the initial expectations. After four years of survey, more than 2.5 million 
species observations have been joined together to map the distributions of the majority 
of European bee species. In August 2014, more than 1200 European bee species have 
been mapped. ‘This extensive mapping has been made available to both the scientific 
and public audiences on the Atlas Hymenoptera website (www.atlashymenoptera. 
net; Rasmont & Haubruge, 2014). A joint effort produced the first comprehensive 
checklist of European bees (Kuhlmann e¢ a/., 2014). Further, a collaboration with the 
IUCN resulted in a first Red List of European bees (Rasmont e¢ a/., 2013). 

Most of the STEP data concerns bumblebees (more than one million bumblebee data 
from all West-Palearctic countries). Thanks to the STEP project, the mapping and the 
IUCN assessments of all European bumblebee species have been published (Rasmont 
et al., 2013; Rasmont & Iserbyt, 2014). This large database can now allow investigation 
of the recent history of bumblebee species. 

4.3. Bumblebee decline and the tomorrow’s bumblebee fauna 

As long ago as the early 1970's, many entomologists pointed out the decline of 
bumblebee species in Europe (Peters, 1972; Williams, 1982; Rasmont & Mersch, 1988; 
Williams et al, 1991, 2007, 2013; Goulson et a/., 2005, 2008b; Rasmont et al., 2005; 
Biesmeijer et al., 2006; Kosior et al., 2007; Williams & Osborne, 2009; Carvalheiro 
et al., 2013). Thank to the advances in the study of bumblebee biogeography, and to 
the sharing of long-term datasets, the comparison of the past and current European 
bumblebee fauna has revealed the scale of the problem (Rasmont & Iserbyt, 2014). 
Moreover this decline is a global phenomenon (e.g. North America, Cameron et al., 
2011; South America, Arbetman et a/., 2013; China, Xie et a/., 2008). 


10 

Climatic Risk and Distribution Atlas of European Bumblebees 

Several hypotheses have been proposed to explain this global decline such as (i) habitat 
fragmentation (Williams, 1982; Williams & Osborne, 2009; Darvill et a/., 2010; Mayer 
et al., 2012; Hatten et al., 2013), (ii) shortage of flower resources (Peters, 1972; Williams, 
1989; Rasmont & Mersch, 1988; Rasmont et al., 1993, 2005; Goulson et al., 2005, 
2008a), (iii) killing by car traffic (Donath, 1986), (iv) overgrazing of bumblebee habitat 
by cattle (Ozbek, 1995; Xie et al., 2008), (vi) parasites and pathogens resulting from 
spillover from domesticated species (Cameron et al., 2011; Arbetman et a/., 2013), (vii) 
urbanization (Ahrné et a/., 2009; Martins et a/., 2013), or (viii) vegetation displacement 
due to nitrogen deposition (Rasmont, 2008). Pesticides have most likely also played a 
role because of their extreme toxicity for some bumblebee species (e.g. Whitehorn et 
al., 2012; Zarevicka, 2013) or to the closely related species Apis mellifera (e.g. Johnson 
et al., 2010, 2013). But their impact remains still largely unevaluated for most of the 
bumblebees. Several alternative factors such as herbicides or helminthicides impacting 
other organism groups could also be a factor in bumblebee decline (Lumaret, 1986; 
Madsen et a/., 1990; Colin & Belzunces, 1992; Vandame & Belzunces, 1998; Hayes 
et al., 2002; Mussen et al., 2004; Simon-Delso et al., 2014). At least one other factor 
is regarded as strongly affecting the bumblebee fauna: the changing climate (Iserbyt & 
Rasmont, 2012; Rasmont & Iserbyt, 2012; Ploquin et a/., 2013; Herrera et al., 2014). 
However, throughout the present work, it should be kept in mind that it is very likely 
that none of the factors potentially explaining bumblebee decline is the unique or 
even the main trigger of current bumblebee regression. As Jeremy Kerr (Toronto, pers. 
comm.) recently wrote “there is no silver bullet that killed the bumblebees”. 

Besides explaining the current decline of bumblebees, their importance in ecosystem 
service provision places a premium on predicting the future of the bumblebee fauna. 
Even if all above cited factors do shape the fate of bumblebees, only the evolution of 
climate change has been assessed thanks to the work of the Intergovernmental Panel on 
Climate Change (www.ipcc.ch). In the present work, we investigate the future for the 
European bumblebee fauna in the light of these climatic change projections. 


Introduction 

5. Introduction 
5.1. Effects of climate change 

Climate is one of the most important determinants of large-scale species distributions 
(Thuiller et a/., 2004). Climate and its changes have shaped the current wild bee 
distribution and biodiversity (e.g. Groom et a/., 2014; for bumblebees see Lecocq et al., 
2013). Likewise, several studies have shown that the current bumblebee decline can be 
attributed to climate change (e.g. Williams et a/., 2007; Bartomeus et al., 2013), which 
can either act via an increasing frequency of extreme events or via gradual changes in 
average conditions. 

Ranta & Vepsalainen (1981) pointed out that a large number of individuals could be 
killed by “catastrophic environmental vicissitudes” prior to a fast population recovery. 
This random cycle of “sudden extermination - fast recovery” is considered as a key 
process of local species diversity, by maintaining all species under a competition level 
(Pekkarinen, 1984; Rasmont, 1989). Recent contributions noticed that bumblebees 
are indeed sensitive to extreme climatic events (Iserbyt & Rasmont, 2012; Ploquin 
et al., 2013; Herrera et al., 2014; Rasmont & Iserbyt, 2014). Year after year, the local 
bumblebee fauna can change (including local species extinctions) due to variations 
in local climatic factors such as heat waves and droughts (Iserbyt & Rasmont, 2012; 
Rasmont & Iserbyt, 2012). However, this seems to drive the species not only along 
a random climatic hazard as proposed by Ranta & Vepsaldinen (1981) but also to a 
temporal and spatial gradient of changes (Ploquin et a/., 2013; Herrera et al., 2014). 

Beside extreme climatic events, gradually changing conditions can also seriously impact 
species (e.g. Parmesan & Yohe, 2003). On the one hand, gradually changing climatic 
conditions can lead to shifts in species ranges which has been observed for many species 
(e.g. Parmesan & Yohe, 2003; Chen eg a/., 2011) including bees (e.g. Kuhlmann et a/., 
2012). On the other hand, the gradual changes can lead to modification of species’ 
phenology (Polgar et a/., 2013; Kharouba et a/., 2014; for wild bees see Bartomeus 
et al., 2011). Indeed, in both cases, species can respond to gradual climate change by 
tracking spatially or temporally their climatic niche (e.g. Tingley et al, 2009; Moo- 
Llanes et al., 2013). 

5.2. Toward a new pollinator community 

There is strong evidence that variations in climatic conditions deeply affect the 
bumblebee fauna (e.g. Williams et a/., 2007; Bartomeus e¢ a/., 2013a; Pradervand 
et al., 2014), and recent projections of expected changes in climatic conditions for 
the 21st century give rise to particular concerns. For instance, the Intergovernmental 
Panel on Climate Change (IPCC) states in its 5th Assessment Report that “a large 


12 

Climatic Risk and Distribution Atlas of European Bumblebees 

fraction of terrestrial and freshwater species face increased extinction risk under projected 
climate change during and beyond the 21st century, especially as climate change interacts 
with other pressures, such as habitat modification, over-exploitation, pollution and 
invasive species (high confidence; IPCC 2014)”. Indeed, bumblebee populations seem 
to be more sensitive to other threats when they reach their climatic limits (Williams 
& Osborne, 2009). Further, there is some indication that future climate change could 
have severe impacts on wild bee faunas (Kuhlmann e¢ a/., 2012) including bumblebees 
(Kirilenko & Hanley, 2007; Herrera et a/., 2014). 

Species-specific responses to future climate change can lead to the generation of new 
communities (e.g. Schweiger et a/., 2010; Lurgi et a/., 2012; Pradervand et a/., 2014) 
with changed functional structures. Indeed, changes in the spatial/temporal occurrence 
of pollinators can lead to spatial gaps/asynchrony between the pollinators and insect 
pollinated plants (Kudo, 2013; Kudo & Ida, 2013; Pradervand et al, 2014). The 
resulting effects could be dramatic for both plants and pollinators (e.g. Kudo & Ida, 
2013; Petanidou et al, 2014), even if several empirical studies suggest that the large 
plant and insect biodiversity could mitigate the expected dramatic consequences (e.g. 
Bartomeus et al., 2011, 2013b; Forrest & Thomson, 2011; Iler et a/., 2013). Such changes 
in pollinator communities may not only affect wild plants but can also impact important 
agricultural crops (e.g. Free, 1993; Klein et a/., 2007). In a study in Britain, Polce et al. 
(2014) found that future climate change can lead to spatial mismatches between orchards 
and their pollinators. This may in turn increase the risks to human society of suffering 
from pollination deficits of economically important crops. 

The impacts of future climate change on the fate of single species, the functioning of 
ecosystems and the sustainable provision of ecosystem services highlights the need for 
efficient assessments of potential future climatic risks for pollinators. So far there is no 
comprehensive assessment of such risks available for Europe or any other continent on 
the world. With this atlas we take the first step for a prominent and important group 
of pollinators — the bumblebees. 

5.3. Objectives of the climatic risk atlas 
The general aims of this atlas are: 

¢ to inform the broader public about the potential risks of climate change for the 
future fate of European bumblebees; 

¢ to aid biodiversity conservation managers and policy makers; 

e to provide background knowledge for critical discussions about the sustainable 
provision of pollination services in the light of food security. 


Methodology 

6. Methodology 
6.1. General approach 

The general approach used in this atlas was to assess the climatic niche of each 
bumblebee species according to its European distribution (from 1970 to 2000) and 
the corresponding climatic conditions. The species-specific climatic requirements 
were then used to project the climatically suitable areas and the corresponding 
changes of these areas under potential future climatic conditions. These future 
conditions were taken from scenarios of climate change which incorporate different 
potential pathways of political, socio-economic and technological development. The 
projected changes in suitable climatic conditions for each species were illustrated on 
a map and assigned to six climate risk categories (see chapter 8). Finally, summary 
statistics about the projected changes and the risk categories were used to provide a 
comprehensive overview about the future climatic risks of the majority of European 

bumblebees (Chapter 10, Appendix 2). 
6.2 Species distribution data 

Species distribution data used for this atlas were collated within the EU FP7 
project STEP (Potts et a/., 2011; http://www.step-project.net) and were published 
in an aggregated way on the website Atlas Hymenoptera (Rasmont & Iserbyt, 
2014; http://www.atlashymenoptera.net/). Original data were kindly provided by 
a vast number of professional and citizen scientists (Tab. 6.1). By 28.12.2014 this 
database had 988,187 observation records for all 69 European bumblebees (for a 
list of species see chapter 7). From this extensive database all records (300,435) 
between 1970 and 2000 within a defined geographical frame (latitude from 35° 
to 72°N; longitude from -12°W to 32°E) were extracted and used in the species 
distribution models. 

6.3 Geographic extent and resolution 

Although the original geographic coverage of the Atlas Hymenoptera data is much 
wider, we restricted the geographic extent of the distribution data to avoid including 
areas with low sampling intensities and thus a likely high proportion of areas where 
a species is falsely assumed to be absent just because it has not been observed. Since 
such false absence data tend to increase with increasingly finer spatial resolutions and 
thus could lead to wrong or biased assessments of the species’ climatic requirements, 
we aggregated the distributional data at a 50 km x 50 km UTM grid to increase the 
reliability of our models (Fig. 6.1). 

13 


14 Climatic Risk and Distribution Atlas of European Bumblebees 

Table 6.1. Major data providers (more than 99.9% data). 

Country/Region | Number of records 

90,053 
21,734 
9,857 
IG.Mahé id France 9,156 
2,551 
1,962 
1,538 
1,070 
M.Cornalba tally 945 

Russia 

Slovenia 
Slovenia 

Portugal 

S. Baile France 

59 
Others —“‘“‘iYSS tt 
Tol C—i—“‘“‘S*C*drLSC‘C$K§WNTCOC‘éCNSC WCCO 88,187 | 


15 

Methodology 

and from 35° to 72° N latitude (Fig. 6.1) and included the whole of Europe and the 

The geographic extent of the considered area ranged from -12° W to 32° E longitude 
northern parts of Morocco, Algeria and Tunisia in Africa. 

: PUT! wee Se rao eam 
Fan See a apehooumuesces. ©) 3g%,0% Segccesgreens. ~soceaeeh 
. ae 99.8 @bdaner Th pad Ae oe 6 ee" “eas 
ad, ©6 COSC RCC OE OOWaheO gee: : " tn dhe 
QU i ovocvscccesoosessoccdael — 2 60 
©000008500080060008«¢ 6 6 6 . ee) e00 
sc eeocosgacoeaseseset | ooF Se eee coon ass 
he gaounl®) Balnce eosoned © tone cesegooesees 
ooeseoom, uy jeaecons 
ee eee 00% gocebascoen Do a hee Seccsooess 
; eo) 228659000009 
@6evooovees owe de sbososeessom, 
ofrdee seoovecenns ] ‘ \ndeprp petty kU 
Cen see ceeoeeeee he pesosoocoes! 
le one ooeece 
ASP ececese én ® 
ee 090 Coeccencetsose ) 
Se della CO FC SF SSSR CODSSS C0000 , Peeecooes 
Ce YITITIVI Tt ae te 

Figure 6.1. The study area and all 50 km x 50 km UTM grids used in the species distribution modelling 

represented by a dot. 


16 

Climatic Risk and Distribution Atlas of European Bumblebees 

6.4 Current climate data 

The climatic niches of bumblebees were modelled using monthly interpolated climate 
data (New e¢ al., 2000; Mitchell et a/., 2004) aggregated to the same 50 km x 50 km 
UTM grid as was used for the species distribution data. Mean values of the following 
27 climate variables (absolute values and annual variations) for the period 1971-2000 
were considered for the analysis of the climatic requirements of the bumblebees: 

e annual temperature (°C); 

¢ temperature seasonality (calculated as the range between hottest and coldest 
month; °C); 

¢ quarterly temperature (e.g. March - May = spring; °C); 

e quarterly temperature seasonality (°C); 

e diurnal temperature range per year (°C); 

e diurnal temperature range per quarter (°C); 

¢ annual precipitation sum (mm); 

¢ precipitation seasonality (calculated as the range between wettest and driest 
month; mm); 

e quarterly summed precipitation (mm); 

e quarterly precipitation seasonality (mm); 

* annual cloudiness (%); 

e quarterly cloudiness (%). 

Climatic variables, especially those measuring similar entities such as for instance 
mean annual temperature and mean summer temperature, are often highly correlated 
and their information content is thus highly redundant. Such collinearities among 
environmental variables can cause problems for the assessment of the climatic 
requirements of single species (Dormann et al., 2013). To avoid such biases we selected 
ecological relevant and least correlated variables by means of cluster analysis. The 
threshold for variable selection was a Pearson correlation coefficient lower than 0.3 

(Graham, 2003). 

The selected variables to assess the climatic requirements of each bumblebee species 
were: 

¢ mean annual temperature (Fig. 6.2); 

¢ annual precipitation sum (Fig. 6.3); 

¢ temperature seasonality (reflecting continentality ; Fig. 6.4); 
¢ precipitation seasonality (reflecting oceanity; Fig. 6.5). 


Methodology 

3 

Temperature (°C) 

Figure 6.2. Mean annual temperature. (a) Current conditions (1971-2000); (b, c, d) future conditions for 
2050; (e, f, g) future conditions for 2100; (b, e) moderate change scenario (SEDG); (c, f) intermediate 
change scenario (BAMBU); (d, g) severe change scenario (GRAS); (h) Boxplot of temperature conditions 
and projected changes across all 10 min x 10 min grid cells of the selected geographic window for 2050 
and 2100 under three climate change scenarios. The black bar within the box represents the median 
value; box boundaries show the interquartile range. Whiskers show data points that are no more than 
1.5 times the interquartile range on both sides. Open circles identify outliers. Horizontal dashed line 
shows the median value for current conditions. 

17 


18 Climatic Risk and Distribution Atlas of European Bumblebees 

h 1400 

mm) 
—_, — 
(am) pe) 
[o) oO 
oO fo} 

6 800 

= 600 
ao 8 

9 400 
[yy asz0i-81z 8 

a 

DD) 612.01 - 775 a 
(775.01 - 1030 e 
(1030.01 - 1400 

DH) 1400.01 - 2060 
ti By > 2060 

Figure 6.3. Annual precipitation sum. (a) Current conditions (1971-2000); (b, c, d) future conditions for 
2050; (e, f, g) future conditions for 2100; (b, e) moderate change scenario (SEDG); (c, f) intermediate 
change scenario (BAMBU); (d, g) severe change scenario (GRAS); (h) Boxplot of precipitation conditions 
and projected changes across all 10 min x 10 min grid cells of the selected geographic window for 2050 
and 2100 under three climate change scenarios. The black bar within the box represents the median 
value; box boundaries show the interquartile range. Whiskers show data points that are no more than 
1.5 times the interquartile range on both sides. Open circles identify outliers, note that extreme outliers 
have been cutt off for means of better visualisation. Horizontal dashed line shows the median value for 
current conditions. 


Methodology 

30 

h 
O 25 
eae 
> 
= 20 
© 
Cc 
QB 15 
ae : 
a peris 8, 
| 1501-18 
J] 18.41: -20 
a *- © D Hh © D DH 
L | 20,01 -22 Pg 2s @ 2 = 
(ny 2.01 -2 g 3 8 z 
Sal > 2 2050 2100 

Figure 6.4. Temperature seasonality. (a) Current conditions (1971-2000); (b, c, d) future conditions for 
2050; (e, f, g) future conditions for 2100; (b, e) moderate change scenario (SEDG); (c, f) intermediate 
change scenario (BAMBU); (d, g) severe change scenario (GRAS); (h) Boxplot of temperature seasonality 
and projected changes across all 10 min x 10 min grid cells of the selected geographic window for 2050 
and 2100 under three climate change scenarios. The black bar within the box represents the median 
value; box boundaries show the interquartile range. Whiskers show data points that are no more than 
1.5 times the interquartile range on both sides. Open circles identify outliers. Horizontal dashed line 
shows the median value for current conditions. 

19 


20 

Climatic Risk and Distribution Atlas of European Bumblebees 

LJ 38.01 - 52 
ees 0" 70 
7) 70.01 -95 

Seasonality (mm) >= 

| 95.01 - 130 Ce ae ee ee ae ae ee 
= Ww > ao Ww = ao 
) 130.01 - 175 On Sse nae “Se ae 
‘ P . j > 175 2050 2100 

Figure 6.5. Precipitation seasonality. (a) Current conditions (1971-2000); (b, c, d) future conditions for 
2050; (e, f, g) future conditions for 2100; (b, e) moderate change scenario (SEDG); (c, f) intermediate 
change scenario (BAMBU); (d, g) severe change scenario (GRAS); (h) Boxplot of precipitation seasonality 
and projected changes across all 10 min x 10 min grid cells of the selected geographic window for 2050 
and 2100 under three climate change scenarios. The black bar within the box represents the median 
value; box boundaries show the interquartile range. Whiskers show data points that are no more than 
1.5 times the interquartile range on both sides. Open circles identify outliers, note that extreme outliers 
have been cutt off for means of better visualisation. Horizontal dashed line shows the median value for 
current conditions. 


Methodology 

6.5 Scenarios of climate change 

Current and future climatic conditions are predominantly determined by anthropogenic 
activities which affect the concentrations of greenhouse gases in the atmosphere (IPCC, 
2013). To assess the effects of future climate change on biodiversity, we need to rely on 
scenarios. Since future changes in greenhouse gas emissions depend on a large variety of 
factors such as political decisions, demographic change and sociological, economic and 
technical developments, nobody can actually foresee future conditions. In this context 
scenarios can be a strong tool but they must not be mistaken. ‘They are not predictions but 
they can help to illustrate possible future developments (European Environment Agency, 
2009) following a “what — if” approach. For instance, what will happen if we continue 
like we do now, or establish successful mitigation actions, or follow a path of even higher 
greenhouse gas emissions? By applying such different scenarios of different potential 
future human developments and corresponding effects on the climate, we can get an idea 
on the range of resulting risks for biodiversity but also on the scope and need of action in 
many fields starting from local conservation management to EU-level policies. 

During the production process of the atlas it was not possible to integrate the most 
recent global change scenarios (Representative Concentration Pathways, RCPs) as they 
have been used in the 5th Assessment Report of the IPCC, but we used three scenarios 
which are based on storylines developed within the EU FP6 project ALARM (Settele 
et al., 2005; Spangenberg et al., 2012). These scenarios integrated the IPCC (2001) 
Special Report on Emission Scenarios (SRES). These future climate scenarios were 
developed on the basis of a coupled Atmosphere-Ocean General Circulation Model 
(HadCM3; New et al., 2000). 

The three scenarios were: 

1. SEDG, Sustainable European Development Goal scenario — a storyline for 
moderate change. A policy primacy scenario focused on the achievement of a socially, 
environmentally and economically sustainable development. It includes attempts to 
enhance the sustainability of societal developments by integrating economic, social 
and environment policies. Aims actively pursued include a competitive economy, a 
healthy environment, social justice, gender equity and international cooperation. As a 
normative back-casting scenario, policies are derived from the imperative of stabilising 
atmospheric greenhouse gas concentrations and ending biodiversity loss. This scenario 
approximates the IPCC B1 climate change scenario. Mean expected temperature 
increase in Europe until 2100 is 3.0°C. 

2. BAMBU, Business-As-Might-Be-Usual scenario — a storyline for intermediate 
change. A continuation into the future of currently known and foreseeable socio- 
economic and policy trajectories. Policy decisions already made are implemented 

21 


22 

Climatic Risk and Distribution Atlas of European Bumblebees 

and enforced. At the national level, deregulation and privatisation continue except in 
“strategic areas”. Internationally, there is free trade. Environmental policy is perceived 
as another technological challenge, tackled by innovation, market incentives and 
some legal regulation. The result is a rather mixed bag of market liberalism and socio- 
environmental sustainability policy. This scenario approximates the IPCC A2 climate 
change scenario. Mean expected increase in temperature until 2100 is 4.7°C. 

3. GRAS, GRowth Applied Strategy scenario — a storyline for maximum change. A 
future world based on economic imperatives like primacy of the market, free trade, and 
globalisation. Deregulation (with certain limits) is a key means, and economic growth 
a key objective of politics actively pursued by governments. Environmental policy will 
focus on damage repair (supported by liability legislation) and some preventive action. 
The latter are designed based on cost-benefit calculations and thus limited in scale 
and scope. This scenario approximates the IPCC A1FI climate change scenario. Mean 
expected increase in temperature until 2100 is 5.6°C. 

Projections of future climatic changes resulting from each scenario were developed on 
a 10 min x 10 min grid and intersected with the geographic window used in this atlas. 
Relevant monthly projected climate data were averaged for the two periods 2021-2050 
and 2071-2100. 

6.6 Species distribution models 

To assess the climatic niche of the bumblebee species, we related the presences 
and absences of the species, aggregated to the 50 km x 50 km UTM grid, to the 
respective climatic conditions per grid cell by means of statistical species distribution 
models (SDMs). SDMs were developed with generalised linear models (GLMs) with 
a binomial error distribution and a logit link function. Since GLMs can be sensitive 
to false absence data, where a species has not been observed although it is actually 
present in a grid cell, we excluded grids without any bumblebee observation and 
an additional 51 grids with observations of one species. In total we used 2160 grid 
cells (Fig. 6.1). For the development of the SDMs we used species records from 
1970 to 2000 to match the temporal resolution of the current climate data. For 
the parameterisation of the SDMs we allowed for additive and curvilinear effects 
by incorporating second order polynomials. Models were checked for spatial 
autocorrelation with Moran's I correlograms of model residuals, but none was 
detected. Initial models were simplified by stepwise regression, while minimizing 
Akaike’s information criterion. Models were calibrated on an 80% random sample 
of the initial data set and model accuracy was evaluated on the remaining 20%. 
Agreements between observed presences and projected distributions were evaluated 
by true skill statistic (TSS) and the area und the curve (AUC) of the receiver 
operating characteristic. TSS is a simple and intuitive measure for the performance 


Methodology 

of species distribution models when predictions are expressed as presence-absence 
maps and handle shortcomings of other measures such as kappa (Allouche ez al., 
2006). Thresholds for calculating presence-absences and projections were obtained by 
maximizing TSS. To allow comparability with Settele et a/. (2008) we also calculated 
AUC which is a threshold-independent measure of model performance. While the 
climatic niche models were developed at the 50 km x 50 km UTM grid, the current 
climatic niche and future climatic niche were projected to 10 min x 10 min grid 
cells. According to the projected future conditions of climatically suitable areas in 
comparison with the predictions for current conditions, we mapped the resulting 
changes indicating areas of potential loss, potential gain and remaining suitable 
conditions. These changes were mapped within the geographical window across 
Europe used for this atlas. SDMs were developed in the statistical environment R (R 
Development Core Team, 2013). All maps were based on the WGS1984 coordinate 
system with a Miller cylindrical projection using ArcGIS software (ESRI, 2013). 

6.7 Change categories 

To assess the projected changes in climatically suitable areas, we provide tables with the 
net changes in numbers of grid cells and percentage changes. To ease the interpretation 
of these values, we also provide a colour code where we aggregated the projected changes 
into groups ranging from strong expansion to strong regression (Tab. 6.2). 

Table 6.2. Colour codes to assess the severity of projected changes in climatically suitable 
areas. 

Change intensity Percentage change colour code 

Strong expansion > +80% 
Expansion 

No ot low changes -20 to +20% 

Moderate regression 
Strong regression -50 to -80% 

Very strong regression, with extinction risks -80 to -100% 
6.8 Dispersal abilities 

The severity of geographical changes in the areas of suitable climatic conditions critically 
depends on the ability of the species to keep track with these changes. However, detailed 
data on the dispersal ability do not exist for most of the species. Thus, it is not possible 
to explicitly include dispersal in the assessments of potential future distributions of the 
bumblebees. Consequently, we provide information on the severity of the effects of 
climate change based on two extreme assumptions: 

23 


24 

Climatic Risk and Distribution Atlas of European Bumblebees 

e Unlimited dispersal, in which the entire projected future climatically suitable area 
can be colonised in principle. 

e No dispersal, in which the future climatically suitable area results from the overlap 
of current and future suitable area and the species can only lose areas with suitable 
conditions. 

However, based on ecological behaviour of each species, it is possible to provide a 
rough indication about the potential dispersal abilities of each species. Based on the 
criteria below two authors (PR and TL) performed expert classifications of each species 
into either low or high dispersal ability (Tab. 7.1). This classification can be used as an 
aid to decide which of the both assumptions in the future projections is more likely — 
full or no dispersal abilities. 

Low dispersal ability was assigned for species exhibiting the following characteristics: 

e Species restricted to high altitudes or high latitudes in mountain areas 

e Insular species 

e Species with a highly fragmented distribution with obvious subspecific differen- 
tiation 

e Habitat specialist species 

e Dietary specialist species 

e Parasites of species with low dispersal abilities 

e Species that have been unable to colonise islands 

High dispersal ability was assigned for species exhibiting the following characteristics: 
e Species living in low altitude areas 

¢ Continental species 

e Species with low subspecific differentiation 

e Species with apparently continuous distribution 

e Habitat generalists 

e Dietary generalists 

e Parasites of species with high dispersal abilities 

e Species with recent range expansions 

6.9 Definitions of climate change risk categories for European bumblebees 

We also adapted the system of Settele et a/. (2008) and placed each bumblebee 
species assessed in a risk category according to the loss of grid cells with suitable 
climatic conditions in each climate change scenario. Categories were only assigned 
for species whose distributions were modelled reasonably accurately (AUC > 0.75). 
Species whose distributions were not modelled reasonably accurately were assigned 
to the category “PR — Potential climate change risk”. The categories of model quality 
are as follows: 


Methodology 

AUC > 0.95: Present distribution can be very well explained by climatic variables 
AUC > 0.85 — 0.95: Present distribution can be well explained by climatic variables 
AUC > 0.75 — 0.85: Present distribution can be explained by climatic variables to a 
moderate extent 

AUC < 0.75: Present distribution can be explained by climatic variables to only a 
limited extent 

The climate risk categories which have been defined based on from the analysis and 
which are used throughout the atlas are as follows: 

climate change risk > /50:— 70 
TRL «| besiesas dange a 

The overall risk categories are integrated across all scenarios and time steps and are 

defined as follows: 

HHHR (extremely high climate change risk): Climate change poses a very high risk 
to the species because more than 95% of the grids with currently suitable climate may 
no longer be suitable in 2100 under at least one scenario (under the “no dispersal” 
assumption). Present distribution can be explained by climatic variables at least to a 

moderate extent (AUC > 0.75). 

HAR (very high climate change risk): Climate change poses a very high risk to the 
species because more than 85% of the grids with currently suitable climate may 
no longer be suitable in 2100 under at least one scenario (under the “no dispersal” 
assumption). Present distribution can be explained by climatic variables at least to a 

moderate extent (AUC > 0.75). 

AR (high climate change risk): Climate change poses a high risk to the species because more 
than 70% of the grids with currently suitable climate may no longer be suitable in 2100 
under at least one scenario (under the “no dispersal” assumption). Present distribution can 
be explained by climatic variables at least to a moderate extent (AUC > 0.75). 

R (climate change risk): Climate change poses a risk to the species because more than 
50% of the grids with currently suitable climate may no longer be suitable in 2100 

25 


26 

Climatic Risk and Distribution Atlas of European Bumblebees 

under at least one scenario (under the “no dispersal” assumption). Present distribution 
can be explained by climatic variables at least to a moderate extent (AUC > 0.75). 

LR (lower climate change risk): Climate change poses a lower risk to the species 
because 50% or less of the grids with currently suitable climate may no longer be 
suitable in 2100 under at least one scenario (under the “no dispersal” assumption). 
Present distribution can be explained by climatic variables at least to a moderate extent 

AUCs 045). 

PR (potential climate change risk): At the moment, climate change can only be 
regarded as a potential risk for the species’ long-term survival in Europe. All species 
whose present distribution can be explained by climatic variables to only a limited 
extent (AUC: < 0.75) have been categorised as PR, independent of the rate of decline 
of their climatic niche distribution. 

Bombus polaris. This species currently has a restricted range in the Scandinavian mountains and Arctic tundra. Even the most 
optimistic scenario projects that the species will lose the largest part of its climatically suitable area. It is at risk of extinction in 
Europe as soon as 2050. Photo G. Holmstrém. 


Checklist of the European bumblebee species 

7. Checklist of the European bumblebee species 

According to current taxonomic knowledge we recognise 79 West-Palaearctic bumblebee 
species (Tab. 7.1). Detailed, and up-to-date, distribution maps are provided at the end 
of this atlas and can also be found online (Rasmont & Iserbyt, 2014). 

Eleven West-Palaearctic species do not occur in our defined European window. From 
the remaining 69 species we could not model a further 13 species either because their 
range is too small (five species), their distribution cannot be modelled reliably with 
climate data only (four species) or because of taxonomic issues (four species; Tab. 7.1). 
Range maps of all 13 non-modelled species are presented and discussed in chapter 9. 
In total we modelled 56 species. 

There is presently no general key to allow the identification of all European bumblebee 
species. However several regional keys or keys specific to some subgenera are available 
(Pittioni, 1938; Loken, 1973, 1984; Alford, 1975; Rasmont & Adamski, 1995; Amiet, 
1996; Ornosa & Ortiz-Sanchez, 2004; Edwards & Jenner, 2009; Intoppa et al., 2009; 
Williams et a/., 2011, 2012; Prys-Jones & Corbet, 2011). 

Table 7.1. West-Palaearctic bumblebees. Subgeneric taxonomy follows Williams et al. 
(2008). Species are sorted alphabetically. 

Scientific name 

istribution outside modelling frame 
bution map page 

>) 
on 
SJ 
= 
N 
oy 
om 
3) 
oy 
=") 
N 
uo) 
vo 
=| 
S) 
uo) 
° 

istri 

Taxonomically problematic species 

Not modelled 
nquiline species 

D 

— 

= fa 
Bonbas (Wilnabon alagianas Reig, 30____| | [177] | [an 
[Bom (Apnaborbus apna an 1758) | 2{ | || [an 
[Bonbus Meabonbus arias Scopoh, 78) || | | | [an 
Borba (Toracbombus armenaas Radostkowae, 1671 | [ae | _[_[a13 
Bombs (Apincbobn baltats Dakbom, 183236. | | | [aus 
Bom Pays) bards (iby, 1802) [38 | | |e [a 

27 


28 Climatic Risk and Distribution Atlas of European Bumblebees 

Bones (any boi ad, 1638 ———~dY H/T) (i a 
Bombus (Rhodobombus) brodmanni Skorikov, 1911 | fas} || p23 | 
Bonus Crlonbu) bredmamins Vox. 1909 | [146] | | [ais 
Bonus ayn) eps Ponce, 001) | |_| | [eas 
Bonus (ekbonns) eaasis Radowhowa, 139 | | [aif | [ate 
Bonus rlonbus snus Wablberg 1854 | {|__| | [as 
Bombus (Bombias) confusus Schenck, 1861 f46{ | fe] | 2ts. 
Bonus (enn onsbrins Deklbow, Wes | 48 || | | [are 
Bombus (Bombus) cryptarum (Fabricius, 1775) 1504) ed] ete 
Bonus (Callmambonbn allman Kisey 1803) | 52 | | fe | [216 
CBombus (lorlonbus) dvtropyns Schule, 187 | [148] [~ | [207 
Bombus Cheremabombu ditngendes Mora, 1065 | 54 |_| | [lair 
Bonus ala) fas Rversnasn, 1852 | 56. | | | fe [arr 
Bonus eberrmabonbus fraps Palas, TA) | 58 || | | _[are 
Bonus (ean) gree Nora, 1881 | || [| [ate 
Bonus Clon links Feese, 902 |__| sap [ate 
Bombes (rolantas) humains Kechiaones, 1670 | @ |__| | [are 
Bonus (Menaconlus hndisinsVoxe, 1909 | [ai =| [218 
Gomes (Metonius) lorena sis) +d | | pe | ai. 
Bonus (haraclonbus bonis Hier, 1806 | 6 | | |_| [ao 
Comba: Afnbonbay petornsSehioher, 9 [6 | | | [pe [aio 
CGomlus Cronus bom (175 ‘|| | | | [ao 
Bombs (eanbonbns) inerins Mora, | 2 | | | [are 
| Bumbus (Thoracobombus) inexopectatus (Teale, 1963) | 74 | ||| | 218 | 
Bombus (Pyrobombus) jonellus (Kirby, 1802) ee ee ed eee 
Bomba (Thrall Norwvi, 815 |__| [| [20 
CGombus (ean) iid (1758) | |_| fe | ar 
CGombus (rotons) ipo Wabsicus, 195) | 60 | | Pe | aan 
ee 
Combs Bomba magus Nowe, 91) |p | 
CGombus Cbernabonus) milnur Lapses, oe | | |i) | [ae 
Bombs Mdm) mend Gesvicle, 1869 [66 | |__| [om 
Comba; (luton) mets Gerster, 869 [88 | | | | [2 
Bombus (Thoracobombus) mlokostevitxu Radoszkowski, 1877 ae ee a ed ed ey) 
Bomba (Thrwnbonin) mani Keeckonumess677 [| [49] [* | [23 
Comba Eyton mods Evessmann, 1852 | | [are] | [22 
iomtus Crlonbus) montol Sith, 1649 0. | Le | [2s 


Checklist of the European bumblebee species 29 

Bamba (Toten) mais Gomis, 18 [| | | 2 
Bombe (Tlrantonbn) meron 1758) [4 | | |» ] [aan 
Boma (itrobonins) niveau Kecchbauon, 1670 | 96[ | Te | [2s 
Bonus (Paya) rvs SpecreSebocider,1918) | 98 | | | _[e [aoe 
Bombe (Tlranionbn) param Genpak, 768) [100] [| | [225 
Bonus Bombay potas Nance, 1848 | _—‘[iae| | [ae 
Bombus (Psithyrus) perexi (Schulthess-Rechberg, 1886) | ff fe | @ | 232, 
Bombe (Tlranlonbn) peels Gkor 12) | | [|] [25 
Bonus (Horclonbus persis Radowshowsk, 1681 | | [ai |_| aaa 
Bonus (Anon) pols Cures, 1895_— ‘| 1021 |__| | 22s. 
Bomba (Thranbonns) pono Pasa, Ww) [oe | | || 8 
Bombus (eons) porshinsy Nadoschowsk, 1685 |__| [ai | [| 225, 
Bombus (Pyrobombus) pratorum (1.., 1761) ri ame 
Comba Paton) pyran Pen, 187) ————*([ow | || 
CGombus (aly) quainbr (uepeleies 1832) | 10 | | [« [226 
Bombus (eau) rigs Rasmon, 1983) | __[1a7| fe | [27 
Comba Bombs) nnd Radoschowsl, 1881 [| [1st] Le | [2a 
CGombus (loeolonbus radar (ill, 177 [|_| |_| 2r 
Bombus (Megabombus) ruderatus (Fabricius, 1775) rime me: 
Bombus (Psithyrus) rupestris (Fabricius, 1793) f1t6| | | | @ | 228) 
ions (eons) salaries Skoikow 1951) |__| fai | | 
Bonus (horaolonbas) shrewki Mori, 1881 [118] | [2a 
Bomba Culmanlon) series Soak, 310 [120 | | || [8 
Bonus (ean) sel Radowhowki 1859 [122 | L~ | [20 
CGombus (Ralbon) croens Casco, 171, |e | |_| [aa 
Bombus (Bombus) sporadicus Nylander, 1848 f126{ | | | | 230) 
Bomba abtrancabonba) abr: 1158) [28 | [| [230 
Bomba ibis) afr Reese, 1905 |__| [ai || [20 
Comba (Thranboin) harm, 176) [so |__| | 0 
Bombus (Psithyrus) sylvestris (Lepeletier, 1832) (a2: | ale, almer Fost 
CGombus Bonus iret, 15) —_——————~ids | pe | 
Bombus (Thoracobombus) velox (Skorikov, 1914) Ce ee ea 
Comba Ptr) ats Got, 105) [Ba | f° | 232 
Bombe: (hoon) weranns (brits 155) [ws | |_| [as 
Bomba (Apisnlonus wun Radosh, 1859 [wo | | | [a 
Bombus (Bombus) xanthopus Kriechbaumer, 1873 Jama 5 iemel|25 5 feel FA 
Bombus (Thoracobombus) zonatus Sroith, 1854 ae ae a ie es 


30 Climatic Risk and Distribution Atlas of European Bumblebees 

8. Climatic risks of European bumblebees 

8.1 Colour codes 

Scenario tables 

From 80% gained area 

Between -20% and +20% area change 

Between -50% and -80% lost area 

From -80% lost area 

8.2 Risk categories 


Climatic risks of European bumblebees 

Bombus alpinus. The global distribution of this species is presently restricted to high levels of the Alps, the Carpathian and 
Scandinavian mountains and to Arctic tundra of northern Fennoscandia. The species is expected to lose a substantial part of its 
climatic suitable area already in 2050 and could be driven to the verge of extinction in 2100. Photo G. Holmstrém. 

Bombus niveatus. The global distribution of this species includes presently the Balkan Peninsula and Near Orient. Its climatically 
suitable area is expected to increase dramatically already by 2050 and still further by 2100. Depending to its seemingly high 
dispersal abilities, it is expected to expand its distribution in a large part of Europe. Photo P. Rasmont. 

31 


32 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus alpinus (L., 1758) 
= Bombus (Alpinobombus) alpinus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus alpinus is a large bumblebee, present only in the Arctic tundra and high alpine grass- 
lands. Its coat colour is quite invariable, with a black thorax and a largely reddish abdomen. It 
lives in small colonies and is a generalist forager. Bombus alpinus occurs at the highest elevations 
in the Alps and in the Scandinavian Mountains. It can also be found at the sea level along the 
northern coast of Norway. The modelled distribution shows that its climatic niche would be 
larger than its actual distribution. Indeed, despite the presence of available climatic conditions, 
it is absent from Pyrenees, British Isles and Iceland. All scenarios project that suitable areas 
will disappear from its southernmost lo- 

Present distribution can be well explained by climatic cations in the Carp athians. The GRAS 
variables (AUC = 0.95) scenario projects a strong reduction of 
suitable climate space in the Alps and in 
inate as Ecase ot ate the Scandinavian mountains, resulting 
IUCN Red List status: Vulnerable in an increasingly fragmented distribu- 
tion. Currently, the species seems to be in 

decline especially in the Alps and in the 
[Scenario _| Bulliciispersal Doidispersal Carpathians and it is assessed as Vulner- 
SEDG able in the IUCN Red List of European 
BAMBU Bees. The dispersal ability is unknown, 
but might be low as the species is associ- 
GRAS : ; 
ated with cold temperatures and occurs in 
-2241 (-57%) -2246 (-57%) small populations. It is projected to suffer 
-3036 (-77%) -3038 (-78%) considerably from global warming in all 
scenarios. It is projected to be at the verge 
of extinction by the year 2100. 

-3450 (-88%) -3450 (-88%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
ME gain 
stable 
~_#loss 

BAMBU 
(2050) 

(2050) 

SEDG 
(2100) 

BAMBU 
(2100) 

Bombus alpinus 

33 


34 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus argillaceus (Scoro.t, 1763) 
= Bombus (Megabombus) argillaceus 

© Photo: G. Holmstr6m 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus argillaceus is a very large bumblebee. The large queens show a unique colour pattern, 
being the only European species with a completely black abdomen. Males and workers show the 
same yellow bands but the tail is white. It is considered to be a generalist species but it prefers to 
forage from flowers with a long corolla which are best suited for its long tongue. It is abundant 
south of latitude 45 N; from south-east France to Ukraine and Turkey in the west to Iran in 
the east. It mainly lives in Mediterranean and sub-Mediterranean habitats where it produces 
large colonies. It is generally absent from high mountains. ‘The species is not considered to be 
threatened: Least Concern in the IUCN Red List of European Bees. The modelled distribution 
shows that its climatic niche would include a wider area in eastern Europe, the Iberian pen- 
insula and even in southern Scandinavia, 
from where it is absent at the moment. 

Present distribution can be well explained by climatic 
variables (AUC = 0.86) All scenarios project a large expansion. 

SEDG and GRAS project that its suitable 
areas could even reach as far north as the 

IUCN Red List status: Least Concern Arctic Circle by 2100. The GRAS scenar- 
io indicates that suitable areas could in- 

clude all central Europe and a large part 
| | Scenario _| Full dispersal No dispersal of western Europe and subarctic Scandi- 

SEDG navia. The British Isles and Brittany seem 

Climate risk category: LR 

to remain out of reach in all scenarios. As 
BAMBU 

it is an unspecialised lowland species, we 
GRAS 

could assume that it would have a good 
SEDG 9724 (92%) -22 (0%) dispersal capacity. The species is expected 
BAMBU 10980 (104%) -184 (-2%) to benefit from climate change and will 
most likely expand its distribution range 

GRAS 13428 (127%) -193 (-2%) 

dramatically. 
Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


35 

Bombus argillaceus 


36 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus balteatus (DAHLBOM, 1832) 
= Bombus (Alpinobombus) balteatus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus balteatus is a large bumblebee, with several colour forms, with or without yellow bands 

and with a white or red tail. It is a generalist forager. It is found only in alpine and subalpine 

areas of Scandinavia, northern Finland and northern Russia, with a circum-boreal distribu- 

Present distribution can be very well explained by 
climatic variables (AUC = 0.99) 

Climate risk category: HHR 

IUCN Red List status: Least Concern 

| | Scenario Full dispersal No dispersal 

-3169 (-75%) -3170 (-75%) 
-3915 (-92%) -3926 (-92%) 
-3973 (-94%) -3982 (-94%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

tion. Bombus balteatus mainly lives in the 
taiga and tundra where it produces medi- 
um-sized colonies. The species is not con- 
sidered to be threatened: Least Concern 
in the IUCN Red List of European Bees. 
The modelled distribution shows that its 
climatic niche presently includes most of 
the southern European mountains, from 
where it is absent. All scenarios project a 
reduction of suitable areas. By 2100, it 
would be restricted to alpine areas, dis- 
appearing from lower altitudes even at 
northern latitudes along the Barents Sea 
shore. Regardless of its dispersal ability, as 
it is adapted to cold temperatures, B. bal- 
teatus is projected to suffer considerably 
from global warming. 


Bombus balteatus 37 

SEDG 
(2050) 
ma gain 
stable 
loss 

SEDG 
(2100) 

BAMBU 
(2050) 


38 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus barbutellus (Kirpy, 1802) 

= Bombus (Psithyrus) barbutellus; Psithyrus barbutellus; Psithyrus maxillosus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus barbutellus is a medium-sized bumblebee. While populations close to its northern range 
margin have a coat colour with 3 yellow bands and a white tail, some southern populations can 
be nearly all black with very dark wings (ssp. maxillosus). It is a social parasite species (cuckoo- 
bumblebee) invading the nests primarily of B. argillaceus, B. hortorum and B. ruderatus. The 
species occurs across a large area from Spain in the south to Stockholm and Helsinki in the 
north and from Ireland in the west as far as to the Pacific coast in the east. It is however never 
abundant. The species has disappeared from most of its historic locations in the west- and 
central European lowlands. Despite this regional regression, the species is not considered to 
be threatened at a continental scale: Least Concern in the IUCN Red List of European Bees. 
The modelled distribution shows that 
Present distribution can be explained by climatic its climatic niche includes a wider area 
variables to a moderate extent (AUC = 0.76) along thresh lantieconat ot west Norway 
Climate risk category: HHR from where it is absent. All scenarios 
Pe at eee project a fragmentation of the climatic 
space in central and south Europe and 
an expansion of its suitable areas into the 
|_| Scenario | Fulldispersal_| No dispersal | Arctic Circle to the north. The GRAS 
SEDG scenario projects that suitable areas could 
er completely disappear from all lowlands 
south of 55° N by 2100. As it is a cuckoo- 

GRAS bumblebee associated with only a few host 
SEDG -6665 (-60%) species and has a scattered distribution, its 
BAMBU 6645 (60%) 8710 (-78%) dispersal ability is expected to be low and 
B. barbutellus would suffer considerably 

GRAS -7445 (-67%) -9617 (-87%) 

from climatic warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


39 

Bombus barbutellus 


40 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus bohemicus SEIDL, 1837 

= Bombus (Psithyrus) bohemicus; Psithyrus bohemicus; Psithyrus distinctus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus bohemicus is a medium-sized bumblebee. Its coat colour is rather constant, generally 

with one large yellow band and a white tail. It is a social parasite species (cuckoo-bumblebee) 

specialising primarily on B. /ucorum and probably also B. magnus, B. cryptarum and B. terrestris. 

Its southernmost location is in the southern Italian mountains. All locations south of latitude 

45° N are in the mountains. In the lowlands, B. bohbemicus occurs from this latitude northwards 
to 70° N. It is distributed from Ireland in the west to the Pacific coasts in the east. It is also 
the most common cuckoo-bumblebee. The species is not considered to be threatened: Least 

Present distribution can be explained by climatic 
vatiables to a moderate extent (AUC = 0.81) 

Climate risk category: HHR 

IUCN Red List status: Least Concern 

|| Scenario Full dispersal No dispersal 

SEDG 

BAMBU -2978 (-19%) 

GRAS 
S 

EDG -8947 (-57%) 
BAMBU -8245 (-53%) -10482 (-67%) 
GRAS 

|GRAS | -10053 (-65%)  -12262 (-79%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

Concern in the IUCN Red List of 
European Bees. The modelled distribution 
corresponds very well to the actual one. 
All scenarios project a fragmentation of 
the range in central and southern Europe 
and an expansion of its suitable areas to 
the Barents Sea coast and to the highest 
altitudes of the Scandinavian mountains. 
The GRAS scenario projects that suitable 
areas could completely disappear from 
the lowlands south of latitude 60° N by 
2100. Regardless of its dispersal ability, 
as it is a cuckoo-bumblebee, specialised 
to few host species with a seemingly low 
dispersal ability, B. bohemicus would 

suffer significantly from global warming. 


Bombus bohemicus 

41 


Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus campestris (PaNnzER, 1801) 
= Bombus (Psithyrus) campestris; Psithyrus campestris 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus campestris is a medium-sized bumblebee. Its coat colour is very variable, generally with 
yellow bands and a yellow tail. The tail can also be reddish. Some specimens are completely 
black. It is a social parasite species (cuckoo-bumblebee) mostly of B. pascuorum, B. humilis, B. 
ruderarius, B. syluarum, B. muscorum and B. subterraneus. Most locations south latitude 45° N 
are in the mountains. In the lowlands, it occurs from this latitude in the south up to 65° N, near 
the Arctic Circle. It is distributed from Ireland in the west to the Pacific coasts in the east. It is also 
one of the most abundant and widespread cuckoo-bumblebees. ‘The species is not considered 
to be threatened: Least Concern in the 

Present distribution can be explained by climatic IUCN Red List of Europ ean Bees. The 
variables to a moderate extent (AUC = 0.77) modelled distribution corresponds very 
well to the actual one. All scenarios project 
Ee eee a fragmentation of the range in central 
IUCN Red List status: Least Concern and southern Europe and an expansion 
of its suitable areas to the Barents Sea 

coast without reaching the highest levels 

| | Scenario | Full dispersal | No dispersal of Scandinavian mountains. The GRAS 
SEDG scenario projects that suitable areas could 
BAMBU completely disappear from all lowlands 
south of latitude 60° N by 2100. As it is 
a cuckoo-bumblebee specialised on a few 
-5989 (-54%) -7633 (-69%) host species, with a scattered distribution, 
-6475 (-59%) -9191 (-83%) and with a seemingly low dispersal ability, 
B. campestris would suffer significantly 

GRAS 

-7420 (-67%) -9879 (-89%) 

from global warming. 
Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
Ma gain 
—__ stable 
> __#loss 

(2050) 

GRAS 
(2100) 

Bombus campestris 

43 


Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus cingulatus WAHLBERG, 1854 
= Bombus (Pyrobombus) cingulatus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus cingulatus is a medium-sized bumblebee. Its coat colour is quite constant, with a 
brownish thorax with a more or less wide black thoracic band and with a white tail. It is a 
species restricted to the boreal taiga where it prefers to forage from Ericaceae and Epilobium 
angustifolium flowers. It occurs in Europe from the latitude of Stockholm in the south close to 
the Barents Sea coast in the north. In the 

Present distribution can be explained by climatic west from Norway to the Pacific coasts 
variables to a moderate extent (AUC = 0.97) in the east. The species is not considered 
to be threatened: Least Concern in the 

IUCN Red List of European Bees. The 

IUCN Red List status: Least Concern modelled distribution corresponds very 

Climate risk category: HHHR 

well to the actual one, bearing in mind 
that the species does not occur in the 
| |Scenario | Full dispersal | No dispersal southern mountains. All scenarios project 
shrinkage of suitable areas by 2050. By 
2100 suitable conditions for this species 
would be restricted to mountain areas, 
this tendency being the most extreme 
-3246 (-57%) -3307 (-58%) with the GRAS scenario. Regardless of its 
5093 (-89%) 5141 (-90%) dispersal capability, as it is a species linked 
to boreal conditions, B. cingulatus would 

-5328 (-94%) -5372 (-94%) 

suffer considerably from global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
EE gain 
stable 
~~ kloss 

(2100) 

Bombus cingulatus 

45 


46 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus confusus SCHENCK, 1859 
= Bombus (Bombias) confusus; Bombus (Confusibombus) confusus; Bombus 
paradoxus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus confusus is a medium-sized bumblebee. It includes two very conspicuously different 
colour forms: the nominal subspecies, black with a red tail, and the ssp. paradoxus, with 3 yellow 
bands and a white tail. The coat also shows a very typical velvet-like aspect. This species is mostly 
present in steppes or dry grasslands with scattered trees and shrubs. The queens and workers forage 
mainly on Fabaceae while males forage for nectar on thistles (Asteraceae). B. confusus occurs from 
the Pyrenees and northern Balkans in the south to Estonia in the north. It is absent from the 
British Isles and Fennoscandia. Its westernmost location is in south-east France while it reaches 
Novosibirsk in the east. The modelled distribution does not perfectly fit with its actual one. 
The species is one of the most threatened 

Present distribution can be well explained by climatic European bumblebees, and is assessed 
variables (AUC = 0.87) as Vulnerable in the IUCN Red List of 
Giimaenemoresory DIMER European Bees. All scenarios project a 
reduction of areas with suitable conditions 
IUCN Red List status: Vulnerable by 2050, with an expansion toward the 
north. By 2100, the suitable areas of the 

species would reach the Arctic Circle 

|| Scenario Full dispersal No dispersal with only a fragmented distribution in 
SEDG -96 (-1%) the south. The GRAS scenario projects 
BAMBU an almost complete shift of suitable areas, 
a with the exception of Fennoscandia. The 
SEDG -5438 (-68%) dispersal ability of B. confusus is likely to 
BAMBU 5375 (-68%) 7618 (-96%) be low. It would therefore considerably 
suffer from global warming, which could 

disappearing from the lowlands of Europe, 

GRAS -4501 (-57%) -7807 (-98%) 

eventually lead to its extinction. 
Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus confusus 

SEDG 
(2050) 
a gain 

___ stable 
loss 

aes 

Bey 
a 

€ 
| gt = 

Fs 
LY pti nis 
= oy 

Me 

(2050) 

(2100) 

47 


48 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus consobrinus SCHENCK, 1859 

= Bombus (Megabombus) consobrinus 

% *, : ’ ‘i 
=. 
oa 
+A 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus consobrinus is a large bumblebee. Its coloration is quite constant, being brown on thorax 
and the basal part of abdomen, the mid part is black gradually becoming whitish grey towards 
the tip. In Europe, this species exclusively inhabits the boreal taiga where it is highly specialised 
in flower choices: it forages almost exclusively on Aconitum spp., even though it occasionally 
forages for nectar on other flowers. It occurs from Norway in the west to the Pacific coasts 
in the east. Its modelled distribution indicates that climatic conditions would be well suited 
in most of the southern European mountains, even though the species is absent there. The 
species is not threatened: Least Concern 
Present distribution can be very well explained by in the IUCN Red List of European 
climatic variables (AUC = 0.97) Bees. All scenarios project a reduction of 
Climate risk category: HR suitable areas by 2050. This tendency is 
projected to continue, and by 2100 the 
GRAS scenario indicates that suitable 
climatic conditions would persist only 

| | Scenario _| Full dispersal No dispersal in the high Scandinavian mountains. 

SEDG Movement of this species to southern 

IUCN Red List status: Least Concern 

European mountains is very unlikely. 
As it is adapted to rather cold climates 
and is highly specialised in its habitat 
SEDG -2030 (-57%) and food choices, the dispersal ability 
of this species is likely to be low. Thus, 

BAMBU 
GRAS 

BAMBU -2843 (-80%) -2970 (-84%) 
GRAS -2898 (-82%) -2999 (-85%) 

B. consobrinus would suffer considerably 
from global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDGE 
(2050) 
= gain 
stable 
loss 

(2100) 

GRAS 
(2100) 

Bombus consobrinus 

49 


50 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus cryptarum (Fasricius, 1885) 

= Bombus (Bombus) cryptarum; Bombus lucocryptarum 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus cryptarum is quite a small species of bumblebee. With some variations, the coat colour 
always shows two yellow bands and a white tip to the abdomen. The more or less developed 
prothoracic yellow band usually has a black “comma” at the height of the tegulae. The 
identification can be very difficult and confusions could occur with B. magnus and B. lucorum. 
In Europe, this species generally inhabits heaths and moors with abundant Ericaceae flowers 
which are its main food resource (e.g. Vaccinium spp., Erica spp., Rhododendron spp.). It occurs 
from the northern Balkan, the Alps and Massif Central in the south to the Barents Sea shore in 
the north and from Ireland in the west to the Pacific coast in the east. The actual and modelled 

distribution might be potentially blurred 

ao ae by numerous identification mistakes. The 
Present distribution can be explained by climatic y 

tarbles trout a Werte Catone AU@ = 0.72) species is not threatened: Least Concern 

in the IUCN Red List of European 
Bees. All scenarios project a reduction of 
IUCN Red List status: Least Concern suitable areas already in 2050, especially 
in the lowlands of west and central 

Climate risk category: PR 

Europe. In 2100 all scenarios project 
|| Scenario | Full dispersal | No dispersal that the suitable climatic conditions 

SEDG would persist only in mountains of 
South Europe, Fennoscandia, Ireland 

BAMB , 
BMY and Scotland (not even in these latter 

GRAS areas following GRAS). As B. cryptarum 
-7183 (-58%) -8112 (-66%) is quite specialised in its habitat and food 
preferences, it could suffer from global 

-7133 (-58%) -8818 (-72%) 
-9260 (-75%) -9880 (-80%) 

warming if these resources are altered, 

regardless of its dispersion capability. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


GRAS 
(2100) 

Bombus cryptarum 

51 


52 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus cullumanus (Kirsy, 1802) 

= Bombus (Cullumanobombus) cullumanus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus cullumanus is a medium-sized bumblebee. The coat colour can show three different regional 
patterns: all black with a red tail (ssp. cu//umanus), with three yellowish bands and a red tail (ssp. 
serrisquama Morawitz), with three white bands and a red tail (ssp. apollineus Skorikov). ‘The ssp. 
cullumanus once occurred in chalky grasslands across the more Atlantic parts of the continent 
from the Pyrenees in the south to the Isle of Oland (Baltic sea) in the north. This ssp. seems 
to be completely extinct, with the last specimen being seen in 2004 in the Massif Central. Ssp. 
serrisqguama was once found in steppe areas of Spain, central and eastern Europe, the Caucasus, 
southern Siberia and Mongolia. For Europe, this ssp. only persists still in a few locations in central 
Spain and the Volga valley, while it can be abundant in some parts of eastern Turkey, Siberia and 
Mongolia. The ssp. apollineus is restricted to eastern Turkey, Georgia, Azerbaijan and Iran where it 
remains abundant in some locations. The identification can be difficult and confusions could occur 
with B. /apidarius. The queens and workers of B. cullumanus forage mainly on Trifolium spp. while 
males visit thistles (Asteraceae). The species 
is highly threatened: Critically Endangered 
in the IUCN Red List of European Bees. 

Climate risk category: HHR All scenarios project a reduction of suitable 

IUCN Red List status: Critically Endangered areas by 2050. By 2.100 all scenarios project 

that the suitable climatic conditions would 

| | Scenario | Full dispersal No dispersal 
where this species has already vanished. 
SENG As this species is highly specialised, has a 

BAMBU -2139 (-51%) scattered distribution, has already become 

GRAS -2115 (50%) 2480 (-59%) extinct from most of its original range, 
and is likely to have a very low dispersal 

Present distribution can be well explained by climatic 
variables (AUC = 0.91) 

persist only in northern Europe from 

SEDG -2264 (-54%) -2833 (-67%) 

capability, B. cullumanus would seriously 
BaMeU -2885 (-68%) -3543 (-84%) suffer from global warming likely leading 
GRAS -3284 (-78%) -3942 (-94%) to its extinction by 2050. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus cullumanus 53 

SEDG 
(2050) 
ma gain 
stable 
-—_#loss 

BAMBU 
(2050) 

BAMBU 
(2100) 


54 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus distinguendus Moraw11z, 1869 

= Bombus (Subterraneobombus) distinguendus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus distinguendus is a large bumblebee. The coat colour is very constant: gold-yellow with a 
black thoracic band. The species occurs from a latitude of 45°N in the south to the Arctic Circle 
in the north and from Ireland in the west to Kamchatka and even to the Aleutian Islands in the 
east. It forages mainly on Trifolium spp. (queens and workers) and thistles (males). This species is 
scarce throughout Europe and it seems threatened: Vulnerable in the IUCN Red List of European 
Bees. The distribution model shows that its climatic niche includes southern mountains such as 
the Cantabrian mountains, Pyrenees, Apennines, and the Balkans from where it is not known to 
occur. Most of its original distribution from a century ago is now out of its modelled range, possibly 
meaning that suitable climatic conditions have already moved significantly. All scenarios project a 
reduction of suitable areas by 2050, where all lowland locations south of 55° N become unsuitable. 
By 2100 all scenarios project that suitable climatic conditions would only persist in northern Europe 
and in the mountains of central and eastern 
Europe. The GRAS scenario indicates that 
the suitable areas would remain in only a 
Climate risk category: HHR very restricted area of the Alps, in scattered 
MCN ed coca uinenble locations in Scotland and north of 60° 
N in Scandinavia, reaching the highest 

| | Scenario | Fullidispersal Nerdisseisal altitudes in the Scandinavian mountains. 
As B. distinguendus is quite specialised in 

SEDG 

Present distribution can be explained by climatic 
variables to a moderate extent (AUC = 0.81) 

its food preference and its range in the 
BAMBU central and western European lowlands 
GRAS is already scattered, low dispersal abilities 
may be assumed (even if it the species is 

SEDG -9656 (-67%) 

able to forage on coastal islands). Thus, B. 
BAMBU -8038 (-56%) -11484 (-79%) distinguendus would suffer considerably 

GRAS -9252 (-64%)  -12608 (-87%) from global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


BAMBU 
(2100) 

Bombus distinguendus 

55 


56 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus flavidus EVERSMANN, 1852 
= Bombus (Psithyrus) flavidus; Psithyrus flavidus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus flavidus is a small- to medium-sized bumblebee. The coat colour generally shows a 
yellow prothoracic band and a whitish to yellowish tail. The rest of the body is black more or less 
intermixed with yellowish hairs. It is a social parasite species (cuckoo-bumblebee) most likely 
of B. monticola, B. lapponicus, B. jonellus, B. cingulatus and B. pyrenaeus. The species occurs in 
the alpine and subalpine zones in the Pyrenees and the Alps, the Scandinavian mountains, the 
boreal taiga and the arctic tundra. To the east, its distribution reaches the Pacific coast. It is 
locally numerous and in some places, it can even be the most abundant cuckoo-bumblebee. The 
species is not considered to be threatened: Least Concern in the IUCN Red List of European 
Bees. The modelled distribution shows that its climatic niche would include Durmitor, the 
Balkans mountains, Carpathians, Tatra, 
Present distribution can be well explained by climatic Massif Central and even Scotland, 
ee (from where the species has never been 
Climate risk category: HHR observed). All scenarios project a small 
shift of suitable climatic conditions by 
2050 while by 2100 the suitable areas 
of the species would be much more 
| | Scenario | Full dispersal | No dispersal restricted in the Scandinavian mountains, 

SEDG Alps and Pyrenees. GRAS projects that 

unsuitable climatic conditions will 

IUCN Red List status: Least Concern 

“ieee exclude B. flavidus from the Pyrenees. 

GRAS Regardless of its dispersal ability, as it is 
-3730 (-50%) -3730 (-50%) a cuckoo-bumblebee specialised to few 
host species, with scattered distribution, 

-5189 (-69%) -5189 (-69%) 
-6354 (-85%) -6354 (-85%) 

B. flavidus would suffer greatly from 

global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2100) 

Bombus flavidus 

57 


58 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus fragrans (Paias, 1771) 
= Bombus (Subterraneobombus) fragrans 

© Photo: G. Holmstr6m 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus fragrans is a very large bumblebee, the largest in Europe. The coat colour is yellowish 
with a black interalar band. It occurs only in true steppes where it is a generalist forager. It 
nests mainly in suslik burrows (steppic colonial rodents of the genus Spermophilus). The species 
occurs in the steppes of central and eastern Europe and the Anatolian plateau, where it is 
generally rare. To the east, it reaches Mongolia. As the species has not been observed recently in 
most of its former central European locations, it is considered to be threatened: Endangered in 
the IUCN Red List of European Bees. The modelled distribution shows that suitable climatic 
conditions include the areas of central Europe from which the species had already disappeared. 
All scenarios project a small shift of its climatic niche space by 2050 with no significant gains 
or losses in area. By 2100, all scenarios project a clear fragmentation of the species’ range south 
of latitude 45°N , while GRAS projects a complete shift of suitable climatic conditions to the 
north of latitude 55°N,where the species 

Present distribution can be well explained by climatic does not live presently. To cope with 

variables (AUC = 0.94) such a major shift would require high 

Climate risk category: HHHR mobility of the species, which would be 
quite unlikely as it only lives in habitats 

IUCN Red List status: Endangered 
that are generally suffering considerably 

| | Scenario _| lid al 7 7 from agricultural intensification. As 
e F e ts 
Sia Set hae peeps it is a species that is restricted to true 

-510 (-9%) steppes (a habitat that is not expected to 
expand) and is already very localised or 
absent from much of its former range, 

-670 (-12%) 

the dispersal ability of the species can be 
assumed to be low. Therefore, B. fragrans 
-4791 (-87%) would suffer considerably from global 

2777 (-50%) 5446 (-98%) warming, the worst scenario leading to 

the extinction of the species in Europe. 
Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus fragrans 59 

SEDG SEDG 
(2050) | | (2100) 
= gain 

—__ stable 

loss 


60 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus gerstaeckeri Morawt1z, 1875 
= Bombus (Megabombus) gerstaeckeri 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus gerstaeckeri is a large bumblebee. ‘The coat colour is brownish on both the thorax and the 
base of the abdomen. The abdomen has a whitish tail. It is a highly specialist forager, visiting almost 
exclusively monkshood (Aconitum spp.). Often, each colony is small, including only a few workers 
and a very low number of new queens and males are produced in a year. Bombus gerstaeckeri is 
a very rare species, endemic to the high mountains of southern Europe: The Pyrenees, Alps, 
Carpathians, and the Caucasus. As it is a very conspicuous species, it is usually regarded as more 
abundant than it actually is. The species is considered to be threatened: Vulnerable in the IUCN 
Red List of European Bees. The modelled distribution includes most of the mountainous areas 

of Europe but its actual distribution is 

ae we Ee icted. All scenarios proj 
Present distribution can be well explained by climatic much more restricted ivan ey er 

variables (AUC = 0.95) smaller shifts of its climatic niche space 

by 2050 with little change in climatically 
Ge Ree On gas suitable area. By 2100, the BAMBU and 
IUCN Red List status: Vulnerable SEDG scenarios also project smaller shifts 

of the climatic niche space but GRAS 

projects more drastic range contractions 
| | Scenario _| Full dispersal No dispersal leading to extinction in both the Pyrenees 
SEDG and Carpathians. As this highly specialised 

BAMBU species shows low dispersal abilities, a 

GRAS 

move to new suitable areas in Scandinavia 
is very unlikely. Thus, B. gerstaeckeri would 
-1661 (-50%) suffer considerably from global warming, 
2134 (-64%) with the worst scenario leading to the 
extinction of the species in the Pyrenees 

-2448 (-74%) 

and Carpathians. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
= gain 
stable 
loss 

GRAS 
(2100) 

Bombus gerstaeckeri 

61 


62 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus haematurus KRIiECHBAUMER, 1870 
= Bombus (Pyrobombus) haematurus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus haematurus is a small bumblebee. The coat colour shows a large yellow prothoracic 

band. The abdomen has a large basal yellow band and a small red tail. It is a generalist species 

occurring in forests, orchards and park-like landscapes of Turkey and south-east Europe. It 

reaches Slovenia in the west and Iran in the esast, southern Greece to the south, and Slovakia 

and Romania to the north. Bombus haematurus has recently expanded its range by about 1000 
km westwards. The species is not considered to be threatened: Least Concern in the IUCN 

Present distribution can be well explained by climatic 
variables (AUC = 0.93) 

Climate risk category: HR 

IUCN Red List status: Least Concern 

-4404 (-69%) 
-5110 (-81%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

Red List of European Bees. Its modelled 
distribution includes a much larger area 
than its actual distribution. All scenarios 
except GRAS project an expansion of 
its climatic niche space by 2050 and 
2100. GRAS projects fragmentation of 
areas with suitable climatic conditions 
in the lowlands south of latitude 48° N 
and also an expansion to the north-west, 
reaching Belgium, the Netherlands and 
Scandinavia.Its recent expansion across 
the Balkans shows that the species is a 
good disperser. Thus, B. haematurus could 
take advantage from global warming 
leading to its expansion towards western 
and northern Europe. 


63 

Bombus haematurus 

SEDG 
(2050 

) 

Ma gain 

a) 
fo} 
£ 
7) 

n 
a 
eo) 


64 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus hortorum (L. 1761) 

= Bombus (Megabombus) hortorum 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus hortorum is a medium-sized bumblebee. The coat colour shows three large yellow bands 
and a white tail. Other coloration patterns can be seen in Corsica (black with a red tail) and here 
and there as a melanic form which is all black but with white tail. It is a generalist species even 
if it forages mainly on flowers with long corollas which are well suited to its very long proboscis. 
It is widely distributed across Europe, from Sicily and southern Spain in the south (where it 
occurs only in the mountains) to the extreme north, occurring in the coldest tundra along 
the Barents Sea coasts. To the east, it reaches the Pacific coast. The species is not considered 
to be threatened: Least Concern in the 

Present distribution can be explained by climatic IUCN Red List of Europ ean Bees. Its 
vatiables to a moderate extent (AUC = 0.77) modelled distribution includes an area 

; somewhat more restricted than its actual 
Climate risk category: HR ; : : 
one. All scenarios project a reduction of 

IUCN Red List status: Least Concern the climatic niche space of the species in 

the south. By 2100, the GRAS scenario 

would make all lowland areas in the 
| | Scenario _| Full dispersal No dispersal European mainland unsuitable. Suitable 
SEDG conditions would only remain in the Alps, 

BAMBU Wales, Ireland, Scotland, Scandinavia 

and northern Finland. Even though 
GRAS 

B. hortorum is quite abundant, and a 
SEDG -7072 (-55%) generalist species, with high dispersal 

BAMBU 7078 (-55%) 8391 (-65%) capability, it would lose a considerable 
amount of climatically suitable area 

GRAS -7997 (-62%) -9146 (-71%) 

under warming conditions. 
Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


GRAS 
(2100) 

Bombus hortorum 

65 


66 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus humilis TLu1GER, 1806 
= Bombus (Thoracobombus) humilis; Bombus variabilis 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus humilis is a medium-sized bumblebee. The coat colour is extremely variable depending on 
region (often classified as different subspecies), but also populations more close to each other show 
considerable variation. Frequent colour patterns are brownish, or black with a red tail, the most 
typical colour character being the brown hairs on the 2nd tergum. It is a generalist species although 
it forages mainly on Lamiaceae and Fabaceae flowers. In Europe, it can be found from southern 
Spain, Greece and Turkey in the south, (where it lives in mountains only), to a latitude of 65° N 
in Scandinavia and Russia in the north. To the west, it reaches Scotland and north-west Spain (not 
Ireland) and the Pacific coast to the east. The species is absent from all the Mediterranean islands. It 
has become scarce in most lowland areas of west and central Europe. Despite this regional regression, 
at a continental scale the species is not considered to be threatened: Least Concern in the IUCN Red 
List of European Bees. Its modelled distribution includes some areas in which the species does not 

occur, such as Morocco, Ireland, Corsica, 
Present distribution can be explained by climatic 

Sardinia, Sicily and western Norway. All 
variables only to a limited extent (AUC = 0.75) 

scenarios project a moderate reduction 
Climate risk category: HHR of suitable areas in the south and some 
extension to the north, depending on its 

dispersion capability. By 2100, the GRAS 
scenario would make all lowland areas in 

| [Scenario | Full dispersal | No dispersal _| the European mainland unsuitable. Suitable 
SEDG conditions would only remain in the Alps, 
BAMBU Wales, Ireland, Scotland, Scandinavia and 
northern Finland. As B. humilis seems to 
oes have quite a low dispersal ability (being 
SEDG -6223 (-51%) unable to reach islands), and as it is already 

BAMBU 6893 (57%) 9741 (-80%) becoming scarce in most lowland areas, it 
would lose a noticeable suitable area because 

IUCN Red List status: Least Concern 

GRAS -7980 (-66%) -10623 (-87%) 

of global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


67 

Bombus humilis 


68 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus hyperboreus SCHONHERR, 1809 
= Bombus (Alpinobombus) hyperboreus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus hyperboreus is a very large bumblebee. The coat colour is constant and very typical: 
with 3 yellow bands and a black tail. It is a social parasite of B. polaris, B. jonellus and probably 
other Bombus species. It is a generalist forager. It lives in the Scandinavian mountains and 
along the northern tundra, reaching the Novaya Zemlya in Russia towards the north. It is 
a circumpolar species, present also in Russia, Alaska, Canada and even the north coast of 
Greenland where it reaches 84° N. The species is considered to be threatened: Vulnerable 
in the IUCN Red List of European 
Present distribution can be very well explained by Bees. Its modelled distribution includes 
eine gees ye Gn a oe the Alps where the species has never 
Climate risk category: HHHR been observed. All scenarios project a 
Pr aA Caen ae strong reduction of suitable areas in the 
lowlands. Only the coldest areas of the 
Scandinavian mountains would remain 
|| Scenario | Fall dispersal | No dispersal | suitable. All three scenarios project severe 

SEDG -1382 (-51%) -1384 (-51%) losses of areas with suitable conditions 
by 2100. Since Bombus hyperboreus 
shows a specialised way of life and as it 
-1496 (-55%) -1496 (-55%) is already rare, has a patchy distribution 
-2217 (-81%) -2217 (-81%) and is restricted to cold areas, it would 
lose a considerable amount of suitable 

-2584 (-94% -2584 (-94%) 
-2688 (-98%) -2688 (-98%) 

area which could lead to its extinction 

in Europe. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
== gain 
stable 
—__loss 

SEDG 
(2100) 

Bombus hyperboreus 

69 


70 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus hypnorum (L., 1758) 
= Bombus (Pyrobombus) hypnorum 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus hypnorum is a medium-sized bumblebee. ‘The coat colour typically shows a brown thorax, 
intermixed (to a greater or lesser extent) with an admixture of black hairs. The abdomen has a 
white tail. It builds large colonies, and nests in tree cavities, buildings, and bird-nest boxes. It is 
a generalist forager often associated with habitats strongly influenced by human activities. To the 
south, it reaches the Pyrenees and Balkan mountains and to the north, it reaches the Barents Sea 
coast. Eastwards, its range extends to the Pacific coast and to the west it has expanded its distribution 
considerably in recent times. Thirty years ago, it was absent from the coast of Brittany and from 
the British Isles. It arrived in England in 2001 and expanded its range very quickly, reaching 
Scotland in 2012. Since 2010, it has also been found in Iceland. The species is not threatened: 
Least Concern in the IUCN Red List of 
Present distribution can be well explained by climatic European Bees. Its modelled distribution 
variables (AUC = 0.80) includes the mountains of central Spain, 
Climate risk category: HR the Apennines and Ireland. All scenarios 
project a significant reduction of suitable 
areas in the lowlands. By 2100, the 
scenarios project a near extinction of 
| | Scenario | Full dispersal | No dispersal the species in all lowlands of Europe 

SEDG south of latitude 55° N. As B. hypnorum 

shows a clear and recent expansion of its 

IUCN Red List status: Least Concern 

BAMBU 
GRAS 

distribution area and as it is more or less 
synanthropic, it seems not threatened by 
-9007 (-54%) -9265 (-56%) global warming. However, even if this 
species presents clearly a high mobility, 

-10118 (61%)  -10374 (-62%) ' 
all scenarios project a noticeable future 

-11766 (-71%) -12022 (-72%) 

reduction of its suitable area. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
Es gain 
stable 
loss 

BAMBU 
(2050) 

Bombus hypnorum 

71 


72 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus incertus Morawt11z, 1882 

= Bombus (Melanobombus) incertus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus incertus is a medium-sized bumblebee. The coat colour shows 3 white bands and a red 
tipped abdomen. It is a generalist forager recorded from Armenia, Iran and Turkey, where it is 
one of the most abundant and ubiquitous bumblebees, but does not reach the Caucasus. It is not 
a threatened species but, as it is not a sensu stricto European species, it has not been assessed in 
the IUCN Red List of European Bees. Its 
Present distribution can be very well explained by modelled distribution shows that limited 
climatic variables (AUC = 1.00) areas in Europe are climatically suitable 
Climate risk category: HHHR for the species. Some scenarios project an 
expansion of its climatic niche space in 
IUCN Red List status: Not Evaluated : 
Europe or, alternatively, a strong regres- 
sion, depending on the dispersal ability 

| | Scenario _| Full dispersal No dispersal which remains unsettled, even if its small 

distribution area suggests low. As B. incer- 

SEDG 79 (20%) -294 (-74%) 

BAMBU -68 (-17%) -359 (-90%) in Turkey (but that never colonized Great 

GRAS -65 (-16%) -372 (-93%) Caucasus), it looks unlikely that the spe- 
SEDG -373 (-93%) cies would go extinct but global warming 

tus is an abundant and ubiquitous species 

could in the same time lead to a possible 
BAMBU -346 (-87%) -399 (-100%) . : ‘ 
expansion of its range in parts of Europe 

GRAS -314 (-79%) -399 (-100%) while it is vanishing in west Turkey. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus incertus 73 

SEDG 
(2050) 
ME gain 
Ey stable 
loss 

GRAS” 
(2050) 


74 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus inexspectatus (TKa.cv, 1963) 

= Bombus (Thoracobombus) inexspectatus 

© Photo: G. Mahé 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus inexspectatus is a medium-sized bumblebee. Its coat colour generally shows three 
greyish bands and a reddish tail. It is a cuckoo bumblebee which is hosted by B. ruderarius 
and maybe also by other closely related species. It occurs only in the Cantabrian Moun- 
tains and the Alps where it is extremely 
Present distribution can be very well explained by rare. Because of its low abundance and 
climatic variables (AUC = 0.98) restricted distribution, it is listed as En- 
dangered in the IUCN Red List of Euro- 
pean Bees. Its modelled distribution in- 
IUCN Red List status: Endangered cludes most of the mountain massifs of 

Climate risk category: HR 

Europe where it does not actually occur. 
All scenarios project a significant reduc- 

|| Scenario | Full dispersal | No dispersal tion of suitable areas by 2050. By 2100, 
SEDG areas with suitable climatic conditions 

BAMBU 266 (-15%) are projected to decrease drastically. As 

B. inexspectatus is a rare and highly spe- 
GRAS -329 (-18%) 

cialised species, with a restricted distri- 
-1070 (-60%) bution, and apparently a low dispersal 
-1215 (68%) capability, the considerable reduction of 
suitable conditions by global warming 

-918 (-51%) -1417 (-79%) 

could drive it to total extinction. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
| gain 
stable 
~—_#loss 

SEDG 
(2100) 

BAMBU 
(2100) 

Bombus inexspectatus 

75 


76 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus jonellus (Kirsy, 1802) 
= Bombus (Pyrobombus) jonellus 

© Photo: P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus jonellus is a small bumblebee. The coat colour is quite constant with 3 yellow bands and a 
white tail. It can be found in moors and heathlands in small colonies. Together with with B. terres- 
tris, it is one of the only two European bumblebees that are bivoltine. It forages on numerous flower 
species but when possible it prefers Ericaceae (Vaccinium spp., Erica spp., Rhododendron spp.). Its 
southernmost populations occur in the Pyrenees and Cantabrian mountains, while it reaches the 
Barents Sea coast to the north. It reaches Iceland in the west and Kamchatka in the east. In Europe, 
it lives in the lowlands north of latitude 50° N and in mountains and hills north of latitude 41° N. 
It can be very abundant in the northern parts of its range while it is very rare in the southern range 
margins such as in the Pyrenees. ‘The species is not considered to be threatened at the continental 
scale: Least Concern in the IUCN Red List of European Bees. Its modelled distribution includes 
some mountains in the south (Balkan, Carpathians) where it has never been observed. However, 
it is inconspicuous with a high chance of 
Present distribution can be well explained by climatic remaining unrecorded as has been the case 
eee reo! in Pyrenees for a long time. Regardless of 
Climate risk category: HR its dispersal capability, all scenarios project 
a reduction of suitable areas space mainly 
in the lowlands. In the worst case, GRAS 
projects a considerable reduction of the 

IUCN Red List status: Least Concern 

| | Scenario _| Full dispersal No dispersal 

SEDG and Scandinavia. As B. jonellus can be very 

suitable area mainly in the Alps, Scotland 

ARPT abundant, bivoltine and able to forage over 

GRAS 

such a long period of the year, it would be 
not directly threatened by global warming 
SEDG even if the area of its suitable areas could 
BAMBU 8131 (60%) 8131 (60%) considerably be reduced. Moreover, heath- 
lands and moors are habitats that could 

GRAS -9463 (-70%) -9463 (-70%) 

suffer a lot from warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
Ma gain 

—__ stable 
7_—_#loss 

GRAS 
(2100) 

Bombus jonellus 

77 


78 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus lapidarius (L., 1758) 
= Bombus (Melanobombus) lapidarius 

yw 

© Photo: A. Pauly 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus lapidarius is a medium-sized bumblebee. Its coat colour is quite constant in most 
parts of Europe: all black with a red tail (female) the head and thorax being more or less 
intermixed with yellow hairs in males. In Spain and Italy, the females can show three greyish 
yellow bands and a red tail (ssp. decipiens Pérez). It lives in large underground colonies in 
nearly all habitats. It is also a generalist forager. Its distribution extends from north Morocco, 
southern Spain, Sicily and southern Greece in the south to northern Sweden in the north. It 
occurs from Ireland in the west to the Ural Mountains in the east. It is generally abundant, 
and is not considered to be threatened at the continental scale: Least Concern in the IUCN 
Red List of European Bees. Its modelled 
Present distribution can be explained by climatic distribution more or less fits with its ac- 
variables to a moderate extent (AUC = 0.79) tual one except in north Scandinavia. All 
scenarios project a reduction of suitable 
Climate risk category: HHR c , : 

areas, mainly in the lowlands. This re- 
IUCN Red List status: Least Concern duction would still be inconspicuous in 

2050 but more drastic by 2100. In the 
worst case, GRAS projects that by 2100 

|| Scenario | Full dispersal | No dispersal there will be a considerable reduction of 
SEDG the suitable area in the Alps, Scotland 
BAMBU and Scandinavia and in some areas of the 
Pyrenees and central European moun- 

ea tains and hills. Even if B. lapidarius is a 
SEDG -7924 (-54%) ubiquitous species, generally abundant, 

BAMBU -11052 (-75%) with a potentially high dispersal capabil- 
ity, the area of its suitable areas would be 

GRAS -8604 (-59%) -12780 (-87%) 

considerably reduced. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus lapidarius 79 


Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus lapponicus (Fasricius, 1793) 
= Bombus (Pyrobombus) lapponicus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus lapponicus is a small bumblebee. Its coat colour is quite variable but always shows 
greyish hairs on the thorax and a largely red abdomen. It is very difficult to separate from 
B. monticola with which it coexists in most areas. It has small colonies in the Fennoscandi- 
an and Russian taiga and tundra where it can be abundant. It extends to the east all along 
the north-Siberian lowlands to the Pacific coast. It is a generalist forager. The species is not 
considered to be threatened: Least Concern in the IUCN Red List of European Bees. Its 
modelled distribution more or less fits 
with its actual one in northern Europe. 
Present distribution can be very well explained by However, the species has never been 
See Sa EN rE) observed in the British Isles or any Eu- 
Climate risk category: HHR ropean mountains south of latitude 60° 
N. All scenarios project a conspicuous 
reduction of suitable areas by 2050 
which becomes even more accentuat- 

ed by 2100. In the worst case scenario, 

| | Scenario | Full dispersal No dispersal GRAS projects that by 2100 the climat- 
SEDG ically suitable area will be restricted to 
BAMBU the Scandinavian mountains and will 

IUCN Red List status: Least Concern 

exclude all lowland areas. Regardless of 
GRAS its dispersal capability, as B. lapponicus is 
SEDG -3544 (-59%) -3555 (-59%) a typical northern species, closely linked 

-5051 (-85%) -5052 (-85%) with boreal taiga and arctic tundra, the 
area of its suitable climatic areas would 

-5177 (-87%) -5184 (-87%) 

be significantly reduced. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
ss gain 
stable 
____#l0ss 

SEDG 
(2100) 

BAMBU 
(2100) 

Bombus lapponicus 

81 


82 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus lucorum (L., 1761) 

= Bombus (Bombus) lucorum 

Doypn' if # m é 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus lucorum is a medium-sized bumblebee. Its coat colour generally shows two yellow bands and 
a white tip to the abdomen. The head and thorax of the males are very typically intermixed with nu- 
merous yellow and greyish hairs. Some specimens are extremely difficult to separate from B. magnus, 
B. cryptarum and B. terrestris. It lives in small- to medium-sized underground colonies in all habitats 
with a clear preference for forests and forest-edges. It is a generalist forager occurring from southern 
Europe north to the Barents Sea coast. To the west, it reaches Iceland and to the east, it occurs across 
northern Asia to the Pacific coast. It can be very abundant, especially towards the north of its range. 
However, since 2000, it becomes obviously much less abundant in Belgium and western France. De- 
spite this regional regression, the species is not considered to be threatened at the continental scale: 
Least Concern in the IUCN Red List of European Bees. Its modelled distribution more or less fits 
with its actual one. All scenarios project a reduction of suitable areas which will already be significant 
by 2050, especially in the south of England, 
Present distribution can be explained by climatic France, Belgium, the Netherlands and 
variables to a moderate extent (AUC = 0.82) across the central European lowlands. ‘This 
reduction is projected to be more drastic 
still by 2100. In the worst case, GRAS proj- 
TUCN Red List status: Least Concern ects that by 2100 a reduction of the suitable 
area will leave only the north of British Isles, 
; ; ; the Scandinavian mountains and the Alps, 
[Scenario | Full dispersal | No dispersal and exclude all European lowlands south 
SEDG of latitude 60° N. As B. /ucorum is a ubig- 
BAMBU uitous species that can be very abundant, 
possibly with a high dispersal capability, it 
is unlikely that it would become extinct in 
SEDG -8143 (-54%) Europe. However, the area of suitable cli- 
BAMBU -9468 (-63%) -9913 (-66%) matic conditions would be considerably 
reduced, leading to a substantial decrease in 
most European countries. 

Climate risk category: HR 

GRAS 

GRAS -10491 (-70%) -10652 (-71%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
MS gain 
stable 
loss 

BAMBU 
(2050) 

Bombus lucorum 

83 


84 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus magnus Vocrt, 1911 
= Bombus (Bombus) magnus 

© Photo P. Rasmont 

Bombus magnus is a medium- to large-sized bumblebee. 

Its coat colour generally shows two yellow bands, extend- 
ing low on the side of the thorax, and a white tail. The 
head and thorax of the males are always intermixed with Dots: actual distribution 1970-2000; yellow areas: 
numerous yellow (but no greyish) hairs. Some individuals HEE ere alae, & pat ome wore epee 
are difficult to separate from B. lucorum and B. cryptarum 
and this taxonomic uncertainty certainly explains why there are relatively few verified records for this species. 
It makes small- to medium-sized underground colonies in heath lands and moorlands. It is clearly most 
common in the region with oceanic influences. As it is a quite difficult species to identify with certainty, its 
distribution is not completely known. As far as we know, it occurs from north Portugal in the south to the 
north of the Arctic Circle along the west Norwegian coast. It reaches Ireland to the west, where it is abun- 
dant, to isolated locations around Moscow to the east, where it is very rare. It is a generalist forager but has a 
preference for Ericaceae (e.g. Vaccinium spp., Erica spp., Rhododendron spp.). The species is not considered to 
be threatened at the continental scale: Least Concern in the IUCN Red List of European Bees. Its modelled 
distribution more or less fits with its actual one, even if it is absent from Corsica, the Apennine mountains 
and probably from Durmitor and the Balkans. 
Present distribution can be explained by climatic All scenarios project a reduction of suitable areas 
variables to a moderate extent (AUC = 0.85) that is already significant by 2050, especially in 
the French lowlands. This reduction is projected 
to be more drastic by 2100. In the worst case, 
TG Nioeduletatatie Tecan @oncern GRAS projects that by 2100 a reduction of the 
climatically suitable area will confine the species 
to the north of the British Isles and Scandina- 
| [Scenario | Full dispersal | No dispersal via. Reductions are also projected for the Alps 
SEDG but here the species is already extremely rare. 
Regardless of its dispersal capability, as B. mag- 

nus is a bumblebee with quite narrowly defined 

Climate risk category: HR 

BAMBU 

GRAS habitat preferences, clear flower preferences 
and climatic preferences linked with oceanic 
4401 (51%) 9029 59%) climates, it would be strongly affected by a sig- 
-4658 (-54%) -5682 (-66%) nificant reduction of suitable areas. It is unlikely 
that this species would become extinct but it 

-5312 (-62%) -6321 (-74%) 

could disappear from most European countries. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus magnus 85 

SEDG 
(2100) 


86 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus mendax GERSTAECKER, 1869 
= Bombus (Mendacibombus) mendax 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus mendax is a medium-sized bumblebee. Its coat colour generally shows three greyish 
bands and a reddish tail. The male has conspicuously enlarged eyes. Bombus mendax lives in 
small underground colonies restricted to high alpine and subalpine areas. It is a generalist for- 
ager but prefers flowers with long corollas such as Trifolium spp. or Aconitum spp. It occurs 
only in the Cantabrian Mountains (where it is very rare), Pyrenees and Alps. Because of its 
restricted distribution, the species is considered to be Near Threatened in the IUCN Red List 
of European Bees. Its modelled distribution includes the Apennines and Scandinavian moun- 
tains where the species has never been ob- 

Present distribution can be very well explained by served. All scenarios P roject a reduction 
climatic variables (AUC = 0.97) of its climatic niche that will already be 
een significant by 2050. In the worst case, the 
GRAS scenario projects reduction of the 

IUCN Red List status: Nearly Threatened suitable climatic conditions to a restricted 

area in the Alps and an even smaller area 

in the Pyrenees by 2100. Such a situation 
|| Scenario Full dispersal No dispersal could lead it close to extinction. As B. 
SEDG -376 (-18%) mendax is a bumblebee that is restricted 

BAMBU to the highest alpine and subalpine levels 
of the Cantabrian Mountains, Pyrenees 
hata and Alps, is likely to have a low dispersal 
-1158 (-55%) capability, global warming is projected to 
-1403 (-66%) reduce the area of suitable climatic condi- 
tions considerably, leaving the species on 

-1162 (-55%) -1676 (-79%) 

the verge of extinction by 2100. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus mendax 87 

SEDG 
(2050) 
= gain 
stable 
loss 

BAMBU 
(2050) 

GRAS 
(2100) 


88 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus mesomelas GERSTAECKER, 1869 
= Bombus (Thoracobombus) mesomelas; Bombus (Rhodobombus) mesomelas; 
Bombus elegans (partim) 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus mesomelas is a medium- to large-sized bumblebee. Its coat colour is grey with a black 
thoracic band. In the Balkan peninsula and Turkey, the abdomen shows a conspicuous reddish 
tinge (ssp. alboluteus). Bombus mesomelas lives in large underground colonies and is restrict- 
ed to mountain meadows. It is a generalist forager but it clearly prefers flowers with long 
corollas such as Trifolium spp. (Fabaceae). It occurs in the Cantabrian Mountains, Pyrenees, 
Alps, Apennines, Balkans, and Carpathians. It has disappeared from lower montane regions 
of central Europe, such as Harz mountains and Krkonose were it was living one century ago. 
Despite this regression, the species is not considered to be threatened at the continental scale: 
Least Concern in the IUCN Red List of European Bees. Its modelled distribution includes 
large areas of European mountains and hills where the species has never been observed. All 
scenarios project a significant reduction of suitable areas by 2050, especially in the low moun- 
tains of central Europe (from where it 
Present distribution can be well explained by climatic already disappeared). The expansion 
A eee area in Scandinavia that could appear as 
Climate risk category: HR a counterbalance is nevertheless out of 
IUCN Red List status: Least Concern reach. This reduction is even more Bro 
nounced by 2100. As B. mesomelas is a 

bumblebee restricted to mountain mead- 

|| Scenario _| Full dispersal No dispersal ows, and is likely to have a low dispersal 

SEDG capability, global warming would lead 

BAMBU to considerable reductions of its suitable 
areas, especially in the central Europe- 
an mountains, then in the Balkans and 
-1268 (-16%) Carpathians, and even in the Cantabri- 
TAS (61%) an mountains and Pyrenees. However, 

it is quite unlikely that global warming 

-4136 (-53%) -5983 (-77%) 

would lead to extinction of the species. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus mesomelas 89 

SEDG 
(2100) 


90 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus monticola Smitn, 1851 
= Bombus (Pyrobombus) monticola 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus monticola is a small bumblebee. Its coat colour is quite variable but always with an 
abdomen that is largely red and with more or less greyish hairs intermixed with black on 
thorax and head. It is difficult to separate from B. lapponicus with which it coexists in Fen- 
noscandia. It lives in small colonies in Fennoscandian taiga and tundra and in alpine and 
subalpine meadows in Wales, Scotland, the Cantabrian mountains, Pyrenees, Alps, Apen- 
nines, Balkans and the Olympus range (Greece). It seems to recently have colonised Ireland 
in the late 1970’. Bombus monticola can be locally abundant across its range and does not 
occur outside Europe. It is a generalist forager. The species is not considered to be threatened 
at the continental scale: Least Concern in the [UCN Red List of European Bees. Its modelled 
distribution more or less fits with its actual one but it is clear that the species does not occur 
in highly isolated mountains and hills. It is noticeable that the modelled distribution does 

not include either the Olympus range or 
Present distribution can be well explained by climatic 

the Apennines, where the species has ex- 
vatiables (AUC = 0.88) 

perienced a significant and recent reduc- 
Climate risk category: HR tion in abundance. All scenarios project 
TaN aD nen a significant reduction of suitable areas 
by 2050 and an even more drastic re- 
duction by 2100. In the worst scenario, 
|| Scenario | Full dispersal | No dispersal GRAS projects a reduction by 2100 of 

SEDG the suitable area which will exclude this 

BAMBU species from Mt Olympus, the Apen- 

GRAS 

nines, Cantabrian mountains and even 
the Pyrenees. Regardless of its disper- 
SEDG -4556 (-57%) -4616 (-58%) sal capacity, as B. monticola is a typical 

4916 (62%) 5019 (-63%) mountain species, linked with cold ar- 
eas, the area of suitable climatic condi- 

-5777 (-72%) -5915 (-74%) 

tions would be significantly reduced. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
ME gain 
stable 
loss 

SEDG 
(2100) 

Bombus monticola 

91 


Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus mucidus GERSTAECKER 1869 

= Bombus (Thoracobombus) mucidus; Bombus (Mucidobombus) mucidus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus mucidus is a small bumblebee. Its coat colour is generally greyish with a black tho- 
racic band; the thoracic and abdominal hairs can be more or less intermixed with black and 
some specimens from the Alps are mainly black with a greyish tail. It lives in small colonies in 
subalpine and alpine meadows and forages mainly on Fabaceae (queens, workers) or thistles 
(males). The species occurs only in the Cantabrian Mountains, Alps, Apennines, Carpathians 
and Balkans. It is generally a rare species, and is endemic to Europe. ‘The species is considered 
as Near Threatened in the IUCN Red List of European Bees. Its modelled distribution covers 
more mountainous areas in which the species does not occur, perhaps due to a lack of sufficient 
dispersal ability. All scenarios project a 

Present distribution can be well explatned by climatic substantial reduction of suitable areas by 
rae Se iat a) 2050 and a more drastic reduction by 
Climate risk category: HR 2100. In the worst case, GRAS projects 

that by 2100 the reduction of the suit- 
able area would exclude the species from 
the Balkan, Carpathians, and Cantabrian 

| | Scenario | Full dispersal No dispersal mountains and probably the Pyrenees. 
SEDG -543 (-17%) Bombus mucidus is already considered as 

a rare species. It is associated with high 
BAMBU 

IUCN Red List status: Least Concern 

mountains and is assumed to have a low 
GRAS dispersal ability. A significant reduction 
in the area of its suitable climatic space 
could lead to extinction in the Cantabri- 

-2013 (-62%) 
-1730 (-53%) -2435 (-75%) 

an mountains, Pyrenees, Appenines, Bal- 

kan mountains and the Carpathians. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 

ma gain 
stable 
loss 

Bombus mucidus 

93 


94 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus muscorum (L., 1758) 

= Bombus (Thoracobombus) muscorum 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus muscorum is a medium-sized bumblebee. Its coat colour is generally completely orange. 
Most populations of western islands, northern Scandinavia and northern Russia have black hairs 
on the legs (ssp. bannitus) while the Corsican ssp. is very dark, with the thorax almost entirely 
black (ssp. pereziellus). It lives in large colonies in grass-tussocks in various landscapes but with 
a clear preference for areas close to the sea coasts. It prefers to forage on Fabaceae flowers. The 
species occurs in most parts of Europe but it is very rare south of latitude 40° N. It has not been 
observed in many of the locations of west, central and south-east Europe where it occurred one 
century ago. Outside Europe, it reaches Mongolia in the east. The species is considered to be 
threatened: Vulnerable in the IUCN Red List of European Bees. Its modelled distribution fits 
with its present distribution but excludes large areas of former occurrence. All scenarios project a 
significant reduction of suitable areas by 2050 and an even more drastic reduction by 2100. Most 
of the areas from which the species has already disappeared, but were modelled as suitable under 
current conditions, are areas that are pro- 
jected to become unsuitable in the future. 
Suitable climatic conditions remain only 
Climate risk category: HHR in Ireland, north of the British Isles (where 
MUON Reduliceerinis. Valneeble it is already scarce) and in Scandinavia 
north of latitude 60° N (where it is not 
| | Scenario | Full dispersal | No dispersal — very abundant). Bombus muscorum is cur- 
rently declining where its extinction is pro- 
jected for the future. It is assumed to have 

-5058 (-51%) a low dispersal capability. The area of its 
-5536 (-56%) -5851 (-59%) suitable climatic niche would be consid- 

Present distribution can be explained by climatic 
variables to a moderate extent (AUC = 0.83) 

erably reduced leading to local extinctions 
across large parts of Europe. The potential 
of the species to persist in the remaining, 

-5816 (-59%) -6990 (-71%) 
-6775 (-69%) -7926 (-80%) 
-7080 (-72%) -8470 (-86%) much smaller area is questionable. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus muscorum 

SEDG 
(2050) 

= gain 
stable 
loss 

BAMBU 
(2050) 

95 


96 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus niveatus KRIECHBAUMER, 1870 

Bombus (Sibiricobombus) niveatus; Bombus vorticosus 

© Photo G. Pisanty 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus niveatus is a large bumblebee. On the European continent, its coat colour is very con- 
stant with three yellowish bands and a red tail (ssp. vorticosus Gerstaecker). In Turkey, Arme- 
nia, Georgia, Azerbaijan, the Caucasus and Iran, a form occurs where the yellowish bands are 
replaced by white ones (ssp. niveatus s.s.). The males have large eyes and fly very fast. It lives in 
sizeable colonies that are often established in cavities, sometimes after ousting the bird that was 
occupying the nest site. It’s a generalist forager even if (as a long-tongued species) it prefers flow- 
ers with deep corollas. In Europe, the species occurs in the Balkans only. Elsewhere it is found 
in Turkey, Georgia, the Caucasus, Azerbaijan and Iran. It is abundant throughout most of its 
range. Ihe species is not considered to be threatened at the continental scale: Least Concern in 
the IUCN Red List of European Bees. Its 

Present distribution can be well explained by climatic modelled distribution is much larger than 
variables (AUC = 0.95) its actual one. All scenarios project a con- 
siderable enlargement of its suitable area 
by 2050. This tendency would be even 
IUCN Red List status: Least Concern more pronounced by 2100, with suitable 
area even extending to western Europe 

Climate risk category: HHR 

and Scandinavia. However, if the species 

| | Scenario _| Full dispersal No dispersal is limited by dispersal, it would be at risk, 

SEDG 523 (-10%) as its original area could become unsuit- 
able (under the GRAS scenario). As it is 
an abundant species, with an assumed 
GRAS high dispersal capability, and adapted to 

warm and dry climatic conditions, B. ni- 

BAMBU 

veatus could benefit from global warming 

-4381 (-84%) 
-4586 (-87%) 

as its climatic suitable area would expand 
considerably. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
Me gain 
—_ stable 
loss 

(2050) 

(2100) 

Bombus niveatus 

97 


Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus norvegicus SPARRE SCHNEIDER, 1918 

= Bombus (Psithyrus) norvegicus; Psithyrus norvegicus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus norvegicus is a small- to medium-sized bumblebee. Its coat colour generally shows a 
large prothoracic band and a white tail. Confusions are frequent with the closely related species 
B. sylvestris. \t is a specialised social parasite species (cuckoo-bumblebee) mostly of B. hypno- 
rum. Its southernmost populations are in the Cantabrian mountains and Pyrenees. It occurs 
as far north as the Arctic Circle, but is not known from the British Isles or in the Balkan 
peninsula. To the east, it reaches the Pacific coast. It is everywhere a rare species. The species is 
not considered to be threatened at a continental scale: Least Concern in the [UCN Red List 
of European Bees. The modelled distribution is much wider than its actual one, extending to 
central Spain, the Apennines, British Isles 

Present distribution can be explained by climatic and the Balkans, where the sp ecies has 
variables to a moderate extent (AUC = 0.81) never been observed. All scenarios project 
a substantial decline in suitable climatic 
Climate risk category: HR ; 
area in central and southern Europe by 

IUCN Red List status: Least Concern 2050. By 2100, no suitable area would 
remain in the lowlands of west, central 
and eastern Europe. The only climatically 

| | Scenario _| Full dispersal No dispersal suitable areas would remain in the Alps 

SEDG and in Scandinavia. Bombus norvegicus is 

a cuckoo-bumblebee specialised on a sin- 
BAMBU -1855 (-14%) 

gle host-species, and is believed to have a 

GRAS low dispersal capability. It already occurs 

SEDG -8345 (-63%) in a much more restricted area than cli- 
mate conditions would allow. One could 

BAMBU -6761 (-51%) -9412 (-71%) 
GRAS -8510 (-64%) -10611 (-80%) 

expect a considerable reduction of its dis- 

tribution caused by global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDGE 
(2050) 
Es gain 
stable 
loss 

(2050) 

SEDG 
(2100) 

Bombus norvegicus 

99 


100 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus pascuorum (Scopo.t, 1763) 

= Bombus (Thoracobombus) pascuorum; Bombus agrorum auctt. 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus pascuorum is a medium-sized bumblebee. The coat colour is extremely variable from region to 
region (as different subspecies), making B. pascuorum the most polytypic of the European bumblebees. 
Colour schemes are typically brownish, with a greater or lesser amount of black and grey hairs intermixed. 
It is a generalist forager that visits all available flowers. In Europe, it can be found from southern Spain, 
Greece and Turkey in the south, where it can reach the Mediterranean coast (even if it is mainly confined 
to hills and mountains), to the Barents Sea coast in the north. To the west, it reaches Ireland and the Pacif- 
ic coast in the east. It has recently colonised Iceland. ‘The species is present on some Mediterranean islands, 
such as Corsica and Sicily. It occurs in any kind of habitats, but it prefers woody landscapes. The species 
is not considered to be threatened: Least Concern in the IUCN Red List of European Bees. It is, by far, 
the most widespread and the most abundant European bumblebee and even in heavily human-influenced 
landscapes, where other bumblebee species are scarce, B. pascuorum remains abundant. The species distri- 
bution model underestimated the current range to some extent and was not able to reproduce occurrences 
in parts of Finland, Sweden, Poland, southern 
Spain, southern Italy and Sicily. All scenarios 

Present distribution can be explained by climatic i : : 
project a moderate reduction of suitable areas 

vatiables to a moderate extent (AUC = 0.75) 
especially in western France. By 2100, in the 

Climate risk category: R worst case, the GRAS scenario would lead to 
unsuitable conditions in all lowland regions of 
the European mainland. Suitable conditions 
would remain only in the Alps, Wales, Ireland, 
| | Scenario | Full dispersal | No dispersal Scotland, Scandinavia and northern Finland. 
However, B. pascuorum is highly polytypic 

SEDG (one potential reason for the just moderate 

BAMBU -1713 (-13%) -2283 (-17%) model performance) and it can be expected 
that the plasticity of the regional populations 
GRAS -2443 (-18%) 

SEDG matic variation. Bombus pascuorum seems to 
have a high dispersal capability, as it is very 

IUCN Red List status: Least Concern 

would allow the species to adapt to local cli- 

BAMBU -7477 (-55%) 

abundant. However, it could lose significant 
GRAS -8253 (-61%) -8912 (-66%) parts of its climatically suitable area because of 
global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus pascuorum 101 


102 Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus polaris Curtis, 1835 

= Bombus (Alpinobombus) polaris, Bombus arcticus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus polaris is a large bumblebee occurring in the Arctic tundra and alpine grasslands in 
Scandinavia. Its colour coat is quite invariable: it has a black thorax and an abdomen that 
is largely covered in faded reddish hairs. It lives in very small colonies and it is a generalist 
forager. It can be found at the sea level in Arctic tundra along the northern coast of Norway 
and Russia and in Novaya Zemlya, North America and Greenland. ‘The species is currently 
declining: assessed as Vulnerable in the IUCN Red List of European Bees. Its modelled 
distribution shows that its climatic niche space might be larger than its actual distribution, 
and includes the higher elevations of the 
Present distribution can be very well explained by Alps and also Iceland, where it has nev- 
climatic variables (AUC = 0.99) er been found. Even in Scandinavia, the 
Climate risk category: HHHR total area of its presently suitable area 

is larger than its actual distribution. All 
IUCN Red List status: Least Concern 

| | Scenario Full dispersal No dispersal 

SEDG -2038 (-57%) -2038 (-57%) By 2100, the reduction of suitable area 
would reach 90% or even more. As the 

scenarios project that its suitable areas 
will be considerably reduced by 2050 

(more than 50% of the current area). 

BAMBU -1955 (-55%) -1955 (-55%) 

species is rare and linked to cold climates 
-2184 (-61%) -2184 (-61%) 

in high Alpine and Arctic habitats, there 
-3115 (-87%) -3115 (-87%) is little chance that the populations will 
3459 (-97%) 3459 (-97%) remain numerous enough to allow for 
the survival of the species in Europe, re- 

gardless of its dispersal capability. 

-3524 (-99%) -3524 (-99%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus polaris 103 

SEDG 
(2050) 
ma gain 
stable 
loss 

SEDG 
(2100) 

BAMBU 

BAMBU 
(2050) 

(2100) 


104 Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus pomorum (Panzer, 1805) 

= Bombus (Thoracobombus) pomorum; Bombus (Rhodobombus) pomorum 

© Photo D. Genoud 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus pomorum is a medium- to large-sized bumblebee. In Europe, its coat colour is black 
with a reddish hind part of the abdomen. In males, the black coat can be more or less inter- 
mixed with grey. Bombus pomorum lives in large underground colonies generally in dry bushy 
grasslands. It is a generalist forager but it clearly prefers flowers with long corollas like Trifolium 
spp. (Fabaceae). It occurs from latitude 42° in the Balkan to 55° N in southern Scandinavia 
and from north-west France and the Massif Central in the west to the Ural and Caucasus 
mountains in the east. It is absent from the Pyrenees, and the Iberian and Italian peninsulas. It 
is considered to be extinct in England, Sweden, Denmark, The Netherlands, Belgium and Lux- 
embourg. Its modelled distribution already integrates this actual reduction of its distribution 
area. This modelled distribution also shows that its suitable climatic area includes the Pyrenees, 

Iberian and Italian peninsula and south- 
Present distribution can be explained by climatic 

ern Turkey, too, despite that, it has never 
variables to a moderate extent (AUC = 0.85) 

been found there. The species is consid- 
Climate risk category: HHHR ered to be threatened: Vulnerable in the 
IUCN Red List of European Bees. It 

never reached some parts of its potential 

IUCN Red List status: Vulnerable 

area, meaning that its dispersion capabil- 

|| Scenario Full dispersal No dispersal ity is likely to be quite low. All dispersal 

SEDG -725 (-10%) scenarios project a considerable reduction 
of the climatically suitable area by 2050, 
BAMBU 
and further reductions by 2100. Because 
Gee cee ey) 4387 (-62%) of its present scarcity in its already declin- 
SEDG -5304 (-75%) ing range, allied with the seemingly low 

BAMBU 5335 (75%) 6856 (97%) dispersal capability of B. pomorum, the 
considerable reduction of its suitable area 

GRAS -4716 (-67%) -6952 (-98%) 

could drive the species to extinction. 
Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus pomorum 105 

SEDG 
(2100) 


106 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus pratorum (L., 1756) 

= Bombus (Pyrobombus) pratorum 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus pratorum is a small bumblebee. Its coat colour is quite variable but always black with 
a red tail and with or without one to three yellow bands. B. pratorum has small colonies which 
are often established in grass tussocks in forests or forest edges. It is a generalist forager. It occurs 
from southern Spain and southern Italy (where it is found only in the mountains) northwards to 
the Barents Sea coasts and from Ireland in the west to the Pacific coast in the east. Its modelled 
distribution fits with the current distribu- 

Present distribution can be explained by climatic tion. ‘The species is one of the most wide- 
variables to a moderate extent (AUC = 0.80) spread and abundant of the European 
bumblebees. Like Bombus pascuorum, it 
Climate risk category: HR ts hems 
can survive in urban and surburban areas 

IUCN Red List status: Least Concern where other bumblebees are scarce. The 
species is not considered to be threatened: 

Least Concern in the IUCN Red List of 
|| Scenario | Full dispersal | No dispersal European Bees. All scenarios project a re- 
SEDG duction of its climatically suitable area. 
BAMBU As it often occurs in high abundances, 
it is a generalist forager, and can occupy 

a most habitats, a high dispersal capability 
SEDG -9377 (-52%) -9472 (-52%) is expected. Bombus pratorum would not 

PAWVUOMEE 10864 (60%) — -10896 (-60%) be threatened by global warming, despite 
a possible significant reduction of its suit- 

GRAS -13243 (-73%) — -13244 (-73%) 

able climatic area. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus pratorum 107 


108 = Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus pyrenaeus PEREZ, 1879 
= Bombus (Pyrobombus) pyrenaeus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus pyrenaeus is a small bumblebee. Its coat colour is quite variable but generally with a large 
greyish thorax with a black thoracic band, with a grey base of the abdomen and with a large part 
of the remaining abdomen being reddish. B. pyrenaeus lives in small colonies in grass tussocks 
or under rocks in subalpine and alpine meadows. It is a generalist forager with some preferences 
for Ericaceae (Vaccinium spp., Rhododendron spp.). It occurs in the Pyrenees, Alps, Tatra, Car- 
pathians, Durmitor and Balkan mountains, with quite conspicuously different subspecies. Its 
modelled distribution includes much more mountain areas than its actual distribution (e.g. in 
UK where it never occurred). The species 
Present distribution can be well explained by climatic is not considered to be threatened: Least 
variables (AUC = 0.93) Concern in the IUCN Red List of Euro- 
Climate risk category: HR pean Bees. As the species seems strongly 
differentiated from mountain to moun- 
IUCN Red List status: Least Concern . : . : 
tain and as larger parts of suitable climatic 
areas are currently not populated, its dis- 

| | Scenario _| Full dispersal No dispersal persal ability is assumed to be low. There- 

SEDG 633 (-18%) fore, the no dispersal scenarios seem to be 
the most likely. Both BAMBU and GRAS 
scenarios show considerable reductions in 
GRAS the Balkan and Carpathian mountains. 
SEDG Despite such reduction of climatically 

BAMBU 

suitable area, B. pyrenaeus would not be 

threatened to the point of extinction by 
-2512 (-71%) global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

-2239 (-63%) 


Bombus pyrenaeus 109 

SEDG 
(2050) 
Me gain 
stable 
loss 

SEDG 
(2100) 

BAMBU 
(2050) 

BAMBU 
(2100) 


110 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus quadricolor (LEPELETIER, 1832) 
= Bombus (Psithyrus) quadricolor; Psithyrus quadricolor 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus quadricolor is a medium-sized bumblebee. Its colour is quite variable: with or without a 
large yellow prothoracic band, with or without yellow band at the base of the abdomen, with a 
red or a white and red tail. It is a specialised social parasite species (cuckoo-bumblebee) mainly 
of B. soroeensis. Its southernmost populations occur in the Cantabrian mountains, Pyrenees and 
Balkan mountains and it reaches a latitude of 65° N. To the east, it reaches the Altai mountains. 
The species is rare and recently declined in most areas. This decline is probably associated with 
the decline of its main host B. soroeensis. Despite this regression, the species is not considered 
to be threatened at the continental scale: Least Concern in the IUCN Red List of European 

Bees. The modelled distribution is some- 

Present distribution can be explained by climatic what wider than the actual one. All sce- 

climatic space in the lowlands of central 
Climate risk category: HR : 

and southern Europe and an expansion of 
IUCN Red List status: Least Concern its climatic niche space to the Barents Sea 

coast and to the Scandinavian mountains. 

By 2100, all scenarios project that its suit- 

| | Scenario | Full dispersal | No dispersal — able areas would completely disappear 
SEDG from all lowlands south of latitude 60° N. 
BAMBU As it is a species which is likely to have a 

low dispersal capability, is specialised on 

eae only one main host species, is currently 
SEDG -5388 (-67%) declining and as its climatically suitable 

BAMBU -6245 (-78%) area would be considerably reduced, B. 
quadricolor would suffer considerably 

GRAS -4116 (-51%) -6773 (-84%) 

from global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


111 

Bombus quadricolor 


112 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus ruderarius (MULLER, 1776) 

= Bombus (Thoracobombus) ruderarius 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus ruderarius is a medium-sized bumblebee. ‘The coat colour is generally black with a reddish 
tail with more or less greyish hairs intermixed on both thorax and abdomen (f. montanus Lepele- 
tier). North African populations have three yellow bands and a white tail (ssp. tunensis (Tkalct)). 
It is a generalist species even if it prefers to forage on Lamiaceae, Fabaceae and Asteraceae flowers. 
In Europe, it can be found from central Spain, southern Italy and northern Greece in the south, 
(where it is restricted to the mountains), to the Arctic Circle in the north. To the west, it reaches 
Ireland and north-west Spain and in the east it occurs to the Altai. There is an isolated population 
in Tunisia and north-east Algeria. It has declined in lowland areas of western and central Europe. 
Despite this regional regression the species is not considered to be threatened at a continental scale: 
Least Concern in the IUCN Red List of European Bees. Its modelled distribution includes some 
areas which the species does not reach, such as northern Scandinavia, but it excludes north Africa. 
By 2050, all scenarios project a moderate reduction of the suitable areas in the south and some 
extension to the north, depending on its 
dispersion capability. By 2100, all scenarios 
project a significant reduction of suitable 
Climate tisk category: HHR areas in the lowlands of Europe. In the 
worst case, the GRAS scenario would lead 
to unsuitable conditions in all lowland re- 

gions of the European mainland and even 

|| Scenario | Full dispersal | No dispersal _ in most of the mountains, depending on 
SEDG its dispersal ability. Bombus ruderarius is 
already becoming scarce in most lowland 

Present distribution can be explained by climatic 
variables only to a limited extent (AUC = 0.75) 

IUCN Red List status: Least Concern 

BAMBU 
GRAS 

locations, the projected losses of suitable 
climatic conditions can be assumed to 
SEDG have severe consequences for this species 
and might lead to extinction from most of 
the temperate lowland areas and southern 

-12331 (-88%) 
-13187 (-94%) 

mountains of Europe. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 
Ma gain 
—_ gtable 
C__§loss 

GRAS 
(2100) 

Bombus ruderarius 

113 


114 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus ruderatus (Fasricius, 1775) 
= Bombus (Megabombus) ruderatus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus ruderatus is a large bumblebee with a variable coat colour. In the main part of Europe and north 
Africa, it shows three yellow bands and a white tail, however, it can also be all black, as in some English 
populations (ssp. perniger (Harris)), all black with a red tail, as in Corsica (ssp. corsicola Strand) or with 
mixed black and reddish hairs and with a white tail as in Sardinia (ssp. sardiniensis Tournier). It is a generalist 
species even if it prefers to forage on flowers with deep corollas which it can access thanks to its very long 
tongue. It has been recorded from the Azores, north Africa, and the Iberian and Italian peninsulas in the 
south, reaching northern England and southern Scandinavia in the north. To the east, it reaches Ukraine 
but it does not occur at all in south-eastern Europe where it is replaced by its sibling species B. argillaceus. 
It has become very rare in most parts of its range. Despite this regression, the species is not considered to 
be threatened: Least Concern in the IUCN Red List of European Bees. Its modelled distribution includes 
some areas where the species does not occur, such as Romania, the Balkans and Turkey. By 2050, all scenarios 
project a moderate reduction of suitable areas of the species in the south and some extension to the north, 
depending on its dispersal capability. In 2100, all scenarios project a significant regression of suitable areas in 
the lowlands of Europe, with some expansion 
Present distribution can be explained by climatic towards the east. However, in most of the lo- 
variables to a moderate extent (AUC = 0.78) cations where suitable climatic conditions are 
projected to persist the species already suffers 

Slat ack A een au presently from strong declines, meaning that 

THIGN: Recolasieetanaslicieneonceen other parameters are affecting the population 
dynamics. In the worst case, the GRAS scenario 
projects unsuitable conditions for all lowlands 

|| Scenario | Full dispersal | No dispersal south of latitude 48° N, and its extinction in all 
SEDG the Mediterranean countries where it presently 
thrives. As B. ruderatus is already becoming rare 

BAMBU in most lowland locations, it is projected to lose 
GRAS a noticeable amount of suitable area because 

of global warming, regardless of its dispersal 
SEDG ability. Additionally, some other ecological or 

BAMBU -6148 (-53%) -8045 (-69%) anthropogenic factors seem to play a role in its 
present regression, thus adding threatening risk 

GRAS -6912 (-60%) -9029 (-78%) 

in a quite unpredictable but negative way. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus ruderatus 

SEDG 
(2050) 

Ma gain 
stable 
loss 

(2100) 

BAMBU 
(2050) 

BAMBU 
(2100) 

GRAS 
(2100) 

115 


116 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus rupestris (Fasricius, 1793) 
= Bombus (Psithyrus) rupestris; Psithyrus rupestris 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus rupestris is a medium- to large-sized bumblebee. Its colour coat is mainly black with a red 
tail. The thorax could be more or less intermixed with grey. The wings are conspicuously darkened. 
It is a specialised social parasite species (cuckoo-bumblebee) mainly on B. lapidarius and B. sichelii. 
It can be found from central Spain, Sicily and the northern Balkans in the south up to latitude 65° 
N in Scandinavia, remaining absent from Scotland and also from all Mediterranean islands, except 
Sicily. To the east, it reaches the Pacific coast. It might be one of the most common cuckoo-bum- 
blebees, but it seems to become rarer in some locations, especially in The Netherlands, Belgium 
and Germany. Despite this regional regression, the species is not considered to be threatened at a 
continental scale: Least Concern in the IUCN Red List of European Bees. The modelled distri- 
bution corresponds quite well to the actual one, with some discrepancies along the Norwegian 
coasts, central Spain and Turkey. All sce- 
Present distribution can be explained by climatic narios project fragmentation in central and 
variables to a moderate extent (AUC = 0.78) Southern Europe avd expansion aie cle 
Climate risk category: HHR matic niche space to the Barents Sea coast. 
By 2100, this tendency would lead to 
IUCN Red List status: Least Concern ‘ P : ; : 
considerable reduction of suitable climatic 

conditions in the lowlands of western Eu- 
| | Scenario | Full dispersal | No dispersal rope. Ihe GRAS scenario projects that its 

SEDG suitable areas would completely disappear 
from all lowlands south of latitude 60° N 
by 2100. As it is a cuckoo-bumblebee and 
GRAS is restricted to just 2 host-species and as it 

SEDG -6702 (-51%) -8699 (-66%) seems already regressing in temperate Eu- 
8170 (-62%) 11149 (-84%) ropean lowlands, B. rupestris would suffer 
- -62% - -84% 

-9631 (-73%) -12224 (-92%) 

BAMBU 

significantly from global warming, regard- 

less of its dispersal abilities. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus rupestris 117 

SEDG 
(2050) 
Me gain 
stable 
> Bless 

GRAS 
(2050) 


118 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus schrencki (Moraw1z, 1881) 

= Bombus (Thoracobombus) schrencki 

© Photo O. Korsun 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus schrencki is a medium-sized bumblebee. The coat colour is brown on the thorax and 
on the base of the abdomen and black with thin grey bands on the rest of the abdomen. It 
occurs in north-eastern Europe, from latitude 55° N in the south to 70° N in the north and 
from eastern Poland in the west eastwards to the Pacific coasts. The species is not considered to 
be threatened: Least Concern in the IUCN Red List of European Bees. It seems to have ex- 
panded recently towards the west, now reaching Poland and Finland. Its modelled distribution 
includes some areas which the species does not reach, especially in Scandinavia and Finland. By 
2050, all scenarios project an extension of 

Present distribution can be very well explained by its suitable climatic niche Space towards 
climatic variables (AUC = 0.96) the north, in Scandinavia and Finland (to 
where the species expanded recently) and 
a noticeable reduction of suitable areas in 
IUCN Red List status: Least Concern the south. These tendencies would con- 
tinue by 2100, leading to an almost total 

extinction if the species cannot disperse 

|| Scenario | Full dispersal | No dispersal sufficiently. The recent and quite rapid 
SEDG expansion of the species suggests a high 
dispersal ability. However, even with the 

Climate risk category: HHHR 

BAMBU -3001 (-55%) 
GRAS -3528 (-65%) 

SEDG -4814 (-88%) would become much smaller than its 
BAMBU 5357 (-98%) present range. Even if it is not threatened, 

best expansion capability, the remain- 
ing suitable climatic area of B. schrencki 

the species would considerably suffer 

GRAS -2729 (-50%) oe) from the global warming 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


119 

Bombus schrencki 

SEDG 
(2050) 

Me gain 

a) 
a 
ia) 
wn 

n 
7) 
Ee) 


120 = Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus semenoviellus Skorikov, 1910 

= Bombus (Cullumanobombus) semenoviellus 

© Photo W. Kornmilch 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus semenoviellus is a medium-sized bumblebee. The coat colour shows three yellow bands 
and a white tail. It occurs in central and eastern Europe, from a latitude of 48° N (in the south) 
to 64° N (in the north) and from Central Germany in the west to Central Siberia in the east. 
It is a generalist forager. It has expanded recently towards the west, reaching Germany and 
the Czech Republic. Recently also found in Norway (2013). The species is not considered to 
be threatened: Least Concern in the IUCN Red List of European Bees. Its modelled distri- 
bution includes areas which the species does currently occupy, especially in Scandinavia and 
Finland, but also in south-eastern Europe, Turkey and the Iberian peninsula, regions that are 
presently out of reach for the species. By 2050, all scenarios project an expansion of its suitable 
climatic areas towards the north in Scan- 
Present distribution can be well explained by climatic dinavia and Finland (to where the species 
variables (AUC = 0.90) expanded recently) and in northern Rus- 
Climate risk category: HHHR sia. The climatically suitable area would 
ere Sere ee severely contract in central Europe. These 
tendencies would continue to 2100, lead- 
ing to an almost total extinction if the 
| | Scenario | Faull dispersal | No dispersal — species cannot disperse sufficiently. How- 
ever, the recent, considerable and rapid 
expansion of the species to the west in- 

-4571 (-55%) 

dicates good dispersal abilities. Even with 
-4893 (-58%) “5675 (-68%) high dispersal capability, the remaining 
-5467 (-65%) -7066 (-84%) suitable climatic area of B. semenoviellus 
6230 (-74%) 8060 (-96%) would become much smaller compared 
to its present range. The species would 

-6236 (-75%) “8112 (-97%) suffer considerably from global warming 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus semenoviellus 121 

SEDG SEDG 
(2050) (2100) 
Es gain 

—__ stable 

loss 


122 = Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus sichelii RADOSZKOWSKI, 1859 
= Bombus (Melanobombus) sichelii 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus sichelii is a medium-sized bumblebee. Its colour coat shows three more or less ex- 
tended greyish bands and a reddish tail. It is a species with a disjunct distribution: one part is 
in north-eastern European boreal taiga in Russia (not mapped); the other part in the highest 
elevations of the Pyrenees, Alps and Balkan mountains. To the east, it occurs in north-eastern 
Turkey, Iran and the Caucasus and eastwards across Siberia to the Pacific coast. It is a general- 
ist forager. The species is not considered 
Present distribution can be very well explained by to be threatened: Least Concern in the 
climatic variables (AUC = 0.96) IUCN Red List of European Bees. The 
Climate risk category: HR modelled distribution appears somewhat 

larger than its actual one, including the 
IUCN Red List status: Least Concern 

Apennines, Scandinavia and the British 
isles. All scenarios project a moderate 

| | Scenario | Full dispersal No dispersal reduction of its suitable areas by 2050. 
SEDG 530 (-19%) By 2100, the climatic niche space would 

RANGEL be even more restricted, especially in the 
BAMBU and GRAS scenario where the 
Pyrenees would become unsuitable. With 
low dispersal capability and as B. sichelii 

is a species linked to cold boreal and al- 

-1871 (-66%) 
-1693 (-60%) -2189 (-77%) 

pine-subalpine conditions, it would suffer 

significantly from global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 

ma gain 
stable 
loss 

Bombus sichelii 

123 


124 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus soroeensis (FABRICIUS, 1776) 

= Bombus (Kallobombus) soroeensis 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus soroeensis is a small- to medium-sized bumblebee. Its colour coat is highly variable. 
Most colour patterns show one, two or three more or less extended yellow bands and a white to 
red tail. This species occurs from northern Spain, southern Italy and the Balkans in the south 
to beyond the Arctic Circle in the north and from north-west Spain, (but not in Ireland), in 
the west eastwards to the Altai and Sajan mountains in Central Siberia. It is restricted to the 
highest elevations of the mountains in the south while it lives also in the lowlands of the north. 
It is a generalist forager with a preference for Campanulaceae. The modelled distribution ap- 
pears somewhat smaller than its actual one, especially in England, Belgium, The Netherlands, 
Poland and southern Italy. However, the species is becoming increasingly rare in these regions. 
Despite this regional regression, the species is not considered to be threatened at a continental 
scale: Least Concern in the IUCN Red List of European Bees. All scenarios project a moder- 
ate reduction of suitable areas by 2050, 
Present distribution can be explained by climatic accentuating the present tendencies. By 
variables to a moderate extent (AUC = 0.82) 2100, its suitable areas would exclude all 
Climate risk category: HR lowlands south of latitude 60° N. In the 
GRAS scenario, the species would only 
IUCN Red List status: Least Concern Pa ee ; 
remain in the Pyrenees and Alps, with 
the remaining areas in the Tatra and Car- 
| | Scenario | Full dispersal | No dispersal pathians becoming too small to allow the 

SEDG survival of the species. Even though B. 

soroeensis is a species which currently has 
BAMBU Lk y 

GRAS 

a large distribution range, and is likely to 
have a high dispersal ability, it is already 
-6530 (-55%) becoming scarce in some regions which 
7942 (67%) are projected to become unsuitable. Its 
climatically suitable area would decrease 

-6786 (-57%) -8639 (-73%) 

considerably under warming conditions. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus soroeensis 125 

SEDG 
(2050) 

= gain 
stable 

(2050) 

(2050) 


126 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus sporadicus NYLANDER, 

= Bombus (Bombus) sporadicus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

1848 

Bombus sporadicus is a large bumblebee. Its coat colour is constant, with three yellow bands 

and a white tail. Its wings are clearly darkened. It is a species restricted to the boreal taiga 

where it forages on the most abundant flowers, mainly Ericaceae and Epilobium angustifolium 

(Onagraceae). It occurs in Europe from the latitude of Stockholm in the south northwards 

towards the Barents Sea coast in the north and from western Norway eastwards to the Pacif- 

Present distribution can be very well explained by 
climatic variables (AUC = 0.97) 

Climate risk category: HR 

IUCN Red List status: Least Concern 

| | Scenario | Full dispersal No dispersal 

SEDG 

BAMBU 

GRAS 

SEDG 

BAMBU -5039 (-65%) -5230 (-68%) 
GRAS -6115 (-79%) -6297 (-82%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

ic coasts. The species is not considered 
to be threatened at a continental scale: 
Least Concern in the IUCN Red List of 
European Bees. The modelled distribu- 
tion corresponds very well to the actual 
one, except that the species does not oc- 
cur in southern mountains. All scenarios 
project a reduction of suitable areas by 
2050. By 2100, the climatic niche space 
of the species would be restricted to 
mountain areas, this tendency being the 
most extreme under the GRAS scenar- 
io. Since this species is linked to boreal 
conditions, B. sporadicus would suffer 
considerably from global warming, re- 
gardless of its dispersal abilities. 


Bombus sporadicus 127 

SEDG 
(2100) 


128 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus subterraneus (L., 1758) 

= Bombus (Subterraneobombus) subterraneus 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus subterraneus is a large bumblebee. The coat colour shows two very different forms. In most of Eu- 
rope it has three yellow bands and thin grey abdominal bands and a white tail (ssp. /atreillellus (Kirby)). In 
Sweden and northern Italy it is all black with or without an admixture of dark brown hairs on the thorax 
and with or without a brownish tail (ssp. subterraneus); it can even be all-black. The coat also appears to be 
conspicuously short and velvet-like. It is a generalist species although it forages especially on Fabaceae and 
other flowers with long corollas. In Europe, it can be found from central Spain, southern Italy, Greece and 
Turkey in the south, (where it is restricted to the mountains) to latitude 62° N in Fennoscandia and Russia 
in the north. In the west, it used to reach Wales and north-western Spain (not Ireland) and to the east it 
reaches Mongolia. The species is absent from all Mediterranean islands. It has become rare in most lowland 
locations of western and central Europe. It is now considered extinct in the British Isles and it has not 
been seen for a long time in Belgium, The Netherlands and in most parts of Germany. Despite this strong 
regional regression, the species is not considered to be threatened at a continental scale: Least Concern in 
the IUCN Red List of European Bees. Its modelled distribution includes precisely some of these areas from 
where the species has vanished recently. All 
scenarios project a moderate reduction of the 
climatic niche space in the south and some ex- 

Present distribution can be explained by climatic 
variables to a moderate extent (AUC = 0.82) 

Climate risk category: HHHR tension to the north, depending on its disper- 
sion capability. In 2100, in the worst case, the 
IUCN Red Last status: Least Concern GRAS scenario projects unsuitable conditions 

for all lowlands and most of the mountains in 
the European mainland. Suitable conditions 
would only remain in the Alps and Scandina- 
via and Finland. As B. subterraneus seems to 

| | Scenario _| Full dispersal No dispersal 

SEDG 
BAMBU have a low dispersal capability (being unable 

GRAS -5176 (-51%) to reach islands), and as it is already becoming 
rare or has even vanished in most lowland lo- 

SEDG -5580 (-55%) -7568 (-74%) cations, it would lose a considerable amount of 

BAMBU 6813 (67%) 9423 (-92%) climatically suitable area under global warm- 
. ing likely leading to extinction in Europe un- 

GRAS -7375 (-72%) -9829 (-96%) der the most severe scenario (GRAS). 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


SEDG 
(2050) 

= gain 
stable 
loss 

| 
I 

? 

GRAS 
(2050) 

BAMBU 
(2100) 

Bombus subterraneus 

129 


130 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus sylvarum (L., 1758) 

= Bombus (Thoracobombus) sylvarum 

Fey 

J ie aee 

© Photo J. Carteron 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus sylvarum is a small- to medium-sized bumblebee. The coat colour shows two different forms: with 
three very large greyish bands and with thin grey bands and a reddish tail intermixed with grey (f. sylvarum); 
or alternatively all black with or without grey intermixed hairs on the thorax and with a reddish tail (f. 
nigrescens Pérez). It is a generalist species but it forages principally on Fabaceae and other flowers with long 
corollas, e.g. Lamiaceae and Boraginaceae. In Europe, it can be found from central Spain, Sicily, southern 
Italy, Greece and Turkey in the south, (where it is restricted to the mountains), to the Arctic Circle in 
Scandinavia in the north. To the west, it reaches Ireland and northern Portugal and to the east it reaches 
Mongolia. The species is absent from all Mediterranean islands except from Sicily. It expanded recently in 
Sweden, with a progression of 5° latitude northwards, now nearly reaching the Arctic Circle. At the same 
time, it has become rare in most lowland locations of western and central Europe. Despite this regional 
regression, the species is not considered to be threatened at a continental scale: Least Concern in the IUCN 
Red List of European Bees. Its modelled dis- 
Present distribution can be explained by climatic tribution fits moderately to its actual one. All 
variables to a moderate extent (AUC = 0.77) scenarios project a moderate reduction of suit- 
able areas in the south and some extension to 
CRU AG ee see egere pialalals the north, depending on its dispersal ability. 
IUCN Red List status: Least Concern In 2100, in the worst case, the BAMBU and 
GRAS scenario project unsuitable conditions 
for most lowlands (including the British Isles). 

|| Scenario | Full dispersal | No dispersal Suitable conditions would only remain in the 
SEDG Alps, Scandinavia and Finland. As B. sylvarum 
seems to have a quite good dispersal ability 

BAMBU (as it is present in Sicily and as it expanded 
GRAS recently toward the north), it would not be 
threatened much by global warming, even if 

SEDG -7026 (-50%) 

its already precarious situation in the lowlands 
BAMBU -8115 (-57%) -11430 (-81%) would become worse. Nevertheless, the most 
GRAS 9627 (-68%) 12672 (-90%) severe scenario projects a considerable reduc- 

tion of its climatically suitable area. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus sylvarum 131 


132 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus sylvestris (LEPELETIER, 1832) 
= Bombus (Psithyrus) sylvestris; Psithyrus sylvestris 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus sylvestris is a small- to medium-sized bumblebee. Its colour is not very variable, general- 
ly with one large yellow band and a white and black tail intermixed with red. It is a specialised 
social parasite species (cuckoo-bumblebee) mostly of B. pratorum and probably also B. jonellus. 
Its southernmost populations are in southern Spain, southern Italy and Greece, (where it is re- 
stricted to the mountains). To the north it reaches the coast of Barents Sea and it is distributed 
from Ireland in the west eastwards across Siberia to the Pacific coasts. It is also the second most 
common cuckoo-bumblebee. The species is not considered to be threatened at a continen- 
tal scale: Least Concern in the IUCN Red List of European Bees. The modelled distribution 

corresponds quite well to the actual one. 

aa ae All scenarios project fragmentation in the 
Present distribution can be explained by climatic EFo) 8 

acs aemodente ener le Ore lowlands of central and southern Europe 

and expansion of its climatic niche space 
Climate risk category: HR in Fennoscandia up to the highest ele- 
IUCN Red List status: Least Concern vations of Scandinavian mountains. The 

GRAS scenario projects that suitable ar- 

eas could completely disappear from all 

| | Scenario _| Full dispersal No dispersal lowlands south of latitude 60° N by 2100 
SEDG -2987 (-18%) and from southern mountain areas (ex- 

FANE “1806 (-11%) cept for the Alps). As it is a cuckoo-bum- 
blebee specialised on only one (possibly 
two) host-species and despite that it is 

-9527 (-57%) presently one of the most abundant cuck- 
8279 (-50%)  -10659 (-64%) oo-bumblebees, B. sylvestris might suffer 

GRAS -2495 (-15%) 

substantially from global warming, re- 

-10840 (-65%) — -12613 (-76%) gardless of its dispersal capability. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus sylvestris 133 


134 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus terrestris (L., 1758) 

= Bombus (Bombus) terrestris 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus terrestris is a medium- to large-sized bumblebee. Its most widespread coat colour pattern shows 
two yellow bands and a white tail, or a faded orange tail in the British isles (ssp. audax (Kirby)). It is all 
black with a white tail in the Canary Islands (ssp. canariensis Pérez). There are still other conspicuous 
colour patterns in the Iberian Peninsula (ssp. /usitanicus Kriiger) and Sardinia (ssp. sassaricus Tournier). 
It lives in very large underground colonies in all habitats. It is also a generalist forager. B. terrestris is the 
only European bumblebee able to change its phenology completely according to the seasonal conditions, 
being able to produce one to three generations per year in any of the four seasons, with or without 
hibernation or aestivation. Its distribution extends from the Azores, Madeira, Canary Islands, southern 
Spain, Morocco, Sicily, north Libya, Crete, Cyprus, Israel and central Iran in the south northwards to 
north Sweden. It has recently made a dramatic advance towards the north, crossing the Arctic Circle 
in Sweden and Norway. It reaches the Altai Mountains to the east. It is abundant or even dominant in 
most of its locations and it is even considered as invasive where it has been introduced (e.g. in Argen- 
tina, Chile, Japan, Tasmania). The species is 
Present distribution can be explained by climatic not considered to be threatened by IUCN. Its 
variables to a moderate extent (AUC = 0.82) modelled distribution fits well with its actual 

Climate tisk category: HR one. All scenarios project a reduction of suit- 
able areas together with an expansion towards 

TUCN Red List status: Least Concern the north. This reduction could be inconspic- 

uous by 2050 but more dramatic by 2100. In 

|| Scenario | Full dispersal | No dispersal | the worst case, the GRAS scenario projects by 
2100 a considerable reduction of the suitable 

are area in southern Europe and North Africa. As 

B. terrestris is a very ubiquitous and generalist 
species, highly polymorphic, very abundant, 
with conspicuous and dramatic dispersal abil- 
ities, it would not be threatened by global 
-10351 (-60%) warming but, nevertheless, in some scenarios, 
its climatic suitable areas would be noticeably 

-12309 (-71%) 

or even considerably reduced in the south. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus terrestris 

SEDG SEDG 
(2050) (2100) 
= gain 
stable 

loss 

BAMBU 
(2050) 

BAMBU 
(2100) 

GRAS 
(2050) 

GRAS 
(2100) 

135 


136 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus vestalis (GEOFFROY, 1785) 
= Bombus (Psithyrus) vestalis; Psithyrus vestalis 

© Photo J. Michailowski 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus vestalis is a large bumblebee. The specimens from the European mainland generally 
have a colour pattern with one large yellow prothoracic band and a white tail intermixed with 
some yellow hairs (ssp. vestalis). There are some conspicuous subspecies in southern areas, as 
in Sardinia, where there is no yellow band on the thorax (ssp. sorgonis (Strand)) or in Corsica, 
where specimens have generally an all-black coat with a red tail (ssp. perezi (Schulthess-Rech- 
berg), which is often considered to be a good endemic species). It is a social parasite species 
(cuckoo-bumblebee) specialised primarily on B. terrestris. The species occurs from north Mo- 
rocco northwards to southern Sweden (where it recently expanded its distribution) and Latvia 
in the north and from Ireland and Portugal eastwards to the Urals. As a cuckoo-bumblebee it is 
clearly much scarcer than its host Bombus terrestris but it can be locally abundant. The species 
is not considered to be threatened: Least Concern in the [UCN Red List of European Bees. 
The modelled distribution shows that its 

Present distribution can be explained by climatic climatic niche space includes a wider area 

variables to a moderate extent (AUC = 0.79) along all margins of the species’ actual 

Climate risk category: HR range. All scenarios project fragmenta- 

tion in southern Europe and expansion 
IUCN Red List status: Least Concern 

| | Scenario _| Full dispersal No dispersal 

of suitable conditions northwards to the 
Arctic Circle and the Barents Sea coast by 
2100. The GRAS scenario projects that its 

SEDG -1785 (-17%) 
BAMBU -2114 (-20%) 
GRAS -2386 (-23%) 

-4744 (-45%) 
-7681 (-73%) 
-8897 (-84%) 

-5663 (-53%) 
-7026 (-66%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

suitable areas could completely disappear 
from all lowlands south of latitude 52° N 
by 2100. As it is a social parasite special- 
ised on a single host-species and despite 
that it is a common cuckoo-bumblebee, 
it would suffer noticeably from global 
warming, depending on its dispersal ca- 

pability (which we assume to be high). 


Bombus vestalis 137 

SEDG SEDG 
(2050) (2100) 
© gain 

stable 

loss 

od 

& 

BAMBU 
(2100) 


138 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus veteranus (Fasricius, 1793) 

= Bombus (Thoracobombus) veteranus, =Bombus arenicola. 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus veteranus is a medium-sized bumblebee. The coat colour is very constant: all grey with a black 
thoracic band. The species occurs from the Massif Central (44° N) northwards to beyond the Arctic 
Circle in Finland and northern Russia. To the west, it reaches Britanny (western France) and to the 
east, it is found across Siberia to the Pacific coast. It does not occur in any of the large European islands 
or in the major peninsulas (Iberian, Italian, Balkan). It prefers to forage mainly on flowers with long 
corollas, especially Fabaceae or Lamiaceae (queens and workers) or Asteraceae (males). It is known to 
be a facultative social parasite of B. sylvarum and B. humilis. It has recently expanded its range towards 
the north and it recolonised Sweden via Finland. It has clearly become rare in most of the western and 
central European parts of its range. Surprisingly, it can also be very abundant during some years in 
some locations (e.g. in The Netherlands). Despite its increasing scarcity the species is not considered 
to be threatened at a continental scale: Least Concern in the IUCN Red List. The distribution model 
shows that its climatic niche space includes large European islands and peninsulas. Most of the places 

where it occurred a century ago, are now out 

Present distribution can be explained by climatic of its modelled distribution, i.e. areas of suit- 
variables to a moderate extent (AUC = 0.80) able climatic conditions have already shifted 

; significantly. All scenarios project a reduction 
eae peateeen ss of suitable areas by 2050. By 2100, all scenar- 
IUCN Red List status: Least Concern ios project that suitable climatic conditions 

would persist only in northern Europe and in 
the mountains of central and eastern Europe. 

| | Scenario _| Full dispersal No dispersal The BAMBU and GRAS scenarios indicate 
SEDG 2014 (-15%) that suitable areas of the species would only 
remain in a very restricted area of the south- 

BAMBU -1978 (-15%) ern mountains and north of latitude 60° N 
GRAS -2504 (-19%) in Fennoscandia, reaching the Scandinavian 

; mountains. As this species is already scattered 
ere e a) in most of its lowland locations in west and 
-10474 (-80%) central Europe, indicating a low dispersal ca- 

7006 (53%) -11487 (87%) pability, B. veteranus would suffer consider- 

ably from global warming. 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


Bombus veteranus 139 

SEDG 
(2050) 
f= gain 
—__ stable 
Bless 

SEDG 
(2100) 

(2050) 

(2100) 


140 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus wurflenii Raposzkowsk1, 1859 
= Bombus (Alpigenobombus) wurflenii; Bombus wurfleini 

© Photo P. Rasmont 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus wurflenii is a medium-sized bumblebee. Its coat colour is mainly black with a reddish 
tail. More or less grey hairs can be intermixed with the black ones on the thorax and abdomen, 
and the amount of grey is characteristic for different isolated subspecies. It is a species with a 
disjunct distribution: (i) Scandinavian mountains, Ural, Cantabrian Mountains, Pyrenees, Alps, 
Apennines; and (ii) Carpathians, Balkans, Little and Great Caucasus. ‘This species is highly adapt- 
ed for “nectar robbing”, i.e. perforating holes at the base of long corollas to reach the nectar, and 
to do this it uses its modified mandibles. The species is not considered to be threatened: Least 
Concern in the IUCN Red List of Eu- 
ropean Bees. The modelled distribution 

Present distribution can be well explained by climatic 
vatiables (AUC = 0.89) 

appears somewhat larger than its actual 
Climate risk category: R one, including parts of central Spain, the 

TUCN Red List status: Least Concern British Isles and eastern Baltic countries. 

Full dispersal No dispersal 

SEDG 

All scenarios project a moderate reduction 
of suitable areas by 2050. By 2100, the 
climatic niche space of the species would 

be even more restricted, especially in the 

BAMBU 

| GRAS 
| SEDG 
BAMBU 
GRAS 

-3731 (-52%) 
-4123 (-57%) 

-3809 (-53%) 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 

GRAS scenario where the Balkan would 
become unsuitable. With low dispersal 
capability and since B. wurflenii is a spe- 
cialised species linked to cold boreal and 
alpine-subalpine conditions, it would suf- 

fer substantially from global warming. 


Bombus wurflenii 141 

SEDG 
(2100) 

BAMBU 
(2100) 


142 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus zonatus SMITH, 1854 

= Bombus (Thoracobombus) zonatus 

} * >. “a , 
© Photo G. Holmstrom 

Dots: actual distribution 1970-2000; yellow areas: 
modelled suitable climatic conditions in 2000 

Bombus zonatus is a medium-sized bumblebee. Its coat colour presents one or two broad yellow 
bands on the thorax and a largely yellow abdomen with a black tail. The species occurs in the 
Balkans, Romania, Moldova, Ukraine, southern Russia, Turkey, Caucasian countries and Iran. 
It is a generalist forager most often found on thistles (Asteraceae). The species is considered to 
be threatened at the European scale: Endangered in the IUCN Red List of European Bees. The 
modelled distribution appears much larger than its actual one, including Spain, Italy, and cen- 

tral Europe. This restricted actual range compared to its potential area, and the recent fragmen- 

tation along the Black Sea coast might 

Present distribution can be well explained by climatic indicate low dispersal ability and/or ad- 
vatiables (AUC = 0.93) 

ditional hostile conditions other than the 
Climate risk category: HR climatic ones. Depending on its dispersal 
IUCN Red List status: Endangered capability (that remains unsettled because 
we have too few European data at hand 
to assess it), the scenarios display consid- 
erably different results. If B. zonatus has 
a high dispersal ability, it could consider- 
ably expand its distribution. If the species 
-965 (-17%) would not be able to use this opportunity, 
it could suffer considerably from reduc- 
tions in the area of climatically suitable 

-4520 (-80%) ; 
space, especially under the most severe 

-4756 (-84%) scenario (GRAS). 

Changes in climatic niche distribution (in 10’ x 10’ grid cells) 


143 

Bombus zonatus 


144 

Climatic Risk and Distribution Atlas of European Bumblebees 

9. Non-modelled European bumblebee species 

Given the methodological restrictions of statistical species distribution modelling, we were 
not able to model 13 species. These were (i) rare species with a very narrow distributional 
range (too few data points) or (ii) species for which additional environmental or 
anthropogenic factors seem to override the climatic limitations or (iii) species with some 
taxonomic uncertainties or with recently modified status. Since most of the species 
included in the two first categories are either rare, endemic or habitat specialists with 
assumed low ability to disperse and thus low ability to follow the changing climates, 
many of them are very likely to be highly vulnerable to climate change. However, warm- 
adapted species might also profit from climate change and an assessment of the actual 
climatic risk of such data-insufficient species remains speculative. For such species better 
data would be needed with a much higher spatial resolution for rare and endemic species 
or from the entire range for species that occur only marginally in Europe. 

9.1. Rare species and/or species with narrow distributional range 

Some European bumblebee species are extremely rare and occur only in small geographic 
areas with few records. Irying to model such species is often not possible because of too 
little information content provided by the very few data points at the coarse resolution 
we used for modelling and consequent overfitting of the SDMs. Further, some species 
have the centre of their distribution in Central Asia and occur only occasionally or on 
the margins of the considered geographic window. For such species we do not cover 
either the whole distributional range or the likely relevant range boundaries (most 
importantly the north and south) necessary for the development of realistic niche 

models. For these reasons the following species were not modelled. 

Bombus reinigiellus is the rarest bumblebee species in Europe. It also shows the most restricted distribution range, living only 
at the highest level of the Sierra Nevada, in south-east Spain. Pictures of this species are exceptional. Photo A.G. Maldonado 


Non-modelled European bumblebee species 

Bombus brodmanni Skorikov, 1911 

=Bombus (Thoracobombus) brodmanni, =Bombus (Rhodobombus) brodmanni 

Distribution of 

B. brodmanni in the 
chosen geographic 
window (red dot in 
Turkey). 

_ 

© Photo P. Rasmont 

This species is extremely rare and endemic to the mountains of north Turkey and west Cauca- 
sus. It occurs only marginally in the south-east of the considered area (Fig. 9.1). The species is 

poorly known and was not assessed by the IUCN Red List of European Bees. 

Bombus brodmannicus Vogt, 1909 

=Bombus (Pyrobombus) brodmannicus 

Distribution of 

B. brodmannicus 

in the chosen 
geographic window 
(red dots in France). 

< 

© Photo P. Rasmont 

This species lives in two well separated areas. In the Caucasian region the species can be locally 
very abundant. In the south west Alps (France) it is restricted to very few locations (Fig. 9.2). 
The Caucasian population forages on many different plant species while the population from 
the Alps seems to be specialised on Cerinthe flowers (Boraginaceae). The restricted distribution 
and food specialisation led to an assessment as Endangered in the IUCN Red List of European 
Bees. Because of this highly specialised foraging requirements and its already localised distribu- 
tion in a small area of the Alps, the western population seems extremely vulnerable to warming. 
On the other hand, the eastern population is rather widespread in the Caucasian region with no 
apparent food specialisation. It is likely much less vulnerable to climate change. 

145 


146 Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus mlokosievitzii Radoszkowski, 1877 

=Bombus (Thoracobombus) mlokosievitzii 

Distribution of 

B. mlokosievitzii 

in the chosen 
geographic window 
(red dots in Turkey 
and the Balkans). 

5 til 

© Photo P. Rasmont 

This species mainly occurs in forests of northern Turkey, the Caucasus and northern Iran. In 
Europe, it occurs only in some scattered locations in the Balkans, where it is locally quite rare. 
The present rarity of the species in the Balkans might make it vulnerable to climate change. It 
has not been assessed (Data Deficient) in IUCN Red List of European Bees. 

Bombus patagiatus Nylander, 1848 

=Bombus (Bombus) patagiatus 

Distribution of 

B. patagiatus in the 
chosen geographic 
window (red dot in 
«| Finland). 

© Photo P. Rasmont 

This species is abundant in the boreal taiga from the Ural to the Pacific coast of Russia. It occurs 
also west of the Ural, reaching the Finnish border near Lake Ladoga. Only very few individuals 
have been found in the area studied here. It has not been assessed (Data Deficient) in IUCN 
Red List of European Bees. 


Non-modelled European bumblebee species 

Bombus renigiellus (Rasmont, 1983) 
=Bombus (Megabombus) renigiellus 

Distribution of 

B. renigiellus in the 
chosen geographic 
window (red dots in 
Spain). 

© Photo P. Rasmont 

This species is endemic to high altitudes of the Sierra Nevada (southern Spain), where it is rare 
and restricted to a few locations above 1800 m asl. Bombus renigiellus has the smallest distribu- 
tion of any Bombus species in Europe. As the species is already restricted to the cooler areas in 
the highest regions of the Sierra Nevada, it seems very likely that any further warming could 
drive the species to extinction. It is assessed as Endangered by IUCN Red List of European Bees. 

9.2. Species with a distribution poorly explained by climatic variables 

The impact of climatic variables on the distributions of some species can sometimes 
be overridden by other environmental factors. This is the case when a species is, for 
instance, highly bound to a particular habitat type that occurs only occasionally and 
most often only with small patches compared to the size of our grid cells and is highly 
scattered across Europe such as steppic grasslands or moors. For these species we were 
not able to develop reliable SDMs indicated by poor model performance. 

The Scandinavian mountain tundra is inhabited by several bumblebee species threatened by global warming (N. Sweden, 
Abisko, 2013. Photo P. Rasmoni). 

147 


148 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus armeniacus Radoszkowski, 1877 
= Bombus (Thoracobombus) armeniacus; Bombus (Rhodobombus) armeniacus 

Distribution of 

B. armeniacus in the 
chosen geographic 
window (red dots). 

© Photo P. Rasmont 

Bombus armeniacus lives in steppic areas in Hungary, Romania, the Balkan peninsula, Moldo- 
va, Ukraine, southern Russia and Turkey. Its range extends into Central Asia. It is a generalist 
forager with a marked preference for flowers with long corollas that can be accessed by its long 
tongue. This species is clearly declining and it was classified as Endangered in IUCN Red List 

of European Bees. 

Bombus deuteronymus Schulz, 1906 

Bombus (Thoracobombus) deuteronymus; Bombus bureschi; Bombus 
superequester 

Distribution of 

B. deuteronymus 
in the chosen 
geographic window 
(red dots). 

© Photo P. Rasmont 

Bombus deuteronymus occurs in dry woods and forest edges in the Balkans and in Russia where 
it is a very rare species. To the east, its distribution reaches the Pacific coasts and Japan. Its floral 
resources are unknown. Because of its scarcity, the IUCN Red List of European Bees does not 
provide an assessment (Data Deficient). 


Non-modelled European bumblebee species 

Bombus laesus Morawitz, 1875 
= Bombus (Thoracobombus) laesus; Bombus (Laesobombus) laesus 

Distribution of 

B. laesus in the 
chosen geographic 
window (red dots). 

© Photo P. Rasmont 

Bombus laesus is a steppic species of Hungary, Romania, Moldova, Ukraine, southern Russia and 
Turkey. To the east, its distribution includes the steppes of central Asia. Its floral resources princi- 
pally include flowers with long corollas which it can access easily with its medium-sized tongue. 
This species was quite abundant in true steppes of Europe, especially in Hungary and Ukraine. 
However, it has recently declined considerably and it is now only present in a small fraction of its 
former distribution area. It is assessed as Near Threatened in the IUCN Red List of European Bees. 

Bombus mocsaryi Kriechbaumer, 1877 

= Bombus (Thoracobombus) mocsaryi; Bombus (Laesobombus) mocsaryi; Bombus 
laesus mocsaryi; Bombus maculidorsis 

Distribution of 

B. mocsaryi in the 

chosen geographic 
window (red dots). 

© Photo P. Rasmont 

Bombus mocsaryi lives in wooded-steppes and dry grasslands on the Iberian peninsula, south of 
France, Hungary, Serbia, Romania, Poland, Belarus, Moldova, Ukraine, central Russia. To the 
east, its distribution includes the extensive grasslands in central Asia. This species forages mainly 
on flowers with long corolla which it can access easily with its medium-sized tongue. Bombus 
mocsaryi was formerly quite abundant in many dry grasslands. However, it recently declined 
considerably, remaining only in few locations in Spain, Hungary, Balkans, and Ukraine. It is 

assessed as Endangered by the IUCN Red List of European Bees. 

149 


150 Climatic Risk and Distribution Atlas of European Bumblebees 

9.3. Taxonomically problematic species 

For the following species from Corsica and adjacent islands, there are some taxonomic 
issues that could potentially affect the modelling. ‘These issues are discussed in chapter 12. 

Bombus perezi Schulthess-Rechberg, 1886 

=Bombus (Psithyrus) perezi; Psithyrus vestalis perezi 

Distribution of 

B. pereziin the 
chosen geographic 
window (red dots). 

© Photo M. McGlinchey 

This species is endemic to Corsica and few other islands of the Tuscan archipelago. It is abun- 
dant within its restricted range where it is a cuckoo-bumblebee of B. xanthopus and, possibly, 
also of B. renardi. Its distribution is too restricted to allow the development of reliable species 
distribution models. It is assessed as Least Concern in the IUCN Red List of European Bees. 

Bombus pereziellus (Skorikov, 1922) 

=Bombus (Thoracobombus) pereziellus; Bombus muscorum pereziellus 

Distribution of 

B. pereziellus in the 
chosen geographic 
window (red dots). 

© Photo P. Rasmont 

This species is endemic to Corsica, where it is quite rare. Its distribution is too restricted to 
allow the development of reliable species distribution models. It is assessed as Least Concern by 

IUCN Red List of European Bees. 


Non-modelled European bumblebee species 

Bombus renardi Radoszkowski, 1881 

=Bombus (Bombus) renardi; Bombus lucorum renardi 

Distribution of 

B. renardi in the 
chosen geographic 
window (red dots). 

© Photo P. Rasmont 

This species is strictly endemic to Corsica. Its distribution is too restricted to allow the develop- 
ment of reliable species distribution models. It has not been assessed by the IUCN Red List of 
European Bees as it has been only recently restored to species status. 

Bombus xanthopus Kriechbaumer, 1877 

= Bombus (Bombus) xanthopus; Bombus terrestris xanthopus 

Distribution of 

B. xanthopus in the 
chosen geographic 
window (red dots). 

© Photo P. Rasmont 

This species is endemic to Corsica and few other islands of Tuscan archipelago where it is very 
abundant. Its distribution is too restricted to allow the development of reliable species distri- 
bution models. It has not been assessed by the IUCN Red List of European Bees as it has been 

only recently been restored to species status. 

151 


152 

Climatic Risk and Distribution Atlas of European Bumblebees 

10. General patterns of future risk 
10.1. General overview 

A common trend across all scenarios is that for the majority of species their climatically 
suitable areas are projected to shrink moderately to strongly, while the suitable areas 
for only few species are projected to expand (Table 10.1). Out of the 56 modelled 
species five species may expand their ranges by 2050 and four to six species, depending 
on the scenario, may expand by 2100 under the assumption of full dispersal. Four to 
17 species, depending on the scenario and the dispersal assumptions, are projected to 
maintain more or less their status quo up to 2050 and zero to one up to 2100. Suitable 
climatic conditions are projected to decrease for 34 to 52 species up to 2050 and for 49 
to 55 species up to 2100. The 13 non-modelled species are all very rare and localised 
and their area can be expected to shrink considerably in any situations and potential 
extinction of many of these species seems most likely in any cases of climate warming. 

Table 10.1 Projected changes in climatically suitable areas for European bumblebee 
species by the years 2050 and 2100. The values represent the number of species in each 
change category. Thirteen species were too rare to be modelled. 

Full dispersal No dispersal 
2050 2100 2050 2100 

3 Percentage 
Change categories fenee 

Non-modelled 

Strong expansion : >+80% 

Status quo -20 to +20% 

Strong regression -50 to -80% 

Very strong regression to 
extinction 

‘ -80 to 100% 

69 69 69 69 69 69 69/69 69 69 69 69 69 69 


General patterns of future risk 

10.2. No dispersal vs full dispersal 

Under the assumption of no dispersal, which is actually the assumption of negligible 
dispersal ability within the next 40 or 90 years respectively, no range expansion is pos- 
sible per definition. Thus, large differences between both assumptions are visible for 
the assessment of potential future reductions of suitable areas. The number of species 
which might retain their status quo in terms of range size, but not necessarily according 
to their current distribution, ranges between zero and 17, when full dispersal ability is 
assumed. But only between one and seven species will not change under the assump- 
tion of no dispersal, which means they could more or less sustain their current ranges. 
The number of species potentially losing suitable areas is higher with no dispersal (be- 
tween 48 and 55), compared to full dispersal (between 34 and 52). Dispersal plays a 
particular role with respect to potential extinction at the European level. While only 
between two and eight species are at a particularly high risk of extinction when full 
dispersal is assumed, between five and 34 species are at risk of extinction under the 
assumption of no dispersal by the year 2100. 

Based on the assumed dispersal ability for each species (as described in chapters 6-7), we 
used the likely more realistic assumption for each species (full or no dispersal if assumed 
dispersal ability is high or low) and assessed the overall effects of projected climate change 
on the potential future range changes for the European bumblebees (Table 10.2). 

Considering the assumed dispersal abilities, three species can potentially expand their 
ranges by 2050 and also by 2100. Five to eight species could keep their status quo up 
to 2050 and none by 2100. Forty-five to 48 species are projected to lose suitable areas 
by 2050 and 53 by 2100 (see also Fig. 13.2, p. 174). 

= 
| 

Bombus glacialis is large species that could 
be found only in two distant locations in Arctic 
ocean: Novaya Zemlya and Wrangel islands. 
Its coat colour strongly recalls the near 
species Bombus lapponicus, greyish on the 
back of the pro- and metathorax and tergite 
1, nearly all the remaining abdomen reddish. 
It differs by the admixture of numerous 
black hairs in the middle of tergite 4 and 5. 
This species only lives in cold arctic tundra 
with permafrost where it nests in lemming’s 
abandoned burrows. Photo P. Rasmont. 

153 


154 

Climatic Risk and Distribution Atlas of European Bumblebees 

Table 10.2 Projected changes in climatically suitable areas for European bumblebee 
species in 2050 and 2100 considering rough assessments of species-specific dispersal 
abilities. The values represent the number of species in each change category. Dispersal 
ability has been assessed for each species according to ecological characteristics (see 
chapters 6-7). Dispersal abilities for two species (B. incertus and B. zonatus) were un- 
known. We therefore assume these two species to have full dispersal abilities (providing 
optimistic results for potential range shifts). 

2050 2100 

Change categories Percentage change 
=) 
oY v & 
: 2|8/2| 4 
an pa ie) ma ~) 

a A 

Strong expansion >+80% 

aon |e | | s | o | 0 [0 

_ strong regression to extinc- -80 to 100% 14 

10. 3. Climate change scenarios 

As expected, the three scenarios do not show strong divergence for 2050. However, 
under the most severe change scenario (GRAS) eight species are projected to suffer 
from strong or very strong reductions of suitable areas compared to four species in 
the intermediate change scenario (BAMBU) or three species for the moderate change 

scenario (SEDG; Table 10.2). 

For 2100, the three models diverge considerably, even if they all project reduction of 
climatically suitable area for the majority of the species. Especially under the GRAS 
scenario a large fraction of European bumblebee species (25 species) are pre projected 
to lose nearly their entire suitable area, leaving them to the verge of extinction. In com- 
parison, the BAMBU and SEDG scenarios project such drastic reductions for only 14 
and 3 species respectively. 


General patterns of future risk 

10. 4. Effects on subregional scales 

To assess the effects of climate change on the bumblebees and their potential variability 
at subregional scales, we also examined the number of species that could find suitable 
climatic conditions within the surrounding (50 km radius) of 30 selected European 
cities (Table 10.3). The actual observed number of species and the number of species 
that could find suitable climatic conditions increases with latitude (Pearson R=0.43 
and 0.31). In most of the cases more species are predicted to find climatically suitable 
conditions in the selected areas than are actually present. For instance, the median 
number of actually present species across the selected areas is 21.5. In contrast, the 
median of the modelled number of species that could find suitable climatic conditions 
is 26 (Table 10.3). This indicates the presence of other limiting factors apart from cli- 
matic conditions. One other major factor might be dispersal limitation for a number 
of species, either as a consequence of species-specific movement abilities, or caused by 
the presence of hard dispersal barriers. Other factors such as limited amounts of other 
essential resources such as the availability of pollen and nesting or hibernation sites 
might play a role, too (see chapter 11 for a more detailed discussion about the principal 
limitations of species distribution modelling). 

For 2050, the median modelled number of species finding suitable climatic conditions 
varies from 17.5 to 19.5, depending on the scenario, meaning a remaining median 
diversity of 64 to 76% compared to the modelled situation in 2000. 

For 2100, the median modelled number of species finding suitable climatic conditions 
varies only between 1.5 and 10. At this time, irrespective of the scenario, the diversity 
reduction will be considerable with a remaining median diversity ranging from 10 to 

46% (Table 10.3). 

Most worryingly, the projected loss of suitable climatic conditions is not evenly dis- 
tributed across Europe. Clearly, southern Europe, which already harbours a poor 
bumblebee fauna, will be most strongly affected by potential further species losses 
(Fig. 10.1). For instance, in areas around Granada, Seville, Athens and Lisbon the 
present number of bumblebee species is already very low. For these regions, severe 
regressions are projected for almost all scenarios. For Lisbon and Seville, most of the 
scenarios project that only one bumblebee species may remain by 2050. In the worst 
case, it could lead to complete extinction of the native bumblebee fauna, as it is pro- 
jected in 2100 by the GRAS scenario. The remaining species will be Bombus terrestris 
or alternatively Bombus argillaceus, as it is likely to be the case already in the warmer 
regions of southern Greece and Turkey. 

155 


156 

Climatic Risk and Distribution Atlas of European Bumblebees 

(o) (jo) 
(o>) oO 
Tr rT 
<a 0.2 
oO oo 
>. ae 
Zo Zo 
Ow Ow 
oO oO 
N N 

0 
0 

40 45 50 55 60 65 40 45 50 55 60 65 
Latitude Latitude 

Figure 10.1 Projected changes of bumblebee diversity in selected areas across a latitudinal gradient. Projected changes were 
calculated relative to the number of species that would find suitable climatic conditions in the respective areas (100%). Selected 
areas represent the surroundings of larger cities in Europe (50 km radius). Black dots and regression lines represent changes 
in 2050. Red dots and regression lines represent changes in 2100. (a) Projected changes in the best case according to scenario 
and dispersal assumptions (see Table 10.3). Linear regression model for 2050 (black): P = 0.004, R? = 0.23 and for 2100 (red): 
P =0.06, R? = 0.09). (b) Projected changes in the worst case according to scenario and dispersal assumptions (see Table 10.3). 
Linear regression model for 2050 (black): P = 0.02, R? = 0.17 and for 2100 (red): P = 0.03, R? = 0.13). 

Table 10.3 Number of bumblebee species that could find suitable climatic conditions 
in several representative areas across Europe. Actual sp. Nb indicates the number of 
species that have been actually observed in the region between 1970 and 2000. Project- 
ed nb of sp provides the modelled numbers of species that could find suitable climates 
in the year 2000. The two first 2050 and 2100 categories display the modelled number 
of species that could find suitable climatic conditions in 2050 or 2100 under 3 climate 
change scenarios for the assumptions of either full or no dispersal abilities. The two 
last 2050 and 2100 categories provide the percentage of remaining species numbers in 
2050 or 2100 for the best or worst cases, according to scenarios and dispersal abilities, 
relative to the model predictions for 2000. Green background indicates an increased 
number of species (remaining number of species > 120%); white background indicates 
an approximate status quo (remaining number of species < 120% and > 80%); yellow 
background indicates a moderate decrease (remaining number of species < 80% and > 
49.9%); indicates a strong to very strong diversity decrease (remaining number of spe- 
cies < 50% and >19.9%); shows a considerable loss in diversity (remaining number of 
species < 20%).In the regions around Madrid, Rome, Bucharest, Belgrade, Bordeaux, 
Budapest and Kiev, the present number of species could be somewhat higher than it 
actually is according to climatic suitability, but there again the diversity loss is project- 
ed to be severe, depending on the scenario. Also in these regions, only one species is 
projected to remain in most case or even none as it is projected in Bordeaux. 


157 

General patterns of future risk 

ST SiG 

cal 
al 

0001 
128 I I 
£8 €801 I S 
£°€6 L9CT Ic (a4 
O00I 6 POI Iv Iv 

O°00T 

ION =: SET IOM = So SVYD NANVE 

SUTUTEUIOI % SUTUILUIAI % [estodstq ON 

O0O0I7 0S07 

91 
€1 

9c 
Iv 

Odds 

Sv 

al 

87 

SVuD 

wn Oo MW WN & 

NINVaA 

jessodsiq [JNJ 

OL 

rR N OM 

Odds 

Iv 

svadD 

87 
Iv 

NIWVA 

yestodstq ON 

8¢ 
Iv 

Odds 

61 

9¢ 

SsvaoD  NaAWVd 

Tessodsiq [NJ 

S'6L 

LI 

WwW 

Odds 

97 

OP 

ds jo qu 
pooloig 

satoeds jo 
‘qu [enjoy 

09°€- 
66°S- 
ELT 
€1'6- 
89°€- 
871 
(awe 
CEET 
OL'9c 
€8°0- 
0S°07 
AID 
80°61 
8ST 
LE9Or 
eC 
LVVL 
cS'0€ 
€CV 
€1'0- 
001? 
887 
Gaal 
ST9- 
00°01 
OL ve 
S0°8T 
€6VT 
€e'S 
vVLI 

opnyisu0T 

SIZE 
8ELE 
86'LE 
CL'8E 
Or'0P 
0617 
CS°CP 
89°CP 
CV Vy 
8° VV 
€8 VV 
0c OP 
0S'LP 
SIT8P 
OC 87 
L880 
80°0S 
€V'0S 
€8°0S 
OS'TS 
STCS 
LETS 
0S°S 
€ees 
8S°€S 
S69 
€£'6S 
81°09 
€€09 
bV'89 

opnyneyT 

UeIpoy, 
epeursly 
aTTFA2S 
suouIy 
wogs!T 
PHpeWw 
sWIOY 
sIno']-JUO| 
eYOs 
ysoreyong 
xneoplog 
apessjag 
RAIUay 
jsadepng 
qoranyy 
UaT MA 
sie 
eyelid 
AOTY 
sjassnig 
uopuoy 
MPSIEAA 
wrepso}sury 

ulpog 

ayqnd 
Sinqurepy] 
esry 
TWOYYPOIS 
PUSH 
uasiog 

yIAreN 


158 

Climatic Risk and Distribution Atlas of European Bumblebees 

In most of the regions around Sofia in the south to around Helsinki in the north 
quite a large number of bumblebee species can be observed currently. However, a con- 
siderable loss of species is projected here under all scenarios. In most of the cases, a 
reduction of bumblebee richness by 40% is projected by 2050. For 2100, already the 
moderate change scenarios project a reduction by 50% while under the most severe 
change scenario only one or two remaining species are projected in the areas around 
Paris, Brussels, Stockholm and Helsinki. Some places like Vienna, London, Amster- 
dam, Warsaw, Hamburg and Riga may suffer from similarly drastic reductions with 
only 3 to 4 remaining species by 2100 or even none in the worst case in Riga. Areas 
around Prague, Dublin and Berlin would also suffer from considerably high diversity 
reduction with 5 to 9 remaining species at worst. 

In the northernmost localities around Narvik and Bergen, the number of species is 
already very high and in most of the scenarios this diversity could be maintained or 
even increased by 2050 and 2100. Of course some of the most sensitive species are at 
risk in these regions but there could be a gain thanks to colonisation from the south. 

The species richnesss of bumblebees in mountainous areas (Mont-Louis, Sofia, Gene- 
va, Munich and Vienna) is currently very high (34 to 36 species). It is noticeable that 
these high species numbers are very well reflected by the modelled potential number of 
species according to climatic suitability for 2000 (38 to 43 species). Iserbyt et al. (2010) 
showed that such a high diversity can be found in very restricted mountain areas (e.g. 
33 species in a small valley in the Pyrenees). Pradervand et a/. (2011) also found such 
a high diversity in the Valais (26 species). The projected future of these areas could be 
quite different: a strong reduction of the diversity in areas around Sofia, Munich and 
Vienna but a higher number of surviving species in Geneva and Mont-Louis. The two 
northernmost areas considered here (around Narvik and Bergen) are characterised by 
mountains, which could explain the sustainability of their diversity in all scenarios. On 
the contrary, most of the areas that are located in the lowlands are projected to suffer the 
most, even if they are far to the north, such as the Riga, Stockholm and Helsinki areas. 

In conclusion, a reduction in bumblebee diversity could already be noticeable in most 
of the considered areas as soon as 2050, and this reduction will become more drastic 
under all scenarios by 2100. Only few areas which include mountains would be able to 
conserve a substantial part of their present diversity. 


Methodological limitations 

11. Methodological limitations 

Species distribution models (SDMs; Guisan & Zimmermann, 2000; Guisan & Thuill- 
er, 2005) are increasingly used in studies of biogeography, conservation biology, ecol- 
ogy and palaeoecology. One way to develop such SDMs, as utilised in this atlas, is 
to assess statistically the relationship of species occurrences and absences to environ- 
mental conditions. Although such models are purely correlative, compared to more 
process-driven dynamic population models (Morin & Thuiller, 2009), they can be a 
powerful tool, especially when a large number of species is modelled for which detailed 
mechanistic understanding of the actual processes that determine occurrence are lack- 
ing. However, several limitations of the statistical SDM approach remain (Araujo & 
Guisan, 2006); and knowledge of these limitations is crucial for the level of certainty 
and interpretation of the results. These limitations can be basically grouped in relation 
to (i) the conceptual approach, (ii) data issues, (iii) modelling techniques and (iv) the 
interpretations of the results. 

11. 1. Conceptual background of SDMs 

SDMs rely on the concept of the ecological niche (Hutchinson, 1957) but there is still 
some discussion about which aspect of the ecological niche, fundamental or realised, 
is assessed by SDMs. Since observed species distributions, which are used for the de- 
velopment of the SDMs, can be constrained by non-climatic factors, many scientists 
consider the outcomes of SDMs as an approximation of the realised niche (Guisan & 
Zimmermann, 2000; Pearson & Dawson, 2003). But in a recent example Soberon & 
Peterson (2005) concluded that niche models can also provide an approximation of the 
fundamental niche. However, Aratijo & Guisan (2006) ask “... whether the distinction 
between fundamental and realized niches is useful ..” and suggest accepting that “... any 
characterization of the niche is an incomplete description of the abiotic and biotic factors 
allowing species to satisfy their minimum ecological requirements.” In terms of climatic 
niche modelling, we thus have to accept the possibility that other biotic or abiotic 
factors may limit the distribution of the focal species in addition to climate. However, 
this could only turn into a problem for the assessment of the climatic niche and the 
consequences of future change when these additional limiting factors are also related 
to climatic conditions and thus would introduce a systematic bias in the assessment of 
the climatic niche. 

In the case of the bumblebees, biotic interactions such as nest and resource availability 
could limit large-scale distributions especially of specialised bumblebees and thus bias 
the assessment of the respective climatic niches. While studies on butterflies have shown 
that limitations by biotic interactions are possible (Schweiger et al, 2008; Hanspach 
et al., 2014) it has also been shown that in most cases distributions are determined by 
climatic conditions rather than by biotic interactions (Schweiger et a/., 2012). 

159 


160 

Climatic Risk and Distribution Atlas of European Bumblebees 

Another assumption of SDMs is that species are in equilibrium with their current en- 
vironment (Pearson & Dawson, 2003). However, there is evidence that some species 
groups actually are in disequilibrium with current climate. It seems that, for instance 
trees suffer from postglacial dispersal limitation and currently do not fill their areas of 
suitable climatic condition (Svenning & Skov, 2007). Such climatic disequilibrium 
could have severe consequences for the reliability of the SDMs since in this case they 
would systematically underestimate the climatic niche space of the species. However, 
it was also suggested that the ability to fill the potential ranges, defined by suitable 
climates, depends on the dispersal abilities of the species (Aratjo & Pearson, 2005). 
For bumblebees we assume either low or high dispersal abilities (see also Kraus et al., 
2009; Lepais et al, 2010), but even bumblebees considered as low dispersers might 
have had enough time since the last glacial period to fill their ranges (Hines, 2008). Re- 
cent phylogeographic studies suggest that Bombus lapidarius expanded its distribution 
from its Ice-Age refugia to the whole of Europe (e.g. Lecocq et a/., 2013a) while other 
species have remained rather stable since the 20th century (Rasmont & Iserbyt, 2014). 
This allows us to expect a climatic equilibrium for bumblebees. However, for a small 
number of species which have recently expanded their ranges into Europe from the 
East, such as B. hypnorum, we might underestimate the climatic requirements, since it 
is most likely that they have not yet filled their potential climatic range. Thus, future 
projections for these species need to be interpreted with great caution. 

Table 11.1. Sampled countries sorted by decreasing order of sampling effort (all 
periods). Data: number of entries in the database; Nspec; number of specimens; 
D/10kmSQ: number of data by 10 km square; N/10kmSQ: number of specimens by 
10 km square. 

ae D/10kmSQ | N/10kmSQ. 

The Netherland 

Sweden 


Methodological limitations 161 

Code Country Data | Nspec | D/10kmSQ | N/10kmSQ 

(km’) 
65303 791 11383 
312679 | 25660 | 47785 
338144 7186 37676 
301336 | 11314 | 32885 
78870 3986 6850 
357026 6605 29140 
43698 15 3228 
13812 hs 74 868 
20273 718 Ti 
2586 tS 141 
110944 2867 4221 
236 Viole: Pale 
505911 8291 13944 
49035 416 1297 
ae, 716 3175 
883061 414 1630 
3100 148 48 
238391 1656 2872 
31067 202 266 
64597 498 aL 
783562 6959 20914 
93029 2D 434 
51197 71 229 
92201 129 Did 
28748 36 56 
33843 12 14 0.04 
3960000 | 1253 [525 0.03 
603549 213 214 0.04 
207600 58 8 0.03 
163610 12 4 0.01 
2381741 31 210 

BE {2 
8.2 
al 
3.8 
Sal 
1.9 

0.3 

Lithuania 
PL | Poland 

FI | Finland 

IT | Italy 

CZ_ | Czech Republic 
DE | Germany 
EE 
ME 
I 
LU 
BG 
MK 
S 

17.4 
1353 
11.1 
10.9 
8.7 

Estonia 

Montenegro 
Slovenia 3.5 
Luxembourg 
Bulgaria 2.6 
FYROM 
Spain 
Slovakia 
Greece 

Serbia 

Denmark 

1.6 

NJ 
© 

0.5 
0.5 
0.3 
0.7 

rs) 

S 
K 

iN 
ON 

Romania 
Croatia 
LV 
R 
U 
BA 

Latvia 

Turkey (whole country) 

Hungary 

Bosnia Herzegovina 
Portugal 
AL 

< 
o 

Moldova 
RU_ | Russia (Europe) 
UA | Ukraine 
BY 
T™N 
DZ 

Belarus 

Oo 

le 
SISlSo S i ae = 
— —_ — N aN —_ 

Tunisia (partim) 

= olo|[colo Stele lel rN [r[rNpwfwlut[u{aln | o 
oe, NVlwWP AT ofNMTNLofR]_alololofUuUfUufopA]d 

ene] cme | 
PUT [Lithuania | 65303 _| 
| PL [Poland | 312679 | 
PFI | Finland | 338144 | 
pir |iealy | 301336 | 
| CZ |Cuech Republic | 78870_| 
| DE [Germany | 357026 | 
PEE [Estonia | 43698 _| 
[ME [Montenegro | 13812_ 
| St_|Slovenia | 20273 | 
PLU [Luxembourg | 2586_| 
|BG | Bulgaria | 110944 
|MK [FYROM | 25713 _| 
PES [Spain | 505911 | 
| SK [Slovakia | 49085_| 
PGR | Greece | 131957 | 
PRS [Serbia 88361_ 
| DK [Denmark | 43100_ 
[RO [Romania | 238391 
PAR [Croatia | 31067_| 
PV |tawia | 64597_| 
| TR | Turkey (whole country) | 783562 | 
PHU [Hungary | 93029 | 
| BA | Bosnia Herzegovina | 51197_| 
PPT |Porugl | 92201 
PAL [Albania | 28748 
[MD | Moldova | 33843_| 
| RU _| Russia (Europe) | 3960000 | 
LUA | Ukraine | 603549 | 
| BY [Belarus | 207600 | 
LIN | Tunisia (partim) | 163610 | 
[| DZ [Algeria (partim) | 2381741 | 

ia (partim) 


162 

Climatic Risk and Distribution Atlas of European Bumblebees 

11. 2. Data issues 
11. 2.1. Data quality 

The quality of SDMs largely depends on the quality of the environmental and species 
data. While climatic data are of high quality, the quality of large-scale species distri- 
bution data usually varies considerably in space and time. While species presence data 
are usually more accurate, although they can suffer from misidentifications, species 
absence data are generally less reliable, e.g. because of low accessibility of some re- 
gions, generally poor knowledge of certain areas, inconspicuousness of the species or 
unavailability of data. Similar to the potential effects of other biotic or abiotic factors 
on the assessment of the climatic niche, variability in sampling effort can lead to biased 
SDMs, especially when sampling effort correlates with climate. 

For the bumblebees analysed in this atlas, misidentifications (false presences) can be 
considered as negligible. The species distribution data used in this atlas was extracted 
from a database (Atlas Hymenoptera) which is thoroughly cross-checked by the leading 
experts in bumblebee ecology. However, the sampling effort has differed considerably 
across Europe (Tab. 11.1) but we did not use grids with no data in the species distribu- 
tion models (Lobo et a/., 2010), thus the influence of false absences are expected to be 
rather low for most species even if some species that are difficult to identify or recently 
separated taxonomicly will have a less well known distribution and thus more false ab- 
sences present in our data such as B. magnus. Further uncertainties in the development 
of SDMs may arise from imprecise information on the sampling date. However, for 
the considered time period (1970-2000) most data in the Atlas Hymenoptera database 
include precise dates (day, month, and year). In some cases, only an interval is given. In 
these cases only the most recent data were considered as sampling data. In cases when 
no sampling date was provided but the information was extracted from a publication, 
the year before the publication occurred was taken as sampling date. Museum data 
with no sampling dates were not considered. 

11.2.2. Polytypic species 

Many European bumblebee species display a large geographic intraspecific variation (i.e. 
polytypic species) (Reinig, 1935; Pittioni, 1938; Rasmont, 1983a; Lecocq et al., 2014) 
most probably fostered by historical biogeographic events (Reinig, 1937, 1939; Lecocq 
et al., 2013a,b). These geographic variations have been used by several authors to define 
many subspecies mainly based on variation in colour patterns (Vogt, 1909, 1911; Pittioni, 
1938; Kriiger, 1951, 1958; Rasmont, 1983a; Rasmont et al., 2008). For example, B. 
pascuorum includes 23 different subspecies in Europe (Rasmont, 1983a,b). In several Eu- 
ropean bumblebee species, this polytypism is related to local eco-climatic and behavioural 
adaptations (Rasmont & Adamski, 1995; Chittka et a/., 2004; Velthuis & van Doorn, 


Methodological limitations 

2006; Rasmont et al., 2008; Lecocq et al., 2013a,b, 2014). Therefore conspecific pop- 
ulations/subspecies could have different climatic requirements, while our modelling ap- 
proach is based on the assessment of the climatic niche at the species level. This homogeni- 
sation of the climatic requirements across the species could underestimate the resistance 
to climate change of some populations, especially populations at species range margins. 
For example, a projected range contraction at the southern range margin according to spe- 
cies-level distribution models must not necessarily affect southern populations when these 
are genetically distinct from their northern conspecifics and moreover are better adapted 
to warmer conditions. The integration of this intraspecific variation in predictive models 
requires an a priori definition of genetically distinct sub-units based on phylogeographic 
lineages (Lecocq et a/., 2013a) or on evolutionary significant units (Lecocq et al., 2014) 
with specific climatic requirements. However, since the necessary data for defining these 
sub-units are not available in most of the case, we developed our SDMs at the species level 
and thus some level of uncertainty about potential effects of local adaptations remain. 

11.3. Modelling techniques 

There are now many different methodological approaches available to develop SDMs 
and all have their strengths and weaknesses (Elith ez a/., 2006; Heikkinen et a/., 2006). 
Since model performance and predictive ability have been shown to differ among these 
techniques (Thuiller et a/., 2003; Thuiller, 2004; Araujo et a/., 2005a; Elith & Graham, 
2009), using ensembles of different models to reach consensus among the different 
models has been suggested (Thuiller, 2004; Araujo et al., 2005a). However, building 
a consensus across a large variety of models bears the danger that models providing 
the most realistic future projections are in a minority and would contribute only little 
to the consensus. Thus, the challenge to develop and discriminate better models still 
remains. In the vast amount of literature comparing different modelling techniques, 
generalised linear models (GLMs) often appear, together with other approaches, such 
as generalised additive models (GAMs), boosted regression trees (BRTs) or MAXENT, 

as the most powerful approaches. 

In this atlas, we used GLMs despite the fact that they did not always provide the best 
model performance compared to general GAMs or BRI (Heikkinen et a/., 2006). 
However, GLMs had the overall best performance and their clear and simple math- 

ematical formulation allows highly accurate extrapolations into new environmental 
space (Elith et a/., 2006; Heikkinen et a/., 2012). 

An elegant model should also have a low level of complexity (i.e. number and complex- 
ity of terms used to explain the variability in the response variable) while maintaining 
a high level of performance (i.e. decreasing the residual variance). Thus the number 
of terms (including linear, second or higher order terms and interactions) must be 
reduced to a reasonable number (Harrell et a/., 1996). 

163 


164 

Climatic Risk and Distribution Atlas of European Bumblebees 

To reduce the complexity of our SDMs, we pre-selected ecological relevant and least 
correlated variables by means of cluster analysis. The threshold for variable selection 
was a Pearson correlation coefficient lower than 0.3 (Graham 2003). In this way we 
avoided statistical problems due to high levels of collinearity between climate variables. 
Further reduction of cemplexity was then undertaken by removing less important vari- 
ables when their removement minimised the AIC. The final model must be evaluated 
in terms of its prediction ability to assess model credibility. At best, the model predic- 
tions should be evaluated against an independent data set. However, assessing predic- 
tive ability of a model for future conditions is not possible, but transferring models 
from historical conditions to current conditions and vice versa could be a solution 
(Dobrowski et a/., 2011). However, in many cases, which apply also to the bumblebee 
data, historical data often suffer from lower sampling intensities, at least in some re- 
gions. Consequently, false absence data are likely increased in historic data sets, which 
make the interpretation of such back-casting evaluations difficult. Thus, splitting the 
data by random resampling the original dataset into a training and an evaluation data- 
set is a commonly applied alternative approach (Fielding & Bell, 1997; Olden & Jack- 
son, 2000; Araujo et al., 2005a). However, such a procedure can only be used to assess 
the ability of the model to predict current conditions, but this does not necessarily 
imply that these models are also able to accurately transfer their predictions in space 
(Heikkinen et a/., 2012) or time (Randin et a/., 2006). Since no proper independent 
dataset was available for the evaluations of our SDMs for European bumblebees, we 
calibrated our models on an 80% random sample of the initial data set and model 
accuracy was evaluated on the remaining 20%. 

Many measures are available for model evaluation (Fielding & Bell, 1997). Most wide- 
ly used are Cohen’s Kappa (Cohen, 1960) and the Area Under the Receiver Charac- 
teristic Curve (AUC; Hanley & McNeil, 1982). However, it was shown that values of 
Cohen’s Kappa and AUC should be interpreted with caution, since they depend on 
species prevalences (the fraction of occurrences relative to all data points) which makes 
model evaluations unreliable for species with very high or low prevalences (Allouche 
et al., 2006; Lobo et al., 2008, 2010). However, the True Skill Statistic (Peirce, 1884; 
Allouche et al., 2006) is independent of prevalence and represents a powerful measure 

of predictive ability. 

To maintain consistency with the Climatic Risk Atlas of European Butterflies (Settele 
et al., 2008), we nevertheless keep the AUC values in the species pages and for the as- 
signment of species to the risk categories. In addition, we also provide TSS and Kappa 
values together with values for specificity (proportion of correctly predicted occurrenc- 
es) and specificity (proportion of correctly predicted absences; Appendix 1). 


Methodological limitations 

11.4. Interpretation of the results 
11.4.1. General remarks 

Projections of species distribution models to future climatic conditions are often mistak- 
en as predictions of future species ranges. However, this is not what SDMs can provide. 
Moreover, they rely on scenarios of potential ways how environmental conditions might 
change in the future (see chapter 6). Under the assumptions of the different scenarios 
the SDMs project assessments of current suitable climatic conditions to the future, thus 
they only indicate areas where a species could in principle occur according to its climatic 
requirements. However, these projections do not allow drawing conclusions whether the 
species will actually be able to colonise the new areas or necessarily have to vanish at 
once in areas of increasingly unsuitable climatic conditions. Thus, the resulting projected 
changes in suitable climatic conditions cannot be translated one-to-one into actual range 
changes but they can be used to assess potential risks of climate change. To obtain a more 
realistic assessment of actual changes in species ranges two main processes might be dis- 
criminated (i) potential colonisation of new areas with suitable climates (leading edge) 
and (ii) extinction in the areas which are projected to become unsuitable (trailing edge). 

11.4.2. Processes at the leading edge 

The ability of a species to colonise new areas with suitable climatic conditions first of all 
depends on the likelihood that it can reach these areas, which basically is determined 
by the dispersal ability of the species but also on the frequency of anthropogenic dis- 
placement. 

In contrast to the assumed high levels of range filling during the long time period since 
the last glacial maximum, bumblebees, especially those with low dispersal capacity, 
might be assumed to lag considerably behind changing climatic conditions. Birds and 
butterflies have been shown to be unable to follow changing climatic conditions sufh- 
ciently during the last twenty years (1990-2008) (Devictor et a/., 2012). Moreover, the 
observed climatic debts of birds and butterflies correspond to a 212 and 135 km lag 
behind climate (Devictor et a/., 2012). Thus, it might be assumed that climate change 
velocities are also much higher than colonisation rates of many bumblebees. Further, 
natural barriers might additionally hamper the colonisation of new areas. For instance, 
some bumblebee species which are restricted to southern mountains and have not yet 
colonised suitable areas in Scandinavia (e.g. B. pyrenaeus). Thus, it is highly unlikely 
that such bumblebee species profit from gains in suitable climatic conditions there. 

Intentional anthropogenic displacements, on the other hand have the potential to lead 
to quick and massive species translocations. For instance, the current bumblebee in- 
ternational trade leads to the importation of nests by over 50 countries for pollination 

165 


166 

Climatic Risk and Distribution Atlas of European Bumblebees 

services to agriculture (Velthuis & van Doorn, 2006), an industry now worth billions 
of dollars annually (Goulson, 2003; Winter et a/., 2006). More than two million B. ter- 
restris colonies (the main bumblebee species used in crop pollination) produced each 
year are shipped throughout the world (Goka e¢ a/., 2001; Velthuis & van Doorn, 
2006). Such commercial translocation has resulted in several introductions around the 
world (Buttermore e¢ a/., 1998; Goulson & Hanley, 2004; Torretta et a/., 2006; Nu- 
katsuka & Yokoyama, 2010; Williams et a/., 2012). Beside the potential low efficiency 
of translocations for the bumblebee conservation (see chapter 14), only one European 
species is currently traded making quite unlikely the translocation of e.g. threatened 
species by the international trade. Moreover, international trade appears more as a 
major threat for the bumblebee fauna rather than a hope for their conservation (Inoue 
et al., 2008; Kanbe et al., 2008; Yoon et al., 2009; Nagamitsu et a/., 2010; Aizen et al., 
2011; Meeus et a/., 2011; Arbetman et a/., 2013a,b; Murray et al., 2013). 

Non-intentional anthropogenic translocations of species could also improve the likeli- 
hood of moving them into novel suitable areas. ‘This is usually seen as a problem when 
alien species are concerned, especially when they are causing problems within ecosys- 
tems or for human health or economy (Richardson et a/., 2000; Jeschke et a/., 2014), 
but might be beneficial for European species which would otherwise seriously suffer 
from range contractions. 

Once a species has reached a new area, survival will depend on the successful establish- 
ment and growth of populations. However, a large range of preconditions have to be 
met in terms of abiotic and biotic requirements of a species in addition to climatic suit- 
ability (Davis et a/., 1998; Heikkinen et al., 2004; Schweiger et al., 2008, 2010, 2012). 

Key resources for bumblebees are pollen and nesting or hibernation sites. However, the 
most specialised species (e.g. B. gerstaeckeri) may not find relevant resources in the oth- 
erwise climatically suitable novel areas. We also cannot expect that the relevant resources 
move simultaneously with the respective species. For instance, it has been shown for but- 
terflies and their larval host plants, that their climatic niches could overlap only to some 
extent (Hanspach et a/., 2014) and due to these differences in the climatic niches, future 
climate changes could lead to drastic spatial mismatches between areas suitable for but- 
terflies or host plants (Schweiger et a/., 2008). Further, even if the climatic niche spaces of 
bumblebees and their pollen plants would change similarly, other constraints can restrict 
the pollen sources to colonise the novel areas successfully. Dispersal limitations might be 
one limiting factor, hostile soil conditions another. For instance, deciduous forests which 
usually grow on brown soils will shift towards the current taiga and podzol soil. Steppes 
which are associated with sierozem or chernozem soils will move towards brown soils and 
podzols. Of course, these soils themselves will change their chemistry as a response to the 
new climatic conditions but this is a very slow process, taking typically not centuries but 
thousands of years. Thus, it will most likely not be entire vegetation complexes or plant 


Methodological limitations 

communities which shift their ranges, but single species will move individualistically. 
This also means that finding proper pollen resources might get increasingly difficult the 
farther the range has to be shifted, usually towards the north or upwards in altitude, to 
keep track with climate change. Thus, many species are likely to considerably lag behind 
future climate change due to a lack of proper pollen resources in the novel areas. 

The potential problem of spatial mismatch between interacting species will even be 
more pronounced when one species entirely depends on one or few others (Schweiger 
et al., 2008). This is the case for many parasitic (inquiline) bumblebee species. Here the 
same principle of individualistic responses of the host and the social parasite to chang- 
ing climates because of differences in the climatic niche, dispersal ability or colonisa- 
tion success could lead to a decreased, or impeded, ability to colonise new climatically 
suitable areas for Psithyrus species. 

11.4.3. Processes at the trailing edge 

There are also several reasons why a species might not become extinct immediately from 
areas where climatic conditions are projected to become unsuitable. Extinction of these 
populations can be avoided if they move to favourable refugia (Stewart & Lister, 2001) 
or if the individuals overcome the climatic stress through plastic changes (see below) or 
evolutionary adaptation (Williams et a/., 2008). Far from spread over several million years, 
evolutionary change can be rapid in a number of taxa (West-Eberhard, 1983), especially 
in fragmented populations (Blondel, 2000; Millien, 2006) or in populations under an- 
thropogenic pressure (Hendry et a/., 2008). Many studies show that species can display 
rapid evolutionary adaptations that help them to counter stressful conditions or realise 
ecological opportunities arising from climate change (review in Hoffmann & Sgré 2011). 

The adaptive capacity of bumblebee species has not been integrated into our models. 
This could probably bias our projections and lead to a more pessimistic picture of the 
future fate of European bumblebees. This limitation can be overruled by using the 
approaches developed for predicting and describing evolutionary responses to recent 
climate change in natural populations (review in Hoffmann & Sgrd 2011). However 
all of these methods have their own limitations (Hoffmann & Sgr, 2011) and cannot 
be easily applied to a large number of species. ‘Thus, when interpreting our results, one 
has to bear in mind that there is the possibility of rapid adaptation and that the actual 
losses at the trailing edge might be lower than projected for some species. So far, we 
can neither estimate the adaptive potential of the bumblebees nor potential differences 
therein among the species, but it is more likely that evolutionary adaptation potentially 
occurs under moderate change scenarios than under severe change scenarios. 

Another reason why species might persist even when climates are projected to become 
hostile are time lag effects. For some species it can take a considerable amount of time 

167 


168 

Climatic Risk and Distribution Atlas of European Bumblebees 

before declining populations disappear (Tilman et al/., 1994). These extinction debts 
have been observed in the course of habitat loss and fragmentation (Krauss et al, 
2010) but can in principle also occur under climate change (Jackson & Sax, 2010). 
Changing climatic conditions must not necessarily lead to instant extinction but can 
lead to unsustainable populations due to reduced fitness or competitive success. As a 
consequence, even moderate projections of range losses can ultimately lead to severe 
but delayed declines of species ranges (Dullinger et a/., 2012). It has also been shown 
that extinction debts depend on the longevity of the organisms (Krauss et a/., 2010) 
and that short-lived animals respond more rapidly to climate change (Thomas et al., 
2004; Morris et al., 2008; Devictor et al., 2012). Thus, for bumblebees we could expect 
a minor impact of extinction debts at the trailing edge of distribution. 

A further reason for potentially sustaining populations under projected unsuitable 
conditions is a more trivial one and concerns the resolution of our spatial data and 
the model output. For reasons of model reliability, we used a rather coarse grid of 50 
km x 50 km. To increase the information content for our projections we downscaled 
the models to a 10 min x 10 min grid. Downscaling to a certain extent is justifiable 
(Araujo et al., 2005b) but there are obvious limits to how far downscaling can go. 
Thus a resolution of 10 min x 10 min is still coarse enough not be overly precise but 
also to ignore climatic variability within the grid cells. As a consequence, there might 
still be smaller areas within a grid cell where suitable climatic conditions remain, e.g. 
by shifting such conditions from a south-facing side of a hill to the north-exposed 
side. Such micro-refuges could ensure the persistence of a species within a grid cell 
in which the average conditions are projected to turn hostile (Austin & Van Niel, 
2011; Lawson et al., 2014). With the current resolution of our projections we cannot 
assess the importance of such micro-refuges for the future fate of bumblebees at their 
trailing edges. However, since such small-scale beneficial conditions are more like- 
ly to occur in more heterogeneous areas, the likelihood of bumblebee populations 
surviving is higher in mountainous areas and areas with a larger number of different 
types of habitats. 

11.5. Conclusions on SDM limitations 

Here we highlighted the major limitations of SDMs and future projections. Given all 
these limitations it is obvious that SDMs cannot represent the entire complexity of re- 
al-world systems. Moreover SDMs depend on and are only valid under simplifying as- 
sumptions. These limitations might question the usefulness of such an approach. How- 
ever, if the simplifications are accepted and the limitations considered as good as possible, 
such simplified assessments can even help to gain a better understanding of the basic 
patterns and underlying natural processes, while not getting lost in all the species- and 
context dependencies. Thus, SDMs cannot predict the future fate of bumblebees, but 


Taxonomic issues 

they are a strong tool to assess their climatic risks in terms of potential changes in climat- 
ically suitable areas. They can help to identify areas of particular conservation concern, 
e.g. areas with an increased risk of species loss or areas with an increased level of colonisa- 
tion credits, i.e. areas that species could colonise and thus maintain large enough ranges 
or even ensure survival but are unlikely to reach them in time on their own. 

12. Taxonomic issues 

For the following species, there are some taxonomic issues that could potentially affect 
the interpretation of the results. Prior to computation, we had to make some assump- 
tions and simplifications according to taxonomic knowledge and data availability. 

12.1. Bombus confusus and B. paradoxus 

Bombus confusus includes two well differentiated taxa (Reinig, 1939): (i) B. confusus 
confusus Schenck 1859, with an all-black coat and a red tail; (ii) B. confusus par- 
adoxus Dalla Torre 1882, with three yellow bands and a white tail. Some authors 
assumed that B. confusus confusus and B. confusus paradoxus were different species 
(e.g. Pittioni, 1938). Following most authors (Rasmont, 1983; Williams, 1998) and 
due to limited data availability at subspecies level, we assumed here that B. confusus 
confusus and B. confusus paradoxus are conspecific. However, we should keep in mind 
that these two taxa are conspicuously different and that their conspecificity is poorly 
grounded. These two taxa seem also to have a quite distinctly different fates during 
the 20" century. While both taxa were widespread across the entire mainland of 
Europe at the beginning of the 20" century, B. confusus paradoxus disappeared quite 
early in most parts of western and central Europe. In the Volga basin and in Siberia, 
on the other hand, B. confusus paradoxus remains by far the more abundant subspe- 
cies. We cannot exclude that these taxa could be two separate species and that they 
could have quite different eco-climatic requirements. If this is the case, then their 
climatic risk should be assessed separately. 

12.2. Bombus cryptarum, B. lucorum, B. magnus, and B. terrestris 

In the majority of their ranges, these four taxa constitute a group of cryptic species (see 
e.g. Williams et a/., 2012). Until recently, several confusions in species identification 
between these four species occurred (Rasmont, 1983; Rasmont et al., 1986; Bertsch, 
1997; Williams et af, 2012). This means that most of the historical data could be 
based on misidentifications: old observations of one of these species could include 
observations of other species. As a result, we cannot exclude that the distribution and 
abundance of B. /ucorum are defined with the same accuracy (i.e. overestimation of the 
species distribution) as it could be for easily identified species. 

169 


170 

Climatic Risk and Distribution Atlas of European Bumblebees 

12.3. Bombus lucorum and B. renardi, B. terrestris, and B. xanthopus 

Bombus renardi and B. xanthopus are closely related to B. lucorum and B. terrestris, respec- 
tively (Rasmont & Adamski, 1995; Lecocq et al., 2013, 2014). In contrast to other taxa 
from the group of cryptic species including B. /ucorum and B. terrestris (see above), B. 
renardi and B. xanthopus are phenotypically well differentiated from their sibling species 
(Rasmont & Adamski, 1995). Both species have been described as distinct species (Kriech- 
baumer, 1870; Radoszkowski, 1884) but later included in B. lucorum and B. terrestris, 
respectively (Vogt, 1909; Kriiger, 1951; Rasmont & Adamski, 1995). However, Lecocq 
et al. (2013, 2014) recently demonstrated their species status according to their differenti- 
ation in genetics, morphology and species specific male secretions from B. /ucorum and B. 
terrestris. Even if we recognise that B. renardi and B. xanthopus are two distinct species, we 
considered them together with B. /ucorum and B. terrestris respectively for means of data 
reliability. This could marginally affect the B. /ucorum and B. terrestris modelling. 

12.4. Bombus cullumanus 

Until recently, the specific status of B. cullumanus (Kirby 1802), B. serrisquama Moraw- 
itz 1888, and B. apollineus Skorikov 1910 remained doubtful. Recent evidence from 
COI barcodes and male species identification secretions is consistent with the three 
taxa being part of a single species (Rasmont et a/., 2012; Williams et al, 2013). We 
used this new lumped taxonomic status in our analyses. 

12.5. Bombus laesus and B. mocsaryi 

Bombus laesus has been until recently considered as conspecific with B. mocsaryi (= 
maculidorsis Panfilov). Brasero et al. (2012) recently showed that these taxa diverge 
not only in their conspicuously different colour patterns but also by their morphol- 
ogy, their genetics and the composition of their species recognition male secretions. 
Reinig (1939) already showed that these two taxa also occupy quite distinctive 
biogeographical areas. B. /aesus seems to be associated with true steppes, while B. 
mocsaryi lives in woody-steppes and in grasslands, reaching almost to the Arctic Cir- 
cle in north Russia. Both species seem to have regressed considerably during the last 
decades. However, since this split has occurred very recently, we do not have reliable 
data to model these two species separately and thus we consider and model them as 
a single species. 

12.6. Bombus muscorum and B. pereziellus 
Bombus muscorum is a polytypic species with quite numerous distinct allopatric sub- 

species. Some authors (e.g. Kruseman, 1964) have lumped all the subspecies with 
black-haired legs in a distinct species for which the priority name would be B. bannitus 


Taxonomic issues 

Skorikov (= smithianus auct.). All the taxa associated with bannitus are mainly insular, 
living on the Atlantic littoral from La Coruna, in Spain to Namsos in Norway and in 
most of the small British islands (but not on the mainland). All taxa associated with 
muscorum s.s. are living on the mainland, more often along the sea coast but also on the 
continent, and even in central Asia, reaching eastward to Mongolia. 

Bombus pereziellus has been described as an endemic Corsican subspecies of B. mus- 
corum (under the name B. cognatus nigripes Pérez 1909). Later it was considered as 
a distinct species (Skorikov, 1922), while in the following it was considered to be 
a subspecies of B. muscorum again (Delmas, 1976; Rasmont, 1983). More recently 
Rasmont & Adamski (1996) considered pereziellus as a distinct species. A recent 
study nevertheless suggested that pereziellus is an insular subspecies of B. muscorum 

(Lecocq et al. 2014). 

Here we assumed that they are conspecific since most of studies considered them as a 

single species (Loken, 1973; Alford, 1975; Rasmont, 1983; Williams, 1998). 
12.7. Bombus niveatus and B. vorticosus 

Bombus niveatus includes two taxa: niveatus Kriechbaumer 1870 and vorticosus Ger- 
staecker, 1872. In Europe (the Balkans, Romania, Ukraine), only the ssp. vorticos- 
us occurs. In Turkey, the Caucasian region and Iran, both subspecies niveatus and 
vorticosus can be present together. However, while vorticosus occurs from sea level 
in Greece, ssp. niveatus only occurs in mountains. Numerous authors considered 
vorticosus to be a distinct species (e.g. Pittioni, 1938). However, these taxa have been 
more recently considered to be conspecific (Williams, 1998; Rasmont et al., 2005). 
Following this recent status revision, we assumed here that vorticosus is a subspecies 
of B. niveatus. 

12.8. Bombus perezi and B. vestalis 

Lecocq et al. (2013, 2014) recently considered that B. perezi is conspecific with B. ves- 
talis. However, several authors have considered B. perezi as a distinct species (review in 
Rasmont & Adamski, 1995). We assumed here that it is a good species. Therefore the 
distribution of B. perezi is not included in our B. vestalis modelling. 

12.9. Bombus reinigiellus 

Bombus reinigiellus has been described as a distinct species endemic to the Sierra Ne- 
vada (south-east Spain). However, Castro (1987) synonymised it with B. hortorum 
until more material showed that B. reinigiellus is a separate species (Castro, 1988). We 
assumed here that B. reinigiellus is a good species. 

171 


172 

Climatic Risk and Distribution Atlas of European Bumblebees 

12.10. Bombus sichelii and B. erzurumensis 

Bombus erzurumensis and B. sichelii are two closely related taxa considered as con- 
specific (Williams 1998) or as distinct species (Rasmont et al. 2000). Nevertheless, 
recent taxonomic revision based on genetic, morphology, and species-specific attractive 
compounds suggest that the two taxa are conspecific (Lecocq et al. in press). We here 
followed these recent statuses. 

12.11. Bombus handlirschianus and B. shaposhnikovi 

These two taxa should be better considered as conspecific (Cameron et al. 2007; De 
Meulemeester et a/. 2011). 

12.12. Bombus lapponicus, B. monticola and B. glacialis 

Bombus lapponicus was considered as a species distinct from B. monticola by Svensson 
(1979) while B. glacialis was also regarded as a different species from the two other 
species by Berezin (1990). We follow these authors and separate these species. 

12.13. Bombus lapidarius and B. caucasicus 

Bombus lapidarius included five subspecies (Rasmont 1983; Reinig 1935, 1970; 
Tkalctit 1960): (i) dapidarius (L.) in the European plains, Balkans and West Anatolia, 
(ii) decipiens Pérez 1890 in the Iberian Peninsula and in Southern Italy, (iii) cauca- 
sicus Radoszkowski 1859 in the North East Anatolia and Caucasus, (iv) eriophorus 
Klug 1807 in Caucasus, and (v) B. lapidarius atlanticus Benoist 1928 in the Moroc- 
can Atlas. However, recent genetic and eco-chemical studies showed that caucasicus 
is a different species from B. lapidarius (Lecocq et al. 2013a, in press). Nevertheless, 
eriophorus (not studied by Lecocg et al. 2013a) and B. caucasicus have been consid- 
ered as two forms of the same taxon by Reinig (1935) while Rasmont (1983) regard- 
ed them as two different taxa. If eriophorus and B. caucasicus are to be considered 
conspecific, B. eriophorus (Klug, 1807) would be the oldest available name for the 
species. Further analyses on B. lapidarius eriophorus and B. caucasicus are needed to 
assess their conspecificity. 

12.14. Bombus barbutellus and B. maxillosus 

Bombus barbutellus and B. maxillosus were previously considered by most authors 
as two closely related species (review in Lecocg et al. 2011). However, the two taxa 
have been shown to be conspecific by Lecocq et al. (2011). We follow these authors 
and considered the taxa as conspecific, Bombus barbutellus being their senior name. 


Climate change and bumblebee conservation 

13. Climate change and bumblebee conservation 
13.1 Climatic risks of European bumblebees 

Bumblebees are clearly cold-adapted species (Fig. 13.1; Heinrich, 1979). While the mean 
annual temperature ranges between -3.6°C and 22.1°C with a median value of 9.2°C for 
the analysed geographic window (Fig. 13.1A), the average temperature requirements of 
all bumblebee species did not reflect this broad range but are concentrated at intermedi- 
ate to cold conditions (Fig. 13.1B). Average temperature requirements of the bumblebees 
were calculated as the mean of the mean annual temperature values across the grid cells in 
which a particular species occurred. This value is also known as species temperature index 
(STI; Devictor et a/., 2008) and has successfully been used for the assessment of commu- 
nity changes in response to recent climate change (Devictor et al, 2012). The STI values 
for the European bumblebees range between -1.6°C (Bombus hyperboreus) and 10.4°C 
(B. ruderatus) with a median of 7.0°C. The species with the lowest STI values (<5°C) 
formed a separate group (Fig. 13.1B) and were dominated by boreal and Arctic species 
with particularly small distributions (Appendix 3). Given these climatic preferences, Eu- 
ropean bumblebees can be considerably affected by climate warming. 

Taking into account their most likely dispersal abilities, we can project the severity of 
suitable area changes for 2100 (see Appendix 3 for species values). 13 species have not 
been modelled. In all three scenarios, only 3 species are expected to expand their suit- 
able climatic area. With the less severe scenario (SEDG), 3 species (4%) are expected to 
loose more than 80% of their suitable area (meaning severe risk of extinction); 27 spe- 
cies (39%) from 50 to 80% of their suitable area and 23 (33%) species from 20 to 50% 
suitable area lost. With the intermediate scenario (BAMBU), 14 species (20%) should 
loose more than 80% of suitable area; 33 species (48%) from 50 to 80% of suitable area 
and 6 species from 20 to 50% lost. For the most extreme scenario (GRAS), as much as 
25 species (36%) should loose more than 80% of their suitable area while 28 species 
(41%) would loose from 50 to 80% of suitable area. This last case means that a total of 
77% of the bumblebee species would lose the largest part of their suitable climatic area, 
with more than a third of the total number of species driven at the verge of extinction. 

173 


174 

Climatic Risk and Distribution Atlas of European Bumblebees 

3 
w 
Se o 
oO ce 
2 o 
LL OY 
8 
5 8 18: 16, Bel 25 G@ «By 4 4 
Temperature (°C) Temperature (°C) 

Figure 13.1 Temperature conditions across Europe and climatic requirements of the European bumblebees. (A) Frequency 
distribution of mean annual temperature at a grid resolution of 50 km x 50 km. (B) Species temperature index (STI) of all 
modelled bumblebee species and their respective range size in numbers of UTM 50 km grid cells. The species with the lowest 
STI are Arctic and Boreal species that are the most sensitive to climate warming (see Appendix 3). 

i more than 80% los 

@ 50 io 80% lost 

G 20 to 50% lost 

B from 20 to 20% changes 
© 20 to 80% gain 

@ more than 80% gain 

OD not modelled 

Figure 13.2 Severity of projected changes in 2100 for the 69 studied European bumblebee species. For 21 species we 
assumed full dispersal, for the remaining no dispersal (see Appendix 3). Thirteen species have not been assessed (white 
background). Dark green background indicates a large expansion (more than 80% gain in suitable area); light green indicate 
expansion (between 20 and 80% gain in suitable area); yellow background indicates regression (between 20 and 50% loss of 
suitable area); red background indicates strong regression (from 50 to 80% loss of suitable area); dark background indicates 
very strong regression with extinction risk (more than 80% loss of suitable area). A. SEDGE scenario; B. BAMBU scenario; 
C. GRAS scenario. 

13.2. Potential mitigation strategies 

As mentioned in chapter 11, the actual response of species to changing climatic condi- 
tions depends on whether the species will be able to colonise new climatically suitable 
areas or can survive, at least for a while, in areas of increasingly unsuitable climatic con- 
ditions. These two ways of responding render different conservation actions possible. 
In principle they should aim at (i) guaranteeing the unrestricted, or even aid, move- 
ment of the species through the landscape to new areas, (ii) facilitate the colonisation 
success in the new areas, (iii) improve habitat conditions and microclimatic protection 
in the areas indicated to become unsuitable at average. 


Climate change and bumblebee conservation 

13.2.1 Are translocations of threatened species possible? 

‘There is an ongoing debate about whether species threatened by climate change should 
be actively translocated to regions which are projected to become suitable in the near 
future (Thomas, 2011; Vila & Hulme, 2011; Webber et a/, 2011; Miiller & Eriksson, 
2013). Thomas (2011) argues that the benefits of translocation will outweigh the asso- 
ciated risks. This point of view has received strong criticism, especially given the expe- 
riences from invasion ecology (Webber et a/. 2011; Vila & Hulme 2011). A further ar- 
gument concerns the dependency of translocation success on the level of specialisation. 
Webber et a/. (2011) argue that the chances of successful translocations are highest for 
generalist species which do not depend largely on prey or mutualists. However these 
species are known to be likely to cause severe problems in recipient locations. On the 
other hand, Miller and Eriksson (2013) concluded in a recent study that translocation 
can prevent more global extinctions than it can cause and thus claim that translocation 
should be more widely accepted as a conservation tool. 

For most of the threatened species in the Alps or in the Pyrenees, one solution might 
be to translocate them to the Scandinavian mountains, while for the most threatened 
arctic species (like B. alpinus, B. balteatus, B. hyperboreus, and B. polaris), the survival 
on the European mainland is unlikely. Three of these species nevertheless also occur 
outside Europe where they may survive in the extreme north of the Siberian Arctic 
(e.g. Taymir and Anadyr peninsulas), in Alaska, northern Canada or in Greenland. 
Bombus alpinus, on the other hand, is endemic to Europe. Thus disappearance from 
the European mainland would mean total extinction. To avoid the total extinction of 
this species, translocation to some northern Archipelagos, like Svalbard, Franz Josef 
Land or Novaya Zemlya might be a solution. 

Undoubtedly, several bumblebee species have already been successfully translocated 
to different countries or even different continents. At the end of the 19" century, four 
species (B. terrestris, B. hortorum, B. ruderatus, and B. subterraneus) were moved from 
England to New Zealand, where they thrive (Buttermore et a/., 1998; Goulson et al, 
2002; Velthuis, 2002; Goulson & Hanley, 2004; Torretta et al, 2006; Yokoyama & 
Inoue, 2010). From New Zealand, B. ruderatus has then been moved with success to 
Chile, where it is now more abundant than native species (Goulson & Hanley, 2004). 
From New Zealand again, some colonies of B. terrestris have been moved to Tasmania 
where the species settled and expanded very rapidly (Buttermore et a/., 1998; Goulson 
et al., 2002). The domestication of B. terrestris has led to its translocation to numerous 
countries where it has successfully established, e.g. in Japan, Argentina and Chile (Tor- 
retta et al., 2006; Yokoyama & Inoue, 2010). This species could be even considered as 
invasive and one could expect that sooner or later, all climatically suitable areas in the 
world will be colonised by B. terrestris (Peredo-Alvarez et al. 2014). 

175 


176 

Climatic Risk and Distribution Atlas of European Bumblebees 

The extreme success of B. terrestris in invading new areas results from the extremely 
high number of translocated individuals (hundreds to thousands per event) but also 
from the high level of adaptability of this species (Rasmont et al., 2008). Bombus terres- 
tris is able to produce up to three generations per year, with a very adaptive phenology. 
It is also one of the most generalist foragers. Very few other bumblebee species share 
all these characteristics. 

Nevertheless, in some cases, successful colonisation could not be supported even by 
the import of thousands of B. terrestris colonies, as has been the case for the Sardinian 
B. terrestris sassaricus which has been exported to areas in Southern France (Ings et al., 

2010). 

Taking these experiences into account, we can assume that any successful translocation 
would require moving very large numbers of threatened species to the new targeted 
area. However, most of the species at high climatic risk are already rare and finding a 
sufficient number for translocation in the wild would be difficult if not impossible. For 
instance, for the most threatened A/pinobombus species, it would already be difficult 
to collect even one or two dozens of queens. Further, a recent attempt to reintroduce 
B. subterraneus from New Zealand to England, from where it disappeared, did not 
succeed (Gammans et al., 2009) and also another try with Swedish strains has not yet 
proven successful (Sears, 2014). 

If we, nevertheless, assume that such translocations can be done, several other prob- 
lems arise. As the example of B. terrestris has shown, translocated species can cause 
severe disruption to existing ecosystems. Novel species can lead to the reduction of 
native species when they are better competitors (Stout & Morales, 2009; Nagamitsu et 
al., 2010), worse pollinators (Kenta et al, 2007) or introduce novel pathogens (Stout 
& Morales, 2009), and thus put additional pressures on the native bee and plant com- 
munity which might already be suffering from climate change (Schweiger et a/., 2010). 
Even if the translocated species are readily integrated into the local communities, they 
may not find their required pollen and nectar plants or preferred nesting sites and 
material in the new areas. ‘Thus, simple translocations can easily fail unless the ecology 
of the species with all their relevant resources, and important interactions with other 
species, are well known and their consequences can be assessed and evaluated. 

13.2.2 Supporting species on the move 
While species, whose climatically suitable conditions will just retract, such as Alpine 

species, might depend on active translocations by humans to overcome large areas of 
hostile climates, there are quite many species whose climatic conditions are projected 


Climate change and bumblebee conservation 

to move continuously. Such species might be able move along with their suitable cli- 
mates and thus expand their current ranges. There is evidence that some bumblebee 
species are able to spread fast and even cross narrow sea channels (Mikkola, 1978). In 
one of the most famous case, the rapid settlement and expansion of B. hypnorum in the 
UK indicates that this species was able to cross the channel between mainland Europe 
and the UK (Goulson et a/., 2011). However, the ability of species to successfully keep 
track with climate change depends on the general dispersal ability of the species and, 
moreover, on the landscape structures that have to be crossed (Hill et a/., 1999). 

To support bumblebees in keeping track with changing climates the landscapes should 
be managed in a way that moving bumblebees can find enough and species-specific 
resources such as wild flowers and nesting sites. Unfortunately, there are large parts of 
Europe which are heavily dominated by intensive agriculture. In such cases, a shortage 
of food and additional pressures from pesticides might considerably hamper the abil- 
ity of the species to move through the landscape and thus follow changing climates. 
Flower strips as part of European environmental schemes are very likely to support the 
moving species. 

13.2.3 Supporting species at their trailing edge 

As mentioned in chapter 12, the projected retractions at the southern and lower alti- 
tudinal edges of bumblebee distributions were done for average conditions within a 
grid cell of 10 min x 10 min. This ignores microclimatic variability due to habitat and 
topographic heterogeneity. Thus, the areas of loss might be better regarded as deterio- 
rating average conditions which provide some space for successful conservation action. 
Management in a way that maximises microclimatic heterogeneity could allow some 
species to survive in micro-refugia, as has been observed with some vertebrate species 
(Willis & Bhagwat, 2009; Willis et a/., 2010; Morelli et a/., 2012) or suggested for 
butterflies (Lawson et al, 2014). 

Special attention should also be paid to natural features with major influence on the lo- 
cal microclimate. A very good example is the Forét de la Sainte-Baume in south-eastern 
France (Fig. 13.6).Thanks to the shelter of a high cliff, a large beech (Fagus sylvatica) 
and yew (Taxus baccata) forest persists there since at least two thousand years (Chalvet 
2013), while the surrounding area is characterized by a dry Mediterranean vegetation, 
deeply impacted by recent droughts (Villa et al. 2008). Here Bombus pratorum and 
other rare wild bees can survive (Terzo & Rasmont, 2003), whilst they do not occur 
in the surrounding area. Most hills and mountains include such areas of potential mi- 
crorefugia and thus they should be of particular conservation concern and should be 
monitored with appropriate programmes. 

177 


178 

Climatic Risk and Distribution Atlas of European Bumblebees 

Figure 13.6 The Forét de la Sainte-Baume, near Marseille (Photo Georges Millet). On the right, the canopy of the beech forest 
sheltered by the cliff; on the left, the dry Mediterranean vegetation. 

Species at their range margins can be particularly sensitive to additional threats apart 
from climatic limitations (Williams, 1988; Thomas et a/., 2001; Williams et a/., 2007; 
Oliver et al, 2009). In such cases, management strategies should aim at reducing ad- 
ditional threats while providing as many micro-refugia as possible. Particularly im- 
portant threats at the range margins could be malnutrition (due to reduced availability 
of floral resources) or increased stress (due to pathogens or pesticides). Again, proper 
management in agricultural areas could contribute enormously to the persistence of 
bumblebees in areas where the average climatic conditions are projected to get less suit- 
able in the future. Conserving populations at the trailing edge of their distribution can 
be of particular importance because they can act as long-term stores of species genetic 
diversity and foci of speciation (Hampe & Petit, 2005). 

Bombus alagesianus is a medium to large species that Bombus modestus is a small species typical of Siberian 
inhabits the steppes of alpine and subalpine levels in east taiga where it could be very abundant. In Europe, it only lives 
Turkey, north Iran and Caucasian region. It also lives in high in few locations in boreal forest between Moscou and Ural. 
steppes of Central Asia. It is a rare species and few are known Its coat colour could be very variable but generally shows a 
about its way of life. It presents a typical colour pattern, with largely yellow thorax and tergites 1 and 2 and with black and 
yellowish bands and a reddish abdomen tip. Photo P. Rasmont. grey on the remaining of abdomen. Photo P. Rasmont. 


Conclusions 

14. Conclusions 

Thanks to the EU FP7 project STEP (Potts et a/, 2011), over one million bumblebee 
records from all over Europe have been collated. Based on data from 1970 to 2000 
we modelled the current climatic niche for almost all European species (56 out of 69) 
and projected future climatically suitable conditions based on three climate change 
scenarios (SEDG, BAMBU and GRAS) for the years 2050 and 2100. Due to limited 
knowledge of actual bumblebee dispersal, we made two extreme assumptions: (i) the 
species has full dispersal abilities (meaning that the species is able to spread all over its 
suitable area) or (ii) the species is unable to disperse at all (i.e. that changes in climatic 
conditions can only lead to projected range retractions; see chapter 6). However, to aid 
the assessment as to which of these two extreme assumptions are more likely to meet 
reality, we also provide a rough indication of the species’ potential dispersal ability 

based on the ecology of the different bumblebees. 

Since bumblebees are mainly adapted to colder conditions, they appear as highly vul- 
nerable to climate change. In 2100, depending on the scenario of climate change, up 
to 36% of the European bumblebees are projected to be at an high climatic risk (i.e. 
losing more than 80% of their current range), 41% will be at risk (loss between 50% 
and 80%). In addition to the projections of the modelled species, the 13 non-modelled 
species have a restricted distribution and their ranges are most likely to be shrinking 
considerably under all of the scenarios. Only three species are projected to benefit 
from climate change and can potentially enlarge their current distributions in Europe, 
B. argillaceus, B. haematurus and B. niveatus. 

As expected, the three scenarios considered provide considerably different projections 
for 2100. While under the moderate change scenario (SEDG) only three species are 
projected to be at the verge of extinction by 2100. 14 species are at a particularly high 
risk under the intermediate change scenario (BAMBU). Under the most severe change 
scenario (GRAS) as many as 25 species are projected to lose almost all of their climat- 
ically suitable area. 

Also the ability to keep track with climate change has a considerable impact of the pro- 
jected changes. For instance, under the most severe climate change scenario (GRAS) 
eight species are at an extremely high climatic risk when full dispersal is assumed. 
However, under the assumption of no dispersal within the next 100 years, 34 species 
would fall into this category. When potential dispersal abilities, inferred from species 
traits and their autecologies, are considered to decide for which species no or full dis- 
persal assumptions are more realistic, it seems that three to four species might expand 
their ranges by 2100, no species is likely to remain at the status quo, and 25 species 
would be at an extremely high climatic risk. 

179 


180 

Climatic Risk and Distribution Atlas of European Bumblebees 

We also found that for many species (about 30%), especially the cold-adapted ones in 
Alpine and Arctic regions (e.g. B. alpinus, B. balteatus, B. hyperboreus and B. polaris) 
their dispersal abilities are actually irrelevant for the assessment of their future fate 
because climate change will only lead to reductions of areas with suitable climatic con- 
ditions while no extra suitable regions will emerge. 

Given the great sensitivity of bumblebees to climate change and further considering 
the severe projected changes in the light of the great relevance of bumblebees as polli- 
nators, designing management plans to sustain the highest level of pollination services 
on the one hand and to ensure the survival of as many bumblebee species on the other 
hand is of utmost importance. Given the different mechanisms leading to change, es- 
pecially at the leading versus the trailing edge of species distributions and the geograph- 
ical differences in the severity of climate change, management actions must be well 
and target-specific designed. One important issue would be to prioritise management 
actions across different geographic regions in Europe. We have seen that the expected 
species loss due to climate change increases with decreasing latitudes, i.e. that regions 
in the south of Europe will be most affected by the loss of important pollinators. 

Important means to support European bumblebees would be to facilitate the move- 
ment of species trying to keep track with changing climates at the trailing edge and to 
prolong the persistence in regions where climatic conditions are deteriorating. Land- 
scape management can be of particular help in this context. Increased connectivity and 
quality of bumblebee habitats can help colonising species, while habitat heterogeneity 
will generate heterogeneity in the microclimate and can thus increase population per- 
sistence at the trailing edge as a kind of “Noah's Ark”. Areas with naturally high levels 
of microclimatic heterogeneity (such as mountainous areas) can be of particular im- 
portance and deserve special attention. Finally, the idea of assisted migration, i.e. pur- 
poseful anthropogenic translocations, seems appealing at first sight for species whose 
original distributional areas are projected to shrink tremendously and cannot move to 
suitable areas because of natural or anthropogenic barriers. However, the feasibility of 
such actions is still questionable. 

To conclude, climatic risks for bumblebees can be extremely high, depending on the 
future development of human society, and the corresponding effects on the climate, 
strong mitigation strategies are needed to preserve this important species group and 
to ensure the sustainable provision of pollination services, to which they considerably 
contribute. 


References 

15. References 

Ahrné, K., Bengtsson, J. & Elmqvist, T. (2009) Bumble bees (Bombus spp) along a 
gradient of increasing urbanization. PLoS ONE, 4, e5574. 

Aizen, M.A., Lozada, M. & Morales, C.L. (2011) Comparative nectar-foraging behav- 
iors and efficiencies of an alien and a native bumble bee. Biological Invasions, 13, 
2901-2909. 

Alford, D. V (1975) Bumblebees, David Poynter, London. 

Allouche, O., Tsoar, A. & Kadmon, R. (2006) Assessing the accuracy of species dis- 
tribution models: prevalence, kappa and the true skill statistic (TSS). Journal of 
Applied Ecology, 43, 1223-1232. 

Amiet, FE (1996) Hymenoptera Apidae, 1. Teil. Allgemeiner Teil, Gattungsschlissel, die 
Gattungen Apis, Bombus und Psithyrus. Insecta Helvetica 12, Schweizerische Ento- 
mologische Gesellschaft, Neuchatel. 

Araujo, M.B. & Guisan, A. (2006) Five (or so) challenges for species distribution mod- 
elling. Journal of Biogeography, 33, 1677-1688. 

Araujo, M.B. & Pearson, R.G. (2005) Equilibrium of species’ distributions with cli- 
mate. Ecography, 28, 693-695. 

Araujo, M.B., Pearson, R.G., Thuiller, W. & Erhard, M. (2005a) Validation of spe- 
cies-climate impact models under climate change. Global Change Biology, 11, 
1504-1513. 

Araujo, M.B., Thuiller, W., Williams, PH. & Reginster, I. (2005b) Downscaling Eu- 
ropean species atlas distributions to a finer resolution: implications for conserva- 
tion planning. Global Ecology and Biogeography, 14, 17-30. 

Arbetman, M., Meeus, I., Morales, C., Aizen, M. & Smagghe, G. (2013a) Co-in- 
troduction and spillover of parasites by invasive Bombus terrestris. Bumblebee’s 
chauffer hitchhike alien parasites when traveling to Patagonia, Argentina. Bio- 
logical Invasions. Biodiversity and Conservation, 15, 489-494. 

Arbetman, M.P., Meeus, I., Morales, C.L., Aizen, M.A. & Smagghe, G. (2013b) Alien 
parasite hitchhikes to Patagonia on invasive bumblebee. Biological Invasions, 15, 
489-494. 

Austin, M.P. & Van Niel, K.P. (2011) Improving species distribution models for cli- 
mate change studies: variable selection and scale. Journal of Biogeography, 38, 
1-8. 

Bartomeus, I., Ascher, J.S., Gibbs, J., Danforth, B.N., Wagner, D.L., Hedtke, S.M. 
& Winfree, R. (2013a) Historical changes in northeastern US bee pollinators 
related to shared ecological traits. Proceedings of the National Academy of Sciences 
of the United States of America, 110, 4656-4660. 

Bartomeus, I., Ascher, J.S., Wagner, D., Danforth, B.N., Colla, S., Kornbluth, S. & 
Winfree, R. (2011) Climate-associated phenological advances in bee pollinators 
and bee-pollinated plants. Proceedings of the National Academy of Sciences of the 
United States of America, 108, 20645-9. 


182 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bartomeus, I., Park, M.G., Gibbs, J., Danforth, B.N., Lakso, A.N. & Winfree, R. 
(2013b) Biodiversity ensures plant-pollinator phenological synchrony against 
climate change. Ecology Letters, 16, 1331-8. 

Berezin M.V. (1990) Ekologiya i gnezdovanie shmelej na ostrove Wrangel’ [Ecology 
and nesting of bumblebess of Wrangel Island]. p. 19-28 in: Proceedings of Collo- 
quia of All-Union Entomological Society. Section for the Study of Social Insects, 
Leningrad, 2-8 October 1990. [in Russian]. 

Bertsch, A. (1997) Discrimination of the bumblebee species Bombus cryptarum and B. 
lucorum by means of male labial gland secretions and morphological characters 
(Hymenoptera : Apidae) Entomologia Generalis, 22, 129-145. 

Biesmeijer, J.C., Roberts, S.PM., Reemer, M., Ohlemiiller, R., Edwards, M., Peeters, 
T., Schaffers, A.P., Potts, S.G., Kleukers, R., Thomas, C.D., Settele, J. & Kunin, 
W.E. (2006) Parallel declines in pollinators and insect-pollinated plants in Brit- 
ain and the Netherlands. Science, 313, 351-354. 

Blondel, J. (2000) Evolution and ecology of birds on islands: Trends and prospects. Vie 
et Milieu, 50, 205-220. 

Brasero, N., Lecocq, T., De Meulemeester, T., Urbanova, K., Rami, M., Valterova, I., 
Rasplus, J.-Y. & Rasmont, P. (2012) The male sexual marking pheromones of the 
Bombus laesus group: toward an integrative approach. Eurbee 5th European Confer- 
ence of Apidology, p. 1. Halle an der Saale, Germany. 

Buttermore, R.E., Pomeroy, N., Hobson, W., Semmens, T. & Hart, R. (1998) As- 
sessment of the genetic base of Tasmanian bumble bees (Bombus terrestris) for 
development as pollination agents. Journal of Apicultural Research, 37, 23-25. 

Cameron, S.A., Lozier, J.D., Strange, J.P, Koch, J.B., Cordes, N., Solter, L.E & Gris- 
wold, T.L. (2011) Patterns of widespread decline in North American bumble 
bees. Proceedings of the National Academy of Sciences of the United States of Amer- 
ica, 108,662-667. 

Carvalheiro, L.G., Kunin, W.E., Keil, P., Aguirre-Gutiérrez, J., Ellis, W.N., Fox, R., 
Groom, Q., Hennekens, S., Van Landuyt, W., Maes, D., Van de Meutter, E, 
Michez, D., Rasmont, P., Ode, B., Potts, S.G., Reemer, M., Roberts, S.PM., 
Schaminée, J., Wallisdevries, M.E & Biesmeijer, J.C. (2013) Species richness 
declines and biotic homogenisation have slowed down for NW-European polli- 
nators and plants. Ecology Letters, 16, 870-878. 

Castro, L. (1987) Nuevas citas de tres Bombinae (Hym. pidae) de la Peninsula Ibérica. 
Boletin de la Asociacion Espanola de Entomologia, 11, 413. 

Castro, L. (1988) Nuevas citas de tres Bombinae (Hym. APidae) de la Peninsula Ibéri- 
ca. Boletin de la Asociacion Espanola de Entomologia, 12, 281-289. 

Chen, I.-C., Hill, J.K., Ohlemiiller, R., Roy, D.B. & Thomas, C.D. (2011) Rapid 
range shifts of species associated with high levels of climate warming. Science, 
333, 1024-6. 

Chalvet M. (2013) La forét domaniale de la Sainte-Baume : un espace exceptionnel et 
protégé en Provence. Les Cahiers de Framespa, 13, 1-20. http://framespa.revues. 
org/2250 


References 

Chittka, L., Ings, T.-C. & Raine, N.E. (2004) Chance and adaptation in the evolu- 

tion of island bumblebee behaviour. Population Ecology, 46, 243-251. 

Cohen, J. (1960) A coefficient of agreement for nominal scales. Educational and Psy- 
chological Measurement, 20, 37-46. 

Colin, M.E. & Belzunces, L.P. (1992) Evidence of synergy between prochloraz and 
deltamethrin in Apis mellifera L.: a convenient biological approach. Pesticide Sci- 
ence, 36, 115-119. 

Darvill, B., O'Connor, S., Lye, G.C., Waters, J., Lepais, O. 8& Goulson, D. (2010) 
Cryptic differences in dispersal lead to differential sensitivity to habitat fragmen- 
tation in two bumblebee species. Molecular Ecology, 19, 53-63. 

Davis, A.J., Jenkinson, L.S., Lawton, J.H., Shorrocks, B. & Wood, S. (1998) Making 
mistakes when predicting shifts in species range in response to global warming. 
Nature, 391, 783-6. 

De Meulemeester, T., Aytekin, M., Cameron, S. & Rasmont, P. (2011) Nest architec- 
ture and species status of the bumble bee Bombus (Mendacibombus) shaposhnikovi 
(Hymenoptera: Apidae: Bombini). Apidologie, 42, 301-306. 

Delmas, R. (1976) Contribution a l'étude de la faune francaise des Bombinae (Hyme- 
noptera, Apoidea, Bombidae). Annales de la Société Entomologique de France 
(N.S.), 12, 247-290. 

Devictor, V., Julliard, R., Couvet, D. & Jiguet, E (2008) Birds are tracking climate 
warming, but not fast enough. Proceedings of the Royal Society B: Biological Sci- 
ences, 275, 2743-2748. 

Devictor, V., van Swaay, C., Brereton, T., Brotons, L., Chamberlain, D., Heliola, J., 
Herrando, S., Julliard, R., Kuussaari, M., Lindstrém, A., Reif, J., Roy, D.B., 
Schweiger, O., Settele, J., Stefanescu, C., Van Strien, A., Van Turnhout, C., Ver- 
mouzek, Z., WallisDeVries, M., Wynhoff, I. & Jiguet, E (2012) Differences in 
the climatic debts of birds and butterflies at a continental scale. Nature Climate 
Change, 2, 121-124. 

Dobrowski, S.Z., Thorne, J.H., Greenberg, J.A., Safford, H.D., Mynsberge, A.R., 
Crimmins, S.M. & Swanson, A.K. (2011) Modeling plant ranges over 75 years 
of climate change in California, USA: temporal transferability and species traits. 
Ecological Monographs, 81, 241-257. 

Donath, H. (1986) Der Strassentod als bestandsgefahrdender Faktor fir Hummeln 
(Insecta, Hymenoptera, Bombidae). Naturschutzarbeit in Berlin und Branden- 
burg, 22, 39-43. 

Dormann, C.F, Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G., Marquéz, 
J.R.G., Gruber, B., Lafourcade, B., Leitao, P.J., Miinkemiiller, T., McClean, C., 
Osborne, PE., Reineking, B., Schréder, B., Skidmore, A.K., Zurell, D. & Laut- 
enbach, S. (2013) Collinearity: a review of methods to deal with it and a simula- 
tion study evaluating their performance. Ecography, 36, 27-46. 

Dullinger, S., Gattringer, A., Thuiller, W., Moser, D., Zimmermann, N.E., Guisan, 
A., Willner, W., Plutzar, C., Leitner, M., Mang, T., Caccianiga, M., Dirnbéck, 
T., Ertl, S., Fischer, A., Lenoir, J., Svenning, J.-C., Psomas, A., Schmatz, D.R., 

183 


184 

Climatic Risk and Distribution Atlas of European Bumblebees 

Silc, U., Vittoz, P. & Hiilber, K. (2012) Extinction debt of high-mountain 
plants under twenty-first-century climate change. Nature Climate Change, 
2, 619-622. 

Edwards, M. & Jenner, M. (2009) Field Guide to the Bumblebees of Great Britain and 
Treland, Ocelli Limited, London, UK. 

Elith, J. & Graham, C.H. (2009) Do they? How do they? WHY do they differ? On 
finding reasons for differing performances of species distribution models. Ecog- 
raphy, 32, 66-77. 

Elith, J., Graham, C.H., Anderson, R.P., Dudik, M., Ferrier, S., Guisan, A., Hijmans, 
R.J., Huettmann, E, Leathwick, J.R., Lehmann, A., Li, J., Lohmann, L.G., Lo- 
iselle, B.A., Manion, G., Moritz, C., Nakamura, M., Nakazawa, Y., Overton, 
J.McC., Peterson, A.T., Phillips, S.J., Richardson, K.S., Scachetti-Pereira, R., 
Schapire, R.E., Soberén, J., Williams, S., Wisz, M.S. & Zimmermann, N.E. 
(2006) Novel methods improve prediction of species’ distributions from occur- 
rence data. Ecography, 29, 129-151. 

ESRI (2013) ArcGIS Desktop: Release 10.2. 

European Environment Agency (2009) Looking back on looking forward: a review of 
evaluative. EEA Technical Report No 3/2009., Copenhagen. 

Fielding, A.H. & Bell, J.E (1997) A review of methods for the assessment of prediction 
errors in conservation presence/absence models. Environmental Conservation, 24, 
38-49. 

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

Free, J.B. (1993) Insect Pollination of Crops, 2nd edn. Academic Press, London, UK. 

Gammans, N., Banks, B. & Edwards, M. (2009) The return of the native: Loss and 
repatriation of the short-haired bumblebee Bombus subterraneus. British Wildlife, 
21, 116-118. 

Goka, K., Okabe, K., Yoneda, M. & Niwa, S. (2001) Bumblebee commercialization 
will cause worldwide migration of parasitic mites. Molecular Ecology, 10, 2095- 
2099. 

Goulson, D. & Hanley, M.E. (2004) Distribution and forage use of exotic bumblebees 
in South Island, New Zealand. New Zealand Journal of Ecology, 28, 225-232. 

Goulson, D. (2003) Bumblebees: their behaviour and ecology, Oxford University Press, 
Oxford. 

Goulson, D., Hanley, M.E., Darvill, B., Ellis, J.S. & Knight, M.E. (2005) Causes of 
rarity in bumblebees. Biological Conservation, 122, 1-8. 

Goulson, D., Kaden, J.C., Lepais, O., Lye, G.C. & Darvill, B. (2011) Population 
structure, dispersal and colonization history of the garden bumblebee Bombus 
hortorum in the Western Isles of Scotland. Conservation Genetics, 12, 867-879. 

Goulson, D., Lye, G.C. & Darvill, B. (2008a) Decline and conservation of bumble 
bees. Annual Review of Entomology, 53, 191-208. 


References 

Goulson, D., Stout, J.C. & Kells, A.R. (2002) Do alien bumblebees compete with native 
flower-visiting insects in Tasmania? Journal of Insect Conservation, 6, 179-189. 

Graham, M.H. (2003) Confronting multicollinearity in ecological multiple regres- 
sion. Ecology, 84, 2809-2815. 

Groom, S.V.C., Stevens, M.I. & Schwarz, M.P. (2014) Parallel responses of bees to 
Pleistocene climate change in three isolated archipelagos of the southwestern Pa- 
cific. Proceedings of the Royal Society B: Biological Sciences, 281, 20133293. 

Guisan, A. & Thuiller, W. (2005) Predicting species distribution: offering more than 
simple habitat models. Ecology Letters, 8, 993-1009. 

Guisan, A. & Zimmermann, N.E. (2000) Predictive habitat distribution models in 
ecology. Ecological Modelling, 135, 147-186. 

Hampe, A. & Petit, R.J. (2005) Conserving biodiversity under climate change: the 
rear edge matters. Ecology Letters, 8, 461-7. 

Hanley, J.A. & McNeil, B.J. (1982) The meaning and use of the area under a receiver 
operating characteristic (ROC) curve. Radiology, 143, 29-36. 

Hanspach, J., Schweiger, O., Kihn, I., Plattner, M., Pearman, P.B., Zimmermann, 
N.E. & Settele, J. (2014) Host plant availability potentially limits butterfly dis- 
tributions under cold environmental conditions. Ecography, 37, 301-308. 

Harrell, EE., Lee, K.L. & Mark, D.B. (1996) Multivariable prognostic models: issues 
in developing models, evaluating assumptions and adequacy, and measuring and 
reducing errors. Statistics in Medicine, 15, 361-87. 

Hatten, I-D., Looney, C., Strange, J.P. & Bosque-Pérez, N.A. (2013) Bumble bee fau- 
na of Palouse Prairie: Survey of native bee pollinators in a fragmented ecosystem. 
Journal of Insect Science, 13, 26. 

Hayes, T’B., Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart, A.A. & Vonk, 
A. (2002) Hermaphroditic, demasculinized frogs after exposure to the herbicide 
atrazine at low ecologically relevant doses. Proceedings of the National Academy of 
Sciences of the United States of America, 99, 5476-80. 

Heath, J. & Leclercq, J. (1969) The European Invertebrates Survey. Preliminairy 
notice - Cartographie des Invertébrés Européens, Notice préliminaire - Erfas- 
sung der Europdischen Wirbellosen, vorldufige Mitteilung, Biological Records 
Centre, Abbots Riplon and Faculté des Sciences agronomiques de Gembloux, 
Belgium. 

Heath, J. & Leclercq, J. (1981) Provisional Atlas of The Invertebrates of Europe. Maps 
1-27, Institute of Terrestrial Ecology, Cambridge (UK) and Faculté des Sciences 
Agronomiques, Gembloux, Belgium. 

Heikkinen, R.K., Luoto, M., Aratjo, M.B., Virkkala, R., Thuiller, W. & Sykes, M.T. 
(2006) Methods and uncertainties in bioclimatic envelope modelling under cli- 
mate change. Progress in Physical Geography, 30, 751-777. 

Heikkinen, R.K., Luoto, M., Virkkala, R. & Rainio, K. (2004) Effects of habitat cover, 
landscape structure and spatial variables on the abundance of birds in an agricul- 

tural-forest mosaic. Journal of Applied Ecology, 41, 824-835. 

185 


186 

Climatic Risk and Distribution Atlas of European Bumblebees 

Heikkinen, R.K., Marmion, M. & Luoto, M. (2012) Does the interpolation accuracy 
of species distribution models come at the expense of transferability? Ecography, 
35, 276-288. 

Heinrich, B. (1979) Bumblebee economics, Harvard University Press, Cambridge, Mas- 
sachusetts. 

Hendry, A.P., Farrugia, T.J. & Kinnison, M.T. (2008) Human influences on rates of 
phenotypic change in wild animal populations. Molecular Ecology, 17, 20-9. 

Herrera, J.M., Ploquin, E.E, Rodriguez-Pérez, J. & Obeso, J.R. (2014) Determining 
habitat suitability for bumblebees in a mountain system: a baseline approach for 
testing the impact of climate change on the occurrence and abundance of species. 
Journal of Biogeography, 41, 700-712. 

Hill, J.K., Thomas, C.D. & Huntley, B. (1999) Climate and habitat availability deter- 
mine 20th century changes in a butterfly’s range margin. Proceedings of the Royal 
Society B: Biological Sciences, 266, 1197-1206. 

Hines, H.M. (2008) Historical biogeography, divergence times, and diversification 
patterns of bumble bees (Hymenoptera: Apidae: Bombus). Systematic Biology, 57, 
58-75. 

Hoffmann, A.A. & Sgré, C.M. (2011) Climate change and evolutionary adaptation. 
Nature, 470, 479-85. 

Hutchinson, G.E. (1957) Concluding remarks. Cold Spring Harbor Symposium on 
Quantitative Biology, 22, 415-427. 

Iler, A.M., Inouye, D.W., Hoye, T-'T., Miller-Rushing, A.J., Burkle, L-A. & Johnston, 
E.B. (2013) Maintenance of temporal synchrony between syrphid flies and floral 
resources despite differential phenological responses to climate. Global Change 
Biology, 19, 2348-59. 

Ings, I-C., Ings, N.L., Chittka, L. & Rasmont, P. (2010) A failed invasion? Commer- 
cially introduced pollinators in Southern France. Apidologie, 41, 1-13. 

Inoue, M.N., Yokoyama, J. & Washitani, I. (2008) Displacement of Japanese native 
bumblebees by the recently introduced Bombus terrestris (L.) (Hymenoptera: Ap- 
idae). Journal of Insect Conservation, 12, 135-146. 

Intoppa, FE, Piazza, M.G., Bolchi Serini, G. & Cornalba, M. (2009) J Bombi; Guida al 
riconoscimento delle specie italiene. CRA-Unita di Ricerca di Apicoltora e Bachic- 
oltura, Bologna, 174 p. 

IPCC (2001) Climate Change 2001: The Scientific Basis. Contribution of Working Group 
I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, 
Cambridge University Press, Cambridge, UK. 

IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working 
Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate 
Change, (ed. by T-E. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. 
Boschung, A. Nauels, Y. Xia, V. Bex, and PM. Midgley) Cambridge University 
Press, Cambridge, United Kingdom and New York. 


References 

Iserbyt, S. & Rasmont, P. (2012) The effect of climatic variation on abundance and 
diversity of bumblebees: A ten years survey in a mountain hotspot. Annales de la 
Société Entomologique de France (N.S.), 48, 261-273. 

Jackson, S.T. & Sax, D.E (2010) Balancing biodiversity in a changing environment: 
extinction debt, immigration credit and species turnover. Trends in Ecology & 
Evolution, 25, 153-60. 

Jeschke, J.M., Bacher, S., Blackburn, T.M., Dick, J.T-A., Essl, E, Evans, T:, Gaertner, 
M., Hulme, PE., Kiihn, I., Mrugata, A., Pergl, J., Pysek, P., Rabitsch, W., Ric- 
ciardi, A., Richardson, D.M., Sendek, A., Vila, M., Winter, M. & Kumschick, 
S. (2014) Defining the impact of non-native species. Conservation Biology, 28, 
1188-1194. 

Johnson, R.M., Dahlgren, L., Siegfried, B.D. & Ellis, M.D. (2013) Acaricide, fungi- 
cide and drug interactions in honey bees (Apis mellifera). PLoS ONE, 8, 54092. 

Johnson, R.M., Ellis, M.D., Mullin, C.A. & Frazier, M. (2010) Pesticides and honey 
bee toxicity — USA. Apidologie, 41, 312-331. 

Kanbe, Y., Okada, I., Yoneda, M., Goka, K. & Tsuchida, K. (2008) Interspecific mat- 
ing of the introduced bumblebee Bombus terrestris and the native Japanese bum- 
blebee Bombus hypocrita sapporoensis results in inviable hybrids. Naturwissen- 
schaften, 95, 1003-1008. 

Kenta, T., Inari, N., Nagamitsu, T., Goka, K. & Hiura, T. (2007) Commercialized 
European bumblebee can cause pollination disturbance: An experiment on seven 
native plant species in Japan. Biological Conservation, 134, 298-309. 

Kharouba, H.M., Paquette, S.R., Kerr, J.T. & Vellend, M. (2014) Predicting the sensi- 
tivity of butterfly phenology to temperature over the past century. Global Change 
Biology, 20, 504-14. 

Kirilenko, A.P. & Hanley, R.S. (2007) Using multiple methods to predict climate change 
impacts on bumblebees in North America. Proceedings of the Third IASTED In- 
ternational Conference on Environmental Modelling and Simulation, EMS 2007, 
42-47. 

Klein, A.M., Vaissi¢re, B.E., Cane, J.H., Steffan-Dewenter, I., Cunningham, S.A, 
Kremen, C. & Tscharntke, T. (2007) Importance of pollinators in changing 
landscapes for world crops. Proceedings of the Royal Society B: Biological Sciences, 
274, 303-313. 

Kosior, A., Celary, W., Olejniczak, P., Fijat, J., Krol, W., Solarz, W. & Plonka, P. (2007) 
The decline of the bumble bees and cuckoo bees (Hymenoptera: Apidae: Bombi- 
ni) of Western and Central Europe. ORYX, 41, 79-88. 

Kraus, FB., Wolf, S. & Moritz, R-EA. (2009) Male flight distance and population 
substructure in the bumblebee Bombus terrestris. Journal of Animal Ecology, 78, 
247-252. 

Krauss, J., Bommarco, R., Guardiola, M., Heikkinen, R.K., Helm, A., Kuussaari, M., 
Lindborg, R., Ockinger, E., Partel, M., Pino, J., Poyry, J., Raatikainen, K.M.., 
Sang, A., Stefanescu, C., Teder, T., Zobel, M. & Steffan-Dewenter, I. (2010) 

187 


188 

Climatic Risk and Distribution Atlas of European Bumblebees 

Habitat fragmentation causes immediate and time-delayed biodiversity loss at 
different trophic levels. Ecology Letters, 13, 597-605. 

Kriechbaumer, J. (1870) Vier neue Hummelarten. Verhaudlungen der Kaiserlich-Konigli- 
chen Zoologischen-Botanischen Gesellschaft in Wien, 20, 157-160. 

Kriiger, E. (1951) Phanoanalytische Studien an einigen Arten der Untergattung Ter- 
restribombus O. Vogt (Hymen. Bomb.). I. Teil. Tijdschrift voor Entomologie, 93, 
141-197. 

Kriiger, E. (1958) Phanoanalytische Studien an einigen Arten der Untergattung Ter- 
restribombus O. Vogt (Hymenoptera, Bombidae). III. Teil. Tijdschrift voor Ento- 
mologie, 101, 283-344. 

Kruseman, G. (1964) Bombus muscorum et ses sous-espéces de la Manche. Entomolo- 
gische Berichten, Amsterdam, 24, 245-247. 

Kudo, G. & Ida, TY. (2013) Early onset of spring increases the phenological mismatch 
between plants and pollinators. Ecology, 94, 2311-2320. 

Kudo, G. (2013) Vulnerability of phenological synchrony between plants and pollina- 
tors in an alpine ecosystem. Ecological Research, 29, 571-581. 

Kudrna, O., Harpke, A., Lux, K., Pennerstorfer, J., Schweiger, O., Settele, J. & 
Wiemers, M. (2011) Distribution Atlas of Butterflies in Europe, Gesellschaft ftir 
Schmetterlingsschutz, Halle, Germany. 

Kuhlmann, M., Ascher, J.S., Dathe, H.H., Ebmer, A.W., Hartmann, P, Michez, D., 
Miiller, A., Patiny, S., Pauly, A., Praz, C., Rasmont, P., Risch, S., Scheuchl, E., 
Schwarz, M., Terzo, M., Williams, PH., Amiet, E, Baldock, D., Berg, W., Bo- 
gusch, P, Calabuig, I., Cederberg, B., Gogala, A., Gusenleitner, F, Josan, Z.., 
Madsen, H.B., Nilsson, A., Odegaard, E, J. Ortiz-Sanchez, J. Paukkunen, T., 
Pawlikowski, M., Quaranta, M., Roberts, S.P.M., Saropataki, M., Schwenninger, 
H.-R., Smit, J., Séderman, G. & Tomozei, B. (2014) Checklist of the Western Pa- 
laearctic Bees (Hymenoptera: Apoidea: Anthophila). http://westpalbees.myspe- 
cies.info. 

Kuhlmann, M., Guo, D., Veldtman, R. & Donaldson, J. (2012) Consequences of 
warming up a hotspot: Species range shifts within a centre of bee diversity. Di- 
versity and Distributions, 18, 885-897. 

Lawson, C.R., Bennie, J., Hodgson, J.A., Thomas, C.D. & Wilson, R.J. (2014) Topo- 
graphic microclimates drive microhabitat associations at the range margin of a 
butterfly. Ecography, 37, 732-740. 

Lecocq, T:, Brasero, N., De Meulemeester, T., Michez, D., Dellicour, S., Lhomme, P, 
de Jonghe, R., Valterova, I., Urbanova, K. & Rasmont, P. (2014) An integrative 
taxonomic approach to assess the status of Corsican bumblebees: implications for 
conservation. Animal Conservation, doi:10.1111/acv.12164. 

Lecocq, T., Dellicour, S., Michez, D., Dehon, M., Dewulf, A., De Meulemeester, T., 
Brasero, N. Valterova, I., Rasplus, J-Y. & Rasmont, P. Methods for species delim- 
itation in bumblebees (Hymenoptera, Apidae, Bombus): towards an integrative 

approach. Zoological Scripta, in press. 


References 

Lecocq, T., Dellicour, S., Michez, D., Lhomme, P., Vanderplanck, M., Valterova, I., 
Rasplus, J.-Y. & Rasmont, P. (2013) Scent of a break-up: phylogeography and 
reproductive trait divergences in the red-tailed bumblebee (Bombus lapidarius). 
BMC Evolutionary Biology, 13, 263. 

Lecocq, T., Lhomme, P., Michez, D., Dellicour, S., Valterova, I., & Rasmont, P. 
(2011). Molecular and chemical charaters to evaluate species status of two cuck- 
oo bumblebees: Bombus barbutellus and Bombus maxillosus (Hymenoptera, Api- 
dae, Bombini). Systematic Entomology 36, 453-469. 

Lecocq, T., Vereecken, N.J., Michez, D., Dellicour, S., Lhomme, P., Valterova, L., 
Rasplus, J.-Y. & Rasmont, P. (2013) Patterns of genetic and reproductive traits 
differentiation in Mainland vs. Corsican populations of bumblebees. PLoS ONE, 
8, e65642. 

Lepais, O., Darvill, B., O’Connor, S., Osborne, J.L., Sanderson, R.A., Cussans, J., 
Goffe, L. & Goulson, D. (2010) Estimation of bumblebee queen dispersal dis- 
tances using sibship reconstruction method. Molecular Ecology, 19, 819-831. 

Lobo, J.M., Jiménez-Valverde, A. & Hortal, J. (2010) The uncertain nature of absences 
and their importance in species distribution modelling. Ecography, 33, 103-114. 

Lobo, J.M., Jiménez-Valverde, A. & Real, R. (2008) AUC: a misleading measure of the 
performance of predictive distribution models. Global Ecology and Biogeography, 
17 145=151). 

Loken, A. (1973) Studies on Scandinavian bumble bees (Hymenoptera, Apidae). Norsk 
Entomologisk Tidsskrift, 20, 1-218. 

Loken, A. (1984) Scandinavian species of the genus Psithyrus Lepeletier (Hymenop- 
tera: Apidae). Entomologica Scandinavia (Supplement), 23, 1-45. 

Lumaret, J.-P. (1986) Toxicité de certains helminthicides vis-a-vis des insectes copro- 
phages et conséquences sur la disparition des excréments de la surface du sol. Acta 
Oecologica/Oecologia Applicata, 7, 313-324. 

Lurgi, M., Lépez, B.C. & Montoya, J.M. (2012) Novel communities from climate 
change. Philosophical Transactions of the Royal Society B: Biological Sciences, 367, 
2913-22. 

Madsen, M., Nielsen, B.O., Holter, P., Pedersen, O.C., Jespersen, J.B., Jensen, K.- 
M.V., Nansen, P. & Gronvold, J. (1990) Treating cattle with ivermectin: Effects 
on the fauna and decomposition of dung pats. The Journal of Applied Ecology, 27, 
1-15. 

Martins, A.C., Gongalves, R.B. & Melo, G.A.R. (2013) Changes in wild bee fauna of 
a grassland in Brazil reveal negative effects associated with growing urbanization 
during the last 40 years. Zoologia (Curitiba), 30, 157-1706. 

Mayer, C., Michez, D., Chyzy, A., Brédat, E. & Jacquemart, A.-L. (2012) The abun- 
dance and pollen foraging behaviour of bumble bees in relation to population 
size of whortleberry (Vaccinium uliginosum). PLoS ONE, 7, e50353. 

Meeus, I., Brown, M.J.E, De Graaf, D.C. & Smagghe, G. (2011) Effects of invasive 
parasites on bumble bee declines. Conservation Biology, 25, 662-671. 

189 


190 

Climatic Risk and Distribution Atlas of European Bumblebees 

Michener, C.D. (2000) The bees of the world, The Johns Hopkins University Press, 
Baltimore, MD. 

Mikkola, K. (1978) Spring migrations of wasps and bumble bees on the southern coast 
of Finland (Hymenoptera, Vespidae and Apidae). Annales Entomologici Fennici, 
44, 10-26. 

Millien, V. (2006) Morphological evolution is accelerated among island mammals. 
PLoS Biology, 4, 1863-1868. 

Mitchell, T.D., Carter, T.R., Jones, PD., Hulme, M. & New, M. (2004) A compre- 
hensive set of high-resolution grids of monthly climate for Europe and the globe: the 
observed record (1901-2000) and 16 scenarios (2001-2100), Tyndall Centre for for 
Climate Change Research, University of East Anglia, Norwich, UK. 

Moo-Llanes, D., Ibarra-Cerdeha, C.N., Rebollar-Téllez, E.A., Ibafez-Bernal, S., 
Gonzalez, C. & Ramsey, J.M. (2013) Current and future niche of North and 
Central American sand flies (Diptera: psychodidae) in climate change scenarios. 
PLoS Neglected Tropical Diseases,'7, e2421. 

Morelli, T-L., Smith, A.B., Kastely, C.R., Mastroserio, I., Moritz, C. & Beissinger, 
S.R. (2012) Anthropogenic refugia ameliorate the severe climate-related decline 
of a montane mammal along its trailing edge. Proceedings of the Royal Society B: 
Biological Sciences, 279, 4279-86. 

Morin, X. & Thuiller, W. (2009) Comparing niche- and process-based models to re- 
duce prediction uncertainty in species range shifts under climate change. Ecology, 
90, 1301-1313. 

Morris, W.E., Pfister, C.A., Tuljapurkar, S., Haridas, C. V., Boggs, C.L., Boyce, M.S., 
Bruna, E.M., Church, D.R., Coulson, T., Doak, D.E, Forsyth, S., Gaillard, J.- 
M.., Horvitz, C.C., Kalisz, S., Kendall, B.E., Knight, T.M., Lee, C.T. & Menges, 
E.S. (2008) Longevity can buffer plant and animal populations against changing 
climatic variability. Ecology, 89, 19-25. 

Miiller, H. & Eriksson, O. (2013) A pragmatic and utilitarian view of species trans- 
location as a tool in conservation biology. Biodiversity and Conservation, 22, 
1837-1841. 

Murray, T\E., Coffey, M.F., Kehoe, E. & Horgan, E.G. (2013) Pathogen prevalence in 
commercially reared bumble bees and evidence of spillover in conspecific popu- 
lations. Biological Conservation, 159, 269-276. 

Mussen, E.C., Lopez, J.E. & Peng, C.Y.S. (2004) Effects of selected fungicides on 
growth and development of larval honey bees, Apis mellifera L. (Hymenoptera: 
Apidae). Environmental Entomology, 33, 1151-1154. 

Nagamitsu, T., Yamagishi, H., Kenta, T., Inari, N. & Kato, E. (2010) Competitive 
effects of the exotic Bombus terrestris on native bumble bees revealed by a field 
removal experiment. Population Ecology, 52, 123-136. 

Neff, J.L. & Simpson, B.B. (1993) Bees, pollination systems and plant diversity. Hyme- 
noptera and Biodiversity (ed. by J. LaSalle and 1.D. Gauld), pp. 143-167. CAB 
International, Wallingford, UK. 


References 

New, M., Hulme, M. & Jones, P. (2000) Representing twentieth-century space-time 
climate variability. Part HI: Development of 1901-96 monthly grids of terrestrial 
surface climate. Journal of Climate, 13, 2217-2238. 

Nukatsuka, Y. & Yokoyama, J. (2010) Environmental factors and land uses related to 
the naturalization of Bombus terrestris in Hokkaido, northern Japan. Biological 
Invasions, 12, 795-804. 

Olden, J.D. & Jackson, D.A. (2000) Torturing data for the sake of generality: How 
valid are our regression models? Ecoscience,’7, 501-510. 

Oliver, T., Hill, J.K., Thomas, C.D., Brereton, T. & Roy, D.B. (2009) Changes in 
habitat specificity of species at their climatic range boundaries. Ecology Letters, 
12, 1091-102. 

Ornosa, C. & Ortiz-Sanchez, FJ. (2004) Hymenoptera: Apoidea I. Introduccién. Fa- 
milias Colletidae, Melittidae y Apidae. Fauna Ibérica, vol. 23., Museo Nacional de 
Ciencias Naturales, CSIC, Madrid. 

Ozbek, H. (1995) The decline of wild bee populations in Turkey. Changes in Fauna of 
Wild Bees in Europe (ed. by J. Banaszak), pp. 33-36. Pedagogical University, 
Bydgoszcz, Poland. 

Panfilov, D.V. (1978) 7. Sem. Apidae - Apidy. Opredelitel’ nasekomykh evropeyskoy chasti 
SSSR. Tom III. Pereponchatokrylye. Pervaya chast’ “. "Nauka’, pp. 508-519. Len- 
ingradskoe Odtelenie, Leningrad. 

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

Pearson, R.G. & Dawson, T:P. (2003) Predicting the impacts of climate change on the 
distribution of species: are bioclimate envelope models useful? Global Ecology and 
Biogeography, 12, 361-371. 

Peirce, C.S. (1884) The numerical measure of the success of predictions. Science, 
Vol. 4, No. 93, 453-454. 

Pekkarinen, A. (1984) Resource partitioning and coexistence in bumblebees (Hyme- 
noptera, Bombinae). Annales Entomologici Fennici, 50, 97-107. 

Peredo-Alvarez, V., Rasmont, P., Scriven, J. J. & Goulson, D. Predicting the global 
spread of an invasive pollinator, the bumblebee Bombus terrestris. Ecology and 
Evolution, in press. 

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

Peters, G. (1972) Ursachen fiir den Riickgang der seltenen heimischen Hummelarten 
(Hym., Bombus et Psithyrus). Entomologische Berichte, Berlin, 9, 85-90. 

Pittioni, B. (1938) Die Hummeln und Schmarotzerhummeln der Balkan-Halbinsel 
mit besonderer Berucksichtigung der Fauna Bulgariens. I: Allgemeiner Teil. /z- 
vestiya na Tsarskite Prirodonauchni Instituti v Sofiya, 11, 12-69. 

Pittioni, B. (1942) Die boreoalpinen Hummeln und Schmarotzerhummeln (Hymen., 
Apidae, Bombinae). I. Teil. Mitteilungen aus den Koninglichen Naturwissenschaft- 
lichen Instituten in Sofia, 15, 155-218. 

191 


192 

Climatic Risk and Distribution Atlas of European Bumblebees 

Pittioni, B. (1943) Die boreoalpinen Hummeln und Schmarotzerhummeln (Hymen., 
Apidae, Bombinae). II. Teil. Mitteilungen aus den Koninglichen Naturwissen- 
schaftlichen Instituten in Sofia, 16, 1-78. 

Ploquin, E.E, Herrera, J.M. & Obeso, J.R. (2013) Bumblebee community homoge- 
nization after uphill shifts in montane areas of northern Spain. Oecologia, 173, 
1649-60. 

Polce, C., Garratt, M.P., Termansen, M., Ramirez-Villegas, J., Challinor, A.J., Lap- 
page, M.G., Boatman, N.D., Crowe, A., Endalew, A.M., Potts, S.G., Somerwill, 
K.E. & Biesmeijer, J.C. (2014) Climate-driven spatial mismatches between Brit- 
ish orchards and their pollinators: increased risks of pollination deficits. Global 
Change Biology, 20, 2815-28. 

Polgar, C.A., Primack, R.B., Williams, E.H., Stichter, S. & Hitchcock, C. (2013) Cli- 
mate effects on the flight period of Lycaenid butterflies in Massachusetts. Biolog- 
ical Conservation, 160, 25—31. 

Potts, S.G., Biesmeijer, J.C., Bommarco, R., Felicioli, A., Fischer, M., Jokinen, P,, 
Kleijn, D., Klein, A.-M., Kunin, W.E., Neumann, P, Penev, L.D., Petanidou, T., 
Rasmont, P., Roberts, S.PM., Smith, H.G., Sorensen, P.B., Steffan-Dewenter, I., 
Vaissiére, B.E., Vila, M., Vujic, A., Woyciechowski, M., Zobel, M., Settele, J. & 
Schweiger, O. (2011) Developing European conservation and mitigation tools 
for pollination services: Approaches of the STEP (Status and Trends of European 
Pollinators) project. Journal of Apicultural Research, 50, 152-1064. 

Pradervand, J.-N., Pellissier, L., Randin, C. & Guisan, A. (2014) Functional homoge- 
nization of bumblebee communities in alpine landscapes under projected climate 
change. Climate Change Responses, 1, 1-10. 

Pr¥s-Jones, O.E. & Corbet, S. (2011) Bumblebees - Naturalists Handbook 6, 3rd edn. 
Pelagic Publishing, London, UK. 

R Development Core Team (2013) R: A language and environment for statistical comput- 
ing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051- 
07-0, URL http://www.R-project.org. 

Radoszkowski, O. (1884) Révision des armures copulatrices des males du genre Bom- 
bus. Byulletin’ Moskovskogo Obshchestva Ispytatelei Prirody, 59, 51-92. 

Randin, C.F, Dirnbock, T., Dullinger, S., Zimmermann, N.E., Zappa, M. & Guisan, 
A. (2006) Are niche-based species distribution models transferable in space? Jour- 
nal of Biogeography, 33, 1689-1703. 

Ranta, E. & Vepsalainen, K. (1981) Why are there so many species? Spatio-temporal 
heterogeneity and northern bumblebee communities. Oikos, 36, 28-34. 

Rasmont, P. & Adamski, A. (1995) Les Bourdons de la Corse (Hymenoptera, Apoidea, 
Bombinae). Notes Fauniques de Gembloux, 31, 1-87. 

Rasmont, P., Roberts, S.P.M., Cederberg, B., Radchenko, V. & Michez, D., (2013) 
Bombus. IUCN 2013. IUCN Red List of Threatened Species. Version 2013.2, 1. 
Downloaded on 01 Januray 2014. http://www.iucnredlist.org. 


References 

Rasmont, P. & Haubruge, E. (2014) Atlas Hymenoptera. Université de Mons, Gem- 
bloux Agro Bio Tech, Mons, Gembloux, Belgium. 

Rasmont, P. & Iserbyt, S. (2012) The Bumblebees Scarcity Syndrome: Are heat waves 
leading to local extinctions of bumblebees (Hymenoptera: Apidae: Bombus). An- 
nales de la Société Entomologique de France (N.S.), 48, 275-280. 

Rasmont, P. & Iserbyt, S. (2014) Atlas of the European bees: genus Bombus. 3rd 
Edition. STEP Project. Status and Trends of European Pollinators, Atlas Hymenop- 
tera, Mons, Gembloux (Belgium). http://www.zoologie.umh.ac.be//hymenoptera/ 
page.asp?ID=169 

Rasmont, P. & Mersch, P. (1988) First estimation of faunistic drift by bumblebees of 
Belgium, (Hymenoptera: Apidae). - Premiere estimation de la derive faunique chez 
les bourdons de la Belgique (Hymenoptera: Apidae). Annales de la Société Royale 
Zoologique de Belgique, 118, 141-147. 

Rasmont, P. (1983a) Catalogue commenté des Bourdons de la région ouest-paléarc- 
tique (Hymenoptera, Apoidea, Apidae). Notes fauniques de Gembloux, 7, 1-72. 

Rasmont, P. (1983b) La notion d’exerge appliquée 4 Megabombus (Thoracobombus) 
pascorum (Scopoli) (Hymenoptera, Apidae). Bulletin annuel de la société royale 
belge d entomologie, 119, 185-195. 

Rasmont, P. (1989) Centre de richesse et centres de pauvreté de la faune des Bourdons de 
France (Hymenoptera, Apidae) - Théorie d’Inouye contre théorie de Ranta & Ves- 
péalainen. LUtilisation des Inventaires informatisés d'Invertébrés pour l'Tdentification 
et la Surveillance d'Espaces de grand Intérét faunistique (ed. by J. de Beaufort and 
H. Maurin), p. 12. Paris, France. 

Rasmont, P. (2008) La régression massive des espéeces dabeilles sauvages et de bourdons 
d Europe : un effet de la perturbation mondiale du cycle de Vazote. ctes du colloque 
Insectes et Biodiversité, 6 octobre 2006, Saint-Léons en Lévézou (France, Aveyron), 
pp. 43-60. Conseil général de l’Aveyron, Rodez, France. 

Rasmont, P., Aytekin, A.M., Valterova, I. & Williams, PH. (2012) The fate of Bombus 
cullumanus: regression and species status. Eurbee5, 5th european conference of api- 
dology, 3-7th September 2012, Halle an der Saale, Germany (ed. by B. Barth, H. 
Scharpenberg, and R.EA. Moritz), p. 185. Halle an der Saale, Germany. 

Rasmont, P., Coppée, A., Michez, D. & De Meulemeester, T. (2008) An overview of 
the Bombus terrestris (L. 1758) subspecies (Hymenoptera: Apidae). Annales de la 
Société Entomologique de France (N.S.), 44, 243-250. 

Rasmont, P., Leclercq, J., Jacob-Remacle, A., Pauly, A. & Gaspar, C. (1993) The faunis- 
tic drift of Apoidea in Belgium. Bees for pollination (ed. by E. Bruneau), pp. 65-87. 
Commission of the European communities, Brussels, Belgium. 

Rasmont, P, Pauly, A., Terzo, M., Patiny, S., Michez, D., Iserbyt, S., Barbier, Y. & 
Haubruge, E. (2005) The survey of wild bees (Hymenoptera, Apoidea) in Belgium 
and France, Food & Agriculture Organization of the United Nations, Rome, 
Italy. 

193 


194 

Climatic Risk and Distribution Atlas of European Bumblebees 

Rasmont, P., Scholl, A., de Jonghe, R., Obrecht, E. & Adamski, A. (1986) Identité 
et variabilité des males de bourdons du genre Bombus Latreille sensu stricto en 
Europe Occidentale et Centrale (Hymenoptera, Apidae, Bombinae). Revue Suisse 
de Zoologie, 93, 661-682. 

Rasmont, P., Terzo, M., Aytekin, A.M., Hines, H., Urbanova, K., Cahlikova, L. 
& Valterova, I. (2005) Cephalic secretions of the bumblebee subgenus Szbiri- 
cobombus Vogt suggest Bombus niveatus Kriechbaumer and Bombus vorticosus 
Gerstaecker are conspecific (Hymenoptera, Apidae, Bombus). Apidologie, 36, 
571-584. 

Reinig, W.E (1935) On the variation of Bombus lapidarius L. and its cuckoo, Psithyrus 
rupestris Fabr., with notes on mimetic similarity. Journal of Genetic, 30, 321-356. 

Reinig, W.E (1937) Die Holarktis. Ein Beitrag zur diluvialen und alluvialen Geschichte 
der zirkumpolaren Faunen- und Florengebiete, Verlag von Gustav Fischer, Jena, 
Germany. 

Reinig, W.E (1939) Die Evolutionsmechanismen erlautert an den Hummeln. Verhan- 
dlungen der Deutschen Zoologischen Gesellschaft (supplement), 170-206. 

Reinig, W.E. (1939) Die genetisch-chorologischen Grundlagen der gerichteten geogra- 
phischenVariabilitat. Zeitschrift fiir Induktive Abstammungs- und Vererbungslehre, 
76, 260-308. 

Richardson, D.M., PySek, P., Rejmanek, M., Barbour, M.G., Panetta, ED. & West, 
C.J. (2000) Naturalization and invasion of alien plants: concepts and definitions. 
Diversity Distributions, 6, 93-107. 

Schweiger, O., Biesmeijer, J.C., Bommarco, R., Hickler, T;, Hulme, PE., Klotz, S., 
Kuhn, I., Moora, M., Nielsen, A., Ohlemiiller, R., Petanidou, T., Potts, S.G., 
PySek, P., Stout, J.C., Sykes, M.T., Tscheulin, T., Vila, M., Walther, G.-R., West- 
phal, C., Winter, M., Zobel, M. & Settele, J. (2010) Multiple stressors on biotic 
interactions: how climate change and alien species interact to affect pollination. 
Biological Reviews of the Cambridge Philosophical Society, 85, 777-95. 

Schweiger, O., Heikkinen, R.K., Harpke, A., Hickler, T., Klotz, S., Kudrna, O., Kithn, 
I., Poyry, J. & Settele, J. (2012) Increasing range mismatching of interacting 
species under global change is related to their ecological characteristics. Global 
Ecology and Biogeography, 21, 88-99. 

Schweiger, O., Settele, J., Kudrna, O., Klotz, S. & Kiihn, I. (2008) Climate change 
can cause spatial mismatch of trophically interacting species. Ecology, 89, 3472- 
3479. 

Sears, J. (2014) Short-haired bumblebee reintroduction. http://www.rspb.org.uk/our- 
work/projects/details.aspx?id=299380 

Settele, J., Hammen, V., Hulme, P, Karlson, U., Klotz, S., Kotarac, M. & AL, E. 
(2005) ALARM: Assessing LArge-scale environmental Risks for biodiversity with 
tested Methods. Gaia-Ecological Perspectives for Science and Society, 14, 69-72. 

Settele, J., Kudrna, O., Harpke, A., Kihn, I., Swaay, C. van, Verovnik, R., M., W., 
Wiemers, M., Hanspach, J., Hickler, T., Kihn, E., Halder, I. van, K, V., Vliegen- 


References 

thart, A., Wynhoff, I. & Schweiger, O. (2008) Climatic Risk Atlas of European 
Butterflies, Biorisk 1 (special issue). BioRisk, 10, 1-710. 

Simon-Delso, N., San Martin, G., Bruneau, E., Minsart, L.-A., Mouret, C. & Hauti- 
er, L. (2014) Honeybee colony disorder in crop areas: the role of pesticides and 
viruses. PLoS ONE, 9, e103073. 

Skorikov, A.S. (1922) Shmeli palearktiki. Chast I. Obshchaya biologiya (so vklyuche- 
niem zoogeografii). Les bourdons de la faune paléarctique. Partie I. Biologie gé- 
nérale (la zoogéographie y compris). lzvestiya Severnoi Oblastnoi Stantsii Zashchi- 
ty Rastenii ot Vreditelei, 4, 1-160. 

Soberon, J. & Peterson, A.T. (2005) Interpretation of models of fundamental ecologi- 
cal niches and species’ distributional areas. Biodiversity Informatics, 2. 

Spangenberg, J.H., Bondeau, A., Carter, T.R., Fronzek, S., Jaeger, J., Jylha, K., Kihn, 
I., Omann, I., Paul, A., Reginster, I., Rounsevell, M., Schweiger, O., Stocker, A., 
Sykes, M.T. & Settele, J. (2012) Scenarios for investigating risks to biodiversity. 
Global Ecology and Biogeography, 21, 5-18. 

Stewart, J.R. & Lister, A.M. (2001) Cryptic northern refugia and the origins of the 
modern biota. Trends in Ecology and Evolution, 16, 608-613. 

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

Svenning, J.-C. & Skov, E (2007) Could the tree diversity pattern in Europe be gener- 
ated by postglacial dispersal limitation? Ecology Letters, 10, 453-60. 

Svensson, B. G. (1979) Pyrobombus lapponicus auct., in Europe recognized as two spe- 
cies: P lapponicus (Fabricius, 1793) and P monticola (Smith, 1849) (Hymenop- 
tera, Apoidea, Bombinae). Entomologica Scandinavica, 10, 275-296. 

Terzo, M. & Rasmont, P. (2003) Xylocopa cantabrita Lepeletier en France (Hymenop- 
tera, Apoidea). Bulletin de la Société Entomologique de France, 108(5), 441-445. 

Thomas, C.D. (2011) Translocation of species, climate change, and the end of trying to 
recreate past ecological communities. Trends in Ecology & Evolution, 26, 216-21. 

Thomas, C.D., Bodsworth, E.J., Wilson, R.J., Simmons, A.D., Davies, Z.G., Musche, 
M. & Conradt, L. (2001) Ecological and evolutionary processes at expanding 
range margins. Nature, 411, 577-81. 

Thomas, J.A., Telfer, M.G., Roy, D.B., Preston, C.D., Greenwood, J.J.D., Asher, J., 
Fox, R., Clarke, R.T. & Lawton, J.H. (2004) Comparative losses of British but- 
terflies, birds, and plants and the global extinction crisis. Science, 303, 1879-81. 

Thuiller, W. (2004) Patterns and uncertainties of species’ range shifts under climate 
change. Global Change Biology, 10, 2020-2027. 

Thuiller, W., Aratjo, M.B. & Lavorel, S. (2003) Generalized models vs. classification 
tree analysis: Predicting spatial distributions of plant species at different scales. 
Journal of Vegetation Science, 14, 669-680. 

Thuiller, W., Araujo, M.B. & Lavorel, S. (2004) Do we need land-cover data to model 
species distributions in Europe? Journal of Biogeography, 31, 353-3061. 

195 


196 

Climatic Risk and Distribution Atlas of European Bumblebees 

Tilman, D., May, R.M., Lehman, C.L. & Nowak, M.A. (1994) Habitat destruction 
and the extinction debt. Nature, 371, 65—G6. 

Tingley, M.W., Monahan, W.B., Beissinger, S.R. & Moritz, C. (2009) Birds track their 
Grinnellian niche through a century of climate change. Proceedings of the Nation- 
al Academy of Sciences of the United States of America, 106 Suppl., 19637-19643. 

Torretta, J.P, Medan, D. & Arahamovich, A.H. (2006) First record of the invasive 
bumblebee Bombus terrestris (L.) (Hymenoptera, Apidae) in Argentina. Transac- 
tions of the American Entomological Society, 132, 285-289. 

Tutin, T.G., Heywood, V.H., Burges, N.A., Valentine, D.H., Walters, S.M. & Webb, 
D.A. (2001) Flora Europaea, Cambridge University Press, Cambridge, UK. 
Vandame, R. & Belzunces, L.P. (1998) Joint actions of deltamethrin and azole fungi- 

cides on honey bee thermoregulation. Neuroscience Letters, 251, 57-60. 

Velthuis, HLH.W. & van Doorn, A. (2006) A century of advances in bumblebee do- 
mestication and the economic and environmental aspects of its commercializa- 
tion for pollination. Apidologie, 37, 421-451. 

Velthuis, H.H.W. (2002) The historical background of the domestication of the bum- 
ble-bee, Bombus terrestris, and its introduction in agriculture. Pollinating Bees - The 
conservation link between agriculture and nature (ed. by P. Kevan and V.L. Imper- 
atriz Fonseca), Ministry of Environment, Sao Paulo, Brasil. 

Vila, B.,Vennetier, M., Ripert, C., Chandioux, O., Liang, E., Guibal, E & Torre, 
F (2008). Has global change induced divergent trends in radial growth of Pi- 
nus sylvestris and Pinus halepensis at their bioclimatic limit? The example of the 
Sainte-Baume forest (south-east France). Annals of Forest Science, 65(709), 1-9. 

Vila, M. & Hulme, PE. (2011) Jurassic Park? No thanks. Trends in Ecology & Evolu- 
tion, 26, 496-7. 

Vogt, O. (1909) Studien tiber das Artproblem. 1. Mitteilung. itiber das Variieren der 
Hummeln. I. Teil. Schriften der Berlinischen Gesellschaft Naturforschender, Freun- 
de, Berlin, 1909, 28-84. 

Vogt, O. (1911) Studien iiber das Artproblem. 2. Mitteilung. iiber das Variieren 
der Hummeln. 2. Teil (Schluss). Schriften der Berlinischen Gesellschaft Natur- 
forschender, Freunde, Berlin, 1911, 31-74. 

Webber, B.L., Scott, J.K. & Didham, R.K. (2011) Translocation or bust! A new ac- 
climatization agenda for the 21st century? Trends in Ecology & Evolution, 26, 
495-496; author reply 497-498. 

West-Eberhard, M.J. (1983) Sexual selection, social competition, and speciation. 
Quarterly Review of Biology, 58, 155-183. 

Whitehorn, P.R., O’Connor, S., Wackers, EL. & Goulson, D. (2012) Neonicotinoid 
pesticide reduces bumble bee colony growth and queen production. Science, 336, 
351-352. 

Williams PH. (1989) Why are there so many species of bumble bees at Dungeness? 
Botanical Journal of the Linnean Society, 101, 31-44. 


References 

Williams, PH. & Osborne, J.L. (2009) Bumblebee vulnerability and conservation 
world-wide. Apidologie, 40, 367-387. 

Williams, PH. (1982) The distribution and decline of British bumble bees (Bombus 
Latr.). Journal of Apicultural Research, 21, 236-245. 

Williams, PH. (1988) Habitat use by bumble bees (Bombus spp.). Ecological Entomol- 
ogy, 13, 223-237. 

Williams, PH. (1998) An annotated checklist of bumble bees with an analysis of pat- 
terns of description (Hymenoptera: Apidae, Bombini). Bulletin of The Natural 
History Museum (Entomology), 67, 79-152. 

Williams, PH., An, J. & Huang, J. (2011) The bumblebees of the subgenus Sub- 
terraneobombus: Integrating evidence from morphology and DNA barcodes 
(Hymenoptera, Apidae, Bombus). Zoological Journal of the Linnean Society, 163, 
813-862. 

Williams, PH., An, J., Brown, M.J.E, Carolan, J.C., Goulson, D., Huang, J. & Ito, 
M. (2012) Cryptic bumblebee species: consequences for conservation and the 
trade in greenhouse pollinators. PLoS ONE, 7, e32992. 

Williams, PH., Aratjo, M.B. & Rasmont, P. (2007) Can vulnerability among British 
bumblebee (Bombus) species be explained by niche position and breadth? Biolog- 
ical Conservation, 138, 493-505. 

Williams, PH., Brown, M.J.E, Carolan, J.C., An, J., Goulson, D., Aytekin, A.M., 
Best, L.R., Byvaltsev, A.M., Cederberg, B., Dawson, R., Huang, J., Ito, M., 
Monfared, A., Raina, R.H., Schmid-Hempel, P.,, Sheffield, C.S., Sima, P. & Xie, 
Z. (2012) Unveiling cryptic species of the bumblebee subgenus Bombus s. str. 
worldwide with COI barcodes (Hymenoptera: Apidae). Systematics and Biodi- 
versity, 10, 21-56. 

Williams, P.H., Byvaltsev, A., Shefheld, C. & Rasmont, P. (2013) Bombus cullumanus 
- An extinct European bumblebee species? Apidologie, 44, 121-132. 

Williams, PH., Cameron, S.A., Hines, H.M., Cederberg, B. & Rasmont, P. (2008) A 
simplified subgeneric classification of the bumblebees (genus Bombus). Apidolo- 
gie, 39, 46-74. 

Williams, PH., Colla, S. & Xie, Z., (2009) Bumblebee vulnerability: Common cor- 
relates of winners and losers across three continents. Conservation Biology, 23, 
931-940. 

Williams, PH., Humphries, C.J. & Vane-Wright, R.I. (1991) Measuring biodiversity: 
taxonomic relatedness for conservation priorities. Australian Systematic Botany, 
4, 665-679. 

Williams, S.E., Shoo, L.P, Isaac, J.L., Hoffmann, A.A. & Langham, G. (2008) To- 
wards an integrated framework for assessing the vulnerability of species to cli- 
mate change. PLoS Biology, 6, 2621-2626. 

Willis, K.J. & Bhagwat, S.A. (2009) Ecology. Biodiversity and climate change. Science, 
326, 806-807. 

197 


198 

Climatic Risk and Distribution Atlas of European Bumblebees 

Willis, S.G., Hole, D.G. & Huntley, B. (2010) Climate change and conservation, (ed. 
by Leader-Williams, N, Adams, WM & Smith, RJ (Eds) (2010) Tiade-offs in 
Conservation: deciding what to save. Oxford: Wiley-Blackwells, Chichester, UK. 

Winter, K., Adams, L., Thorp, R., Inouye, D., Day, L., Ascher, J. & Buchmann, S. 
(2006) Importation of non-native bumble bees into North America: potential conse- 
quences of using Bombus terrestris and other non-native bumble bees for greenhouse 
crop pollination in Canada, Mexico, and the United States,. Ed. Laurie Adams, San 
Francisco, 35 p. 

Xie, Z., Williams, PH. & Tang, Y. (2008) The effect of grazing on bumblebees in the 
high rangelands of the eastern Tibetan Plateau of Sichuan. Journal of Insect Con- 
servation, 12, 695-703. 

Yokoyama, J. & Inoue, M.N. (2010) Status of the invasion and range expansion of an 
introduced bumblebee, Bombus terrestris (L.), in Japan. Applied Entomology and 
Zoology, 45, 21-27. 

Yoon, H.J., Sohn, M.R., Choo, Y.M., Li, J., Sohn, H.D. & Jin, B.R. (2009) Defensin 
gene sequences of three different bumblebees, Bombus spp. Journal of Asia-Pacific 
Entomology, 12, 27-31. 

Zarevucka, M. (2013) Insecticide Resistance of Bumblebee Species. Biodegradation - Life 
of Science (ed. by R. Chamy and E Rosenkranz), pp. 207-228. Intech, Rijeka, 

Croatia. 


Appendices 

16. Appendices 

Appendix 1. Measures of model performance. AUC, area under the curve; TSS, true 
skill statistic. Empty cells for non-modelled species. 

Species Range | Modelled TSS | Sensi- 
size (50 | range size 
km grid) | (10' grid) 

86 

132 

Bombus alpinus 

ot Ke 
SS 
So 
— . 
00 
Qo 

Bombus argillaceus 10590 

Bombus armeniacus 

o;]o 
WW | CO 
NLD 

jen) 
STS 
NI | \O 
Oo eRe 

S 
ON 
ro 
° 
00 
No) 

a 
No) 
aS 

120 
D2 
933 

4245 
1110 
yoy: 

Bombus balteatus 

fn) 
= 
Oo 
— 
jo) 
aN 
— 

Bombus barbutellus 
Bombus bohemicus 
Bombus brodmanni 

Bombus brodmannicus 

se et loa 
en NID 
N AlN 

N 

So 
™N 
Oo 
\9 
So 
aN 
aN 

670 11067 : 
192 5694 7. 
Bombus confusus 157 7946 87 
ae. 3536 97 
326 12298 Fo 
8 4215 wal 

Bombus campestris 

— 

= 3 
O IX ag 
Q — HY o Pd 
a 
aa 
aN 
\9 
= 
WN 
— 
a as ‘ 
\O IN CON | \o 
SJIIN KRISTA 

= 
= 
N 
S 
= 
No) 
S 
, 
No) 
Oo 

Bombus cingulatus 

o 
= 
SS 
oO 
= 
ON 
tw 
oo 
nn 
= 
N 
ore 

Bombus consobrinus 

= 

oS 

on 

=) 

=) 

00 
olo|lolto 

00 

ox 
ay] Poa 
QD] 
BS |X 

om) 

atl Pecaaal (Bs 

— | 

HX | \O 

Se eS 

NJ Go 

Qo J GO 
; So |S 
(oe) CO TON 
\9 \9 J \o 

=) 

(oe) 

aN 

olo]o 
\o 
_—_ 
oS 
Co 
™ 

Bombus cryptarum 

an) 

jo) 
— 
5] | 
ON 

o|slo 
nN 
WW Fo 
jm) 
| 
\9 

Bombus cullumanus 

N 

Bombus deuteronymus 

= : 
io 
— 

502 
221 
36 
52 
34 
1331 12958 
Bombus humilis 707 1218 
33 
961 

14470 8 

ADD 94 
5530 94 
3324 wD 
99 
wh 

wD 

Bombus distinguendus 
Bombus flavidus 
Bombus fragrans 

cael Peart : 
NW 
co} 

o;loO 
Nn 

sod lose Rs 
00 | oo 
No) 

Bombus gerstaeckeri 

an) 
) 
me FRO TR 
ON 
So 
eal esecaal fs 
\S | \O JO 
NY AR AS 

j) 
jn) 
~ 
(oe) 
= 
(oe) 
— 

Bombus haematurus 

Bombus hortorum 

oS 
Ni 
S 
ey) 
co |N 
al Road lease fe 
aN 
S 
olo|lo 
ON 
N 
oo) 
N 
ey) 

Oo 
an) 

Sead ea 2 
&» | & 
Oo | AX 

= 

SS) 

No) 
OTN 

aN 

oS 

ON 

WN 

\9 
\9 
— 
i) 
jon) 
See | ee 
WN 
ioe) 
jo) 
\9 
OoyrRejy]o 
i) 
jo) 
i) 
\9 
\9 

ce 
\o 
6 

Bombus hyperboreus 

° 
No) 
oa cael bs 

00 

ON 
© 
No) 
SX 

Sills 
O 
N 

So 
A 
Nn 
So 
io 

16632 .80 
i) 
1791 98 
1354 90 

Bombus hypnorum 

io 

Bombus incertus 

Ot 
el 
00 
alts 
No) 
Oo 

Bombus inexspectatus 

So 

Bombus jonellus 
Bombus laesus 
14644 yy 
98 

— — 
N 

= : 
ON 
No) 

Oo oO 
oy ON J 
N Ne ea 
= = 
—~ oN 
N 
= SealPos allt 
NI CO | \O 
co Wn J BS 

oO 

oo 

oO 

So 
\9 

1233 
Bombus lapponicus 206 5976 

Bombus lapidarius 

° 
° 
oN 
ON 
° 
00 
oO 
oe 
No) 
A 
ro 
No) 
S 

Bombus cullumanus | 3 

199 


200 = Climatic Risk and Distribution Atlas of European Bumblebees 

335 
[Bombus mlokosievitaii_| 5 | | 
; —__} __ 

80 | 0. 
93 [041 [077 | 088 
85 | 0.32 | 0.58 | 0.80_| 
2 
5 
-ise9_[o78 [029 [0.43 | 079 
o_| 13260_| 0.78 | 0.36 | 0.45 | 0.85 _| 
| 37_| 5446 | 0.96 | 0.16 | 0.82 | 0.97 _| 
| 61 | 8366 | 0.90 | 0.13 | 0.67 | 0.92 | 
. 


201 

Appendices 

(%99-) 
Z118- 

(%89-) 

8cyC- 
(%F6-) 
ZLES- 

na 
Odds -WVd 
[essod [essod 
-sIp ou | -sIp ou 

(%1¥-) 
COFT- 

Jesiod 
“SIP [PY 

[essod 
“SIP [PY 

(%F6") 
87ES- 

(%¥6-) 
€L6E- 

(%ZTI-) 
S7PEI 

Jessod 
“SIP [IY 

(%FE-) 
€S6l- 

(%I1¥-) 
I€ZI- 

(%64-) 
0L0C- 

na 
Odds -WVd 
[essed [essod [essod 
-sIp ou | -sIp ou | -sip ou 

(%I¥-) 
O€ZI- 

[essod 
“STP [IY 

(%9F-) 
78Y 

Jessod 
“SIP JY 

(%IS-) 
GSES 

(%SE-) 
OLET- 

[essod 
“SIP [JPY 

snupunyns snquog 
wnsmgdas snquog 
snulAgosuos snguog 
snsnfuos snquiog 
snqvynsu19 snquog 
stuqsaq@ups snquLog 
snaimayog snqulog 
smpaqngavg snquiog 
snqvayvg snquog 
SNaIDITIDAV SNQULOg 
snuidjy snquog 

saqoadg 

OOLZ Puke OSOT 10} UONduINsse [essodsrp ou pur [Ny pue 
SOLIBUDIS IBULYD IICUTT] DdIYI JOPUN SUOTITPUOD SNeUTT]D dTqeIMS YITM s]]I9 prs Jo Jaquinu ay) ul sasueyD porwaloig *Z xipueddy 


Climatic Risk and Distribution Atlas of European Bumblebees 

202 

(%8¥-) 
60S9- 
(%09-) 
OLOI- 

(%89-) 
GLEE 

(%S9-) 
T6€8- 

(%¥9-) 
VEIC 

(%9E-) 
OSey 

(%9F-) 
6009- 

(%9V-) 
6869- 

[essod 

“STP TTY 

(%7¥-) 
Cyl 

(%Z8-) 
OFE- 

(%9S-) 
8€08- 

[essod 

“STP TITY 

(%VE-) 
OS9F- 

(%PE-) 
Abs 

(%06-) 
6S¢- 

(%99°) 
086Z- 

(%PE-) 
8LEV- 

(%9¥-) 
SISI- 

(%¥7-) 
FI8I- 

(%¥9-) 
S76" 

[essed 

“STP TIS 

9696- 
Odds 
[essod 
-sIp ou 

(%I1¥-) 
697S- 

(%FE-) 
8Y9V- 

(%FI-) 
8S7- 

(%0Z-) 

(%9I1-) 
ZL6I- 

(%¥C-) 

VI8I- 

Jesiod 
“SIP TTY 

[essod 
“SIP [TAY 

(%9Z-) 
ZEIE- 

[essod 
“SIP [TAY 

smpjauol snquog 
snqvjaadsxaul snquog 
snqaaoul snguiog 
wnsoudty snquog 
snasogsadly snquog 
sytuny snguog 
WNLOJOY snguOg 

SNANJDULIVY SnqQuog 

1494 IAVISAAB SnqQuog 

suplspef snguiog 
snpianyf snquog 
snpuansuysip sngquog 

satoadg 


203 

Appendices 

(%9¥-) 
0€79- 

(%¥-) 
CYYC 

(%C¥-) 
Once- 

(%FS-) 
YC6L- 

[essod 

-SIp ou 

Ue 

(%bF-) 
916S- 

(%99-) 
7168" 

(%¥8-) 
18¢r- 

(%19-) 
Crly- 
(%99-) 
COVI- 

(%9I-) 
89TI- 

“SEP TIPS 

(%9F-) 
8779- 

(%bF-) 
€CFI- 

(%9E-) 
78L7- 

(%9¥-) 
CL6- 

[essod 
“SIP TY 

(%99-) 
O1S8- 

Jessod 
“SIP TITY 

Odds 

[essed 

-SIp ou 

(%9E-) 
O9L- 

“WV! 

qd 
Jessod 
-sIp ou 

(%04-) 
CyIE- 

(%PF-) 
S199 

re) 
[essed 
-sIp ou 

%FyI-) 
C60I- 

[essod 
“STP TTY 

%¥I-) 
CC8I- 

Jesiad 
“SIP [JPY 

(%9T-) 
5661- 

(%S-) 
609% 

Py 

Jessod 
“SIP [PY 

wnsonasvd snquog 
SNIIBANLOU SNQULO 
snqpaaiu snquog 
UNAOISNUL SNQULOg 
SnpLInUL snquLog 
DIOIIUOUL sNQULOg 

SPJIULOSIUL snquiog 

— 
aati 

uinsoony snguiog 
sniuoddyy snguog 

snippidy] snquog 

aie 


Climatic Risk and Distribution Atlas of European Bumblebees 

204 

(%99-) | (%¥8-) 
6698- | 6YITT- 
(%67-) 
99S: 

(%bF-) 
COSE- 

Odds 

[essod [essod [essod 
-sIp ou -sIp ou | -sip [Ny 

(%79-) 
OLI8- 

(%09-) 
FO80I- 

[essod 
“SIP [IY 

(%L9-) 
9TLY- 

Fa 

[ested 
“SIP JAF 

Odds 

[essod 

-SIp ou 

(%07-) 
LYDG- 

(%8¥-) 
PLEE- 

(%L¥-) 
60%¢- 

SVu 
[essed 

-sIp ou 

Jesiod 
~SIP [PY 

[essod 
“SIP TPF 

[essod 
“SIP TY 

Iayais snquiog 
smpjataouauas snqulog 
1yIUALYIS SNQULOT 
stagsadnas sngquog 
snqmuapnas snguog 
SNIADLIPNA SNQULO 
Lojonapunbh snquog 
snapuashd snquog 
wnsoqvd snquiog 
wnsouod snquog 
stavjod snquiog 

satoadg 


205 

Appendices 

(%I¥-) 
HCET- 

(%bF-) 
06¢E- 

Odds 
Jessod 

-SIp ou 

[essod 
-sIp ou | -sIp ou 

(%F8-) 
9SLY- 

(%¥rE-) 
TS8¢- 
(%9Z-) 
E19ZI- 

(%¥9-) 
6S90I- 

(%I¥-) 
LLI€- 

(%9-) (%I¥-) 
CHOL- Cys 

“STP TTY 

(%I1¥-) 
FECT 

(%99-) 
9TOL- 

(%9S-) 
I1L6- 

(%S9-) 
OF80I- 

(%9¥-) 
08%S- 

[essod [essed 

“SIP TTS | -SEP IIMS 

(%¥T-) 
960F- 

Odds 
[essod 

-STp ou 

-SIp ou 

(%8¥-) 
9697 

(%¥I-) 
SLFT- 
(%¥T-) 
7L6E- 

(%FE-) | (%9C-) 
080% | ¥cle- 

—— Jesiod Jessod [essod 
“SIp ou | -SIpP [TAF | -SIP TTY | -SIP ITY 

snqpuoz snguiog 
muafinm snquog 
snupsajaa snquiog 
sypysan snquog 
Stlqsadday snguiog 
stagsands snguog 
wnavajds snquog 
SnNaUplsaggns snguLog 
snoippsods snquog 
sisuaao4os snqULOg 

satoodg 


206 

Climatic Risk and Distribution Atlas of European Bumblebees 

Appendix 3. Species characteristics and projected changes (in percent of modelled 
number of occupied 10 min grid cells) under three climate change scenarios and full 
and no dispersal assumption for 2050 and 2100. 

=) 

(1) a} ® [6 7) | 3) | @) | Go) 
ae [sR r 
esrConam | P| 33 [200 | o56 [ms [2 
SS [idee |? bo |e fa [a | 
Lest Conem | 8 ARERR 570 [ ae [ar [0 
Last Concem | 1 | 806 | oars | 2000 | a2 | a25 

est Coneen [| 1 | 69 [2006s | 5687 [ 1064 
ieee 
B 

IUCN status 2014 

Taxonomical issues 
STI: Species Temperature 
Index (°C) 

Number of records in the 
database across the study area 
Number of records in the 
database for 1970 - 2000 
Range size (number of 50km 
grid units) 

Range size (number of 50km 
grid units) 1970 - 2000 

5 
5 
Qa 
= 
f 
oe. 
fa) 
“—~ 

Ee am) FC 

patos Ps [as fo fan  P 
> oe 
[vote |v | 76 | 2ese [aa | 66 [57 
pesGwes Pe [aa fre [on [oo [a 
ee 
ively Eadangeed | [71 | 78 | 105 | 0 | 28 
a 
Lew Conca) ARN 582 [729 [2 [on 
ee a ER 
Ralnenble | BER 81 | se0 |e |e 
eee CON Se 
Lew Concern | P| 02 [s0es0] eres | 547 | 331 
Lew Concern | -F | a1 [2000s] sooo | 951 | 707 
er 
[Lew Concen |S A 
er re 
a ° ee 
Lew Concem |S RRR 1866 5909 [i020 | 03 
<_ gSe a E 

& 

B. flavidus 

B 
B. 
B 

brodmanni 

brodmannicus 

consobrinus 

deuteronymus 

distinguendus 
. gerstaeckeri 


Appendices 207 

Full dispersal No dispersal 

Most realistic dispersal 
assumption 

7D) GFP) -88.06 3733) F754) -88.06 
ee Ee Ee a es 

en 92.23 -93.59 DGS) 92.49 -93.8 

59.84 -67.05 002 Fe) 26.0 
52.94 -64.55 By lez awies A 
a Se 
es a — 
EENEN 54.12 -58.51 -67.05 68.97 -83.05 -89.27 [J 
EEPSM 5701) 89.45 -93.57 YEN 58108) -90.29 -94.34 
Siete MaGe 68.44 -95.37 -98.25 
80.4  -81.96 43.38 SYA No 
Zea. | ete oe 65.96 |-71.7_ -30.34 
5018) 53:71 |268:45 7751 BARRY 5075/5882 0727) 84.06 -93.52 
WEEE 55.55 -63.94 166.73 |-7936 -37.13 BR 

oO 
oO 
oO 
oO 

PN 
al 
Pea 

93.55 80.31 69.42 -80.55 [EM 

-46.37 Owl 54.58 |-64.76 -70.58 

G58 eGae RENen silos) 7996) -87.2 
5051 PRIA S408) -81.03 -94.44 -98.25 (5058 BARR S408) -81.03 -94.44 -98.25 
| Full | 

WEREN 54.15 -60.83 -70.74 WYAaa 55.71 6237 -7223 EM 

-86.72 78.7 73.68) -89.97  -93.23 -93.48 -100 -100 

51,26 RIPEN 59.74 67.84 -79.12 
| Fall | 

[Ee [ee A (SF (a |” ESF 

N 
N 
N 
N 


208 ~—=_ Climatic Risk and Distribution Atlas of European Bumblebees 

IUCN status 2014 

Taxonomical issues 
STI: Species Temperature 
Index (°C) 
Number of records in the 
atabase across the study area 
Number of records in the 
database for 1970 - 2000 
Range size (number of 50km 
grid units) 
Range size (number of 50km 
grid units) 1970 - 2000 

~~ 
0 
2359 

110 

Least Concern 
Least Concern 
Least Concern 
Least Concern 
Near Threatened 
Least Concern 
Data Deficient 
Endangered 
Least Concern : 
Near Threatened 

Vulnerable 

Least Concern 
Least Concern 
Least Concern 
Data Deficient 
x 
x 

Least Concern 
Least Concern 
Least Concern 
Vulnerable 

Least Concern 
Least Concern 
Least Concern 

Endangered 

Not Evaluated 
Least Concern 
x 

Least Concern 

Least Concern 

Least Concern 
Least Concern 

Least Concern 

a 
[ [ease Concer 
| [ease Concer 
LX [Lease Concern 
|X [Lease Concern 
[ [Near Threatened 
| [ease Concern 
[| [Daca Deficient 
PX [Endangered 
| [hease Concer 
| [Near‘Threatened 
PX [Vulnerable 
|X [Least Concer 
| [heast Concern 
| [east Concern 
| [Dac Deficient 
|X [Least Concern 
|X [Least Concern 
| [east Concern 
| [Nalnerable 
| |east Concern 
| [heast Concern 
| |east Concern 
PX [Endangered 
PX [No Braluated 
[| |heast Comoe 
|X [Least Concen 
| |heast Concem 
[ [east Concem 
| [heast Concern 
| [east Concern 
| [heast Conoen 
| |east Concern 

S 

7.8 

9 


Appendices 209 

Full dispersal No dispersal 

Most realistic dispersal 
assumption 

=a) a) =a) 
23) 

BHBY 84.52 -86.63 5949) 84.54 -86.75 

Pep B) 69.62 WEE 54.04 -65.78 -70.69 EM 
BEM 51.26 54.26 -61.88 34.78 [iets 
54.71 40.25 eeew) 

5708/6159 |-7238 57.83 -62.88 

56.18 -59.02 -68.75 -71.85 PERRY 51.33 -59.38 -70.94 -80.43 

WON 51.15 -64.38 Bec eae 

a CS I 
SESS] EEE EE] eS esi 
ne SS eS eS SS ee 
57107 5275) We) -87.23 -96.86 -98.68 MSAO7)SA7S) NCIS” -87.23 -96.86 -98.68 
50.74 BEREM -75.39  -66.64 61.99 74.95 -96.88 -98.23 
51.69 -59.89  -73 RM 52.21 -60.06 -73.01 
"2187 51.14 66.94 9759) -84.15 
=i Sy a | a a ar ae a 
SSS te a 

yo Pecen bons. F218 -87.92 -94.02 

RPT 53.05 -59.64 69.42 -77.91 

50.54 161.61 072.63 516" -84.08 -92.19 

50.11 Be ay 88.4 -98.37 -98.81 
58145) 20535) 7447 7454 ARYA S164) 20783) 84.46 -96.34 -96.96 
56.97 54.82 |-66.67 |-72.52 
65.31 |-79.26 ee -81.62 


210 = Climatic Risk and Distribution Atlas of European Bumblebees 

ae 

IUCN status 2014 

Taxonomical issues 
database for 1970 - 2000 

Number of records in the 
database across the study area 
Number of records in the 
grid units) 1970 - 2000 

STI: Species Temperature 
Index (°C) 
Range size (number of 50km 

Ries 1 
eared 
Ditton 5 on 
eco — 
eel 
pia 
es. 
i 

(2) See chapter 7 

(3) See chapter 12 

(4) See Rasmont et al. (2014) 

(5) H=Honey-comb maker, I=Inquiline, P=Pocket-maker, S=Pollen-storer, Vs=Unknown 

(6) STI (Species Temperature Index, Devictor et al. 2008); *= indicative value to assess potential response of 
non-modelled species; **= not computed; STI under 5°C are in red; See chapter 13. 

(7-8) Number of specimens in the modelling frame (latitude from 35° to 72°N; longitude from -12°W to 32°E), 
30.XII.2014 

(11) to (22) See chapter 6 

(23) See chapter 8 


Full dispersal No dispersal 

Appendices 

Most dispersal 
assumption 

20) 23) 

-54.52 -66.57 -72.06 

50.58 | -73.95 -92.08 -96.04 
-80.97  -89.76 
57.18 -63.98 -75.7 

PIB 34.03 
| as 

-52.87 

51.79 -57.23 

SSS SS fe na EE aE 

Bombus portschinsky is a large species endemic to east 
Turkey, north Iran and Caucasian region, where it lives 
mainly at forest-edges of subalpine level. It forages mainly 
flowers with long corolla, like Aconitum spp. or the endemic 
Lallemantia canescens. Its colour closely recalls the very 
common Bombus hortorum but here with greyish bands 
instead of yellow. Photo P. Rasmont. 

Bombus saltuarius is a species that only lives in Europe 
in the north-east of Russia: the Pechora basin and the Ural 
mountains where it seems extremely rare. It could be found 
here and there in Siberia, Mongolia and north China. Nothing 
is known about its way of life. Photo P. Rasmont. 

211 


212 

Climatic Risk and Distribution Atlas of European Bumblebees 

17. Distribution maps of West-Palaearctic bumblebees 

Bombus caucasicus is a species endemic to mountain 
forests of Caucasian region. It has been very recently 
resurrected to the species status. Photo P. Rasmont. 

Bombus velox is a small species endemic to east Turkey 
and Caucasian region, where it is very rare, with a patchy 
distribution. Its thorax is grey with a large dorsal black band. 
Its abdomen is yellowish. Photo P. Rasmont. 

Bombus handlirschianus is a medium-sized species that 
lives in the highest mountain levels in east Turkey, north Iran 
and Caucasian region. Two different colour patterns could be 
found, with grey or yellowish bands and a reddish abdomen 
tip. Thanks to its long tongue, it forages mainly flowers with 
long corolla, like Astragalus spp. Photo P. Rasmont. 

Bombus melanurus is a very large high mountain species 
that occurs in the whole Central Asia. To the west, it reaches 
Caucasus and eastern Turkey. Photo P. Rasmont. 

Vo. oo Fol - 
< re rs eee } 

~ 

| 

Bombus persicus is an endemic species of mountains 
steppes in east-Turkey, north Iran and Caucasian region. 
Photo P. Rasmont. 

Bombus sulfureus is a large species endemic to mountain 
steppes in east Turkey and Iran. It shows a very conspicuous 
colour pattern, bright yellow with a black thoracic band and 
a reddish tergite 6. Beside that it is a very rare species, the 
males fly extremely fast and are therefore rarely observed. 
Photo P. Rasmont. 


Distribution maps of West-Palaearctic bumblebees 

meee: 

= 

Bombus alpinus, 1424 specimens , o* 

We 

=* 

Bombus armeniacus, 2088/specimens 

213 


214 

Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus barbutellus, 6095,specimens ) is eg 

ee ee 
Z o 
eo 
tee 

oe te 

om 


215 

Distribution maps of West-Palaearctic bumblebees 

= 

cimens 

spe 

619 

Bombus brodmannicus,, 

x 

a! 

x 

Bombus campestris, 10454 specimens 

t 

f 

Ee 
‘speginie 

2s 
= 

"el 

Bombus cingulatus, 1048 

imens 

Bombus, confusus, 2793 


216 

Climatic Risk and Distribution Atlas of European Bumblebees 

= _ 

f aad 
‘ t >" ‘ ‘ey = " 
= = : f t a } i = a s a bt $ ee 
Bombus cryptarum, 8391 specimens i ¢ : bt ea oe i Pe 
IP 3) oa tome a te ae a oh 
— =, =] 


217 

Distribution maps of West-Palaearctic bumblebees 

€ 
cimens 

ee 

63,5 pe 

Bombus deuteronym Us, I 

ee 

ken 
Mi 

oie 
pecimens 

* 

Bombus distinguendus, 9500's 

So Jo 

= 
3 
Fc 
j 

3 speci 

Ge 

+ 

imens™ 
r 

Bombus flavidus, 178: : 


218 = Climatic Risk and Distribution Atlas of European Bumblebees 

specimens 

Bombus gerstaeckeri, 1783 specimens 
Fl LS 

ames incertis*2163 specimens) \ * Bombus inexspectatus; 2 P-speciniens 


Distribution maps of West-Palaearctic bumblebees 

Bombus humilis, 21 37/5,specimens 

— 

. tel 
fees. 3 
Bombus hyperboreus, 370%specimens a, ; 

219 


Climatic Risk and Distribution Atlas of European Bumblebees 

220 

== 

ecimens 

12 

Bombus hypnorum, 25902 s 

S 

ime 

Bombus jonellus, 19738) spec 

a 

ki 

ecimens 

ae 

Bombus laesus, 777 sp 

+ 
== 


Distribution maps of West-Palaearctic bumblebees 

p or = 4 f 

Bombus lapidarjuss 97808) specimens | 

7 

- 

\ 

Bombus lapponicus, 571 HeGimens 
a 

= 

221 


222 

Climatic Risk and Distribution Atlas of European Bumblebees 

oz 

Bomb 

Fy 
ib 

Bombus melanurus, 30/7 specimens Lbnt mlokosievitzii, 334;specimens 

# 

== 

Borabons mesomelas, 6054 

mens 

Bombus mendax; 


Distribution maps of West-Palaearctic bumblebees 223 

Bombus monticola, 10550 specimens 

Bombus mucidus, 


224 ~—_ Climatic Risk and Distribution Atlas of European Bumblebees 

| 
v5, . 

ee 
¢ 

i a7 
Bombus muscorum, 12938 Specimens ‘ ‘ ; 

, pad 
4 7 Oe. eames 9 
‘\Bombus pataciatus,3 specimens 
Woe I? & ( 31 R wie 


Distribution maps of West-Palaearctic bumblebees 225 

ee ee 
Talal 

‘oa 
af F aoe t 3 : : 7 - 5 ye 
ri o 7 ; 3 2 2 Ss we? ° = 
Bombus pascuorum, 16457 2*specimens . tite 
a = = mame 
Leg Ay ) af A 

e 7 
4 yes \ f 

Bovibus pereziellus, 215 specimens a 

iy nf 
- | Bombus portchinsky, AI) specimens: ne 

Bombus pyrenaeus, 71 76,specimens 


Climatic Risk and Distribution Atlas of European Bumblebees 

226 

4 

ecimens 

Bombus pomorum, 3997 sp 

e a 3, ager i a F 
La ae | rp “a =| 
an ay ee Oy: 
@ - \ : 4. Pe 
o hg eo a eT 
i = i, } a \ r 
‘ ; a r os" —s 
he Z B Mow oe" o a e \ a. / = Me 
ae Tere” | aan “ , D 
0 o = ae 

== 

a DN 
SI ss 
a 8 
i) ®, 
(D) : Cul 
(Sy | on 
Ss Co) 
Mt oe 
SI 2) 
oO cl 
S 
& Ly 
S g 
S : 
S BS 
S S 
=| = 
ha Q SS 
r] SS i SS 
4 S S 
Sa) Sa) 


227 

Distribution maps of West-Palaearctic bumblebees 

ee: 
2 
: 7° 
2 
ee rsh ee sa 
Bombus reinigiellus, 5/)specimens 
oe r ne Bee Leas 
ah 
a é 
e 
2 m4 
\ 
oO rem 
ae 
oO ( = 
Lee 
\, 
(ex 
mn eee | 
"i | " . 
tal 
bee = on | 
oS 

\ 

\ 

‘ 
specimens 

Bombus ruderarius, 23896, 

Bombus ruderatus, JO94s spechmans 


228 = Climatic Risk and Distribution Atlas of European Bumblebees 

— re 

Li Fs — f 
; | iC 
i 
i! 
a u \y 
ie me 

=" - a ea 
7. <a > 
= ( 
¥/. i } i : —— i a i" bd c My 
34 specimen t Ne + a 
Bombus rupestris, 1093 specimens ' ye etc ee 

Bombus schrencki, 2861 specimens 

_ 

a 

Bombus semenoviellus, 435 specim 

@ 


Distribution maps of West-Palaearctic bumblebees 229 

yt ey, 
Bese 
es 
| Bombus sulfureus, 89 specimens 
14 — ic ee ee | T "ea 

f Fan 

=o 

2 - af 
Bombus sichelit, 7329 specimens | 4 : 


Climatic Risk and Distribution Atlas of European Bumblebees 

230 

ENS, 

703*sp. 

Bombus sporadicus, 2 

= 

ecim 
——— 

= 

\ 

ecimens 

Pas, 

Bombus subterraneus, 7P84 sp 

== 

aa 

Bombus sylvarum, 22775 specimens 


‘fe fe f of 

Bombus sylbestris, 13257 specimens 

Bombus terrestris} 104290.specimens 

Distribution maps of West-Palaearctic bumblebees 

“Ss 
ae 

a 

C 

231 


232 Climatic Risk and Distribution Atlas of European Bumblebees 

Bombus vestalis, 8461 specimens 

E= 

a) 

= 

S, 2255 specimens 


Distribution maps of West-Palaearctic bumblebees 

only a 
Coe 
cro 

{ 

Bombus wurflenii, 25866 specimens) ~ 
Let 4 a 

eal 

og Ff | 
Bombus zonafus 1148 specimens ‘i "i 


234 

Climatic Risk and Distribution Atlas of European Bumblebees 

18. Summary 

P. Rasmont, M. Franzén, T. Lecocq, A. Harpke, S.P.M. Roberts, 

J.C. Biesmeijer, L. Castro, B. Cederberg, L. Dvorak, U. FitzPatrick, 

Y. Gonseth, E. Haubruge, G. Mahé, A. Manino, D. Michez, J. Neumayer, 
FE. Odegaard, J. Paukkunen, T. Pawlikowski, S.G. Potts, M. Reemer, 

J. Straka, J. Settele, O. Schweiger. 2015. 

Climatic Risk and Distribution Atlas of European Bumblebees. 
Pensoft publishing, Sofia. 

Thanks to the EU FP7 project STEP (Potts et a/. 2011), over one million bumblebee 
records from all over Europe have been collated. Based on data from 1970 to 2000 
we modelled the current climatic niche for almost all European species (56 out of 69) 
and projected future climatically suitable conditions based on three climate change 
scenarios (SEDG, BAMBU and GRAS) for the years 2050 and 2100. Due to limited 
knowledge of actual bumblebee dispersal, we made two extreme assumptions: (i) the 
species has full dispersal abilities (meaning that the species is able to spread all over 
its suitable area) or (ii) the species is unable to disperse at all (i.e. that changes in 
climatic conditions can only lead to projected range retractions). However, to aid the 
assessment as to which of these two extreme assumptions are more likely to meet 
reality, we also provide a rough indication of the species’ potential dispersal ability 

based on the ecology of the different bumblebees. 

Since bumblebees are mainly adapted to colder conditions, they appear as highly 
vulnerable to climate change. In 2100, depending on the scenario of climate change, 
up to 36% of the European bumblebees are projected to be at an high climatic risk (i.e. 
losing more than 80% of their current range), 41% will be at risk (loss between 50% 
and 80%). Non-modelled species are all very rare and localised and their ranges are 
most likely to be shrinking considerably under all of the scenarios. Only a few species 
are projected to benefit from climate change and can potentially enlarge their current 
distributions in Europe, such as B. argillaceus and B. haematurus. 

As expected, the three scenarios considerably differed in their projections for 2100. While 
under the moderate change scenario (SEDG) only five species are projected to be at 
the verge of extinction by 2100, twenty species are at particularly high risk under the 
intermediate change scenario (BAMBU). Under the most severe change scenario (GRAS) 
as many as 34 species are projected to lose almost all of their climatically suitable area. 

Also the ability to keep track with climate change has a considerable impact of the 
projected changes. For instance, under the most severe climate change scenario (GRAS) 
nine species are at an extremely high climatic risk when full dispersal is assumed for all 
of them. However, under the assumption of no dispersal within the next 100 years 34 


Summary 

species would fall into this category. When potential dispersal abilities, inferred from 
species traits and their auto-ecologies, are considered to decide for which species no 
or full dispersal assumptions are more realistic, it seems that only three might expand 
their ranges by 2100, no species is likely to remain at the status quo, and 25 species 
would be at an extremely high climatic risk. 

We also found that for many species (about 30%), especially the cold-adapted ones in 
Alpine and Arctic regions (e.g. B. alpinus, B. balteatus, B. hyperboreus and B. polaris) 
their dispersal abilities are actually irrelevant for the assessment of their future fate 
because climate change will only lead to reductions of areas with suitable climatic 
conditions while no extra suitable regions will emerge. 

Given the great sensitivity of bumblebees to climate change and further considering 
the severe projected changes in the light of the great relevance of bumblebees as 
pollinators, designing management plans to sustain the highest level of pollination 
services on the one hand and to ensure the survival of as many bumblebee species on 
the other hand is of utmost importance. Given the different mechanisms leading to 
change, especially at the leading versus the trailing edge of species distributions and 
the geographical differences in the severity of climate change, management actions 
must be well and target-specific designed. One important issue would be to prioritise 
management actions across different geographic regions in Europe. We have seen that 
the expected species loss due to climate change increases with decreasing latitudes, i.e. 
that regions in the south of Europe will be most affected by pollinator loss. 

Important means to support European bumblebees would be to facilitate the movement 
of species trying to keep track with changing climates at the trailing edge and to prolong 
the persistence in regions where climatic conditions are deteriorating. Landscape 
management can be of particular help in this context. Increased connectivity and 
quality of bumblebee habitats can help colonising species, while habitat heterogeneity 
will generate heterogeneity in the microclimate and can thus increase population 
persistence at the trailing edge as a kind of “Noah's Ark”. Areas with naturally high 
levels of microclimatic heterogeneity (such as mountainous areas) can be of particular 
importance and deserve special attention. Finally, the idea of assisted migration, i.e. 
purposeful anthropogenic translocations, seems appealing at first sight for species 
whose original distributional areas are projected to shrink tremendously and cannot 
move to suitable areas because of natural or anthropogenic barriers. However, the 
feasibility of such actions is still questionable. 

To conclude, climatic risks for bumblebees can be extremely high, depending on the 
future development of human society, and the corresponding effects on the climate, 
strong mitigation strategies are needed to preserve this important species group and 
to ensure the sustainable provision of pollination services, to which they considerably 
contribute. 

235 


236 

Climatic Risk and Distribution Atlas of European Bumblebees 

List of authors 

RasMonrt Pierre, Laboratoire de Zoologie, Institut Biosciences, Université de Mons, B-7000 

Mons, Belgium; pierre.rasmont@umons.ac.be 

FRANZEN Markus, Helmholtz-Zentrum fir Umweltforschung GmbH - UFZ, Theodor-Lie- 
ser-Strasse 4 / 06120 Halle / Germany; markus.franzen@ufz.de, markus.franzen@ 
biol.lu.se 

Lecocg Thomas, Laboratoire de Zoologie, Institut Biosciences, Université de Mons, B-7000 
Mons, Belgium; thomas.lecocq@umons.ac.be 

Harpxe Alexander, Helmholtz-Zentrum fiir Umweltforschung GmbH - UFZ, Theodor-Lie- 
ser-Strasse 4 / 06120 Halle / Germany; alexander. harpke@ufz.de 

Roserts Stuart P.M., Visiting Research Fellow; University of Reading, PO Box 237, Read- 
ing,RG6 GAR, UK; stuart.roberts@cantab.net 

BIESMEJJER Jacobus, Naturalis Biodiversity Center, postbus 9517, 2300 RA, Leiden, The 
Netherlands; koos.biesmeijer@naturalis.nl 

Castro Leopoldo, LE.S. Vega del Turia, C/ Victor Pruneda 1, E-44001 Teruel, Spain; discoe- 
lius@discoelius.jazztel.es 

CEDERBERG Bjorn, Swedish Species Information Centre, Swedish University of Agricultural 
Sciences, PO Box 7007, 750 07 Uppsala, Sweden; bjornceder@gmail.com 

Dvor&x Libor, Municipal museum Marianské Lazné, Goethovo namésti 11, 35301 
Marianské Lazné, Czech Republic; lib.dvorak@seznam.cz 

Frrzpatrick Una, National Biodiversity Data Centre, Carriganore, Waterford, Eire; ufitzpat- 
rick@biodiversityireland.ie 

GONSETH Yves, Centre Suisse de Cartographie de la Faune (CSCF), Passage Maximilien de 
Meuron 6, CH-2000 Neuchatel, Switzerland; yves.gonseth@unine.ch 

Havsruce Eric, Université de Li¢ge - Gembloux Agro-Bio Tech, Unité d’Entomologie 
Fonctionnelle et Evolutive, Passage des Déportés, 2, B-5030 Gembloux Belgium; 
e.haubruge@ulg.ac.be 

Mane Gilles, 320 chemin du velin F-44420 Mesquer, France; gilles.mahe.fr44@gmail.com 

Manno Aulo, Dipartimento di Scienze Agrarie, Forestali e Alimentari, Universita di Torino, 
Largo Paolo Braccini 2, I-10095 Grugliasco TO, Italy; aulo.manino@unito. it 

Micuez Denis, Laboratoire de Zoologie, Institut Biosciences, Université de Mons, B-7000 
Mons, Belgium; denis.michez@umons.ac.be 

Neumayer Johann, ObergrubstrafSe 18, 5161 Elixhausen, Austria; jneumayer@aon.at 

(ODEGAARD Frode, Norwegian Institute for Nature Research - NINA, Postal address: PO.Box 
5685 Sluppen, NO-7485 Trondheim, Norway; frode.odegaard@nina.no 

PAUKKUNEN Juho, Finnish Museum of Natural History, Zoology Unit, P.O. Box 17, FI- 
00014 University of Helsinki, Finland; juho.paukkunen@helsinki.f 

PawLikowski Tadeusz, Chair of Ecology and Biogeoraphy, Nicolaus Copernicus University, 
Lwowska 1, 87-100 Torun, Poland; pawlik@biol.uni.torun.pl 

Potts Simon G, Centre for Agri-Environmental Research, School of Agriculture Policy and 
Development, Reading, University, Reading, RG6 6AR, UK; s.g.potts@reading.ac.uk 

REEMER Menno, European Invertebrate Survey - the Nederlands, p/a Naturalis Biodiversity 
Center, Postbus 9517, 2300 RA Leiden, The Netherlands; menno.reemer@naturalis.nl 

SETTELE Josef, Helmholtz-Zentrum ftir Umweltforschung GmbH - UFZ, Theodor-Lie- 
ser-Strasse 4 / 06120 Halle, Germany; josef.settele@ufz.de 

StRAKA Jakub, Department of Zoology, Faculty of Science, Charles University in Prague, 
Vinicna 7, 128 44 Praha 2, Czech Republic; jakub.straka@aculeataresearch.com 

SCHWEIGER Oliver, Helmholtz-Zentrum ftir Umweltforschung GmbH - UFZ, Theodor-Lie- 
ser-Strasse 4 / 06120 Halle, Germany; oliver.schweiger@ufz.de