Consequences of warming up a hotspot: species range shifts within a centre of bee diversity

BIODIVERSITYRESEARCH

Consequences of warming up a hotspot:species range shifts within a centre of beediversity

Michael Kuhlmann1*, Danni Guo2, Ruan Veldtman3,4 and John Donaldson3

INTRODUCTION

About 25,000 bee species globally (Michener, 2007) are

the major pollinators of flowering plants and of more than

half of the 3000 crop species (Klein et al., 2007). As such, they

are essential ecological keystone species contributing to

the integrity of most terrestrial ecosystems. Numerous studies

have demonstrated their economic value to agricultural (Klein

et al., 2007; Allsopp et al., 2008) and to natural ecosystems

(Kearns et al., 1998). Aizen et al. (2009) show that pollination

shortage will intensify demand for agricultural land, a trend

that will be more pronounced in the developing world

potentially leading to decreased food security and increasing

pressure on supply of agricultural land.

South Africa and especially the Cape Floristic Region (CFR)

is known as a centre of bee (Kuhlmann, 2009) and plant

1Department of Entomology, The Natural

History Museum, Cromwell Road, London,

UK SW7 5BD, 2Climate Change and Bio-

Adaptation Division and 3Applied Biodiversity

Research, South African National Biodiversity

Institute, Private Bag X7, Claremont 7735,

South Africa, 4Department of Conservation

Ecology and Entomology, Stellenbosch

University, Stellenbosch, Private Bag X1,

Matieland 7602, South Africa

*Correspondence: Michael Kuhlmann,

Department of Entomology, The Natural

History Museum, London, UK.

E-mail: [email protected]

ABSTRACT

Aim Bees are the most important pollinators of flowering plants and essential

ecological keystone species contributing to the integrity of most terrestrial

ecosystems. Here, we examine the potential impact of climate change on bees’

geographic range in a global biodiversity hotspot.

Location South Africa with a focus on the Cape Floristic Region (CFR) diversity

hotspot.

Methods Geographic ranges of 12 South African bee species representing

dominant distribution types were studied, and the climate change impacts upon

bees were examined with A2 and B2 climate scenarios of HadCM3 model, using

MaxEnt for species distribution modelling.

Results The predicted levels of climate change-induced impacts on species ranges

varied from little shifts and range expansion of 5–50% for two species to

substantial range contractions between 32% and 99% in another six species. Four

species show considerable range shifts. Bees of the winter rainfall area in the west

of South Africa generally have smaller range sizes than in the summer rainfall area

and generally show eastward range contractions toward the dry interior. Bee

species prevalent in summer rainfall regions show a tendency for a south-easterly

shift in geographic range.

Main conclusions The bee fauna of the CFR is identified as the most vulnerable

to climate change due to the high level of endemism, the small range sizes and the

island-like isolation of the Mediterranean-type climate region at the SW tip of

Africa. For monitoring climate change impact on bees, we suggest to establish

observatories in the coastal plains of the west coast that are predicted to be worst

affected and areas where persistence of populations is most likely. Likely impacts

of climate change on life history traits of bees (phenology, sociality, bee-host plant

synchronization) are discussed but require further investigation.

Keywords

Bees, climate sensitivity, future climate scenario, geographic range shift,

pollination, South Africa.

Diversity and Distributions, (Diversity Distrib.) (2012) 1–13

DOI: 10.1111/j.1472-4642.2011.00877.xª 2012 Blackwell Publishing Ltd http://wileyonlinelibrary.com/journal/ddi 1

A J

ourn

al o

f Co

nser

vati

on B

ioge

ogra

phy

Div

ersi

ty a

nd D

istr

ibut

ions

diversity. Because of high levels of endemism, the CFR is of

global importance (Myers et al., 2000), and pollinators are

believed to have played an important role in plant speciation

especially (van der Niet & Johnson, 2009). Recent studies have

both highlighted the importance of a geographic mosaic

of pollinator availability for plant diversification (Anderson

& Johnson, 2008) and shown the vulnerability of the CFR

to climate change with predicted high risks of large-scale plant

extinctions and transformation of the vegetation structure of

vast areas (Rutherford et al., 1999; Erasmus et al., 2002;

Midgley et al., 2002, 2003; Yates et al., 2010a). The extant

areas of the Fynbos, Nama- and Succulent Karoo biomes areas

are predicted to decrease markedly and transform into an

unknown vegetation type (Midgley et al., 2003; Botes et al.,

2006) with between 0.3% and 42.4% of the plant species

becoming extinct within protected areas of these biomes

(Rutherford et al., 1999).

Climate change is likely to have a significant impact on bee

populations as well, because of both changes in temperature

and general meteorological conditions (Brown & Paxton,

2009). Nevertheless, despite the enormous economic and

ecological importance of bees as pollinators, we are currently

aware of only two studies investigating the interaction between

bee species range changes and climate. Williams et al. (2007)

assessed the vulnerability of three bumble bee (Bombus) species

to extinction in Britain by studying climatic niches that

determine their ranges at a regional scale. In a study spanning

across Europe, Roberts et al. (2011) investigated the impact of

climate change on three generalist and three specialist Colletes

species. They found that both specialist and generalist species

showed an increased risk of extinction with shifting climate,

but this was particularly high for the most specialized species.

Warburton et al. (2005) reports a general warming in parts

of the South African winter rainfall area and the southern

section of the early summer rainfall area, both known as bee

diversity centres (Kuhlmann, 2009). A statistically significant

increase of mean annual daily minimum and maximum

temperatures as well as summer means of daily maximum

temperatures was observed in both regions (Warburton et al.,

2005). Now that climate change is a reality for South Africa

and other biodiversity hotspots globally, we see the urgent need

to investigate its impact on bees as the most important group

of pollinators to understand the potential consequences for

pollination as a vital ecosystem function and service.

In this study, we modelled the present (1960–2000) climate

suitability and explored the potential impact of projected

future (2080–2100) climate change upon the geographic

distributions of 12 South African solitary bee species. Sensitive

areas were identified where bee populations should be

monitored to detect range contractions at an early stage.

Finally, we discuss the potential impacts of climate change on

other life history traits of bees such as seasonal emergence.

METHODS

Selection of bee species and origin of distribution

data

From the SouthABees database (M. Kuhlmann, unpublished

data), we selected 12 ground nesting bee species with 18–100

distribution records (Table 1) occurring in various parts of the

Table 1 Bee species used in the study and their distribution types

Bee species (N records) Distribution Source

Amegilla atrocincta (Lepeletier) (N = 55) Similar to Anthophora vestita but centred a bit more to the

west, Namibia, Botswana, Zimbabwe

Eardley (1994)

Anthophora vestita Smith (N = 61) Widespread in the early to mid-summer rainfall area and

southern Cape region with rain all year, Zimbabwe,

Mozambique

Eardley & Brooks (1989)

Colletes antecessus Cockerell (N = 43) Endemic of greater Cape region M. Kuhlmann

(unpublished data)

Colletes marleyi Cockerell (N = 90) Restricted to the early summer rainfall area, Uganda,

Tanzania, Zambia, Malawi, Zimbabwe, Mozambique

M. Kuhlmann

(unpublished data)

Fidelia braunsiana Friese (N = 27) Rare endemic of winter rainfall area and southern Cape

region with rain all year, Namibia

Whitehead & Eardley (2003)

Melitta arrogans (Smith) (N = 94) Distribution centered in winter rainfall area and area with

rain all year, SW Namibia

Eardley & Kuhlmann (2006)

Nomia (Leuconomia) candida Smith (N = 46) Restricted to the early summer rainfall area and directly

adjacent regions, Zimbabwe

Pauly (2000)

Rediviva aurata Whitehead & Steiner (N = 23) Narrow endemic of western part of Fynbos region Whitehead & Steiner (2001)

Rediviva intermixta (Cockerell) (N = 100) Endemic of southern winter rainfall area Whitehead & Steiner (2001)

Rediviva rufipes (Friese) (=bicava Wh. &

Steiner) (N = 80)

More common endemic of winter rainfall area and

southern Cape region with rain all year

Whitehead & Steiner (2001)

Rediviva rufocincta (Cockerell) (N = 24) Endemic of the greater Drakensberg region Whitehead et al. (2008)

Scrapter chloris Eardley (N = 18) Endemic of Succulent Karoo region, SW Namibia Eardley (1996)

M. Kuhlmann et al.

2 Diversity and Distributions, 1–13, ª 2012 Blackwell Publishing Ltd

winter and early to mid-summer rainfall areas (Fig. 1) and

representing the most important South African bee distribu-

tion types (Kuhlmann, 2009). Stem, wood and surface-nesters

form only a minor element of the South African bee fauna and

have been little studied. Hence, they have been excluded from

this data set. The present study is focused on South Africa,

Lesotho and Swaziland. Distribution records of the seven

species also occurring outside this region (Table 1) are omitted

because of the lack of suitable data for modelling. For small

and inconspicuous organisms like most bee species, little

distribution data are available for most parts of Africa. This is

especially a problem with range restricted and endemic species

in the west of South Africa for which relatively few records are

available. However, these species are particularly interesting

with respect to the impact of climate change. For this reason,

we gathered all available data from published sources and

museum collections and used all records irrespective of their

age. The oldest bee records are from 1884, and the most recent

ones are from 2008 with 74.8% of all data collected after 1980.

We have assumed that all records are representative of the

current distribution following the approach of Williams et al.

(2007).

Distribution modelling to determine risk under

climate change

The bee species’ potential future geographic ranges in South

Africa were modelled using MaxEnt (Phillips et al., 2004,

2006). MaxEnt applies Bayesian methods to estimate the

potential geographic distribution of species by finding the

probability distribution of maximum entropy and is an

effective method for modelling species distributions from

presence-only data (Elith et al., 2006). The programme can

also work with a small number of samples and with records

that have not been collected as a part of systematic biological

surveys, which is particularly useful for processing the bee data

presented here that are based on museum collections data

(Elith et al., 2011). Sampling bias that is common in

geographic records based on museum collections because

collectors usually favoured the most accessible areas (e.g. along

roads, nature reserves) can also be addressed in MaxEnt

(Phillips et al., 2009) making it a suitable tool for the present

study. We used MaxEnt to first relate current environmental

conditions to occurrence data for the 12 bee species and

subsequently made spatial predictions for two climate change

scenarios.

The climate and environmental data used were from the

South African Atlas of Climatology and Agrohydrology

(Schulze & Kruger, 2007; Schulze & Maharaj, 2007), the

Worldclim database (Hijmans et al., 2005) and from the

National Land-cover Project 2006. To examine the possible

future climate change impacts on bee species’ geographic

distribution and range shifts in South Africa, projected future

(2080–2100) climate data were used with the present (1960–

2000) climate data, both with a resolution of 1 km · 1 km.

The HadCM3 (Hadley Centre climate model) general circu-

lation model (GCM) or global climate model was used,

because it is a fairly conservative model preferred by South

African scientists as a GCM. No information about predicted

future land use is available for South Africa and thus could not

be included in the model despite its presumably high impact

on bee distribution.

To explore how sensitive the bee species are to climate

changes, the A2 and B2 climate scenarios were used to allow

between model comparisons. The A2 climate scenario assumes

self-reliance, preserving local identities, with moderate regional

Figure 1 Rainfall seasonality in South

Africa, Lesotho and Swaziland (Schulze &

Maharaj, 2007).

Range shifts in a centre of bee diversity

Diversity and Distributions, 1–13, ª 2012 Blackwell Publishing Ltd 3

economic development, but with high increase in population,

and higher CO2 emissions (Carter et al., 2007). The B2 climate

scenario assumes environmental preservation, with local

solutions to economic, social, environmental sustainability,

with slower increase in global population, and less CO2

emissions (Carter et al., 2007). These two scenarios were not

used as absolute possible scenarios for the future, but more as a

tool to explore the sensitivity of bee species to climate changes

in the future, whether it is a natural fluctuation in the climate

or an artificial disturbance.

The present and future geographic distributions and the

seasonal temperature and rainfall variables which influence

their geographic ranges were examined for 12 bee species

(Table 2). In South Africa, seasonality in rainfall plays

an important role in the nature and distribution of plant

and animal species, with in the case for bee species, winter

and summer seasonal variables being of particular importance.

The climatic variables used were: summer minimum temper-

ature, summer maximum temperature, winter minimum

temperature, winter maximum temperature, summer mean

temperature, winter mean temperature, summer rainfall and

winter rainfall (see also McLeish et al., 2011). The environ-

mental variables used for this study act as constraints on the

prediction of distribution of the bees and are comprised of:

altitude (elevation above sea level in metres), terrain mor-

phology (plains, hills, mountains, pans, etc.), rainfall season-

ality (winter rainfall, summer rainfall, etc.) and land use

(urban, cultivated, natural, etc.). Fire has been excluded as a

variable because it is relevant only for a small part of South

Africa and its frequency is predominantly altered by human

activities than by climatic factors. Each MaxEnt model

generates the percentage contribution for every predictor

variable used (climate and environmental variables) (Table 2)

with percentage contribution being estimated from the

iterations of the training algorithm by adding or subtracting

regularized gain to the variable in question.

Modelling was performed using default parameters including

regularization multiplier = 1 and maximum iterations = 500,

but random test percentage was set to 20% because of a low

number of samples for some of the bee species. The MaxEnt

model provides test calculations of model performance via area

under curve (AUC) of the training and test data for all bee

species (Heikkinen et al., 2006): Amegilla atrocincta

(AUC = 0.896), Anthophora vestita (AUC = 0.917), Colletes

antecessus (AUC = 0.987), Colletes marleyi (AUC = 0.956),

Fidelia braunsiana (AUC = 0.986), Leuconomia candida

(AUC = 0.956), Melitta arrogans (AUC = 0.953), Rediviva

aurata (AUC = 0.998), Rediviva intermixta (AUC = 0.990),

Rediviva rufipes (AUC = 0.975), Rediviva rufocincta

(AUC = 0.976), Scrapter chloris (AUC = 0.994). The average

AUC was 0.9653 and ranged from 0.896 to 0.998 which is high

indicating an excellent predictive ability of the models (e.g.

Yates et al., 2010b).

In this study, a habitat is considered suitable for a bee

species if the probability of persistence is ‡ 0.3, while regions

with a probability < 0.3 are assumed to be unlikely for long-

term survival. MaxEnt produces two graphs of omission versus

predicted area and sensitivity versus specificity and calculates

model performance which were all used to decide on

the persistence threshold for the bee species. Here, we used

a threshold of 0.3 that maximized both training sensitivity and

specificity under the current climate (Liu et al., 2005). While

there were some extreme outliers of bee samples not covered

by the predicted geographic range, this could be due to

mislabelling or misinterpretation of the collecting site of a bee

species, especially considering these bee records go back several

decades (Guo, 2010).

RESULTS

Modelling predicts that climate change will have a substantial

impact on the geographic ranges of all 12 studied bee species

regardless of the climate scenario (Table 3; Figs 2–5).

The contribution of the different variables to the bee species

distribution (Table 2) shows that rainfall seasonality is the

main factor in predicting the geographic range of the bee

Table 2 Contribution of variables for the present distribution of bee species

Species Alt. Landuse Terrain

Rain

season

Sum. max.

temp.

Sum. min.

temp.

Sum.

rain.

Sum.

mean temp.

Win. max.

temp.

Win. min.

temp.

Win.

rain.

Win. mean

temp.

Amegilla atrocincta 7.81 36.01 25.18 5.19 4.20 0.00 7.89 3.67 1.69 5.60 2.25 0.51

Anthophora vestita 0.77 44.08 6.71 11.88 0.45 2.99 4.30 9.56 0.14 2.40 16.72 0.00

Colletes antecessus 0.65 1.46 2.57 33.67 1.86 0.19 2.72 0.07 0.00 0.55 56.25 0.00

Colletes marleyi 1.45 20.08 13.79 8.53 0.30 0.76 47.31 5.31 0.36 0.72 1.26 0.13

Fidelia braunsiana 3.04 2.20 12.74 66.91 0.10 0.00 2.53 0.13 0.00 6.07 6.27 0.00

Leuconomia candida 0.06 25.31 28.25 2.10 0.09 1.11 31.22 1.17 1.35 6.18 0.02 3.15

Melitta arrogans 0.83 1.76 10.26 45.52 0.04 0.53 34.36 2.25 2.55 0.12 1.65 0.12

Rediviva aurata 4.65 3.07 29.88 47.99 1.68 0.72 5.10 0.06 0.12 0.00 6.74 0.00

Rediviva intermixta 1.77 1.22 13.04 26.42 2.33 0.36 25.62 0.00 1.45 0.92 26.84 0.04

Rediviva rufipes 0.31 0.93 7.73 43.19 0.38 0.11 8.47 0.27 1.46 0.57 36.29 0.27

Rediviva rufocincta 0.89 2.57 22.75 2.81 8.45 0.00 31.84 11.56 0.00 4.94 14.12 0.05

Scrapter chloris 5.02 0.00 7.20 64.32 7.88 0.00 11.89 0.00 1.17 0.00 2.52 0.00

Text in bold indicates dominant variables for each species.

M. Kuhlmann et al.


species, while altitude and landscape morphology are con-

straining the landscape-scale distribution.

The size of the present suitable area in South Africa for the

12 bees species investigated varies by two orders of magnitude

from 9234 km2 for the narrow endemic R. aurata Whitehead

& Steiner (Melittidae) to about 280,000 km2 for the

widespread A. atrocincta (Lepeletier) (Apidae) (Table 3).

The average size of the present suitable area of the bees from

the western South African winter rainfall area (44,333 km2,

seven species) is only 32.9% of that of the species from the

summer rainfall areas in the east of the country (134,419 km2,

five species). Our model predicts a net reduction of range size

for eight of the 12 species that is dramatic for five of them

(Table 3). The overall gain of future suitable area under the

more likely emission scenario A2 averages 26.6% of the

modelled present distributions but is exceeded by an average

63.1% future loss resulting in an overall range contraction

of 36.5% to less than two-thirds of their present size.

For the lower emission B2 scenario, the predictions show the

same general trend but are less severe (Table 3).

There are conspicuous differences in average predicted losses

of habitat between species of winter (59.8%) and summer

rainfall regions (67.6%) (Figs 2 & 4), but the forecasted gains

are much smaller for the former (11.7%) than for the latter

(47.5%) resulting in a greater range contraction for the species

of the winter rainfall region (48.2%) compared to the summer

rainfall region (20.1%) that is dominated by range shifts rather

than contractions. However, the change of range sizes varies

considerably between species from almost balanced gains and

losses in C. marleyi Cockerell (Colletidae) and S. chloris

Eardley (Colletidae), to almost complete loss in habitat in

R. intermixta Cockerell (Melittidae) and 50% net gains in

M. arrogans (Smith) (Melittidae) (Table 3).

Five species of the winter rainfall area show net losses of

climatically suitable habitats between 49% and 99% mainly in

the drier coastal plains and northern parts of their range

resulting in massive range contractions, while net gains for

S. chloris and M. arrogans are predicted for areas east of the

present distribution ranges leading to eastward range exten-

sions (Figs 2 & 4). In the summer rainfall area, only

R. rufocincta (Cockerell) (Melittidae) shows a dramatic net

loss of suitable habitat and a range contraction into the higher

altitudes of the Drakensberg Mountains (Fig. 4). The other

four species show variable net habitat gains or losses (Table 3),

but pronounced range shifts generally away from the dry

interior of South Africa towards the coasts and the Drakens-

berg Mountains (Figs 4 & 5).

DISCUSSION

Range shifts in South African bee species

The bee fauna of the winter rainfall area is arguably in terms

of climate change the most vulnerable and highly threatened

taxonomic group because of its confined to an area of only

about 150,000 km2 (ranging from western South Africa to

the extreme southwest of Namibia). In contrast to this, the

distribution area of most bees of the summer rainfall area

extends far north in some cases to East Africa (Kuhlmann,

2009). Because of lack of sufficient climate and distribution

data, we neglected all records outside South Africa.

However, despite the predicted dramatic loss of suitable

habitats for bees in eastern South Africa, their overall range

sizes of many species even within the country are about

an order of magnitude larger than those of their congeners

in the west (Table 3). In principle, they also have the chance

Table 3 Projected percentage change of geographic range for 12 South African bee species under the HadCM3 GCM, A2 and B2 climate

scenarios in 2080–2100

Bee species

Present suitable

area (km2)

A2 scenario predicted 2080–2100

distribution

B2 scenario predicted 2080–2100

distribution

% Gain % Loss

% Net

suitable area % Gain % Loss

% Net

suitable area

Rediviva aurata 9234 0.0 87.4 )87.4 0.0 64.0 )64.0

Scrapter chloris 28,632 20.8 6.1 +14.7 13.9 9.3 +4.6

Rediviva intermixta 19,336 0.0 99.4 )99.4 0.0 93.8 )93.8

Colletes antecessus 34,823 0.0 75.4 )75.4 0.0 51.9 )51.9

Fidelia braunsiana 48,479 0.1 91.4 )91.3 0.1 83.0 )82.9

Rediviva rufipes 68,533 2.4 51.5 )49.1 4.5 36.9 )32.4

Melitta arrogans 101,299 58.3 7.9 +50.4 32.2 6.4 +25.8

Nomia candida 49,186 92.5 51.2 +41.3 87.0 43.9 +43.1

Rediviva rufocincta 68,087 0.1 93.2 )93.1 0.1 90.9 )90.8

Colletes marleyi 77,943 78.5 74.6 +3.9 58.3 70.6 +12.3

Anthophora vestita 196,407 49.3 61.2 )11.9 44.6 49.0 )4.4

Amegilla atrocincta 280,472 16.9 57.6 )40.7 23.9 46.3 )22.4

GCM, general circulation model.

Species ordered by distribution type (see Table 1) and range size.



to evade the impact of climate change by north and

eastward range shifts which has happened several times in

the past by using the so-called arid corridor (Poynton, 1995)

and subsequently making them potentially less vulnerable to

climate change.

There is a broad consensus that the CFR and its flora and

fauna are highly susceptible to climate change (Rutherford

et al., 1999; Midgley et al., 2003; Hannah et al., 2005). Erasmus

et al. (2002) investigated the impact of climate change on a set

of 179 animal species including mammals, birds, reptiles,

butterflies, dung beetles, jewel beetles, antlions and termites.

Species of the winter rainfall area had the highest losses and

generally showed range shifts in an easterly direction, reflecting

the east–west aridity gradient across the country. In the east of

the country, species partly showed a westward shift. The high

vulnerability of the CFR to climate change was also confirmed

by a study of the ant fauna along an altitudinal transect from

the coast to the Cederberg mountains (Botes et al., 2006) and

the general eastward range shift of mammals in southern Africa

(Thuiller et al., 2006).

Climate change does not only affect the distribution of

organisms but also a number of other life history traits

(Parmesan, 2006). The potential impact of climate change on

some aspects of bee biology and ecology is briefly discussed in

the following because they can seriously alter the model

predictions (Parmesan, 2006; Willis & Bhagwat, 2009).

Climate change and sociality in bees

Bee species show all gradations between solitary ways of life

and highly social behaviour with morphologically differenti-

ated queen and worker castes (Michener, 2007). The subfamily

Halictinae is unique in this respect because it exhibits all

intermediate forms of sociality including species like Halictus

Figure 2 Projected changes in geographic

range for bee species endemic to the winter

rainfall area under A2 climate scenario.

M. Kuhlmann et al.


rubicundus that is socially flexible and can show social and

solitary behaviour even within a single population (Yanega,

1993). The level of sociality and demography of primitively

social bees is influenced by environmental factors, mainly

temperature, with sociality generally occurring in warmer

environments on lower altitude or lower latitude, respectively

(Soucy, 2002). For Lasioglossum baleicum, the threshold

temperature for a shift from solitary to social behaviour

was estimated to be 10.33 �C, and the expected degree-day

accumulation was 340 degree days (Hirata & Higashi, 2008).

Primitively social halictids are among the most common

bees in many ecosystems and as generalist flower visitors they

are potentially very important pollinators (Michener, 2007),

but despite this, their role in natural and semi-natural

ecosystems has hardly been investigated. With more than 50

described species, Patellapis is the most species-rich halictid

bee genus in the CFR (Kuhlmann, 2009). Most Patellapis bees

are endemic to the CFR, and they can be particularly abundant

in some places suggesting that they play an important role as

pollinators. Because at least some of the species are primitively

social (Timmermann & Kuhlmann, 2008), it is likely that their

life cycle is altered by climate change.

Furthermore, most bee species in the CFR are active in

winter and spring during the peak flowering season (Mayer &

Kuhlmann, 2004). Cold and rainy weather in this time of the

year provide an unfavourable environment for pollinating

insects like bees leading to reduced flight activity, low rate of

reproduction and widespread pollen limitation of the flora

(Johnson & Bond, 1997). A general warming of the South

African winter rainfall area as reported by Warburton et al.

(2005) can, therefore, be expected to increase reproduction

especially in primitively social bees potentially resulting in

higher abundances. This hypothesis is in accordance with

anecdotal observations that the populations of the communal


range for bee species endemic to the winter

rainfall area under B2 climate scenario.



Patellapis doleritica increased drastically in some places since

2002 (M. Kuhlmann, unpublished data).

Climate change and plant–pollinator interactions

Little is known how climate change might impact on other

aspects of the bee life cycle, from winter hibernation through

foraging to reproduction and host plant synchronization

(Brown & Paxton, 2009). Studies on dragonflies (Dingemanse

& Kalkman, 2008), hoverflies (Graham-Taylor et al., 2009) and

butterflies (Roy & Sparks, 2000) revealed phenological shifts as

a consequence of global warming. Although for plants

changing flowering phenologies under climate change are well

documented (Menzel et al., 2006), the impact on plant–

pollinator mutualisms is little understood. By reviewing

published data for plant–pollinator mutualisms, Hegland et al.

(2009) found that phenological responses to climate warming

often seem to occur at parallel magnitudes in plants and

pollinators, rarely resulting in temporal mismatches among

these mutualistic partners. However, Memmott et al. (2007)

found that climate change-induced phenological shifts reduced

the floral resources available to pollinators by 17–50%, thereby

increasing extinction risks and disrupting plant–pollinator

interactions. Because of the key role that plant–pollinator

interactions play in the CFR (see above) and because bees in

this region, unlike in other parts of the world, seem to show no

or little synchronization with their host plants (Mayer &

Kuhlmann, 2004), more pronounced negative effects because

of climate change are likely. Dreyer et al. (2006) demonstrated

that specifically exendospermous Oxalis (Oxalidaceae) species

(with more than 200 described species, Oxalis is a major

constituent of the Cape flora) tend to have a flowering

sensitivity to alterations in temperature and delayed onset of

winter rains potentially leading to desynchronization with their

pollinators, many of them specific visitors of Oxalis. Special-

ized plant–pollinator mutualisms are common in the CFR, and

the loss of a specialized pollinator can have dramatic negative

effects on plant reproduction. This was demonstrated for the


range for bee species of various

distribution types under A2 climate

scenario.

M. Kuhlmann et al.


rare shrub Ixianthes retzioides (Scrophulariaceae) whose

oil-secreting flowers are exclusively pollinated by the oil-

collecting bee Rediviva gigas (Steiner & Whitehead, 1996).

These observations are fuelling concerns that plant–pollinator

mutualisms in the CFR could potentially breakdown on a

larger scale because of climate change, with negative effects for

plant reproduction, vegetation structure and the ecological and

evolutionary processes that maintain and generate the extant

species (Cowling & Pressey, 2001; Parmesan, 2006).

Implications for conservation

For successful conservation of biodiversity under climate

change-induced habitat fragmentation, availability of migra-

tion corridors and reserves, dispersal rates and colonization

ability can be crucial factors as demonstrated for the Cape

Proteaceae (Hannah et al., 2005).

Bees are excellent flyers and have been show to rapidly

colonize restored inland sand dunes if source populations are

in close proximity (Exeler et al., 2009). But their capability

for colonization is correlated with distance from source and

habitat fragmentation (Steffan-Dewenter & Westphal, 2008).

Even areas neighbouring ancient suitable bee habitats

are slowly recolonized, and isolated habitats are hardly reached

by most bee species, limiting their potential to successfully

establish new populations even in suitable habitats (Franzen

et al., 2009). Colonization success is often limited by resource

availability because bees are central place foragers and require

resources like suitable nesting sites, food plants and nesting

materials (e.g. resin, mud) for specialized species or sufficiently

large host bee populations within a short range for cleptopar-

asitic cuckoo bees (Murray et al., 2009). Thus, for many

species, local habitat structure appears to be more important

than large-scale landscape composition (Gathmann

& Tscharntke, 2002). This is in accordance with the findings

of Donaldson et al. (2002) who showed that bee species

composition did not vary in different sized renosterveld

fragments in the Cape, but the abundance of some species


range for bee species of various distribu-

tion types under the B2 climate scenario.



was affected. However, the authors stressed the need for more

data on pollinator habitat requirements to assess the effect of

habitat fragmentation on pollinators and plant–pollinator

interaction. Caution may be required in applying results based

on studies in Europe to the CFR as in the latter bee

populations generally seem to be much smaller and more

isolated (M. Kuhlmann, unpublished data). This is supported

by the level of population genetic subdivision (Fst) within the

honeybee that was almost twice as high in South Africa (0.105)

compared to Germany (0.064), despite that the sampled

subpopulations in South Africa were geographically closer

together (Moritz et al., 2007).

Understanding and monitoring the effects of rising temper-

atures and changing precipitation regimes on pollination

service and plant–pollinator interactions in the CFR are vital

for the development of efficient conservation strategies for this

global biodiversity hotspot (Hoffman et al., 2009). This is a

pressing problem because climate change is accompanied by

rapid habitat transformation and landscape fragmentation

(Rouget et al., 2003). Key to success is the availability of the

still largely missing baseline data on species ecology and

processes involved in pollination service in natural ecosystems,

including field monitoring for validating the predictions of

species distribution models (Yates et al., 2010a).

Based on our results, suitable regions for monitoring bee

populations are the coastal low lands of the winter rainfall

area. This area has been identified to have the highest risks

for extinction and areas where persistence of populations

is most likely, for example, on the higher grounds of the

escarpment (Figs 2–5). It is likely that here the impact and

dynamics of climate change on population level can probably

be observed best. Priority should be given to endemic bees

and especially the basal genera within the families Colletidae

(Scrapter), Melittidae (Afrodasypoda, Haplomelitta) and

Megachilidae (Afroheriades, Aspidosmia, Fidelia, Fideliopsis)

that form a unique part of South Africa’s rich bee fauna in

the winter rainfall area (Kuhlmann, 2009). The conservation

of evolutionary processes and phylogenetic diversity has been

recognized as a priority (Cowling & Pressey, 2001), and

Forest et al. (2007) recently demonstrated for the Cape

flora that it is the best strategy for preserving its unique

diversity.

CONCLUSIONS

Our models predict that in course of climate change, most bee

species in South Africa might experience considerable range

shifts and declines in suitable habitats. The CFR, a bee diversity

hotspot of global significance, is likely to be particularly

negatively affected increasing the risk of extinction for a

significant proportion of its unique fauna. Rising temperatures

and altered rainfall regimes potentially also have a huge impact

on bee life history traits like sociality in halictid bees and host

plant synchronization and, thus in turn, on pollination service

and plant reproductive success. Given the relevance of plant–

pollinator mutualisms for ecosystem integrity and evolution in

the Cape (Johnson & Bond, 1997; van der Niet & Johnson,

2009), it is crucial that future studies on the impact of climate

change include plant–animal interactions, with a particular

focus on vital ecosystem functions and services like pollination.

Our results provide a starting point for further investigations

in this direction, and we recommend that bee population

dynamics is monitored across species ranges to validate our

model predictions.

ACKNOWLEDGEMENTS

We would like to thank the South African Atlas of Climatology

and Agrohydrology, the Worldclim database, the National

Land-cover Project 2006, and the South African National

Biodiversity Institute, for providing some of the data used in

this study. This material is based upon work partially

supported financially by the National Research Foundation

of South Africa (Ref. No. IFR2009090800013).

REFERENCES

Aizen, M.A., Garibaldi, L.A., Cunningham, S.A. & Klein, A.-M.

(2009) How much does agriculture depend on pollinators?

Lessons from long-term trends in crop production. Annals of

Botany, 103, 1579–1588.

Allsopp, M.H., de Lange, W.J. & Veldtman, R. (2008) Valuing

insect pollination services with cost of replacement. PLoS

ONE, 3, e3128.

Anderson, B. & Johnson, S.D. (2008) The geographical mosaic

of coevolution in a plant-pollinator mutualism. Evolution,

62, 220–225.

Botes, A., McGeoch, M.A., Robertson, G.H., van Niekerk, A.,

Davids, H.P. & Chown, S.L. (2006) Ants, altitude and change

in the northern Cape Floristic Region. Journal of Biogeogra-

phy, 33, 71–90.

Brown, M.J.F. & Paxton, R.J. (2009) The conservation of bees:

a global perspective. Apidologie, 40, 410–416.

Carter, T.R., Jones, R.N., Lu, X., Bhadwal, S., Conde, O.,

Mearns, L.O., O’Neill, B.C., Rounsevell, M.D.A. & Zurek,

M.B. (2007) New assessment methods and the characteri-

sation of future conditions. Climate change 2007: impacts,

adaptation and vulnerability. Contribution of working group II

to the fourth assessment report of the Intergovernmental Panel

on climate change (ed. by M.L. Parry, O.F. Canziani, J.P.

Palutikof, P.J. van der Linden and C.E. Hanson), pp. 133–

171. Cambridge University Press, Cambridge, UK.

Cowling, R.M. & Pressey, R.L. (2001) Rapid plant diversifica-

tion: planning for an evolutionary future. Proceedings of the

National Academy of Sciences USA, 98, 5452–5457.

Dingemanse, N.J. & Kalkman, V.J. (2008) Changing temper-

ature regimes have advanced the phenology of Odonata in

the Netherlands. Ecological Entomology, 33, 394–402.

Donaldson, J., Nanni, I., Zachariades, C. & Kemper, J. (2002)

Effects of habitat fragmentation on pollinator diversity and

plant reproductive success in renosterveld shrublands of

South Africa. Conservation Biology, 16, 1267–1276.

M. Kuhlmann et al.


Dreyer, L.L., Esler, K.J. & Zietsman, J. (2006) Flowering

phenology of South African Oxalis – possible indicator of

climate change? South African Journal of Botany, 72,

150–156.

Eardley, C.D. (1994) The genus Amegilla Friese (Hymenoptera:

Anthophoridae) in southern Africa. Entomology Memoir

Department of Agricultural Development Republic of South

Africa, 91, 1–68.

Eardley, C.D. (1996) The genus Scrapter Lepeletier & Serville

(Hymenoptera: Colletidae). African Entomology, 4, 37–92.

Eardley, C.D. & Brooks, R.W. (1989) The genus Anthophora

Latreille in southern Africa (Hymenoptera: Anthophoridae).

Entomology Memoir Department of Agriculture and Water

Supply Republic of South Africa, 76, 1–55.

Eardley, C.D. & Kuhlmann, M. (2006) Southern and East

African Melitta Kirby (Apoidea: Melittidae). African Ento-

mology, 14, 293–305.

Elith, J., Graham, C.H., Anderson, R.P. et al. (2006) Novel

methods improve prediction of species’ distributions from

occurrence data. Ecography, 29, 129–151.

Elith, J., Phillips, S.J., Hastie, T., Dudık, M., Cheel, Y.E. &

Yates, C.J. (2011) A statistical explanation of MaxEnt for

ecologists. Diversity and Distributions, 17, 43–57.

Erasmus, B.F.N., van Jaarsveld, A.S., Chown, S.L., Kshatriya,

M. & Wessels, K.J. (2002) Vulnerability of South African

animal taxa to climate change. Global Change Biology, 8,

679–693.

Exeler, N., Kratochwil, A. & Hochkirch, A. (2009) Restoration

of riverine inland sand dune complexes: implications for the

conservation of wild bees. Journal of Applied Ecology, 46,

1097–1105.

Forest, F., Grenyer, R., Rouget, M., Davies, T.J., Cowling, R.M.,

Faith, D.P., Balmford, A., Manning, J.C., Proches, S., van der

Bank, M., Reeves, G., Hedderson, T.A. & Savolainen, V.

(2007) Preserving the evolutionary potential of floras in

biodiversity hotspots. Science, 445, 757–760.

Franzen, M., Larsson, M. & Nilsson, S.G. (2009) Small local

population sizes and high habitat patch fidelity in a

specialised solitary bee. Journal of Insect Conservation, 13,

89–95.

Gathmann, A. & Tscharntke, T. (2002) Foraging ranges of

solitary bees. Journal of Animal Ecology, 71, 757–764.

Graham-Taylor, L.G., Stubbs, A.E. & Brooke, M.D. (2009)

Changes in phenology of hoverflies in a central England

garden. Insect Conservation and Diversity, 2, 29–35.

Guo, D. (2010) Contributions to spatial uncertainty modelling in

GIS. Lambert Academic Publishing, Saarbrucken.

Hannah, L., Midgley, G.F., Hughes, G. & Bomhard, B. (2005)

The view from the Cape: extinction risk, protected areas and

climate change. BioScience, 55, 231–242.

Hegland, S.J., Nielsen, A., Lazaro, A., Bjerknes, A.-L. &

Totland, O. (2009) How does climate warming affect plant-

pollinator interactions? Ecology Letters, 12, 184–195.

Heikkinen, R.K., Luoto, M., Araujo, M.B., Virkkala, R.,

Thuiller, W. & Sykes, M.T. (2006) Methods and uncertain-

ties in bioclimatic envelope modelling under climate change.

Progress in Physical Geography, 30, 751–777.

Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. & Jarvis,

A. (2005) Very high resolution interpolated climate surfaces

for global land areas. International Journal of Climatology, 25,

1965–1978.

Hirata, M. & Higashi, S. (2008) Degree-day accumulation

controlling allopatric and sympatric variations in the

sociality of sweat bees, Lasioglossum (Evylaeus) baleicum

(Hymenoptera: Halictidae). Behavioral Ecology and Sociobi-

ology, 62, 1239–1247.

Hoffman, M.T., Carricka, P.J., Gillsona, L. & West, A.G. (2009)

Drought, climate change and vegetation response in the

Succulent Karoo, South Africa. South African Journal of

Science, 105, 54–60.

Johnson, S.D. & Bond, W.J. (1997) Evidence for widespread

pollen limitation of fruiting success in Cape wildflowers.

Oecologia, 109, 530–534.

Kearns, C.A., Inouye, D.W. & Waser, N.M. (1998) Endangered

mutualisms: the conservation of plant-pollinator interac-

tions. Annual Review of Ecology and Systematics, 29, 83–

112.

Klein, A.-M., Vaissiere, 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.

Kuhlmann, M. (2009) Patterns of diversity, endemism and

distribution of bees (Insecta: Hymenoptera: Anthophila) in

southern Africa. South African Journal of Botany, 75, 726–

738.

Liu, L., Berry, P.M., Dawson, T.P. & Pearson, R.G. (2005)

Selecting thresholds of occurrence in the prediction of spe-

cies distributions. Ecography, 28, 385–393.

Mayer, C. & Kuhlmann, M. (2004) Synchrony of pollinators

and plants in the winter rainfall area of South Africa –

observations from a drought year. Transactions of the Royal

Society of South Africa, 59, 55–57.

McLeish, M., Guo, D., van Noort, S. & Midgley, G.F. (2011)

Life on the edge: rare and restricted episodes of a pan-

tropical mutualism adapting to drier climates. New Phytol-

ogist, 191, 210–222.

Memmott, J., Craze, P.G., Waser, N.M. & Price, M.V. (2007)

Global warming and the disruption of plant-pollinator

interactions. Ecology Letters, 10, 710–717.

Menzel, A., Sparks, T.H., Estrella, N. et al. (2006) European

phenological response to climate change matches the

warming pattern. Global Change Biology, 12, 1969–1976.

Michener, C.D. (2007) The bees of the world, 2nd edn. Johns

Hopkins University Press, Baltimore.

Midgley, G.F., Hannah, L., Millar, D., Rutherford, M.C. &

Powrie, L.W. (2002) Assessing the vulnerability of

species richness to anthropogenic climate change in a

biodiversity hotspot. Global Ecology & Biogeography, 11,

445–451.



Midgley, G.F., Hannah, L., Millar, D., Thuiller, W. & Booth, A.

(2003) Developing regional and species-level assessments of

climate change impacts on biodiversity in the Cape Floristic

Region. Biological Conservation, 112, 87–97.

Moritz, R.F.A., Kraus, F.B., Kryger, P. & Crewe, R.M. (2007)

The size of wild honeybee populations (Apis mellifera) and

its implications for the conservation of honeybees. Journal of

Insect Conservation, 11, 391–397.

Murray, T.E., Kuhlmann, M. & Potts, S.G. (2009) Conserva-

tion ecology of bees: populations, species and communities.

Apidologie, 40, 211–236.

Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca,

G.A.B. & Kent, J. (2000) Biodiversity hotspots for conser-

vation priorities. Nature, 403, 853–858.

van der Niet, T. & Johnson, S.D. (2009) Patterns of plant

speciation in the Cape floristic region. Molecular Phyloge-

netics and Evolution, 51, 85–93.

Parmesan, C. (2006) Ecological and evolutionary responses to

recent climate change. Annual Review of Ecology, Evolution,

and Systematics, 37, 637–669.

Pauly, A. (2000) Classification des Nomiinae africains: le genre

Leuconomia Pauly, 1980 (Hymenoptera, Apoidea, Halicti-

dae). Bulletin de l¢Institut Royal des Sciences Naturelles de

Belgique, Entomologie, 70, 165–188.

Phillips, S.J., Dudık, M. & Schapire, R.E. (2004) A maximum

entropy approach to species distribution modelling. Pro-

ceedings of the Twenty-First International Conference on

Machine Learning, 2004, 655–662.

Phillips, S.J., Anderson, R.P. & Schapire, R.E. (2006) Maxi-

mum entropy modelling of species geographic distributions.

Ecological Modelling, 190, 231–259.

Phillips, S.J., Dudik, M., Elith, J., Graham, C.H., Lehmann, A.,

Leathwick, J.R. & Ferrier, S. (2009) Sample selection bias and

presence-only models of species distributions: implications

for selection of background and pseudo-absences. Ecological

Applications, 19, 181–197.

Poynton, J.C. (1995) The ‘‘arid corridor’’ distribution in

Africa: a search for instances among amphibians. Madoqua,

19, 45–48.

Roberts, S.P.M., Potts, S.G., Biesmeijer, K., Kuhlmann, M.,

Kunin, B. & Ohlemuller, R. (2011) Assessing continental-

scale risks for generalist and specialist pollinating bee species

under climate change. BioRisk, 6, 1–18.

Rouget, M., Richardson, D.M., Cowling, R.M., Lloyd, J.W. &

Lombard, A.T. (2003) Current patterns of habitat transfor-

mation and future threats to biodiversity in terrestrial eco-

systems of the Cape Floristic Region, South Africa. Biological

Conservation, 112, 63–85.

Roy, D.B. & Sparks, T.H. (2000) Phenology of British butter-

flies and climate change. Global Change Biology, 6, 407–416.

Rutherford, M.C., Powrie, L.W. & Schulze, R.E. (1999) Climate

change in conservation areas of South Africa and its

potential impact on floristic composition: a first assessment.

Diversity and Distributions, 5, 253–262.

Schulze, R.E. & Kruger, G.P. (2007) Terrain morphology.

South African atlas of climatology and agrohydrology (ed. by

R.E. Schulze). WRC Report 1489/1/06, Section 3.2, pp. 13–

16. Water Research Commission, Pretoria.

Schulze, R.E. & Maharaj, M. (2007) Rainfall seasonality. South

African atlas of climatology and agrohydrology (ed. by R.E.

Schulze). WRC Report 1489/1/06, Section 6.5, pp.39–40.

Water Research Commission, Pretoria.

Soucy, S.L. (2002) Nesting biology and socially polymorphic

behavior of the sweat bee Halictus rubicundus. Annals of the

Entomological Society of America, 95, 57–65.

Steffan-Dewenter, I. & Westphal, C. (2008) The interplay of

pollinator diversity, pollination services and landscape

change. Journal of Applied Ecology, 45, 737–741.

Steiner, K.E. & Whitehead, V.B. (1996) The consequences of

specialization for pollination in a rare South African shrub,

Ixianthes retzioides (Scrophulariaceae). Plant Systematics and

Evolution, 201, 131–138.

Thuiller, W., Broennimann, O., Hughes, G., Alkemade, J.R.M.,

Midgley, G.M. & Corsi, F. (2006) Vulnerability of African

mammals to anthropogenic climate change under conser-

vative land transformation assumptions. Global Change

Biology, 12, 424–440.

Timmermann, K. & Kuhlmann, M. (2008) The biology of

a Patellapis (s. str.) species (Hymenoptera: Apoidea:

Halictidae): sociality described for the first time in this bee

genus. Apidologie, 39, 189–197.

Warburton, M., Schulze, R.E. & Maharaj, M. (2005) Is South

Africa’s temperature changing? An analysis of trends from

daily records, 1950–2000. Climate change and water resources

in southern Africa: studies on scenarios, impacts, vulnerabilities

and adaptation (ed. by R.E. Schulze), pp. 275–295. WRC

Report 1430/1/05, Chapter 16. Water Research Commission,

Pretoria.

Whitehead, V.B. & Eardley, C.D. (2003) African Fideliini:

genus Fidelia Friese (Hymenoptera: Apoidea: Megachilidae:

Fideliinae). Journal of the Kansas Entomological Society, 76,

250–276.

Whitehead, V.B. & Steiner, K.E. (2001) Oil-collecting bees of the

winter rainfall area of South Africa (Melittidae, Rediviva).

Annals of the South African Museum, 108, 143–277.

Whitehead, V.B., Steiner, K.E. & Eardley, C.D. (2008)

Oil collecting bees mostly of the summer rainfall area of

southern Africa (Melittidae, Rediviva). Journal of the Kansas

Entomological Society, 81, 122–141.

Williams, P.H., Araujo, M.B. & Rasmont, P. (2007) Can vul-

nerability among British bumblebee (Bombus) species be

explained by niche position and breadth? Biological Conser-

vation, 138, 493–505.

Willis, K.J. & Bhagwat, S.A. (2009) Biodiversity and climate

change. Science, 326, 806–807.

Yanega, D. (1993) Environmental influences on male pro-

duction and social structure in Halictus rubicundus (Hyme-

noptera: Halictidae). Insectes Sociaux, 40, 169–180.

Yates, C.J., Elith, J., Latimer, A.M., Le Maitre, D., Midgley,

G.F., Schurr, F.M. & West, A.G. (2010a) Projecting climate

change impacts on species distributions in megadiverse

South African Cape and Southwest Australian Floristic

M. Kuhlmann et al.


Regions: opportunities and challenges. Austral Ecology, 35,

374–391.

Yates, C., McNeill, A., Elith, J. & Midgley, G. (2010b) Assessing

the impacts of climate change and land transformation on

Banksia in the South West Australian Floristic Region.

Diversity and Distributions, 16, 187–201.

BIOSKETCHES

All the authors are interested in the effects of climate change on

insect biodiversity in southern Africa and in the application of

species distribution models for understanding potential

vulnerabilities.

Michael Kuhlmann is interested in bee systematics,

biogeography and ecology. Recently, he has focussed on the

impact of climate change and plant–pollinator interactions on

distribution patterns in the biodiversity hotspot of the

South African winter rainfall area and its implications for

conservation.

Author contributions: All authors contributed to the develop-

ment of models and writing of the manuscript.

Editor: Jessica Hellmann



Consequences of warming up a hotspot: species range shifts within a centre of bee diversity

Documents

Transcript of Consequences of warming up a hotspot: species range shifts within a centre of bee diversity