ArticlePDF AvailableLiterature Review

Abstract and Figures

Bees pollinate most of the world's wild plant species and provide economically valuable pollination services to crops; yet knowledge of bee conservation biology lags far behind other taxa such as vertebrates and plants. There are few long-term data on bee populations, which makes their conservation status difficult to assess. The best-studied groups are the genus Bombus (the bumble bees), and bees in the EU generally; both of these are clearly declining. However, it is not known to what extent these groups represent the approximately 20,000 species of bees globally. As is the case for insects in general, bees are underrepresented in conservation planning and protection efforts. For example, only two bee species are on the global IUCN Red List, and no bee is listed under the U.S. Endangered Species Act, even though many bee species are known to be in steep decline or possibly extinct. At present, bee restoration occurs mainly in agricultural contexts, funded by government programs such as agri-environment schemes (EU) and the Farm Bill (USA). This is a promising approach given that many bee species can use human-disturbed habitats, and bees provide valuable pollination services to crops. However, agricultural restorations only benefit species that persist in agricultural landscapes, and they are more expensive than preserving natural habitat elsewhere. Furthermore, such restorations benefit bees in only about half of studied cases. More research is greatly needed in many areas of bee conservation, including basic population biology, bee restoration in nonagricultural contexts, and the identification of disturbance-sensitive bee species.
Content may be subject to copyright.
Ann. N.Y. Acad. Sci. ISSN 0077-8923
The Year in Ecology and Conservation Biology
The conservation and restoration of wild bees
Rachael Winfree
Department of Entomology, Rutgers University, New Brunswick, New Jersey, USA
Address for correspondence: Rachael Winfree, Department of Entomology, 93 Lipman Drive, Rutgers University,
New Brunswick, NJ 08901 USA.
Bees pollinate most of the world’s wild plant species and provide economically valuable pollination services to crops;
yet knowledge of bee conservation biology lags far behind other taxa such as vertebrates and plants. There are few
long-term data on bee populations, which makes their conservation status difficult to assess. The best-studied groups
are the genus Bombus (the bumble bees), and bees in the EU generally; both of these are clearly declining. However, it
is not known to what extent these groups represent the approximately 20,000 species of bees globally. As is the case for
insects in general, bees are underrepresented in conservation planning and protection efforts. For example, only two
bee species are on the global IUCN Red List, and no bee is listed under the U.S. Endangered Species Act, even though
many bee species are known to be in steep decline or possibly extinct. At present, bee restoration occurs mainly
in agricultural contexts, funded by government programs such as agri-environment schemes (EU) and the Farm
Bill (USA). This is a promising approach given that many bee species can use human-disturbed habitats, and bees
provide valuable pollination services to crops. However, agricultural restorations only benefit species that persist in
agricultural landscapes, and they are more expensive than preserving natural habitat elsewhere. Furthermore, such
restorations benefit bees in only about half of studied cases. More research is greatly needed in many areas of bee
conservation, including basic population biology, bee restoration in nonagricultural contexts, and the identification
of disturbance-sensitive bee species.
Keywords: agri-environment scheme; ecosystem service; Farm Bill; global change; land use change; pollination;
pollinator; pollinator conservation; pollinator restoration; restoration ecology
The importance of bees
As the world’s primary pollinators, bees are a criti-
cally important functional group. Roughly 90% of
world’s plant species are pollinated by animals,1,2
and the main animal pollinators in most ecosystems
are bees.3Although other taxa including butterflies,
flies, beetles, wasps, bats, birds, lizards, and mam-
mals can be important pollinators in certain habitats
and for particular plants (e.g., Refs. 4 and 5), none
achieves the numerical dominance as flower visi-
tors worldwide as bees.3The likely reason for this
is that unlike other taxa, bees are obligate florivores
throughout their life cycle, with both adults and lar-
vae dependent on floral products, primarily pollen
and nectar.6
In addition to their crucial role for wild plants,
bees are the main pollinators of agricultural crops,
75% of which benefit from animal pollination.7,8
Honey bees, primarily Apis mellifera and to a lesser
extent Apis cerana,arewidelymanagedinhivesfor
crop pollination and are presumably the most im-
portant agricultural pollinators worldwide (as do-
mesticated species, these are not covered in this re-
view, except for regions where they are native or
feral). Declines in the number of managed honey
bee hives in the United States over the past 50
years,9in conjunction with recent losses due to
Colony Collapse Disorder,10 have raised concern
about the extent to which global agriculture re-
lies on a single-managed bee species. Although at
a global scale neither managed honey bees nor the
yield of the crops they pollinate has declined over
the past few decades, dependence on bee-pollinated
crops is increasing faster than the supply of honey
bees, which suggests that problems may occur in the
may be substantial, but is more debated than is
doi: 10.1111/j.1749-6632.2010.05449.x
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 169
Wild bee conservation Winfree
their importance in natural ecosystems. Non-Apis
species are equally effective or better pollinators
than are honey bees for many crops.14–20 The chal-
lenge in using these species for crop pollination is
not quality but rather quantity, and management
techniques exist only for a small number of non-
Apis taxa.16–18,21–23 Wild, unmanaged native bees
also provide crop pollination as an ecosystem ser-
vice. Unmanaged bees alone can fully pollinate crops
in some agricultural contexts19,20,24,25 and are fre-
quent flower visitors in others,26 thereby contribut-
ing to meeting the crop’s pollination needs. In ad-
dition, when present with honey bees, native bees
can enhance honey bee effectiveness.27,28 The role
of native bees as crop pollinators helps to generate
support for bees’ conservation.
Bee diversity and biogeography
Bees as a monophyletic group constitute the Api-
formes.6Roughly 18,000 bee species have been de-
scribed, with the true total number of species likely
near 20,000.6In contrast to most other taxa, bee
biodiversity peaks not in the tropics but in arid
temperate areas.6,29–31 Global hotspots in recorded
bee diversity include the southwestern USA and the
Mediterranean. In contrast, warm temperate areas
such as eastern North America, Europe, and south-
ern South America are intermediate in diversity,
and the moist tropics are relatively depauperate31
—although recent collecting work in the neotropics
suggests that bee diversity may be higher there than
previously assumed,6and at present only a third
of neotropical bee species have been described.32
None of these biogeographic conclusions is based
on sampling that is standardized for either effort or
area sampled; all are therefore subject to sampling
biases. However, the general patterns have held up
for almost three decades, which suggests they have
some validity. If tropical bee diversity is indeed low,
then bees may provide an exception to the general
rule that the best conservation values, in terms of
protecting the most biodiversity for the least cost,
are to be found in the tropics.33
Several hypotheses have been proposed to explain
the low diversity of tropical bees. First, ground-
nesting bees constitute the majority of species in
many communities, and they may be largely ex-
cluded from the wet tropics because their nests
would flood and/or their larval food supplies would
be subject to fungal attack.6,34,35 Second, tropical
bee communities tend to be strongly dominated by
a small number of eusocial species, primarily from
the groups Apidae and Meliponini.36 These super-
abundant, perennially active and floral generalist
bees may use a large fraction of the available floral
resources, thereby excluding other species.6,35 This
hypothesis raises the question of why social bees
dominate tropical bee communities, and whether
their dominance is a cause or an effect of the low
species richness there.
Sociality may help explain the exceptionally high
bee diversity observed in deserts as well. One hy-
pothesis is that social species, which have long flight
from deserts, where bloom in temporally patchy.
This makes floral resources available for a greater
variety of less abundant, solitary species.37 Asec-
ond hypothesis centers on the role of oligolectic (di-
etary specialist) species, which constitute a high pro-
portion of the desert fauna.30,38 Oligolectic species
may be able to time their emergence to the tempo-
rally erratic bloom of their host plant species better
than polylectic species, thus creating selection for
dietary specialization on the part of desert bees,
and subsequently high diversity.39,40 Third, vari-
able rainfall at relatively small geographic scales,
as found in deserts, provides a possible mechanism
for speciation: if conspecific populations in neigh-
boring localities emerge at different times, the pop-
ulations would not interbreed and may diverge ge-
netically.40 All of these hypotheses about the causes
of bee biodiversity patterns remain to be rigorously
Several phylogenetically important bee lineages
comprise small numbers of species and are
geographically restricted, making them prime
candidates for conservation efforts. The family
Stenotritidae includes just 21 species and is primar-
ily restricted to western Australia. The megachilid
tribe Fideliini represents an ancient lineage of host-
plant specialist bees that are restricted to arid re-
gions of Chile (2 species), Morocco (1 extremely
rare species), and southern Africa (11 species). Fi-
nally, the melittid tribe Macropidini represents a
highly specialized lineage of mostly oil-collecting,
host-plant specialist bees. Afrodasypoda plumipes,
for example, is known from just a handful of spec-
imens collected in the Richtersveld National Park
in northern South Africa. Such lineages highlight
the need to consider conservation efforts directed at
170 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
geographically restricted and phylogenetically im-
portant bee lineages (entire paragraph, Bryan Dan-
forth, pers. com.).
The extent and causes of bee decline
Conservation status of bees
The question of whether we are in the midst of
a global pollinator decline has received much at-
tention in the media as well as the academic lit-
erature,9,41–47 but is difficult to answer empirically
due to a lack of pollinator monitoring programs
and long-term data series. The need for establish-
ing pollinator monitoring programs was recognized
internationally in 1993 when pollinators were in-
corporated into the Convention on Biological Di-
versity, which has been signed by 168 countries
( Polli-
nator monitoring is an important goal of the
EU’s ALARM program (http://www.alarmproject.
net/alarm/objectives.php), which is now collecting
monitoring data in several countries.48 Other re-
gional pollinator protection initiatives are in place,45
but have not yet collected large-scale data.
The best data for entire bee communities come
from the EU and provide strong evidence of de-
clines. Citizen science data from the United King-
dom and the Netherlands show significant declines
in bee species richness when comparing data from
before and after 1980.49 In Belgium, 25% of bee
species have declined during the second half of the
20th century whereas only 11% have increased.50
Across European countries, 37–65% of bee species
are on lists of conservation concern,51–53 although
none is yet Red Listed with IUCN, due to lack of the
required documentation about conservation status.
However, northwest Europe, where bee communi-
ties have been best studied, is one of the most in-
tensively human-used regions of the world and has
been for many centuries.49,54,55 Relative to the rest
of the world, results from this region could therefore
overestimate declines, because bees are responding
to more intensive human land use than elsewhere;
or underestimate them, if the remaining fauna on
which studies are based are already the subset of
species that persist well in agricultural environ-
Thebumblebees(thegenusBombus) are the best-
studied bee taxon and the only taxon that has been
globally assessed for its endangerment status. Eleven
percent of Bombus species should probably be listed
as “near threatened” or above by IUCN.56 How-
ever, only one species is currently listed, because the
others lack the documentation required by IUCN.
Most studies of Bombus have taken place in Europe,
where many species are declining.57,58 Half of the
Bombus species historically known from Britain are
either extinct, or in danger of extinction.57 Of the
60 Bombus species known from west and central
Europe, 30% are now threatened throughout their
range according to IUCN criteria, and 7% went ex-
tinct in this region between 1951 and 2000.59 The
main cause of Bombus decline in the UK and western
Europe is widely agreed to be the agricultural inten-
sification that took place in the 20th century.57 Sev-
eral components of agricultural intensification are
likely important including the decline of preferred
bumble bee forage plants in the landscape,60,61 the
loss of relatively unmanaged grasslands and other
uncropped habitats such as hedgerows, and the
development of synthetic fertilizers that replaced
bee-friendly leguminous cover crops such as clover
as a means for restoring nitrogen to agricultural
soils.57,58 Although many Bombus species have de-
clined, others are doing well despite these changes in
land use. Life history factors associated with species
decline vary somewhat across studies but include
floral specialization, later emergence times, range
extent, and climatic niche.57,61–67
Bombus species are less well studied in North
America but some species are clearly declining there,
for somewhat different reasons than in Europe.
Three formerly common North American species
in the subgenus Bombus sensu strictu,B. affinis,B.
terricola,andB. occidentialis, have all declined dra-
matically, while a fourth which was always rare, B.
franklini, is now close to extinction.68,69 For exam-
ple, B. affinis, which was once common across much
of eastern North America, disappeared from 42 of
43 sites between the early 1970s and mid 2000s.68
The working hypothesis proposed to explain these
declines is parasite infection from commercially
reared congeners. In particular, the fungal pathogen
Nosema bombi mayhavespreadtowildNorthAmer-
ican bees from commercial B. occidentalis and B.
impatiens raised for greenhouse pollination in Eu-
rope, and then imported into the United States.70
In support of this hypothesis, commercial Bombus
are known to have higher pathogen burdens than
wild bees, and to forage outside the greenhouses.71
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 171
Wild bee conservation Winfree
Furthermore parasite loads on individual Bombus
(whether wild or commercial) increase with prox-
imity to greenhouses, and a spatially explicit model
of pathogen spillover from commercial to wild in-
dividuals predicts observed parasite loads well.71,72
In North America other causes of Bombus de-
cline, and other Bombus species, are less well stud-
ied. In contrast to the case in Europe, neither floral
specialization nor habitat and range size effects ex-
plain North American declines across 14 Bombus
species.68 Consistent with the British case, however,
Bombus declines in Illinois cooccurred with large-
scale agricultural intensification.69 As in Britain,
roughly half of the Bombus species in Illinois are
now either extirpated or in broad-scale decline.69
Given that Bombus is the best studied bee genus,
what do Bombus declines tell us about the status of
the other 442 bee genera?6Bombus might be particu-
larly vulnerable because they are social (see below),
whereas most bee species are solitary. Bombus are
also larger than most other bee species, although
this might bias their extinction risk in either di-
rection (see below). In Belgium, Bombus have de-
clined more than have most other genera.50 How-
ever, the published literature as a whole shows no
significant difference between Bombus and all other
non-Apis,non-Bombus species in terms of their sen-
sitivity to human disturbance.73 If Bombus are not
notably different from other genera, then their de-
cline does not bode well for the other 99% of the
world’sbee species w hose conservation status is even
more poorly known.
Long-term data for non-Bombus bee communi-
ties outside of northwest Europe are sparse. Roughly
half of the 60 Hylaues species endemic to Hawaii are
74 How-
ever, islands in general and Hawaii in particular are
well-known hotspots of extinction,75 so these re-
sults cannot be generalized to continental faunas.
A 21-year time series exists for Euglossines (orchid
bees) in a tropical forest, during which populations
showed high variability but few consistent trends in
abundance, but data were collected in a relatively
undisturbed area and therefore do not reflect the
effects of anthropogenic changes.76
Threats to the conservation of bees
Habitat loss, invasive species, and (potentially) cli-
mate change are considered the main causes of
species loss for taxa other than bees.77–80 Most work-
ers consider these to be the most important causes
for bees as well81
Habitat loss and fragmentation
Habitat loss is currently the leading cause of species
endangerment77,78 andispredictedtobeinthefu-
ture.79 A recent meta-analysis shows that habitat loss
and fragmentation negatively affects the abundance
and species richness of wild bees.73 However, this ef-
fect is only significant in studies for which analyses
included at least one site that was extremely iso-
lated, variously defined (depending on the criteria
used in the study) as a habitat fragment of less than
1 ha, a site more than 1 km from the nearest natural
habitat, or a site with less than 5% natural habitat re-
maining in the surrounding landscape. Studies that
did not include such extreme sites showed a negative
trend, but it was not significant.73 Most (61%) of the
studies contributing to the meta-analysis included
an extreme site. This raises the possibility that there
is a research bias in the existing literature, in that
habitat loss has been studied where it is more ex-
treme than would be found in a random sample of
global ecosystems. To assess whether research bias
exists, we would need to compare the land cover
surrounding the sites included in the meta-analysis,
to land cover surrounding a random sample of sites
globally. This has not been done.
The high variability in bees’ response to land use
change73 may result, in part, from the fact that some
bee species appear to do well in human-disturbed
habitats. In fact, some studies of bees and habitat
loss define “bee habitat” to include anthropogenic
habitats such as suburban gardens and agricultural
grasslands, and to exclude the native vegetation type
(e.g.,82,83). Some types of temperate forests, in par-
ticular,appear to suppor t relatively few bees,6,82,84–87
although the needs of forest-obligate bee species
have not been sufficiently researched. Agricultural
lands, when not too intensively managed, can pro-
vide good habitat for many bee species,19,88–91 as can
urban/suburban areas.85,92–94
Bees’ use of human-disturbed habitats, in com-
bination with the ecosystem services they pro-
vide, may make them especially well suited to con-
servation planning that combines ecological and
economic criteria, and includes both preserved
and human-used habitats. These planning meth-
ods can be more effective biologically and also less
expensive than traditional conservation using
172 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
nature reserves.95 Furthermore, in contrast to
better-studied vertebrate taxa, small habitat patches
may be sufficient to support insects, including bees,
in otherwise disturbed landscapes.96–98
Climate change
Climate change could cause widespread extinctions
of bees, as it could for other organisms, if bees are
unable to migrate fast enough to keep up with the
regions within their thermal tolerances.80 As yet
there are almost no published data on this question
for bees.99,100 An as yet unpublished, comprehen-
sive analysis of 527 European bee species suggests
depending on the climate change scenario, Europe
could lose 14–27% of its bee species by 2050 due to
climate change.101 Climate change could negatively
affect oligolectic bees in particular if the phenology
of bees and their host plants do not change in con-
cert. This appears to be the case, as bees advance
their emergence times faster than plants as temper-
ature increases.101 The effects of climate change may
be exacerbated by habitat loss. For example, Warren
et al.102 found that among British butterflies, habi-
tat specialists and less mobile species were less able
to track climate changes. Bees with similar charac-
teristics will likely be at greater risk due to climate
Nonnative species
Species invasions, along with habitat loss and cli-
mate change, rank among the top causes of species
endangerment globally.77,79,103 Bees could be nega-
tively affected by nonnative plants, and/or by non-
native bees including the pathogens and parasites
they carry. As yet, it is not possible to general-
ize about how nonnative plants affect bees. Many
studies of nonnative plant–pollinator interactions
have focused on single plant species known apri-
ori to be particularly attractive to pollinators,104
which may introduce a research bias. Studies that
have examined entire plant–pollinator webs should
not suffer from a research bias, and have found
that the nonnative plants have either fewer,105 sim-
ilar,106 or more107 insect species visiting them, as
compared to native plants. The net effect of non-
native plants on bee populations will depend not
only on the bee species that the nonnatives cur-
rently support, but also on what native plants the
nonnatives displaced. I am not aware of any stud-
ies that accounted for this aspect with experimental
or historical data. Last, nonnative plants may ben-
efit generalist bees more than they benefit special-
ists,108 thereby adding to the list of risk factors for
The role of nonnative bees in native bee declines
has generated much interest, especially given the
human-subsidized spread of the honey bee to all
continents except Antarctica. Competition, how-
ever, is notoriously difficult to demonstrate in an
ecological context. Most studies of competition be-
tween native and nonnative bees have been obser-
vational and based on forager densities at flowers,
and have found generally negative effects,109 but for-
ager densities may be unrelated to native bee repro-
duction.109,110 Thesolefullyexperimentalstudyto
monitor native bee reproduction in the presence
and absence of honey bees found a significant neg-
ative effect of honey bee density.111 However, the
study took place in a system with strong bottlenecks
in floral resource availability, which may have in-
creased the chances of finding competition.111 On
European grasslands, wild bee reproduction is not
negatively correlated with the observed density of
honey bee foragers.112 There are few studies of na-
tive bee competition with nonnative taxa other than
honey bees, but the limited evidence suggests that
competition can occur, for example, between na-
tive and exotic Bombus species.109,113 The spread of
pathogens from nonnative or commercially reared
bees, to native wild bees, is emerging as a significant
cause of native Bombus decline in North America
(see above).
Apis mellifera is widely used as a model organism in
studies of pesticide toxicity and is highly sensitive
to many insecticides.114,115 Honey bees, and likely
other bees as well, have relatively few detoxication
genes, which increases their susceptibility to pes-
ticides.116 Relative to honey bees, wild bees might
experience less pesticide exposure since they do not
forage as exclusively on agricultural crops. On the
other hand, native bees nesting near crops might
experience more exposure since they forage at times
of day and times of year when honey bees are not
present. While growers often reduce or avoid spray-
ing pesticides during periods of honey bee activity
there is less consideration for wild, native bees.117
Pesticide labeling, if it mentions bees at all, gener-
ally states that bees should be closed into their hive
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 173
Wild bee conservation Winfree
before spraying, which is obviously not relevant to
native species.
Few studies have compared pesticide toxicity in
non-Apis species to Apis, and the results have been
variable.118 Laboratory colonies of Bombus impa-
tiens fed spinosad-contaminated pollen at concen-
trations they are likely to encounter in the wild ex-
perienced few lethal effects, but showed impaired
foraging behavior.119 There are only a few field- or
landscape-scale studies of pesticide effects on native
bee abundance, and in meta-analysis they do not
show a significant negative effect (Ref. 73; but see
also Ref. 120, which shows significant negative ef-
fects of phosmet on the reproduction of a nonnative,
non-Apis bee). Clearly more studies of this topic are
Genetically modified crops
The effects of genetically modified (GM) crops on
bees were reviewed by Morandin.121 Crops modi-
fied for increased herbicide resistance account for
72% of global GM acreage, and this trait is unlikely
to negatively affect bees directly, although it could
affect them indirectly if higher herbicide use in GM
fields results in fewer floral resources for bees.121 In
contrast crops modified for insect resistance could
harm bees if the relevant proteins are both toxic
to bees and expressed in pollen. To date, 99% of
the commercialized insect-resistant GM crops have
contained genes for the insecticidal Bacillus thu-
rigiensis, which is not toxic to bees.121 Other types
of genetic sequences conferring insect resistance are
being developed, however, and should be tested on
both honey bees (which is generally done) and non-
Apis bees (which is rarely done) prior to commercial
Features of bees that affect their extinction
Genetic effects
The genetic effective population size (Ne), which
determines a population’s rate of loss of genetic di-
versity over time, is on average an order of magni-
tude smaller than the census population size (N).122
Bees probably have an even smaller Ne/N ratio than
most taxa because they are haplodiploid,123 and be-
cause their population sizes are highly variable over
time.124,125 At present there are too few studies of Ne
in bees to rigorously assess their Ne/N ratio. How-
ever, published values of Neeven for nonthreatened
bees in mainland habitats are low, relative to the
Neof 50–500 thought to be necessary to avoid in-
breeding effects and loss of evolutionary potential,
respectively:122 40–102,126 and 20.127 This suggests
that from a genetic perspective, bee populations are
even smaller than they appear.
In principle, haplodiploids might be able to purge
deleterious recessives through exposure in haploid
males, and thereby avoid the negative fitness conse-
quences that generally accompany reduced genetic
diversity.128,129 While haplodiploids suffer less from
inbreeding depression than diploids, inbreeding de-
pression is still substantial for them.130 The few
studies of inbreeding effects in bees show mixed
Another reason why bees as a group may be vul-
nerable to genetic decline is their complementary
sex determination system. Individual bees that carry
two different alleles at the sex-determining locus
develop as females whereas individuals with only
one allele, or two copies of the same allele, develop
as males. All unfertilized haploid eggs develop into
males. However if heterozygosity is low and a fertil-
ized egg is homozygous at the sex-determining lo-
cus, it will develop as a diploid male. Diploid males
are generally inviable or at least infertile. They there-
fore reduce population growth, making already ge-
netically impoverished populations even more vul-
nerable to the vortex of extinction associated with
negative genetic, demographic, and stochastic ef-
fects. Monte Carlo simulations suggest that bee pop-
ulations with diploid male production are an order
of magnitude more vulnerable to extinction than
are diploid populations, or even haplodiploid pop-
ulations without diploid male production.132 Some
studies have found high diploid male production
in wild bee populations,127,133 but others have not,
even in highly inbred populations.134
Bee species that are oligolectic or rare appear to
be more vulnerable to genetic effects. Oligolectic
bees have more genetically isolated populations and
lower genetic diversity,135–137 likely because their
distributions are limited by the distributions of their
host plants. Rare bee species also have more ge-
netic differentiation and/or smaller Neas compared
to common species.58,131,138 For example, popula-
tions of the rare Bombus sylvarum persisting in frag-
mented British habitats has Nevalues of only 21–72,
suggesting that they fall near or below the limit of ge-
netic viability (insofar as a Neof 50 is thought to be
174 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
necessary to avoid inbreeding effects;122). Similarly,
the rare B. muscuorum shows significant genetic dif-
ferentiation between populations only 3 km apart,
and in all populations at least 10 km apart, whereas
studies of widespread, common Bombus species do
not detect genetic differentiation even in popula-
tions separated by hundreds of kilometers.138
Social bee species may be particularly vulnera-
ble to genetic effects because for them Neis more
closely related to the number of nests than to the
number of individuals.139 This means that census
estimates, which are largely based on worker den-
sities, are likely to greatly overestimate Neand may
not even be correlated with it.140 Several recent syn-
thetic analyses have found that in bees, sociality is
associated with sensitivity to human disturbance. In
a meta-analysis of 54 published studies, the abun-
dance and species richness of social bee species is
significantly, negatively affected by human distur-
bance, whereas effects on solitary species are non-
significant.73 In a species-level analysis of 19 data
sets, social species are more sensitive to disturbance
and in particular to pesticide use.141 Across 23 stud-
ies of crop flower visitation by wild bees, visitation
rate declines more steeply with increasing distance
from noncrop habitat for social as compared with
solitary species.26 The cause of social bees’ increased
sensitivity is not known, although multiple mecha-
nisms can be postulated.26,73,141 Low Neand genetic
effects should be added to the list of possibilities.
Reliance on mutualist partners
Because bees are dependent on plants and vice versa,
it would seem logical that both are more vulnera-
ble to extinction, since the loss of one taxon leads
to the loss of the other.142–144 There is some evi-
dence for this. In intensively human-used regions,
declines in bees and the plants they pollinate are
positively correlated.49,61 Among animal-pollinated
plants, species that require outcrossing are more
sensitive to habitat fragmentation, suggesting a role
for mutualist loss in local extinctions.145 Models
and data for specialist herbivores and pollinators,
and for obligate body parasites for whom the host is
also the habitat, suggest that widespread extinction
of these groups could occur should hosts become
On the other hand, in terms of comparing the vul-
nerability of bees to other organisms, it is not clear
what the appropriate null is, given that most organ-
isms are dependent upon others in complex ways.
Most bee species are floral generalists,38,148 making
bees as a group less reliant on single mutualist part-
ners than are specialist herbivores or obligate body
parasites. Furthermore, plant–pollinator networks
have two features that might make them relatively
robust to species loss. First, the distribution of the
number of partners per species is highly skewed,
such that a minority of “core” species have many
partners and interact largely among themselves,
while most species have few partners.149 This makes
the network more robust to species loss in general,
although it is sensitive to the loss of the highly in-
teracting core species.149 Furthermore, core species
may be the most abundant species,150,151 in which
case they are less likely to go extinct. Second, plant–
pollinator networks are generally asymmetrical with
regard to specialization, meaning that specialist pol-
linators interact with generalist plants, and specialist
plants with generalist pollinators.152,153 Therefore,
the loss of a specialist from the system is unlikely to
result in the loss of its mutualist partner.
Last, many published studies may overestimate
the dependence of particular pollinators on par-
ticular plants due to undersampling. First, in many
sampling designs, rarity is confounded with special-
ization in that pollinators for which few specimens
were collected will of necessity be collected from a
small number of plant species. This bias can be cor-
rected with rarefaction or by using an appropriate
null model in analyses,150,151 but until recently most
investigators did not make this correction. Second,
most published studies are based on only 1–2 years
of data, but examination of long-term data shows
that pollinators visit different plants over time.154
For both of these reasons, many bee species may be
less dependent on particular plants, and therefore
more robust to plant species extinction, than has
been assumed.
Use of partial habitats
Bees require multiple resources tocomplete their life
cycle, including pollen,155 nectar, and nest substrates
and nest-building materials.156 These resources are
often gathered from different locations, making bees
reliant on multiple, “partial habitats.”157 This might
make bees vulnerable to disturbance insofar as they
would be negatively affected by the loss of any of
these habitats. On the other hand, if resources are
provided by the disturbed habitats themselves and
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 175
Wild bee conservation Winfree
bees are facultative in their use of such habitats,
bees might be less vulnerable to disturbance than
are other, more habitat specialist, taxa. For exam-
ple, bees use floral resources from both agricultural
and natural habitats in mosaic landscapes,158,159 and
models that incorporate this complementarity be-
tween habitat types have high explanatory value in
predicting bee abundance and species richness.160
Floral specialization
Dietary specialization is associated with a higher
extinction rate and/or with sensitivity to distur-
bance for a variety of nonbee taxa.161–166 Oligolectic
bee species gather pollen from a small number of
related flower species, whereas polylectic bees are
pollen generalists (even oligolects are dietary gen-
eralists for nectar;167). Oligolectic species probably
account for a large fraction of global bee diver-
sity, since they constitute about 30% of species in
temperate communities and up to 60% of species
in the more species-rich deserts.148 Oligolecty is
a significant predictor of bee species’ decline over
time in northwestern Europe,49 and of sensitiv-
ity to fragmentation in a desert ecosystem.94 Even
among European Bombus,allofwhicharepolylec-
tic, species with more specialized diets show greater
population declines over time.61 Presumably the
risk of decline is heightened by being more reliant
on a smaller number of food sources. In addition,
oligolectic bees have more genetically isolated pop-
ulations and lower genetic diversity (see above),
which further increases their susceptibility to
Other life history traits
Species that nest above ground, and species that use
previously established nest cavities, are more sensi-
tive to disturbance than are species that nest in the
ground or excavate their own nests.141 These species
may be more sensitive because they are more likely
to be nest-site limited. In contrast to other taxa,
body mass does not predict sensitivity to distur-
bance across bee species.141 Perhaps the lack of rela-
tionship is not surprising given contrasting predic-
tions about body size and extinction risk for bees.
For vertebrates, large body size is associated with
greater extinction risk.168,169 However for butter-
flies, the most mobile species have lower extinction
risk.170 Inbees,bodymassispositivelycorrelated
with mobility.171
Strategies for bee conservation
Formal protection of threatened species
Insect conservation generally lags far behind in-
sects’ functional and numerical importance, and
bees largely share the fate of other insects in this
regard. Insects account for an estimated 73% of
the animal species on earth,172 yet only 5–20% of
insect species have even been named, much less
had their natural history described.173 Only 70 in-
sect species have been recorded as going extinct to
date, but several lines of evidence suggest that this
number reflects our inadequate knowledge more
than it reflects reality.174 First, extinctions can only
be recorded for described species, and these are
likely to be the more common and widespread
species, which have lower probabilities of extinc-
tion as compared to undescribed species.175 Second,
most recent recorded insect extinctions are from
Lepidoptera, the best-studied insect order,174 which
constitutes only 15% of described insect species.172
Third, 78% of the recorded insect extinctions are
from the United States,174 which is high in tax-
onomic expertise but low in biodiversity, relative
to other nations. Even within the conservation re-
search community, there is a bias against insects:
despite accounting for 63% of all described species,
insects account for only 7% of the published papers
in leading conservation journals.176
Insects are underrepresented in species protec-
tion programs as well. Countries that have carefully
inventoried their insects find that at least 10% are
vulnerable or endangered,177 which would corre-
spond to at least 95,000 insect species being vulner-
able or endangered globally (based on the 950,000
scientifically described insect species globally;178 the
true number of threatened species might be an order
of magnitude greater). Yet only 771 insect species
have been evaluated for candidacy on the global
IUCN Red List—73% of which were subsequently
determined to be threatened.179 Even for inverte-
brates that achieve listing under the US Endangered
Species Act, the allotted funding per species is more
than an order of magnitude less than that received
by mammals and birds.180
Recent evidence from the few insect taxa that have
been monitored suggests that insects may be declin-
ing even more rapidly than better-studied taxa such
as plants and birds.181 Over the past 2–4 decades,
71% of British butterfly species declined, compared
176 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
to 54% of birds and 28% of plants.181 Over the past
35 years, 54% of British moth species have declined
significantly.182 Based on their rates of decline, 21%
of the moths in this study would be considered
threatened nationally according to IUCN criteria,
yet none is currently listed by the British Red Data
Bees share the fate of insects generally in being
poorly known and poorly protected, although the
estimated proportion of bee species that are scientif-
ically described is thought to be higher than for most
insect taxa (17,500 out of >20,000, or up to 88%;6).
Currently, no bee species is listed as threatened or
endangered under the US Endangered Species Act,
even though many species are known to be very
rare and/or steeply declining, or likely extinct.68,70,74
Similarly, two bee species are listed on the global
IUCN Red List.183
Economic reasons for conserving bees
Because bees provide valuable ecosystem services
the question arises to what extent economic argu-
ments alone can motivate bee conservation. The use
of economic, ecosystem-service-based arguments to
justify conservation is controversial. Some believe
that such arguments undermine the moral legit-
imacy of the conservation movement, which has
historically been based on ethical arguments kept
distinct from questions of economic gain.184 From
a practical standpoint, if conservationists adopt eco-
nomic arguments they could then find that in many
cases, the most profitable course of action is to con-
vert natural areas to human use. In addition, the
cost-benefit analysis of a given situation is likely
to fluctuate over time, with changing commod-
ity prices, property values, and alternative methods
of providing the ecosystem service in question,184
whereas biodiversity conservation is a long-term
commitment. On the other hand, even though
global estimates of the value of ecosystem services
have been criticized for their economic methodol-
ogy,185,186 by any accounting natural areas and the
species they harbor provide extensive and often un-
derappreciated services to humanity. It seems wise
to include these services when considering the rel-
ative merits of conservation versus alternative land
uses. Crop pollination services from native pollina-
tors have featured prominently in this debate.184,187
From an ecological point of view, two issues have
emerged as important challenges to valuing crop
pollination. (The economic aspects of valuing pol-
lination services are outside the scope of this re-
view but are covered elsewhere.188 ) First, in order
to estimate the economic benefit of a given level
of pollination, one must know how pollen deposi-
tion translates into fruit production. This requires
knowing not only pollen deposition per flower in
the field, and the dose-response curve for pollen de-
position versus fruit set per flower, but also the dose-
response curve for the number of flowers fully polli-
nated versus fruit set per plant or per unit area.19,189
Asymptotic fruit set at the plant or field scale may
be reached at lower levels of pollination than would
be estimated at the flower scale because many plants
produce more flowers than they can set into fruit,
even when resources are not limiting.189 Another
reason why changes in pollen deposition may not
translate into changes in crop production is that
production can be limited by other factors, such as
fertilization, pest or weed control, and available wa-
ter. Pollination will only have direct economic value
when it is the factor limiting production. Pollina-
tion limitation can be measured experimentally in
the field.190,191 Or if the pollination requirement of
the plant is known, pollination services can be val-
ued relative to this threshold, on the assumption that
pollination will be limiting at some point(s) across
space or time. In nature, 62–73% of plant popula-
tions show pollination limitation,190 and crops are
even more likely to be pollination-limited because
other potentially limiting factors such as sunlight,
soil fertility, pest and weed control, and water are
provided in abundance in most commercial agricul-
tural settings (although this point is debated;192,193).
Pollination can also not be limiting because it is al-
ready being provided by honey bees. Many studies
value native bee pollination independently of the
pollination provided by managed pollinators, but
methods for valuing the two simultaneously exist.188
A second critical issue for pollination service val-
uation is calculating not only the economic bene-
fit of conservation, but also its opportunity cost—
which in agricultural contexts generally means the
profits foregone by not converting native bee habi-
tat to crop production. One of the first empirical
studies of crop pollination service value found that
wild bees from forest fragments contributed $62,000
per year, or 7% of the farms’ annual income, to one
Costa Rican coffee plantation.194 Since the study was
conducted, however, the price of coffee fell and the
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 177
Wild bee conservation Winfree
plantation was converted to pineapple, which does
not require insect pollination, indicating the critical
role of commodity price fluctuations and opportu-
nity costs involving alternative land uses.184 When
tropical forest in Indonesia is valued for the pol-
lination services its resident bees provide to coffee
plantations, the result (46 per ha) is lower than
that found in the Ricketts et al.studybyafactorof
six.195 The authors attribute this difference to forest
fragmentation, in that Indonesian plantations are
surrounded by large blocks of forest, which reduces
the per ha value, whereas two forest fragments pro-
vided all the pollination services in Costa Rica.195
The economic optimum for pollinator habitat
conservation could be found by modeling the trade-
offs between ecosystem service provision to existing
crop fields, and the opportunity costs of forego-
ing alternative land uses, that is, converting nat-
ural habitat to crops. Two published models exist
for such situations. In Canada, canola seed set in-
creases with increasing wild bee abundance, which
is in turn a function of the amount of seminatural
habitat surrounding crop fields. The model predicts
that the economic optimum is reached when 32%
of the land area is left uncultivated;196 when changes
in land use were implemented experimentally, the
landowner found that the optimum was closer to
15%.121 In the most thorough evaluation to date of
the economic trade-offs between crop pollination
services and land use, Olschewski et al.197 calculated
the marginal loss curve for pollination services as
a function of forest loss for coffee plantations in
Ecuador and Indonesia. The authors included other
potential crops in addition to coffee as alternative
land uses, as well as subtracting the variable costs of
crop production from scenarios where production
was reduced. In all modeled scenarios, the economic
optimum involved deforestation, that is, the value
of pollination services was not sufficient to preserve
existing forests on economic grounds alone. The
value of forest conversion only equaled the value of
preservation when forests were almost gone.197
In sum, based on the limited research to date we
can’t conclude that the economic value of pollina-
tion services alone will provide sufficient incentive
for farmers to preserve native bee habitat in the long
term. This is even more likely to be the case when
a substitute for native bee crop pollination services,
namely pollination by managed honey bees, is added
to the equation. There will always be an element of
risk involved in relying on a single managed species
to pollinate all agricultural crops, and having native
bees available provides a valuable backup against
this risk. But farmers may not consider this insur-
ance value to be a sufficient reason to alter their
land use practices, when honey bee rental costs can
be more economical route to meeting current pol-
lination needs.
This does not obviate the need to evaluate the
pollination services provided by wild bees, and to
include their value in policy decisions. In order to
optimize land use decisions, it is essential to sum all
of the types of ecosystem services provided by the
same land area,198 and the economic value of wild
bee pollination remains an important component of
this summation. Even when data on other ecosystem
services are lacking, the value of crop pollination can
contribute significantly to decisions when the mul-
tiple benefits of conserving pollinators (not just the
economic benefits) are weighed against alternative
land uses.
Restoring bee communities
The context of bee restoration so far has been pre-
dominantly agricultural, likely because significant
governmental funding exists for pollinator restora-
tion on agricultural lands. Although the limited re-
search on pollinator restoration in natural areas is
regrettable from an ecological point of view, the
agricultural emphasis is potentially a powerful ap-
proach given that agriculture currently accounts
for 33% of global terrestrial land area,199 and an-
other billion ha will likely be converted to agricul-
ture by 2050 as crop production expands to feed
a growing human population.200 In addition, the
pollination services that bees can provide to crops
increases their suitability for agricultural restora-
tion programs and the appeal of such programs to
What factors limit bee population size?
In order to design effective restorations, it would be
useful to know what factor(s) most often limit bee
population size, so that these factor(s) could be re-
stored. The resources bees require to complete their
life cycle can be roughly divided into those related to
nesting (the appropriate substrate, such as bare soil,
stems, or cavities, and for some species the materials
necessary to create the nest interior, such as leaves
178 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
or resin), and those related to foraging on flow-
ers (pollen and nectar).157 As yet no experimental
restoration has evaluated the relative effectiveness
of restoring floral and nesting resources. However, a
number of studies have suggested that either floral
or nest site availability can limit bee reproduction or
population size. Population size of a floral special-
ist, Andrena hattorfiana, closely tracks the availabil-
ity of pollen resources provided by its host plant.201
The likelihood of this species being limited by flo-
ral resources is probably higher than average, how-
ever, because it is a ground-nesting floral specialist.
Another floral specialist, Dieunomia triangulifera,
shows evidence of population limitation by both
floral resources and other factors.202 Within a nat-
ural system of isolated mountain meadows, Bom-
bus colony reproduction is higher in meadows with
more floral resources.203
Two studies have provided nest sites experimen-
tally, and then examined the role of floral resource
availability in bee reproduction. The reproduction
of Osmia lignaria in agricultural landscapes ex-
ceeded replacement at sites where floral resources
were more available within the species’ flight dis-
tance, and was likely below replacement at sites with
fewer floral resources.158 Similarly, it took Osmia
caerulescens and Megachile versicolor twice as long
to provision their nests in fields with fewer floral re-
sources.204 This difference probably translates into
lifetime fecundity because solitary bees are thought
to continue provisioning nests until the end of their
lifetime.205 These studies provide weaker evidence
for the generality of floral resource limitation, how-
ever, since nest site limitation was at least partially
removed as a factor.
The one experimental study of nest site limitation
found that Osmia rufa populations increased by a
factor of 35 when nest sites were augmented.206 Ob-
servational data from a similar system also suggest
nest site limitation, in that old meadows similar in
floral resource availability have more wood-nesting
bees when old trees are present.207 In an applied
context, the provision of nest sites for Nomia me-
landeri and Megachile rotundata, which are used for
alfalfa pollination in the western USA, allows for
much larger population sizes than would otherwise
be present;21 however, floral resources are unlikely
to be limiting in this agricultural context. All but
one of the species reported above are cavity-nesting,
and their populations might be more often limited
by nest site availability than is the case for ground
Bee populations could be limited by other factors
such as predation or parasitism,57,208 or, at the egg
and larval stages of the life cycle, by fungal pathogens
in nests.35 For example, Bombus vagans workers
have a 14% chance per day of being attacked by
a crab spider (Thomisidae), and 13–20% of Bom-
bus workers are lethally parasitized by Conopid fly
parasitoids (reviewed in208). However, there is lit-
tle research on the overall importance of these fac-
tors to population growth for wild bee species. The
one experimental study to measure parasitism as a
function of bee nest density found little evidence of
top-down regulation; in fact, parasitism was inverse-
density dependent in most years.206 In any event, it
is not clear how to control parasites and predators
within a restoration context.
Floral restorations
Pollinator restoration to date has focused on restor-
ing floral resources within an agricultural context.
ported by evidence that large-scale declines in for-
age plants are associated with large-scale declines in
pollinators, particularly for Bombus species,49,60,61
and by the studies of bee reproduction and floral
resources, although nest site restoration may also be
critical and merits further study.
A critical element of restoration plantings for pol-
linators is the choice of plant species to include in the
mixes. Mixes ought to include plant species that in
combination provide a long period of bloom, and
are preferred by a diverse pollinator community.
Relatively few studies have used quantitative infor-
mation to determine the best species; however, ef-
forts are progressing in that direction. In the United
Kingdom, bee preference has been studied primar-
ily by comparing bee visitation to the different
restoration protocols available to farmers through
government-subsidized restoration programs. Not
surprisingly, bees prefer planting mixes that are
specifically designed to produce flowers, as com-
pared to grass-based restoration protocols, or less
intensively managed crop areas.209–212 The relative
attractiveness of floral planting mixes and natural
regeneration varies across studies; however, natu-
ral regeneration often involves agricultural weeds
that can be more acceptable to pollinators than
to farmers.209–211,213 In the United States, far less
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 179
Wild bee conservation Winfree
research on restoration protocols has been done.
Bee preference for different flowering plant species
suitable for agricultural restorations has been exper-
imentally and/or statistically tested, and then incor-
porated into restoration protocols, only in Califor-
nia,214 Michigan215 and New Jersey.216
An important finding to emerge from studies of
floral restorations is that often only a few plant
species are responsible for the great majority of
bee visits.64,209,210,212 This suggests that restorations
can be made more efficient and cost-effective by
focusing on a subset of highly attractive species,
rather than simply increasing floral diversity. Unfor-
tunately for North American restoration ecologists,
most of the key bee plants so identified in EU stud-
ies are exotic weeds in North America, highlighting
the need for analogous research on this continent.
In addition, studies of the entire bee community are
needed, as most research to date has considered only
A limitation of many studies assessing which
flowers are attractive to bees is that they are based
on use rather than preference. When a field re-
searcher surveys bees visiting different flowering
plant species, the plant receiving the greatest num-
ber of bee visits could achieve this through being
preferred by bees (the variable that researchers seek
to assess) and/or because its flowers are more abun-
dant than those of other plant species (a statistical
outcome not relevant to bee preference). Preference,
as opposed to use, can be calculated from observa-
tional data on both bee visitation rates and floral
abundance,62,217,218 or in experiments in which the
different plant species are offered simultaneously at
standard densities.215
Nest site restoration
Although nesting resources may be critical in de-
termining bee densities, this aspect of bee restora-
tion has received less attention than have floral re-
sources. There is limited information on the micro-
habitats preferred by nesting bees. British Bombus
queens nest-searching in agricultural habitats pre-
fer sites with banks or tussocky vegetation,219 and
Swedish Bombus queens prefer tussocks or withered
grass.220 Guidelines for creating nest sites for dif-
ferent types of bees are available from the Xerces
Society.214 Studies of the relative efficacy of restor-
ing different types of bee nests sites, analogous to the
comparisons done for floral resources, and studies
of the population-level consequences of nest site
restoration, are greatly needed.
The farm bill and agri-environment schemes
Bee restoration on agricultural lands has taken place
largely within the United States and the EU, both of
which have significant funding in place for such pro-
grams. In the United States, federal funding for habi-
tat restoration on agricultural lands is channeled
largely through the Farm Bill (formally the Food,
Conservation, and Energy Act) and administered at
the state level by the Natural Resource Conservation
Service and the Farm Service Agency. Government
spending for Farm Bill conservation programs av-
eraged $3.5 billion per year from 2002–2007.221,222
Although “conservation” is broadly defined within
the Farm Bill to include many goals in addition
to biodiversity conservation, Farm Bill funding still
dwarfs many forms of government funding for con-
servation on nonagricultural lands. For example,
in 2003 only $0.8 billion was spent on the con-
servation and restoration of all 1335 threatened
and endangered species listed under the Endan-
gered Species Act—none of which was a bee (http:// The
Farm Bill offers around a dozen programs in which
landowners can voluntarily enroll to receive finan-
cial benefits for restoring habitat, primarily on for-
merly agricultural lands. Many of these programs
are suitable for bees; furthermore, the 2008 version
of the Farm Bill explicitly prioritized pollinators as
a target for restorations.223
In the EU, government-sponsored agricultural
land conservation falls largely under the aegis of
agri-environment schemes (AES), for which annual
funding in 2003 was 3.7 billion.55 Participation in
AES programs is mandatory for EU counties un-
der the Common Agricultural Policy.212 As of 2005,
AES cover roughly 25% of the farmland in the 15
older EU countries.55 AES offer farmers many op-
tions for which they are compensated financially,
including restoring habitat on buffer areas or set-
aside fields, and/or farming in-production fields less
Given the large amount of taxpayer money being
spent on agricultural habitat restoration, and the
increasing role of pollinators in such programs, a
critical question is whether these programs are ef-
fective in restoring pollinators. In the United States
little research has been done on this issue. Of the
180 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
Farm Bill programs, the Conservation Reserve Pro-
gram (CRP) is the largest, with roughly 4% of na-
tional cropland area being enrolled.224 Historically,
the goals of the CRP program have been control-
ling erosion and agrochemical runoff, as well as
regulating crop production volume. More recently,
the goals of carbon storage and habitat creation
for birds has been emphasized.225 In 2008 pollina-
tors became a high priority wildlife taxon for CRP
projects.223 Butterflies benefit from CRP restora-
tions, but there have not yet been any studies of
CRP effects on bees.226 The practice of sowing CRP
restorations with nonnative grasses is widespread225
and likely diminishes the value of these habitats for
The two Farm Bill programs most suited to polli-
nator restoration are the Environmental Quality In-
centives Program (EQIP) and the Wildlife Habitat
Incentives Program (WHIP). EQIP is the second-
most funded program, after the CRP, and its goals
include both improving the environmental qual-
ity of lands associated with livestock production,
and habitat restoration for wildlife on agricultural
lands.224 WHIP receives less funding, but unlike
other Farm Bill programs it is focused exclusively
on wildlife habitat.224 Both EQIP and WHIP can
reimburse private landowners for up to 75% of the
costs of restoring wildlife habitat. As of 2008, polli-
nators are a priority taxon for EQIP restorations,223
and pollinators are prioritized in some states (e.g.,
New Jersey) for the WHIP program as well. There
are currently no published studies of the effec-
tiveness of EQIP or WHIP protocols in restoring
bees or other pollinators, although a study of EQIP
pollinator restorations is in progress (C. Kremen,
Unpublished data).
In the United Kingdom and Europe there is
a larger base of research on the effectiveness of
government-sponsored agricultural programs AES
in restoring biodiversity in general, as well as pol-
linators in particular. Biodiversity is one of several
stated goals of AES, with the others including the
historical and esthetic value of landscape preserva-
tion, and improving soil and water quality.227 The
first quantitative assessment of the broad-spectrum
biodiversity benefits of AES found no significant
effect on target taxa, although there were weak pos-
itive effects on nontarget taxa including bees.228 A
meta-analysis 2 years later concluded that only 54%
of 62 studies comparing AES and non-AES fields
found significant biodiversity effects of AES.229 A
common experimental design flaw was noted that
could artificially inflate the perceived benefits of
AES: the locations chosen for AES enrollment may
have higher biodiversity prior to AES implemen-
tation, as growers often choose fields that are less
suitable for intensive agriculture to begin with.55,229
Furthermore, AES predominantly benefitted com-
mon species that may be in less need of protection
than rare species55—although it is important to note
that AES were not designed to benefit rare species,
which may be absent from agricultural habitats in
the first place.230 The findings on inconsistent bio-
diversity benefits have had a significant political im-
pact given the large amount of government funding
spent on AES programs.227
AES management significantly benefits bee com-
munities as compared with conventionally managed
controls in about half of the studies done to date,
consistent with the mixed biodiversity benefits re-
ported for other taxa. The increase in Bombus ter-
restris colony weight, a proxyfor reproduction, is not
significantly different between colonies placed on
conventional farms and those placed on farms with
AES-types restorations.231 AES management in in-
tensively farmed Dutch landscapes significantly in-
creases bee species richness, although the bee fauna
was poor throughout the study with only three
species recorded.228 Swisshaymeadowsenrolledin
AES have significantly greater bee abundance and/or
species richness than do conventionally managed
hay fields.55,232,233 In England, bee abundance is
significantly higher in fields with 6-m wide grass
margins, as compared to fields without margins.234
Various other forms of AES management in three
other studies done in Spain, the Netherlands and
the United Kingdom, however, show no significant
benefit to bees.55 In terms of the benefits they re-
ceive from AES restorations, bees appear about av-
erage relative to other taxa that have been studied
(Table 1).
Organic farming as a method
for restoring bees
Although the exact requirements for organic farm-
ing certification differ by country, all are based on
guidelines issued by the International Federation
of Organic Agriculture Movements, and involve
foregoing synthetic fertilizers, pesticides, and herbi-
cides.235 In the United States, biodiversity standards
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 181
Wild bee conservation Winfree
Tab l e 1. Rank of bees relative to other taxa examined in the same study in terms of response to agri-environment
scheme field-scale habitat restoration protocols. 1 =most positive response, 5 =least positive response
Study design Rank of bees Other taxa studied Reference
Paired comparison of AES
versus control fields;
outcome =species
1.5 (tied with
of 4
Plants, hoverflies, birds (Kleijn et al.228 )
Paired comparison of AES
versus control fields;
outcome =species
4 of 5 Plants, orthoptera,
spiders, birds
(Kleijn et al.55), Spain
Paired comparison of AES
versus control fields;
outcome =species
2 of 5 Plants, orthoptera,
spiders, birds
(Kleijn et al.55),
Paired comparison of AES
versus control fields;
outcome =species
4 of 5 Plants, orthoptera,
spiders, birds
(Kleijn et al.55), UK
Paired comparison of AES
versus control fields;
outcome =species
4 of 5 Plants, orthoptera,
spiders, birds
(Kleijn et al.55),
the Netherlands
Paired comparison of AES
versus control fields;
outcome =species
3 of 4 Plants, grasshoppers,
(Knop et al.232)
Paired comparison of
fields with and without
6-m grass margin strips;
outcome =abundance
and/or species richness
In top 3 of 6 Plants, grasshoppers,
spiders, carabid beetles,
(Marshall et al.234)
were added to the organic certification program ad-
ministered by the USDA in 2009 (Eric Mader, Xerces
Society, There is also a suite of farm
characteristics associated with organic farming but
not required for organic certification. When com-
pared with conventional farms, organic farms of-
ten have smaller field sizes, greater crop diversity,
greater area of seminatural or fallow habitat, and
higher abundance and diversity of weedy flowers,
and these features may be important in supporting
bees.236–238 Recent reviews have found that organic
as compared to conventional farming generally sup-
ports greater biodiversity across a range of nonbee
taxa, with plants being the most strongly benefit-
ted.236,239 At the time these reviews were done, there
were too few studies of bees to assess bees’ response
as a taxon.
Organic farming might be expected to benefit
bees, first due to reduced insecticide use, and sec-
ond because reduced herbicide use can lead to a
greater abundance and diversity of floral resources.
On the other hand, some pesticides used by or-
ganic farmers are highly toxic to bees, and the in-
creased tillage that organic farmers often use as a
replacement for herbicides can destroy nests
of ground-nesting species. Studies investigating
182 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
Tab l e 2. Studies comparing the abundance, species richness, and/or reproduction of wild bees as a function of farm
management (organic vs. conventional)a
Study design Result Significance level Reference
Wild bee species richness in
winter wheat fields
Higher species richness in
organic fields
∗∗∗ (Clough et al.235 )
Wild bee pollination services
to watermelon
No difference NS (Kremen et al.243)
Wild bee visitation rate to
four crops
No difference NS (Winfree et al.91)
Wild bee abundance in canola
Greater abundance in organic
∗∗∗ (Morandin and Winston242)
Wild bee abundance and
species richness in fallow
strips near organic versus
conventional winter wheat
Greater abundance and
species richness near
organic fields
(Holzschuh et al.241)
Wild bee species richness in
winter wheat fields
Higher species richness in
organic fields
∗∗∗ (Holzschuh et al.240)
Reproduction of a solitary
bee, Osmia lignaria
Higher reproduction on
organic farms, but only in
landscapes lacking natural
(Williams and Kremen158)
aSeveral studies finding positive effects are not independent because they were done at the same sites (Clough et al.235;
Holzschuh et al.240; Holzschuh et al.241 ). NS P>0.10, P0.05, ∗∗P0.01, ∗∗∗ P0.001
changes in wild bee communities and/or polli-
nation services as a function of farm manage-
ment have obtained mixed results. Bees are sig-
nificantly more abundant in and near organic as
compared with conventional winter wheat fields in
Germany,235,240,241 and also in organic as compared
with conventional canola fields in Canada.242 Asoli-
tary bee, Osmia lignaria, provisions significantly
more nest cells on organic as compared to conven-
tional farms when farms are set within agriculturally
intensive landscapes.158 However, the difference is
not significant when farms are surrounded by more
natural and seminatural habitat cover, because in
that case the bees can forage outside of the farm and
are not so dependent on local farm management.158
Organic farming has no effect on native bee pollina-
tion services to watermelon in California,243 or on
wild bee abundance on several crop plants in New
Jersey and Pennsylvania91 (Table 2).
Several studies have partially separated the com-
ponents of organic farming to better isolate the vari-
ables that affect bee communities. In one of the stud-
ies finding no significant benefit of organic farming,
conventional and organic farms were distinguished
only by the criteria for organic certification (use of
synthetic fertilizers, herbicides, and pesticides); the
two classes of farms did not differ in other vari-
ables often associated with organic farming, includ-
ing field size, crop diversity, or weedy flower abun-
dance or species richness.91 The lack of significance
in this study suggests that the habitat heterogene-
ity often associated with organic farming may be
more important to bee communities than organic
certification per se, as is the case for some nonbee
taxa.244–246 Wild bees may be particularly benefitted
by weedy flowers and a variety of crops that provide
forage for a longer period, since few bee species have
flight seasons short enough to be supported by a
single monoculture crop.241 In contrast, insecticide
use has had surprisingly weak effects on wild bee
communities in the small number of studies that
have explicitly quantified this factor.241,243 This
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 183
Wild bee conservation Winfree
Figure 1. A hypothesized interaction between landscape
context and the effectiveness of AES restorations. Cleared
landscapes are defined as <1% noncrop cover, simple
as 1–20%, and complex as >20%.Notetheshapeof
the curve would be strongly asymmetrical if the X-axis
values were evenly spaced. The hypothesis suggests that
the benefits of a given restoration effort will be greatest
in landscapes that are already highly agricultural. From
Tsch a r n t k e et al.247 ; used with permission.
similarly suggests that organic farming require-
ments per se may be less important than other land
use practices associated with organic farming. A fi-
nal consideration is that the effectiveness of organic
farming may be contingent on the larger landscape
surrounding the farm (see below).
Where should restorations be done?
Restoration of bee habitat within agricultural land-
scapes is generally done at small scales, ranging from
2–6 m buffer strips to fields of a few ha. Where
should such restorations be done, in order to maxi-
mize their effectiveness? Tscharntke et al .247 hypoth-
esized an asymmetrical, hump-shaped relationship
between landscape heterogeneity and restoration ef-
fectiveness (Fig. 1), such that restorations are less
effective when done in heterogeneous landscapes
(defined as <80% cropland) where pollinators are
present without restorations, most effective in in-
termediate landscapes (defined as 80–99% crop-
land), and less effective in homogeneous landscapes
(defined as >99% cropland) where pollinators are
largely extirpated and few sources of colonists for
restorations exist.
Several studies have since tested the relationship
between local- and landscape-scale factors and have
confirmed that the two interact, and that the ef-
fectiveness of local bee restorations increases con-
sistently with increasing cover of cropland (which
most authors have interpreted as arable, i.e., row
crops) in the surrounding landscape. As yet no
study has tested the hypothesis that effectiveness de-
clines in the most intensively managed landscapes
(>99% cropland). In a system where all sites are
set within highly heterogeneous landscapes (<40%
arable cropland), neither local- nor landscape-scale
factors explains crop visitation by native bees; rather,
native bees are abundant throughout the entire sys-
tem.19,91 This is consistent with the hypothesis that
in highly heterogeneous landscapes, bees are sup-
ported by the landscapes themselves and restora-
tion is not required. In a system where the propor-
tion of arable cropland in the landscape varies from
20–95%, bumble bee density in restored patches in-
creases more than linearly with increasing arable
crop cover.248 Similarly, there is an interaction be-
tween bee species richness in organic versus con-
ventional wheat fields and surrounding land cover,
such that the organic/conventional difference in-
creases with the proportion of arable croplands over
a range of roughly 20–85%.240 Last, the reproduc-
tion of a solitary bee species is similar on organic
and conventional farms when both are near patches
of seminatural habitat, but diverges on farms set
within intensively agricultural landscapes.158 Stud-
ies of nonbee taxa have also found that the benefit
of organic farming is greatest in the most inten-
sively agricultural landscapes.249,250 These studies
are broadly consistent with the work finding that
the economic value of pollination services provided
by natural habitat outweighs the value of land con-
version only in the most degraded landscapes (see
Restorations can also be accomplished by reduc-
ing the intensity of a single land use variable, in
which case the biodiversity gains can be plotted
against land use intensity as a bivariate relation-
ship. The steepness of the resulting slope indicates
where biodiversity gains are greatest for a given
incremental change in land use intensity (Fig. 2).
A study of plant species richness and nitrogen in-
puts (a proxy for land use intensity) shows that the
benefits of reducing nitrogen inputs are greatest in
the least intensive systems54—the opposite of the
conclusion reached by the studies of organic farm-
ing and arable crop cover reviewed above. In reality,
184 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
Figure 2. The relationship between plant species rich-
ness (per 100 m2) and annual nitrogen input (a proxy for
land use intensity) on agricultural grasslands in Europe.
Curved lines indicate the best fit that was found with a
curvilinear function; straight lines resulted from a less
explanatory linear function. The biodiversity benefits of
reducing N inputs by a given amount will be greatest
where the curve is steepest, in the least intensively farmed
landscapes. From Kleijn et al.54; used with permission.
the optimal location for a restoration is determined
not only by relative benefits, as in Figure 2 or the
studies of organic farming above, but also by rel-
ative costs. This full cost-benefit approach has not
yet been applied to the question of what landscape
context offers the best restoration value.
The cost-benefit approach has been used for a
larger-scale question: whether biodiversity conser-
vation and restoration should be focused on agri-
cultural lands at all. In an influential paper, Green
et al.251 contrasted two approaches to biodiver-
sity conservation: wildlife-friendly farming, which
involves integrating conservation into agricultural
landscapes through, for example, AES and Farm Bill
restorations; and sparing land for nature, which en-
tails concentrating agricultural production in high-
intensity, low-biodiversity areas while protecting
more natural areas elsewhere for biodiversity. Green
et al. propose that the relative efficacy of these two
approaches can be evaluated by considering how
rapidly agricultural yield declines when wildlife-
friendly farming is implemented—specifically, by
plotting the density of a given species of conserva-
tion concern against agricultural yield. If this curve
is concave, then wildlife-friendly farming is pre-
dicted to be the best conservation approach, because
species declines are slower than yield increases as
agricultural intensification increases (Fig. 3A). Con-
versely, if the curve is convex, then intensive agricul-
ture combined with land sparing is predicted to be
Figure 3. Two species density versus agricultural yield
relationships that lead to different conservation strate-
gies. (A) When species density decreases slowly with ini-
tial increases in yield, wildlife-friendly farming can be an
effective conservation approach. (B) Conversely, when
species density decreases rapidly at low levels of yield
increase, land sparing is predicted to be the best conser-
vation approach. After Green et al.251
the best approach because species declines are rapid
even when yields are low (Fig. 3B). Note that Green
et al.251 compare the shape of the biodiversity–yield
relationship across entire study systems to iden-
tify the optimal system for conservation projects
(Fig. 3), whereas Kleijn et al.54 seek the optimal
location for restoration within a given system by
finding the area with the steepest slope (Fig. 2).
sumed by the model of Green et al.,251 then greater
yields will tautologically lead to less land area be-
ing used for agriculture because yield is defined
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 185
Wild bee conservation Winfree
as food production per unit area. However, on a per
capita caloric basis enough food is already produced
globally, which suggests that factors other than the
need for food, such as distribution inequities, are
driving agricultural land conversion.252–254 Two ad -
ditional factors make it difficult to evaluate the
relative effectiveness of the wildlife-friendly farm-
ing and land sparing approaches. First, empirical
density-yield relationships of the type shown hy-
pothetically in Fig. 3 are not yet known for any
species.251,255 Although relationships are generally
negative for the few taxa that have been investi-
gated,256,257 the shape of the relationship is not
clear. In addition, the extent to which biodiversity-
friendly agriculture reduces crop yields is contro-
versial. Restorations that take land out of produc-
tion presumably reduce yields, but the transition
to organic farming can either reduce or increase
yield.254 Organic farming is, however, more expen-
sive, which suggests that another variable—the cost
of production—should be considered in the cost-
benefit analysis.
Second, there is as yet little evidence that us-
ing land for intensive agriculture leads to sparing
land for nature elsewhere.258 Yield and deforesta-
tion rates can be negatively correlated,255 but this is
not necessarily a causal relationship. At a local scale,
both agricultural yields and the extent of land un-
der production can be limited by the same factors—
capitalization and technology—such that when lim-
its on yield are removed, it becomes profitable for
farmers to farm more land, not less.259
Differences between developed temperate and de-
veloping tropical systems need to be kept in mind
when comparing among approaches to conserva-
tion and restoration Agricultural expansion over the
next few decades is predicted to occur largely in the
developing world.255 Yetwhatweknowaboutbee
restoration through AES-type approaches is based
largely on northwest Europe, which is one of the
most agriculturally developed areas of the world.54
Tropical bees that have only recently encountered
agriculture may be less robust to it and in greater
need of land-sparing approaches, as compared to the
bee fauna that persists in areas with a long history of
agricultural land use. Last, in terms of global con-
servation planning it is important to keep in mind
that the per area costs of conservation in USA and
UK, including AES-type restorations, are among the
highest in the world.33
For pollinators specifically, several factors weigh
in favor of focusing restoration on agricultural
lands. First, significant funding for such restora-
tions already exists, at least in the EU and USA,
whereas less funding is currently available for nona-
gricultural restorations. Second, ecosystem services
arguments for pollinator conservation are most rel-
evant in agricultural areas. And third, agricultural
systems have the potential toprovide suitable habitat
for at least some bee species. One study has quantita-
tively evaluated how AES restorations might affect
both bee biodiversity and crop yield. Based on a
study of bees in winter wheat fields, an increase in
organic farmland from 5% to 20% is predicted to
increase the species richness of bees in fallow strips
by 50%, and the abundance of solitary bees by 60%
and of bumble bees by 150%.241 These benefits can
be compared to the 40% decrease in yield (kg/ha
of wheat) incurred by changing from conventional
to organic agriculture.240 In this study, 100% of the
bee species were polylectic, indicating that the di-
etary specialists, which may be in the greatest need of
conservation, have likely been lost from the system
already.241 This serves as an important reminder
that only a subset of bees, namely those found in
agricultural settings, are benefitted by agricultural
Do bee restorations restore ecosystem
services to crops?
This is an important question about which we know
surprisingly little. Restoration protocols that re-
store pollinator biodiversity may not restore ecosys-
tem services, and vice versa, because a small subset
of species commonly provide the majority of the
ecosystem services (e.g., Ref. 260). For example, sin-
gle, common bumble bee species provided 49% of
the pollination services to watermelon, out of 46
native bee species found pollinating the crop (Fig.
4;19). It may be that agricultural habitat restoration
programs, which tend to protect common species,55
may be effective for the restoration of ecosystem ser-
vices even if they are not effective for the conserva-
tion of biodiversity. It is striking, given the poten-
tial benefits of agricultural pollinator restorations to
crop pollination, that no published study has inves-
tigated this question using actual cropping systems.
A study investigating the restoration of pollination
services to crops as a function of habitat restora-
tion has been in progress in California since 2006
186 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
Figure 4. Fraction of all watermelon pollen grains de-
posited on female flowers by the different components
of the native bee community. “Small bees” included 28
species, “other large bees” 12 species, and “green bees”
5 species. Data averaged across the 23 farms reported in
Winfree et al.19
but is not yet completed (C. Kremen, Unpublished
data). Several studies have shown the potential for
crop pollination benefits by monitoring potted phy-
tometers or noncrop plants situated near pollinator
restorations. Seed set is higher in AES versus control
(conventional) hay meadows for 2 of 3 potted, non-
crop plant species.233 Seed set of potted phytometers
levels found within 100 m of the restoration; how-
ever, the difference was not significant.261
Bee restoration outside the agricultural context
Given the fact that restorations generally focus on
the vegetative community, yet plants and pollina-
tors are interdependent, it is important to know
to what extent pollinator restoration follows nat-
urally from vegetative restoration. There is only
one published study of nonagricultural bee restora-
tion, which found that bee communities on an-
cient and restored British heathlands were similar
in species richness and dominant species identity.
Species composition was not similar between an-
cient and restored sites, but composition was harder
to assess as it also varied across sites within a restora-
tion class and across time,262 as it typical of bee com-
munities.124 In California, remnant riparian frag-
ments and vegetatively restoredsites have similar bee
abundance and species richness, but species com-
position differs significantly between the restored
and control sites.263 In particular, ground-nesting
species and floral generalists were more abundant at
the restored sites. Pollination function also differed
in that native plant species received fewer visits from
native bees at restored sites.263
Summary and conclusions
Striking gaps in our knowledge of bee conservation
and restoration became apparent in the process of
writing this review. The following topics are partic-
ularly in need of scientific attention.
First, there is a great need for monitoring of bee
populations to provide information about long-
term population trends. Data from regions other
than northwest Europe, and genera other than Bom-
bus, are particularly needed. We also need studies
that assess how different bee species are affected by
land use change, so that conservation planners can
prioritize the needs of the most sensitive species,
while not basing conservation programs on the bee
species that do well in disturbed areas. Studies from
both temperate and tropical systems show that even
when aggregate bee abundance and species rich-
ness are not negatively affected by land conversion,
species composition can change dramatically,85,88,94
indicating that species-level analyses are important.
Last, there are almost no studies of bees and climate
change and these are clearly needed.
Second, we lack basic information about the pop-
ulation biology of bees. To my knowledge, a life ta-
ble analysis or population viability analysis (PVA)
has not yet been done for any bee species. Soli-
tary bees have unusually low fecundity for an insect,
with studies reporting 2–30 eggs or offspring per
female lifetime.202,205,264,265 Presumably, this means
that survivorship rates for juveniles and/or adults are
unusually high. Studies that measure these rates and
then perform sensitivity analyses to assess which life
stages most strongly determine population growth
rate would enable conservation plans to focus on
the most critical aspects of bee biology. PVA could
determine the population sizes necessary for bee
species persistence as well as the land area required
for reserves.
Third, we need to know what factors most often
limit bee populations. This is a challenging question,
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 187
Wild bee conservation Winfree
but could perhaps be addressed in the restoration
context using experimental additions of nest sites
and/or floral resources. A related point is that we
need more studies that measure bee reproduction
as opposed to merely forager density. Forager den-
sity, reproduction, and genetic population size can
be uncorrelated, at least for social bees.110,140,266 An-
other problem is that bees are generally sampled at
flowers, or using floral mimics such as pan traps, yet
bees assess the attractiveness of a given flower patch
relative to the alternative floral resources available
in the larger landscape. Because researchers rarely
have data on all the alternatives, this behavior can
make studies based on forager density alone difficult
to interpret, and even lead to erroneous outcomes
such as concluding that bee abundance is highest
in degraded landscapes, when actually the relative
attractiveness of a standardized flower patch is high-
248,261 Yet the great majority
of published studies measure forager density as the
outcome variable.
Fourth, given the funding and effort going into
Farm Bill and AES-type restorations, we need more
research evaluating the effectiveness of these restora-
tions. I am not aware of any published studies of
the efficacy of Farm Bill restorations in restoring
bees. The research on AES restorations is strongly
dominated by studies from the United Kingdom
and the Netherlands, which presents a scope of
inference problem insofar as these are among the
most human-dominated agricultural landscapes in
the world.54 Determining the relationship between
agricultural yield and bee density would also be very
useful as it would allow land managers to use the
model of Green et al.251 to identify the optimal lo-
cation for bee restorations.
Fifth, the success of habitat restorations in restor-
ing ecosystem services to crops has not yet been
studied and would provide important information
for conservation planning and policy. Last, there
are only two studies of bee restoration in nonagri-
cultural settings. The paucity of studies makes it
clear that much more work is needed in order to
understand the restoration ecology of this critical
functional group.
I thank Neal M. Williams, Scott Hoffman-Black,
Daniel P. Cariveau, and William H. Schlesinger
the manuscript, Bryan N. Danforth and Stuart
Roberts for sharing their knowledge of bee bio-
geography and bee conservation in the EU, re-
spectively, and Wilhelmenia Ross for bibliographic
Conflicts of interest
The authors declare no conflicts of interest.
1. Linder, H.P. 1998. Morphology and the evolution of
wind pollination. In Reproductive Biology.S.J.Owens
& P.J. Rudall, Eds.: 123–135. Royal Botanic Gardens,
Kew. R ichmond , UK.
2. Bawa, K.S. 1990. Plant-pollinator interations in tropical
rain forests. Annu. Rev. Ec ol. System.21: 399–422.
3. Neff, J.L. & B. B. Simpson. 1993. Bees, pollination sys-
tems and plant diversity. In Hymenoptera and Biodi-
versity. J. LaSalle & I.D. Gauld, Eds.: 143–167. CAB
International. Wallingford, UK.
4. Anderson, S. 2003. The relative importance of birds and
insects as pollinators of the New Zealand flora. NZ J.
Ecol.27: 83–94.
5. Kearns, C.A. 2001. North American dipteran
pollinators: assessing their value and conserva-
tion status. Conservation Ecology. 5: http://www.
6. Michener, C. 2007. The Bees of the World, 2nd edition.
Johns Hopkins University Press. Baltimore and Lon-
7. Klein, A-M. et al. 2007. Importance of pollinators in
changing landscapes for world crops. Proc. R. Soc.
Lond., Ser. B. 274: 303–313.
8. Free, J.B. 1993. Insect Pollination of Crops, 2nd Edition.
Academic Press. London.
9. National Research Council. 2007. Status of Pollinators in
North America. The National Academies Press. Wash-
ington, DC.
10. Oldroyd, B.P. 2007. What’s killing American honey
bees? Plos Biol.5: 1195–1199.
11. Aizen, M.A. & L.D. Harder. 2009. The global stock of
domesticated honey bees is growing slower than agri-
cultural demand for pollination. Curr.
12. Aizen, M.A. et al. 2008. Long-term global trends in
crop yield and production reveal no current pollination
shortage but increasing pollinator dependency. Curr.
Biol.18: 1572–1575.
188 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
13. Winfree, R. 2008. Pollinator-dependent crops: an in-
creasingly risky business. Curr. Biol.18: 968–969.
14. Javorek, S.K. 2002. Comparative pollination effective-
ness among bees (Hymenoptera: Apoidea) at lowbush
blueberry (Ericacea: Vaccinium angustifolium Ait.).
Ann. Entomol. Soc. Am.95: 345–351.
15. Richards, K.W. 1996. Comparative efficacy of bee
species for pollination of legume seed crops. In The
Conservation of Bees,Vol.No.18.A.Matheson,S.L.
Buchmann, C. O’Toole, P.Westrich, I.H. Williams, Eds.:
81–103. Academic Press. London.
16. Parker, F., S.W.T. Batra & V.J. Tepedino. 1987. New
pollinators for our crops. Agric. Zool. Rev. 2: 279–304.
17. Heard, T.A. 1999. The role of stingless bees in crop
pollination. Ann. Rev. Entomol.44: 183–206.
18. Kevan, P.G., E.A. Clark & V.G. Thomas. 1990. Insect
pollinators and sustainable agriculture. Am. J. Altern.
Agric.5: 12–22.
19. Winfree, R. et al. 2007. Native bees provide insurance
against ongoing honey bee losses. Ecol. Lett. 10: 1105–
20. Kremen, C., N.M. Williams & R.W. Thorp. 2002. Crop
pollination from native bees at risk from agricul-
tural intensification. Proc. Natl. Acad. Sci.99: 16812–
21. Bohart, G.E. 1972. Management of wild bees for polli-
nation of crops. Annu.Rev.Entomol.17: 287–312.
22. Torchio, P.F. 1991. Bees as crop pollinators and the
role of solitary species in changing environments. Acta
Horticulturae 288: 49–61.
23. Richards, K.W. 1993. Non-Apis bees as crop pollinators.
Revue Suisse De Zoologie. 100: 807–822.
24. Klein, A.-M., I. Steffan-Dewenter & T. Tscharntke. 2003.
Fruit set of highland coffee increases with the diversity
of pollinating bees. Proc.R.Soc.Lond.Ser.BBiol.Sci.
270: 955–961.
25. Klein, A.-M., I. Steffan-Dewenter & T. Tscharntke. 2003.
Pollination of Coffea canephora in relation to local and
regional agroforestry management. J. Appl. Ecol.40:
26. Ricketts, T.H. et al. 2008. Landscape effects on crop
pollination services: Are there general patterns? Ecol.
Lett. 11: 499–515.
27. Greenleaf, S.S. & C. Kremen. 2006. Wild bees enhance
honey bees’ pollination of hybrid sunflower. Proc. Natl.
Acad. Sci.103: 13890–13895.
28. Chagnon, M., J. Gingras & D. de Oliveira. 1993. Com-
plementary aspects of strawberry pollination by honey
and indigenous bees (Hymenoptera). Ecol. Behav.86:
29. Ollerton, J., S.D. Johnson & A.B. Hingston. 2006.
Geographical variation in diversity and specificity of
pollination systems. In Plant-Pollinator Interactions:
From Specialization to Generalization.N.M.Waser&J.
Ollerton, Eds.: 283–308. The University of Chicago
Press. Chicago.
30. Moldenke, A.R. 1979. Host-plant evolution and the di-
versity of bees in relation to the flora of North America.
Phytologia 43: 357–419.
31. Michener, C.D. 1979. Biogeography of the bees. Ann.
Missouri Botanical Garden 66: 278–347.
32. Freitas, B.M. et al. 2009. Diversity, threats and conser-
vation of native bees in the Neotropics. Apidologie 40:
33. Balmford, A. et al. 2003. Global variation in terres-
trial conservation costs, conservation benefits, and un-
met conservation needs. Proc. Natl. Acad. Sci. USA 100:
34. Malyshev, S.I. 1935. The nesting habits of solitary bees.
Eos 11: 201–309.
35. Wcislo, W.T. & J.H. Cane. 1996. Floral resource uti-
lization by solitary bees (Hymenoptera: Apoidea) and
exploitation of their stored foods by natural enemies.
Annu. Rev. Entomol.41: 257–286.
36. Heithaus, E.R. 1979. Community structure of neotrop-
ical flower visiting bees and wasps: diversity and phe-
nology. Ecology 60: 190–202.
37. Minckley, R. 2008. Faunal composition and species
richness differences of bees (Hymenoptera: Apiformes)
from two North American regions. Apidologie 39: 176–
38. Waser, N.M. et al. 1996. Generalization in pollination
systems, and why it matters. Ecology 77: 1043–1060.
39. Minckley, R.L., J.H. Cane & L. Kervin. 2000. Origins and
ecological consequences of pollen specialization among
desert bees. Proc. R. Soc. B-Biol. Sci.267: 265–271.
40. Danforth, B.N. 1999. Emergence dynamics and bet
hedging in a desert bee, Perdita portalis. Proc.R.Soc.
Lond. Ser. B-Biol. Sci.266: 1985–1994.
41. Cane, J.H. & V.J. Tepedino. 2001. Causes and ex-
tent of declines among native North American
invertebrate pollinators: detection, evidence, and
consequences. Conservation Ecol .5: [online] URL:
42. Allen-Wardell, G. et al. 1998. The potential conse-
quences of pollinator declines on the conservation of
biodiversity and stability of food crop yields. Conserva-
tion Biol.12: 8–17.
43. Kearns, C.A., D.W. Inouye & N.M. Waser.
1998. Endangered mutualisms: the conservation of
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 189
Wild bee conservation Winfree
plant-pollinator interactions. Annu. Rev. Ecol. Syst. 29:
44. Kremen, C. & T. Ricketts. 2000. Global perspectives on
pollination disruptions. Conservation Biol.14: 1226–
45. Ghazoul, J. 2005. Buzziness as usual? Questioning the
global pollination crisis. Tr en d s Ec ol . Ev ol .20: 367–373.
46. Ghazoul, J. 2005. Response to Steffan-Dewenter et al.:
Questioning the global pollination crisis. Trends E c o l .
Evol.20: 652–653.
47. Steffan-Dewenter, I., S. G. Potts & L. Packer. 2005. Polli-
nator diversity and crop pollination services are at risk.
Tre n ds E c ol . Ev o l .20: 651–652.
48. Westphal, C. et al. 2008. Measuring bee diversity in dif-
ferent European habitats and biogeographical regions.
Ecol. Monogr.78: 653–671.
49. Biesmeijer, J. C. et al. 2006. Parallel declines in polli-
nators and insect-pollinated plants in Britain and the
Netherlands. Science 313: 351–354.
50. Rasmont, P. et al. 2006. The survey of wild bees (Hy-
menoptera, Apoidea) in Belgium and France. In Status
of the World’s Pollinators: 1–18. Food & Agriculture Or-
ganization of the United Nations. Rome.
51. Mohra, C., M. Fellendorf & R.J. Paxton. 2004. The
population dynamics and genetics of solitary bees:
aEuropeancasestudy,Andrena vaga (Hymenoptera,
andrenidae). In Solitary Bees: Conservation, Rearing
and Management for Pollination. B.M. Freitas & J.O.P.
Pereira, Eds.: 85–95. University Dederal do Ceara.
Ceara, Brazil.
52. Patiny, S., P. Rasmont & D. Michez. 2009. A survey and
review of the status of wild bees in the West-Palaearctic
region. Apidologie 40: 313–331.
53. Fitzpatrick, U., T. E. Murray, A. Byrne, et al. 2006. Re-
gional red list of Irish bees.
54. Kleijn, D. et al. 2009. On the relationship between farm-
land biodiversity and land-use intensity in Europe. Proc.
R. Soc. B-Biol. Sci.276: 903–909.
55. Kleijn, D. et al. 2006. Mixed biodiversity benefits of
agri-environment schemes in five European countries.
Ecol. Lett. 9: 243–254.
56. Williams, P.H. & J.L. Osborne. 2009. Bumblebee vul-
nerability and conservation worldwide. Apidologie 40:
57. Goulson, D. 2003. Bumblebees: their Behavior and Ecol-
ogy. Oxford University Press. New York.
58. Goulson, D., G.C. Lye & B. Darvill. 2008. Decline and
conservation of bumble bees. Annu. Rev. Entomol. 53:
59. Kosior, A. et al. 2007. The decline of the bumble bees
and cuckoo bees (Hymenoptera: Apidae: Bombini) of
Western and Central Europe. Orzx 41: 79–88.
60. Carvell, C. et al. 2006. Declines in forage availability for
bumblebees at a national scale. Biol. Conservation 132:
61. Kleijn, D. & I. Raemakers. 2008. A retrospective analysis
of pollen host plant use by stable and declining bumble
bee species. Ecology 89: 1811–1823.
62. Williams, P. H. 2005. Does specialization explain rarity
and decline among British bumblebees? A response to
Goulson et al. Biol. Conservation 122: 33–43.
63. Goulson, D. et al. 2005. Causes of rarit y in bumblebees.
Biol. Conservation 122: 1–8.
64. Goulson, D. & B. Darvill. 2004. Niche overlap and
diet breadth in bumblebees; are rare species more spe-
cialized in their choice of flowers? Apidologie 35: 55–
65. Goulson, D. et al. 2006. Biotope associations and the
decline of bumblebees (Bombus spp.). J. Insect Conser-
vation 10: 95–103.
66. Fitzpatrick, U., T.E. Murray, R.J. Paxton, et al. 2007.
Rarity and decline in bumblebees – a test of causes and
correlates in the Irish fauna. Biol. Conservation 136:
67. Williams, P., S. Colla& Z. Xie. 2009. Bumblebee vulner-
ability: common correlates of winners and losers across
three continents. Conservation Biol .23: 931–940.
68. Colla, S.R. & L. Packer. 2008. Evidence for decline in
eastern North American bumblebees (Hymenoptera:
Apidae), with special focus on Bombus affinis Cresson.
Biodiversity Conservation 17: 1379–1391.
69. Grixti, J.C. et al. 2009. Decline of bumble bees (Bom-
bus) in the North American Midwest. Biol. Conserva-
tion 142: 75–84.
70. Evans, E. et al. 2008. Status review of three formerly
common species of bumble bee in the subgenus Bom-
bus. Xerces Society for Invertebrate Conservation. Port-
land. OR.
71. Colla, S.R. et al. 2006. Plight of the bumble bee:
pathogen spillover from commercial to wild popula-
tions. Biol. Conservation 129: 461–467.
72. Otterstatter, M.C. & J.D. Thomson. 2008. Does
pathogen spillover from comercially reared bumble
bees threaten wild pollinators? PLoS ONE 3: 1–9.
73. Winfree, R. et al. 2009. A meta-analysis of bees’ re-
sponses to anthropogenic disturbance. Ecology 90:
74. Magnacca, K.N. 2007. Conservation status of the en-
demic bees of Hawai’i, Hylaeus (Nesoprosopis) (Hy-
menoptera : Colletidae). Pac. Sci. 61: 173–190.
190 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
75. Wilcove, D. 1999. The Condor’s Shadow: The Loss and
Recovery of Wildlife in America. Freeman & Co. New
Yo r k , N Y .
76. Roubik, D.W. 2001. Ups and downs in pollinator pop-
ulations: when is there a decline? Conservation Ecol.5:
[online] URL:
77. Wilcove, D.S. et al. 1998. Quantifiying threats to imper-
iled species in the United States. Bioscience 48: 607–615.
78. Venter, O. et al. 2006. Threats to endangered species in
Canada. Bioscience 56: 903–910.
79. Sala, E.O. et al. 2000. Global biodiversity scenarios for
the year 2100. Science 287: 1770–1774.
80. Thomas, C.D. et al. 2004. Extinction risk from climate
change. Nature 427: 145–148.
81. Brown, M.J.F. & R.J. Paxton. 2009. The conservation of
bees: a global perspective. Apidologie 40: 410–416.
82. Steffan-Dewenter, I. et al. 2002. Scale-dep endent effects
of landscape context on three pollinator guilds. Ecology
83: 1421–1432.
83. Steffan-Dewenter, I., U. M¨
unzenberg & T. Tscharntke.
2001. Pollination, seed set and seed predation ona land-
scape scale. Proc. R. Soc. Lond. Ser. B.268: 1685–1690.
84. Williams, P.H. 1986. Environmental change and dis-
tributions of British bumble bees (Bombus Latr.). Bee
Wor l d 67: 50–61.
85. Winfree, R., T. Griswold & C. Kremen. 2007. Effect of
human disturbance on bee communities in a forested
ecosystem. Conservation Biol. 21: 213–223.
86. Williams, P.H. 1988. Habitat use by bumble bees (Bom-
bus spp.). Ecol. Entomol. 13: 223–237.
87. Klemm, M. 1996. Man-made bee habitats in the an-
thropogenous landscape of central Europe: substitutes
for threatened or destroyed riverine habitats? In The
Conservation of Bees. A. Matheson, S.L. Buchmann, C.
O’Toole, P. Westrich, I.H. Williams, Eds.: 17–34. Aca-
demic Press. London, UK.
88. Brosi, B.J., G.C. Daily & P.R. Ehrlich. 2007. Bee com-
munity shifts with landscape context in a tropical coun-
tryside. Ecol. Appl.17: 418–430.
89. Brosi, B.J. et al. 2008. The effects of forest fragmen-
tation on bee communities in tropical countryside.
J. Appl. Ecol.45: 773–783.
90. Tylianakis, J.M., A.-M. Klein & T. Tscharntke. 2005.
Spatiotemporal variation in the diversity of hy-
menoptera across a tropical habitat gradient. Ecology
86: 3296–3302.
91. Winfree, R. et al. 2008. Wild bee pollinators provide the
majority of crop visitation across land use gradients in
New Jersey and Pennsylvania.J. Appl. Ecol.45: 793–802.
92. Chapman, R.E., J. Wang & A.F.G. Bourke. 2003. Ge-
netic analysis of spatial foraging patterns and resource
sharing in bumble bee pollinators. Mol. Ecol.12: 2801–
93. McFrederick, Q.S. & G. LeBuhn. 2006. Are urban parks
refuges for bumble bees Bombus spp. (Hymenoptera :
Apidae)? Biol. Conservation 129: 372–382.
94. Cane, J.H. et al. 2006. Complex responses within a
desert bee guild (Hymenoptera: Apiformes) to urban
habitat fragmentation. Ecol. Appl. 632–644.
95. Polasky, S. et al. 2005. Conserving species in a work-
ing landscape: land use with biological and economic-
objectives. Ecol. Appl.15: 1387–1401.
96. Tscharntke, T. et al. 2002. Contribution of small habi-
tat fragments to conservation of insect communities
of grassland-cropland landscapes. Ecol. Appl. 12: 354–
97. Tscharntke, T. et al. 2002. Characteristics of insect pop-
ulations on habitat fragments: a mini review. Ecol. Res.
17: 229–239.
98. Matteson, K.C., J.S. Ascher & G.A. Langellotto. 2008.
Bee richness and abundance in New York City urban
gardens. Ann. Entomol. Soc. Am.101: 140–150.
99. Hegland, S.J. et al. 2009. How does climate warming
affect plant-pollinator interactions? Ecol. Lett.12: 184–
100. Wilson, R.J., Z.G. Davies & C.D. Thomas. 2007. Insects
and climate change: processes, patterns and implica-
tions for conservation. In Insect Conservation Biology
(Proceedings of the Royal Entomological Society’s 23rd
Symposium). A.J.A. Stewart, T.R. New & O.T. Lewis,
Eds.: CABI Publishing. Wallingford, UK.
101. Potts, S.G. 2009. Climate change impacts on pollina-
tors – risks in space and time. Conference presentation,
Entomological Society of America, 15 December 2009.
Indianapolis, IN.
102. Warren, M.S. et al. 2001. Rapid responses of British
butterflies to opposing forces of climate and habitat
change. Nature 414: 65–69.
103. Vitousek, al . 1996. Biological invasions as global
environmental change. Am. Sci .84: 468–478.
104. Bjer knes, A.L. etal . 2007. Do alien plant invasions really
affect pollination success in native plant species? Biol.
Conservation 138: 1–12.
105. Memmott, J. & N.M. Waser. 2002. Integration of alien
plants into a native flower-pollinator visitation web.
Proc. R. Soc. Lond. Ser. B Biol. Sci.269: 2395–2399.
106. Morales, C.L. & M.A. Aizen. 2006. Invasive mutualisms
and the structure of plant-pollinator interactions in the
temperate forests of north-west Patagonia, Argentina.
J. Ecol.94: 171–180.
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 191
Wild bee conservation Winfree
107. Bartomeus, I., M. Vil`
a & L. Santamar´
ıa. 2008. Con-
trasting effects of invasive plants in plant-pollinator
networks. Oecologia 155: 761–770.
108. Tepedino, V., B. Bradley & T. Griswold. 2008. Might
flowers of invasive plants increase native bee carrying
capacity? Intimations from Capitol Reef National Park,
Utah. Na t. Areas J.28: 44–50.
109. Goulson, D. 2003. Efffects of introduced bees on native
ecosystems. Annu. Re v. Ecol. Sy st. 34: 1–26.
110. Thomson, D.M. 2006. Detecting the effects of intro-
duced species: a case study of competition between Apis
and Bombus. Oikos 114: 407–418.
111. Thomson, D. 2004. Competitive interactions between
the invasive European honey bee and native bumble
bees. Ecology 85: 458–470.
112. Steffan-Dewenter, I. & T. Tscharntke. 2000. Resource
overlap and possible competition between honey bees
andwildbeesincentralEurope.Oecologia 122: 288–
113. Inoue, M.N., J. Yokoyama & I. Washitani. 2008. Dis-
placement of Japanese native bumblebees by the re-
cently introduced Bombus terrestris (L.) (Hymenoptera:
Apidae). J. Insect Conservation 12: 135–146.
114. Riedl, H. et al. 2006. How to reduce bee poisoning from
pesticides. Pac. Northwest Extension 591: 1–24.
115. Johansen, C.A. 1977. Pesticides and pollinators. Annu.
Rev. Entomol.22: 177–192.
116. Weinstock, G.M. et al. 2006. Insights into social insects
from the genome of the honeybee Apis mellifera.Nature
443: 931–949.
117. Thompson, H.M. & L.V. Hunt. 1999. Extrapolating
from honeybees to bumblebees in pesticide risk assess-
ment. Ecotoxicology 8: 147–166.
118. Johansen, C.A. et al. 1983. Pesticides and bees. Environ.
Entomol.12: 1513–1518.
119. Morandin, L.A. et al. 2005. Lethal and sub-lethal ef-
fectsofspinosadonbumblebees(Bombus impatiens
Cresson). Pest Manag. Sci.61: 619–626.
120. Alston, D.G. et al. 2007. Effects of the insecticide phos-
met on solitary bee foraging and nesting in orchards
of Capitol Reef National Park, Utah. Environ. Entomol.
36: 811–816.
121. Morandin, L.A. 2008. Genetically modified crops: ef-
fects on bees and pollination. In Bee Pollination in Agri-
cultural Ecosystems. R.R. James & T.L.Pitts-Singer, Eds.:
203–218. Oxford University Press. New York.
122. Frankham, R., J.D. Ballou & D.A. Briscoe. 2002. Intro-
duction to Conservation Genetics. Cambr idge University
Press. New York.
123. Zayed, A. 2004. Effective population size in Hy-
menoptera with complementary sex determination.
Heredity 93: 627–630.
124. Williams, N.M., R.L. Minckley & F.A. Silveira. 2001.
Variation in native bee faunas and its implications for
detecting community changes. Conservation Ecol.5:
[online] URL:
125. Hanski, I. 1990. Density dependence, regulation and
variaiblity in animal populations. Philos. Trans. R. Soc.
Lond.330: 141–150.
126. Kraus, F.B., S. Wolf & R.F.A. Moritz. 2009. Male flight
distance and population substructure in the bumblebee
Bombus terrestris.J. Anim. Ecol.78: 247–252.
127. Zayed, A. & L. Packer. 2001. High levels of diploid male
production in a primitively eusocial bee (Hymenoptera:
Halictidae). Heredity 87: 631–636.
128. Reed, D.H. & R. Frankham. 2003. Correlation between
fitness and genetic diversity. Conserv ation Biol. 17: 230–
129. Keller, L.F. & D.M. Waller. 2002. Inbreeding effects in
wild populations. Tre n ds E c ol . Ev o l .17: 230–241.
130. Henter, H.J. 2003. Inbreeding depression and hap-
lodiploidy: experimental measures in a parasitoid and
comparisons across diploid and haplodiploid insect
taxa. Evolution 57: 1793–1803.
131. Ellis, J.S. et al. 2006. Extremely low effective population
sizes, genetic structuring and reduced genetic diversity
in a threatened bumblebee species, Bombus sylvarum
(Hymenoptera: Apidae). Mol. Ecol.15: 4375–4386.
132. Zayed, A. & L. Packer. 2005. Complementary sex deter-
mination substantially increases extinction proneness
of haplodiploid populations. Proc. Natl. Acad. Sci. USA
102: 10742–10746.
133. Zayed, A., D.W. Roubik & L. Packer. 2004. Useof diploid
male frequency data as an indicator of pollinator de-
cline. Proc. R. Soc. Lond.271: S9–S12.
134. Paxton, R.J. et al. 2000. Microsatellite DNA analysis
reveals low diploid male production in a communal
bee with inbreeding. Biol. J. Linnean Soc.69: 483–
135. Zayed, A. et al. 2005. Increased genetic differentiation
in a specialist versus a generalist bee: implications for
conservation. Conservation Genet.6: 1017–1026.
136. Packer, L.Z., A. Grixti, et al. 2005. Conservation genetics
of potentially endangered mutualisms: reduced levels
of genetic variation in specialist versus generalist bees.
Conservation Biol. 19: 195–202.
137. Zayed, A. & L. Packer. 2007. The population genet-
ics of a solitary oligolectic sweat bee, Lasioglossum
(Sphecodogastra) oenotherae (Hymenoptera: Halicti-
dae). Heredity 99: 397–405.
192 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
138. Darvill, B. et al. 2006. Population structure and in-
breeding in a rare and declining bumblebee, Bombus
muscorum (Hymenoptera: Apidae). Mol. Ecol .15:601–
139. Chapman, R.E. & A.F.G. Bourke. 2001. The influence of
sociality on the conservation biology of social insects.
Ecol. Lett. 4: 650–662.
140. Herrmann, F. et al. 2007. Genetic diversity and mass
resources promote colony size and forager densities of
a social bee (Bombus pascuorum) in agricultural land-
scapes. Mol. Ecol. 16: 1167–1178.
141. Williams, N.M., Crone, E. E., Roulston, T.H., et al.Inre-
view. Ecological and life history traits predict bee species
responses to environmental disturbances. Biol. Conser-
vation (in press).
142. Memmott, J., N.M. Waser& M.V. Price. 2004. Tolerance
of pollination networks to species extinctions. Proc. R.
Soc. Lond. B, Biol. Sci. 271: 2605–2611.
143. Bond, W.J. 1994. Do mutualisms matter? Assessing the
impact of pollinator and disperser disruption on plant
extinction. Philos. Trans. R. Soc. Lond. B.344: 83–90.
144. Bond, W.J. 1995. Assessing the risk of plant extinction
due to pollinator and disperser failure. In Extinction
Rates. J.H. Lawton & R.M. May, Eds.: 131–146. Oxford
University Press. Oxford, UK.
145. Aguilar, R. et al. 2006. Plant reproductive susceptibility
to habitat fragmentation: review and synthesis through
a meta-analysis. Ecol. Lett. 9: 968–980.
146. Koh, L.P. et al. 2004. Species coextinctions and the bio-
diversity crisis. Science 305: 1632–1634.
147. Dunn, R.R. et al. 2009. The sixth mass coextinction:
are most endangered species parasites and mutualists?
Proc. Royal Soc. B-Biol. Sci.276: 3037–3045.
148. Minckley, R.L. & T.H. Roulston. 2006. Incidental mutu-
alisms and pollen specialization among bees. In Plant-
Pollinator Interactions: From Specialization to General-
ization. N.M. Waser & J. Ollerton, Eds.: 69–98. The
University of Chicago Press. Chicago.
149. Bascompte, J. & P. Jordano. 2007. Plant-animal mutu-
alist networks: the architecture of biodiversity. Annu.
Rev. Ecol., Evol. Syst.38: 567–593.
150. Vazquez, D.P. et al. 2009. Uniting pattern and process
in plant–animal mutualistic networks: a review. Ann.
Bot.103: 1445–1457.
151. V´
azquez, D.P. et al. 2007. Species abundance and asym-
metric interaction strength in ecological networks.
Oikos 116: 1120–1127.
152. Bascompte, J., Pedro Jordano & Jens M. Olesen. 2006.
Asymmetric coevolutionary networks facilitate biodi-
versity maintenance. Science 312: 431–433.
153. V´
azquez, D.P. & M.A. Aizen. 2004. Asymmetric special-
ization: a pervasive feature of plant-pollinator interac-
tions. Ecology 85: 1251–1257.
154. Petanidou, T. et al. 2008. Long-term observation of a
pollination network: fluctuation in species and inter-
actions, relative invariance of network structure and
implications for estimates of specialization. Ecol. Lett.
11: 564–575.
155. Muller, A. et al. 2006. Quantitative pollen requirements
of solitary bees: implications for bee conservation and
the evolution of bee-flower relationships. Biol. Conser-
vation 130: 604–615.
156. Potts, S.G., Betsy Vulliamy, Stuart Roberts, et al. 2005.
Role of nesting resources in organising diversebee com-
munities in a Mediterranean landscape. Ecol. Entomol.
157. Westrich, P. 1996. Habitat requirements of central Eu-
ropean bees and the problems of partial habitats. In The
Conservation of Bees. A. Matheson, S.L. Buchmann, C.
O’Toole, P. Westrich, I.H. Williams, Eds.: 1–16. Aca-
demic Press for the Linnean Society of London and
IBRA. London, UK.
158. Williams, N.M. & C. Kremen. 2007. Resource distribu-
tions among habitats determine solitary bee offspring
production in a mosaic landscape. Ecol. Appl. 17: 910–
159. Dailey, T.B. & P.E. Scott. 2006. Spring nectar sources
for solitary bees and flies in a landscape of deciduous
forest and agricultural fields: production, variability,
and consumption. J. Torrey Botanical Soc.133: 535–
160. Lonsdorf, E. et al. 2009. Modeling pollination services
across agricultural landscapes. Ann. Bot.103: 1589–
161. Koh, L.P., N.S. Sodhi & B.W. Brook. 2004. Ecological
correlates of extinction proneness in tropical butterflies.
Conservation Biol. 18: 1572–1578.
162. Steffan-Dewenter, I. & T. Tscharntke. 2000. Butterfly
community structure in fragmented habitats. Ecol. Lett.
3: 449–456.
163. McKinney, M.L. 1997. Extinction vulnerability and
selectivity: combining ecological and paleonto-
logical views. Annu. Rev. Ecol. Syst. 28: 495–
164. De victor, V., R. Julliard & F. Jiguet. 2008. Distribution of
specialist and generalist species along spatial gradients
of habitat disturbance and fragmentation. Oikos 117:
165. Hamback, P.A. et al. 2007. Habitat specialization,
body size, and family identity explain lepidopteran
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 193
Wild bee conservation Winfree
density-area relationships in a cross-continental com-
parison. Proc. Natl. Acad. Sci. USA 104: 8368–8373.
166. Kotiaho, J.S. et al. 2005. Predicting the risk of extinction
from shared ecological characteristics. Proc. Natl. Acad.
Sci. USA 102: 1963–1967.
167. Cane, J.H. & S. Sipes. 2006. Characterizing floral spe-
cialization by bees: analytical methods and a revised
lexicon for oligolecty. In Plant-Pollinator Interactions:
From Specialization to Generalization.N.M.Waser&J.
Ollerton, Eds.: 99–122. The University of ChicagoPress.
168. Zavaleta, E. et al. 2009. Ecosystem responses to com-
munity disassembly. Year Ecol. Conservation Biol., Ann.
NY Acad. Sci. 1162: 311–333.
169. Cardillo, M. et al. 2005. Multiple causes of high extinc-
tion risk in large mammal species. Science 309: 1239–
170. Thomas, C.D. 2000. Dispersal and extinction in frag-
mented landscapes. Proc. R. Soc. Lond. B.267: 139–145.
171. Greenleaf, S.S. et al . 2007. Bee foraging ranges and their
relationship to body size. Oecologia 153: 589–596.
172. Wilson, E.O. 1999. The Diversity of Life.Norton.New
Yo r k , N Y .
173. Stork, N.E. 2007. World of insects. Nature 448: 657–658.
174. Dunn, R.R. 2005. Modern insect extinctions, the
neglected majority. Conservation Biol. 19: 1030–
175. McKinney, M.L. 1999. High rates of extinction and
threat in poorly studies taxa. Conservation Biol. 13:
176. Clark, J.A. & R.M. May. 2002. Taxonomic bias in con-
servation research. Science 297: 191–192.
177. Hoffman Black, S., M. Shepard & M. Mackey Allen.
2001. Endangered Invertebrates: the case for greater
attention to invertebrate conservation. Endangered
Species Update 18: 42–50.
178. Hawksworth, D.L. & M.T. Kalin-Arroyo. 1995. Magni-
tude and distribution of biodiversity. In United Nations
Environment Program: Global Biodiversity Assessment.
V.H. Heywood, Ed.: 105–191. Cambridge University
Press. Cambridge, UK.
179. Warren, M.S. et al. 2007. What have red lists donefor us?
The values and limitations of protected species listing
for invertebrates. In Insect Conservation Biology (Pro-
ceedings of the Royal Entomological Society’s 23rd Sym-
posium). A.J.A. Stewart, T.R. New & O.T. Lewis, Eds.:
76–91. CABI Publishing. Wallingford, UK.
180. Bean, J.M. 1993. Invertebrates and the Endangered
Species Act. Wing s 17: 12–15.
181. Thomas, J.A. et al. 2004. Comparative losses of British
butterflies, birds, and plants and the global extinction
crisis. Science 303: 1879–1881.
182. Conrad, K.F. et al. 2006. Rapid declines of common,
widespread British moths provide evidence of an insect
biodiversity crisis. Biol. Conservation 132: 279–291.
183. IUCN. 2009. IUCN Red List of threatened species, Vol.
184. McCauley, D.J. 2006. Selling out on nature. Nature 443:
185. Turner, K. et al. 2003. Valuing nature: lessons learned
and future research directions. Ecol. Econ. 46: 493–510.
186. Costanza, R. et al . 1997. The value of the world’s ecosys-
tem services and natural capital. Nature 387: 253–260.
187. Ghazoul, J. 2007. Challenges to the uptake of the ecosys-
tem service rationale for conservation. Conservation
Biol. 21: 1651–1652.
188. Winfree, R. & B. Gross. In revision. A new method for
valuing crop pollination. Ecol. Econ.
189. Bos, M.M. et al. 2007. Caveats to quantifying ecosys-
tem services: fruit abortion blurs benefits from crop
pollination. Ecol. Appl. 17: 1841–1849.
190. Ashman, T. et al. 2004. Pollen limitation of plant repro-
duction: ecological and evolutionary causes and conse-
quences. Ecology 85: 2408–2421.
191. Knight, T.M., J.A. Steets & T.L. Ashman. 2006. A quanti-
tative synthesis of pollen supplementation experiments
highlights the contribution of resource reallocation to
estimates of pollen limitation. Am. J. Bot .93: 271–277.
192. Ghazoul, J. 2007. Recognizing the complexities of
ecosystem management and the ecosystem service con-
cept. Gaia 16: 215–221.
193. Klein, A.-M., R. Olschewski & C. Kremen. 2008. The
ecosystem service controversy: is there sufficient evi-
dence for a “pollination paradox”? Gaia 17: 12–16.
194. Ricketts, T.H. et al. 2004. Economic value of tropical
forest to coffee production. Proc. Natl. Acad. Sci.101:
195. Priess, J.A. et al. 2007. Linking deforestation scenarios
to pollination services and economic returns in coffee
agrofoestry systems. Ecol. Appl. 17: 407–417.
196. Morandin, L.A. & M.L. Winston. 2006. Pollinators
provide economic incentive to preserve natural land
in agroecosystems. Agric. Ecosyst. Environ.116: 289–
197. Olschewski, R. et al. 2006. Economic evaluation of
pollination services comparing coffee landscapes in
Ecuador and Indonesia. Ecol. Soc.11: 7[online]URL:
198. Chan, K.M.A. et al. 2006. Conservation planning for
ecosystem services. PLoS Biol. 4: 2138–2152.
194 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
199. Defries, R., J.A. Foley & G.P. Asner. 2004. Land-use
choices: balancing human needs and ecosystem func-
tion. Front. Ecol. Environ.2: 249–257.
200. Tilman, D. et al. 2001. Forecasting agriculturally driven
global environmental change. Science 292: 281–284.
201. Larsson, M. & M. Franzen. 2007. Critical resource lev-
els of pollen for the declining bee Andrena hattorfiana
(Hymenoptera, Andrenidae). Biol. Conservation 134:
202. Minckley, R.L. et al . 1994. Behavior and phenology of a
specialist bee (Dieunomia) and sunflower (Helianthus)
pollen availability. Ecology 75: 1406–1419.
203. Bowers, M. 1986. Resources availability and timing of
reproduction in bumble bee colonies (Hymenoptera:
Apidae). Environ. Entomol .15: 750–755.
204. Gathmann, A., H.J. Greiler & T. Tscharntke. 1994. Trap-
nesting bees and wasps colonizing set-aside fields: suc-
cession and body size, management by cutting and sow-
ing. Oecologia 98: 8–14.
205. Frohlich, D.R. & V.J. Tepedino. 1986. Sex ratio, parental
investment, and interparent variability in nesting suc-
cess in a solitary bee. Evolution 40: 142–151.
206. Steffan-Dewenter, I. & S. Schiele. 2008. Do resources
or natural enemies drive bee population dynamics in
fragmented habitats? Ecology 89: 1375–1387.
207. Tscharntke, T., A. Gathmann & I. Steffan-Dewenter.
1998. Bioindication using trap-nesting bees and wasps
and their natural enemies: community structure and
interactions. J. Appl. Ecol.35: 708–719.
208. Dukas, R. 2001. Effects of predation risk on pollinators
and plants. In Cognitive Ecology of Pollination: Animal
Behaviour and Floral Evolution. L. Chittka & J. Thom-
son, Eds.: 214–236. Cambridge University Press. Cam-
bridge, UK.
209. Carvell, C. et al. 2004. The response of foraging bum-
blebees to successional change in newly created arable
field margins. Biol. Conservation 118: 327–339.
210. Pywell, R.F., E.A. Warman, L. Hulmes, et al. 2006. Ef-
fectiveness of new agri-environment schemes in pro-
viding foraging resources for bumblebees in intensively
farmed landscapes. Biol. Conservation 129: 192–206.
211. Potts, S.G. et al. 2009. Enhancing pollinator biodiversity
in intensive grasslands. J. Appl. Ecol.46: 369–379.
212. Carvell, C. et al. 2007. Comparing the efficacy of agri-
environment schemes to enhance bumble bee abun-
dance and diversity on arable field margins. J. Appl.
Ecol.44: 29–40.
213. Pywell, R.F. et al. 2005. Providing foraging resources
for bumblebees in intensively farmed landscapes. Biol.
Conservation 121: 479–494.
214. Vaughan, M. et al. 2004. Farming for bees: guidelines
for providing native bee habitat on farms. The Xerces
Society. Portland, OR.
215. Tuell, J. K. et al. 2008. Visitation by wild and managed
bees (Hymenoptera: Apoidea) to eastern U.S. native
plants for use in conservation programs. Environ. En-
tomol.37: 707–718.
216. Williams, N.M., R. Winfree & E. McGlynn. 2009. Native
217. Johnson, D.H. 1980. The comparison of usage and
availability measurements for evaluating resource pref-
erence. Ecology 61: 65–71.
218. Kells, A.R., J.M. Holland & D. Goulson. 2001. The value
of uncropped field margins for foraging bumblebees.
J. Insect Conservation 5: 283–291.
219. Kells, A.R. & D. Goulson. 2003. Preferred nesting sites
of bumblebee queens (Hymenoptera : Apidae) in agroe-
cosystems in the UK. Biol. Conservation 109: 165–174.
220. Svensson, B., J. Lagerlof & B.G. Svensson. 2000. Habi-
tat preferences of nest-seeking bumble bees (Hy-
menoptera: Apidae) in an agricultural landscape. Agric.
Ecosystems Environ. 77: 247–255.
221. OECD. 2003. Agri-environmental policies in OECD
countries. In Agricultural Policies in OECD Countries:
Monitoring and Evaluation 2003: 67–80. OECD Publi-
cations. Paris, France.
222. OECD. 2003. Analysis of the 2002 Farm Act in the
United States. In Agricultural Policies in OECD Coun-
tries: Monitoring and Evaluation 2003: 45–66.OECD
Publications. Paris, France.
223. Vaughan, M. & M. Skinner. 2008. Using Farm Bill pro-
grams for pollinator conservation. NRCS/The Xerces
Society/San Francisco State University.
224. Gray, R.L. & B.M. Teels. 2006. Wildlife and fish con-
servation through the Farm Bill. Wildl. Soc. Bull.34:
225. Burger, L.W. 2006. Creating wildlife habitat through
federal farm programs: an objective-driven approach.
Wildl. Soc. Bull. 34: 994–999.
226. Davros, N.M. et al. 2006. Butterflies and continuous
conservation reserve program filter strips: landscape
considerations. Wildl. Soc. Bull.34: 936–943.
227. Whitfield, J. 2006. How green was my subsidy? Nature
439: 908–909.
228. Kleijn, D. et al. 2001. Agri-environment schemes do
not effectively protect biodiversityin D utchagr icultural
landscapes. Nature 413: 723–725.
229. Kleijn, D. & W.J. Sutherland. 2003. How effec-
tive are European agri-environment schemes in
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 195
Wild bee conservation Winfree
conserving and promoting biodiversity? J. Appl. Ecol.
40: 947–969.
230. Potts, S.G. et al. 2006. Commentary on: mixed bio-
diversity benefits of agri-environment schemes in five
European countries. Ecol. Lett.9: 254–256.
231. Goulson, D. et al. 2002. Colony growth of the bumble-
bee, Bombus terrestris, in improved and conventional
agricultural and suburban habitats. Oecologia 130: 267–
232. Knop, E. et al. 2006. Effectiveness of the Swiss agri-
environment scheme in promoting biodiversity. J. Appl.
Ecol.43: 120–127.
233. Albrecht, M. et al. 2007. The Swiss agri-environment
scheme enhances pollinator diversity and plant repro-
ductive success in nearby intensively managed farm-
land. J. Appl. Ecol.44: 813–822.
234. Marshall, E.J.P., T.M. West & D. Kleijn. 2006. Impacts of
an agri-environment field margin prescription on the
flora and fauna of arable farmland in different land-
scapes. Agric. Ecosyst. Env iron.113: 36–44.
235. Clough, Y. et al. 2007. Alpha and beta diversity of
arthropods and plants in organically and convention-
ally managed wheat fields. J. Appl. Ecol.44: 804–812.
236. Bengtsson, J., J. Ahnstrom & A-C Weibull. 2005. The
effects of organic agriculture on biodiversity and abun-
dance: a meta-analysis. J. Appl. Ecol.42: 261–269.
237. Fuller, R.J. et al. 2005. Benefits of organic farming to
biodiversity vary among taxa. Biol. Lett. 1: 431–434.
238. Gibson, R.H. et al. 2007. Plant diversity and land use
under organic and conventional agriculture: a whole-
farm approach. J. Appl. Ecol.44: 792–803.
239. Hole, D.G. et al. 2005. Does organic farming benefit
biodiversity? Biol. Conservation 122: 113–130.
240. Holzschuh, A. et al. 2007. Diversity of flower-visiting
bees in cereal fields: effects of farming system, landscape
composition and regional context. J. Appl. Ecol.44: 41–
241. Holzschuh, A., I. Steffan-Dewenter & T. Tscharntke.
2008. Agricultural landscapes with organic crops sup-
port higher pollinator diversity. Oikos 117: 354–361.
242. Morandin, L.A. & M.L. Winston. 2005. Wild bee abun-
dance and seed production in conventional, organic,
and genetically modified canola. Ecol. Appl. 15: 871–
243. Kremen, C. et al. 2004. The area requirements of
an ecosystem service: crop pollination by native
bee communities in California. Ecol. Lett. 7: 1109–
244. Weibull, A.C., O. Ostman & A. Granqvist. 2003. Species
richness in agroecosystems: the effect of landscape,
habitat and farm management. Biodiversity Conserva-
tion 12: 1335–1355.
245. Weibull, A.-C., J. Bengtsson & E. Nohlgren. 2000. Di-
versity of butterflies in the agricultural landscape: the
role of farming system and landscape heterogeneity.
Ecography 23: 743–750.
246. Benton, T.G., J.A. Vickery & J.D. Wilson. 2003. Farm-
land biodiversity: is habitat heterogeneity the key?
Tre n ds E c ol . Ev o l .18: 182–188.
247. Tscharntke, T. et al. 2005. Landscape perspectives on
agricultural intensification and biodiversity – ecosys-
tem service management. Ecol. Lett.8: 857–874.
248. Heard, M.S. et al. 2007. Landscape context not patch
size determines bumble-bee density on flower mixtures
sown for agri-environment schemes. Biol. Lett. 3: 638–
249. Rundl¨
of, M. & H.G. Smith. 2006. The effect of organic
farming on butterfly diversity depends on landscape
context. J. Appl. Ecol. 43: 1121–1127.
250. Roschewitz, I. et al . 2005. The effects of landscape com-
plexity on arable weed species diversity in organic and
conventional farming. J. Appl. Ecol.42: 873–882.
251. Green, R.E. et al. 2005. Farming and the fate of wild
nature. Science 307: 550–555.
252. Vandermeer, J. & I. Perfecto. 2007. The agricultural
matrix and a future paradigm for conservation. Con-
servation Biol.21: 274–277.
253. Matson, P.A. et al. 1997. Agricultural intensification
and ecosystem properties. Science 277: 504–509.
254. Badgley, C. et al. 2007. Organic agriculture and the
global food supply. Renew. Agric. Food Syst .22: 86–108.
255. Balmford, A., R.E. Green & J.P.W. Scharlemann. 2005.
Sparing land for nature: exploring the potential impact
of changes in agricultural yield on the area needed for
crop production. Global Change Biol. 11:1594–1605.
256. Donald, P.F., R.E. Green & M.F. Heath. 2001. Agricul-
tural intensification and the collapse of Europe’s farm-
land bird populations. Proc. R. Soc. Lond.268: 25–29.
257. Perfecto, I. et al. 2005. Biodiversity, yield, and shade
coffee certification. Ecol. Econ. 54: 435–446.
258. Rudel, T.K. et al. 2009. Agricultural intensification and
changes in cultivated areas, 1970–2005. Proc.Natl. Acad.
Sci.106: 20675–20680.
259. Perfecto, I. & J. Vandermeer. 2008. Biodiversity conser-
vation in tropical agroecosystems: a new conservation
paradigm. Ann. NY Acad. Sci.1134: 173–200.
260. Balvanera, P., C. Kremen & M. Martinez-Ramos. 2005.
Applying community structure analysis to ecosystem
function: examples from pollination and carbon stor-
age. Ecol. Appl.15: 360–375.
196 Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences.
Winfree Wild bee conservation
261. Kohler, F. et al. 2008. At what spatial scale do high-
quality habitats enhance the diversity of forbs and
pollinators in intensively farmed landscapes?. J. Appl.
Ecol.45: 753–762.
262. Forup, M.L. et al. 2008. The restoration of ecological
interactions: plant-pollinator networks on ancient and
restored heathlands. J. Appl. Ecol. 45: 742–752.
263. Williams, N.M. In review. Restoration of non-target
species: Pollinators and pollination function in riparian
forests. Restor. Ecol.
264. Michener, C.D. & C.W. Rettenmeyer. 1956. The ethol-
ogy of Andrena e ry thronii with comparative data on
other species (Hymenoptera, andrenidae). Univ. Kans.
Sci. Bull.37: 645–684.
265. Torchio, P.F. 1990. Osmia ribifloris,anativebeespecies
developed as a commercially managed pollinator
of highbush blueberry (Hymenoptera: Megachilidae).
J. Kans. Entomol. Soc.63: 427–436.
266. Westphal, C., I. Steffan-Dewenter & T. Tscharntke.
2009. Mass flowering oilseed rape improvesearly colony
growth but not sexual reproduction of bumblebees.
J. Appl. Ecol.46: 187–193.
Ann. N.Y. Acad. Sci. 1195 (2010) 169–197 c
2010 New York Academy of Sciences. 197
... If not all of these resources are available in a single habitat, wild bees may complementarily use both habitats. For example, wild bees use food resources in open landscapes while using nesting resources in forests (Mandelik et al. 2012;Winfree 2010). Therefore, forest road verges may offer critical food resources not only to forest-associated species but also to ubiquitous species and species associated with open landscapes. ...
... Gathering information about the specific use of flowering resources within habitats is essential for conservation efforts for wild bees (Winfree 2010;Jha, Stefanovich, and Kremen 2013). In this study, we took a closer look at the pollen use of the common carder bee to illustrate its use of forest road verge resources and to clarify which resources are especially important for them. ...
Forests in Germany are occupied with roads, paths, and trails with a density of 5.03 km/km². Their construction and maintenance create a network of verges promoting flowering plants. Whether these verges are visited by bees, which factors are determining their abundance, diversity, and composition, and which flowering resources are used is unknown. We selected 13 verges in the Black Forest (Germany), sweep-netted wild bees along transects, calculated the flowering area of all herbs, and measured the area (hectares) of grassland within 1 km around the transects. To evaluate the resource use of a common bumblebee species, we analyzed the pollen load of common carder bees (Bombus pascuorum) using microscopes. The abundance and diversity of wild bees was positively related to flowering area. With an increasing area of grassland, the abundance of ubiquitous species increased. Wild bee community composition was driven by flowering area. Common carder bees collected pollen from several flower resources but mainly used few species, such as the common hemp nettle (Galeopsis tetrahit L.). As the flowering area influenced wild bee abundance, diversity, and composition, we suggest creating road verges that favor the occurrence of native flowering plants to support wild bees in forest ecosystems. Study Implications: Forest road verges generally have higher light availability than the forest interior and therefore have higher availability of flowering plants. Although the importance of verges for wild bee conservation in agricultural landscapes is known, forest road verges are understudied. Our study demonstrates that forest road verges are important habitats for many ubiquitous bees and that the flowering area on these verges is the key determinant for the abundance and diversity of wild bees. Therefore, creating road verges that favor the occurrence of native flowering plants is key to support bees on these verges.
... Wild bees (Hymenoptera: Anthophila) are a diverse group of pollinators that provide essential pollination services vital to the stability of global plant-pollinator communities (Potts et al 2010). Many species of bees are presumed to be in decline, likely as a result of synergistic properties from multiple environmental pressures experienced by a species (Potts et al. 2010, Winfree et al. 2010, Koh et al. 2016. In response, state and federal government agencies have taken initiative to mitigate species loss through various governmental policies, increased funding to promote healthy wild bee populations, and in some cases, providing legal protection for species of greatest conservation concern (I.e., listing under the Endangered Species Act) (United States 1983, Taylor et al. 2005. ...
... These decisions, in part, are informed by contemporary knowledge regarding species long-term population viability and geographic occupancy trends, specific habitat requirements, and life-history traits, which are supported by robust collection records and data sets for species of interest. However, in many parts of the United States, the necessary information to make inferences on a species' conservation status are lacking (Winfree 2010, Woodard et al. 2020, Fischer et al. 2021. This is particularly evident for rare species, primarily due to inadequate population level data to inform conservation decision making practices. ...
Technical Report
Full-text available
Michigan’s Lakeplain Prairie and Prairie Fen natural communities contain refugia for many at-risk species of insects, including species that are federally and state listed. Wild bees are a group of insects that are crucial for maintaining robust plant-pollinator communities. Numerous species of bees are presumed to be in decline. However, baseline community surveys are lacking but are needed to document the status of species occupying these natural communities. In 2021, Michigan Natural Features Inventory completed baseline wild bee surveys in Lakeplain Prairie (Lakeplain Wet-mesic Prairie and Lakeplain Wet Prairie) and Prairie Fens in Michigan using a combination of aerial netting and bowl trapping. A total of 1118 wild bees, representing 104 unique species, were collected during 2021 surveys, including new state records for Dufourea marginata and Sphecodes nigricorpus. The baseline inventory of wild bees in 2021 provide valuable information on species presence and floral resources used by the wild bee communities at these sites. Continued management of Lakeplain Prairie and Prairie Fen natural communities should take into consideration wild bee communities and strive to ensure populations maintain stable numbers. Additional baseline surveys are needed to document the wild bee communities of other natural communities in Michigan and would benefit the long-term conservation of species at risk of population decline.
... Over the last two decades, there has been substantial interest in the status of pollinators (Allen-Wardell et al. 1998;Kevan and Phillips 2001;Marlin and LaBerge 2001;Biesmeijer et al. 2006;Berenbaum et al. 2007;Potts et al. 2010;Winfree 2010;Colla et al. 2012;Bartomeus et al. 2013;Lebuhn et al. 2013;Senapathi et al. 2015). With the possible exception of bumble bees (Cameron et al. 2011;Kerr et al. 2015), few wild bee taxa have been sufficiently well documented in North America to provide effective baseline data to reliably measure conservation status. ...
Full-text available
We record 392 species or morphospecies of bees (Hymenoptera: Apoidea) for Manitoba, Canada, which is 154 more species than reported in 2015 and includes five new generic records since 2015 ( Ashmeadiella , Brachymelecta , Eucera, Neolarra, and Triepeolus ). Thirteen new records reported here are new for Canada: Calliopsis ( Nomadopsis ) australior Cockerell, Perdita ( Perdita ) tridentata Stevens, Brachymelecta interrupta (Cresson), Diadasia ( Dasiapis ) ochracea (Cockerell), Melissodes bidentis Cockerell, Nomada crawfordi crawfordi Cockerell, Nomada fuscicincta Swenk, Nomada sphaerogaster Cockerell, Nomada xantholepis Cockerell, Triepeolus cf. grindeliae Cockerell, Dianthidium ( Dianthidium ) parvum (Cresson), Coelioxys ( Xerocoelioxys ) nodis Baker, and Megachile ( Megachiloides ) dakotensis Mitchell. We remove the following species from the list of Manitoba bees based on re-examination of voucher material: Andrena ( Ptilandrena ) geranii Robertson, Andrena ( Rhacandrena ) robertsonii Dalla Torre, Andrena ( Simandrena ) nasonii Robertson, Andrena ( Trachandrena ) ceanothi Viereck, Andrena ( Trachandrena ) quintilis Robertson, Lasioglossum ( Hemihalictus ) pectoraloides (Cockerell), Lasioglossum ( Lasioglossum ) forbesii (Robertson), and Dianthidium ( Dianthidium ) concinnum (Cresson). We propose that Nomada alpha paralpha Cockerell, 1921 and N. alpha dialpha Cockerell, 1921 are junior synonyms of N. alpha Cockerell, 1905. Nomada arenicola Swenk, 1912 is considered a junior synonym of N. fervida Smith, 1854. Protandrena albertensis (Cockerell) and Neolarra mallochi Michener are recognised as valid species. We provide additional notes on taxonomy, nomenclature, and behaviour for select species in the list.
... Wild (+): living in a state of nature and not tame or domesticated (Merriam-Webster 2022); uncontrolled, violent, or extreme (Cambridge, 2022); [bees] not managed by humans (Mallinger et al., 2017;Winfree, 2010). ...
Full-text available
Effectively promoting the stability and quality of ecosystem services involves the successful management of domesticated species and the control of introduced species. In the pollinator literature, interest and concern regarding pollinator species and pollinator health dramatically increased in recent years. Concurrently, the use of loaded terms when discussing domesticated and non-native species may have increased. As a result, pollinator ecology has inherited both the confusion associated with invasion biology’s lack of a standardized terminology to describe native, managed, or introduced species as well as loaded terms with very strong positive or negative connotations. The recent explosion of research on native bees and alternative pollinators, coupled with the use of loaded language, has led to a perceived divide between native bee and managed bee researchers. In comparison, the bird literature discusses the study of managed (poultry) and non-managed (all other birds) species without an apparent conflict with regard to the use of terms with strong connotations or sentiment. Here, we analyze word usage when discussing non-managed and managed bee and bird species in 3614 ecological and evolutionary biology papers published between 1990 and 2019. Using time series analyses, we demonstrate how the use of specific descriptor terms (such as wild, introduced, and exotic) changed over time. We then conducted co-citation network analyses to determine whether papers that share references have similar terminology and sentiment. We predicted a negative language bias towards introduced species and positive language bias towards native species. We found an association between the term invasive and bumble bees and we observed significant increases in the usage of more ambiguous terms to describe non-managed species, such as wild . We detected a negative sentiment associated with the research area of pathogen spillover in bumble bees, which corroborates the subjectivity that language carries. We recommend using terms that acknowledge the role of human activities on pathogen spillover and biological invasions. Avoiding the usage of loaded terms when discussing managed and non-managed species will advance our understanding and promote effective and productive communication across scientists, general public, policy makers and other stake holders in our society.
... These ecosystems have higher levels of bee abundance when compared to ecosystems dominated by anthropogenic activities and contain crucial habitats for bee communities (Figure 1) (Koh et al. 2016;Carril et al. 2018). Yet restoration strategies for pollinators have primarily been developed within agroecosystems as opposed to seminatural ecosystems (Winfree 2010 ...
Individual plant species play valuable roles in meeting restoration goals for pollinators. However, the selection of plant species for pollinator restoration is rarely informed using empirical evidence and is usually developed in agroecosystems, which experience frequent human interventions to ensure plant success as compared to seminatural ecosystems. We highlight concepts and future research needs to design planting mixes that fulfill the ecological requirements of pollinators in seminatural ecosystems. Native plants that are attractive to pollinators, increase the stability of pollination services, and provide consistent floral resources across the landscape and growing season should be prioritized in pollinator restoration projects in seminatural ecosystems. Furthermore, condensing criteria of desirable plant traits into a composite score can aid managers in selecting plant species that meet restoration goals. Developing restoration strategies for pollinators on seminatural lands is important for preserving organisms essential for biodiversity maintenance and ecosystem function.
... However, in our opinion, this obviously rough approach is still effective in outlining a real difference in diversity between eusocial and solitary species. One of the underlying issues is that many ecological and behavioural data are still lacking for most European wild bee species (Antoine & Forrest, 2021;Winfree, 2010) and very few studies analysed possible geographical clines in such traits across populations within species (Knerer, 1992;Lawson et al., 2018). Even parasitism rate, which affects productivity, is certainly variable across species (Wcislo, 1996), years (Wcislo et al., 1994) and also across populations/locations (Knerer, 1973); so we are aware that our assumption that such variation is similar across species represents a limitation. ...
1. Bees provide important ecosystem services and are subjects of extensive studies on their α-diversity, which is generally calculated with indices that integrate the number of species with their abundances. Variation in social behaviour, though expected to impact genetic diversity, is still largely neglected in such studies. 2. We propose a simple method to show how sociality may affect diversity indices, when a surrogate of genetic diversity is taken into account. This method weighs the number of sampled females (N) to obtain new abundance values (N W), by taking into account relevant biological traits affecting genetic structure of populations, that is, the number of (natal) nests from which the sampled females originated (which depends on brood productivity and sex ratio) and the genetic relatedness among such females. 3. Solitary species tend to have greater N W than eusocial ones especially at larger sample sizes. Across studies on 121 bee communities, we found that Taxonomic dis-tinctness, Shannon-Wiener diversity and Gini-Simpson dominance tended to be greater when based on N W rather than on N. Such differences increased in communities with decreasing number of eusocial species and with increasing proportion of individuals from eusocial species. 4. The results suggest that taking into account the social organisation of wild bees may have important consequences in estimating α-diversity, thus claiming for future efforts in collecting biological data on as many wild bee species as possible to improve the precision of N W estimations. For example, stressors affecting more solitary bees could impoverish communities dominated by few but abundant eusocial species, more than expected by using unweighted abundances.
... Empirical evidence shows that nesting resources affect the abundance of bees. There is a need to study how the availability of natural nesting resources affects solitary bee populations [82][83][84] since this resource is also essential for bee existence. ...