Content uploaded by James Edward Bennett
Author content
All content in this area was uploaded by James Edward Bennett on Oct 22, 2018
Content may be subject to copyright.
An assessment of bumblebee (Bombus spp) land use and floral
preference in UK gardens and allotments cultivated for food
Gemma Foster
1
&James Bennett
2
&Tim Sparks
3
#Springer Science+Business Media New York 2016
Abstract There is increasing interest in the value of domestic
gardens for supporting biodiversity. While it is well
established that bumblebees exploit urban green spaces, this
is the first study to explore the land use and floral preferences
of the UK’s seven most common bumblebees in gardens and
allotments cultivated for food. A citizen science survey was
carried out at 38 sites, between 1st June and 30th September
2013. At the landscape scale, bumblebee abundance and spe-
cies richness was not significantly correlated with surrounding
land use characteristics (both p>0.05).Bombus pratorum was
the only species to show correlations with surrounding land
use, demonstrating a positive relationship with built areas and
gardens and allotments, and a negative correlation with
greenspace and agriculture. At the local site-level scale, bum-
blebee abundance was negatively correlated with areas culti-
vated for vegetables and fruits, and positively correlated with
areas cultivated for flowers, although neither correlation was
statistically significant (p= 0.070 and p= 0.051 respectively).
Bumblebee species richness was not correlated with either
land use (p> 0.05). All bumblebee species were negatively
correlated with areas cultivated for vegetables and fruit, bare
ground and hard paving. Several flowering plants were visited
by all bumblebee species, although relative preferences varied
between bumblebee species. Results emphasise the impor-
tance of including floral resources within garden and allot-
ment areas cultivated for food, and the need for a mosaic of
different flowering plants to cater for varying floral prefer-
ences demonstrated by bumblebee species.
Keywords Garden .Allotment .Bombus .Bumblebee, land
use .Floral resources
Introduction
There have been global declines in populations of insect pol-
linators, including the honeybee (Apis mellifera) and bumble-
bees (Bombus spp) –a trend which has been well documented
across Europe and North America (FAO 2008; Goulson et al.
2008;UNEP2010; Potts et al. 2010). This is of concern be-
cause bees play an essential ecological role as pollinators of a
large number of wild flowers and crops (Klein et al. 2007),
including those that are commonly cultivated in domestic,
community and roof-top gardens and allotments (Matteson
and Langellotto 2009).
The main driver of insect pollinator decline is thought to be
habitat loss, as a result of agricultural intensification and in-
creasing urbanisation (Goulson et al. 2008; Winfree et al.
2009; Potts et al. 2010). Pollinator species richness has been
found to decline with increasing intensification of agricultural
practices (Kremen et al. 2002; Steffan-Dewenter 2003), and
levels of urbanisation (Hernandez et al. 2009; Bates et al.
2011). However, some studies have indicated that urban hab-
itats can actually support high bee species richness (Saure
*Gemma Foster
gemma.foster@coventry.ac.uk
James Bennett
james.bennett@coventry.ac.uk
Tim Sparks
tim.sparks@coventry.ac.uk
1
Centre for Agroecology, Water and Resilience, Coventry University,
Ryton Gardens, Wolston Lane, Coventry CV8 3LG, UK
2
Centre for Agroecology, Water and Resilience, Coventry University,
Ryton Gardens, Wolston Lane, Coventry CV8 3LG, UK
3
Faculty of Engineering, Environment and Computing, Coventry
University, Priory Street, Coventry CV8 3LG, UK
Urban Ecosyst
DOI 10.1007/s11252-016-0604-7
1996;Frankie et al. 2009;Baldock et al. 2015), and certain
taxa such as bumblebees have been found to demonstrate a
positive response to urbanisation (Carré et al. 2009; Bates
et al. 2011).
The value of well managed urban green spaces for native
biodiversity is therefore becoming increasingly recognised
(Goddard et al. 2010). Although the number of urban bee
studies has increased in recent years, urban areas remain
understudied (Baldock et al. 2015). Consequently, little is
known about the diversity and abundance of bees in domestic
gardens and allotments (Hernandez et al. 2009;Mattesonand
Langellotto 2009) or the variables that influence their pres-
ence and diversity (Shwartz et al. 2013).
Domestic gardens are private spaces surrounding dwell-
ings, which comprise a range of features such as lawns, flower
beds and vegetable patches. Allotments are areas reserved for
horticulture where plots are let to individuals for growing
vegetables and flowers. These spaces have high spatial het-
erogeneity as a result of varying land use, management and
plant cultivation at small spatial scales (Thompson et al.
2003). As a result they can provide bees with a wide diversity
of forage (Loram et al. 2007; Frankie et al. 2009). They have
also been found to support high densities of bumblebee
nests compared to agricultural land (Osborne et al.
2008b; Ahrné et al. 2009) and higher bumblebee abun-
dance and species richness than parks and cemeteries
because gardeners often employ techniques that enhance
pollinator habitats (Anderson et al. 2007).
There is great potential for domestic gardens, which ac-
count for between 21 and 27 % of the land in UK cities
(Loram et al. 2007), to be managed for conservation.
Despite this, many gardening recommendations in popular
literature are based on assumptions or informal observation
rather than empirical data (Garbuzov and Ratnieks 2013).
There is also a lack of understanding about how to support
and encourage bee populations in gardens and allotments
(Frankie et al. 2009). This is especially true of areas cultivated
for food, as the majority of advice about enhancing floral
provision focuses on the addition of flowers to ornamental
borders. As food cultivation is becoming an increasingly com-
mon practice in gardens and allotments in industrialised coun-
tries (Lawson 2005), the effect this has on the presence of bees
warrants investigation, especially as evidence suggests that
increasing intensity of garden management reduces bumble-
bee species richness (Smith et al. 2006).
The aim of this study was to explore the presence and
foraging activity of the UK’s seven most common bumblebee
species in gardens and in allotments cultivated for food. We
hypothesised that bumblebee abundance, bumblebee species
richness and the individual bumblebee species would 1) be
positively correlated with built areas and gardens and allot-
ments in the surrounding landscape, 2) be positively correlat-
ed with the cultivation of flowers within allotment and garden
survey sites and that 3) bumblebees would demonstrate rela-
tive preferences in the flowering plants they visited both over-
all and between species.
Materials and methods
Study design
A citizen science approach was used to gather information
about the presence of the UK’s seven most common bumble-
bee species in garden spaces and allotments used for food
cultivation. This enabled us to gather data from a wide variety
of locations across the UK (Fig. 1). A total of 121 participants
were contacted through Garden Organic, a UK charity pro-
moting organic growing. These participants were selected be-
cause they cultivated their gardens or allotments for food
(growing vegetables and/or fruits), and had experience of
bumblebee identification, having previously taken part in a
Garden Organic bumblebee survey. These participants were
therefore ideally placed for completing this survey as they
were confident in participating in citizen science projects,
and were familiar with the seven species of bumblebee we
asked them to monitor. Results were received from 38 partic-
ipants (representing 31 % of those contacted).
Participants were provided with instructions, recording
forms and a bumblebee identification field guide which
contained images and descriptions of the species that formed
part of the study. The species surveyed, which are often re-
ferred to as the ‘big seven’,wereB. terrestris and B. lucorum
(the Buff-tailed Bumblebee and White-tailed Bumblebee,
which were grouped because of the difficulty separating the
species in the field), B. hortorum (Garden Bumblebee),
B. pratorum (Early Bumblebee), B. pascuorum (Common
Carder Bee), B. lapidarius (Red-tailed Bumblebee), and
B. hypnorum (Tree Bumblebee). These species were selected
as they are the most common bumblebee species in the UK
and account for the majority of sightings (Goulson 2010).
With the exception of B. terrestris/lucorum, they are also dis-
tinct enough in appearance to be readily identified.
Bumblebee survey
The survey was undertaken from 1st June 2013 to 30th
September 2013. This period was selected to capture the peak
activity of the survey species (Prys-Jones and Corbet 2011).
Participants were instructed to select a survey site (3 m × 3 m,
rising to 3 m above the ground), which was representative of
the areas cultivated for food in their garden or allotment. They
were asked to make a note of, or mark out, its location so that
they could return to it each week during the survey period.
Participants were instructed to observe and record the bum-
blebees entering the survey site once a week, during a 10 min
Urban Ecosyst
period, when there was sunshine or scattered cloud, and the
temperature was above 15 °C.
Participants were asked to record the total number of each
species of bumblebee observed. In addition, they were asked
to record which flowering plant species each bumblebee vis-
ited and the number of visits made to flowers on that plant.
Participants were asked to record visits to all flowering plants,
including ornamental flowers, fruits, vegetables, herbs and
weeds. For each of the flowering plants visited, they were
asked to estimate the number of floral units of that species
within the survey site, using the following scale: 1 = 1 flower;
2=2–10 flowers; 3 = 11–50 flowers; 4 = 51–100 flowers;
5 = over 100 flowers. One floral unit was defined as a collec-
tion of flowers that a bumblebee could walk around when
foraging, without the need for flight. At the end of the season,
participants submitted their data either by mail or via an online
form. Since bumblebee tongue length affects floral prefer-
ences, the tongue lengths of the survey species were
characterised as long, medium or short, using descriptions in
Prys-Jones and Corbet (2011).
Landscape scale characteristics
The latitude of survey sites was determined from the garden
and allotment addresses provided by participants. The land
use characteristics within 2500 m of the midpoint of each
garden or allotment were categorised using digital data from
the Ordnance Survey MasterMap dataset (accessed through
Fig. 1 Locations of the 38 UK
gardens and allotments surveyed
for bumblebees
Urban Ecosyst
EDINA Digimap in February 2014), and analysed with the
ArcGIS 10.1 (ESRI Redlands, USA) geographical informa-
tion system (GIS). This radius was selected because bumble-
bee species may travel kilometres from their nest sites to for-
age within the surrounding landscape.For example, field stud-
ies have reported B. terrestris workers to forage at least 1.5 km
(Osborne et al. 2008a) and 2.2 km (Kreyer et al. 2004)from
their nests.
Based on previous work by Bates et al. (2011) and Ahrné
et al. (2009), five broad land use types were defined and attri-
butes from several OS MasterMap fields were grouped to
represent each category. The five categories (and their OS
MasterMap fields) were: Built (manmade buildings, glass-
houses, roads, rail, manmade tracks and paths), Gardens and
Allotments (multiple make general surface - which was com-
prised of lawns, vegetable/flower beds, trees, garages/sheds/
greenhouses, paving), Greenspace and Agriculture (natural
land, rough grassland, scrub, rock, boulders, natural tracks
and paths), Trees (coniferous trees, non-coniferous trees, or-
chards), and Water (tidal water, inland water, marsh reeds,
saltmarsh). A small percentage of the areas remained unclas-
sified (mean 0.4 % of the landscape) as the MasterMap data
included unclassified areas and minor gaps in the data.
Site-level characteristics
Participants were asked to provide details about site-level
characteristics by estimating the percentage ground cover ded-
icated to each of the following land use types within their
survey site: vegetable and fruit cultivation, cultivated flowers,
hard paving, lawn, trees, shrubs/hedges, waste ground/rough
habitat, water and bare soil. Participants were also asked to
score their gardening practices on a scale ranging from one to
five, with one representing frequent management with syn-
thetic chemicals, and five being exclusively organic.
Statistical analysis
For the purposes of analysis the B. terrestris/lucorum group
was treated as one species. Thus analyses considered six spe-
cies groups. Analysis was based on standard correlation and
regression techniques as described below, and carried out
using Minitab 17, Minitab Inc. Redundancy Analysis ordina-
tion was carried out in Canoco 5.
Weighted regression was conducted to investigate the ef-
fect of time on the abundance of each bumblebee species, total
bumblebee abundance and bumblebee species richness (the
number of bumblebee species) using mean values on week
number (weights in regression = number of recorders).
Correlation analysis was used to investigate the relation-
ships between latitude and the abundance of each bum-
blebee species, total bumblebee abundance and bumble-
bee species richness.
Redundancy Analysis ordination (RDA) was carried out to
compare the abundances of the six species groups with the
land use characteristics within 2500 m of the survey sites,
and with the land use characteristics within the 3 m × 3 m
survey sites. Total bumblebee abundance and species richness
were compared to the same land use characteristics at each
scale using Pearson correlations, data were checked for nor-
mality prior to analysis and no significant deviation was
detected.
To identify which flowering plants were preferred by bum-
blebees, flower abundance categories were converted to the
midpoint of their range and summed across each plant species.
Similarly, the total number of bumblebees feeding on each
plant species was summed. The log (x + 1) number of bum-
blebees was then regressed on log (x + 1) flower abundance
(R
2
=64.2%,p< 0.001), using only those flower species/
cultivars (n= 94) with at least 3 records, and standardised
residuals stored. A cut-off of +1.645 (i.e. the expected 95th
percentile) for the standardised residuals was used to identify
plant species on which bumblebee feeding was greater than
expected given its flower abundance.
Preferences relative to overall feeding patterns were esti-
mated for individual bumblebee species. This involved
regressing the number of feeding records of the bumblebee
species on the total number of bumblebee feeding records
for each plant species. Again standardised residuals of
+1.645 were used as a cut-off point to identify relative prefer-
ences. This process was repeated for each bumblebee species
and was restricted to the same plant species as for total
abundance.
Results
Bumblebee surveys
The majority of participants (34, 89 %) surveyed gardens,
while the remaining four participants surveyed allotments.
Participants’survey sites were located across the United
Kingdom –the majority were located in England, two were
in Scotland and one was in Wales (Fig. 1). The 38 participants
submitted data from a total of 363 survey occasions undertak-
en between June and September 2013. The majority of survey
sites (57 %) were visited by five or six of the six bumblebee
species groups monitored (Fig. 2).
A total of 2621 individual bumblebees were observed
(Table 1). B. terrestris/lucorum was the mostly frequently ob-
served species group, accounting for 50 % of sightings and
was the only species group present at all 38 sites. All other
species were relatively common, being present at the majority
of survey sites, and accounting for 7 to 16 % of observations.
Between 3 and 16 (mean 9.6) weekly observations were made
by each participant. There were significant reductions in
Urban Ecosyst
species richness (p< 0.05), and in the number of both
B. pratorum (p<0.001)andB. hypnorum (p< 0.05) as the
season progressed (Fig. 3).
Landscape scale characteristics
Latitude was not found to have a significant effect on the
abundance of each bumblebee species, total bumblebee abun-
dance, or bumblebee species richness (all p>0.05).Landuse
characteristics within 2500 m of survey sites were not signif-
icantly correlated with total bumblebee abundance or bumble-
bee species richness (both p > 0.05).
The first two axes of RDA explained 16.3 % (13.0 % and
3.3 % respectively) of the relationship between bumblebee
species abundances and surrounding land use characteristics
(Fig. 4). The more important first axis may be considered to be
a gradient from open green spaces to urban areas and gardens
and allotments; the second axis is difficult to interpret. Most
species had short vectors indicating the lack of a relationship
with surrounding land use characteristics, although the vector
for B. pratorum was close to (and hence positively correlated
with) vectors for built areas, and gardens and allotments, and
opposed to (and hence negatively correlated with) the vector
for greenspace and agriculture.
Site-level characteristics
The majority of participants (33, 87 %) rated their garden
management practices as organic (4 or 5 on the scale), and
so the range was not sufficient to allow further investigation
into the effects of management practices. The most common
land use within the survey sites was food growing (mean 42 %
of the area), followed by the cultivation of flowers (mean 26 %
of the area).
Total bumblebee abundance was positively correlated with
the percentage of the survey site used for cultivated flowers,
and negatively correlated with percentage of the survey site
used for food production, although neither correlation was
statistically significant (p= 0.051 and p= 0.070 respectively).
Correlations between bumblebee species richness and the per-
centage of the survey site used for flower cultivation and food
production were not significant (both p>0.05).
The first two axes of RDA explained 15.9 % (10.2 % and
5.7 % respectively) of the relationship between bumblebee
species abundances and site-level land use characteristics
(Fig. 5). The first axis can be interpreted as a gradient from
heavily managed sites to those containing more woody spe-
cies; the second axis as a gradient from flower rich gardens to
those dominated by lawns and vegetable production. Vectors
for all bumblebee species were opposed to (and hence nega-
tively correlated with) vectors for fruit and vegetable cultiva-
tion, bare ground and hard paving. Vectors for B. lapidarius
and B. pratorum were also close to (and hence positively
correlated with) vectors for flowers and waste ground.
Fig. 2 Frequency distribution of
bumblebee species richness
across the 38 UK survey sites
Tabl e 1 Presence of the bumblebees surveyed in 38 UK gardens and
allotments cultivated for food
Number (%) of sites with
species present (n=38)
Number (% of total) of
bumblebees recorded
B. terrestris/
lucorum
38 (100 %) 1313 (50 %)
B. hortorum 25 (68 %) 300 (11 %)
B. pratorum 22 (58 %) 169 (6 %)
B. pascuorum 32 (84 %) 406 (16 %)
B. lapidarius 32 (84 %) 239 (9 %)
B. hypnorum 23 (61 %) 194 (7 %)
2621
Urban Ecosyst
Floral preferences
There were 12,105 observations of bumblebees visiting 164
flowering plant species. The majority of bumblebee visits
were made to a relatively small proportion of the flowers
available, with 51 % of bumblebee visits being made to 37
plant species (24 % of the plants surveyed), which accounted
for 10 % of the floral units available. A small selection of plant
species were visited by all six bumblebee species groups, in-
cluding Meadow Cranesbill (Geranium pratense), Blackberry
(Rubus fruticosus), Borage (Borago officinalis) and Lavender
(Lavandula angustifolia).
Regression analysis identified 32 plant species where differ-
ences in feeding preferences could be detected (Table 2). The
plant species which were relatively preferred by bumblebees
overall were Meadow Cranesbill (Geranium pratense), Green
Alkanet (Pentaglottis sempervirens), Sage (Salvia officinalis),
Fig. 3 Plots between weeks 22
and 39 (1st June and 29th
September 2013) showing mean
numbers per location of each
bumblebee species, mean
bumblebee abundance and mean
number of species
Fig. 4 RDA biplot examining relationship between bumblebee
abundances and land use characteristics within 2500 m. Most species
had short vectors indicating the lack of a relationship with surrounding
land use characteristics, although B. pratorum was positively correlated
with built areas, and gardens and allotments, and negatively correlated
with greenspace and agriculture. Species abbreviations: B. terr/luco = B.
terrestris/lucorum group, B. hort = B. hortorum, B. prat = B. pratorum,
B. pasc = B. pascuorum, B. lap = B. lapidarius, B. hyp = B. hypnorum.
Land use abbreviations: Built = built areas, G & A = gardens and
allotments, Greenspace = greenspace and agriculture, Trees etc. = trees/
woodlands/ orchards, and Water = water
Fig. 5 RDA biplot examining relationship between bumblebee
abundances and land use characteristics within the 3 m × 3 m survey
sites in gardens and allotments. All species were negatively correlated
with fruit and vegetable cultivation, bare ground and hard paving, and
B. lapidarius and B. pratorum were positively correlated with cultivated
flowers and waste ground. Species abbreviations: B. terr/luco = B.
terrestris/lucorum group, B. hort = B. hortorum, B. prat = B. pratorum,
B. pasc = B. pascuorum, B. lap = B. lapidarius, B. hyp = B. hypnorum.
Land use abbreviations: Veg = vegetable and fruit cultivation, Flowers =
cultivated flowers, Hard paved = hard paved, Lawn = lawn, Trees = non-
crop trees, Shrubs = shrubs and hedges, Waste = waste ground/ rough
habitat, Water = water, and Bare soil = bare soil
Urban Ecosyst
Mallow (Lavatera spp), and Marjoram (Origanum vulgare).
Relative preferences varied between bumblebee species.
Discussion
Our results indicate that the bumblebee species surveyed are
frequent visitors to spaces cultivated for food in UK gardens
and allotments. The majority of survey sites were visited by
five or six of the six bumblebee species groups surveyed,
indicating that these common species are fairly ubiquitous
within these spaces. The number of bumblebee sightings,
and the presence of each bumblebee species varied through
the season, probably as a result of phenological differences in
the bumblebee species studied.
In contrast to expectation, bumblebee abundance and spe-
cies richness were not correlated with surrounding land use
characteristics. A similar study of Swedish allotments also
reported landscape to have limited on bumblebee abundance,
but did find bumblebee species richness to be negatively cor-
related with increasing proportions of built areas/impervious
surfaces (Ahrné et al. 2009). If our study had surveyed all 25
UK bumblebee species a similar pattern may have emerged, as
evidence suggests that urban areas generally contain fewer
Tabl e 2 Standardised residuals for floral preferences of bumblebees relative to the overall mean. Standardised residuals of +1.645 were used as a cut-
off to identify relative preferences and these are shown in bold. Tongue length categories: Lt = Long tongue, Mt. = Medium tongue, St = Short tongue
Plant species All bumble-
bees
B. terrestris/
lucorum
B. hortorum B. pratorum B. pascuorum B. lapidarius B. hypnorum
(St) (Lt) (St) (Lt) (Mt) (St)
Meadow Cranesbill (Geranium pratense)3.08 0.01 0.18 2.47 2.72 3.05 0.21
Green Alkanet (Pentaglottis sempervirens)2.66 -0.55 -1.19 1.59 1.36 2.16 3.75
Sage (Salvia officinalis)2.49 -0.14 -1.12 -0.81 2.79 -1.00 -0.89
Mallow (Lavatera spp) 2.17 2.09 1.00 2.00 1.02 1.17 -0.32
Marjoram (Origanum vulgare)1.70 1.57 1.46 -0.87 1.26 0.18 2.48
Blackberry (Rubus fruticosus) 1.43 0.93 -0.26 2.55 0.66 0.46 2.32
Oriental Poppy (Papaver orientale)1.301.251.98 -0.07 -0.45 -0.39 -0.22
Opium Poppy (Papaver somniferum)1.241.68 1.06 -0.30 -0.29 0.65 -0.42
Borage (Borago officinalis) 1.04 0.5 4 0.74 2.13 1.61 1.82 1.47
Perennial Cornflower (Centaurea montana) 0.99 -0.81 -0.66 -0.38 0.80 2.11 -0.50
Cornflower (Centaurea cyanus) 0.92 0.5 2 0.29 2.21 -1.16 1.59 0.65
Russian Comfrey (Symphytum x uplandicum) 0.89 1.12 -1.02 2.37 -1.33 0.50 0.76
Foxglove (Digitalis purpurea)0.860.801.90 -1.14 1.35 -1.28 0.29
Common Bugle (Ajuga reptans)0.780.191.96 -0.77 0.92 1.13 -0.85
Geranium ‘Rozeanne’(Geranium ‘Rozeanne’
Gerwat)
0.77 1.19 -0.98 1.51 -1.92 1.85 1.03
Small Globe Thistle (Echinops ritro) 0.73 1.20 -1.09 -0.78 -1.41 1.70 -0.34
Red Campion (Silene dioica)0.720.322.49 -1.04 -0.44 -1.19 -1.09
Common Snowberry (Symphoricarpos albus) 0.53 -0.20 -1.06 1.90 -1.36 -0.95 2.26
Clover (Trifoliu m spp) 0.50 -0.13 -0.71 0.84 0.55 1.68 -0.54
Bear’sBreeches(Acanthus mollis) 0.43 -1.96 -0.45 -0.18 1.67 -0.48 -0.32
Sedum (Sedum spp) 0.37 0.92 0.06 0.39 -0.59 1.76 -1.20
White Campion (Silene latifolia subsp.Alba) 0.37 -2.00 1.98 0.66 -0.63 -0.50 -0.34
Poached Egg Plant (Limnanthes douglasii) 0.22 0.26 -1.42 2.02 -0.08 1.62 -1.14
Lavender (Lavandula angustifolia) 0.21 0.15 0.90 -0.39 1.27 0.16 1.72
Larkspur (Delphinium spp) 0.14 -2.05 1.85 -0.76 0.95 -0.96 -0.02
Raspberry (Rubus idaeus) 0.08 -0.69 0.30 1.93 -0.25 -1.43 2.50
Rock crane’s-bill (Geranium phaeum) 0.04 0.53 -1.21 2.20 -1.56 -1.07 -0.96
Broad bean (Vicia faba) -0.13 -1.26 -1.52 1.69 1.24 -0.35 0.78
Hollyhock (Alcea rosea) -0.23 -1.79 1.88 -0.77 0.00 -0.21 -0.85
Field Scabious (Knautia arvensis) -0.41 -1.58 1.75 -0.77 0.15 -0.97 -0.85
Honeywort (Cerinthe major ‘Purpurascens’) -0.48 -0.79 -1.38 1.74 -0.28 1.38 -1.10
Montbretia (Crocosmia spp) -0.68 -0.96 -0.92 -0.62 -0.26 -0.85 1.70
Urban Ecosyst
floral specialists and rare species (Frankie et al. 2009;
Hernandez et al. 2009; Bates et al. 2011). In particular, the
rare species of bumblebee in the UK have been found to be
less generalist/polylectic than the common species which are
able to exploit the wide range of floral resources available in
gardens (Goulson et al. 2005).
Bombus pratorum was the only species to show correla-
tions with the surrounding land use characteristics, demonstrating
a positive relationship with built areas, gardens and allotments,
and a negative relationship with greenspace and agriculture. This
species appears to be becoming increasingly urbanised due to its
intolerance of agriculture (Benton 2000), and our results suggest
urban green spaces are important habitats for this species. This
could be because B. pratorum is a small species with a short
foraging distance. It also often nests in cavities and bird boxes
(Prys-Jones and Corbet 2011; Lye et al. 2012) and these features
are common in urban areas and gardens.
Many studies have suggested that local factors such as site
management, flower abundance and plant diversity have a
greater effect on bee presence than surrounding landscape
(Ahrné et al. 2009; Frankie et al. 2009; Pardee and Philpott
2014). Our results indicate that bumblebee abundance is neg-
atively correlated with increasing proportions of food cultiva-
tion, and all bumblebee species demonstrated a negative re-
sponse to this local land use. Bumblebee abundance was pos-
itively correlated with increasing flower cultivation. This is
supported by other studies which have reported bumblebee
abundance to be positively related to the proportion of
flowering plants (Ahrné et al. 2009), as well as floral abun-
dance and plant diversity (Hennig and Ghazoul 2012).
Bumblebee species richness however was not affected by
site-level land use, again possibly because the survey was
limited to the most common bumblebee species. Other studies
have reported an increase in bee species richness with increas-
ing floral diversity (Smith et al. 2006) and floral area
(Matteson and Langellotto 2010), and a reduction in bumble-
bee species richness with increasing intensity of garden man-
agement (Smith et al. 2006). This highlights the need for fur-
ther detailed assessments of garden management practices,
also called for by Shwartz et al. (2013).
Our results suggest the importance of including flowers in
areas cultivated for food, in order to maximise floral resource
availability within these areas. This message is likely to be-
come increasingly important if food production continues to
grow in popularity in domestic gardens and allotments. This
does not necessarily need to be at the expense of production in
vegetable patches or allotments, as bumblebees demonstrated
relative preferences for several herbs, fruits and
vegetables. The provision of floral resources may also
contribute to improved yields of insect pollinated crops.
Recent research by Potter and LeBuhn (2015) found
that tomato pollination in gardens was positively corre-
lated with floral resource density within survey gardens,
as an indirect effect of increased abundance of bees or
increased visitation rates.
Although 164 plant species were visited during this study, the
majority of visits were made to a relatively small number of
flowering plants. Half of all bumblebee visits were made to 37
plant species representing only 10 % of the available floral units.
Similar figures were reported by Ahrné et al. (2009), where al-
though 168 plant species were visited by bumblebees, half of the
visits were made to just 14 plant species. Some of the flowering
plant species monitored in this study were visited by all bumble-
bee species groups studied, including Meadow Cranesbill
(Geranium pratense), Blackberry (R. fruticosus), Borage
(B. officinalis)and Lavender (L. angustifolia). These plants can
therefore be recommended as useful species for supporting the
UK’s common bumblebee species.
As expected, the relative preferences varied between bum-
blebee species, and the majority of the variation can be ex-
plained by the morphological differences in tongue length
(Prys-Jones and Corbet 2011). Meadow Cranesbill
(G. pratense), the most visited flower overall, was especially
popular with B. pratorum, B. lapidarius and B. pascorum
(short, medium and long-tongued species respectively), which
all showed a significant relative preference for the flower.
Many Geranium species are known to be popular with bees
because their open flowers enable a range of species with
varying morphology to access pollen and nectar resources
within the flower (Kirk and Howes 2012).
Few weed species were visited by bumblebees during this
study, and only four (Thistle (Cirsium spp), Red Deadnettle
(Lamium purpureum), Common Bugle (Ajuga reptans)and
Clover (Trifolium spp)) featured in the list of the 37 most visited
plant species. This is probably a result of the intensive weeding
practices commonly employed in spaces cultivated for food. As
native plants are an important source of forage in the wider
landscape, especially in the spring, gardeners may benefit from
education about the value of certain flowering weeds, which
could be removed after flowering but before seed set.
Although this study did not explore the huge variety of
plants available to gardeners, with over 75,000 being listed
in the RHS Plant Finder (Armitage 2016), it does highlight
some of the more commonly grown plants and how they are
used by bumblebees. Although plants such as Meadow
Cranesbill (G. pratense) had general appeal, many bumblebee
species demonstrated relative preferences for other flowers.
This emphasises the need for a mosaic of different flowering
plants that can cater for the needs of the various bumblebee
species in both space and time.
Conclusion
This is the first study to explore the land use and floral pref-
erences of bumblebees in gardens and allotments cultivated
Urban Ecosyst
for food across the UK. Although bumblebee abundance and
species richness was not significantly correlated with sur-
rounding land use characteristics, our results suggest that gar-
dens and allotments are important habitats for B. pratorum.
Our results also highlight the importance of local site-level
land use, as all bumblebee species were negatively correlated
with vegetable and fruit cultivation, while total bumblebee
abundance was positively correlated with flower cultivation.
This emphasises the importance of including flowers in areas
cultivated for food, in order to maximise floral resource avail-
ability within these areas.
Many gardeners are keen to support pollinators by growing
appropriate flowers and there are numerous lists available to
guide them in their plant choices. The preferred plants identi-
fied in this study generally compared well with popular lists of
plants designed to support pollinators (such as Royal
Horticultural Society 2016a;2016b). However, several plant
species that bumblebees demonstrated relative preferences for
in our study do not feature in the lists. This highlights the need
for expanding current lists or creating lists designed to support
pollinators in spaces cultivated for food, for which there is
limited information available.
Although the survey period covered the flight activity of all
the common bumblebee species, future studies would benefit
from an earlier start date to capture information about the
activity of emerging queens. There is also a need to study all
25 UK bumblebee species, as rarer species are in greater need
of conservation support than the common species, which tend
to be more generalist, with broader tolerances.
Acknowledgments We thank the Garden Organic members who kind-
ly volunteered their time to collect data for this project. We also thank
Stuart Gill for his work carrying out GIS analysis, as well as Barbara
Smith, Judith Conroy and two anonymous reviewers for their comments
which improved this manuscript. This study was funded by Coventry
University.
References
Ahrné K, Bengtsson J, Elmqvist T (2009) Bumble bees (Bombus spp)
along a gradient of increasing urbanization. PLoS One 4:e5574.
doi:10.1371/journal.pone.0005574
Anderson E, Barthel S, Ahrné K (2007) Measuring social-ecological
dynamics behind the generation OF ecosystem services. Ecol Appl
17:1267–1278. doi:10.1890/06-1116.1
Armitage J (2016) RHS Plant Finder 20 16, 30th edn. Royal Horticultural
Society, London
Baldock KCR, Goddard MA, Hicks DM et al (2015) Where is the UK’s
pollinator biodiversity ? The importance of urban areas for flower-
visiting insects. Proc R Soc B Biol Sci. doi:10.1098/rspb.2014.2849
Bates AJ, Sadler JP, Fairbrass AJ et al (2011) Changing bee and hoverfly
pollinator assemblages along an urban-rural gradient. PLoS One 6:
e23459. doi:10.1371/journal.pone.0023459
Benton TG (2000) The Bumblebees of Essex. Lopinga Books
Carré G, Roche P, Chifflet R et al (2009) Landscape context and habitat
type as drivers of bee diversity in European annual crops. Agric
Ecosyst Environ 133:40–47. doi:10.1016/j.agee.2009.05.001
FAO (2008) Rapid assessment of pollinators’status a contribution to the
international initiative for the conservation and sustainable use of
pollinators. Food and Agriculture Organization of the United
Nations, Rome 2008.
Frankie GW, Thorp RW, Pawelek JC et al (2009) Urban bee diversity in a
small residential garden in northern California. J Hymenopt Res 18:
368–379
Garbuzov M, Ratnieks FLW (2013) Quantifying variation among garden
plants in attractiveness to bees and other flower-visiting insects.
Funct Ecol 28:364–374. doi:10.1111/1365-2435.12178
Goddard M a, AJ D, Benton TG (2010) Scaling up from gardens: biodi-
versity conservation in urban environments. Trends Ecol Evol 25:
90–98. doi:10.1016/j.tree.2009.07.016
Goulson D (2010) Bumblebees: Behaviour, ecology and conservation.
Oxford University Press, USA
Goulson D, Hanley ME, Darvill B et al (2005) Causes of rarity in bum-
blebees. Biol Conserv 122:1–8. doi:10.1016/j.biocon.2004.06.017
Goulson D, Lye GC, Darvill B (2008) Decline and conservation of bum-
ble bees. Annu Rev Entomol 53:191–208. doi:10.1146/annurev.
ento.53.103106.093454
Hennig EI, Ghazoul J (2012) Pollinating animals in the urban environ-
ment. Urban Ecosyst 15:149–166. doi:10.1007/s11252-011-0202-7
Hernandez JL, Frankie GW, Thorp RW (2009) Ecology of urban bees: a
review of current knowledge and directions for future study. Cities
Environ 2:1–15
Kirk WDJ, Howes FN (2012) Plants for Bees. A Guide to the Plants that
Benefit the Bees of the British Isles, International Bee Research
Association
Klein A-M, Vaissière BE, Cane JH et al (2007) Importance of pollinators
in changing landscapes for world crops. Proc R Soc B Biol Sci 274:
303–313. doi:10.1098/rspb.2006.3721
Kremen C, Williams NM, Thorp RW (2002) Crop pollination from native
bees at risk from agricultural intensification. Proceedings of the
National Academy of Sciences 99(26):16812–16816
Kreyer D, Oed A, Walther-Hellwig K, Frankl R (2004) Are forests po-
tential landscape barriers for foraging bumblebees? Landscape scale
experiments with Bombus terrestris agg. and Bombus pascuorum
(Hymenoptera, Apidae). Biol Conserv 116(1):111–118
Lawson LJ (2005) City bountiful: a century of community gardening in
America. University of California Press, Berkeley
Loram A, Tratalos J, Warren PH, Gaston KJ (2007) Urban domestic
gardens (X): the extent & structure of the resource in five major
cities. Landsc Ecol 22:601–615
Lye GC, Osborne JL, Park KJ,Goulson D (2012)Usingcitizen science to
monitor Bombus populations in the UK: nesting ecology and rela-
tive abundance in the urban environment. J Insect Conserv 16:697–
707. doi:10.1007/s10841-011-9450-3
Matteson KC, Langellotto G (2009) Bumble bee abundance in New York
City Community gardens: implications for urban agriculture. Cities
Environ 2:1–12
Matteson KC, Langellotto GA (2010) Determinates of inner city butterfly
and bee species richness. Urban Ecosyst 13:333–347
Osborne JL, Martin AP, Carreck NL et al (2008a) Bumblebee flight
distances in relation to the forage landscape. J Anim Ecol 77:406–
415. doi:10.1111/j.1365-2656.2007.01333.x
Osborne JL, Martin AP, Shortall CR et al (2008b) Quantifying and com-
paring bumblebee nest densities in gardens and countryside habitats.
J Appl Ecol 45:784–792. doi:10.1111/j.1365-2664.2007.01359.x
Pardee GL, Philpott SM (2014) Native plants are the bee’sknees:local
and landscape predictors of bee richness and abundance in backyard
gardens. Urban Ecosyst 17:1–19. doi:10.1007/s11252-014-0349-0
Urban Ecosyst
Potter A, LeBuhn G (2015)Pollination service to urban agriculture in San
Francisco, CA. Urban Ecosyst 18:885–893. doi:10.1007/s11252-
015-0435-y
Potts SG, Biesmeijer JC, Kremen C et al (2010) Global pollinator de-
clines: trends, impacts and drivers. Trends Ecol Evol 25:345–353.
doi:10.1016/j.tree.2010.01.007
Prys-Jones OE, Corbet SA (2011) Bumblebees, 3rd edn. Pelagic
Publishing
Royal Horticultural Society (2016a) RHS Perfect for Pollinators: Garden
Plants.
Royal Horticultural Society (2016b) RHS Perfect for Pollinators:
Wild flowers.
Saure C (1996) Urban habitats for bees: the example of the city of Berlin.
In: Matheson A, Buchmann S, Toole C, et al. (eds) The conservation
of bees. Linnean Society Symposium Series No. 18. New York, NY:
Academic Press, pp 47–53
Shwartz A, Muratet A, Simon L, Julliard R (2013) Local and manage-
ment variables outweigh landscape effects in enhancing the diversity
of different taxa in a big metropolis. Biol Conserv 157:285–292.
doi:10.1016/j.biocon.2012.09.009
Smith RM, Warren PH, Thompson K, Gaston KJ (2006) Urban domestic
gardens (VI): environmental correlates of invertebrate species rich-
ness. Biodivers Conserv 15:2415–2438. doi:10.1007/s10531-004-
5014-0
Steffan-Dewenter I (2003) Importance of habitat area and landscape con-
text for species richness of bees and wasps in fragmented orchard
meadows. Conserv Biol 17:1036–1044. doi:10.1046/j.1523-
1739.2003.01575.x
Thompson K, Austin KC, Smith RM et al (2003) Urban domestic gardens
(I): putting small-scale plant diversity in context. J Veg Sci 14.
doi:10.1658/1100-9233(2003)014[0071:UDGIPS]2.0.CO;2
UNEP (2010) Global honey bee colony disorders and other threats to
insect pollinators. United Nations Environment Programme.
Winfree R, Aguilar R, Vazquez DP et al (2009) A meta-analysis of bees’
responses to anthropogenic disturbance. Ecology 90. doi:10.1890
/08-1245.1
Urban Ecosyst
- A preview of this full-text is provided by Springer Nature.
- Learn more
Preview content only
Content available from Urban Ecosystems
This content is subject to copyright. Terms and conditions apply.
