文档库 最新最全的文档下载
当前位置:文档库 › A comparison of bee communities

A comparison of bee communities

Landscape and Urban Planning 103 (2011) 102–108

Contents lists available at ScienceDirect

Landscape and Urban

Planning

j o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /l a n d u r b p l a

n

A comparison of bee communities of Chicago green roofs,parks and prairies

Rebecca Tonietto a ,b ,?,Jeremie Fant b ,John Ascher c ,Katherine Ellis d ,Daniel Larkin b

a

Plant Biology and Conservation,Northwestern University,Evanston,IL 60208,USA

b

Division of Plant Science and Conservation,Chicago Botanic Garden,Glencoe,IL 60022,USA c

Division of Invertebrate Zoology,American Museum of Natural History,Central Park West at 79th St.,New York,NY 10024-5192,USA d

Department of Biology,Kalamazoo College,Kalamazoo,MI 49006,USA

a r t i c l e

i n f o

Article history:

Received 11April 2011

Received in revised form 30June 2011Accepted 7July 2011

Available online 2 August 2011

Keywords:Green roofs Native bees

Pollinator conservation Urban ecology

Hymenoptera:Apoidea:Anthophila

a b s t r a c t

Due to loss of natural habitats,human-dominated green spaces are likely to increase in importance for biodiversity support.We assessed the potential value of urban “green roofs”for native pollinator conservation in the Chicago region,comparing them with reference habitats of tallgrass prairie natural areas and traditional city-park green spaces.We found that native bees are present on green roofs,though at lower abundance and diversity than in reference habitats.Green-roof and prairie bee communities were distinct from each other,while those in parks were intermediate and similar to the other two habitat types.Bee-community patterns were related to habitat characteristics at both the site and landscape scales.Overall,bee abundance and species richness increased with greater proportions of green space in the surrounding landscape.However,this relationship disappeared in cases where green space was dominated by turf grass.At the site scale,bees bene?ted from greater plant diversity,and bee and plant-community composition were signi?cantly correlated.Green roofs are potentially valuable sites for bee conservation in urban areas,particularly if planted with diverse native forbs to provide foraging resources,and designed to accommodate bees with different nesting habits.

? 2011 Elsevier B.V. All rights reserved.

1.Introduction

Certain pollinators have declined globally due to habitat loss and other land-use changes (Murray,Kuhlmann,&Potts,2009;Winfree,Aguilar,Vasquez,LeBuhn,&Aizen,2009).Long-term declines of honey bees and some wild bees,particularly bum-ble bees,have been documented in North America (Grixti,Wong,Cameron,&Favret,2009;National Research Council,2006).Accord-ing to some,this decline has already reached a crisis stage (Klein,Steffan-Dewenter,Buchori,&Tscharntke,2002;Kremen &Ricketts,2000;Larsen,Williams,&Kremen,2005;Steffan-Dewenter,Potts,&Packer,2005;Williams &Kremen,2007).However,studies in urban,agricultural,and natural systems have continued to ?nd most wild bees in expected diversity and abundance according to historical records (Cane,2001;Giles &Ascher,2006;Marlin &LaBerge,2001;Tuell,Ascher,&Issacs,2009).To better under-stand,monitor and support native pollinators habitat needs,more pollinator research should be included within restoration ecol-?Corresponding author at:Chicago Botanic Garden c/o Plant Science Center,1000Lake Cook Rd.,Glencoe,IL 60022,USA.Tel.:+18478356991.

E-mail addresses:rebeccatonietto2008@https://www.wendangku.net/doc/118960670.html,

(R.Tonietto),jfant@https://www.wendangku.net/doc/118960670.html, (J.Fant),ascher@https://www.wendangku.net/doc/118960670.html, (J.Ascher),katherine.ellis09@https://www.wendangku.net/doc/118960670.html, (K.Ellis),dlarkin@https://www.wendangku.net/doc/118960670.html, (https://www.wendangku.net/doc/118960670.html,rkin).

ogy,especially within anthropogenically altered landscapes (Dixon,2009).

There is an emerging recognition that properly designed and managed human-dominated landscapes can play an important role in biodiversity support,allowing native species to continue to col-onize sites that have been altered from their natural state (Daily,1997;Rosenzweig,2003).The protection and restoration of nat-ural areas are critical.Remnant habitats are often insuf?cient to conserve biodiversity in urban areas making these managed,or cre-ated habitats even more valuable (Daily,2006;Rosenzweig,2003).There is growing evidence that substantial components of native bee communities can persist in anthropogenic landscapes (Cane,2001;Hernandez,Frankie,&Thorp,2009;Matteson,Ascher,&Langellotto,2008).

Worldwide,native bee abundance and diversity in agricultural systems are generally positively correlated with proximity to and proportion of natural areas in the surrounding landscape (Kremen,Williams,Bugg,Fay,&Thorp,2004;Ricketts et al.,2008).However,anthropogenic habitats can also be of high value to native bees.For example,wild bee abundance and diversity were greater in agri-cultural and suburban areas of New Jersey than within forested areas,the dominant pre-settlement land cover (Winfree,Griswold,&Kremen,2007).Other agricultural habitats,such as blueberry ?elds,are known to support very diverse communities of native bees (Tuell et al.,2009).

0169-2046/$–see front matter ? 2011 Elsevier B.V. All rights reserved.doi:10.1016/https://www.wendangku.net/doc/118960670.html,ndurbplan.2011.07.004

R.Tonietto et al./Landscape and Urban Planning103 (2011) 102–108103

While bee communities associated with agriculture are well-documented,few replicated studies have compared bee com-munities across multiple urban habitats(Cane,Minckley,Kervin, Roulston,&Williams,2006;Hernandez et al.,2009).Urban bee studies have tended to focus on one habitat type,such as natural-area remnants(Cane et al.,2006;Hisamatsu&Yamane,2006)or gardens or parks(Fetridge,Ascher,&Langellotto,2008;Frankie et al.,2009;Hernandez et al.,2009;Matteson et al.,2008; McFrederick&LeBuhn,2006)without detailed assessment of plant community composition or surrounding land uses(Hernandez et al.,2009;but see Winfree et al.,2007).One recent stud-ies from the Neotropics found lower bee diversity at an urban site than a nearby ecological reserve(Dalmazzo,2010)whereas another found the impact of the urban matrix to be minimal and that?oral resource abundance and distribution were cor-related with bee abundance in an urban landscape(Wojcik, 2011).

The conservation value of novel urban green spaces should be investigated further,as their value for biodiversity support is often unknown(Rosenzweig,2003).This is especially true of green roofs, a rare example of an urban green space rapidly increasing in area in North America(Green Roofs for Healthy Cities,2009).Green roofs are typically?at or slightly sloped rooftops with soil substrates to support vegetation.While green roofs often house honey-bee hives(Shevory,2010),we know of only two published study that assessed green roofs as potential habitat for native bees in North America(Colla,Willis,&Packer,2009;MacIvor&Lundholm,2010). Bee species diversity was not statistically compared by Colla et al. (2009),and MacIvor and Lundholm(2010)focused on insect counts, and compared collections from green roofs and ground sites with-out species identi?cation.

We chose the city of Chicago and surrounding metropolitan areas in northeastern Illinois to test whether green roofs support wild bee communities.In Illinois,loss of natural land cover has been severe.From1972to1997,the proportion of developed land in the metropolitan region increased by49%(Wang&Moskovits,2001). Over80%of the state’s total land cover is agricultural,and only 0.001%of tallgrass prairie,the dominant pre-settlement ecosys-tem,remains(Illinois Department of Agriculture,2009;National Research Council,2006).However,Chicago has become the lead-ing city in the United States for green roof implementation;as of 2010,over?ve hundred were extant or in development(Kamin, 2010).

We investigated green roofs as habitat for native bees by com-paring them against a reference natural habitat(tallgrass prairie) and traditional urban green space(ecologically managed areas in city parks).Within these three habitat types,we assessed the abundance,diversity and bee community composition.We tested the extent to which bee community patterns could be explained by habitat characteristics at the site and landscape scales in the form of plant community composition and land-use cover,respec-tively.

2.Materials and methods

2.1.Site descriptions

We collected and observed bees at six green roofs,six city parks, and six prairies from June to October2008.All sites were within the greater metropolitan Chicago(Cook,Lake,and Will counties; IL).Entire park area ranged from6to485ha,yet the natural area plantings were all less than2ha in area.The smallest prairie was also6ha,but the largest was slightly over6000ha.

The vegetation of two green roofs was dominated by plants native to Illinois.The other four were dominated by Sedum spp.;short-statured succulents that quickly produce dense mats of?ow-ering vegetation.Planted areas of green roofs ranged from0.01to 0.23ha,and were located on buildings ranging from2to15sto-ries tall.All park sites had small,managed areas of native prairie plants.Prairie sites were located outside of Chicago in Cook,Lake,or Will Counties and have undergone active management and varying degrees of restoration.

We characterized the surrounding landscape and vegetation of each site.We imported Google Earth(Version4.3.7204.0836,2008 Google)images into ArcMap(ArcGIS version9.2,2006ESRI)and quanti?ed land cover categories(urban,suburban,water,and green space)within a500-m radius of each site.Green space was further divided into turf grass and natural area.

To characterize foraging resources for pollinators,we recorded the identities of all blooming species within a5-m radius surround-ing focal plants.We also measured the density of all blooming species in0.25-m2plots surrounding focal plants.We did not record grasses or non-blooming forbs,as they were not available foraging resources for bees.

2.2.Bee sampling

We performed bee observations using two bee-pollinated native forbs:an early summer-blooming foxglove(Penstemon digitalis), and a late summer-blooming Asteraceae;either Echinacea or Rud-beckia spp.We planted two P.digitalis or native cone?ower (Echinacea purpurea)on the three green roofs lacking these?ower types,and used existing plants on the other roofs.

We observed each focal plant for three non-consecutive,sunny, warm days at each site.Observations were performed for15-min periods starting at9:00am,10:30am,12:00pm,and1:30pm. We recorded bees using size and color morphological criteria(e.g., small and dark).Bumble bees(Bombus spp.)and honey bees(Apis mellifera)were recorded to genus and species,respectively.

We employed two collection methods.At each site,we cap-tured bees that landed on observed?owers with insect nets for 15min at the end of one early-season(May–early July)and one late-season(late July–September)observation day.And we deployed bee bowls:3-oz pan traps coated with UV-re?ective paint and?lled with a water/dish-detergent solution(LeBuhn et al.,2003).We hap-hazardly placed15bowls,5each of3different colors(blue,white, yellow),1–5-m apart at each site.At sites with vegetation>1-m tall, we mounted bowls on a1.5-m dowel.We placed bowls at each site for approximately24h,once in the early and once in the late bloom seasons.

Bees were identi?ed to genus using Michener,McGinley,and Danforth(1994),and later veri?ed and identi?ed to species by J.S.Ascher and J.Gibbs(Lasioglossum).Specimens are housed at the Chicago Botanic Garden.Bees’nesting and foraging traits were obtained from Michener(2000)and Giles and Ascher(2006).

2.3.Statistics

Data were analyzed in R version2.12.0(R Development Core Team,2010)except as noted.

To assess relationships between land-cover and bee-community attributes,we used linear models and multiple regressions for each habitat type with proportion of land cover as the independent variable and bee abundance or species richness as the dependent variable.We used analysis of covariance(ANCOVA)to test whether relationships differed by habitat type.

Floral species richness and bloom density were analyzed using Wilcoxon rank-sum tests with continuity correction for non-normal variance.Correlations between landscape-scale factors and habitat types were calculated using linear regression.

104R.Tonietto et al./Landscape and Urban Planning103 (2011) 102–108 Generalized linear models(assigned a Poisson distribution to

account for zero-in?ated data)were used to test for effects of habi-

tat type on visitation rates.We used a priori contrasts to compare

visitation rates between habitat types.

Non-metric multidimensional scaling(NMDS)ordinations were

used to compare species composition of bees and plants across

habitat types using the vegan package in R(Oksanen et al.,2010).

Bee species only collected once during the study were removed

from the analysis.Data were relativized by samples and species

due to high coef?cients of variation and the NMS was constructed

using three axes.

We performed permutational multivariate analysis of vari-

ance(PERMANOVA)to test for differences among habitat types

in bee and plant–species composition and bee nesting habits

(soil-dwelling,cavity-dwelling,or cleptoparasitic)using the pro-gram PERMANOVA(Anderson,2005).Analyses were based on Bray–Curtis dissimilarity with1000permutations for each test and pair-wise comparisons between habitat types.

Having found signi?cant differences by PERMANOVA,we eval-uated which species were most responsible for differentiating communities using similarity percentage(SIMPER)analysis.SIM-PER evaluates the contributions of each species to the Bray-Curtis dissimilarity of all pairs of samples between groups(Clarke& Warwick,2001)and was implemented in PRIMER v6(Clarke& Gorley,2006).

We conducted a Mantel test to determine if there was a signif-icant relationship between bee and plant community composition by habitat type using the vegan package in R(Oksanen et al., 2010).

3.Results

https://www.wendangku.net/doc/118960670.html,ndscape-level characteristics

Overall,habitat type was a predictor for the proportion of green space in the surrounding landscape(p=0.002).There was more green space surrounding prairies than green roofs(p=0.003)or parks(p=0.06).There was also a greater proportion of green space around parks than green roofs(p=0.07).The proportion of green space categorized as natural area differed by habitat type(p=0.001) and was greater around prairies than parks(p=0.02)or green roofs (p=0.02),which were similar to each other(p=0.95).Although the proportion of green space composed of turf grass was not related to habitat type(p=0.15),it was greater surrounding parks than green roofs(p=0.07).There were no differences between prairies and parks(p=0.17)or green roofs(p=0.60).

Across habitat types,bee richness and abundance were posi-tively correlated with surrounding green space(species richness: p=0.0004,R=0.55;abundance:p=0.04,R=0.65).However,rela-tionships differed by habitat type:richness and green space were positively correlated for prairies(p=0.02,R=0.7)and the trend was positive,but not signi?cant for green roofs(p=0.12,R=0.45). There was a negative,but not signi?cant,trend for parks(p=0.13, R=0.46).The trend for parks was signi?cantly different from that of green roofs or prairies(p=0.009and0.002,respectively;Fig.1a). Bee abundance was not correlated with green space at prairies (p=0.12,R=0.45),parks(p=0.33,R=0.23)or green roofs(p=0.27, R=0.29).

For all habitats,there was a positive trend between bee species richness and proportion of natural area in the surrounding land-scape,which was signi?cant only at prairies(prairies:p=0.06, R=0.61,parks:p=0.23,R=0.32,green roofs:p=0.22,R=0.34; Fig.1b).Bee abundance was not correlated with natural area in the surrounding landscape for green roofs(p=0.10,R=0.54),parks (p=0.20,R=0.36)or prairies(p=0.18,R=0.38).

5

10

15

20

25

30

35

a

b

B

e

e

s

p

e

c

i

e

s

r

i

c

h

n

e

s

s

Proportion of green space (500-m radius) 0

5

10

15

20

25

30

35

B

e

e

s

p

e

c

i

e

s

r

i

c

h

n

e

s

s

Proportion of natural area (500-m radius)

Fig.1.Multiple regressions of(a)bee species richness against the proportion of green space in the surrounding landscape at green roofs(p=0.12,R=0.45),parks (p=0.13,R=0.46),and prairies(p=0.02,R=0.76),and(b)bee species richness against the proportion of natural area in the surrounding landscape at green roofs (p=0.22,R=0.34),parks(p=0.23,R=0.32),and prairies(p=0.06,R=0.61).

3.2.Site-level characteristics

Blooming-plant species richness differed by habitat type (p=0.002).Green roofs had fewer species in bloom than parks (p<0.001)or prairies(p<0.001).Richness of blooming plants was marginally greater at prairies than parks(p=0.09).Habitat type was a signi?cant predictor of blooming-plant species composition (Fig.2a,PERMANOVA:p=0.002).

3.3.Bee surveys

We performed114h of pollinator observations over59days from9June to6October,2008.We observed fewer bees visiting ?owers on green roofs(n=111)than in parks(n=668)or prairies (n=746)(p=0.024).Visitation rate at parks was slightly lower than at prairies(p=0.08)and was much lower at green roofs compared to parks(p=0.03)or prairies(p=0.0001).

Aggregating bee-bowl and net samples,we collected677bees belonging to5families,23genera,and63species;30%of species were represented by a single individual(see Appendix1).We col-lected more bees in prairies(n=329)than in parks(n=225)or green roofs(n=123).Bee species richness was signi?cantly differ-ent between habitat type(p=0.006).We collected marginally more bee species from prairies(n=46)than parks(n=30),but the differ-ence was not signi?cant(p=0.07).We collected signi?cantly more species from prairies than green roofs(n=19,p=0.005).There was

R.Tonietto et al./Landscape and Urban Planning103 (2011) 102–108

105

Fig.2.Bee species abundance curves at green roofs,parks,and prairies.

not a signi?cant difference between bee species richness at parks and green roofs(p=0.4).Green roofs had a greater proportion of bee species represented by a single individual than parks or prairies (Fig.3).

Bee species composition differed by habitat type(Fig.2b,PER-MANOVA:p=0.008).There was a signi?cant relationship between bee species and blooming-plant species composition by habitat type(Mantel:p=0.032,R=0.17).

Based on SIMPER analysis,four halictine bee species(Agapos-temon virescens,Halictus ligatus,Lasioglossum anomalum and H. ligatus)made particularly strong contributions to differentiating communities by habitat type(Table1).

Most bee species collected(87%)are native to Illinois.Prairies had the greatest proportion and number of native species(91%, n=42),followed by parks(83%,n=25),and green roofs(73%, n=14).Native species represented97%of individuals collected from prairies and94%from parks,but only74%from green roofs.

Nesting habitat distribution did not differ by habitat(PER-MANOVA:p=0.86).Over60%of bee species collected from each

L

o

g

(

r

e

l

a

t

i

v

e

a

b

u

n

d

a

n

c

e

)

Species rank

Fig.3.Nonmetric multidimensional scaling ordination of study sites according to (a)plant species composition(stress of9.31)and(b)bee species composition(stress of15.2),points represent sites,and ellipses represent habitat types.

habitat were ground nesters(green roofs:13spp.,n=200,parks: 20spp.n=270,prairies:29spp.n=90).Of those,more than half of species and individuals were soil-dwelling solitary bees.Almost 30%of individuals collected from green roofs were cavity nesters but these made up only8%of individuals collected from parks and 11%from prairies.Wood-and pith-nesting bees were collected in prairies and parks,but not green roofs.Soft-or rotting-wood nesting bees were found only in prairie sites.We collected only three individual cleptoparasitic bees,one from each habitat,each a different species.

Only3of the63bee species collected are oligolectic(pollen specialists on a particular plant species or group).All specimens of these three species(Colletes latitarsis[n=1],Melissodes desponsa [n=4],and Peponapis pruinosa[n=1])were collected from prairies. All other species collected are generalist foragers without a strict dietary preference for pollen.

Table1

Bee species most responsible for differentiating bee communities by habitat type (SIMPER analysis).

Bee species Average Abundance Contribution%a Prairies vs.parks

Lasioglossum anomalum 1.1711.512.42

Halictus ligatus7.17411.96

Melissodes bimaculata40.67 5.95

Lasioglossum mitchelli 2.67 1.5 5.70

Bombus impatiens21 5.69

Agapostemon virescens9.17 2.17 5.19

Lasioglossum pilosum08 4.73

Halictus confusus 1.330.67 3.86

Apis mellifera0.170.83 3.68

Ceratina calcarata or dupla 1.670.17 3.34

Prairies vs.green roofs

Megachile rotundata0212.16

Agapostemon virescens9.17411.01

Halictus ligatus7.170.6 6.57

Lasioglossum ellisiae0.832 6.40

Lasioglossum mitchelli 2.67 2.8 6.19

Melissodes bimaculata40.4 5.01

Anthidium oblongatum0.833 4.82

Lasioglossum illinoense1 2.8 4.81

Lasioglossum zephyrum0 2.6 4.58

Parks vs.green roofs

Megachile rotundata0.17211.95

Lasioglossum anomalum11.5 1.211.51

Halictus ligatus40.610.28

Agapostemon virescens 2.1749.48

Lasioglossum ellisiae0.52 6.43

Lasioglossum mitchelli 1.5 2.8 6.30

Bombus impatiens10.2 4.80

a Proportion of the bee community difference attributed to each species cumula-tively responsible for the?rst60%of dissimilarity between habitat types.

106R.Tonietto et al./Landscape and Urban Planning103 (2011) 102–108

Five of the species collected were new records for the state of Illinois,four of which were reported by Tonietto and Ascher(2009). We collected one specimen each of two rarely collected native bees: Coelioxys banksi(a cleptoparasite of Megachile spp.)from a prairie, and Lasioglossum michiganense,a presumed social parasite of euso-cial,metallic L.(Dialictus),at a park.Other new records for Illinois were two exotic cavity-nesting megachilids introduced from the Palearctic to the eastern United States and are now widely dis-tributed:Anthidium manicatum and Anthidium oblongatum.All four specimens of A.manicatum were collected from park sites,whereas A.oblongatum was collected from all habitats.Subsequent to pub-lication of Tonietto and Ascher(2009),we identi?ed one male of another new species for Illinois,Megachile inermis.This northern species,well known in Wisconsin,was found at our northernmost prairie site.

4.Discussion and conclusions

Native bees were present on green roofs in Chicago but were represented by fewer species and individuals than in prairies or park natural areas.In all habitat types,>70%of collected species were native to Illinois,and soil-dwelling species were most com-mon.Over30%of species were represented by a single individual, consistent with proportions of singletons found in other studies (Cane,2001).Composition of bees was signi?cantly different across habitat https://www.wendangku.net/doc/118960670.html,pared to parks and prairies,green roofs had fewer species of blooming plants and distinct plant community composition.At the landscape scale,bee richness was positively correlated with the proportion of natural area within a500-m radius.At the site scale,bee diversity increased with the diversity of blooming plants.

North American urban bee studies typically report the pro-portion of green space in the surrounding landscape without differentiating its composition,or report only local?oral compo-nents of sites(Fetridge et al.,2008;Frankie et al.,2005;reviewed by Hernandez et al.,2009).We agree with Hernandez et al.(2009)that urban bee studies should include landscape-scale analyses,with categories derived with bees’habitat and foraging needs in mind. Had we used undifferentiated green space as our land-cover cat-egory,we would have concluded that there was a negative trend between green space and bee diversity in parks.This was an arti-fact of most green space around park sites being turf grass,which is maintained to eliminate bare patches and weeds(including com-mon?owering lawn weeds such as dandelions and clover that would support bees),is frequently mowed,and has compacted soil. As a result,park turf grass at our sites has fewer foraging resources and its maintenance discourages soil nesting.

For many of the landscape and vegetation characteristics mea-sured,parks and prairies were similar,while green roofs differed. Park sites in this study included managed areas of native prairie plants.Such vegetation differs from that in most urban parks,which are typically managed for recreational or aesthetic considerations that do not favor native status(Gobster,2001;Loeb,2006).Wild bees have been shown to prefer native plant species over exotic ornamentals(Frankie et al.,2005;Hinners,2003),so these sites may draw more native bees than typical parks.These natural areas ful-?lled restoration goals(sensu Rosenzweig,2003),as anthropogenic habitats supporting wild bee communities similar in diversity and abundance to natural areas.

Our?ndings are consistent with others’that sites with greater plant diversity generally have greater bee diversity(Hendrix, Kwaiser,&Heard,2010;Potts,Vulliamy,&al,2003;Wojcik,2011). The green roof with the greatest number of bee species and indi-viduals collected was planted with native prairie species,and had the highest plant diversity of all green-roof sites.The two sites most similar in bee community composition,a park and a prairie, had the greatest diversity of blooming plants and bees within their respective habitat types.Planting of green roofs with diverse native species,rather than monotypes of Sedum,would increase their value for bee conservation.This approach would provide a greater variety of foraging resources to attract a wider vari-ety of native bees,and would also provide foraging resources for more of the blooming season.Ideally,green roofs in temperate regions should have diverse native forbs in bloom from May to October.Bees typically live in habitats where foraging resources are patchily distributed and spatially dissociated(Cane,2001),so planted rooftops need not be contiguous or have similar plant composition,but maintaining temporal contiguity of blooms is important(The Xerces Society,2011).

Oligolectic bees were only represented by three species in our study,all collected from prairies.Other urban bee studies have also found few?oral specialists,presumably due to a lack of appropriate host plants(Dalmazzo,2010;Hernandez et al.,2009).Oligolectic bees made up only2–9%of bees collected from urban and sub-urban gardens in the New York City region(Fetridge et al.,2008; Matteson et al.,2008),although a considerable diversity of oligolec-tic species occur in larger parks,botanical gardens,and cemeteries. The oligoleges we collected in prairies all specialize on plant species we never found on green roofs:Physalis pruinosa(ground cherry), the host of Colletes latitarsis;Cirsium spp.(thistle),the host of Melis-sodes desponsa;and Curcurbita spp.(squash),the host of Peponapis pruinosa.Frankie et al.(2005)attributed encountering very few pollen specialists in California gardens to a lack of host plants. If green roofs were intentionally planted with bee-friendly and specialist-attracting forbs,it is likely that more specialist bees would use these habitats.

In a recent review of urban bee ecology,one of the main patterns was an increase in abundance of cavity nesters over other nest-ing types(Hernandez et al.,2009;see also Dalmazzo,2010),and a study of bee diversity in New York City urban gardens recorded nearly twice as many cavity-nesting as soil-dwelling individuals (Matteson et al.,2008).It was therefore surprising that most of the bees collected in our study from parks and roofs were ground nesters and that the proportions of ground-nesting species and individuals were similar to those found in prairies.Over70%of bees collected from green roofs were soil dwellers,as were80% at parks,compared with66%at prairies.However,this may be an artifact of our collection methods,as certain ground-nesting taxa such as Lasioglossum(Dialictus)are more often trapped in bowls than cavity-nesters such as Megachile.

The presence of many soil-dwelling bees on green roofs is encouraging,as it shows that these sites are within their for-aging range(including height above ground level)or that bees utilize green-roof substrates for nesting.It is possible that ground-dwelling bees,especially larger species,have nests on the ground nearby and?y up to rooftops to forage.Although data on bee forag-ing distances are available,there is a lack of data on typical foraging heights(Greenleaf,Williams,Winfree,&Kremen,2007).Detailed observations of bees and nest excavations on green roofs are nec-essary to determine where soil-dwelling bees may be nesting.On typical green roofs used in this study,under a top layer of very small slate pebbles there are one to several layers of well-draining soils,many with a high proportion of sand,that may provide a suitable nesting substrate.Potential sites could also include small ?ower pots,as Lasioglossum(Dialictus)are well known to nest in these.In order to attract and support cavity-nesting bees,site man-agers or landscaping companies could install“bee condos”(wood blocks drilled with numerous small holes).In addition,paper/straw nests could be deployed for mason bees and dead branches could be added to provide habitat for small carpenter bees(Ceratina). Providing appropriate nesting substrates has already been shown

R.Tonietto et al./Landscape and Urban Planning103 (2011) 102–108107

to attract many bird species to green roofs(Fernandez-Canero& Gonzalez-Redondo,2010).

Of the15bee species that collectively contributed over60% of the bee-community differences among habitat types;four had their highest relative abundance on green roofs.The alfalfa leaf-cutter bee Megachile rotundata,was the greatest overall contributor to community differences across habitat types,and was the only species collected from one green roof site.This exotic species was also found to be common in California urban gardens(Frankie et al., 2009)and was collected widely in New York City(Matteson et al., 2008).Two additional exotic species of subgenus Eutricharaea are also locally numerous in urban North America.Other species that made up a large proportion of green-roof bee communities were Lasioglossum tegulare,Lasioglossum mitchelli,and Lasioglossum illi-noense,all of which are native,eusocial soil dwellers,and small even for species of subgenus https://www.wendangku.net/doc/118960670.html,sioglossum(Dialictus)fre-quently occur in urban bee surveys(Frankie et al.,2009;Matteson et al.,2008)and made up45%of individuals in a survey of two green roofs in Toronto,Ontario(Colla et al.,2009).Their relatively small size predicts short?ight ranges(Greenleaf et al.,2007)and small habitat ranges so they would seem less likely than larger species to visit green roofs from distant ground-level nests.For some Dial-ictus,green roofs may serve as self-contained“islands”of suitable habitat.

Green roofs are functioning as novel habitat islands in urban set-tings,providing resources for native bees and other volant fauna. Vertical green corridors using trellises to hold plants against the sides of buildings or green steps along the outside of buildings have been proposed(Fernandez-Canero&Gonzalez-Redondo,2010), and horizontal corridors already exist(e.g.,the High Line in New York City)that enable access by non-?ying animals.The addition of more green roofs,and especially those with diverse plant com-munities and ranges in height would help to support native bees, other insects,and birds(Fernandez-Canero&Gonzalez-Redondo, 2010).

Our study and that of Colla et al.(2009)demonstrate that green roofs provide suitable habitats for North American wild bees including foraging resources and possibly to,an unknown extent,nesting substrates.Although native-bee communities on Chicago green roofs had fewer species and signi?cantly lower abundance than ground-level urban natural areas,their overall bee-community composition did not signi?cantly differ.Green roofs are a potentially important offset to habitat loss.By providing habitat for native species and thus supporting urban biodiversity,green roofs function as novel habitats valuable for conservation from a reconciliation ecology perspective(Rosenzweig,2003).Urban green roofs are not suitable for all native bees,but a subset of them are supported by the foraging and(presumably)nesting resources green roofs provide.Relative to a surrounding,inhospitable matrix of buildings,streets,and traditional non-vegetated rooftops,green roofs are of high habitat value.In the face of land-use change and other anthropogenic disturbances,green roofs and other novel habitats are likely to play increasing roles in pollinator conserva-tion.

Acknowledgments

The Prairie Biotic Research Fund and Northwestern University funded this research through awards to R.K.T.Special thanks to C. Askham for help in the?eld and J.Gibbs,and S.Droege for bee identi?cation,especially of Lasioglossum(Dialictus).We thank the following groups for access to sites and permission to conduct research:Intrinsic Landscaping(green roofs),the City of Chicago Park District,and prairie-preserve land managers.J.S.A.thanks Robert G.Goelet for his generous support of bee research at the AMNH.Appendix1.Bee species abundance,nesting types,and foraging preferences collected from green roofs,parks,and prairies

Bee family

genus

Species Status a Nest b Green roofs Parks Prairies

Andrenidae

Pseudopa-

nurgus

albitarsis N S002 Apidae

Anthophora terminalis N SW001

Apis mellifera E H051

Bombus auricomus N H008

fervidus N H023

griseocollis N H019

impatiens N H1612 Ceratina calcarata or

dupla

N P0110

dupla N P007 Eucera hamata N S0023

Melissodes agilis N S020

agilis?N S018

agilis or

trinodis

N S003

bimaculata N S2424

comptoides N S001

denticulata N S010

desponsa N S004

trinodis N S002

trinodis?N S0110 Peponapis pruinosa N S001

Svastra obliqua N S002

Xylocopa virginica N W011 Colletidae

Colleies latitarsis N S002 Hylaeus

af?nis N C001

af?nis or

modestus

N C007

hyalinatus E C433

leptocephalus E C200

mesillae N C064 Halictidae

Agapostemon

virescens N S201355 Augochlora pura N Sw002 Augochlorella

aurata N S005

Halictus confusus N S148

ligatus N S32443

parallelus N S001 Lasioglossum(Dialictus)

unknown

N S or SW320

albipenne N S032

anomalum N S6653

cf callidum N S001

cinctipes N S001

coriaceum N S002

ellisiae N S1035

illinoense N S1406

leucozonium E S001

michiganense N CP010

mitchelli N S141320

pectorale N S240

perpunctatum N S003

pilosum N S0480

pruinosum N S020

zephyrum N S1310

zophops N S100 Sphecodes unknown N CP100 Megachilidae

Anthidium manicatum E C040

oblongatum E C1515 Coelioxys banksi N CP001

Hoplitis pilosifrons N C006

producta N C001 Megachilie

centuncularis E?C100

inermis N C001

latimanus N C010

108R.Tonietto et al./Landscape and Urban Planning103 (2011) 102–108 Bee family

genus

Species Status a Nest b Green roofs Parks Prairies

mendica N C010

montivaga N C005

rotundata E C1010

Osmia [Chenosmia]N C001 albiventris N C002

Total

bees

col-

lected

123225329

a Status in Illinois:N,native;E,exotic;E?,exotic status uncertain,treated as exotic for purposes of analyses.

b Nest location:S,soil;C,cavity;SW,soft wood;CP,cleptoparasitic. References

Anderson,M.J.(2005).PERMANOVA:a FORTRAN computer program for permuta-tional multivariate analysis of variance.Auckland,New Zealand:Department of Statistics,University of Auckland.

Cane,J.H.(2001).Habitat fragmentation and native bees:A premature verdict?

Conservation Ecology,5(1),art.no.-3.

Cane,J.H.,Minckley,R.L.,Kervin,L.J.,Roulston,T.H.,&Williams,N.M.(2006).Com-plex responses within a desert bee guild(Hymenoptera:Apiformes)to urban habitat fragmentation.Ecological Applications,16(2),632–644.

Clarke,K.,&Gorley,R.N.(2006).PRIMER,PRIMER-E.England:Plymouth.

Clarke,K.,&Warwick,R.(2001).Change in marine communities:an approach to statistical analysis and interpretation(2nd ed.).England:PRIMER-E,Plymouth. Colla,S.R.,Willis,E.,&Packer,L.(2009).Can green roofs provide habitat for urban bees(Hymenoptera:Apidae)?Cities and the Environment,2(1),12,article4. Daily,G.C.(1997).Countryside biogeography and the provision of ecosystem ser-vices.In P.Raven(Ed.),Nature and human society:the quest for a sustainable world (pp.104–113).Washington,DC:National Research Council,National Academy Press.

Daily,G.C.(2006).Countryside biogeography.In M.J.Groom,G.K.Meffe,&C.R.Car-roll(Eds.),Principles of conservation biology.Sunderland,MA:Sinauer Associates, Inc.

Dalmazzo,M.(2010).Diversidad y aspectos biológicos de abejas silvestres de un ambiente urbano y otro natural de la región central de Santa Fe,Argentina.

Revista de la Sociedad Entomológica Argentina,69(1–2),33–44.

Dixon,K.(2009).Pollination and restoration.Science,325(5940),571–573. Fernandez-Canero,R.,&Gonzalez-Redondo,P.(2010).Green roofs as habitat for birds:A review.Journal of Animal and Veterinary Advances,9(15),2041–2052. Fetridge,E.D.,Ascher,J.S.,&Langellotto,G.A.(2008).The bee fauna of residential gardens in a suburb of New York City(Hymenoptera:Apoidea).Annals of the Entomological Society of America,101(6),1067–1077.

Frankie,G.,Thorp,R.W.,Hernandez,J.,Rizzardi,M.,Ertter,B.,Pawelek,J.C.,et al.

(2009).Native bees are a rich natural resource in urban California gardens.

California Agriculture,63(3),113–120.

Frankie,G.W.,Thorp,R.W.,Schindler,M.,Hernandez,J.,Ertter,B.,&Rizzardi,M.

(2005).Ecological patterns of bees and their host ornamental?owers in two northern California cities.Journal of the Kansas Entomological Society,78(3), 227–246.

Giles,V.,&Ascher,J.S.(2006).A survey of the bees of the Black Rock Forest pre-serve,New York(Hymenoptera:Apoidea).Journal of Hymenoptera Research, 15(2),208–231.

Gobster,P.H.(2001).Visions of nature:Con?ict and compatibility in urban park https://www.wendangku.net/doc/118960670.html,ndscape Urban Planning,56,35–51.

Green Roofs for Healthy Cities.(2009).Green roof industry grows16.1per cent in 2009despite economic downturn.In S.W.Peck(Ed.),Green roofs for healthy cities.Toronto,Ontario,Canada:Green Roofs for Healthy Cities.

Greenleaf,S.S.,Williams,N.M.,Winfree,R.,&Kremen,C.(2007).Bee foraging ranges and their relationship to body size.Oecologia,153(3),589–596.

Grixti,J.C.,Wong,L.T.,Cameron,S.A.,&Favret,C.(2009).Decline of bumble bees(Bombus)in the North American Midwest.Biological Conservation,142(1), 75–84.

Hendrix,S.D.,Kwaiser,K.S.,&Heard,S.B.(2010).Bee communities(Hymenoptera: Apoidea)of small Iowa hill prairies are as diverse and rich as those of large prairie preserves.Biodiversity Conservation,19,1699–1709.

Hernandez,J.L.,Frankie,G.W.,&Thorp,R.(2009).Ecology of urban bees:A review of current knowledge and directions for future study.Cities and the Environment, 2(1),15,article3.

Hinners,S.J.(2003).Pollination in an urbanizing landscape:effects of habitat fragmen-tation on wild bee assemblages.Boulder,CO:University of Colorado.,p.6. Hisamatsu,M.,&Yamane,S.(2006).Faunal makeup of wild bees and their?ower utilization in a semi-urbanized area in central Japan.Entomological Science,9(2), 137–145.Illinois Department of Agriculture.(2009).Facts about Illinois agriculture.Spring?eld.

https://www.wendangku.net/doc/118960670.html,/about/agfacts/html

Kamin,B.(2010).Ten years of green roofs in Chicago:Mayor Daley’s green thumb and iron?st have produced impressive gains,but the movement remains in its infancy.

Chicago:Chicago Tribune.

Klein,A.M.,Steffan-Dewenter,I.,Buchori,D.,&Tscharntke,T.(2002).Effects of land-use intensity in tropical agroforestry systems on coffee?ower-visiting and trap-nesting bees and wasps.Conservation Biology,16(4),1003–1014. Kremen,C.,&Ricketts,T.(2000).Global perspectives on pollination disruptions.

Conservation Ecology,14(5),1226–1228.

Kremen,C.,Williams,N.M.,Bugg,R.L.,Fay,J.P.,&Thorp,R.W.(2004).The area requirements of an ecosystem service:Crop pollination by native bee commu-nities in California.Ecology Letters,7,1109–1119.

Larsen,T.H.,Williams,N.M.,&Kremen,C.(2005).Extinction order and altered-community structure rapidly disrupt ecosystem functioning.Ecology Letters, 8(5),538–547.

LeBuhn,G.,Griswold,T.,Minckley,R.,Droege,S.,Roulston,T.a.,Cane,J.,et al.(2003).

A standardized method for monitoring bee populations–The bee inventory(BI)

plot,https://www.wendangku.net/doc/118960670.html,/~beeplot/.

Loeb,R.E.(2006).A comparative?ora of large urban parks:Intraurban and interur-ban similarity in the megalopolis of the northeastern United States.Journal of the Torrey Botanical Society,133(4),601–625.

MacIvor,J.S.,&Lundholm,J.(2010).Insect species composition and diversity on intensive green roofs and adjacent level-ground habitats.Urban Ecosystems, Marlin,J.C.,&LaBerge,W.E.(2001).The native bee fauna of Carlinville,Illinois,revis-ited after75years:A case for persistence.Conservation Ecology,5(1),U91–U116. Matteson,K.C.,Ascher,J.S.,&Langellotto,G.A.(2008).Bee richness and abundance in New York city urban gardens.Annals of the Entomological Society of America, 101(1),140–150.

McFrederick,Q.S.,&LeBuhn,G.(2006).Are urban parks refuges for bumble bees Bombus spp.(Hymenoptera:Apidae)?Biological Conservation,129(3),372–382. Michener,C.D.(2000).The bees of the world.Baltimore:The Johns Hopkins University Press.

Michener,C.D.,McGinley,R.J.,&Danforth,B.N.(1994).In R.Shef?eld(Ed.),The bee genera of North and Central America(Hymenoptera:Apoidea).Washington,DC: The Smithsonian Institution.

Murray,T.E.,Kuhlmann,M.,&Potts,S.G.(2009).Conservation ecology of bees: Populations,species,and communities.Apidologie,40,211–236.

National Research Council.(2006).The status of pollinators in North America.Wash-ington,DC:National Academies of the Sciences.

Oksanen,J.F.,Blanchet,G.,Kindt,R.,Legendre,P.,O’Hara,R.B.,Simpson,G.L., et al.(2010).Vegan:community ecology package.R package version 1.17-4.

https://www.wendangku.net/doc/118960670.html,/package=vegan

Potts,S.G.,Vulliamy,B.,&al,A.D.e.(2003).Linking bees and?owers:How do?oral communities structure pollinator communities.Ecology,84(10),2628–2642.

R Development Core Team.(2010).R:a language and environment for statistical computing.Vienna,Austria.

Ricketts,T.H.,Regetz,J.,Steffan-Dewenter,I.,Cunningham,S.A.,Kremen,C.,Bog-danski,A.,et al.(2008).Landscape effects on crop pollination services:Are there general patterns?Ecology Letters,11(5),499–515.

Rosenzweig,M.L.(2003).Reconciliation ecology and the future of species diversity.

Oryx,37(2),194–205.

Shevory,K.(2010).The beekeeper next door.The New York Times[New York].[09 December2010.New York ed.,Section D1].

Steffan-Dewenter,I.,Potts,S.G.,&Packer,L.(2005).Pollinator diversity and crop pollination services are at risk.Trends in Ecology&Evolution,20(12),651–652. The Xerces Society.(2011).In D.Burns(Ed.),Attracting native pollinators:protecting north america’s bees and butter?ies.North Adams,MA:Storey Publishing. Tonietto,R.K.,&Ascher,J.S.(2009).Occurrence of the Old World bee species Hylaeus hyalinatus,Anthidium manicatum,Anthidium oblongatum,Megachile sculpturalis, and the native Coelioxys banksi,in Illinois(Hymenoptera:Apoidea:Colletidae, Megachilidae).The Great Lakes Entomologist,41(1&2),200–203.

Tuell,J.K.,Ascher,J.S.,&Issacs,R.(2009).Wild bees(Hymenoptera:Apoidea: Anthophila)of the Michigan highbush blueberry agroecosystem.Annals of the Entomological Society of America,102(2),275–287.

Wang,Y.,&Moskovits,D.K.(2001).Tracking fragmentation of natural communities and changes in land cover:Applications of landsat data for conservation in an urban landscape.Conservation Biology,15(4),835–843.

Williams,N.M.,&Kremen, C.(2007).Resource distributions among habitats determine solitary bee offspring production in a mosaic landscape.Ecological Applications,17,910–921.

Winfree,R.,Aguilar,R.,Vasquez, D.P.,LeBuhn,G.,&Aizen,M. A.(2009).A meta-analysis of bees’responses to anthropogenic disturbance.Ecology,90(8), 2068–2076.

Winfree,R.,Griswold,T.,&Kremen, C.(2007).Effect of human disturbance on bee communities in a forested ecosystem.Conservation Biology,21(1), 213–223.

Wojcik,V.(2011).Resource abundance and distribution drive bee visitation within developing tropical urban landscapes.Journal of Pollination Ecology,4(7),p.

48.56.

相关文档