Timberlake, T., Vaughan, I. P., & Memmott, J. (2019). Phenology offarmland floral resources reveals seasonal gaps in nectar availability forbumblebees. Journal of Applied Ecology, 56(7), 1585-1596.https://doi.org/10.1111/1365-2664.13403

Publisher's PDF, also known as Version of record

License (if available):CC BY

Link to published version (if available):10.1111/1365-2664.13403

Link to publication record in Explore Bristol ResearchPDF-document

This is the final published version of the article (version of record). It first appeared online via Wiley athttps://besjournals.onlinelibrary.wiley.com/doi/full/10.1111/1365-2664.13403 . Please refer to any applicableterms of use of the publisher.

University of Bristol - Explore Bristol ResearchGeneral rights

This document is made available in accordance with publisher policies. Please cite only the publishedversion using the reference above. Full terms of use are available: http://www.bristol.ac.uk/pure/user-guides/explore-bristol-research/ebr-terms/

J Appl Ecol. 2019;00:1–12.  | 1wileyonlinelibrary.com/journal/jpe

Received:30August2018  |  Accepted:12March2019DOI: 10.1111/1365-2664.13403

R E S E A R C H A R T I C L E

Phenology of farmland floral resources reveals seasonal gaps in nectar availability for bumblebees

Thomas P. Timberlake1  | Ian P. Vaughan2  | Jane Memmott1

ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycited.©2019TheAuthors.Journal of Applied EcologypublishedbyJohnWiley&SonsLtdonbehalfofBritishEcologicalSociety

1SchoolofBiologicalSciences,UniversityofBristol,Bristol,UK2CardiffSchoolofBiosciences,CardiffUniversity,Cardiff,UK

CorrespondenceThomasP.TimberlakeEmail:thomas.timberlake@bristol.ac.uk

Funding informationNaturalEnvironmentResearchCouncil,Grant/AwardNumber:NE/L002434/1

HandlingEditor:FabriceRequier

Abstract1. Floralresourcesareknowntobeimportantinregulatingwildpollinatorpopula-tionsandarethereforeanimportantcomponentofagri‐environmentandresto-ration schemeswhich aim to support pollinators and their associated services.However,thephenologyoffloralresourcesisoftenoverlookedintheseschemes—afactorwhichmaybelimitingtheirsuccess.

2. Ourstudycharacterizesandquantifiesthephenologyofnectarresourcesatthewhole‐farmscaleonreplicatefarmsinSouthwesternUKthroughoutthefloweringseason.Wequantifythecorrespondingnectardemandsofasubsetofcommonwild pollinators (bumblebees) to compare nectar supply and pollinator demand throughouttheyear,therebyidentifyingperiodsofsupply‐demanddeficit.

3. Werecordstrongseasonalfluctuationsinfarmlandnectarsupplies,withtwomainpeaksofnectarproduction(MayandJuly)andaconsiderable“JuneGap”inbe-tween.MarchandAugust/Septemberarealsoperiodsoflownectaravailability.

4. Comparingthephenologyofnectarsupplywiththephenologyofbumblebeenec-tardemandreveals“hungergaps”duringMarchandmuchofAugust/Septemberwhensupplyisunlikelytomeetdemand.

5. Permanentpastureandwoodlandproducedthegreatestshareoffarmlandnectarbecauseoftheirlargearea;however,linearfeaturessuchashedgerowsandfieldmarginsprovidedthegreatestnectarperunitarea.Fiftypercentoftotalnectarwassuppliedby justthreespecies (Allium ursinum, Cirsium arvense and Trifolium repens),butsome lessproductivespecies (e.g.Hedera helix and Taraxacum agg.) wereimportantinensuringphenologicalcontinuityofnectarsupply.

6. Synthesis and applications.Bycomparingthephenologyoffarmlandnectarsupplywiththephenologyofpollinatordemand,wedemonstratethatthetimingofnec-tarsupplymaybeasimportantastotalnectarproductioninlimitingfarmlandpol-linatorpopulations.Consideringphenologyinthedesignofagri‐environmentorrestorationschemesisthereforelikelytoimprovetheirsuitabilityforpollinators.Plant specieswhich flower during periods of resource deficit (early spring andlatesummer)shouldbeprioritizedinschemeswhichaimtoconserveorrestore

2  |    Journal of Applied Ecology TIMBERLAKE ET AL.

1  | INTRODUC TION

Theservicethatpollinatorsprovidetoamajorityoftheworld'swildflowering plants (Ollerton,Winfree, & Tarrant, 2011) and 75% ofworld crop species (Kleinet al., 2007)makes their conservation ahighpriority.Understandingthefactorsthatlimitpollinatorpopula-tionsonfarmlandiscriticalindesigningconservationschemesthatensuretheirlong‐termsurvival.Wildpollinatorpopulationsarelim-itedbyseveralfactors includingfloralnectarandpollenresources(Goulson,Nicholls,Botias,&Rotheray,2015;Potts,Vulliamy,Dafni,Ne'eman, & Willmer, 2003), nesting sites (Steffan‐Dewenter &Schiele,2008)andvariousotherfactorssuchasdisease,pesticidesandpredators (Goulsonetal.,2015;Roulston&Goodell,2011). IntheUK,nectar levels fellby32%between1930and1978, in linewith trends in pollinator diversity and agricultural intensificationsincetheSecondWorldWar(Baudeetal.,2016).Changesinthelast30years, likelyduetodecreasedacidification,decreasednitrogendepositionandtheuptakeofEnvironmentalStewardshipSchemes,have led tomodest increases innectarproduction.However,nec-tarproductionremainslowerthanpre‐1930slevelsandsignificantlossesinnectardiversityremain(Baudeetal.,2016).

The large‐scalecoverageofagricultural land intheUK(70.8%)(WorldBank, 2015), makes it an important consideration for anyprogramme aiming to conserve biodiversity at a national level. In theUK, Environmental Stewardship Schemes provide annual pay-mentstofarmersandlandmanagersformanagingtheir landinanenvironmentallyfriendlyway,includingforthebenefitofpollinators(NaturalEngland2009).Nectar‐richfieldmarginsareanimportantcomponentoftheseschemesandtherearedataonthebestfloralmixturesforsupportingfarmlandbumblebees,forexample(Carvell,Meek,Pywell,&Nowakowski,2004;Pywelletal.,2005).Itisknownthat the addition of floral resources can increase bumblebee col-onygrowthandnestdensity(Carvelletal.,2017;Crone&Williams,2016;Wood,Holland,Hughes,&Goulson,2015),andincreasespe-cies diversity and abundance of trap nesting bees (Dainese et al.,2018). However, the timing of resource availability (i.e. the phe-nology) is also important (Carvell et al., 2017;Williams,Regetz,&Kremen,2012),butthisaspectismuchlessunderstood.

Forpollinatorstopersistandthriveatthelandscapelevel,theymusthavesufficientfloralresourcesfortheentiredurationoftheirflight season (Menz et al., 2011; Russo, DeBarros, Yang, Shea, &

Mortensen,2013;Scheperetal.,2015).“Phenologicalgaps”ofjust15 days severely affect modelled honeybee colony development(Horn,Becher,Kennedy,Osborne,&Grimm,2016),afindingempir-icallysupportedbyRequier,Odoux,Henry,andBretagnolle(2017).Such gaps are likely to be evenmore detrimental to bee specieswhich do not have honey reserves. The importance of a season‐longsupplyoffloralresourceshassofarnotbeengivensufficientconsiderationinthedesignofEnvironmentalStewardshipschemes(Carvell,Meek,Pywell,Goulson,&Nowakowski,2007).Itissimilarlyoverlookedintherestorationofnaturalhabitatswhichrelyonpol-linatorstoensurethereproductivecontinuityoftherestoredplantcommunity(Dixon,2009).Theseoversightscouldultimatelybelim-itingthesuccessofbothtypesofscheme.

Identifying periods of the year inwhich floral resourcesmoststrongly limit pollinator populationswill be key to addressing thisissue inatargetedandcost‐effectiveway.Thisrequiresanunder-standing of both flowering phenology and pollinator floral needsata landscape‐scaleover theirentire flight season.Our studyad-dressestheseknowledgeneedsviathefollowingthreeobjectives:(a)characterizingandquantifyingthephenologyofnectarresourcesatthewhole‐farmscaleonreplicatefarmsthroughoutthefloweringseason (lateFebruary toearlyNovember); (b)quantifying thecor-responding nectar demands of common farmland bumblebees tocomparenectarsupplyandpollinatordemandthroughouttheyear,therebyidentifyingperiodswhenthereisasupply‐demanddeficit;(c) identifyinghabitatsandplantspecieswhichmayfill thesegapsand thereby provide sufficient resources for the entire pollinatorflight seasonon farmland.Ourmethodsprovideanovelapproachto plant‐pollinator phenological matching (Russo et al., 2013) andenable targeted planting strategies for the restoration of nectarsupplies on farmland, an approach that could be applied to otheranthropogenichabitats.

2  | MATERIAL S AND METHODS

2.1 | Study sites

The study was conducted in 2016 and 2017 on four medium‐sized (140–280ha)mixed (dairy, sheepandarable) farms inNorthSomerset,noneofwhichwereinEnvironmentalStewardship.Siteswere surrounded by mixed farmland and rural villages, typical of

pollinatorpopulations.Maintainingarangeofsemi‐naturalhabitatswithcomple-mentaryfloweringphenologies(e.g.woodland,hedgerowsandfieldmargins)willensureamorecontinuoussupplyofnectaronfarmland,therebysupportingpol-linatorsfortheirentireflightseason.

K E Y W O R D S

agri‐environment,bumblebees,floralresources,floweringphenology,nectar,pollination,pollinatorconservation,restoration

     |  3Journal of Applied EcologyTIMBERLAKE ET AL.

Southwest UK. The substantial time demands of recording floralabundance at a farm scale regularly from late February to earlyNovemberrestrictedfurtherreplication.Thereisatrade‐offinphe-nologystudiesbetweentheamountandresolutionofdatathatcanbegatheredatasiteandthenumberofsitesthatcanbesampled.Here,weadoptedadualapproachwherebyonesitewassampledintensivelytocapturethefine‐scaletemporalvariationinfloweringphenology and three other siteswere sampled less intensively tocapturethespatialvariation.

The intensive study site, Birches Farm in Somerset, England(51°25′19.04″N, 2°40′49.93″W)was sampled twice per week in2016 from lateFebruaryuntilearlyNovember,providing the in-tensivecomponentofthestudy.Thereweretwocomponentstotheextensivepartofourstudy.First,in2016,threefurtherfarmsin Somerset — Eastwood Farm (51°29′41.71″N, 2°60′56.74″W),ChestnutFarm(51°24′22.94″N,2°91′08.96″W)andElmtreeFarm(51°21′58.04″N, 2°85′44.36″W)—were sampled each fortnightfromMarchuntilNovemberin2016.Thefourfarmswere6–20kmfromeachotheranddifferedslightlyintheirhabitatcomposition,withvaryingproportionsofpastureandarablefields,hedgerows,margins andwoodland (see Table S1 and Figure S1). The nectarproductionandhabitatcompositionofallfourfarmswerebroadlyrepresentative of thewider landscape, based upon unpublisheddatafrom11farmsinSouthwestUK(seeAppendixS1andFigureS2).Thesefourfarmswereusedtocomparetheplantspeciesandhabitatcontributionstofarmlandnectarsupply.Second,in2017,threeofthefourfarms(Birches,EastwoodandElmtreefarms,re-ferredtohereafterasthephenologyfarms)weresampledeveryweek throughout the flowering season; thisprovidingbothphe-nologically informativedataandtemporal replicationforBirchesfarm.

2.2 | Objective 1: Characterizing and quantifying the phenology of nectar resources at the whole‐farm scale

2.2.1 | Nectar measurements

On each sampling occasion, six randomly located 50m transectswereconductedwithineachhabitattype(e.g.24transectsintotal,forafarmwithfourhabitattypes).Oneachtransect,thenumberofopenfloralunitsofeachfloweringplantspecieswasrecordedina1 m2quadratat5mintervalsalongitsentirelength(i.e.10quadratspertransect).Fortreesandshrubs,allflowersina5‐mverticalcol-umnabovethequadratwerecounted.Abovethis,thetree'sheightwithintheverticalcolumnwasestimatedwithaclinometerandthefloralabundancevaluesweremultipliedupaccordingly,asinBaudeet al. (2016).Values for the nectar sugar production of each spe-cieswerefromBaudeetal. (2016)whomeasuredormodelledthesugar(sucrose)productionof305plantspeciesintheUK,includingthe175mostcommonspecies.Thesugarproductionofeightspe-ciesencounteredinthestudybutnotcoveredbyBaudeetal.(2016)weremeasuredaccordingtotheirmethods(seeAppendixS2).

2.2.2 | Quantifying flowering phenology

Eachvisittoafarmgeneratedanestimateofthenumberofopenflowers per square metre in each habitat for that point in time.Whenmultipliedbythemeanfloralsugarproductionofeachspe-cies,anestimateofthegramsofsugarperunitareaper24‐hrpe-riodwasobtainedforeachhabitat.Thiswasmultipliedbytheareaofthathabitatonthefarm(calculatedusingQGISv.2.12.3)togiveanestimateofsugaravailabilityonthewholefarm.Ageneralizedadditivemodel(GAM),wasusedtomodelasmooth,nonlineartrendinsugaravailabilitybytime,withseparateanalysesperformedata farm andhabitat level.GAMsprovide a usefulwayof fitting asmoothcurvetodatawithnonlinearpatterns,thusallowinginter-polation between data points. To incorporate uncertainty associ-atedwithestimatesof individualspecies’nectarproduction,highandlowestimatesoffarmlandnectarprovisioningwerecalculatedusing upper (mean + SE)andlower(mean−SE)estimatesofeachspecies’sugarproduction.Thesethreeestimates(upper,lowerandmean)weremodelled separately.AGammaerror familywith loglinkfunctiongavethebestfitforthezero‐inflatedcountdata.Theextent of smoothing was varied between candidate models andguidedbyVaughanandOrmerod(2012)whoadvisevaluesaround0.3ofthenumberoftimepoints,asacompromisetocapturebothseason‐long trends and shorter term variation. Akaike's informa-tion criterion (AIC) was used to compare candidate models andselect thetop‐rankingone (with lowestAICvalue). Inadditiontomodellingsugarproductionatthewhole‐farmscaleandthehabitatlevel,the20mostcommonplantspeciesineachhabitatweremod-elledseparatelyusingtheapproachoutlinedabove.Thisallowedustocomparethesequenceofspeciesfloweringbetweenfarmsandbetweenyearsandidentifyparticularlyimportantspecies—bothin terms of total sugar production and phenological importance.Allstatisticalanalyses,figureplotting,andmodelswereperformedwithrversion3.2.2(RCoreTeam),usingthemgcvpackage(Wood,2011).

2.3 | Objective 2: Quantifying the nectar demands of three common bumblebee species to compare nectar supply and pollinator demand throughout the year

Toidentifyperiodsinwhichfarmlandnectarsuppliesarelikelytobelimitingpollinatorpopulations,wecomparedthetotalsugaravailabil-ityofBirches,EastwoodandElmtreefarms(usingtheGAMpredic-tions)withtheestimatedpopulation‐level,farm‐scalesugardemandsofthethreemostcommonbumblebeesonUKfarmland(Bombus ter-restris, Bombus pascuorum and Bombus lapidarius).Bumblebeeswerechosenasafocalgroupastheyweretheonlytaxonwithsufficientdataonenergyconsumption,colonydensityandphenologytomakethenecessarycalculations.Theyarealsoknowntobeimportantpol-linatorsofwildplants(Kovacs‐Hostyanszkietal.,2013)andarangeofcrops(Garrattetal.,2014)andyetare indeclineacrossvariouspartsoftheworld(Goulson,Lye,&Darvill,2008).

4  |    Journal of Applied Ecology TIMBERLAKE ET AL.

EnergydemanddatacamefromRotheray,Osborne,andGoulson(2017)who recorded thegramsof sugar consumedeachweekbycaptive B. terrestris audax colonies as they grew from singlewild‐collectedqueens to full colonies.Toaccount for theextraenergyexpendedduringforagingflight,0.312gofsugarwasaddedperindi-vidualforagingbeeperday(Rotherayetal.,2017),basedoncalorificcalculationsfromHeinrich(1979).WefollowedtheassumptionsofRotherayetal.(2017),thathalfoftheworkersforage4daysaweek,theothersremaininginthenestashousebees,andthatthequeenforagesuptothepointatwhichfiveworkersareproduced.SugarconsumptiondatawereonlyavailableforB.terrestris,butB. pascuo-rum and B. lapidarius were assumed to have similar consumptionratesbecausetheirbodysizes(Intertegularspan[mm]for:B. terres-tris [3.5];B. lapidarius and B. pascuorum [5.2]) (Greenleaf,Williams,Winfree,&Kremen,2007),andtotalcolonysizes(400individualsforB.terrestris and B. lapidariusand300forB. pascuorum)(Dicksetal.,2015) are broadly similar.

ColonydensitiesweretakenfromDicksetal.(2015)whosum-marize (froma rangeof studies) thenestdensityestimatesof thethreemostcommonBombus species on agricultural land: B. terrestris (meannestdensity:32/km2),B. pascuorum(83/km2) and B. lapidarius (78/km2).WorkernumberspercolonyandtheirchangesthroughtheyearweretakenfromRotherayetal.(2017).

Toestimatethetimingofcolonyfoundationinourstudyarea,weusedBeeWalktransectdata(BumblebeeConservationTrust2016&2017)from31recordingsitesinNorthSomerset.TheproportionsofB. terrestris, B. pascuorum and B. lapidariusqueensemergingindif-ferentmonthsoftheyearwerecalculated,allowingustomatchthetimingofcolonydevelopmentandnectardemandwiththetimingsoffarmlandnectaravailability.

2.4 | Objective 3: Identifying habitats and plant species which fill the gaps in nectar production

TherelativeimportanceofdifferentfarmlandhabitatswasassessedbycomparingtheGAMpredictionsforeachhabitatonthefourfarmsrecordedin2016.Thephenologicalimportanceofeachplantspeciesineachhabitatwascalculatedbysummingtheproportionalcontribu-tiontototalweeklysugarsupplymadebythespecies,foreachweekoftheyear.Themetriccapturesboththetemporaluniquenessofaspecies’nectarsupplyanditslengthoffloweringtime.

3  | RESULTS

3.1 | Objective 1: Characterizing and quantifying the phenology of nectar resources at the whole‐farm scale

During137visitstothefourfarmsover2years,nearlyhalfamillion(494,291)individualfloralunitsfrom176floweringplantspecieswerecounted in 2,664 transects (761 hedgerow transects, 759 pasture,576woodlandand568margins).ThedailysugarproductionofeightnewspecieswasrecordedandaddedtothenectardatabaseofBaudeetal.(2016)(seeTableS2).Thetop‐rankinggeneralizedadditivemodel(seeTableS3)describedanonlineartrendinsugaravailabilitywhichfluctuatedgreatlythroughtheyear,creatingthesixfloweringperiodshighlightedinFigure1.Althoughtotalyearlysugarproductionperkil-ometresquaredvarieduptothreefoldbetweenfarmsin2017(342kgofsugarkm−2 year−1onBirchesFarm,461onEastwoodFarmand131onElmtreeFarm),thephenologicalpatternofsugarproductionwasrelativelyconsistentamongthefarms(Figure2).

F I G U R E 1  DailysugarproductionofBirchesFarm(squares),EastwoodFarm(circles)andElmtreeFarm(triangles)duringindividualvisitsoveranentirefloweringseasonin2017.DataaresmoothedwithaGeneralizedAdditiveModel.Thecurvebaseduponthemeansugarproductionofeachplantspecies(±SE;dashedlines)isshowninblack,whilethecurvesbaseduponlowandhighestimatesofeachspecies’sugarproductionareshowninred(±SE;dottedlines).Theyearisdividedintowhatisvisuallyperceivedasthemainfloweringseasons,withpinkrepresentingtroughsandpurplerepresentingpeaks

     |  5Journal of Applied EcologyTIMBERLAKE ET AL.

3.2 | Objective 2: Quantifying the nectar demands of a subset of common wild pollinators to compare nectar supply and pollinator demand throughout the year

Thestrongseasonalityofnectarsupplydidnotsynchronizewellwiththesugardemandofcommonbumblebeespecies(Figure3).Oneach

ofthethreephenologyfarms,thepollinatorflightseasonwaschar-acterizedbyalternatingperiodsofnectardeficitandsurpluswhichwere relatively consistent in their timings, though differed some-what in themagnitudeof theirpeaksand troughs. InearlyMarchwhen queens emerge, sugar demand per individual bee was highwhilefarmlandnectarproductionwasatitslowestfortheflowering

F I G U R E 2  Nectarphenologyprofilesof(a)BirchesFarm2016,(b)BirchesFarm2017,(c)EastwoodFarm2017and(d)ElmtreeFarm2017.Resultsaretakenfromsummingtheoutputsofindividualspeciesmodelsforeachfarm.Reddottedlinesshowmediandailysugarproductionfortheyear.Peaksofnectarproduction(>median)aremarkedinpurple,whiletroughsorgaps(<median)areshowninpink.Notethedifferentscaleforeachgraph.TheJuneGaponBirchesFarm2016(plotA)isevidentfromthecurvebutdoesnotregisterasaformaltroughasitdoesnotcrossthemedianline

0

2,000

4,000

6,000

8,000

10,000

March April May June July Aug Sept Oct Nov

0

5,000

10,000

15,000

March April May June July Aug Sept Oct Nov

0

500

1,000

1,500

2,000

2,500

March April May June July Aug Sept Oct Nov

0

2,000

4,000

6,000

8,000

10,000

Sug

ar/d

ay/k

m2

(g)

Sug

ar/d

ay/k

m2

(g)

Sug

ar/d

ay/k

m2

(g)

Sug

ar/d

ay/k

m2

(g)

March April May June July Aug Sept Oct Nov

(a)

(b)

(c)

(d)

6  |    Journal of Applied Ecology TIMBERLAKE ET AL.

season.Thisleftameandeficitof12.3gofsugarkm−2 day−1(±1.7SE) betweenwhatwasavailableandourestimatesofbumblebeeneeds.This“hungergap”lastedfromthestartofthepollinatorflightseasonuntillateMarch.Duringthistime,theonlyspeciesproducingecolog-icallymeaningfulquantitiesofsugaronthefarmswereTaraxacum agg., Prunus spinosa, Glechoma hederacea, Ranunculus ficaria and Bellis perennis.Together,thesespeciescontributedameanof13.1gofsugarkm−2 day−1 (±6.8SE)during thehungergap. Justone for-agingqueenrequiresanestimated0.7gofsugarperday,meaningthatformostofMarch,amaximumof19queenbumblebeescouldbesupportedon1km2offarmland.Thisdoesnotaccountforanyyoungworkers that have been produced, or other pollinator spe-ciescompetingfornectarsuchasearlyspeciesofsolitarybeesorhoverflies.

In late summer (August‐October), the three study farmshad ameandeficitof1,053gof sugarkm−2 day−1 (±81.4SE) lasting be-tween1 and3months (Figure3).Although sugarproductionwasrelatively high at this time, Bombus colonies were reaching theirmaximum size, generating a high demand for nectar which could

notbemetbythefarmlandlandscape,resultinginasecondhungergap.AverysmallproportionofthefarmlandsugarwasproducedbyplantsspeciesunlikelytobeutilizedbyBombus (e.g. Stellaria media) implyingnectaravailabilitymaybeevenlowerthanpredicted.

From lateMarchuntilmid‐lateMay, therewasameansurplusof2,196gofsugarkm−2 day−1(±986SE)onthethreestudyfarms.Mass‐floweringoil seedrapewasnotpresentonanyof thestudyfarmsbutnormallyflowersduringthisperiodandwouldthereforebe expected to add to the nectar surplus recorded on our farmsratherthanfillahungergap.

3.3 | Objective 3: Identifying habitats and plant species which fill the gaps in nectar production

Habitatsdifferedgreatlyintheirsugarproductionvalueatafarmscalebuttheirrelativevaluesamongfarmsweresimilar(Figure4).Hedges produced the greatest sugar per unit area (1.88 g ofsugar m−2 year−1;±0.24SE)andwithameancoverageof1%offarmarea,theymadeup9.4%(±3SE)oftotalsugar.Theirphenological

F I G U R E 3  Comparisonbetweendailynectarsupplyanddailydemandofthreecommonbumblebeespeciespresenton1km2offarmlandon:(a)BirchesFarm2016,(b)BirchesFarm2017,(c)EastwoodFarm2017and(d)ElmtreeFarm2017.Blacklinesshowgramsofsugaravailableeachdayon1km2farmland,dividedbythenumberofcommonbumblebeespresentonthelandscapeatthattimethatis,sugaravailableperindividualbee(±SE).TheredlineshowstheestimatedmeandailysugarrequirementofaBombus terrestris individual at eachpointintheyear(±SE),fromRotherayetal.(2017).Notethatenergydemandperindividualishighestinearlyspringwhenqueensareforagingandestablishingcolonies.Shadedregionshighlightperiodsofnectardeficitwheredemand(redline)exceedssupply(blackline).Notethey-axis is plotted on a log10 scale

(a) (b)

(c) (d)

     |  7Journal of Applied EcologyTIMBERLAKE ET AL.

continuitywasalsohighest,beingthemostnectar‐richhabitatperunitarea62% (±3SE)of theyear.Fieldmarginswerealsoa richhabitatfornectar,withameanof1.68gofsugarm−2 year−1(±0.09SE). However, their period of nectar production was relativelyshort‐lived (seeFigureS3).Witha coverageof1%of farmarea,theymadeup3.1%(±4SE)oftotalsugarproduction.Thenectarproductionofpasturewassubstantial(54%oftotalsugarproduc-tion,±12SE)becauseofitslargeareaonthefarm(mean64%cov-erage),butperunitareaitproducedonly0.27gofsugarm−2 year−1

(±0.12SE).Wherewoodlandwaspresent,itcoveredanaverageof8%ofthefarm,producing1.08gofsugarm−2 year−1(±0.06SE) and makingup33.1% (±12SE)of total farmnectar supply.However,approximately90%of this supplywasproduced in just1month(May) and it was almost exclusively provided by Allium ursinum (89%).Figure5showsthesugarcontributionsofthemostproduc-tiveplantspeciesineachofthefourhabitats.

Althoughupto59plantspeciesproducedecologicallymean-ingful quantities of sugar at some point in the year (>0.3 g of

F I G U R E 4  Totalyearlynectarproductionofthefourmainhabitattypespresenton(a)atypical1km2areaoffarmland(includingvaluesfromfarmswherethathabitatwasnotpresent)and(b)asquaremetreofthegivenhabitat.Valuesforeachhabitatareexpressedasameanofthefourstudyfarms(Birches,Eastwood,ElmtreeandChestnut)±SE

(a) (b)

F I G U R E 5  Nectarcontributionsofthemostproductiveplantspeciesin(a)fieldmargins,(b)hedgerows,(c)pastureand(d)woodland.Valuesshownareameanofthefourstudyfarms(Birches,Chestnut,EastwoodandElmtree)

Rubusfru�cosus

25%

Heracleumsphondylium

15%Cirsiumarvense

15%

Prunusspinosa

11%

Anthriscussylvestris

7%

Salix sp6%

Crataegusmonogyna

5%

Other16%

Allium ursinum89%

Rubus fru�cosus3%

Anthriscus sylvestris2%

Ilex aquifolium 2% Other

Heracleumsphondylium

47%

Trifoliumrepens

13%

Anthriscussylvestris

12%

Taraxacumofficinale

8%

Cirsium arvense3%

Convolvulus arvensis2%

Calystegia sepium2% Other

13%

Trifolium repens26%

Cirsium arvense19%

Trifoliumpratense

17%

Ranunculusacris12%

Taraxacumofficinale

6%

Centaurea nigra5%

Heracleum sphondylium4% Other

11%

(a) (b)

(c) (d)

8  |    Journal of Applied Ecology TIMBERLAKE ET AL.

F I G U R E 6  Plantspecies’contributionstototalfarmlandnectarsupplyon(a)BirchesFarm,(b)EastwoodFarm,(c)ElmtreeFarmand(d)ChestnutFarmin2016.Linesshowthecumulativecontributionofeachspecies.Onlythe20mostproductivespeciesoneachfarmareshown

(a) (b)

(c) (d)

Cum

ulat

ive

nect

ar c

ontib

utio

n (%

)C

umul

ativ

e ne

ctar

con

tibut

ion

(%)

TA B L E 1  The10mostphenologicallyimportantspeciesonBirchesfarmin2016,rankedinorderofdecreasingscore.Thephenologicalimportancemetricgivestheproportionalcontributiontototalweeklynectarsupplymadebythespecies,summedacrosseachweekoftheyear.Highscoringspeciesarethosewhichflowerattimeswhenlittleelseisinbloom,contributingaveryhighproportionoftotalnectar.Theirdateofpeakfloweringisshown,alongsidethedateatwhichtheyaremakingthegreatestproportionalcontributiontototalnectarsupply,thatis,thepointatwhichtheirprovisioningismostimportant

Species Phenological importance metric Peak flowering date Peak phenological importance

Hedera helix 8.0 05 October 02 November

Taraxacum officinale 6.6 04May 06 April

Cirsium arvense 3.6 13July 13July

Allium ursinum 3.3 11May 11May

Rubus fruticosus 2.5 20July 17August

Heracleum sphondylium 2.5 15June 08June

Trifolium repens 2.5 13July 03 August

Bellis perennis 1.1 11May 02March

Glechoma hederacea 1.1 18May 02March

Centaurea nigra 1.0 29June 29June

     |  9Journal of Applied EcologyTIMBERLAKE ET AL.

sugarkm−2 day−1),50%of total sugarwas suppliedby just threespecies and 80% of the sugar was supplied by eight species(Figure 6). These were:A. ursinum (18%), Cirsium arvense (16%),Trifolium repens (14%),Trifolium pratense (12%),Heracleum sphon-dylium (6%),Ranunculus acris (5%),Rubus fruticosus agg. (5%)andTaraxacumagg.(4%).Severallessproductivespeciesmadeimport-antcontributionstothephenologicalcontinuityofnectarsupply,due to their unusual flowering times (Table1).Hedera helix pro-videdoverhalfofall sugar frommid‐Septemberuntil theendofthefloweringseason,whileTaraxacumagg.providedthemajorityofsugarfrommid‐MarchuntiltheendofApril.

4  | DISCUSSION

OurstudyquantifiesthefloweringphenologyoffourUKfarmsatahightemporalresolutionthroughoutthefloweringseason.There-sultsshowstrongseasonalfluctuationsinfarmlandnectarsuppliesandsuggestthephenologyofnectarsupplycouldbeasimportantastotalnectarproduction in limitingfarmlandpollinatorpopulations,thoughthisremainstobetested.Comparingnectarsupplywiththeenergy demands of a subset of common Bombus species reveals gapsbetweennectar supplyanddemandduringMarchandmuchofAugustandSeptember.Habitatsonthefarmsdifferedgreatlyintheirpatternofnectarproductionbuttendedtocomplementeachother'snectarsupply.Permanentpastureandwoodlandproducedthegreatest shareof farmlandnectarbecauseof their large area;however, linearfeaturessuchashedgerowsandfieldmarginspro-videdthegreatestnectarperunitarea,reflectingfindingsbyBaudeetal.(2016)intheirUK‐wideanalysis.Mostofthefarmlandnectarwassuppliedbyasmallnumberofplantspecies,butsomelesspro-ductivespecieswereimportantinensuringphenologicalcontinuityofnectarsupply.

4.1 | Limitations

Therewerethreemainlimitationstoourwork.First,thepracticaland time constraints of recording flowering phenology at a highresolutioninmultiplelocationsmeantthatourstudywaslimitedtofourfarmsacrossoneregionoftheUK.Whilethepatternofnec-tar supplywas relatively consistentacross these four farms, thispatternwill differ according to geography, inter‐annual variationandagriculturalpractices.Forexample, farmswithmanyearlier‐flowering tree species or late‐flowering haymeadows, are likelytohaveadifferentphenologicalpatternofnectarproduction.Thephenomenonofnectargapshowever, is likely tobea featureofmanyhuman‐alteredlandscapes,particularlythosethathavebeenheavily simplified. Second, while we model Bombus nectar de-mandsoneachfarm,alackofdatameansthatwecannotincludethedemandsofthemanysolitarybees,honeybees,hoverfliesetc.It is thereforea conservativeestimateofdemandand shouldbeviewedasaminimumbaselinerequirementforbumblebeesalone,rather than an ideal level.However, this approach still allowsus

toidentifythemostseverenectargapswhicharelikelytoaffectall pollinatorgroups.And finally,whilewehavedetaileddataonnectar,wedid not quantify pollen.Althoughboth are importantresources,wefocusonnectarbecauseofitsimportanceasanen-ergysourceinthedietsofadultbeesandotherpollinatorgroups.It also allowsus todirectly compare thenutritional contributionofallplantspeciesandhabitatsthroughthecommoncurrencyoftotalsugars(Willmer,2011).Itispossiblehoweverthatpollenre-sources (which are known to limit brood production and colonysizeofhoneybees,Requieretal.,2017andbumblebees,Rotherayetal.,2017),maydifferfromnectarresourcesintheirphenology,resultinginadifferenttimingofresourcegaps.Thisisanimportanttopicforfutureresearch.

4.2 | Flowering and pollinator phenology

Thehighlyseasonalnectarsupplydetected inourstudyonfarm-landinSouthWestUKislikelytohaveimportantimplicationsforwild andmanagedpollinators.The largedifferencesbetween theflowering phenology of different habitats (see Figure S3), sug-gestthatpollinatorsneedtomovebetweenhabitats,trackingthechangingresourcesupplies,toensureacontinuoussupplyofnectar.ThiseffecthasbeendemonstratedinagriculturalareasoftheU.S.wherecomplementaryhabitatsprovideresourcesatdifferenttimesof the year and the pollinator community tracks these resources(Mandelik,Winfree,Neeson,&Kremen,2012).Thishighlightstheimportanceofhavingarangeofdistincthabitattypespresentonfarmland.

Variousstudieshave identifieda fooddeficit forhoneybees inJune/July(Couvillon,Schurch,&Ratnieks,2014;Requieretal.,2015)whichcoincideswiththeperiodbetweenthespringfloralresources(includingmass‐floweringoilseedrapewhichisknowntobeimport-antforwildpollinators(Westphal,Steffan‐Dewenter,&Tscharntke,2003))andsummerfloralresources.Thisperiodoftheyearhasbeenanecdotallynamedthe“JuneGap”bybeekeepers.Withthelargedipinnectarresourcesrecordedbetweenthespring(May)andsummer(July)wildflowerbloomsandthemodestgapbetweennectarsupplyandbumblebeedemandrecordedinJune,ourstudyprovidesstrongempiricalevidencefortheexistenceofthe“JuneGap”onfarmlandinthisregion.

Theearlyspringseason(lateFebruarytolateMarch)isaperiodofverylownectaravailability.Thiscoincideswithaperiodofhighenergy demand by queen bumblebeeswhich are foraging, estab-lishingnestsandheatingtheirbrood(Heinrich,1972),resultinginanectardeficitformostofMarch.Thismodestgapcouldbehavingamarkedeffectonthesurvivalofqueens—aneffectwhichislikelytocascadethroughtheyearbylimitingthenumberofcolonieses-tablished.Indeed,ourdatahelpexplainthefindingbyCarvelletal.(2017)thatavailabilityofearlyspringresourcesonfarmlandstronglyinfluencesbumblebeecolonydensities.EarlyBombus colonies grow very littleunderfood limitation (Rotherayetal.,2017),suggestingtheeffectsofthisgapmayextendbeyondcolonyestablishment,af-fectingcolonysizetoo.

10  |    Journal of Applied Ecology TIMBERLAKE ET AL.

Compared with the early spring gap, the late‐season gap isgreater inmagnitudeandlasts longer (one‐threemonths),which islikelytothreatenthesurvivalof late‐emergingbumblebeespeciesonfarmland.This isconsistentwithBalfour,Ollerton,Castellanos,and Ratnieks (2018) who found significantly greater numbers ofextinctions in late‐summer flying British pollinator species, andFitzpatrick et al. (2007), who found a disproportionate decline inlate‐emergingbumblebeespeciesinIrelandandBritain.Theyattri-butethesedeclinestoareductioninlate‐summerfloralresources,partiallydrivenbytheshiftinagriculturalpracticesfromhaytosi-lage production. Other wild pollinators such as solitary bees andhoverflies have shorter flight seasons, somay not be affected byallthesameresourcegaps.However,thepopulationsofmostpolli-natorspeciespeakinlatesummer(Balfouretal.,2018),suggestingthismaybeaperiodofnectardeficitformanydifferentpollinatortaxa.Hornetal.(2016)demonstratedthatbadlytimedgapsinnec-tarsuppliescangreatlyaffecttheresilienceofmodelledhoneybeecolonies;bumblebees,whichdonotaccumulatesignificantresourcereserves,arelikelytobemorestronglyaffectedbysuchgaps.Morevulnerablestillwillbespecieswithshort flightseasons (e.g.manysolitarybees)whoseemergencetimescoincidewithanectardefi-cit.Resourcegapsdifferedslightlybetweensamplingyears,withanorderofmagnitudegreaterspringnectardeficitin2017than2016onBirchesFarm(Figure3a,b), likelyduetothewarmerspringandearlieremergencetimesofqueenbumblebeesin2017(BumblebeeConservation Trust 2016 & 2017). Variation in resource gaps be-tweensites(Figure3b–d)waslikelyduetodifferenthabitatcompo-sitionandmanagementofthefarms,particularlypasture,themostvariablehabitat(Figure4a),whichislikelytoofferthegreatestpo-tential for improvement. The effects of inter‐annual variation andlandscapecompositiononnectarphenologyareimportanttopicsforfuturestudy.

Withclimatechangeadvancingthefloweringphenologyofmanyplantspecies(e.g.FitterandFitter(2002)),andthepotentialforre-sulting phenological mismatches between plants and pollinators(Forrest,2015;Hegland,Nielsen,Lazaro,Bjerknes,&Totland,2009),itwillbecomeincreasinglyimportanttounderstandhowthetimingof resource suppliesaffectspollinatorpopulations.Byquantifyingthecurrentphenologyofnectar resources,wecanmakemore in-formed predictions about how this resource supplymight changeandwhichspeciesaremostlikelytobeaffected.

4.3 | Management implications

Wehavedemonstratedthat itmaynotbe just theavailabilityofnectar resources limiting Bombus populations, but also the tim-ingof these resources, though this remains tobe tested.MarchandAugust/Septemberareperiodsofgreatestnectardeficit forBombuspopulationsandshouldthereforebeprioritizedtoensureasufficientannualnectarsupply.Plantspecieswhichflowerduringtheseperiodsofdeficit—so‐called“bridgingspecies”(Menzetal.,2011)—shouldbeprioritizedinschemeswhichaimtoconserveorrestorepollinatorpopulationsonfarmland.Theearlyhungergap

weobservedon the four farms could theoretically be “plugged”byaddingjust12.3extragramsofsugareachdayacross1km2offarmland, the equivalent of c. 1,000willow catkins for example(datafromBaudeetal.,2016).WillowsSalix spp. could be readily addedtoUKfarmingsystems,deliveringpollenandnectarintheearlyspringwhenfloralresourcesareparticularlyscarce(Moquet,Mayer,Michez,Wathelet,& Jacquemart, 2015).The late‐seasongaphoweverwouldrequirebetween500and2,000extragramsofsugarperday,whichequatestoapproximatelyonehectareoflate‐flowering red cloverT. pratense (Rundlof, Persson, Smith, &Bommarco,2014),oranextra40brambleR. fruticosusagg.flowerspermetresquaredofhedgerow(basedonameanfarmcoverageof1%hedgerowarea).

Onallfourstudyfarms,halfofthetotalnectarsupplywaspro-videdbythreespeciesorfewer,afindinginaccordwithBaudeetal.(2016)intheirUK‐wideanalysis.Withjustafewplantspeciesdom-inatingfarmlandnectarsupplyformostoftheyear,thereisthepo-tentialforthesespeciestodominatethedietsofpollinators,reducingtheirdietdiversity.Theimmunocompetenceofhoneybeeshasbeenshowntoreducewithalessvarieddiet(Alaux,Ducloz,Crauser,&LeConte,2010;DiPasqualeetal.,2013)anditislikelythatthesameistrueforbumblebees.Resourcediversityshouldthereforebeconsid-eredalongsidetotalresourceavailabilityinthedesignofanyschemesaimingtorestoreorconservehealthypollinatorcommunities.

5  | CONCLUSIONS

Wild pollinator populations are known to be limited by floral re-sources and we have demonstrated why the timing of these re-sources may be an important factor driving this limitation. Thetemporalmismatchbetweenpollinator resourcedemandandphe-nologyoffarmlandresourcesupplydetectedinthisstudy, is likelytobeafeatureofmanyotherhuman‐alteredlandscapes;thoughthisremainstobetested.Ourresultssuggestthat inanyagri‐environ-mentorrestorationschemewhichaimstosupportpollinatorsandtheprovisioningofpollinationservices,consideringthephenologyofbothplantsandpollinatorswillbecritical.

ACKNOWLEDG EMENTS

This work was supported by the Natural Environment ResearchCouncil through the NERC GW4+ Doctoral Training Partnership(NE/L002434/1).Wewouldalsoliketothankthefollowingfieldas-sistants:EmmaBall,IzzyCarpenterandRowanHookham,alongwiththefourfarmerswhogavepermissiontousetheirland.

AUTHORS’ CONTRIBUTIONS

J.M.andT.P.T.conceivedtheideasanddesignedthemethodology;T.P.T.collectedthedata;T.P.T.andI.P.V.analysedthedata;T.P.T.andJ.M.ledthewritingofthemanuscript.Allauthorscontributedcriti-callytothedraftsandgavefinalapprovalforpublication.

     |  11Journal of Applied EcologyTIMBERLAKE ET AL.

DATA ACCE SSIBILIT Y

Data available via the Dryad Digital Repository https://doi.org/10.5061/dryad.3nk236h (Timberlake, Vaughan, & Memmott,2019).

ORCID

Thomas P. Timberlake https://orcid.org/0000‐0001‐8166‐0825

Ian P. Vaughan https://orcid.org/0000‐0002‐7263‐3822

R E FE R E N C E S

Alaux,C.,Ducloz,F.,Crauser,D.,&LeConte,Y.(2010).Dieteffectsonhoneybeeimmunocompetence.Biology Letters,6,562–565.https://doi.org/10.1098/rsbl.2009.0986

Balfour,N.J.,Ollerton,J.,Castellanos,M.C.,&Ratnieks,F.L.W.(2018).British phenological records indicate high diversity and extinctionratesamong late‐summer‐flyingpollinators.Biological Conservation,222,278–283.https://doi.org/10.1016/j.biocon.2018.04.028

Baude,M.,Kunin,W.E.,Boatman,N.D.,Conyers,S.,Davies,N.,Gillespie,M.A.,…Memmott, J. (2016).Historicalnectarassessment revealsthe fall and rise of floral resources inBritain.Nature,530, 85–88.https://doi.org/10.1038/nature16532

Carvell,C.,Bourke,A.F.G.,Dreier,S.,Freeman,S.N.,Hulmes,S.,Jordan,W.C.,…Heard,M.S. (2017).Bumblebee family lineagesurvival isenhancedinhigh‐qualitylandscapes.Nature,543,547–549.https://doi.org/10.1038/nature21709

Carvell,C.,Meek,W.R.,Pywell,R.F.,Goulson,D.,&Nowakowski,M.(2007).Comparingtheefficacyofagri‐environmentschemestoen-hancebumblebeeabundanceanddiversityonarablefieldmargins.Journal of Applied Ecology,44,29–40.

Carvell,C.,Meek,W.R.,Pywell,R.F.,&Nowakowski,M. (2004).Theresponse of foraging bumblebees to successional change in newlycreatedarable fieldmargins.Biological Conservation,118,327–339.https://doi.org/10.1016/j.biocon.2003.09.012

Couvillon,M.J.,Schurch,R.,&Ratnieks,F.L.W.(2014).Waggledancedistances as integrative indicatorsof seasonal foraging challenges.PLoS ONE,9,e93495.https://doi.org/10.1371/journal.pone.0093495

Crone, E. E., &Williams, N.M. (2016). Bumble bee colony dynamics:Quantifying the importance of land use and floral resources forcolonygrowthandqueenproduction.Ecology Letters,19,460–468.https://doi.org/10.1111/ele.12581

Dainese, M., Riedinger, V., Holzschuh, A., Kleijn, D., Scheper, J., &Steffan‐Dewenter, I. (2018). Managing trap‐nesting bees as croppollinators: Spatiotemporal effects of floral resources and an-tagonists. Journal of Applied Ecology, 55, 195–204. https://doi.org/10.1111/1365-2664.12930

DiPasquale,G.,Salignon,M.,LeConte,Y.,Belzunces,L.P.,Decourtye,A.,Kretzschmar,A.,…Alaux,C.(2013).Influenceofpollennutritiononhoneybeehealth:Dopollenqualityanddiversitymatter?PLoS ONE,8,e72016.https://doi.org/10.1371/journal.pone.0072016

Dicks, L. V., Baude, M., Roberts, S. P. M., Phillips, J., Green, M., &Carvell, C. (2015). How much flower‐rich habitat is enough forwild pollinators? Answering a key policy question with incom-plete knowledge. Ecological Entomology, 40, 22–35. https://doi.org/10.1111/een.12226

Dixon,K.W.(2009).PollinationandRestoration.Science,325,571–573.https://doi.org/10.1126/science.1176295

Fitter,A.H.,&Fitter,R.S.R. (2002).Rapidchanges in flowering timeinBritishplants.Science,296,1689–1691.https://doi.org/10.1126/science.1071617

Fitzpatrick,U.,Murray,T.E.,Paxton,R.J.,Breen,J.,Cotton,D.,Santorum,V.,…Brown,M.J.F.(2007).Rarityanddeclineinbumblebees–Atestof causes and correlates in the Irish fauna.Biological Conservation,136,185–194.https://doi.org/10.1016/j.biocon.2006.11.012

Forrest, J. R. K. (2015). Plant‐pollinator interactions and phenologicalchange: What can we learn about climate impacts from experi-mentsandobservations?Oikos,124,4–13.https://doi.org/10.1111/oik.01386

Garratt,M.P.D.,Coston,D. J., Truslove,C. L., Lappage,M.G., Polce,C., Dean, R., … Potts, S. G. (2014). The identity of crop pollina-tors helps target conservation for improved ecosystem services.Biological Conservation, 169, 128–135. https://doi.org/10.1016/j.biocon.2013.11.001

Goulson,D.,Lye,G.C.,&Darvill,B.(2008).Declineandconservationofbumble bees. Annual Review of Entomology,53,191–208.https://doi.org/10.1146/annur ev.ento.53.103106.093454

Goulson,D.,Nicholls, E., Botias, C., & Rotheray, E. L. (2015). Bee de-clinesdrivenbycombinedstressfromparasites,pesticides,andlackof flowers. Science, 347, 1255957. https://doi.org/10.1126/science.1255957

Greenleaf,S.S.,Williams,N.M.,Winfree,R.,&Kremen,C.(2007).Beeforagingrangesandtheir relationship tobodysize.Oecologia,153,589–596.https://doi.org/10.1007/s00442‐007‐0752‐9

Hegland, S. J., Nielsen, A., Lazaro, A., Bjerknes, A. L., & Totland,O. (2009). How does climate warming affect plant‐pollina-tor interactions? Ecology Letters, 12, 184–195. https://doi.org/10.1111/j.1461-0248.2008.01269.x

Heinrich, B. (1972). Physiology of brood incubation in bumblebeequeen, Bombus vosnesenskii. Nature, 239, 223–225. https://doi.org/10.1038/239223a0

Heinrich,B.(1979).Bumblebee economics.Cambridge:HarvardUniversityPress.

Horn,J.,Becher,M.A.,Kennedy,P.J.,Osborne,J.L.,&Grimm,V.(2016).Multiplestressors:UsingthehoneybeemodelBEEHAVEtoexplorehow spatial and temporal forage stress affects colony resilience.Oikos,125,1001–1016.https://doi.org/10.1111/oik.02636

Klein, A. M., Vaissiere, B. E., Cane, J. H., Steffan‐Dewenter, I.,Cunningham,S.A.,Kremen,C.,&Tscharntke,T.(2007).Importanceofpollinatorsinchanginglandscapesforworldcrops.Proceedings of the Royal Society of London. Series B, Biological sciences,274,303–313.https://doi.org/10.1098/rspb.2006.3721

Kovacs‐Hostyanszki, A., Haenke, S., Batary, P., Jauker, B., Baldi, A.,Tscharntke,T.,&Holzschuh,A.(2013).Contrastingeffectsofmass‐floweringcropsonbeepollinationofhedgeplantsatdifferentspatialand temporal scales. Ecological Applications,23,1938–1946.https://doi.org/10.1890/12-2012.1

Mandelik,Y.,Winfree,R.,Neeson,T.,&Kremen,C.(2012).Complementaryhabitat use by wild bees in agro‐natural landscapes. Ecological Applications,22,1535–1546.https://doi.org/10.1890/11‐1299.1

Menz, M. H., Phillips, R. D., Winfree, R., Kremen, C., Aizen, M. A.,Johnson,S.D.,&Dixon,K.W.(2011).Reconnectingplantsandpol-linators: Challenges in the restoration of pollination mutualisms.Trends in Plant Science, 16, 4–12. https://doi.org/10.1016/j.tplants.2010.09.006

Moquet, L., Mayer, C., Michez, D., Wathelet, B., & Jacquemart, A.‐L.(2015).Early spring floral foraging resources forpollinators inwetheathlands inBelgium. Journal of Insect Conservation,19,837–848.https://doi.org/10.1007/s10841‐015‐9802‐5

Ollerton, J., Winfree, R., & Tarrant, S. (2011). How many floweringplants are pollinated by animals?Oikos,120, 321–326. https://doi.org/10.1111/j.1600‐0706.2010.18644.x

Potts, S.G.,Vulliamy,B.,Dafni,A.,Ne'eman,G.,&Willmer, P. (2003).Linking bees and flowers: How do floral communities structurepollinator communities? Ecology, 84, 2628–2642. https://doi.org/10.1890/02-0136

12  |    Journal of Applied Ecology TIMBERLAKE ET AL.

Pywell, R. F.,Warman, E. A., Carvell, C., Sparks, T. H., Dicks, L. V.,Bennett, D., … Sherwood, A. (2005). Providing foraging re-sources for bumblebees in intensively farmed landscapes.Biological Conservation, 121, 479–494. https://doi.org/10.1016/j.biocon.2004.05.020

Requier,F.,Odoux,J.F.,Henry,M.,&Bretagnolle,V.(2017).Thecarry‐overeffectsofpollenshortagedecrease thesurvivalofhoneybeecolonies in farmlands. Journal of Applied Ecology, 54, 1161–1170.https://doi.org/10.1111/1365‐2664.12836

Requier, F.,Odoux, J. F., Tamic, T.,Moreau,N.,Henry,M.,Decourtye,A.,&Bretagnolle,V. (2015).Honey bee diet in intensive farmlandhabitats reveals anunexpectedlyhigh flower richnessandamajorrole of weeds. Ecological Applications, 25, 881–890. https://doi.org/10.1890/14-1011.1

Rotheray,E.L.,Osborne,J.L.,&Goulson,D.(2017).Quantifyingthefoodrequirementsandeffectsoffoodstressonbumblebeecolonyde-velopment. Journal of Apicultural Research,56,288–299.https://doi.org/10.1080/00218839.2017.1307712

Roulston,T.H.,&Goodell,K.(2011).Theroleofresourcesandrisksinregulating wild bee populations. Annual Review of Entomology,56(56),293–312.https://doi.org/10.1146/annurev‐ento‐120709‐144802

Rundlof,M.,Persson,A.S.,Smith,H.G.,&Bommarco,R.(2014).Late‐seasonmass‐floweringredclover increasesbumblebeequeenandmale densities. Biological Conservation, 172, 138–145. https://doi.org/10.1016/j.biocon.2014.02.027

Russo,L.,DeBarros,N.,Yang,S.A.,Shea,K.,&Mortensen,D. (2013).Supporting crop pollinators with floral resources: Network‐basedphenologicalmatching.Ecology and Evolution,3,3125–3140.https://doi.org/10.1002/ece3.703

Scheper, J., Bommarco, R., Holzschuh, A., Potts, S. G., Riedinger, V.,Roberts,S.P.M.,…Kleijn,D.(2015).Localandlandscape‐levelflo-ralresourcesexplaineffectsofwildflowerstripsonwildbeesacrossfourEuropeancountries.Journal of Applied Ecology,52,1165–1175.https://doi.org/10.1111/1365‐2664.12479

Steffan‐Dewenter,I.,&Schiele,S.(2008).Doresourcesornaturalene-miesdrivebeepopulationdynamicsinfragmentedhabitats?Ecology,89,1375–1387.https://doi.org/10.1890/06‐1323.1

Timberlake,T.,Vaughan,I.,&Memmott,J.(2019).Datafrom:Phenologyoffarmlandfloralresourcesrevealsseasonalgapsinnectaravailabil-ityforbumblebees.Dryad Digital Repository,https://doi.org/10.5061/dryad.3nk236h

Vaughan,I.P.,&Ormerod,S.J.(2012).Large‐scale,long‐termtrendsinBritishrivermacroinvertebrates.Global Change Biol,18,2184–2194.https://doi.org/10.1111/j.1365‐2486.2012.02662.x

Westphal,C.,Steffan‐Dewenter,I.,&Tscharntke,T.(2003).Massfloweringcropsenhancepollinatordensitiesatalandscapescale.Ecology Letters,6,961–965.https://doi.org/10.1046/j.1461‐0248.2003.00523.x

Williams, N. M., Regetz, J., & Kremen, C. (2012). Landscape‐scaleresources promote colony growth but not reproductive per-formance of bumble bees. Ecology, 93, 1049–1058. https://doi.org/10.1890/11-1006.1

Willmer,P.(2011).Pollination and floral ecology.Princeton,NJandOxford:PrincetonUniversityPress.

Wood, S. N. (2011). Fast stable restricted maximum likelihoodand marginal likelihood estimation of semiparametric gen-eralized linear models. Journal of the Royal Statistical Society. Series B (Statistical Methodology), 73, 3–36. https://doi.org/10.1111/j.1467‐9868.2010.00749.x

Wood, T. J., Holland, J. M., Hughes,W. O. H., & Goulson, D. (2015).Targetedagri‐environment schemes significantly improve thepop-ulation size of common farmland bumblebee species. Molecular Ecology,24,1668–1680.https://doi.org/10.1111/mec.13144

WorldBank. (2015).World development indicators – Agricultural landcover.DatafromUNFoodandAgricultureOrganization.Retrievedfromhttps://data.worldbank.org/indicator/AG.LND.AGRI.K2

SUPPORTING INFORMATION

Additional supporting information may be found online in theSupportingInformationsectionattheendofthearticle.

How to cite this article:TimberlakeTP,VaughanIP,MemmottJ.Phenologyoffarmlandfloralresourcesrevealsseasonalgapsinnectaravailabilityforbumblebees.J Appl Ecol. 2019;00:1–12. https://doi.org/10.1111/1365‐2664.13403