Keystone predators PDF

Preview

Summary

biol notes...

Description

|

|

Received:14September2018   Revised:19December2018    Accepted:8January2019 DOI:10.1002/ece3.4953

ORIGINAL RESEARCH

Sex and occupation time influence niche space of a recovering keystone predator Erin U. Rechsteiner1,2

 | Jane C. Watson3 | M. Tim Tinker4,5 | Linda M. Nichol6 |

Matthew J. Morgan Henderson2 | Christie J. McMillan2,7 | Mike DeRoos2 | Marie C. Fournier2 | Anne K. Salomon8 | Leah D. Honka9 | Chris T. Darimont1,2 1

DepartmentofGeography,Universityof Victoria,Victoria,BritishColumbia,Canada

Abstract

2

Predatorsexertstrongeffectsonecologicalcommunities,particularlywhentheyre‐

HakaiInstitute,HeriotBay,British Columbia,Canada

occupyareasafterdecadesofextirpation.Withinspecies,sucheffectscanvaryover

3

VancouverIslandUniversity,Nanaimo, BritishColumbia,Canada 4

DepartmentofEcologyandEvolutionary Biology,UniversityofCaliforniaatSanta Cruz,SantaCruz,California 5

NhydraEcologicalConsulting,St. Margaret’sBay,NovaScotia,Canada 6

FisheriesandOceansCanada,Pacific BiologicalStation,Nanaimo,British Columbia,Canada 7

MarineEducationandResearchSociety, PortMcNeill,BritishColumbia,Canada 8

SchoolofResourceandEnvironmental Management,SimonFraserUniversity, Burnaby,BritishColumbia,Canada 9

SalmonWatershedsProgram,Pacific SalmonFoundation,Vancouver,British Columbia,Canada

timeandbysexandcascadeacrosstrophiclevels.Weusedaspace‐for‐timesubstitu‐ tiontomakeforagingobservationsofseaotters(Enhydra lutris)acrossagradientof reoccupation time (1–30years), and nonmetric multidimensional scaling (nMDS) analysistoaskwhether(a)seaotternichespacevariesasafunctionofoccupation timeand(b)whethernichespacevariesbysex.Wefound that nichespacevaried amongareasofdifferentoccupationtimes.Dietarynichesatshortoccupationtimes weredominatedbyurchins(MesocentrotusandStrongylocentrotus spp;>60%ofdiets) inopenhabitatsat10–40mdepths.Atlongeroccupationtimes,nichesweredomi‐ natedbysmallclams(Veneroida;>30%diet),mussels(Mytilusspp;>20%diet),and crab(Decapoda;>10%diet)inshallow(60%), and urchins (~25%) from deep waters (>40m), and females and territorial males consuming smaller, varied prey from shallow waters (50% clams, whereas territorial males and fe‐

thatniche space significantly differsamong sexclasses (ANOSIM,

males consumed ~15% urchins, ~20% each of small and large

R=0.36, p10% each of small crabs and mussels, and ~10% snails

and territorial males used a different niche space than bachelor

(Turbinidae;Figure3).Meanpreysizewashighestforbachelormales

males(SupportingInformationTableS4).Bubbleplotsdepictingthe

(13.34±0.37cm) and lower for territorial males (10.03±0.26cm)

variablesmostimportantindrivingsimilaritieswithinsexclasses,as

andfemales(9.94±0.26cm)(Figure3).Shannonindicesshowedlow‐

identifiedbySIMPER(>8% contributionto within‐group similarity;

estdiversityinbachelormalediets(H′=2.21)andhigherdiversityin

Figure7b), illustrate differences in niche spacewere drivenby ur‐

territorial male and female diets (H′>2.50; Figure 3). Monte Carlo

chins, clams, geoduck, small crabs, open water, and shallow water

analysisindicatedthatthemeanenergyintakeratewashighestfor

(SupportingInformationTableS5).Posthocvectorcorrelationsin‐

bachelormales(26.67±4.72kcal/min)andlowerforterritorialmales

dicatedthathigherdietdiversityanduseofintertidal,shallow,kelp

(12.89±1.72kcal/min)andfemales(11.29±0.63kcal/min;Figure3).

canopy,andseagrassareaswerecorrelatedwithfemalesandterrito‐ rialmales,whereasuseofdeep,mid,andopenareas,andlargerprey

3.2 | Sea otter niche space

sizes were correlated with bachelor males (Figure 7a, Supporting InformationTableS6).

Otters from different occupation areas strongly diverged in niche space.Clusteranalysisshowedgroupingswith63%similaritybyoc‐ cupationareaandbyInitialandEstablishedsitesattheCalvertoccu ‐ pationarea(Figure4).ResultsofthenMDSshowedtwo‐dimensional stressof0.12anddissimilaritiesamongoccupationareas(Figure5a).

4  |  DI S CU S S I O N 4.1 | The dynamics of sea otter niche space

TheresultsoftheANOSIMsupportedstatisticallysignificantdiffer‐

Inthisstudy,weusedseaottersasamodeltoevaluatesupportfor

ences in niche space among occupation areas (ANOSIM, R=0.74,

thehypothesesthata)predatornichespacechangesduringpopula‐

p8%contribution

urchinsinopenwater,beforetransitioningtosoft‐sedimenthabitats

to within‐group similarity; Figure 5b). We found that niche space

toforageonclamsindeepwaters,andthentokelpcanopyandin‐

similarities within occupation areas were determined primarily by

tertidal areas where smaller and more diverse invertebrates were

urchins, clams,geoduck,mussels, smallcrabs,openwater (i.e.,ab‐

acquired.Bothmeanpreysizeandenergyintakerateswerelower

sence of kelp canopy), and shallow water (Supporting Information

inlong‐occupiedareasthaninrecentlyoccupiedareas.Theseniche

TableS2). Posthoc vectorcorrelations indicatedthat use ofinter‐

spacedifferencesamongsiteswithvaryingoccupationtimeslikely

tidal,shallow,andkelpcanopyareaswascorrelatedwithlongeroc‐

reflectsea otter behaviouralresponsesto variationinpreyavaila‐

cupationtimes,whereasuseofdeep,mid,andopenareas,largeprey

bility—alegacyoftheeffectsthatseaottersexertonbenthicprey

sizes,andhigherratesofenergyintakewerecorrelatedwithshorter

communitiesastheyrecover(i.e.,Estes,Riedman,Staedler,Tinker,

occupationtimes(Figure5a,SupportingInformationTableS3).

&Lyon,2003;Tinker,Bentalletal.,2008).Thepatternweobserved

Sexes also showed strong divergence in niche space. Cluster

at a single occupation area over the four‐year study period offers

analysis identified groupings by sex class with 45% similarities

evidencetosupportthis hypothesis, withseaotterdietschanging

(Figure6).ResultsofthenMDSshowedtwo‐dimensionalstressof

fromurchin‐dominated(>60%)toclam‐dominated(~50%)afteronly

0.13 and dissimilarities among females and territorial males, and

~1yearofoccupation.

RECHSTEINER ET al.

|

      3329

F I G U R E 7   Nonmetric multidimensionalscalinganalysis(nMDS) plotofseaotternichespacewith(a) clustersidentifiedinFigure6with45% similarity,andenvironmentalvectorswith ≥0.5correlationtodissimilarities,and(b) bubbleplotsdepictingthemostcommon preygroups,withbubblesegments approachingsizesofsegmentsinthe legendrepresenting~80%ofthedietby frequencyofoccurrence

Nichespacealsovariedamongfemale,territorialmale,andbach‐

Althoughnichespacehasnotbeenthoroughlyexaminedinsea

elormaleseaotters.Bachelormalesatrecentandmoreestablished

otters(butseeSillimanetal.,2018),seaotterdietshavebeenwell

areas (i.e., 1–8years)foraged predominantlyinopen(nokelpcan‐

studied.InBritishColumbia(Breen,Carson,Foster,&Stewart,1982;

opy or seagrass) habitats, using deeper waters than females, and

Hessing‐Lewis et al., 2017; Honka, 2014), Washington (Laidre &

consumed ~70% clams. In contrast, females and territorial males

Jameson,2006;Haleetal.,inpress),California(Estesetal.,1982,

exhibitedhigherdietdiversityandusedkelpcanopy,seagrass,and

2003;Tinker,Doaketal.,2008;Tinkeretal.,2012),andAlaska(Hoyt,

shallowandintertidalwaterstoforageonadiversesuiteofsmaller

2015;Kvitek,Oliver,DeGange,&Anderson,1992;Weizman,2013),

prey.Together,thesefindingssuggestthatthenicheoccupiedbysea

seaotterdietshavebeenfoundtodifferamongrecentlyandlong‐

ottersonthecentralcoastofBCiscontext‐specific,dependingon

occupiedareas.InCalifornia,whereindividualseaottershavebeen

bothoccupationtimeandsex.

observed for decades, sea otter diets diversify at the population

|

3330   

levelasoccupationtimeincreases(Estesetal.,1982).InWashington

  

sea otter diets and that energy is the primary resource driving sea

State,seaotterdietsareaffectedmorebykelphabitatthanoccu‐

otter prey selection (Oftedal et al., 2007). However, variability in

pationtimeandpopulationdensity(Haleetal.,inpress).Ontheex‐

macronutrient composition among prey species can be an import‐

posedWashingtoncoast,kelphabitatsmaybeanimportantsource

antdriverofdietchoiceandnichespace(Machovsky‐Capuskaetal.,

offood,aswellasprovidingshelterfromstormsandpredators(i.e.,

2018; Machovsky‐Capuska, Senior, Simpson et al., 2016; Mayntz,

Thometzetal.,2016),whereasinBC,theroleofkelpmaybelessim‐

Raubenheimer, Salomon, Toft, & Simpson, 2005; Raubenheimer,

portantbecausethereismoreshelterprovidedbyislandsandinlets

Simpson,&Mayntz,2009;Tait,Raubenheimer,Stockin,Merriman,&

(Hessing‐Lewisetal.,2017).

Machovsky‐Capuska,2014).Accordingly,includingdataonthemac‐

Differencesinthedietsoffemaleandmaleseaottershavealso

ronutrientcontentofpreymayenhanceourcharacterizationofniche

been noted in other regions. In California, individual males over

space,butquantifyingmacronutrientprofilesforthediversesuiteof

1.5years consumed more diverse prey than females, though this

preythatseaottersconsumedwasbeyondthescopeof thisstudy.

likelyincludedmalesintransitionbetweenbachelorandterritorial

However,futureworkshouldincludeamorecomprehensiveexamina‐

stages(Elliot‐Smith,Newsome,Estes,&Tinker,2015).Inourstudy

tionofnutritionaldimensionstoseaotterpreyselection.

bachelormaleshadtheleastdiversediets,withlargegroupsofbach‐

Energy intake was approximated by an established analytical

elormales(n=~50‐150)oftenmovingenmasse(within~1.5km)to

frameworkusedinmanyseaotterforagingstudies(e.g.,Tinker,Bentall

new rafting and foraging areas where prey were large and forag‐

etal.,2008;Tinkeretal.,2012;Hessing‐Lewisetal.,2017).Thisanalysis

ingsuccesswashigh.Incontrast,femalesmayrestricttheirdietsto

makesuseofapublisheddatasetonpreyediblebiomassandcaloric

increasetimespentatthesurfacewiththeirpups(Thometzetal.,

content from 76taxa collected in allseasonsover a4‐yearperiodat

2016).Thesesex‐basednichedifferencesareimportantparticularly

multiplesitesinCaliforniaandAlaska(seeOftedaletal.,2007;Tinker,

becausethefeedingecologyandhabitatuseoflargemalegroups,

2015 for more details). Although these data are broadly representa‐

which occur across thegeographic range ofsea otters,havebeen

tive,werecognizetheirlimitationforassessingfine‐scalevariationin

largelyoverlooked.

preyquality.Whilesuchfine‐scalevariationisundoubtedlyimportant, ourstudywasdesignedtoexaminecoarse‐scalepatternsofnicheuse.

4.2 | Study limitations Our study indicates that changes occur in sea otter niche space

Becausewecollecteddataovera4‐yearperiod and sampledequally acrosswinterandsummer,webelievethatwecapturedvariationinsea otterdietsandenergyintakeatthescaleofinterest.

with increasing occupation time, changes that are likely a prod‐

Disentanglingtherelativeeffectsofsexandoccupationtime

uctoftheecologicaleffectsexertedbytheseaottersthemselves.

in sea otter niche space is difficult because the two are linked

However,theinherentweaknessofspace‐for‐timeframeworkslies

viathenaturalhistoryofseaotters.OnthecentralcoastofBC,

indeterminingtheextenttowhichdifferencesare afunctionof

groupsofbachelormales(typicallyn =>100malesineachraft)oc‐

predatorrecoveryandoccupationtime,orreflectvariationamong

curredattherangeedge(CalvertInitialandEstablished)andinthe

sites in theabsence ofpredator recovery.Our observational de‐

rangecentre(Simonds).Unpublisheddata(50foragingbouts)col‐

signmitigatesthislimitation.Weusedmultipleobservationsites

lectedoveratwo‐weekperiodinAugust2016byEURandJCWin

(n=4–6)withineachoccupationareatoreplicateoccupationtime

KyuquotSound(50.0°N,127.4°W;>20 0kmsouthofthecentral

and monitored changes within each occupation area over the

coastofBC),inanareaoccupiedbybachelormaleseaotterssince

four‐yearstudyperiod.OnlyCalvert,themostrecentlyoccupied

the1980s, corroborated our findings: This bachelor male group

area,showedanyindicationofnichechange,andwesuspectthis

foraged primarily on small clams (~70% of diet) in open waters.

isbecauseinitialchangeshappenrapidly,whereasfurtherchanges

Further unpublished data (50 foraging bouts) collected over a

takelongerthan4yearstoobserve.Tointerpretwhether,forex‐

two‐weekperiodinJuly2017byLMNandCJMintheNuchatlitz

ample,seaotterdietsatlongeroccupationtimestransitionfrom

Islands (49.8°N, 126.9°W; ~40km south of Kyuquot Sound), at

clam‐dominated to mussel‐dominated and smaller invertebrate‐

an area occupied by sea otters since the 1990s, similarly found

dominated(asappearstobethecasewhencomparingMcMullins

that the diet of bachelor males was comprised mostly of clams

toGosling),we wouldneedtoobserveottersand theassociated

(~70%).Ourfindingsindicatethatevenatlong‐occupationtimes,

preycommunitiesineachoccupationareaonatimescaleofdec‐

bachelormalesexploitasoft‐sediment,open‐waternichespace

ades.Thiswouldalsoconfirmwhetherthechangesdetectedfrom

wheretheyfeedonclams;however,atlongeroccupationtimes,

ourspace‐for‐timedesignareindeedafactoroftheecologicalef‐

clamsweresmallandenergyintakerateswerelikelylowerthan

fectsofseaottersonthepreycommunity(e.g.,Tinker,Doaketal.,

onthecentralcoastofBC.

2008;Tinkeretal.,2012). Ourstudyfocusedontheenergyintake,diet,andhabitatuseof seaotters.Seaotters,lackingtheinsulatingblubberofmostmarine mammals,havehighmetabolicratesandconsume~25%oftheirbody‐

4.3 | Ecological and conservation implications of changing niche space in sea otters

weight daily (Costa & Kooyman, 1982). Previous research suggests

Sex‐relateddifferencesinfeedingecologyandhabitatusearerarely

thatmostcriticalmacro‐andmicronutrientsarewellrepresentedin

consideredinhabitatmanagementorecologicalinteractions,despite

|

RECHSTEINER ET al.

      3331

thesedifferenceshavingpotentiallymajoreffects(DuToit,2005).In

could act as a source of disturbance to plants (i.e., Alexandre,

apolygynousspeciessuchastheseaotter,thenicheusedbymales—

Santos, & Serrão, 2005) and invertebrates (Kvitek et al., 1992).

and the associated effects on survival—may be less important to

Recently,theroleofseaotterpredationoncrabsinseagrasscom‐

populationproductivitythanthenicheusedbyfemales,becausein‐

munitieshasbeenshowntoinitiateatrophiccascadethatincreases

dividualfemalescontributemoretopopulationrecoverythanmales

seagrassbiomass(Hughesetal.,2013).Thesestudiesshowthatsea

(Emlen&Oring,1977;Tarjan,2016).Ourfindingthatfemalesgain

ottercommunityinteractionsarecontext‐dependentandthatthe

lessenergyperminuteofforagingthanbachelormalesindicatesthat

ecologicalconsequencesofalternativeniche‐usepatternsmaybe

theymustspendmoretimeforagingthanmales.Moreover,thehigh

importanttoconsiderintermsofbeneficialconservationimpacts,

costoflactation(Thometzet al.,2016)andparental care mayput

fisheriesinteractions,andothermanagementandconservationpri‐

femalesatagreaterconservationriskthanmales.Thisfindingsug‐

orities(Sillimanetal.,2018).

geststhatsitefidelityinfemales,eventolower‐energypreyareas,is important,perhapstoavoidpredatorsorduetootherreproductive constraints. Thus, recovery planning should consider that females and territorial males may require different prey and habitat than bachelormales.

4.4 | The dynamics of niche space in recovering predators Most ofthe world's large predators have suffered widespread ex‐

Understanding how prey selection and the range of habitats

tirpation, limiting our understanding of how ecological communi‐

usedbyseaottersvaryoverthecourseofrecoveryhasimportant

ties function with intact predator populations (Estes et al., 2011;

implicationsforseaotterconservationandaffectsourunderstand‐

Jackson,2001;Silliman etal.,2018)...