RESEA RCHARTICL E Humpback whaleªsuper-groupsº ±A novel low-latitude feedingbehaviour ofSouthern Hemisphere humpbackwhales(Megaptera novaeangliae) inthe Benguela Upwelling System Ken P.Findlay 1 , S. Mduduzi Seakamela 2 , Michael A.Mey Èer 2 , Stephen P.Kirkman 2 , Jaco Barendse 3 , David E.Cade 4 , David Hurwitz 5 , Amy S.Kennedy 6 , Pieter G.H.Kotze 2 , Steven A.McCue 2 , Meredith Thornton 7 , O. Alejandra Vargas-Fonseca 8 , Christopher G. Wilke 9 1 Mamma lResearch InstituteWhaleUnit,University ofPretoria, Pretoria,Wynberg, SouthAfrica/Current address: Research Chair:Oceans Economy, CapePeninsula UniversityofTechnolo gy,Cape Town, South Africa, 2Department ofEnvironm entalAffairs, Branch Oceans andCoasts, Victoria&Alfred Waterfront , Cape Town, SouthAfrica, 3Mamma lResearch InstituteWhaleUnit,Univers ityofPretoria, Pretoria,0001, South Africa[c/o16Ebor Rd,Wynberg ,South Africa] /Current address: MarineStewardsh ipCouncil, 1Snow Hill, London, UnitedKingdom, 4Departm entofBiology, Hopkins MarineStation, Stanford University, 120 Ocean ViewBlvd, Pacific Grove, CA,United StatesofAmerica, 512 Victory Way,Simon's Town.Cape Town, SouthAfrica, 6The Marine Mamma lLaboratory, AlaskaFisheries ScienceCenter,7600Sand Point Way NE.Seattle, WA,United StatesofAmerica, 7Mammal Research InstituteWhaleUnit,University of Pretoria, Pretoria,Wynberg ,South Africa, 8ConserBi oFounda tion,Heredia, CostaRica,9Department of Agriculture, ForestryandFisherie s,Branch :Fisheries Managem ent,Foretrust Building,Foreshore, Rogge Bay, Cape Town, SouthAfrica  findlayk@ cput.ac.za Abstract Southern Hemisphere humpbackwhales(Megaptera novaeangliae) generallyundertake annual migrations frompolar summer feedinggrounds towinter calving andnursery grounds in subtropical andtropical coastalwaters. Evidence forsuch migrations arisesfromseasonal- ity ofhistoric whaling catches bylatitude, Discovery andnatural markreturns, andresults of satellite taggingstudies. Feeding isgenerally believedtobe limited tothe southern polar region, whereAntarctic krill(Euphausia superba)hasbeen identified asthe primary preyitem. Non-migrations and/or suspended migrationstothe polar feeding grounds havepreviously been reported fromasummer presence ofwhales inthe Benguela System,wherefeeding on euphausiids (E.lucens), hyperiidamphipods (Themistogaudichaudii),mantis shrimp (Pterygosquilla armatacapensis) andclupeid fishhas been described. Threerecent research cruises (inOctober/Novem ber2011, October/Novemb er2014 andOctober/Novem ber2015) identified largetightly-spaced groups(20to200 individuals )of feeding humpback whales aggregated overatleast aone-month periodacross a220 nautical mileregion ofthe southern Benguela System.Feedingbehaviour wasidentified bylunges, strongmilling andrepetitive and consecutive divingbehaviours, associatedbirdand seal feeding, defecations andthe pungent ªfishyºsmellofwhale blows. Although nodedicated preysampling couldbecarried out within thetightly spaced feeding aggregations, observationsofE. lucens inthe region of PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 1/18 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 23(1 $&&(66 Citation: FindlayKP,Seakamela SM,MeyÈer MA, Kirkman SP,Barendse J,Cade DE,etal. (2017) Humpback whaleªsuper-groups º± A novel low- latitude feeding behaviour ofSouthern Hemisphere humpback whales(Megaptera novaeangliae) inthe Benguela Upwelling System.PLoSONE12(3): e0172002. doi:10.1371/journal.pone.017200 2 Editor: Alexander J.Werth, Hampden Sydney College, UNITED STATES Received: June7,2016 Accepted: January30,2017 Published: March1,2017 Copyright: Thisisan open access article,freeofall copyright, andmay befreely reproduced, distributed, transmitted,modifie d,built upon, or otherwise usedbyanyone forany lawful purpose. The work ismade available undertheCreative Commons CC0public domain dedication. Data Availabilit yStatement: Allrelevant dataare within thepaper anditsSupport ingInformation file. Funding: DavidHurwitz isemployed byThe Simon's TownBoatCompany. TheSimon's Town Boat Company providedsupportinthe form of salary forauthor DH,butdidnot have any additional roleinthe study design, datacollection and analysis, decisiontopublish, orpreparation of groups andthefullstomach contentsofmantis shrimp frombothaco-occurrin gpredatory fish species (Thyrsites atun)andone entangled humpback whalemortality suggestthesemaybe the primary preyitems ofatleast some ofthe feeding aggregation s.Reasons forthis recent novel behaviour patternremainspeculative, butmay relate toincreasing summerhumpback whale abundance inthe region. Thesenovel,predictable, inter-annual, lowlatitude feeding events provide considerable potentialforfurther investigation ofSouthern Hemisphere hump- back feeding behaviours inthese relatively accessible low-latitude waters. Introduction Southern Hemisphere humpbackwhales(   ) aregenerally understood to migrate seasonally betweensummer high-latitude Antarcticfeedinggrounds andlow-lati- tude winter calving andmating grounds intropical andsubtropical coastalwatersofSouthern Hemisphere continentsandisland archipelagos [1±4].Evidence forsuch migrations arises from thetiming andseasonality ofhistoric whaling catches[5±12]andcontemporary sighting survey results[13,14]bylatitude,   mark([2,15,16] and natural markreturns, [17, 18] aswell asresults ofsatellite taggingstudies[19±25]. Ofthe seven Breeding Stocksrecog- nised inthe Southern Hemisphere bythe International WhalingCommission [26],themigra- tions ofthe BBreeding Stockoccurs between breeding groundsoffthe west coast ofsouthern and central Africa(thecoast between AngolaandGabon) andtheSouthern Ocean.Although historical catcheswererecorded offthe Cape coastofSouth Africa andNamibia inthe austral winter, springandsummer [6,9,12], ithas been argued thatthemajority ofthe population migrates offshorealongtheWalvis Ridgetothe west ofthe southern Benguela region[6,9, 27]. Atthe northern limitofSouthern Hemisphere humpbackwhalemigrations, Rasmussen   . [4] found adirect correlation betweenseasurface temperature andthelocation ofwinter- ing areas, indicative thatwarmer waters(ofbetween 21.1and28.3ÊC) areintegral indefining winter breeding migration patterns.Theprevailing coldtemperate oceanographic conditions in the Benguela Upwelling Systemoffthe west coast ofSouthern Africaresults inthe breeding grounds onthis coast being significantly furthernorth(0±16ÊS) thanoffthe east coast (13± 26ÊS). Feeding haslong been understood onthe whole tobe seasonally limitedtosummer and spatially limitedwithinthesouthern high-latitude polarregions [3,9,10, 28, 29], with analyses of stomach contents ofwhales caughtoutside theSouthern Oceaninthe winter season often showing littleorno food intheir stomachs [3,9,30±32]. InSouthern Oceanregions, Antarctic krill (  )hasbeen identified asthe primary preyitem within stomachs exam- ined onwhaling flensingplatforms [3,9,10, 33] and thedistribution ofhumpback whalesdur- ing thefeeding seasoninSouthern Oceanwatersappears strongly coupledtothat ofthese krill [34±36]. Duringthemigrations tolow latitude watershumpback whalesareunderstood tosur- vive offenergy reserves accumulated duringsummer feedingaswhales caughtatthe end of winter inthe low latitude grounds wererecorded tobe considerably leanerwithlower oil yields thanthose caught earlyinwinter [31].Whereas Corkeron andConnor [37]noted that the summer humpback whalemigrations tothe high latitudes areeasily understood asfeeding migrations [38±40],thereisalso considerable evidencethatsome populations, includingthose off the west coasts ofSouth America [41]andsouthern Africa[42±44], suspendtheirsouth- ward migrations topolar waters whilstothers mayremain inproductive temperatewatersdur- ing thenon-breeding season.Furthermore, anon-migratory humpbackwhalepopulation in Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 2/18 the manuscript. Thespecific roleofthis author is articulated inthe `author contributions' section. Competing interests:We have thefollowing interests: DHisemploye dby The Simon's Town Boat Company. Therearenopatents, products in developm entormarketed productstodeclare. This does notalter ouradherence toall the PLOS ONE policies onsharing dataand materials. the Arabian Seahasbeen described byMinton  . [45]. Records offeeding byhumpback whales during theirmigrations inlower latitudes havebeen reported, althoughmanyofthese records areofindividual animalsorsmall groups inapparent opportunistic feedingbehaviours [46±53]. Themore regular encounters offeeding fromtheSouthern Benguelaregion[42±44], northern Patagonia [41]andtheMagellan StraitsofChile [54]andtheEden coastofNew South Wales, Australia, [55,56]suggest thatmid-latitude feedingandsuspended migrations do occur onallthree ofthese continents, althoughfeedingisat arelatively lowfrequency com- pared tofeeding athigh latitudes. The distributions offeeding baleenwhales aregood indicators ofoceanographic productiv- ity because thesewhales require denseandpredictable preyaggregations forsuccessful forag- ing [34, 57].The productive coastalBenguela Upwelling Systemextends frombetween 14± 17ÊS to37ÊS offthe southern Africanwestcoast. Distinct upwelling cellsoccur within theSys- tem, themost intense ofwhich isthe almost permanent LuÈderitz upwelling cell(inthe region of about 26ÊSto27ÊS) thateffectively dividestheecosystem intophysically andbiologically distinct northern andsouthern subsystems [58].Thisstudy focussed withinthesouthern Ben- guela subsystem betweenStHelena Bayand Cape Point, theoceanography ofwhich isdomi- nated byseasonal windcycles ofpredominantly south-easterlyoffshorewindsinsummer and north-westerly windsinwinter, withthesummer south-easterly winds(October toMarch) resulting inpulsed, wind-driven coastalupwelling [58,59].The high productivity andabun- dant concentrations ofzooplankton associatedwiththeupwelling [60]results inthe southern Benguela areabeing animportant nurserygroundforseveral fishspecies ofecological and commercial importance, includingsmallplanktivorous pelagicshoaling speciessuchassar- dine (   )andanchovy (    ) [58,59]. Historic catches(1909to1928) madeatthe shore-based whalingstationsinthe Saldanha Bay region onthe west coast ofSouth Africa showed abimodal [61]migration patternbelieved to comprise thenorthward (July/August) andsouthward (October/November) migrations[6, 8], with Olsen [6]noting thatthewhaling seasonsfrom1912to1913 wererelatively longcom- pared toother regions lastingtillmid-December. However,Olsen[6]translated inHinton [61] also reported thatªInthecold currents ofSaldanha Bayone could meetsingle young males probably yearlings, throughout thewhole summerº, althoughtheprovenance ofthis informa- tion isquestioned byFindlay andBest [62]asthe whalers onlyoperated betweenthemonths of March andlateDecember. Best . [42] hypothesised thatthesuspension ofthe southward migration ofhumpback whalesonthe west coast ofSouth Africa wasinresponse tolocally abundant prey.Historic reportsofthe presence ofhumpback whaleswithinthenorthern Ben- guela subsystem duringsummer monthsarisefrom Townsend's [63]records of19 th century catches inJanuary offWalvis Bay,Namibia (23ÊS)andacomment ofKeeler's (in[64]) that many humpback whaleswerefound offHollam's BirdIsland, Namibia inJanuary 1829.InJan- uary, themajority ofanimals areexpected tobe on the Antarctic feedinggrounds. Notwithstanding thenumerous reportsofapparent orsuspected feedingbyhumpback whales withintheBenguela Upwelling System[6,9,42, 44, 62], orsome evidence forrelatively large loose feeding associations (.20individuals) atalocal scale [46]thispaper describes a unique humpback whalefeeding behaviour thathasbeen observed duringthemonths ofOcto- ber and November inrecent years(2011, 2014and2015) atamuch widerregional scalewithin the southern Benguela sub-system. Methods andmaterials In this paper, weuse theterm ªsuper-groupº todescribe groupsof20 ormore individual humpback whalesestimated tobe within fivebody lengths oftheir nearest neighbour. Such Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 3/ 18 observations arenovel inthat prior observations offeeding byhumpback whalesinthe region (e.g. [42±44]) wereofloose aggregations ofsmall groups ofwhales (ofupto20 individuals). These records ofthe ªsuper-groupsº ofhumpback whaleswithinthesouthern Benguela region arise from twosources, namelyprimary observations fromdedicated researchcruises,and incidental observations reportedbythe general public. Three dedicated researchcruiseswerecarried outincoastal watersbetween 32Ê20'Sand 34Ê20'S offthe south-western Caperegion ofSouth Africa (Fig1)inlate October±early November of2011, 2014and2015 andfocussed onidentifying thedistribution, relativeabun- dance andmovements ofmigratory largewhales andother cetacean faunainrelation tobiotic and abiotic factors(including upwellingareas)andthetrophic ecology oflarge whales within this coastal area.The2011 cruise wasconducted from10November to20 November, the2014 cruise from28October to8November, whilethe2015 cruise wasconducted from29October to 7November. The2011 cruise wasconducted aboardtheSouth African Department ofEnvi- ronmental Affairs'research vessel  , whilebothofthe 2014 and2015 cruises werecarried out aboard theSouth African Department ofAgriculture ForestryandFisheries researchvessel the     (hereafterbothvessels arereferred toasªmothershipº below).During 2014 and2015 concurrent oceanographic cruises(tothose carried outonthe      ) werecarried outinthe region aboard the   . Duetolimitations inthe launching and recovery ofrigid hullinflatable boats(RHIBs) fromthe    inthe prevailing south-westerly swellconditions during2014and2015 thisvessel generally anchored insheltered conditions atnight withtheovernight positiondefining thearea ofsearch effortthefollowing day. Searching forwhales wasgenerally initiatedeachmorning fromthebridge ormonkey island of the mothership   tothe 100m isobath (some5n. miles offshore) andpositions ofwhale groups werecommunicated byradio toone ortwo RHIBs following asternorparallel tothe     ifconditions forRHIB operations allowed.Onceapproachable groupsofwhales were sighted, research personnel weretransferred tothe RHIBs forapproach ofwhale groups, and RHIBs couldbedirected fromwhale group towhale group bythe     , until late afternoon whenpersonnel weretransferred backtothe ship prior toseeking asheltered anchorage orproviding alee, ifswell conditions allowedforthe recovery ofthe RHIBs. In2011, the RHIBs weredeployed fromthe    atsea only once thefirst approachable whalegroup was sighted, andthereafter travelledasternorparallel tothe mothership betweenwhalegroups. Where weather conditions precludedRHIBoperations, whalegroups wereapproached bythe mothership. Allapproaches towhale groups andallresearch activities reportedinthis study were carried outunder research permits(numbers RES2011/70, RES2014/61, RES2015/94) granted to Mammal ResearchInstitute,University ofPretoria bythe South African Department ofEnvi- ronmental Affairsandresearch permit(number RES2015/DEA) grantedtothe South African Department ofEnvironmental Affairsbythe Department ofAgriculture, ForestryandFisheries under provisions ofthe South African MarineLivingResources Act(Act 18of1998) andunder ethics permits ofthe University ofPretoria, SouthAfrica (Reference numbersÐEC020-12 and EC061-15). Due tothe tight spacing andrepeated divingbehaviour ofwhales withinªsuper-groupsº (at times lessthan 2±3mapart) andthehigh riskofentangling whaleswithinsampling gear,no plankton sampling couldbecarried outfrom theRHIBs inthe near vicinity ofthese groups. Group sizesofdedicated sightingswereestimated independently byexperienced observersas upper, lowerandbest estimates duringtheclose approaches ofthe group, eitherbyRHIBs or by the mothership sothat group sizeestimates werenotcarried outatdistance andconse- quently notinfluenced byBeaufort ScaleorSea State. Bestestimates ofgroup sizewere not necessarily themean ofthe overall upperandlower estimates asindependent estimateswere made byallofthe personnel observingthegroup andthefinal estimated groupsizewas based Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 4/ 18 on consensus ofthese individual estimates.Theestimates ofthe sizes ofªsuper-groupsº pertain only tothose individuals withinfivebody lengths ofthe nearest neighbour atthe time ofthe Fig 1.Locations ofthe Study Region anddedicate d(open circles) andincidental observation s(open squares) offeeding ªsuper-gro upsºoffthe south westCape coast ofSouth Africa. The200m isobath is shown. Theinset expands thearea shaded ingrey. doi:10.1371 /journal.pone. 0172002.g001 Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 5/ 18 largest aggregation duringtheencounter. Fluidityofthe associations withingroups andmove- ment ofidentified individuals betweengroupswithinaggregations (includinginand outof ªsuper-groupsº) wasdifficult todetermine inthe field, duetothe repetitive deepdivebehav- iour ofthe individual whales,whichmadethedetermination ofgroup sizesdifficult. Where such difficulty precluded accurategroupsizeestimation, thelowest estimate wasselected. Incidental observations ofªsuper-groupsº werereported bythe general publicfromaircraft during 2015including duringtwosightseeing flightsmadebyDH between Koeberg andSt Helena Bay(including offshoreofDassen Island)on19October and26October 2015.These two flights werecarried outatover 300m above sealevel inaCessna 185aircraft. Allofthe reported incidental observations wereverified fromphotographs. Groupsizeestimates ofthe incidental sightings wereestimated bythe observers whorecorded thesesightings. Although otherreports were received fromboat-based andshore-based observersthesehavebeen excluded fromthe database asthey were notaccompanied byadequate photographic verification. Results ªSuper-groupsº weredefined asgroups of20 ormore tightly-spaced individualhumpback whales eachestimated tobe within fivebody lengths oftheir nearest neighbour (thoughin reality mostªsuper-groupsº weremore tightly spaced thanthis)(Fig2).Inmost cases our observations ofªsuper-groupsº occurredinthe midst ofmore loosely-spaced aggregationsof less dense, smaller sub-groups whichwerewidely distributed overtheobserved range(Fig3). A total of22 dedicated sightingsofªsuper-groupsº weremade inthe southern Benguela region during thethree cruises (Table1),whilst afurther onewasrecorded duringtheoceano- graphic surveybythe ..    in2015. Atleast seven incidental sightingswerereported bythe public including threesightings madefromanaircraft oneach of16 October and25October 2015 andonesighting madefromanaircraft on4November 2015(Table 2).The locations of these primary sightings andthegeneral areaofthe incidental sightingsareshown inFig 1.Water depths associated withthededicated sightingsrangedbetween 32and 86m. The sightings of these ªsuper-groupsº allarise from alocalised coastalregionsome200nautical milesinextent within thehighly productive southernBenguela Upwelling Systemandranged fromtheColum- bine upwelling cellinthe north tothe Cape Peninsula upwelling cellinthe south. Positions of the ªsuper-groupsº didnot appear tobe related toany bathymetric featuresordepth intervals. Best estimates ofthe sizes ofthe dedicated cruisesightings ofªsuper-groupsº ofhumpback whales ranged between 20and 200individuals (Table1),although therewasoften considerable range inthese estimates madebydifferent observers, reflectingthedifficulty ofgroup sizeesti- mation ofsuch large groups. Despite aminimum independent estimateofgroup sizebeing the number ofphoto-identified individualsencountered inthe group, suchanestimate doesn't take thehigh group fluxinto consideration. Forexample, atleast 67individuals wereidentified from tailfluke colouration patternswithintheªsuper-groupº observationof28 October 2014, at which abest group sizeof60 individuals wasestimated atthe time ofsighting. Estimates of group sizesmade byincidental observersrangedbetween 50and 60whales pergroup, allof which werecarried outfrom aircraft atelevated observations andwhich mayconsequently have greater veracity. Onthewhole theªsuper-groupsº wereoften located withinlargeloose aggregations ofhumpback whalesspreadovertheentire visualareafrom themothership (esti- mated asan extended areaofbetween 10and 20square kilometres). Furthermore, theflux of animals intoandoutofªsuper-groupsº appearedhigh,withsingle animals, pairsandeven sub- groups ofup tofive individuals observedtojoin andleave ªsuper-groupsº duringobservations; some ofthese wereobserved totravel intowards theªsuper-groupº fromconsiderable distances. Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 6/ 18 At least 40individuals arediscernible inFig 2(image ofincidental observation number7 (Table 2)estimated bythe observer as60 individuals). Theproportion ofdiving whales to whales observable ator near thesurface isunknown, butgiven theobserved rapidand repeated divingbehaviour withinthesegroups, isbelieved tobe greater thanparity, sothat a group sizeestimate of60 individuals isconsidered aprobable underestimate. Generally, thebehaviour ofanimals withinallªsuper-groupsº appearedtobe focussed on feeding, withimmediate feedingbeingidentified byobservations ofsurface gapingandlung- ing and tight turning andrepeated strongvertical divingbehaviour whilstextended feeding behaviour wasapparent fromobservations ofdefecations (includingbrickredsolid faeces) and apungent (ªfishyº) odourofthe whale blows, asopposed tothe more normal oilyodour of blows found onbreeding grounds(KFpers. obs.). Therepeated strongvertical diving behaviour predominated inthese encounters, especiallyin2014. Underwater exhalationsand the formation ofbubble cloudsbyindividual whaleswerenoted inanumber ofthe ªsuper- groupsº andareclearly evident inFig 2.Bird (Cape gannet (   ) andtern species) and Cape furseal (     ) feedingwereassociated withsome super- groups. Ontwo occasions asingle finwhale (     )wasassociated with super-groups, andontwo occasions (06November 2014and30October 2015),southern right Fig 2.Aerial incidental observatio n(Observatio n7, Table 2)of aªsuper-gro upºencounte redsome 5km west ofCrayfish Factoryonthe west coast ofthe Cape Peninsula, SouthAfrica. Imagecourtesy ofJean Tresfon. doi:10.1371 /journal.pone. 0172002.g002 Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 7/ 18 whales (   )werepresent withinasuper-group. TheRHIB echo-sounders often noted preyaggregations eitheratthe bottom, mid-water oratthe surface withinthe water column inthe vicinity ofªsuper-groupsº althoughnoverification ofthese targets could be determined throughplankton towsduetothe tight spacing ofthe whales. Whilstsurface lunging wasobserved onoccasion, whalebehaviour withintheªsuper-groupsº oftenincluded a high incidence ofrepetitive fluke-updives,indicative offeeding atsome depth. Oncertain occasions whalescouldbetracked bythe RHIB echo-sounder inthe vicinity ofsuch prey aggregations atthe seafloor (Fig4).Both mantis shrimp (! "  # ) and euphausiids (  )wereobserved onthe surface withinsupergroups. Whilst ªsuper-groupsº asawhole moved duringobservations, thedirection ofwhale move- ments within thegroup wasnotnecessarily tiedtothe direction ofthe group movement, although insome cases(particularly inthe larger super-groups) theindividual's movementon the surface wasinthe general direction ofthe group movement. Inmost cases there wasinde- pendent turningandmilling ofindividuals withinthesuper-groups, althoughsynchronicity in diving andfeeding wassometimes evidentwithinsub-groups ofup tofive whales withinthe super-groups. ªSuper-groupsº wereobserved toform rapidly withsimilar rapiddissociation before forming innew areas reasonably closetothe previous area,presumably dependenton the concentrations offood resources inthe water column. Thedistribution ofªsuper-groupº aggregations in2015 appeared toshift southward overtheperiod 19October to12 November with nogroups beingseeninthe south atthe start andnogroups beingseeninthe north atthe end ofthis period respectively. Whilst largemature humpback whaleswereobserved insuper-groups, manyofthe observed individuals weresmall estimated atbetween 8and 10metres. Onlyonecalfwas sighted inthe over 30observations ofªsuper-groupsº (on5November 2015). Discussion Southern Hemisphere humpbackwhalesaregenerally understood tofeed atthe summer polar termini oftheir annual migrations, withthemajority ofevidence ofsuch feeding arisingfrom investigations ofstomach contents ofwhales examined onflensing platforms (bothin Fig 3.An aerial incidental observation ofaªsuper-gro upº(Observation 3Table 2,circled in backgro und)within awidely distributed loosely-spacedaggregation ofsmaller feeding sub-groups (foregro und)offDassen Island. doi:10.13 71/journal.pone .0172002.g003 Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 8/ 18 Antarctic regionandlower latitude migration corridorsandbreeding grounds) [3,9]. Although evidence forfeeding duringmidandlowlatitude migrations havebeen reported [41,42,44,56, 62, 65], thepredictable seasonalobservations ofnovel feeding behaviour bysuch large groups of Table 1.Dedicated observation sof ªsuper-gro upsºencounte redoncruises inthe southern Benguela in2011, 2014and2015. Observat ion Number Vessel DateGroupsize high Group size low Group size best Latitude (S) Longitude (E) Comment 1 RVAlgoa 12Novemb er 2011 38 2230 -32.616317.8566 2 RVAlgoa 12Novemb er 2011 22 2020 -32.6216617.8538 3 RVAlgoa 14Novemb er 2011 20 1020 -32.594018.03605 4 FRSEllen Khuzway o28 Octobe r 2014 70 5060 -33.393318.02833 5 FRSEllen Khuzway o29 Octobe r 2014 80 6070 -33.458518.02833 6 FRSEllen Khuzway o29 Octobe r 2014 200 150175 -33.45717.99283 7 FRSEllen Khuzway o01 Novemb er 2014  60 -33.472418.02393 8 FRSEllen Khuzway o01 Novemb er 2014  80 -33.333918.00527 9 FRSEllen Khuzway o02 Novemb er 2014  20 -33.37418.03868 Finwhale withingroup 10 FRSEllen Khuzway o02 Novemb er 2014 25 2020 -33.37418.05 Finwhale withingroup 11 FRSEllen Khuzway o05 Novemb er 2014 35 2530 -32.617217.83518 12 FRSEllen Khuzway o05 Novemb er 2014 50 3035 -32.542217.905 13 FRSEllen Khuzway o05 Novemb er 2014  30 -32.904717.781 Southernrightwhale within group 14 FRSEllen Khuzway o05 Novemb er 2014  20 -32.904717.781 15 FRSEllen Khuzway o06 Novemb er 2014  70 -32.888417.8016 16 FRSEllen Khuzway o06 Novemb er 2014  80 -32.888417.8016 17 FRSEllen Khuzway o07 Novemb er 2014 30 2025 -32.821717.77667 18 FRSEllen Khuzway o29 Octobe r 2015 60 4040 -33.410018.08138 19 FRSEllen Khuzway o30 Octobe r 2015 150 100120 -32.785617.78837 TwoSouther nright whales withingroup 20 FRSEllen Khuzway o31 Octobe r 2015 60 4050 -32.914417.83 21 FRSEllen Khuzway o05 Novemb er 2015 180 150150 -34.304318.356 22 FRSEllen Khuzway o06 Novemb er 2015 30 2020+ -34.07518.33604 23 RVAlgoa 31Octobe r 2015 35 2530 -32.874217.76917 ÐNot Recorded. doi:10.137 1/journal.pone. 0172002.t001 Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 9/ 18 feeding humpback whalesacrossatleast fiveyears (2011 to2015) reported herearethe first con- firmed records ofsuch large aggregation behaviourpatternswithinlowlatitude waters,andpos- sibly surpass thesizes offeeding groupsinboth northern andsouthern polarwaters. Neither Best  . [42] norBarendse  . [44] report ofgroup feeding behaviours intheir shore-based observations andassociated smallboatsurveys withinthisregion in1995 andbetween 2001and 2003 respectively. Onegroup of20 whales wasrecorded byBarendse  . [44], although the authors notedthatthisªinreality wasadynamic aggregation ofseveral smaller groupsº anda ªLarge, looselyassociated groupidentified as11 smaller groupsfromLandº: Suchobservations may have been precursors ofthe contemporary super-groups,butgiven thelimited spatial extent ofthe shore-based studiesandmore-limited sightingconditions fromshore, otherlarger groups couldhaveremained undetected. Despitefeedinghumpback whaleshaving been recorded byanumber ofauthors withinthesouthern Benguela region[6,9,42, 44], wepropose that theªsuper-groupº feedingphenomenon (astightly spaced largegroups ofwhales) isarela- tively recent behaviour exhibitedbythese whales. Whilstthedifficulty ofthe group sizeestima- tion isacknowledged, theestimates weremade inclose proximity tothe groups sothat the estimates areinfluenced onlythrough thewhale behavioural patternsratherthanenvironmental variables. Furthermore, althoughaerialplatforms mayprovide someelevated perspective, they do not increase theconfidence ofgroup sizeestimation asrepeated surfacing boutsareimpossi- ble tolink toparticular individuals. The novelty ofthe encountered behavioursistwofold, namelya)the formation oftightly- spaced largegroups ofhumpback whalesfeeding intensely withinconcentrated areas(an intensity whichappears toexceed thatobserved bytwo ofthe authors (MAMandKPF) inAnt- arctic waters), andb)the predictability ofthese feeding behaviours ofhumpback whaleswithin low latitude waters.Nowacek  's. [36] description ofhigh densities ofhumpback whales feeding onAntarctic krillinthe Western Antarctic Peninsula reportwhaledensities of5.1/ km 2 , an order ofmagnitude lessthan thegroup aggregations reportedhere.Furthermore, no such dense feeding aggregations havebeen reported elsewhere inlow ormid latitudes during Southern Hemisphere humpbackwhalemigrations. Indeed,aggregations ofwhales ofthis size have seldom beenreported inthe literature, withªlargeº groups oftennumbering inthe range of 10±20 orless [66±68]. Barendse  . [69] found thatthestudy areaisused during springandsummer byasmall component ofabout 500humpback whaleswhichtheythought werelikely tomigrate and breed within thetropical watersoffGabon inCentral Africa.However, Carvalho . [70] noted thatsome ofthe whales feeding inthe southern Benguela maybebreeding andcalving in an area thatisyet tobe identified. Whilstnoabundance estimateshaveyetbeen generated from thephoto-identification inthe 2011, 2014and2015 surveys, thenumbers ofwhales seen in both large ªsuper-groupsº andaggregations duringthelate October andearly November Table 2.Incidental observation sof ªsuper-gro upsºmade bythe public inthe southern Benguelain 2011, 2014and2015. Allofthese observatio ns were made ataltitude sexceeding 300m. Observat ionObserver Date PlatformGroupsize Locality 1 DavidHurwitz 19October 2015 Aerial 50 5n. miles westofVondeling Island 2 DavidHurwitz 19October 2015 Aerial 100 2n. miles NNW ofDassen Island 3 DavidHurwitz 19October 2015 Aerial 50 1n. mile west ofabove observat ion 4 DavidHurwitz 26October 2015 Aerial 50 JustN-NW ofDassen Island 5 DavidHurwitz 26October 2015 Aerial 50 JustN-NW ofDassen Island 6 DavidHurwitz 26October 2015 Aerial 50 JustN-NW ofDassen Island 7 JeanTresfon 4Novemb er2015 Aerial 60 5km west ofCray®s hFactory onthe west coast ofthe Cape Penins ula doi:10.137 1/journal.pone. 0172002.t002 Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 10/ 18 Fig 4.Photograph ofaRHIB sonar echogr amshowing thetrail ofadiving whaletothe seafloor withinthededicated observation ofthree whales onthe edge ofaªsuper- groupº at32.57Ê S;18.048ÊE on14November 2011(Observatio n 3 in Table 1). doi:10.1371 /journal.pone. 0172002.g004 Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 11/ 18 2014 and2015 surveys mustsuggest abrief temporary immigration intotheregion thatiswell in excess ofthe 500 individuals estimatedtooccur inthe region insummer byBarendse  . [69]. Preliminary resultsofsatellite tracking carriedoutin2014 showing thatsixofeight indi- viduals taggedwithinªsuper-groupsº migratedsouthwards tothe Southern Oceanbyearly summer (November), althoughfourofthese individuals firstmoved northforsome 100to150 n. miles before migrating south[71].However, whereasthereislittle current information on the migratory originofthe whales withintheªsuper-groupsº, thereissome evidence tosuggest that they maynotnecessarily beasuspended southward migrationofwhales fromthebreed- ing grounds offWest Africa asproposed byBest  . [42]. Thesouthward migrationrouteof humpback whalesfromtheWest Africa breeding groundsislikely further tothe west ofthe study area,asappears tobe the case forthe northwards migrationofthese animals basedon historical reports[6,8,61] and more recent scientific observations [44]showing littleevidence for northwards migratinganimalsinthe vicinity ofSaldanha Bay.Furthermore, themigration tracks oftwo humpback whalessatellite-tagged inthe breeding groundsoffGabon asthese animals travelled south-westwards alongtheWalvis Ridgewerewelltothe west ofour study domain duringtheirsouthward migration[27].Theabove andthecombination of(a) the lack of encountered calvesthatwould beexpected atthis time ofyear ifthe whales inªsuper- groupsº wereemanating fromthebreeding groundsandalthough Barendse etal. [69] and Findlay andBest [62]report ofobservations ofsingle calves fromtheregion atthis time of year, theexpected numbers shouldbefar higher basedonobservations ofcalves onthe east coast ofSouth Africa during thesouthward migration[72]orduring southward migrations off other continents [73];(b)that whale disentanglement effortsinthe region haveshown links through gear-type tothe south coastofthe country duringtheprevious months(MAMpers. obs.); and(c)the high incidence ofyellow diatom films(presumed tobe     )on the skin ofthe whales foundinthis area issuggestive ofarecent movement ofwhales froma cold water habitat (see[74]), andnotfrom thetropics. Webelieve theanimals encountered within ªsuper groupsº didnot migrate furthernorththanSouth African watersduring thepre- ceding austral winters andpossibly comprises anon-breeding migrationofyoung animals from Antarctic watersandtheimmigration ofyoung non-breeding animalsfromBreeding Stock Cduring theirsouthern migration. Itshould benoted thatDawbin [2]found thatyoun- ger humpback whalescouldbecaught onmid-latitude groundsduringthemid-winter season when mature whaleswereonthe more northerly breedinggrounds. Furthermore, Olsen[6] reported thatyoung malehumpback whaleswereencountered inthe southern Benguela dur- ing thesummer months. The identification ofthe prey species eliciting ªsuper-groupº feedingbehaviours remains undetermined. Humpbackwhaleshavebeen recorded feedingonanumber ofprey species within theregion. Although Barendse . [44] noted thatthe26observations offeeding behaviour andªapparent feedingbehaviourº (includingdefecations) duringspringandsum- mer inthe southern Benguela upwelling regionbetween 2001and2003 provide evidence that the region mayfunction asan important feedingareaforthese whales, theynote thatlittle evi- dence ofprey species couldbeobtained, apartfrompossible euphausiid exoskeleton remains and ahyperiid amphipod foundwithin collected faecalsamples. Humpback whaleshavebeen recorded feedinginsummer oncopepods (ªrodaateº inNorwegian fromOlsen (1914) [6] translated in[44] andªherringsº [6],mantis shrimp (!.#  ) [62],euphausiids (. ) andamphipods [42,44]including hyperiidamphipod ($#    % [44]in the southern Benguela, whileMatthews [9]found thatofatotal ofeight stomachs ofhumpback whales examined atSaldanha Bayduring winter, fish(ª?Clupeiodsº andªapasty mash offish scales andbonesº) werepresent intwo; oneinJune, 1926andoneinSeptember, 1926.Best   . [42] recorded apparentfeedingbehaviour byhumpback whaleson10occasions over Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 12/ 18 their 38days ofshore-based observations andtheobservation ofproduction offaeces indica- tive ofrecent feeding on7occasions duringsmallboatapproaches. Theircollected faecalsam- ples contained euphausiid remains(ªpossibly . º) on2occasions andamphipods on another. Theavailable evidence pointstoconsiderable opportunistic feedingbehaviour ona wide range ofprey species fromtheregion. Cleardifferences inthe acoustic scattering ofprey patches (measured at200 kHz and38kHz from the    ) werehighly sugges- tive ofdense aggregations ofeuphausiids in2015. Inall the study years, theRHIB echo-sound- ers identified densepreyaggregations almostonthe seafloor under theªsuper-groupsº although thepossibility ofthese aggregations beingpreyspecies otherthan.  cannot be ruled out.During the2014 observations ofthe super-groups, aco-occurring predatoryfish species, Snoek($  ),werecaught atDassen Islandwithfullstomach contents ofsto- matopod mantisshrimps (!.#  % (CWpers.obs.). Although nosuch catches were made in2015, mantis shrimps wereobserved inassociation withhumpback whaleson06 November 2015offHout Bayand amass stranding ofmantis shrimps occurred inthis bay within 14days after thecruise. Furthermore, theonly prey recorded inthe stomach ofa6.4m humpback whaleentangled inrock lobster fishing gearin1990 (see[62]) wasyoung adultsof this stomatopod specieswhichGriffiths andBlaine [75]found tobe the only stomatopod spe- cies occurring tothe west ofCape Point, withthedistribution offthe west coast corresponding to terrigenous mudbeds. Although predominantly abenthic species, theoccurrence of!. #   instomachs ofother pelagic predators suchasAfrican penguins, cormorants and snoek, [75]suggests thatpelagic swarming behaviour makesthespecies available topreda- tors including baleenwhales [62]. The recent observations ofthis novel andintense feeding behaviour isof particular interest in light ofrapidly recovering humpback populations onboth theeast andwest coasts ofsouth- ern Africa andinthe associated SouthernOceanregion [14,43,76,77] with theobservation of this behaviour potentially arisingfromthefollowing potentialfourscenarios: 1. alterations inprey availability leadingtoanovel feeding strategy; 2. increasing humpback whaleabundance intensifying pressureonprey availability elsewhere and aconsequential switchinfeeding strategies orareas; 3. arestoration ofapreviously unobserved feedingstrategy asthe population abundances re- establish; or 4. an increase inthe probability ofdetection ofªsuper-groupº behaviourasthe population abundance increases. Although Verheye . [78] describe long-term changesinneritic zooplankton communi- ties inthe southern Benguela subsystems withashift indomination fromlargetosmaller spe- cies, wecurrently remainuncertain asto the mechanism drivingthisnovel low-latitude behaviour inSouthern Hemisphere humpbackwhales.GiventhatOlsen [6]reported ofasum- mer incidence ofjuvenile humpback whalesinthe region priorto1914 (albeit notatsuch den- sities) theobserved behaviour maywellbearestoration ofapreviously unobserved feeding strategy asthe population abundances re-establish. Furthermore, asnoted byOwen  . [55] the extent offeeding onmigration couldinfluence theextent towhich humpback whalesrely on the Antarctic ecosystem eachyear. Despite theunknown causeofthis recent behaviour, we postulate thatthearea hasdeveloped /is developing intoanimportant seasonalhumpback whale feeding ground thatattracts significant immigration intotheregion inthe late austral spring /early summer. Whilsthumpback whalesinthe region areclearly feeding opportunisti- cally onarange ofprey species, theconcentrations, identityandsizeofthe prey eliciting Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 13/ 18 ªsuper-groupº feedingbehaviours remainunknown andrequire investigation aswe believe this willbeaprimary stepinidentifying thesystem changes (including whalepopulation increases) thathave resulted inthese observations. Owen . [55] suggest thatprey typemay be important ininfluencing theextent offeeding onmigration. Furtherrequired investigations include a)the origin ofthis immigration intotheregion asthe lack ofcalves inthe ªsuper- groupº encounters andthehigh incidence ofcold water diatoms onthe whales isunexpected at this time ofyear, b)observations oftagged ormarked individuals withinªsuper-groupsº to obtain someinformation onthe surfacing timetodive time ratios (along withaerial observa- tions) forthe modelling ofgroup sizeestimation. This novel, predictable (atleast overthe2014±2015 period)andaccessible feedingbehav- iour within lowlatitudes provides considerable opportunityforfurther investigation ofSouth- ern Hemisphere humpbackfeedingaggregations andbehaviours outsideofthe relatively inaccessible ice-edgeregionofthe Southern Ocean.Futureareasofinvestigation should include identifying migrationlinksandthepopulation identityofparticipating whales,multi- scale examination ofthe feeding ecology ofthese humpback ªsuper-groupsº (e.g.[79]) and modelling ofthe energetic advantages ofthe migration suspension. Supporting information S1 Dataset. Dedicated andincidental sightingsofªsuper-groupsº ofhumpback whales reported inthis study. (DOCX) Acknowledgmen ts We thank observers onthe cruises whoprovided sightingsdata,including GwenithPenry, Tess Gridley, KirstyVenter, JeanPurdon, thelate Nompilo Radebe,Noloyiso Gwanqa,Laur- enne Snyders, KuhleHlati,Oyena Masiko, LelethuZepe,Miranda Sofika,Bridget James, Taryn Morris, LiezeSwart, Zyanda Nkumanda, LihleSikolo, KyleLloyd, Sizwekazi Yapi, Abongile Sobekwa, Thubelihle ThwalaandMike Makhale. Weareindebted tothe public for reports ofªsuper-groupsº includingJeanTresfon forsupplying imagery.Wethank thecap- tains andcrews ofthe        fortheir dedication tothis project during the2011, 2014and2015 cruises. Author Contributions Conceptualization: KPFMAM SMS. Data curation: MAMAKMT SMS DH. Funding acquisition: MAMSMS. Investigation: KPFSKDEC MAM PGHK SAMAKJBOAVF MTSMS DHCGW. Methodology: KPFMAM AKSMS. Project administration: KPFMAM SMS. Resources: MAMSMSDH. Supervision: KPFJB. Validation: KPFSMS. Visualization: KPF. Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 14/ 18 Writing ±original draft:KPF. Writing ±review &editing: KPFSKDEC MAM AKJBOAVF MTSMS DHCGW. References 1. Dawbin WH.TheMigration sof Humpba ckWhales whichPasstheNew Zealand Coast.1956.p.147±96. 2. Dawbin WH.Theseasonal migratorycycle ofhumpback whales.1966.p.145±70 . 3. Chittlebor oughRG.Dynami csoftwo popula tionsofthe humpbac kwhale, Megaptera novaeangliae (Borowski ).Australian JournalofMarine andFreshwater Research.1965;16(1):33±1 28. 4. Rasmussen K,Palacio sDM, Calambok idisJ,Saborio MT,Dalla Rosa L,Secchi ER,etal. Southe rn Hemisph erehumpback whaleswintering offCentral America: insightsfromwater temperatu reinto the longest mammalian migration.Biology Letters. 2007;3(3):302±5 .doi: 10.1098 /rsbl.2007.00 67PMID: 17412669 5. Risting S.Knolhvalen NorskFiskeritide nde.1912; 11:437±49 . 6. Olsen O.Hvaler oghvalangst iSydafrika. BergensMuseum Arbok1914±1 9151914; 5:1±56. 7. Harmer SF.History ofwhaling. Proceeding sof the Linnean Society, London(1927±28) 1928:51±9. 8. Harmer SF.Southe rnWhaling. Proceeding sof the Linnean Society, London(1929±30). 1931;42:85±1 63. 9. Matthews LH.The humpback whaleMegapter anodosa. Discover yReports. 1938;17:7±92. 10. Mackintos hNA. The southern stocksofwhalebon ewhales. 1942.p.197±30 0. 11. Omura H.Areview ofpelagic whaling operations inthe Antarcti cbased onthe effort andcatch datain 10Ê squares oflatitude andlongitude ScientificReports ofthe Whales Researc hInstitute, Tokyo.1973; 25:105±203 . 12. Findlay KP.Areview ofhumpback whalecatches bymodern whalingoperations inthe Southern Hemi- sphere. Memoirsofthe Queensland Museum.2002; 47(2):411± 20. 13. Kasamatsu F,Matsuoka K,Hakama daT.Interspecifi crelationships indensity amongthewhale com- munity inthe Antarcti c.Polar Biology. 2000;23(7):466± 73. 14. Branch TA.Hump backwhale abundanc esouth of60 degrees Sfrom three complete circumpolarsets of surveys .Journal ofCetacean Researchand Managemen t.2011;(Sp. Iss.3):53± 69. 15. Rayner GW.Whale marking: progressand results toDecember 1939.Discovery Reports1940;19: 245±84. 16. Brown SG.Whale marking: ashort review. In:Angel M,editor. AVoyage ofDiscover y:George Deacon 70th Annivers aryVolume Oxford:Pergamon Press;1977.p.569±81. 17. Stone GS,Florezgo nzalezL,Katona S.Whale Migratio nRecord. Nature.1990;346(6286) :705-. 18. RockJ,Pastene LA,Kaufman G,Forestell P,Matsuoka K,Allen J.Anote onEast Australia GroupV Stock humpback whalemovement betweenfeedingandbreeding areasbased onphoto-identi fication Journal ofCetace anResearch andManagemen t2006; 8:301±5. 19. Zerbini AN,Andriolo A,Heide-Jor gensenMP,Moreira SC,Pizzorno JL,Maia YG,etal. Migration and summer destinatio nsofhumpbac kwhales (Megapte ranovaean gliae)inthe western SouthAtlantic Ocean. Journal ofCetacean ResearchandManagemen t.2011;(Sp. Iss.3):113±8. 20. DallaRosa L,Secchi ER,Maia YG,Zerbini AN,Heide-Jor gensenM-P.Movements ofsatellite-m oni- tored humpback whalesontheir feeding groundalongtheAntarctic Peninsula. PolarBiology. 2008;31 (7):771±81 . 21. FelixF,Guzman HM.Satellite Tracking andSighting DataAnalyses ofSoutheas tPacific Humpback Whales (Megaptera novaeangliae):Isthe Migrator yRoute Coastal orOceanic? AquaticMamma ls. 2014; 40(4):329± 40. 22. Garrigue C,Clapham PJ,Geyer Y,Kennedy AS,Zerbini AN.Satellite trackingrevealsnovelmigratory patterns andtheimportan ceofseamounts forendangered SouthPacific humpback whales.RoyalSoci- ety open science. 2015;2(11):1504 89-.doi:10.1098 /rsos.15048 9PMID: 267160 06 23. Garrigue C,Zerbini AN,Geyer Y,Heide-Jor gensenM-P,Hanaoka W,Clapham P.Movem entsofsatel- lite-monit oredhumpbac kwhales fromNewCaledoni a.Journal ofMammalogy .2010; 91(1):109± 15. 24. Hauser N,Zerbini AN,Geyer Y,Heide-Jor gensenM-P,Clapham P.Movem entsofsatellite- monitored humpback whales,Megaptera novaeangliae,fromtheCook Islands. MarineMamma lScience. 2010; 26(3):679± 85. 25. Zerbini AN,Andriolo A,Heide-Jor gensenM-P,Pizzorno JL,Maia YG,VanBlarico mGR, etal. Satellite- monitore dmovements ofhumpback whalesMegapter anovaeangliae inthe southwest AtlanticOcean. Marine Ecology Progress Series.2006; 313:295±304 . Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 15/ 18 26. IWC(Internationa lWhaling Commission ).Annex G.Report ofthe Sub-Co mmittee oncomprehe nsive assessmen tof Southe rnHemisp herehumpback whales.InternationalWhaling Commission ,1998. 27. Rosenbaum HC,Maxwell SM,Kershaw F,Mate B.Long-Range MovementofHumpba ckWhales and Their Overlap withAnthropoge nicActivity inthe South Atlantic Ocean.Conservation Biology.2014;28 (2):604±15 .doi: 10.1111 /cobi.1222 5PMID: 24495092 28. Morch JA.Onthe natural historyofwhalebon ewhales. Proceedings ofThe Zoologica lSociety ofLon- don. 1911; 1911:661±76. 29. Kellogg R.What isknown ofthe migrations ofsome ofthe whalebon ewhales. Reports ofthe Smithso- nian Institution. 1929;1928:46 7±94. 30. Banniste rJL, Baker AdC.Observations onfood andfeeding habitsofbaleen whales offDurban. Norsk Hvalfangs ttidende.1967;56(4):78±8 2. 31. Lockyer C.All creatur esgreat andsmaller: astudy incetacean lifehistory energetics .Journal ofthe Marine Biologic alAssoc iationofthe United Kingdom. 2007;87(4):1035 ±45. 32. Lockyer CH,Brown SG.Themigration ofwhales .Society forExperi mental Biology SeminarSeries. 1981:105±3 7. 33. LawsRM.Thesignificance ofvertebrates inthe Antarctic marineecosystem. In:Lano GA,editor. Adap- tation within Antarctic ecosystems Washington,DC: Smithsoni anInstitution,; 1977.p.411±38 . 34. Friedlaende rAS, Halpin PN,Qian SS,Lawson GL,Wiebe PH,Thiele D,etal. Whale distribution inrela- tion toprey abundance andoceanog raphicprocesses inshelf waters ofthe Western AntarcticPenin- sula. Marine Ecology Progress Series.2006;317:297 ±310. 35. Murase H,Matsu okaK,Ichii T,Nishiwaki S.Relationshi pbetween thedistribution ofeuphau siidsand baleen whales inthe Antarcti c(35 degrees E±145degrees W).Polar Biology. 2002;25:135± 45. 36. Nowacek DP,Friedlaende rAS, Halpin PN,Hazen EL,Johnston DW,Read AJ,etal. Super-Ag grega- tions ofKrill and Humpba ckWhales inWilhelmin aBay, Antarcti cPeninsula. PlosOne. 2011; 6(4). 37. Corkeron PJ,Connor RC.Why dobaleen whales migrate? MarineMammal Science.1999;15 (4):1228±4 5. 38. Norris KS.Some observatio nson the migratio nand orientation ofmarine mammals .In: Storm RM,edi- tor. Animal orientationand migration Corvallis,OR.:Oregon StateUniversity Press1967.p.101±25 39. LawsRM.TheEcology ofthe Southern Ocean.American Scientist.1985;73(1):26±4 0. 40. Tynan CT.Ecological importanceofthe Southern Boundary ofthe Antarcti cCircumpolar Current. Nature. 1998;392(6677) :708±10. 41. Hucke-Gae teR, Haro D,Torres-Flore zJP, Montecinos Y,Viddi F,Bedrinana -RomanoL,etal. Ahistori- cal feeding groundforhumpback whalesinthe eastern SouthPacific revisited: thecase ofnorthern Pat- agonia, Chile.Aquatic Conserv ation-Marine andFreshwate rEcosyste ms.2013; 23(6):858± 67. 42. BestPB,Sekiguch iK, Findlay KP.ASusend edMigratio nof humpback whalesMegapter anovaean gliae on the west coast ofSouth Africa. Marine Ecology Progress Series.1995; 118(1±3):1 ±12. 43. Barendse J,Best PB,Thornto nM, Elwen SH,Rosenba umHC, Carvalho I,et al. Transit stationordesti- nation? Attendancepatterns, movements andabundanc eestimate ofhumpback whalesoffwest South Africa fromphotographi cand genotypic matching. AfricanJournal ofMarine Science. 2011;33(3):353± 73. 44. Barendse J,Best PB,Thornto nM, Pomilla C,Carvalho I,Rosenbaum HC.Migratio nredefined? Sea- sonality, movemen tsand group composition ofhumpback whalesMegapter anovaeangliae offthe west coast ofSouth Africa. African Journal ofMarine Science. 2010;32(1):1±22 . 45. Minton G,Collins T,Findlay K,Baldwin R.Cetacean distribution inthe coastal watersofthe Sultanate of Oman. Journal ofCetacean Researchand Managemen t.2010; 11(3):301± 13. 46. Alves LCPdS, Andriolo A,Zerbini AN,Pizzorno JLA.andClapham PJ.Record offeeding byhumpback whales (Megapte ranovaean gliae)intropical watersoffBrazil. Marine Mammal Science.2009;25 (2):416±9. 47. Acevedo J,Plana J,Aguayo-Lobo A,Pastene LA.Surface feedingbehaviorof humpback whalesinthe Magella nStrait. Revista DeBiologia MarinaYOceanografi a.2011; 46(3):483± 90. 48. Danielwicz D,Tavares M,Benites Moreno I,Ott PH, Campo sTrigo C.Evidence offeeding bythe hump- back whale (Megap teranovaean gliae)inmid-lati tudewaters ofthe western SouthAtlantic.JMBA2 Bio- diversity Records. 2009;6348:1±3. 49. Baraff LS,Clapham PJand Matilla DK.Feeding behaviorof ahumpback whaleinlow-latitude waters. Marine Mamma lScience. 1991;7:197±202. 50. Gendron D,and Urban J.Evidence offeeding byhumpback whales(Megaptera novaeangliae)inthe Baja California breedingground,Mexico. MarineMamma lScience. 1993;9:76±81. Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 16/ 18 51. Bortolotto GA,Kolesniko vasCKM, Freire AS,and Sim}es-Lopes PC.Young humpbac kwhale Mega- ptera novaean gliaefeeding inSanta Catarina coastalwaters, Southern Brazil,andaship strike report. Marine BiodiversityRecords .2016 9(29). 52. GillPC, Evans KJ,Wapstra H.Feeding byhumpback whalesinTasmanian waters.Records ofthe Queen Victoria Museum .1998; 107:1±5. 53. Stockin KA,and Burgess EA.Opportuni sticfeeding ofan adult humpback whale(Megaptera novaean- gliae) migrating alongthecoast ofSoutheastern Queensland,Australia. AquaticMammals 2005; 31:120±3. 54. Acevedo J,Rasmuss enK,Felix F,Castro C,Llano M,Secchi E,etal. Migratory destinationsofhump- back whales fromtheMagella nStrait feeding ground, SoutheastPacific. MarineMamma lScience. 2007; 23(2):453± 63. 55. Owen K,Warre nJD, Noad MJ,Donnelly D,Goldizen AW,Dunlop RA.Effect ofprey typeonthe fine- scale feeding behaviourofmigratin geast Austral ianhumpback whales.MarineEcology Progress Series. 2015;541:231±44. 56. Stamation KA,Croft DB,Shaughnes syPD. andWaples KA.Observation sof humpback whales(Mega- ptera novaean gliae)feeding duringtheirsouthward migrationalongthecoast ofsoutheas ternNew South Wales, Australia: identification ofapossible supplemen talfeeding ground.Aquat Mamm 2007; 33:165±74. 57. Moore SE,Waite JM,Friday NA,Honkalehto T.Cetacean distribution andrelative abundance onthe centraleas ternandthesoutheasten BeringSeashelf withreferenc eto oceanog raphicdomains. Prog- ress inOceano graphy. 2002;55:249±61. 58. Hutchings L,van derLingen CD,Shann onLJ, Crawford RJM,Verheye HMS,Bartholom aeCH, etal. The Benguela Current:anecosystem offour componen ts.Progress inOceano graphy. 2009;;83::15± 32. 59. Shannon LV.The Benguela Ecosystem1.Evolution ofthe Benguela, PhysicalFeatures andprocesses . Oceanograp hyand Marine Biology. 1985;23:105-&. 60. Huggett JA,Verhey eHM, Escribano R,Fairweath erT. Copepod biomass,sizecompos itionandpro- duction inthe Southern Benguela: spatio-temporalpatterns ofvariation, andcomparison withother eastern boundaryupwelling systems.Progress inOceanograp hy2009; 83:197±20 7. 61. Hinton MAC.Report onthe papers leftbythe late Major Barrett-H amilton,relatingtothe whales of South Georgia. LondonCrownAgents forthe Colonies 62. Findlay KP,Best PB.Summer incidence sof humpback whalesonthe west coast ofSouth Africa. South African Journal ofMarine Science- Suid-Afrika anseTydskrif VirSeeweten skap.1995; 15:279±82 . 63. Townsend CH.The distribution ofcertain whales asshown bylogbook recordsofAmerican whale ships. Zoologica NY1935; 19:1±50 . 64. BestPB,Shaug hnessy PD.Anindepend entaccount ofCaptain Benjamin Morrell'ssealingvoyage to the south-wes tcoast ofAfrica inthe Antarcti c,1828/29 .Fisheries BulletinofSouth Africa. 1979;12:1± 19. 65. Acevedo J,Mora C,Aguayo-Lobo A.Sex-related sitefidelity ofhumpback whales(Megaptera novaean- gliae) tothe Fueguia nArchipelago feedingarea,Chile. Marine Mamma lScience. 2014;30(2):433± 44. 66. Tyack P,Whitehea dH. Male competition inlarge groups ofwintering humpback whales.Behaviour. 1983; 83:132± 54. 67. Whitehea dH. Structur eand stability ofhumpback whalegroups offNewfou ndlandCanadian Journalof Zoology. 1983;61:1391 ±7. 68. Whitehea dH, Carlson C.Social behavio uroffeeding finbackwhalesoffNewfou ndland: comparison s with thesympatric humpback whale.Canadian JournalofZoology. 1988;66:217±21. 69. Barendse J,Best PB,Carvalho I,Pomilla C.Mother knowsbest:occurrenc eand associations of resighted humpbackwhales suggest maternally derivedfidelitytoaSouthern Hemisphere coastalfeed- ing ground. PLoSOne.2013; 8(12):e812 38.Epub 2013/12/19. PubMedCentralPMCID: PMC385 7176. doi: 10.1371 /journal.pon e.0081238PMID:24349047 70. Carvalho I,Loo J,Collins T,Barends eJ, Pomilla C,Leslie MS,etal. Does temporal andspatial segrega - tion explain thecomplex population structureofhumpback whalesonthe coast ofWest Africa? Marine Biology. 2014;161(4):805 ±19. 71. IWC(Internationa lWhaling Commission ).Annex H:Report ofthe Sub-Commit teeonOther Southern Hemisph ereWhale Stocks. Reportofthe Scientifi cCommitte e,San Diego, CA,USA,. 201622May-3 June 2015. Report No. 72. Findlay KP,Best PB.Estimates ofthe numbers ofhumpback whalesobserved migratingpastCape Vidal, South Africa, 1988±1 991.Marine Mamma lScience. 1996;12(3):354± 70. Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 17/ 18 73. Pitman RL,Totterde llJA, Fearnbac hH, Ballanc eLT, Durban JW,Kemps H.Whale killers:Prevalenc e and ecologica limplication sof killer whale predation onhumpback whalecalves offWestern Australia. Marine Mamma lScience. 2015;31(2):629± 57. 74. Nemoto T,Best PB,Ishimaru K,Takano H.Diatom filmsonwhales inSouth African waters. Scientific Reports ofthe Whales Research InstituteTokyo.1980:97±103. 75. Griffiths CL,Blaine MJ.Distribution ,population structureandbiology ofstomatopod crustaceaoffthe west coast ofSouth Africa. SouthAfrican Journal ofmarine Science. 1988;7:45±50. 76. Findlay K,Meyer M,Elwen S,Kotze D,Johnson R,Truter P,etal. Distribution andabundance ofhump- back whales, Megapter anovaeangliae, offthe coast ofMozambique, 2003.Journal ofCetace an Research andManagem entSpecial Issue.2011;3:163±7 4. 77. Findlay KP,Best PB,Meyer MA.Migratio nsofhumpback whalespastCape Vidal, South Africa, andan estimate ofthe population increaserate(1988±2 002).African Journal ofMarine Science. 2011;33 (3):375±92 . 78. Verheye HM,Lamont T,Huggett JA,Kreiner A,Hampton D.Plankton productivityofthe Benguela Cur- rent Large Marine Ecosyste m(BCLME) .Environm entalDevelopm ent[Internet]. 2015. 79. Hazen EL,Friedlaende rAS, Thompson MA,Ware CR,Weinri chMT, Halpin PN,Wiley DN.Fine-scale prey aggregation sand foraging ecologyofhumpback whalesMegapter anovaean gliae.Marine Ecology Progress Series2009;395:75±89. Novel low-latitu defeeding byhumpback whalesinthe Benguela Upwelling System PLOS ONE|DOI:10.137 1/journal.po ne.0172002March 1,2017 18/ 18