8 pages

A junk-food hypothesis for gannets feeding on fishery waste

of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
World-wide fisheries generate large volumes of fisheries waste and it is often assumed that this additional food is beneficial for populations of marine top-predators. We challenge this concept via a detailed study of foraging Cape gannets Morus
  A junk-food hypothesis for gannets feedingon fishery waste David Gre´millet 1,2, * , Lorien Pichegru 1,2 , Gre´goire Kuntz 1 , Anthony G. Woakes 3 ,Sarah Wilkinson 4 , Robert J. M. Crawford 5 and Peter G. Ryan 2 1 Centre National de la Recherche Scientifique, Centre d’Ecologie Fonctionelle et Evolutive, 1919 Rue de Mende, F-34293 Montpellier cedex 5, France 2 DST/NRF Centre of Excellence, Percy FitzPatrick Institute, University of Cape Town,Rondebosch 7701, Republic of South Africa 3 School of Bioscience, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK  4 CapFish, PO Box 50035, Waterfront, Cape Town 8002, Republic of South Africa 5 Department of Environmental Affairs and Tourism, Marine and Coastal Management, Private Bag X2,Rogge Bay 8012, Republic of South Africa Worldwidefisheriesgeneratelargevolumesoffisherywasteanditisoftenassumedthatthisadditionalfoodis beneficial to populations of marine top-predators. We challenge this concept via a detailed study of foraging Cape gannets  Morus capensis  and of their feeding environment in the Benguela upwelling zone.The natural prey of Cape gannets (pelagic fishes) is depleted and birds now feed extensively on fisherywastes. These are beneficial to non-breeding birds, which show reduced feeding effort and high survival.By contrast, breeding gannets double their diving effort in an attempt to provision their chickspredominantly with high-quality, live pelagic fishes. Owing to a scarcity of this resource, they fail and mostchicks die. Our study supports the junk-food hypothesis for Cape gannets since it showsthat non-breedingbirds can survive when complementing their diet with fishery wastes, but that they struggle to reproduceif live prey is scarce. This is due to the negative impact of low-quality fishery wastes on the growth patternsof gannet chicks. Marine management policies should not assume that fishery waste is generallybeneficial to scavenging seabirds and that an abundance of this artificial resource will automatically inflatetheir populations. Keywords:  biotelemetry; dispersal; fishery discard management; foraging behaviour; industrial fisheries;scavenger 1. INTRODUCTION Human fisheries substantially perturb marine ecosys-tems through destruction of marine habitats, removal of organisms from higher trophic levels, accidentalby-catch of non-target species and dumping of fisherywastes ( Jennings  et al  . 2001). Fishery wastes includeoffal generated when processing fishes at sea, as well asdiscarded undersized fishes and non-target species. Theamount of fishery waste produced is substantial:approximately 7.3 million tonnes of discards arereturned to the sea annually by worldwide fisheries(Kelleher 2005). This additional food source affectsfood web structures, as it favours scavengers throughoutthe water column and on the seabed (Catchpole  et al  .2006). For some marine top-predators such as seabirdsand marine mammals, fishery wastes provide analternative food source; they can forage on both liveprey and fishery wastes, eventually favouring the latterwhen the former becomes scarce (Votier  et al  . 2004).Some seabird populations benefit greatly from currentdumping practices and are artificially inflated (Garthe et al  . 1996; Furness 2003). It is therefore often assumed that fishery waste is beneficial to seabirds (Garthe &Hu¨ppop 1994; Tasker  et al  . 2000; Montevecchi 2002). We challenge this concept through a detailed study of the year-round foraging behaviour of an avian top-predator, the Cape gannet (  Morus capensis ), which feedson live prey and fishery wastes in the Benguelaupwelling ecosystem.The Benguela is one of the four major upwelling zonesof the world’s oceans, located along the Atlantic coasts of Namibia and South Africa. Its productive waters aretraditionally home to vast biomasses of pelagic fishessuch as anchovy ( Engraulis encrasicolus ) and sardines( Sardinops sagax ), which support a large community of predatory fishes, seabirds and marine mammals(Shannon 1985). Anchovies, sardines and predatoryfishes such as hake (  Merluccius  spp.) are also targeted byhuman fisheries (Griffiths  et al  . 2004). The SouthAfrican hake fishery currently produces some 52 500tonnes of waste per year in the form of discards and offal(Walmsley  et al  . 2007).With approximately 500 000 breeding individuals,weighing on average 2.6 kg, Cape gannets are majoravian predators in the Benguela (Crawford 2005). Theyfeed primarily by plunge-diving on pelagic fishes withinthe first 20 m of the water column (Crawford 2005), butalso gather behind hake trawlers to feed on fishery wastes(Abrams 1983; Ryan & Moloney 1988). Such food has Proc. R. Soc. B doi:10.1098/rspb.2007.1763 Published online * Author for correspondence ( Received   22 December 2007  Accepted   28 January 2008 1  This journal is q 2008 The Royal Society  only half the calorific value of Cape gannets’ natural prey(Batchelor & Ross 1984). In this context, the junk-foodhypothesis (JFH) posits that seabirds feeding on prey of low energy and nutrient content have reduced reproduc-tivesuccess(Piatt&Anderson1996),becausesuchadietaffects the growth patterns and the cognitive abilities of their offspring (Kitaysky  et al  . 2005; Wanless  et al  . 2005).The JFH has so far been tested for seabirds and marinemammals facing natural changes in diet quality, forinstance after an ecosystem shift (Rosen & Trites 2000;Litzow  et al  . 2002; Jodice  et al  . 2006). Nevertheless, itmight also apply to marine predators feeding onenergetically poor fishery wastes (Pichegru  et al  . 2007).Furthermore, previous tests of the JFH addressed eitherthe fate of non-breeding adults or that of their youngduring the reproductive phase, but did not study onespecies on a year-round basis.Here we examine the JFH during the reproductive andnon-reproductive phases of a seabird foraging on live preyand/or fishery wastes. We used year-round, paralleldatasets of hake trawler distribution, Cape gannet diet,Cape gannet movements and foraging behaviour recordedby bird-borne miniaturized data loggers, as well asassessments of Cape gannet breeding success. Morespecifically, we test the prediction that abundant fisherywaste is beneficial to non-breeding birds, but not to birdsraising chicks. 2. MATERIAL AND METHODS All measurements were performed in the southernBenguela upwelling zone under permits issued by SouthAfrican National Parks. According to the Cape gannetbreeding phenology (Crawford 2005), two main timeperiods were considered. (i) The breeding phase betweenmid-September and the end of February. The first twomonths of this period are spent mating and incubating,while the following 3.5 months are spent raising thechick. (ii) The non-breeding period between March andmid-September.( a )  Fishing activities The positions of all trawls performed by commercial vesselsoperating in the southern Benguela in 2005 was compiledfrom logbooks by Marine and Coastal Management,Department of Environmental affairs and Tourism andmapped on a 20 0 ! 20 0 (approx. 37 ! 30 km) square grid.Recent investigations by Rademeyer & Butterworth (2006)showed a strong correlation between trawling effort andsuccess in this fishing industry, hence trawling effort is areliable index of offal production.(  b )  Seabird diet  Diet data were collected monthly from Cape gannets atMalgas Island (33 8 03 0 S, 17 8 55 0 E) throughout 2005. Breed-ing and non-breeding gannets are present at the islandthroughout the year. Each month, an average of 37 birds(range 10–50)was caught randomlyon the edge of the colonyas they returned from the sea, upon which they regurgitatedtheir stomach contents. Prey items were identified to thespecies level and allocated to three categories: small pelagics(anchovies and sardines); fishery wastes (more or lessexclusively hake); and other prey items. Fishery waste wasidentified as entrails or sections of hake which would havebeentoolargetobeswallowedalivebygannets.Wecalculatedtherelativeproportionofeachcategorybymass,aswellastheaverage calorific value of the diet for each month in 2005. Tothis end, we used calorific values (Batchelor & Ross 1984) of 4.07 kJ g K 1 for fishery wastes, 8.59 kJ g K 1 for sardines,6.74 kJ g K 1 for anchovies and 6.20 kJ g K 1 for sauries( Scomberesox saurus ).( c )  Seabird behaviour  Cape gannets raising large chicks were caught in January2005 at Malgas Island where approximately 20% of theirworld population breeds (Crawford 2005).(i) Seven birds were fitted with global location sensors(GLSs; GeoLT: Earth and Ocean Technologies,Germany; 14 mm in diameter  ! 45 mm long, 8.2 g,approx. 0.3% of the bird’s body mass) attached to aDarvic leg ring. The loggers were set to record lightintensity every 30 s for up to 10 months. Lightresolution was 4.5% to less than 0.025% of reading(down to approx. 0.001 lux), depending on the level.Therecordingswereusedtoestimatethepositionofthebirds ( G 170 km; Phillips  et al  . 2004) at sunrise andsunset throughout the recording period after Wilson et al  . (1992) using the software M ULTI T RACE  Geoloca-tion (Jensen Software Systems, Laboe, Germany).Kernel analyses were subsequently performed to mapthe distribution of the birds. Ninety per cent concavepolygons were also used under R  ANGES 6 (Anatrack,Wareham, UK) to investigate spatial overlap betweenindividual home ranges as well as between monthlyhome ranges within individual home ranges.(ii) Fourteen birds were fitted with heart rate and depthdata loggers (HRDDLs; Woakes  et al  . 1995), whichmeasure 60 ! 24 ! 7 mm and weigh 20 g, approxi-mately 0.7% of the bird’s body mass. The HRDDLswere programmed to record depth every second andheart rate every 2 s, either continuously (sevendevices) or every second day (seven devices) for upto 12 months. All devices were calibrated before andafter use (depth resolution 0.1 m). Loggers weresurgically implanted under isoflurane anaesthesiafollowing Gre´millet  et al  . (2005). Dive depth andheart rate data were analysed using the softwareM ULTITRACE  while discarding the first two weeks of recordings to ensure that the birds had fully recoveredfrom the surgery. Dive depths above 0.5 m were usedto extract the number of dives per day, the averagemaximum depth of these dives and the duration of thedives. Heart rates were analysed to determine the totaltime spent flying per day. Following Ropert-Coudert et al  . (2006), birds were assumed to rest either on landor at the sea surface for heart rates below220 beats min K 1 , and to be flying for heart ratesabove this threshold.Behavioural differences between breeding and non-breeding gannets were tested using residual maximum-likelihood analyses for repeated measurements in G ENSTAT 8th edition (VSN International Ltd, Rothamsted, UK).Effects of the time periods were determined by comparingWald statistics (expressed as  X  2 throughout the results) with  F  -distributions (5% significance level). This method 2 D. Gre´millet  et al. Junk-food and seabirds Proc. R. Soc. B  accounted for the fact that wewere dealing with time series of different lengths recorded for different individuals.( d )  Seabird breeding success and energyrequirements Cape gannet breeding success at the Malgas colony wasassessed for the 2004–2005 and 2005–2006 breedingseasons via monthly and bimonthly checks (at the beginningand towards the end of the season, respectively) of nestsmarked throughout the colony from egg-laying (September)until fledging (February of the following year). Capegannets usually lay a single egg and raise at most onechick per year.The average energy requirements (kJ d K 1 ) of breedingadult Cape gannets were estimated after Ellis & Gabrielsen(2002, table 11.6) using an average body mass of 2630 g (see  § 3). Average energy requirements of the chick throughoutthe rearing period (kJ d K 1 ) were taken from Cooper (1978). 3. RESULTS ( a )  Fishing activities A total of 45 458 trawls were performed by commercialfishery vessels operating in the southern Benguela in 2005,out of which 26 579 occurred within the home range of Cape gannets from Malgas. Both inshore and offshoretrawlingfleetswereprimarilytargetinghakesalongtheedgeof the continental shelf and on the Agulhas Bank (south of Mossel Bay), with most effort concentrated along the shelf edgebetweenSaldanhaBay(wheretheMalgasCapegannetcolony is situated) and Cape Point south of Cape Town(figure 1 a ). Trawling activities took place throughout theyear (figure 1 b ), with significantly more trawls performedoutsidethanduringthegannetbreedingseason(2523 G 165versus 1896 G 388 trawls per month, respectively,  t  Z 3.68,  p Z 0.01). Trawling effort nonetheless remained extremelyhigh throughout the year, with 45–90 trawls per day withinthe foraging area of Cape gannets from Malgas. JanFebMarAprMayJunJulAugSepOctNovDec    t  r  a  w   l  s  p  e  r  m  o  n   t   h 010002000300014 ° 0'0"E36 ° 0'0"S34 ° 0'0"S32 ° 0'0"S30 ° 0'0"S( a )( b )16 ° 0'0"E18 ° 0'0"E20 ° 0'0"E22 ° 0'0"E24 ° 0'0"ESaldanha200m500m1000mdemersal trawler effortdistribution (2005)Cape TownMossel Baynumber of trawls5 – 99100 – 399400 – 799800 – 14991500 – 3228 Figure1.( a )Distribution of commercial trawling effortinthe southern Benguelain2005by a20 0 ! 20 0 square grid. ( b )Numberof trawls conducted every month in 2005 by industrial fisheries within the home range of Cape gannets from Malgas Island(2005). The horizontal bar shows the Cape gannet breeding (black) and non-breeding (white) seasons.  Junk-food and seabirds  D. Gre´millet  et al.  3 Proc. R. Soc. B  (  b )  Seabird diet  Regurgitations were obtained from 444 birds. Smallpelagic fishes represented only 15% of all prey items andnever comprised more than 40% for any single month(figure 2). Fishery waste represented 43% of all preyitems, reaching very high proportions during the non-breeding phase (81% between May and September).There was nonetheless no significant difference in theproportion offishery waste between the non-breeding andbreeding phases (  Z  Z 1.39,  p Z 0.2).The calculated calorific value of the diet averaged5.43 G 0.77 kJ g K 1 (range 4.45–6.63 kJ g K 1 ) over the yearcycle and remained lower than the calorific value of sardine (8.59 kJ g K 1 ) and anchovy (6.74 kJ g K 1 ) in allmonths (figure 2). The average calorific value of Capegannet diet was consequently 37% lower than that for asardine diet, and 19% lower than that for an anchovy diet.( c )  Seabird behaviour and survival  None of the chicks died immediately after deployment of theloggersontheirparents.Wehavenoinformationaboutthe proportion of those chicks that fledged. However,since average breeding success was low, we suppose thatmanyof them did not survive until the end of the breedingseason. Of the 21 Cape gannets equipped with dataloggers in early 2005, 18 were resighted during thefollowing breeding season (minimum survival rate85.7%).All seven GLSs were recovered. Five devices providedlocations between January and November 2005 and twobetween January and July 2005. All seven birds remainedwithin the Benguela year-round (figure 3 a , b ), with onlyone individual visiting the northern Benguela off theNamibian coast during the winter non-breeding period.No birds ventured into the Indian Ocean, east of CapeAgulhas (figure 3 b ). Half of all locations were within300 km of the breeding colony (figure 3 a , b ). Month-by-month overlap analyses of maximum concave polygonsrevealed that (with the exception of the Namibianexcursion mentioned above) individual birds remainedwithin a well-defined area throughout the study period(average individual between-month overlap 64.8%, range49.4–74.8%).Total home ranges ofthe different birds alsooverlapped widely (average between-bird overlap 61.3%,range 47.8–100%).Of the 14 birds implanted with HRDDLs in January2005,11wereresightedwhilebreedinginNovember2005.Ten birds were caught and the HRDDL recovered. Birdsshowedsimilarbodymassesatthebeginningandtheendof the experiment (2630 G 100 g versus 2620 G 160 g, paired t  -test Z 0.18,  p Z 0.86). One HRDDL was faulty, fiverecorded throughout the study period (10 months), whilethe others logged data for 3, 6, 7 and 9.5 months,respectively. The behavioural parameters compiled fornine breeding and interbreeding Cape gannets showedsubstantialvariability,butbreedingbirdsgenerallyworkedharder than non-breeding individuals (figure 4). Theyperformed more dives per day than non-breedingbirds (69 G 58 versus 39 G 34;  X  2 Z 106,  p ! 0.001;figure 4 a ) and spent more time underwater (6.1 G 5.5versus 3.3 G 2.7 min d K 1 ;  X  2 Z 99,  p ! 0.001; figure 4 b ).When comparing the entire non-breeding and breedingperiods, birds had similar flight times (368 G 281 versus343 G 248 min d K 1 ;  X  2 Z 1.35,  p Z 0.246), but between July and mid-September daily flight time fell to signi-ficantly lower levels (222 G 204 min d K 1 ;  X  2 Z 56,  p ! 0.001; figure 4 c ). The same pattern occurred for divedepth, which was similar when comparing breedingand non-breeding phases (2.3 G 1.3 versus 2.3 G 1.6 m;  X  2 Z 0.02,  p Z 0.894), but fell to significantly lower levelsbetween July and mid-September (1.4 G 1.3 m;  X  2 Z 50,  p ! 0.001; figure 4 d  ). Only average dive duration showedno significant variability throughout the study period(5.3 G 1.9 s;  X  2 Z 2.04,  p Z 0.154).( d )  Seabird breeding success and energyrequirements Breeding success of Cape gannets at the Malgas colonywas 0.42 chicks per nest in 2004–2005 ( n Z 55) and 0.02chicks per nest in 2005–2006 ( n Z 201). The latter value is JanFebMarAprMayJunJulAugSepOctNovDec   p  r  o  p  o  r   t   i  o  n  o   f   f  o  o   d   t  y  p  e   (   %   ) 020406080100   c  a   l  o  r   i   f   i  c  v  a   l  u  e   (   k   J  g   –   1    ) 4567505030305030301050135051non-breedingbreeding Figure2. Proportion by mass of different food types (vertical bars: black bars, small pelagics; light grey bars, fishery wastes; darkgrey bars, others) in the diet of Cape gannets from Malgas Island during 2005 in relation to their breeding season (horizontalbar). Numbers at the top of the vertical bars indicate the number of birds sampled per month. Small pelagics are anchovies( E. encrasicolus )andsardines( S. sagax ),andotherfishspeciesaremainlysaury( S. saurus ).Thesecondaryaxisandlineshowthemonthly average calorific value of the diet. 4 D. Gre´millet  et al. Junk-food and seabirds Proc. R. Soc. B  the lowest ever recorded for this species, and bothestimates are low compared with those from previousstudies (e.g. 0.69 G 0.07 for the 1986–1988 time period;Navarro 1991). Such fledging success is also lowcompared with the average hatching success of gannetsfrom Malgas (82%; Staverees  et al  . in press), indicatingthat breeding failure mainly occurs during the chick-rearing phase.We estimated that breeding adult Cape gannets requirean average of 3720 kJ d K 1 . During this period, the chickrequires an average of 2060 kJ d K 1 . Since two parentsraise one chick, each of them provides 1030 kJ d K 1 .Raising a chick therefore increases adult daily energyrequirements by approximately 28%. 4. DISCUSSION Previous investigations demonstrated that fishery wastesare highly beneficial to a variety of scavenging seabirdssuch as albatrosses, petrels, large gulls and skuas (seereviews in Montevecchi 2002 and Furness 2003). For these species, fishery waste often has higher energycontent and digestibility than their natural prey (Furness et al  . 2007). Some populations consequently become 40 ° 0'0"S35 ° 0'0"S30 ° 0'0"S25 ° 0'0"S( a )( b )NamibiaBotswanaSouth AfricaLesothoBird Islandpercentage of timespent in the area50%75%90%Bird IslandMalgas IslandPossession IslandIchaboe IslandMercury Island0200 km40 ° 0'0"S10 ° 0'0"E20 ° 0'0"E25 ° 0'0"E30 ° 0'0"E35 ° 0'0"E35 ° 0'0"S30 ° 0'0"S25 ° 0'0"Spercentage of timespent in the area50%75%90%BotswanaSouth AfricaLesothoBird IslandBird IslandMalgas Island0200 kmNamibiaPossession IslandIchaboe IslandMercury Island Figure 3.Combined home rangesof seven( a ) breeding and ( b ) non-breeding Cape gannets from Malgas Island asinferred fromlight levels recorded in 2005 by bird-borne GLSs (see  § 2 for details). Lines indicate the 500 and 1000 m isobaths.  Junk-food and seabirds  D. Gre´millet  et al.  5 Proc. R. Soc. B
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!