Corella, 2003, 27\4li 1o2-10s BOOBYNESTING ON RAINEISLAND, GREATBARRIER REEF

P K. DYER

Universityof the SunshineCoast, Sippy DownsDrive, Sippy Downs, 4556

Reeiwd:11February 2002

Collectingreliable field data on bregdingpopulations of boobyspecies on RaineIsland tor temporal comparison is challenging.This trial'saim was to establishbaseline data on nest censusestimates in December1998 to serveas ihJfoundationfor continuedmonitoring. Based On censuses, estimates of numbe.sof 3 700 Brown, 7OOMasked, and 330 Red-footedBooby nests were established. Egg predationby seagullsin December2000, however,prevented the first replicationof the studythus exposingthe limitationsol the method

'a INTRODUCTION 1993); and for the Red-footed Booby, range from few stragglers'in December l9l3 to 300 in February 1959 RaineIsland (144'01'E,11"37'S), a 28 hectarecoral cay (Warhaml96l). that consistsof a vegetatedridge sunounding a bare central few Australian studies detailing Red- depression,previously mined for phosphorous,is located Because therc are numbers or status approximately 100 kilometres ENE of Cape Grenville, footed or breeding (Marchant and Higgins 1990) this study attempted,in northernQueensland. The island supportsbreeding colonies data for continuing of three Booby species: Brown SuLaleucogaster', Masked December1998, to establishbaseline It was intendedto avoid Sukt doctylalrat and Red-footed Sula sula. breedingpopulation estimates. error due to absentbirds during the daytime, by estimating Brown Booby nestson Raine Island are distinguishable breedingpopulations based on nestcensus estlmates. 'loosely from those of the Masked Booby by the assembled'twigs gatheredin and around the nest, whereas METHODS Maked Booby nestshave a small collectionof pebblesor Milton el dl. (in Neil 1l dl. 2000 p.l3) stated that allhough the 'quadra! coral rubble in the centre with a similar collection search' for estimatesof seabird populations was more Practlcal 'line surrounding the outer perimeter of the nest (Marchant and when time was limited. there was less error in the transect'method in conjunction with line uansecl Shearwaler Higgins 1990).Although few MaskedBoobies nest in the This study was conducted data collection (see Dyer 2003). veg€tatedridges and on beaches,Warham (1961) suggests that guano mining, that denuded and extended the central To establish breeding estimates for the three Booby species, Boobies within ter north/south transects, made up of 4 depression,may haveinadvertently enhanced the preterred and their nest contents metre wide x 5 mctr€ long contiguous quadrats, were recorded Data habitatfor MaskedBoobies. collection was limited to a couple of hours in the moming and afternoon ro avoid the risk of heat stress for birds and researchersalike During The majority of the bare central depressionon Raine the day, because many birds are abseflt feeding or Ieave nesls when lsland is occupiedby Masked Boobies with Brown Boobies disturbed by researchers, the nest count was considered to indicate nestingaround the edges,extending into all areasof the breeding populations more reliably than bird counts. vegetatedridges. Red-footed Boobies roost and nest, about The ralio method for eslimating populations from stratified sample 'cases' a foot above ground, on matted vegetation or branchesof counts (Cochran l9?7, scc Dyer 2003) was used to eslimate all low shrubs.Except for Moulter Cay, Raine Island supports (i.e. nests with or wilhout birds, as well as birds without nests).In addition estimates for nests only, and birds only, were established for the only Red-footedBooby colony on the Great Barrier each of the three Booby species. Data were recorded according lo the Reef (Taplinand Blaber 1993). various habitats previously identified by Jeff Miller (J. C\nelius, pers. comm.) for coDtinuing bird studies (see Dyer 2003). In additior, The Raine lsland Brown Booby population has been 'in Pearson's Corelation Coefficient was used to eslablish any relationship recordedvariously as breeding small numbers'in July between sample areas and findings, and Chi-squared tesls were used to 'thousands 1843 (MacGillivray 1846 p. 1478) to all over test for any influence according to habita( type or transect location. 'guano theplace'in October l9l0 (MacGillivrayl9l8 p.l8l), and Stoddarl €l al- (1981) measured the density of birds in the between 8 000 and 16 000 (no dates provided) (Taplin and flats' using five sample areas. This was then extrapolated to estimale Blaber 1993). Without clarification of methods for the number of nests for Brown and Masked Boobies in lhe central according 1o Stoddart €r al., covered 40440 square estimatesor counts, little can be surmized conceming depression that, metres. This esdmate excluded absent feeding birds; those disturbed by population dynamics from the literature; however,Stoddart researchers;and birds nesting in the ridge areas.Although the sampling et al. (1981 pp. 23, 26) provided densitiesfor Brown mcthod was not replicated hcre, a similar mean density measure was (6.835/100mt,2'164 birds)and Masked(1.730/100 m'?, establishedand extrapolated for the ccntral depression, for comparative 'guano for the different 699 birds) Boobies in the whole of the flats' in purposes. Interpretation of comparisons should account research reportcd here was carried out population timing of data collection. The November1973. Reports of the MaskedBooby in the ea.ly morniflgs or late aftemooos belween 4 and 7 December rangefrom 400 to 500 in February1959 (Warham 1961) 1998 inclusive, whereas Sloddart, Gibbs, and Hopley's dala were to I 500-2 500 (no dates provided) (Taplin and Blaber collected at midday on 3 November 1973 (Stoddart er al. l98l). December,2003 P K. Dyet: Booby N€stingon Raine lsland,

RESULTS 5 460 (+ 97o se) cases altogether. The latter estimate allowed for absent birds and compared favourably with Brown Booby ( Entire Islond) total counts of Brown Boobies (by the author and others) The breeding stage for Brown Boobies ranged from using a quadrat method for counting, including visible empty nests, through adults incubating eggs and chicks but involving an unknown enor factor, of 2 073 mid accompaniedand unaccompanieddowny chicks, to almost afternoonon 28 November1998, and 1274 mid morning fledged chicks. This diversity complicatedanalyses so on 10 December 1998. The numbers of cases were censusesfor all cases (aggregating cases with birds and/ positively related to transect areas (Pearson'sCorrelation or nests), for cases involving nests without birds, and for Coefficient(df = 8): birds 0.0.914,p = 0.000;nests 0.872, birds without nests,were established. p = 0.001; all cases,0.926, p = 0.000). No significanr relationship was found between Brown Booby nests, In 31.9 per cent of cases,Brown Boobies were not Brown Boobies apart from nests, and habitat type (X, = obviously on or near a nest (Table l.l). Unaccompanied 1.529,df = 4, p = 0.Szt, n = 335) or transecrs(Ir= 6.551, chicksconstirured the majoriryof rhesecases 1i.e. 14.5 per df = 9, p = 0.684,n = 335). cenlof all casest.Empty neslsconstitured 39 per centof all cases(Table l.2). The remaining28.9 per cent of all Masked.Booby ( Entire Island) casesshowed signs of breeding.Of theseI5.4 per cent (of The breedingstage of all cases)represented birds still incubating one or two eggs the MaskedBooby was similar to that of the Brown Booby in (Table1.2). In additionto the habitatsanalysed, two chicks, that it encompassedthe full range from empty nests one nest,and one nest with an adult, a chick and an egg to almost fledged chicks. Masked Boobies were counted in (i.e. four casesin all) were presentin the vicinity of the those transectsthat traversedthe central depression.In 44.9 per transecttrajectories in the grassy verge that sunounded the cent of casesbirds were not vegetatedridge areas. obviously associatedwith nest (Table 2). Unlike Brown Boobies,lone MaskedBooby chicks were the minority in The censusfbr all habitatscombined resultedin estimates this category,constituting only 9 per cent of all cases.Lone of 3700 (1 (+ 77o se) nesrs, 1760 9o/ose) birds. and adults or pairs constituted the majority of cases not

TABLE I.1 BrowDBoobies no! in nests,foufld in December1998, for habitatson RaineIsland: Central Dcpression; sourh section of the southernvegelated Ridgc (SsvR)r North Sectionof the Southemvegetated Ridse (NSVR): and NorlhemVeseralcd Rirtpe /NVRr Habitat Brown Booby Birds Depression NSVR NVR Nest No. % No. Vo Adult l4 4 1.2 4 1.2 l 0.3 23 Two adulls 6.9 I 0.3 0 0.0 0 0.0 l 0.3 2 0.6 Two adults and chick I 0.3 0 0.0 0 0.0 0 0.0 I Adult and chick 0.3 t2 3.6 l 0.3 4 t.2 3 0.9 Adult and juvenile 20 6.0 I {J 0.0 0 0.0 0 0.0 I 0.3 Chick 2',7 '7 8.2 1 0.3 2.1 13 3.9 48 14.5 T$o chicks I 0.3 0 0.0 0 0.0 0 0.0 I Juvenile 0.3 1 2.1 0 0.0 I 0.3 l Two juveniles 0.3 9 2.',7 0.0 0 0.0 I 0.3 0 0.0 I 0.3 Sub-total 64 r 9.3 6 t.8 17 5.1 19 5.'l 106 31.9

TABLE 1.2 BrownBooby n€sts found in December1998' for habiratson Rainelsland: cenFal Depression;south sectiooof rhesouthem vegeEled Ridge (ssvR): North Sectionof rhe Sourhem , VegeuredRrdge (NSVRt; ;nd NorthemVegetared Ridge iNVnl.- Habitar Brown Bobby Depression NVR SSVR TOTAL Nests No. E No. % No. No. Empty 65 19.6 l) 4.5 Adult 7.3 7.6 t29 39.0 6 L8 0 0.0 2 Adult and 0.6 3 0.9 u 3.3 chick l3 3.9 I 0.3 2 0.6 5 1.5 2l 6.3 Aduh, Chick and egg 4 t.2 I 0.3 I 0.3 0 0.0 6 Adult aod egg 28 NJ 1.8 I 0.3 3 0.9 3 0.9 Adult and two eBgs 8 2.4 35 10.6 0 0.0 0 0.0 I 0.3 Adult,egg and juvenite I 0.3 9 2.',7 0 0.0 0 0.0 0 0.0 Chick 4 1.2 I 0.3 0.0 0 0.0 7 2.1 l1 Egg i 0.3 0 0.0 0 0.0 0.0 I Two eggs 0 0.0 0 0.0 0 0.0 t 0.3 I 0.3 Sub-lotal 130 39.2 l8 5.4 32 9.',7 45 13.6 225 67.9 Toral1.2 and 1.2 r94 58.60 7.30 49 14.8 19.3 t00.0 P.K. Dyer: Booby Nesting on Rainelsland, Great Barrier Reef Corella27(4)

TABLE 2 Masked Boobies and lheir nests found in December1998, for habilats on Raine Island: CentralDepression and Norlh Sectionof the Southemvegetated Ridge (NSVR) Habitat MaskedBooby Depression NSVR TOfAL Birds (no nest) No. Vo No. Ea Adult t2 15.4 I 1.3 13 t6.7 Two Adults 10 12.8 0 0.0 10 12.8 Adult and chick 5 6.4 0 0.0 5 6.4 ,7 '7 Chick 9.0 0 0.0 9.0 Sub-total 43.6 I 1.3 35 44.9 Birds (no nest) No Vo No. 7o No. Vo 'l '7 Empty 9.0 0 0.0 9.0 'l '7 Adult 9.0 0 0.0 9.0 Two adults 2 2.6 0 0.0 2 2_6 Adult and ctuck 3 3.8 0 0.0 3 3.8 Adult and egg 1'l 2t.8 I 1.3 18 23.1 Adult and two eggs 5.1 0 0.0 4 5.I Chick t 1.3 0 0.0 r 1.3 Ecg I 1.3 0 0.0 l 1.3 Sub-total 53.9 I 1.3 43 55.2 '76 ',7 TOTAL 91.4 2 2.6 8 100.0

TABLE3 Break down of Red-footed Booby cases (i.e. bird, nest or both) in December 1998, tor habitats on Raine Island: Central Depression, North Section of the Southem Vegetated Ridge (NSVR); and South Section of th" S"",h".r kg"t"t"d Red-footed Habita! Booby Depression NSVR SSVR TOTAL Birds (no nest) No. 7o No. % No. Vo No. Vo Adult 0 0.0 I 3.',7 I 3.'l 2 7.4 Chick 0 0.0 I 3.7 0 0.0 l 3.7 Juvenile I 3.',7 I 3.7 2 4 14.8 Sub-Total I 3.'7 3 I l.l 3 l l.l 1 25.9 Nests No. % No. EA No. % No. Vo Nest 0 0.0 6 22.2 t4 51.9 20 74.1 TOTAL I 3.1% 9 33.3% t'l 63.OVo 2',7 100.0/o

associatedwith nests (29.57oof all cases).Other cases Red-footedBooby (Entire Island) recorded were made up of empty nests (97o oi all cases) Red-footed Booby with 34.6 per cent of all casesshowing signs of breeding. In December 1998, the only signs of than nests,were one Birds still incubatingone or two eggsconstituted 28.2 per breeding in the sample transectsother four juveniles. Though there cent of all cases(Table 2). well-developedchick and were only 20 nests in the sample area, the clumping of The censusfbr all habitats combined resulted in 700 (i nests in the southern vegetatedridge on the western end 1570se) Masked Booby nests,570 (+ 2l9o se) Masked of the island resulted in relatively large standardenors for Boobies and | 2'70 (+ Dqa se) cases.This compared with the censusestimates. In contrast with the other two species, total counts of Masked Boobies, based on quadrat counts no Red-footed Boobies were recorded on nests: all nests that included visible chicks but involved an unknown error (74.19oof all cases)were empty.A singleadult, chick or factor. of 733 mid afternoon on 28 November 1998, and juvenile constitutedthe remainder of records involving 888 mid morning on l0 December1998. The number of birds (4) (Table 3). Given the small sample size, few Masked Boobies o. their nests tended to be positively conclusionscan be drawn other than that in 1998 the related to transect areas but this trend was not significant breeding stage of the Red-footed Booby was further for those without nests (Pearson'sCorrelation (df = 7): advancedthan that of the other Booby species. birds 0.603, p = 0.065; nests0.807, p = 0.005; all cases, 0.812,p = 0.004).The resultswere influencedby transect The Red-footed Booby censusfor all habitats combined location(12 - 41.612,df= 8, p < 0.001,n = 78); the resultedin 33o (+ 43Eose) nests, ll0 (i 3070se) Red- distribution of Masked Boobies and their nests being footed Boobies, and 440 (t 357o se) cases.This compared aggregatedwith 76 cases(91 .49o) in the central depression with total counts of 30 Red-footed Boobies, including and only two (2.69o) casesin the northern vegetatedddge. visible chicks but involving an unknown error factor, mid December,2003 P K. Dyer:BoobyNesting on Rainelsland, Great Barrier Reef 105 afternoon on 28 November 1998, and 100 mid morning on Iocations,undulations and centralmounds block the view, l0 December 1998. Becausethere were so few birds, and again resulting in under-estimates.It was considered that becausethey nest in aggregationsin the less sensitiveareas establishingthe breedingpopulation by nestcounts would of the island, a total count of Red-footed Boobies or their be more reliablethan daytime bird countsthat are biased nests was feasible and therefore would be the preferred by absentbirds. censusmethod. In December 1998 the Boobies were relatively unalfected Cert ral depre s s ion comp arison by turtle disturbanceas the turtle numbers were extremely low with approximatsly 200-400 Green Turtles Chelonia The estimatefor the central depressionin December mydas a night visiting the beach. Researcherdisturbance 1998was basedon a samplearea that constituted6.56 per to birds in 1998 was also minimal. No predationby gulls cent of the central depression.Because the area of the was observed.However, in 2000 when the first replication central depressionin this study was determinedto be of the nest censuswas intended, it was abandonedbecause 42 680 squaremetres, the Stoddarte/ a/. (1981)densities of severepredation of Booby eggsby Silver Gulls Larls of 0.06835 square metres (Brown) and 0.01730 square novaehollandia associatedwith researcherdisturbance of metres (Masked) were reworked accordingly. This resulted Boobies,thus exposingthe limitationsof this method. in estimatesof 2 917 (previously2764 basedon 40 440 mr) and 738 (previously 699 based on 40 440 m'?)for Unfortunatelythe major problemof establishingreliable Brown and MaskedBoobies respectively. datacollection methods for longitudinalanalyses were not overcomeby this exercise.It had been anticipated,after The densityfor Brown Booby adultsand juveniles in the the successof this studyin 1998,that the methoddescribed centraldepression in December 1998 was 0.0139 square here would be useful for continuingstudies. If the birds 'managed'data metresresulting in an estimateof 594 adultsand juveniles; of the islandare to be collectedover time and for adults,juveniles and chicksthe densitywas 0.0293 in consistentnon-intrusive ways are gssential,but the issue squaremetres resulting in an estimateof I 249 Brown af how to achievethis effectively and efficiently is still Boobiesincluding chicks. The latterfigure is approximatelv unresolved. half that of the reworkedStoddart er al. (1981) estimate for November1973. ACKNOWLEDGMENTS Similarly, the densiry fbr Masked Booby adults (there The author appreciatesthe supportof the Raine Island were no juveniles) in December 1998 in the central Corporation that was constituted by The Meaker Trust depressionwas 0.0178 square metres resulting in an (Raine Island Research)Act 1981 of the Queensland estimateof 761 adults.For adults and chicks the density Parliament,and that organizesand funds regularresearch was 0.0264square metres resulting in an estimateof I 127. trips to Raine Island. Michael Angelides and Duncan The latterfigure is approximatelyone and a half timesthat Thompsonassisted with the data collectionand collation. of the reworked Stoddart er al. (1981) estimate for Their assistanceis appreciated. November1973. REFERENCES DISCUSSION Cochran,W G. (1977). Sarnpling'ltchniqucs' (John Wiley and Sons, Because Stoddarr et al. (1981) do not describe the Nerv York.) nestingin the vegetatedridges it is not possibleto ascertain Dyer, P K. (2003). A Decline ln The Numbcr Of WedgeTailed whetherthere were fewer Brown Boobieson RaineIsland ShcarwarersBreeding On Rarnc Islandl Corella:2'l: l3-l'1 in December1998, or whether the perceivedincrease in Maccillivray, J. (1846). An accounrof Raines Isler, on NE coast of New Holland.Zoobqi\t lt 1413 1481 Masked Boobies has resulted in more Brown Boobies MrcGillivray, w' (1918).Omithologisls in Norrh Queenstand.lj,xr r7: nestingin otherhabitats. Few MaskedBoobies were found 180-212. outsidethe depression(Table 2). Although Stoddarter al. Marchant,S. and Higgins, I, J. (1990) Handbookof Ausrrilian, New describenest occupancy, they did not presentproportions Zealand & Antarctic birds' (Oxford Universiry Press Ausrralia, Mclbourne.) of occupiedvs unoccupiednests so comparisonscannot be Neil, D. T. and Brieze,L C. (Comp.) (2000).Raioc made. IslandCorporation: ResearchReport 199f1998 Raine lsland Corporarion,Brisbane, Establishingthe size of Booby populationson Raine Sloddart.D. R., Gibbs, P E. and Hoplcy, D. (1981).Narurat history of Raine Island, Greal Barrier Reef. Au Island provesdifficult. Many birds Rerearch Bu etit 2541 t41. leave the islandearly Taplin, A. and Blabcr, S. J. M. (1993). Seabirdbrccding poputarion 'Rainc in the morning or roost on the beach during the day, sludresat RaineIsland. In IslandEnvirons, Great Banier Reei returningto nestsperiodically during the day. Birds can quest to Preservea Fragilc Outposl of Nature'. Comp. (Eds A. K be countedwith minimal disturbanceby viewing through Srnylh,K. H. Zeve.in8and C. E. Zevcrin8).pp. 5t-56. (RaineIsland Corporation. Brisbane.) a telescopefrom the toweron the easternend of the island, Warham,J. (1961).The birds of RarneIsland, pandor.r but with no similar vantagepoints Cay and Mura) in westernor central Island Sandbank,North Quecnsland.I:nu 6t:7't 93