POPULATION STUDY

HEBREEDING RANGE OF LEACH'S Storm-Petrel( Oceanodroma leu- The world'slargest known corhoa)in the Northwest Atlan- tic is centered in eastern Newfound- land. Owing to the species'nocturnal nestingcolony of and burrow-nestinghabits, the breed- ing populations and distributions of storm-petrels, first documented in Leach's Storm-Petrels Newfoundlandby Petersand Burleigh (1951), have historicallybeen grossly underestimatedand are not presently on BaccalieuIsland,, well known. Storm-petrel coloniesin are vastly larger than all others in the Northwest Atlantic Newfoundland (Huntington 1963; Montevecchi, in prep.), but few quantitative data on colony populationshave been gener- ated. The colony of Leach's Storm-Pe- trels on Baccalieu Island, Newfound- B. O. Sklepkovych and IV. A. Montevecchi land, hasgone largely unrecognized as an important nestingsite for the spe- des (Brown et al. 1975; Nettleship 1980). In the present paper, we esti- mate the breeding population of storm-petrelsbased on measurements of burrow densitiesin all nestinghab- itats and on quantitative estimatesof the island'ssurface areas. Our popu- lation figure, which is compared with thoseof other coloniesthroughout the world, yields new perspectivesof the energy consumption of these birds in the marine ecosystemof the North- west Atlantic.

Study Area

BaccalieuIsland (-6.3 x I kilome- ter, 48'07'N, 54ø12'W) is situatedoff the northeastern tip of Newfound- land's (Figure 1). Leach'sStorm-Petrels nest in a variety of habitats over the entire island, but principallyoccupy grassy-hummock- ed slopes,boreal forest and Empetrum heaths. Smaller, less defined habitats suchas herb (grassyareas with more than 60 percent fern or tall herb cover), grassy meadow, and heath Leach• Storm-Petrel (Oceanodromaleucorhda). Photograph/GeoffLeBaron/VIREO mixed with herb or grassesare also 105/7/009. utilized by nesting storm-petrels (Sklepkovych1986). The densestcon- centrations of storm-petrelsoccur in the grassy-hummockedareas along

38 AmericanBirds, Spring 1989 TABLE 1. Percentage(number) of burrowoccupancy in grassy-hummocked,forest, and heath habitat plots. Habitat Contour % Occupied% BurrowsWith Burrows Eggs/Chick Grass Level 74.6 (47/63) 69.8 (44/63) Slope 67.5 (56/83) 60.2 (50/83) Total 70.5 (103/146) 64.4 (94/146) Forest Level 71.4 (10/14) 50.0 (7/14) Slope 72.4 (21/29) 72.4 (21/29) Total 72.1 (31/43) 65.1 (28/43) Heath Plots 55.6 (10/18) 50.0 (9/18) Transects 52.4 (11/21) 52.4 (11/21) Total 53.8 (21/39) 51.3 (20/39) Grand Mean (Sum) 68.0 (155/228) 62.3 (142/228)

the eastern side of the island (Figure longesttransverse axis of the islandat 2), and medium and low densitycon- approximatelyequal intervals.Cross- centrations are found in forest and sectionsof the island were plotted and heath regions,respectively. Northern scaledrelative to the planar distance Fulmars (Fulrnarusglacialis), North- (X), and then measuredusing the Nu- ern Gannets (Sula bassanus),Black- monics digitizer (Y). The ratio of Figure 1. Aerial photographof Baccalieu leggedKitfiwakes (Rissa tridactyla), thesevalues (Y/X) providedan esti- Island, located off the northeasterntip of NewfoundlandgAvalon Peninsula between Common and Thick-billed tourres mate of the proportional increasein Conceptionand Trinity bays.Photograph/ ( Uria aalgeand U. lornvia), surfacearea of each segmentresulting Canadian Dept. of Mines, Energy, Re- (Alca torda),Atlantic Puffins(Frater- from the island's contour, and the sources. cula arcfica),Black Guillemots( Cep- mean of these ratios was used as an phusgrylle), Herring Gulls (Larus ar- estimate of overall increase in surface tained from data collected in associa- gentatus), and Great Black-backed area due to slope.This procedureas- tion with studiesof red fox hunting Gulls (L. marinus) also nest on Bac- sumed that habitats are equally dis- strategy(Sklepkovych 1986). Both the calieu Island, making up the greatest tributed over different contours and number of burrowsat fox diggingsites diversityof breedingseabirds at any as a consequencetends to underesti- and at random adjacent control sites site in Newfoundland and Labrador mate grassy-hummockedand forest were included in the estimates of hole (Montevecchiand Tuck 1987). Red habitats, because much of the rela- density, which were similar to those foxes(Vulpes vulpes)are the island's tivelylevel terrain of the islandis com- at randomly selected sites (Sklepko- only resident mammals (Maccarone posedof heath. Owing to differences vych 1986). and Montevecchi 1981; Sklepkovych in the nestingdensities associated with During May and June 1984, five- 1986). these habitats (see Results), the pro- by-ten meter occupancy plots were cedure will also tend to underestimate established in typical grassy-hum- the breedingpopulation. Two-dimen- mocked (n = 4) and forest (n = 4) Methods sional considerations of habitat areas habitats. Each plot was subdivided would generateerrors that greatly ex- into four equal quadrats(12.5m 2) of A Numonics digitizer(Model 1210) ceed those produced by our contour which one was chosen at random. was used to estimate the two-dimen- estimation procedure. Two plots in each habitat were in sionalsurface area and major habitats A census of holes was made in each areas with _>30* slope and two were of Baccalieu Island from a 1:8650 habitat to estimatethe densityof po- on areaswith little or no slope(0-5ø). scaledmap. For comparisonand ad- tential storm-petrel burrows. To cover In heath habitat, the density of holes ditional information, three 1:25,000 most areas of the island, 57 relatively (potentialburrows) was very low and aerial photographs(Canadian Dept. equidistantsites were preselectedon a variable, and therefore, two 100-me- Energy, Mines, Resources)compris- map. The samplingscheme was strat- ter-squareplots were set up in heath ing the southern,central, and north- ified according to slope and geo- areaswith high burrow densities(i.e. ern portions of the island were also graphiclocation. Randomness of the on the island'speriphery), and two 4 used.A grid of 150-1 cm2 quadrats exact location of studyplots and tran- x 100-meter transects were set out was laid over the aerial photographs, sects at these sites was attained by randomly in typical heath habitat. and each quadrat was examined with travelling a random compassdirec- Holes were examined betweenJuly 12 a magnifying lamp or stereoscopic tion (N, E, S, W) for a distancedeter- and 22, early in the chick period ap- glasses,and the percentageof each mined from a random numbers table proximatelyfive to sixweeks after egg- habitat type was visuallyestimated. between 100 and 500 meters in heath, laying (Monteve•chiunpubl. data; To estimate the island's surface area 50 and 100 meters in forest and 25 Morse and Buchheister 1979). Dia- contour analysiswas performedon a and 50 meters both in mixed heath gramsof plots were drawn indicating 1:50,000 map (Canadian Dept. En- and in grassymeadow habitats.Esti- hole position, distance to adjacent ergy,Mifies, Resources). Fifteen tran- matesof hole densitiesin grassy-hum- holes, connections between burrows, sectionswere drawn perpendicularto mocked and herb habitats were oh- burrow lengths,general vegetational

Volume43, Number 1 39 features and the presenceof nest ma- terial, adults, chicks, and eggs.Holes <20 centimeters long were typically inactive and were not considered to be burrows. Nest chambers with mul- tiple entrances were counted once. The number of confirmed nests ex- cluded burrows that were probable nests but in which nest chambers were inaccessible. For heath, forest, and grasshabitats exact occupancyvalues were used in calculating density; for all other sub-habitat estimates, the .$ overall proportion of nestingburrows and grand mean of the occupancy rates of the three sampled habitats were used.

Results

PhysicalHabitat Characteristics-- The total two-dimensional surface area of Baccalieu Island was calcu- lated to be 523 hectares, approxi- mately 60 percent of which is _•75 meters above sea level. Analysis of Figure 2. The grassy-hummockedslopes on the easternside of BaccalieuIsland, where island contour revealed 31-100 per- Leach• Storm-Petrelsnest in abundance.PhotographlB. O. Sklepkovych. centincreases in area(2 __SD = 59 _+19%) due to topography,suggesting that the surface area of Baccalieu Is- and occupancy by known nestersat occupancyx % breeding)of all acces- land is actually 832 hectares,with 688 plots and transects in grassy-hum- sibleburrows (N = 228) wasestimated hectares providing potential nesting mocked, forest, and heath habitats. to be occupied by breeding pairs. habitat for Leach's Storm-Petrels. Sixty-eight percent (155/228) of all Nearly identicaloccupancy rates have Analysis of aerial photographsand accessible nests were occupied by been found on Great Island, Witless maps yielded estimates of 315 hec- either an adult, chick and/or egg.Per- Bay, Newfoundland and on Little tares (38%) of heath, 286 hectares centageof occupancywas similar be- Duck Island, Maine. (R. Butler [De- (34%) of forest,and 46 hectares(6%) tween grassy-hummocked(71%) and partment of Science & Math, Univ. of grassy-hummocked habitats. forest (72%) habitats and lower, Maine, Farmington]pers. comm.). Smaller, lessdefined habitatsmade up though not significantly,within the Study plots varied in size and num- 41 hectares(5%), and non-nestingre- heath habitat (54%). The presenceof ber within and between habitats and gions(including exclusive puffin nest- an egg or chick was used to confirm therefore, two measurements were ing habitat, see Montevecchi and that a nestwas occupiedby a breeding calculated:the grand mean of holesin Tuck 1987) comprised 144 hectares pair, the percentage being approxi- each habitat, in order to obtain a var- 07%). mately equal in each habitat (grassy- iance estimate,and the mean density Burrow Occupancyand Density at hummocked 92%; forest 90%; heath of holes based on the total number of Habitat P!ots.--Table I summarizes 95%;combined X = 92% (142/155). holesencountered/total area sampled. the percentagesof burrow occurrence Therefore, a minimum of 62.3% (% In the former measurement, plots

TABLE 2. Estimates of the numbersof nestingpairs of Leach's Storm-Petrels in different habitats on BaccalieuIsland, Newfoundland. Area(N) Density Est.Density Est.Surface CorrectedEst Habitat Plotsof(m 2) (Holes/m2) OccupiedBurrows Area(ha) (BreedingPairs) Heath 3100 (23) 0.046 0.017 314.8 54,000 Forest 577 (24) 0.941 0.604 286.2 1,730,000 Grassy-Hummocked 379 (83) 4.166 2.131 52.2 1,111,000 Herb 126 (21) 4.010 2.495 15.4 385,000 Meadow 700 (7) 0.143 0.088 10.8 10,000 Mixed 63 (3) 0.704 0.478 14.6 70,000 Non-Nesting ------137.7 -- Total ------831.7 3,360,000

40 American Birds,Spring 1989 _•2m?were excludedfrom analysisin the central heath region (0.008 4- On the basisof the presentcensus, this an effort to reduce bias toward small 0.003m--') alongthe rock backboneof speciesis now estimated to comprise plots, then each plot regardlessof size the island. The heath also provided 80 percent of the seabird population was weighted equally. The grand nestingsubstrate along the island'spe- breeding in Newfoundland (Cairns et means of the hole densities are pre- riphery; this sub-habitat was esti- al. 1986a) and to make up a com- sentedbelow, but the mean densityof mated at five percent of total heath parable high proportion of the breed- holes of all the plots in a habitat was habitat (Sklepkovych1986). ing seabird population of the entire considered a better estimate and was Table 2 delineates the components NorthwestAtlantic region.The major used to calculate storm-petrel num- of our estimateof the breedingpopu- colonies of Leach's Storm-Petrels in bers. lation of Leach's Storm-Petrels on Newfoundlandare shownin Figure 3. Burrow density was the highest in BaccalieuIsland. The vast majority of The Leach's Storm-Petrel popula- herb and grassy-hummockedhabitats nesting burrows are associatedwith tion of Bacealieu Island is now the and differedsignificantly among hab- forestand grassy-hummockedhabitat. largest documented in the scientific itat types(x, 2 -- 126.6, P • .001; see The estimated breeding population literature.The next largestknown col- Table 2). In other large storm-petrel based on contour corrections for the ony is located on the Japaneseisland coloniesin Newfoundland, high nest- island'ssurface area is 3,360.000 nest- of Daikokujima, where a breeding ing densitiesare also found in areas ing pairs, with lower and upper 95 population of 1,000,000 pairs has with thick fern cover (R. Butler [Sci- percentconfidence limits of 3,130,000 been estimated(Hasegawa 1986). Col- ence & Math Dept., Univ. Maine Far- and 3,590,000 nesting pairs, respec- onies on Gull and Great Islands in mington],pers. comm.,Montevecchi, tively. on Newfoundland's east pers. obs.). Overall the mean number coast (Fig. 3) have been estimated to of holesin grassy-hummockedhabitat contain approximately 530,000 and was similar between areas of low (0- 250,000 nesting pairs, respectively 29*) and intermediate (30-59*) slope Discussion (Nettleship 1980), though the num- at 4.3 _4-0.4m -2 and 4.2 _4-0.3m -2, bers of nestingpairs on theseislands respectively,and lower(3.6 _4-1.2m-•') may be much higher (L. M. Tuck in within steep(60-90*) areas.In conif- Leach's Storm-Petrels are the most litt.). Moreover, in Newfoundland erous forests, hole density increased abundant breeding seabirds in the and in Alaska (seeLensink 1986) and with slope(0-29 ø, 0.7 _4-0.05; 30-59*, Northwest Atlantic. Owing to their elsewherethe sizesof many largecol- 1.7 4- 0.3; 60-90 ø, 2.9 4- 1.4). There nocturnal and burrow-nestinghabits, onies still require systematicestima- were no significantdifferences in the they are difficult to census.Their col- tion, and many colonies remain to be mean lengthsof occupiedburrows in ony siteshave often been overlooked "discovered." sloped and level plots within grassy- and their numbers underestimated In recent years, much research at- hummocked(overall X 4- = 38.4 4- (Montevecchi and Tuck 1987). tention has focusedon population en- 10.0cm) or forest(overall • 4-= 37.5 Though data on the breedingpopula- ergetics and ecosystem modelling. 4- 12.0 cm) habitats.Northern (0.036 tions of marine birds in the Northwest With regard to marine systems,the 4- 0.013m-•') and southern(0.050 4- Atlantic has increased tremendously lack of information about breeding 0.008m-:) heath regions had about in recent decades, Leach's Storm-Pe- populations of storm-petrels has led five times the density of holes as did trels remain enigmatic in this regard. to their being virtually ignored in all

Leach• Storm-Petrel(Oceanodroma leucorhda). Photograph/Geoff LeBaron/VIREO 105/07/008

Volume 43, Number I 41 considerationsof the energy reqmre- the field work on Baccaheu Island •n CAIRNS, D K, R D ELLIOT, W ments of breeding seabird communi- 1983. R. Butler and D. K. Cairns pro- THRELFALL, and W. A. MONTE- ties. Based on the present informa- vided useful comments on the manu- VECCHI. 1986a.Researcher's guide to Newfoundland seabird colonies Me- tion, Cairns et al. (1986b) have pro- script.Research was supportedby the morial Univ. Nfid. Occas.Pap. Btol 10 visionally estimated that these small PsychologyDepartment of Memorial CAIRNS, D. K., W. A. MONTEVECCHI, animals consume over 50 percent of University of Newfoundland and and V. L. BIRT. 1986b.Energetics and the energytaken by marine birds that NSERC Individual Operating Grant preyconsumption by seabirdsbreeding breed in Newfoundland. These con- (A0687) awarded to WAM. The Ca- in Newfoundland (abstract). Pacific siderationsallow us to develop more nadian Coast Guard provided trans- Seabird Grp. Bull. 13: 102. HASEGAWA, H. 1986. Status and con- realisticand comprehensivemodels of portation to and from Baccalieu Is- servationof seabirdsin Japan, with the breeding seabird community of land during winter. The lighthouse- special attention to the Short-Tafied the NorthwestAtlantic. In this regard, keepers (R. Hyde, L. Walsh, F. Albatross.Pp. 487-500, in J.P. Crox- the censusingof Leach's Storm-Pe- Noonan, E. Blundon and P. Rice) all, P. G. H. Evans and R. W. Schreiber trels is a very important step in un- provided lodging and hospitality on (eds.),Status and Conservationof Sea- derstanding and defining the avian the island. birds.Internat. Counc. Bird Preserv, component of the marine ecosystem. Cambridge,England. HUNTINGTON, C. E. 1963. Population LITERATURE CITED dynamics of Leach's Storm-Petrel ACKNOWLEDGMENTS (Oceanodromaleucorhoa). Pp. 701- BROWN, R. G. B., D. N. NETTLESHIP, 705, in C. G. Sibley,J. J. Hickey,and P. GERMAIN, C. E. TULL, and T. M. B. Hickey (eds.), Proc. XIII Int H. Butler provided valuable infor- DAVIS. 1975. Atlas of Eastern Cana- Ornith. Congr.,Am. Ornithol. Union, mation and help with topographical dian Seabirds.Canadian Wildl. Serv., Baton Rouge. analysis,and J. A. Becker assistedin Ottawa. LENSINK, C. J. 1986. The status and conservationof seabirdsin Alaska Pp Figure3. The major coloniesof Leach'sStorm-Petrels in Newfoundlandand on St. Pierre 13-27, in J.P. Croxall, P. G. H. Evans et Miquelon. Estimated numbersof breedingpairs are given in parentheses.Sources of and R. W. Schreiber(eds.), Status and Conservation of Seabirds. Internat estimatesare given in Cairns et al. (1986a). 1 -- BaccalieuIsland, 48ø07'N,54ø48'W (3,360,000),2- Gull Island,47ø15.8'N, 52ø46.3'W(530,000), 3- GreatIsland, 47ø11 'N, Counc.Bird Preserv.,Cambridge, Eng- land. 52ø49'W (250,000), 4 -- Corbin Island, 46ø57.9'N, 55ø12.8W (100,000), 5 -- Grand Columbier, 46ø49'N, 56ø10'W (100,000), 6 -- Green Island, 46ø52.7'N, 56ø05.3'W MACCARONE, A.D., and W. A. MON- TEVECCHI. 1981. Predation and (72,000), 7 -- Little Fogo Islands, 49ø48'N, 54ø07'W (38,000), 8 -- Middle Lawn, 46ø52.3'N,55ø37.1 'W and Offer Lawn, 46ø51.4,55ø37.3'W (230,000), 9 -- Iron Island, caching of seabirds by Red Foxes 47ø02.5'N,55ø07.5'W(10,000). (Vulpes vulpes) on BaccalieuIsland, Newfoundland. Canadian Field-Nat d 95: 352-353. MONTEVECCHI, W. A., and L M TUCK. 1987. Newfoundland Birds Exploitation, Study, Conservation Publ. Nuttall Ornithol. Club, Cam- bridge,Massachusetts. MORSE, D. H., and C. W. BUCHHEIS- TER. 1979. Nestingpatterns on Matln- icus Rock, Maine. Bird-Banding 50 145-158. NETTLESHIP, D. N. 1980. A guide to the major seabird colonies of eastern : identity, distribution and abundance. Canadian Wildl. Serv, Dartmouth, . PETERS, H. S., and T. D. BURLEIGH 7 1951. The Birds of Newfoundland NewfoundlandDept. Nat. Resources, St. John's. SKLEPKOVYCH, B. O. 1986. The Pred- atory Behaviour and Impact of Red Foxes (Vulpes vulpes)on the Seabird Colonies of BaccalieuIsland, New- foundland. M.Sc. thesis, Memorial University of Newfoundland,Depart- NEWFOUNDLAND ments of Psychologyand Biology,St John's, Newfoundland.

-Departmentsof Psychologyand Biology,Newfoundland Institute for CoM Ocean Science,Memorial Universtry of Newfoundland,St. John's N Newfoundland,Canada, A1B 3X7

Amencan Birds,Spnng 1989