Seabirds in Southeastern Hawaiian Waters
Total Page:16
File Type:pdf, Size:1020Kb
WESTERN BIRDS Volume 30, Number 1, 1999 SEABIRDS IN SOUTHEASTERN HAWAIIAN WATERS LARRY B. SPEAR and DAVID G. AINLEY, H. T. Harvey & Associates,P.O. Box 1180, Alviso, California 95002 PETER PYLE, Point Reyes Bird Observatory,4990 Shoreline Highway, Stinson Beach, California 94970 Waters within 200 nautical miles (370 km) of North America and the Hawaiian Archipelago(the exclusiveeconomic zone) are consideredas withinNorth Americanboundaries by birdrecords committees (e.g., Erickson and Terrill 1996). Seabirdswithin 370 km of the southern Hawaiian Islands (hereafterreferred to as Hawaiian waters)were studiedintensively by the PacificOcean BiologicalSurvey Program (POBSP) during 15 monthsin 1964 and 1965 (King 1970). Theseresearchers replicated a tracklineeach month and providedconsiderable information on the seasonaloccurrence and distributionof seabirds in these waters. The data were primarily qualitative,however, because the POBSP surveyswere not basedon a strip of defined width nor were raw counts corrected for bird movement relative to that of the ship(see Analyses). As a result,estimation of density(birds per unit area) was not possible. From 1984 to 1991, using a more rigoroussurvey protocol, we re- surveyedseabirds in the southeasternpart of the region (Figure1). In this paper we providenew informationon the occurrence,distribution, effect of oceanographicfactors, and behaviorof seabirdsin southeasternHawai- ian waters, includingdensity estimatesof abundant species. We also document the occurrenceof six speciesunrecorded or unconfirmed in thesewaters, the ParasiticJaeger (Stercorarius parasiticus), South Polar Skua (Catharacta maccormicki), Tahiti Petrel (Pterodroma rostrata), Herald Petrel (P. heraldica), Stejneger's Petrel (P. Iongirostris), and Pycroft'sPetrel (P. pycrofti). STUDY AREA AND SURVEY PROTOCOL Our studywas a piggybackproject conducted aboard vessels studying the physicaloceanography of the easterntropical Pacific. Our transectswere Western Birds 30:1-32, 1999 1 SEABIRDS IN SOUTHEASTERN HAWAIIAN WATERS • I:/ / // ß o SEABIRDS IN SOUTHEASTERN HAWAIIAN WATERS concentratedsouth and southeastof the islandof Hawaii (Figure1). While the ship was underway during daylight, at least two personssurveyed seabirdssimultaneously from the flying bridge. Using the strip-transect surveymethod (see Spear et al. 1992a), they countedall seabirdsseen within a strip of given width off the forequarteroffering the best observation conditions.Strip width, determinedas per Heinemann (1981), was 500 or 600 m dependingon heightof the flyingbridge of the threevessels used (14 and 15 m abovesea level).Alternating between observers, we usedhand- held binocularsto scan the outer portion of the transect strip nearly constantlyfor birdsmissed with the unaidedeye. We useda 25 x 150 mm mountedbinocular to aid speciesidentification. We recordedbehavior and, for birdsflying in a steadydirection, recorded flight direction to the nearest 10 ø. We also recordedspecies and numbersof birdsin feedingflocks (>4 birdspursuing prey) that passedwithin 2000 m. Every 0.5 hr we recordedthe ship'sposition, speed, and course.We conducted289 transects,each of 0.5 hr exceptthose terminated when the ship stopped,and calculatedthe area surveyedas the surveyperiod multi- pliedby stripwidth and ship speed. Our effortcomprised 72.1 hrsof surveys over 971.6 km2 duringspring (18 April through27 June),and 71.9 hrs covering1156.8 km2 duringautumn (7 Octoberthrough 20 November), 1984-1991 (Table 1). Surveyswere not conductedin spring 1985 or autumn 1986 and 1988. Each 0.5 hr we also recorded sea-surfacetemperature and salinity, thermoclinedepth and "slope"(see below),wind direction(nearest 10ø), wiredspeed, and distancefrom land. Thermoclinedepth and slope, indices of mixingin the water column,were monitoredwith expendablebathyther- mographs.Thermocline depth is the point where the warm surfacelayer meets cooler water below; i.e., the shallowestinflection point determined from bathythermographplots of temperaturewith depth. Exceptionswere when there was no inflectionpoint, indicatingthat the thermoclinewas at the surface,or there was more than one inflection,when we assumedthat the thermoclinebegan at the depthof the strongestinflection. We measured thermoclineslope as the temperaturedifference between the thermocline and a point 20 m belowthe thermocline. ANALYSES To estimate bird densities,we correctedobserved (raw) numbersfor the effectof flight speedand directionof birdsrelative to the ship'sspeed and course(Spear et al. 1992a; flight speedsfrom Spear and Ainley 1997). Without these corrections,densities from at-sea survey data are usually overestimated,particularly for fast fliers. The correctionalso is required becauseany patterns in bird or ship direction will bias analyses.For example,if birdsflew east and west at the samespeed and in equalnumbers, uncorrectedcounts from a shiptraveling west would show greater numbers flyingeast because the observerwould count more that were flyingeast than west. We calculateddensities for each transectby dividingthe correctedcount by the numberof squarekilometers surveyed and report densitiesas birds SEABIRDS IN SOUTHEASTERN HAWAIIAN WATERS Table 1 SurveyEffort in Hawaiian Watersby Date Year Day Hoursof survey Areasurveyed (km 2) 1984 8 May 6.5 85.9 1984 9 May 2.0 25.8 1984 18 May 5.0 55.8 1984 19 May 1.0 12.8 1984 5 Nov 5.5 79.2 1985 16 Oct 8.5 132.6 1985 30 Oct 8.5 120.8 1986 18 Jun 4.5 60.7 1987 31 May 2.5 39.0 1987 7 Oct 5.5 92.9 1988 4 Jun 5.8 85.8 1988 19 Jun 5.0 83.1 1989 2 May 9.0 143.2 1989 27 Jun 8.5 107.8 1989 17 Nov 7.0 122.1 1989 20 Nov 6.5 105.2 1990 21 Apr 10.8 101.0 1990 7 Nov 5.4 91.8 1990 15 Nov 8.5 134.7 1991 18 Apr 4.0 66.9 1991 25 Apr 7.5 106.6 1991 12 Nov 11.0 194.8 1991 19 Nov 5.5 82.7 Total 8 23 144.0 2128.4 per 100 km2 of oceansurface. For predominantspecies (definition given below),unless noted otherwise, all abundanceestimates pertain to corrected counts,with variance given as one standarderror. We report only raw numbers in accountsof less frequent species.The POBSP calculated abundancedifferently; precluding between-study comparisons of absolute abundance. We recognizedthree speciesgroups: "breeding residents," species that breedon the HawaiianIslands, "nonbreeding residents," species that do not breedon the Hawaiian Islandsbut residein the studyarea as prebreedersor duringthe nonbreedingseason, and "migrants,"species that migrateacross the studyarea when travelingbetween breeding and winteringareas. We dividedthe studyarea into four zones(Figure 1): zone 1, 0 to 73 km from Hawaii; zone 2, 74 to 172 km; zone 3, 173 to 271 km; and zone 4, 272 to 370 km. Respectivelyby zone,we surveyed130.8 (10.8 hrs),289.7 (22.0 hrs), 291.4 (22.3 hrs),and 259.7 km2 (17.0 hrs)of oceansurface duringspring, and 140.9 (9.0 hrs),324.1 (19.9 hrs),362.4 (22.5 hrs)and 329.1 (20.5 hrs)km • duringautumn. During April, May, June, October, and November,we surveyed274.5, 359.7, 337.4, 346.3, and810.5 km• of oceansurface, respectively. SEABIRDS IN SOUTHEASTERN HAWAIIAN WATERS Using the STATA program (STATACorp. 1995), we used multiple- regressionanalyses and Sidak multiple-comparisontests (an improved versionof the Bonferronitest; SAS Institute1985) to compareseabird densitiesby zone, season,and speciesgroup. The sampleunit was one 0.5- hr transect.We alsouse log-likelihood ratio (G)tests to examineproportional differences. We log-transformeddensities to satisfyassumptions of normality (skew- ness/kurtosistest for normalityof residuals,P > 0.05). Becauseno birds were seenduring 87 (30%) of the 289 transects,densities included values of zero. As a result,transformations were calculatedas the log (density+ 1). Experimentationwith different modifications[e.g., log (density+ 0.5)] showedno appreciableeffect of choiceof modificationson probabilities(?). All analysesof variancewere of the log-transformeddensity values. Normal- ity wasnot achievedin all analyses,but least-squaresregression (ANOVA) is consideredto be very robustwith respectto non-normality(Seber 1977, Kleinbaumet al. 1988). Althoughregression analyses yield the bestlinear unbiasedestimator relating density to independentvariables, even in the absenceof normallydistributed residuals, ? valuesat the lower levelsof significancemust be regardedwith caution (Seber 1977). Therefore, to reducethe chancesof committinga Type I error, we assumedsignificance for ANOVAs at ? < 0.02. We includedtwo- and three-orderpolynomials in regressionanalysis to test for curvilinearity. Unlessnoted otherwise,species accounts pertain only to transectdata; i.e., they do not includefeeding flocks seen outside the transectzone. All referencesto the POBSP referto King (1970). Seasonsare definedas in the northern hemisphere. RESULTS Seasonal Distributions We recorded32 species,including 15 speciesof breedingresidents, 11 nonbreedingresidents, and sixmigrants (Table 2). Duringspring, densities of breedingresidents were significantlyhigher than those of nonbreeding resident;those of migrantswere significantly lower (Sidak tests, all ? < 0.01, Figure 2). During autumn, densitiesamong the three groups differed insignificantly(Sidak test, all ? > 0.1). Duringspring, densities of breeding residentswere significantlyhigher, those of migrantssignificantly lower than duringautumn (Sidak tests, df = 287, both? < 0.01, Figure2). Betweenthe two seasons,densities of nonbreedingresidents differed insignificantly (Sidaktest, P = 0.3). During spring,densities were significantlyhigher in zones 1 and 4 than duringautumn