12 9 Abstract—A ship-based line-transect survey was conducted during the Abundance estimates of cetaceans from a summer and fall of 2010 to obtain line-transect survey within the U.S. Hawaiian abundance estimates of cetaceans in the U.S. Hawaiian Islands Exclusive Islands Exclusive Economic Zone Economic Zone (EEZ). Given the low sighting rates for cetaceans in the 1 study area, sightings from 2010 were Amanda L. Bradford (contact author) pooled with sightings made during Karin A. Forney2 previous line-transect surveys with- Erin M. Oleson1 in the central Pacific for calculating 3 detection functions, which were esti- Jay Barlow mated by using a multiple-covariate approach. The trackline detection Email address for contact author: [email protected] probabilities used in this study are the first to reflect the effect of sight- 1 Pacific Islands Fisheries Science Center ing conditions in the central Pacific National Marine Fisheries Service, NOAA and are markedly lower than esti- 1845 Wasp Boulevard, Building 176 mates used in previous studies. Dur- Honolulu, Hawaii 96818 ing the survey, 23 cetacean species 2 Southwest Fisheries Science Center (17 odontocetes and 6 mysticetes) National Marine Fisheries Service, NOAA were seen, and abundance was esti- 110 Shaffer Road mated for 19 of them (15 odontocetes Santa Cruz, California 95060 and 4 mysticetes). Group size and 3 Southwest Fisheries Science Center Beaufort sea state were the most National Marine Fisheries Service, NOAA important factors affecting the de- 8901 La Jolla Shores Drive tectability of cetacean groups. Across La Jolla, California 92037 all species, abundance estimates and coefficients of variation range from 133 to 72,528 and from 0.29 to 1.13, respectively. Estimated abundance is highest for delphinid species and lowest for the killer whale (Orcinus Twenty-five cetacean species are nificant insight into the occurrence, orca) and rorqual species. Overall, known to occur in the U.S. Hawaiian distribution, abundance, stock struc- cetacean density in the Hawaiian Is- Islands Exclusive Economic Zone ture, and social organization of ceta- lands EEZ is low in comparison with (EEZ). Before the 2000s, most re- ceans in Hawaii waters, the surveys highly productive oceanic regions. search on cetaceans in Hawaii focused were focused primarily on nearshore on humpback whales (Megaptera no- odontocete species associated with the vaeangliae) (e.g., Herman and Antino- main Hawaiian Islands. ja, 1977; Mobley et al., 1999) and spin- In 2002, the Southwest Fisher- ner dolphins (Stenella longirostris) ies Science Center (SWFSC) of the (e.g., Norris and Dohl, 1980; Norris et National Marine Fisheries Service al., 1994) because individuals of these (NMFS) conducted the first Hawaiian species are concentrated (seasonally Islands Cetacean and Ecosystem As- in the case of humpback whales) in sessment Survey (HICEAS), a ship- nearshore waters of the main Ha- based line-transect survey designed waiian Islands. Although there were to estimate the abundance of ceta- studies of rarer or less accessible spe- ceans in the entirety of the Hawaiian Manuscript submitted 6 January 2016. cies, such as the pygmy killer whale Islands EEZ. During the HICEAS in Manuscript accepted 5 December 2016. (Feresa attenuata) and short-finned 2002, 23 cetacean species (18 odonto- Fish. Bull. 115:129–142 (2017). pilot whale (Globicephala macrorhyn- cetes and 5 mysticetes) were encoun- Online publication date: 19 January 2017. chus) (e.g., Pryor et al., 1965; Shane tered, and the abundance of 19 spe- doi: 10.7755/FB.115.2.1 and McSweeney, 1990), more frequent cies (18 odontocetes and 1 mysticete) and directed surveys for a variety of was estimated (Barlow, 2006). These The views and opinions expressed or implied in this article are those of the species were not initiated until 2000 estimates represented the first abun- author (or authors) and do not necessarily (e.g., Baird, 2005; McSweeney et al., dance estimates for most cetacean reflect the position of the National 2007; Baird et al., 2009). Although stocks in Hawaii waters and were Marine Fisheries Service, NOAA. that recent research has provided sig- incorporated in the stock assessment 13 0 Fishery Bulletin 115(2) reports produced by NMFS in ac- cordance with the Marine Mammal Protection Act of 1972 (e.g., Car- retta et al., 2005). Abundance estimates used in marine mammal stock assessment reports are considered outdated af- ter 8 years (NMFS1). Therefore, a second HICEAS was carried out in 2010, as a collaborative effort be- tween the SWFSC and the NMFS Pacific Islands Fisheries Science Center (PIFSC), with objectives, timing, and methods comparable to those of the HICEAS conducted in Pacific Ocean 2002. However, adjustments were made to the data collection protocol for the false killer whale (Pseudorca crassidens) during the HICEAS in 2010—changes that necessitated a separate and specialized abundance analysis for this species (Bradford Figure 1 et al., 2014, 2015). The objective of Locations of cetacean groups (black dots; n=198) sighted by observers on the present study was to estimate systematic line-transect survey effort (fine lines) in Beaufort sea states 0–6 the abundance of the remaining ce- within the U.S. Hawaiian Islands Exclusive Economic Zone (thick black out- tacean stocks encountered during line) during the Hawaiian Islands Cetacean and Ecosystem Assessment Sur- the HICEAS in 2010. Although the vey in 2010. Nine sightings were of mixed-species groups, in which at least resulting abundance estimates are 2 species were seen. The main Hawaiian Islands are shown in gray with a specific to cetacean stock assess- thin black outline. ment in the Hawaiian Islands EEZ, the analytical methods used are ap- plicable to line-transect surveys of cetaceans in other spacing intervals of 85 km. Transect lines for the HI- regions. CEAS in 2010 were placed midway between each of the lines used in 2002 to maximize spatial coverage of the Hawaiian Islands EEZ over the 2 surveys. The survey Materials and methods effort in 2002 was stratified, and a higher density of transect lines occurred within 140 km of the main Ha- Data collection waiian Islands. This stratification was not maintained for the HICEAS in 2010. Therefore, the systematic sur- The HICEAS in 2010 was conducted aboard two 68-m vey effort in 2010 was roughly uniform throughout the NOAA research vessels within the Hawaiian Islands study area. The survey speed of both ships was 18.5 EEZ during the summer and fall (Fig. 1) The study area km/h (10 kt). was surveyed from the NOAA ship McArthur II from Although transits to and from ports and circumnav- 13 August to 1 December 2010 and from the NOAA igations of the Northwestern Hawaiian Islands were ship Oscar Elton Sette from 2 September to 29 October not a part of the systematic survey grid, the observers 2010. The survey design of the HICEAS in 2010 was remained on-effort and followed standard observation similar to that of the HICEAS in 2002 (Barlow, 2006). protocols during these periods. This nonsystematic ef- That is, both surveys were based on a grid of parallel fort differed from effort during periods when the ob- transect lines that provided comprehensive coverage servers were not following standard observation pro- of the study area. These transect lines were the basis tocols—periods that were considered to be off-effort for the daily tracklines of each ship and were oriented (e.g., during inclement weather or diversions from the from west-northwest to east-southeast in order to mini- tracklines). Sightings of cetaceans made during non- mize the effects of dominant regional swells generated systematic effort and off-effort were not applied to by northeasterly to easterly trade winds. The grid used the density estimator (see Eq. 1 later in this section) for the HICEAS in 2002 was established by positioning because those sightings were not detected on the sys- transect lines parallel to a randomly placed baseline at tematic transect lines. However, sightings made dur- ing nonsystematic effort were used in the estimation 1 NMFS (National Marine Fisheries Service). 2005. Revi- of species detection functions because the observation sions to guidelines for assessing marine mammal stocks, 24 protocols did not differ between systematic and non- p. [Available at website.] systematic efforts. Bradford et al.: Abundance estimates of cetaceans within the U.S. Hawaiian Islands EEZ 131 The observation methods used during the HICEAS lected during previous NMFS ship-based line-transect in 2010 were developed by the SWFSC and have been surveys of the eastern Pacific. The estimation of de- in use for the last 3 decades (e.g., Barlow, 2006). To tection functions for the HICEAS in 2002 incorporated summarize these methods, observation teams consisted sightings made throughout the eastern Pacific during of 6 observers who rotated through 3 roles (port and SWFSC surveys conducted from 1986 through 2002, starboard observers and a data recorder) and searched but the sighting pool for the analysis of the 2010 data for cetaceans 180° forward of the vessel by using 25× was restricted to sightings made in the central Pacific binoculars (port and starboard observers) and with un- (defined here as the area of the eastern Pacific north of aided eyes (data recorder) from the flying bridge (ap- 5°S, south of 40°N, west of 120°W, and east of 175°E) proximately 15 m above the sea surface on both ships). during SWFSC and PIFSC surveys from 1986 through When cetaceans were sighted within 5.6 km (3 nmi) 2010. The pooled sightings (collected during both sys- of the trackline by 1 of the 3 on-effort observers, sys- tematic and nonsystematic efforts) were limited to the tematic search effort was suspended and the ship di- central Pacific to minimize heterogeneity resulting verted from the trackline toward the sighting so that from geographical differences in species associations species, species composition (for mixed-species groups), and behavior—complex factors that can be difficult to and group size could be determined.