12 



Fishery Bulletin 104(1) 



we model the historic and current consumption rate by 

 humpback whales within waters of northeastern Kodiak 

 Island in order to assess the impact these whales have 

 as predators on local prey populations. 



Materials and methods 



Study area 



The study area encompassed waters of northeastern 

 Kodiak Island, including Chiniak and Marmot Bays 

 (Fig. 1). The study area was divided into four subareas 

 of approximately equal size in order to equalize sampling 

 effort and maximize coverage of the study area. Subar- 

 eas were also used to separate sightings of humpback 

 whales for the purpose of weighting diet composition 

 in relation to prey availability. An additional subarea, 

 including the waters near Woody and Long Islands, was 

 not considered a survey subarea but was designated in 

 the poststudy period for calculating diet composition 

 ("nearshore," Fig. 1). 



Sightings and abundance of humpback whales 



Data on humpback whale sightings were collected during 

 vessel surveys conducted between June and September 

 in 2001 and 2002. Individual whales were identified from 

 photographs of the black and white pigment patterns 

 (and other natural markings) on the ventral surface 

 of their tail flukes (Katona et al., 1979). A humpback 

 whale sighting was defined as a sighting of an individual 

 whale on a single day. Therefore, no whale was counted 

 twice on one day, but may have been counted multiple 

 times during the study period. Humpback whale sight- 

 ings were summed by month and then by subarea for 

 calculation of whale diet (see "Materials and methods" 

 section: "Composition of simulated diets"). 



These sightings and fluke photographs were used 

 in an associated study to estimate current humpback 

 whale abundance within the study area (Witteveen, 

 2003). The estimate determined from this associated 

 study was used in conjunction with historic catch data 

 from the Port Hobron whaling station to estimate his- 

 toric humpback whale abundance. The whaling grounds 

 of Port Hobron encompassed most of eastern Kodiak 

 waters — an area approximately four times that of the 

 study area. To account for the size difference between 

 whaling grounds and the study area, catch values were 

 divided by four under the assumption of a random har- 

 vest throughout the grounds. The prewhaling and cur- 

 rent estimates of humpback whale population size in 

 the study area are 343 individuals (95% CI: 331, 376) 

 and 157 individuals (95% CI: 114, 241), respectively 

 (Witteveen, 2003). 



humpback whales. The diets were simulated because 

 direct observation of humpback whale feeding behavior 

 is rare and, even when observed, cannot produce a pre- 

 cise account of the prey species being eaten. 



Diet A simulated historic target species and was 

 based on the stomach contents of 39 humpback whales 

 harvested at the Port Hobron whaling station from 

 southeast Kodiak waters between 30 May and 9 August 

 1937 as analyzed by Thompson (1940). 



Diet B simulated current target species and assumed 

 no prey selectivity. It was based on the assumption 

 that humpback whales will eat prey of a suitable size 

 (<30 cm) in proportion to the relative occurrence of the 

 prey in areas used by humpback whales. Euphausiid 

 proportions in the diet were based on historic stomach 

 contents and assumed to be constant over time (no cur- 

 rent euphausiid abundance estimate is available). 



Information on seasonal prey availability was collect- 

 ed from mid-water trawl surveys that were conducted 

 within eastern Kodiak waters in July 2001 and from 

 June through September 2002. Multiple passes with a 

 commercial mid-water trawl net with a 22-mm mesh 

 codend liner were made through acoustic scattering 

 layers, ensuring an accurate representation of mid-wa- 

 ter fish composition and occurrence. Species composi- 

 tion, species counts, and fish size were determined for 

 each tow and grouped within the study subareas. Only 

 data from tows conducted during the study period in 

 2001 and 2002 in areas utilized by humpback whales 

 were included in our analysis. Therefore, prey surveys 

 overlapped humpback whale sightings both temporally 

 and spatially. A separate series of acoustic and purse- 

 seine (center panel with a 3.2-mm mesh net) surveys 

 was used to determine prey availability within the 

 nearshore subarea from June through September 2002 

 (Foy*). Prey composition determined by these surveys 

 was assumed to be homogeneous throughout the near- 

 shore habitat within the study area. 



To calculate diet B, the occurrence of fish smaller 

 than 30 cm was determined from the mid-water trawl 

 surveys within each subarea and month for both 2001 

 and 2002. Tow data were first separated by subarea 

 and month. Percent composition of prey species in each 

 tow was calculated by dividing the total number of 

 fish of each species caught by the total number of all 

 fish caught in each tow, excluding species larger than 

 30 cm (Nemoto, 1959) and species that were not previ- 

 ously documented as prey, such as flatfish and other 

 nonschooling fishes (Nemoto, 1957, 1959; Klumov, 1963; 

 Krieger and Wing, 1984, 1986; Perry et al., 1999). 



To calculate diet B for the entire study area, prey 

 proportions were weighted by the number of whales 

 in each subarea. The weighted proportions were then 

 summed across all months and subareas and multiplied 

 by one minus the percentage of assumed euphausiid 



Composition of simulated diets 



Two diets were simulated: one that reflected the historic 

 diet and the other that reflected the current diet for 



■• Foy, R. 2002. Unpubl. data. Fishery Industrial Technol- 

 ogy Center, University of Alaska Fairbanks, Kodiak, AK 

 99615. 



