HOBSON: FEEDING RELATIONSHIPS OF FISHES 



12/17 X 52.5 = 37.05. The data are tabulated when 

 there are more than a few items in the gut con- 

 tents of a given species. 



In species with a well-defined stomach, usually 

 only stomach contents were analyzed, as materi- 

 als in the intestines generally were too far di- 

 gested for satisfactory analysis. On the other 

 hand, some fishes that do not have a well-defined 

 stomach have much material in their intestine 

 that is suitable for study, and so was included in 

 the analysis. Thus data sometimes are specified as 

 being from stomach contents, but at other times 

 the more general term gut contents is used. 



Transect Counts 



To characterize the fishes inhabiting each of the 

 various inshore habitats described below, 100-m 

 transect lines were established in locations judged 

 typical of each habitat. Twenty-two counts of 

 fishes within 5 m of transect lines at 17 sites rep- 

 resenting five habitat categories (see below) were 

 made between September 1969 and May 1970; at 

 least 1 mo passed between counts in any one 

 habitat. 



Quantifying Relative Abundance 



In the Tables below that present data from the 

 transect counts, the relative abundance of the dif- 

 ferent species is represented by a relative abun- 

 dance index. This is the percentage that species 

 represented of all fishes (individuals) counted 

 along all transect lines in that habitat. 



Assessing Nocturnal Colorations 



A number of species possess distinctive day and 

 night differences in coloration. Earlier (Hobson, 

 1968b), I discussed the problem of distinguishing 

 true nocturnal hues from those elicited as a re- 

 sponse to the diving light — a frequent source of 

 error in literature reports of nocturnal color pat- 

 terns. No color pattern that becomes intensified 

 under the diving light is considered here to be a 

 nocturnal pattern; the vast majority described 

 herein were in fact almost immediately lost when 

 the fish was illuminated. 



Study Area 



The study area extends 7.7 km along the south- 

 western shore of the Island of Hawaii, from 



Keawekaheka Point just north of Kealakekua 

 Bay, to Alahaka Bay, south of Honaunau (Figure 

 1). This is part of what is known as the Kona coast. 

 Except for short stretches of sand and cobble 

 beaches at Napoopoo and Keei, the shoreline is a 

 rough basalt face that drops abruptly into the sea 

 from 2 to 3 m above the water's surface (Figure 2), 

 to a similar depth below. From the base of this face 

 the sea floor slopes down to water depths of about 

 20 to 30 m, about 50 to 600 m from shore, then falls 

 away sharply to much greater depths. Thus, along 

 this coast water less than 20 m deep is limited to a 

 relatively narrow shelf, the outer rim of which 

 provided a convenient natural boundary to the 

 study area (Figure 1). 



Environmental conditions in Kona are remark- 

 ably constant, which greatly aided this study. Sur- 

 face water temperatures ranged from 29°C in the 

 fall to 22°C in the spring, but I noted no marked 

 seasonal variations among the fishes either in 

 their activity or species composition. Conditions 



1 KILOMETER 



NAPOOPOO 



I9°25.0 



' '~3 QjHU 



HONOLULU 



CHART AREA- 



I55°575 



I55°550 



Figure 1. — The study area along the Kona coast, Island of 

 Hawaii. Adapted from C. & G.S. chart 4123. Depth contour in 

 meters. 



917 



