FISHERY BULLETIN: VOL. 81, NO. 3 



other species are possible. Unfortunately, we are only 

 able to speculate about many aspects of the introduc- 

 tion and subsequent spread of this species in 

 Hawaii. 



METHODS 



Herklotsichthys quadrimaculatus were taken at 

 night with small purse seines in water 12-13 m deep 

 in the southern end of Kaneohe Bay on the island of 

 Oahu. All sets were "blind," i.e., the net was set more or 

 less at random within the general area to be sampled, 

 and no lights were used to attract fish prior to the set. 

 Most herring were taken with a 61 m long by 12 m 

 deep seine of ca. 0.3 mm square mesh. Up to several 

 hundred juveniles and up to 1 2 adults were taken per 

 set. Between September 1974 and December 1976, 

 99 sets were made with this net. This series was 

 initially intended to sample the Hawaiian anchovy. 

 Over most of this period, samples were taken at 2-3 

 wk intervals; the longest interval between sampling 

 was slightly more than a month. Most adult herring 

 were taken with a few exploratory purse seines of a 

 larger (153 m long by 11 m deep), coarser mesh (ca. 

 25 mm stretch mesh netting); this purse seine was 

 first tested in May 1975 and set at about monthly in- 

 tervals between December 1975 and December 

 1976. Other sporadic samples were taken with both 

 nets during 1977-79. 



Herklotsichthys quadrimaculatus were found only in 

 shallow water (<3-4 m) during the day and were 

 collected by a variety of methods from several 

 locations on Oahu. Juveniles were readily captured 

 by small beach seines or cast nets over reef flats in 

 Kaneohe Bay; however, to obtain adults from all 

 seasons of the year, particularly when they were ab- 

 sent from night purse seine collections, we used 

 specimens from beach seining on exposed coasts on 

 the northeast side of Oahu as well as on reef tops in 

 Kaneohe Bay, from baited hook or three-pronged 

 hook "snag" fishing in harbors at Haleiwa and in 

 Kaneohe Bay, and from bait captured in Pearl Har- 

 bor by a skipjack tuna vessel. 



In addition to our collections for biological analyses, 

 we have also examined specimens caught by others 

 from a wide variety of habitats and locations on Oahu 

 in order to confirm that//, quadrimaculatus was pres- 

 ent all around the island. We have very few data on 

 the occurrence of H. quadrimaculatus elsewhere in 

 the archipelago and thus cannot be certain that our 

 observations apply to all Hawaiian Islands. 



Inshore areas in southern Kaneohe Bay were sur- 

 veyed visually by day, and the presence of gold spot 

 herring schools, their approximate size composition, 



and approximate numbers noted. Surveys were 

 made at nearly weekly intervals from June 1978 to 

 September 1979; prior to this time they were made 

 sporadically and unsystematically. 



Most fish were immediately preserved and held in 

 ca. 4% formaldehyde seawater solution. Those fish 

 used for age determination were frozen until otoliths 

 were removed. Standard length (SL) was measured 

 to the nearest mm and wet weight (ww) to the nearest 

 0.01 g after blotting dry. Dry weight was measured to 

 the nearest 0.01 mg after 24 hat60°C. Stomach con- 

 tents of the fish were removed before drying. Sex was 

 determined for all adults and a sample of about 200 

 juveniles 50-80 mm SL. Rough estimates of stomach 

 fullness and types of prey eaten were recorded from 

 390 fish. When desired, gonads were dried and 

 weighed separately. The gonad/somatic weight ratio 

 (G/S) was calculated from the dry weight of the 

 gonads and dry weight of the fish excluding gonads 

 (dw s ), rather than total dry weight of the fish includ- 

 ing gonads (dw t ). 



Portions of the ovaries of mature-sized females 

 from each month's sample were examined, and 

 diameters of several of the largest ova measured to 

 the nearest 0.02 mm with an ocular micrometer. For 

 20 females with well-developed ova, a section of the 

 ovary containing about 200 ova >0.20 mm was 

 teased apart and all ova >0.20 mm were measured. 

 Preliminary data indicated no differences in ova den- 

 sity or size composition within or between ovaries of 

 the same fish. The examined sections and remaining 

 portions of the ovaries were dried and weighed 

 separately, and the total number of ova in the largest 

 mode was calculated from the weights of the section 

 and remaining ovary and the ova count from the sec- 

 tion. For 26 other females with a clearly separated 

 mode of large ova, all ova in that mode were counted 

 from both ovaries. 



The growth increments or rings on sagittae of fish 

 less than ca. 50 mm SL faded within a few hours to 

 days after dissection; consequently, counts were 

 made as soon as possible after removal from a fresh 

 or thawed specimen. The sagittae were mounted in 

 glycerine and rings of each counted at least three 

 times under 100-400X magnification. Generally 

 otolith rings from fish of this size were easy to dis- 

 criminate and the repeated counts rarely disagreed; 

 however, if a consistent count could not be made, the 

 data were discarded. 



The denser otoliths from fish >50 mm SL were 

 allowed to air dry and then mounted in EuparoP. The 



'Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



588 



