of the millet-seed butterflyfish, Chaetodon miliaris 

 (Pisces: Chaetodontidae), a Hawaiian endemic. M.S. 

 Thesis, Univ. Hawaii, Honolulu. 

 In press. Anomalous growth and reproductive patterns in 

 populations of Chaetodon miliaria (Pisces: Chaetodon- 

 tidae) from Kaneohe Bay, Oahu. Pac. Sci. 

 Randall, J. E. 



1973. Marine parks seen as key to reef beauty. Honolulu 

 Advertiser, Sept. 5. 

 Reese, E. 



1973. Collectors as a threat to reef fishes. Honolulu Star- 

 Bulletin, May 15. 



StRUHSAKER, P., AND J. H. UCHIYAMA. 



1976. Age and growth of the nehu, Stolephonta purpureas 

 (Pisces: Engraulidae) from the Hawaiian Islands as 

 indicated by daily growth increments of sagittae. Fish. 

 Bull., U.S. 74:9-17* 

 Wass, R. 



1967. Removal and repopulation of the fishes on an isolated 

 patch coral reef in Kaneohe Bay, Hawaii. M. S. Thesis, 

 Univ. Hawaii, Honolulu. 



Stephen Ralston 



Zmlogij Department 

 University of Hawaii 

 Honolulu, HI 96S22 

 Present address: 

 College of Fisheries 

 Unii'ersity of Washington 

 Seattle, WA 9S195 



AN EPIBENTHIC SAMPLER USED TO 



STUDY THE ONTOGENY OF VERTICAL 



MIGRATION OF RANDALL SJORDANI 



(DECAPODA, CARIDEA)' 



Pandalus jordani Rathbun, like many other 

 species of pandalid slirimps, undergo regular diel 

 changes in their vertical distribution (Tegelberg 

 and Smith 1957; Alverson et al. 1960; Pearcy 1970, 

 1972; Robinson in press). Little is known, however, 

 about the vertical distribution and diel migrations 

 of larval and juvenile shrimp, or at what stage of 

 the life history vertical migration and benthic 

 existence are initiated. 



Berkeley (1930) found that size or age of larval 

 P. danae increased with increasing depth in a 

 semienclosed embayment in British Columbia. 

 Pearcy (1972) published the only information on 

 day/night differences in benthic occurrence of 

 juvenile P. jordani. Using a plankton net mounted 

 on a beam trawl, he collected more juveniles (<7.0 



'This research was supported in part by Grant No. 04-5-158-2, 

 Office of Sea Grant, National Oceanic and Atmospheric Admin- 

 istration, U.S. Department of Commerce. 



mm in carapace length) near the bottom during 

 day than night. 



In order to sample the water column completely, 

 it was necessary to supplement plankton tows 

 with a discrete, quantitative sample on or just off 

 the bottom. Various methods have been used for 

 this purpose but we thought that all of them were 

 inadequate for the present study. Many epibenthic 

 samplers do not have an opening/closing device 

 and therefore are subject to contamination from 

 the water column above (Russell 1928; Frolander 

 and Pratt 1962; Pearcy 1972; Beardsley 1973). 

 Others are only capable of collecting small sam- 

 ples, in relatively shallow water (Clutter 1965; 

 Macer 1967). In others the opening/closing device 

 seems inefficient or overly complex (Bossanyi 1951; 

 Wickstead 1953; Macer 1967; Hesthagen 1970). 

 Design criteria for the sampler used in this study 

 were: a simple, substrate activated, opening/ 

 closing device capable of quantitatively sampling 

 in depths greater than 150 m and sampling at least 

 500 m' of water with no loss of filtration efficiency. 



Epibenthic Sampler Design 



The epibenthic sampler consists of a sled and a 

 box, to which are attached a plankton net and a 

 substrate-actuated opening/closing device 

 (Figure 1). The frame of the sled was welded from 

 flat steel strap (5.1 x 0.6 cm). The runners 

 (23 X 0.6 cm mild steel plate) are joined across the 

 front by a piece of the same steel bent to conform 

 to the front of the sled. This serves to carry the 

 sled over small obstructions on the seabed and 

 further protect the door of the box when it is in the 

 open position. A bumper bar (5.1 x 0.6 cm) was 

 also fitted across the front of the sled to prevent 

 large obstacles from entering the mouth of the 

 sampler. Two brackets on either side of the sled 

 serve as attachment points for the box. The six 

 different positions allow the box to be positioned 

 from 2.5 to 22.9 cm off the bottom. Two pieces of 

 strap (5.1 X 0.6 cm) were welded along the top of 

 the frame with nine holes to provide various 

 attachment points for the towing bridle. In addi- 

 tion, four pairs of towing points were placed 

 around the front of the frame. 



The box (106.7 x 45.7 x 53.3 cm), made of 3.2- 

 mm mild steel plate, is reinforced in front by steel 

 strap (2.5 x 0.32 cm), forming a lip around the 

 mouth of the box (Figure IB). The box is further 

 reinforced by L stock (2.5 x 0.32 cm) placed 

 around the box 10 cm from the rear edge. Attach- 



994 



