NOTES 



THE FEEDING HABITS OF TWO DEEP 



SLOPE SNAPPERS, PRISTIPOMOIDES ZONATUS 



AND P. AURICULA, AT PATHFINDER REEF, 



MARIANA ARCHIPELAGO 



The lutjanid snappers belonging to the genus 

 Pristipomoides are among the most prized and 

 valuable commercial fish resources in tropical and 

 subtropical regions of the Pacific Ocean (Polovina 

 and Ralston 1987). These fishes normally inhabit 

 escarpments with high vertical relief. During an in- 

 tensive bottom fish survey conducted at Pathfinder 

 Reef in the Mariana Archipelago, two species of 

 snappers, P. zonatus and P. auricilla, comprised 

 more than 68% of the total catch (Polovina 1985). 

 Depth of capture data on these two species demon- 

 strated overlap in their bathymetric distribution 

 (Polovina et al. 1985; Ralston and Williams 1988). 

 Numerous feeding studies have been conducted 

 on snappers that inhabit shallow (<100 m) water; 

 however, published information on the diets of deep 

 slope snapper species in the tropical Pacific is nearly 

 nonexistent (see review by Parrish 1987). Kami 

 (1973) noted prey items for four species of Pristi- 

 pomoides in Guam with total sample sizes ranging 

 from one to six individuals, and Kluegel (1921) 

 presented information on the diet oiP.filaTnentosus 

 in Hawaii based upon four fish. The present paper 

 examines how two coexisting species, P. zonatus and 

 P. auricilla, partition food resources. With recent 

 efforts to expand and develop commercial fisheries 

 for tropical snappers as well as other deep dwell- 

 ing bottom fishes, there is an increasing need to 

 recognize the resources that support these fishes. 

 The results presented here will therefore be useful 

 for developing fishery management strategies and 

 will lead to a better understanding of the ecology 

 of tropical demersal communities. 



Methods 



Stomach and spew samples from 106 P. zonattis 

 and 72 P. auricilla were collected at Pathfinder 

 Reef during an intensive fishing experiment on 

 10-19 April and 5-7 May 1984. Located in the Mari- 

 ana Archipelago at lat. 16°30'N, long. 143°05'E, 

 Pathfinder Reef is a circular, volcanic pinnacle ris- 



FISHERY BULLETIN: VOL. 86, NO. 4, 1988. 



ing to about 16 m beneath the surface. The fishes, 

 ranging from 30.2 to 44.4 cm fork length (x = 

 38.4 cm, standard deviation = 2.81), were caught 

 with hook and line on hydraulic-powered gurdies. 

 The terminal rig and gurdy specifications are 

 described in Uchida and Uchiyama (1986). While 

 fishing, the vessel was usually allowed to drift over 

 the banks and steep slopes, targeting species in the 

 150-275 m depth range. Our study species were cap- 

 tured mainly in depths between 180 and 220 m. Fish- 

 ing was conducted around the entire perimeter of 

 the bank during daylight hours (Polovina 1986). 



Typical of feeding studies conducted on deepwater 

 species with swimbladders, regurgitation of stomach 

 contents posed serious problems (Bowman 1986). 

 The rapid ascent to the surface forced most of the 

 stomachs to evert and lose an unknown quantity of 

 the contents. Many food items, however, were 

 caught in the throat or gill rakers, picked out, and 

 saved. These items, referred to as spews, and full 

 stomachs were frozen and returned to the labora- 

 tory for analysis. 



Laboratory procedures for examination of the 

 samples were similar to those employed by Harrison 

 et al. (1983). Briefly, food samples were sorted, 

 counted, and identified to the lowest practical taxon. 

 The volume of prey items was measured by water 

 displacement. Fishes were predominantly identified 

 by osteological and external anatomical characters 

 and morphometries. Invertebrates were identified 

 by undigested hard parts and external morpho- 

 logical features including shells for gastropods and 

 exoskeletons for crustaceans. 



To analyze the data, we used percent frequency 

 of occurrence to provide a qualitative picture of the 

 food spectrum and percent volume of prey to de- 

 scribe relative biomass of prey items (Hyslop 1980). 

 Because a high percentage of the diets was com- 

 posed of urochordates, we did not employ any 

 numerical analysis of the prey items. These pelag- 

 ic, colonial tunicates (all Pyrosoma spp.) did not 

 occur in discrete units and were thus difficult to 

 count. 



An interspecific comparison of the two snapper 

 species requires some measure of diet overlap and 

 niche breadth. Diet overlap (Q) was computed by 

 using the formula developed by Morisita (1959) and 



807 



