The important factors of temperature and oxygen concentration have received 

 considerable attention from investigators of fish metabolism. At present, the general nature of 

 the effect of temperature and the importance of temperature acclimation seem clearly established 

 (Ege and Krogh 1914; Wells 1935b, c; Sumner and Lanham 1942; Haugaard and Irving 1943), but 

 observed differences among species make it imperative to analyze the influence of temperature 

 on metabolic rate for each species of fish investigated. As regards the effect of oxygen concen- 

 tration, it would appear that most fish are able to maintain a fairly constant level of metabolism 

 over a wide range in oxygen concentrations (Gaarder 1918; Hall 1929; Keys 1930a, b; Lindroth 

 1940). There is, however, considerable variation among species as regards the critical concen- 

 tration, the point at which the rate of oxygen consumption begins to decline. The role of breath- 

 ing regulation in helping to maintain a constant metabolic rate as the oxygen content falls below 

 normal has been pointed out by several workers (Westerlund 1906; Winterstein 1908; Belding 

 1929; Meyer 1935; van Dam 1938), but breathing movements and rate of oxygen consumption have 

 rarely been nneasured at the same time. 



The effect of aggregation size on metabolic rate in fish has received comparatively 

 little attention since the initial discovery by Schuett (1933) of a "group effect" in goldfish. Essen- 

 tially, he found that groups of four fish consumed less oxygen per unit weight than single fish in 

 the sanne size container. His results received confirmation from Shlaifer (1938) and later Geyer 

 and Mann (1939a, b) reported a "group effect" for the perch ( Perca fluviatilis ). Large groupings 

 of fish have apparently not been investigated in this regard. 



With the notable exception of Hall (1930), alnnost all the lethal values of oxygen reported 

 in the literature pertain to freshwater fish (Chapman 1939: Moore 1942; Wells 1913; Townsend 

 and Earnest 1940). Individual differences in resistance to low oxygen are pronounced, as indica- 

 ted by wide range in lethal oxygen values within a species (Wilding 1939). Furthermore, lethal 

 values of oxygen have been shown to vary with temperature and possibly with the size of the fish 

 (Graham 1949: Keys 1931; Moore 1942). 



Finally, it is apparent that most of the investigations on fish metabolism in general 

 have been concerned with freshwater species and a need exists for more work on marine forms. 



CAPTURE, MAINTENANCE, AND HANDLING OF STOCKS 



Practically all the experimental work was performed between 1950 and 1953 at the 

 Hawaii Marine Laboratory, located on Coconut Island in Kaneohe Bay, Oahu. Baitfish, particu- 

 larly iao, were fairly abundant around the island and could be readily seined in shallow waters. 

 Occasionally fish were attracted to a subnnerged light and caught in a net lifted from below. Ini- 

 tially stocks were kept in large concrete cisterns some distance away and transferred as needed 

 to aquaria at the laboratory. Later a shallow wooden box (fig. 1) was built for keeping large num- 

 bers of fish at the laboratory. Water circulation in this container was efficient and the fish could 

 be easily captured. The fish were fed bread crumbs daily in the early afternoon. 



Although both the nehu and iao are delicate fish, the nehu in particular is highly 

 excitable and suffers hemorrhages and loss of scales from rubbing against the sides of the tajik. 

 Suehiro (1951) mentions this as one of the principal causes of death of the anchovies ( Engraulis 

 japonicus ) and sardines (Sardinia melanostica ) used for bait in the Japanese tuna fishery. By the 

 end of the summer of 1950 it was obvious that the nehu was unsuitable as an experimental animal 

 and attention henceforth was focused alnnost entirely on the iao. 



Stock mortalities were greatest in the summer months, and large nnortalities coincided 

 with severe hot weather, with the water temperature often rising above 29 C. Summer fish, on 

 the whole, behaved more erratically in experiments than winter fish, and seemed particularly 

 susceptible to small increases in tennperature. Several summer experinnents had to be discarded 

 at the outset because of the obviously abnormal behavior of the fish. 



