FISHERY BULLETIN: VOL. 70, NO. 3 



Tank L was provided with two fiber-charcoal 

 filters, each with an electric pump delivering 

 2,800 ml per min; tank S had one of these. 



Aeration was by one airstone in tank L and 

 two in tank S." Oxygen concentrations varied 

 from 4.4 to 7.4 ppm for L and 2.4 to 7.2 ppm 

 for S. 



Illumination was by overhead pink fluorescent 

 lamps, which screened out violet-blue light, found 

 by Perlmutter and White (1962) to be lethal to 

 some fish eggs. Lights were controlled by an 

 automatic switch, turned on from 6 AM to 6 PM 

 each day. 



Each tank had shelter in the form of cords 

 suspended from plastic floats. Forty-four cords 

 30 cm long were attached to a 49- by 15-cm float 

 in tank L; numbers and dimensions of the shelter 

 were proportional in tank S. A floating fiber 

 brush type shelter, 4 cm in diameter by 25 cm 

 long, was in L, and a proportional one was in S. 

 The back, one side, and part of the front of each 

 tank were rendered opaque with black plastic; 

 the front area was 44 by 20 cm in L and propor- 

 tional in S. 



Refuges for young fish were enclosed at the 

 end of each tank by fences consisting of 5-mm 

 diameter plastic rods spaced 3 mm apart. The 

 enclosed area was 14 cm wide in L and propor- 

 tional in S. Aquarium gravel was placed in the 

 bottom of each tank for nesting activity. 



Both tanks were maintained at room temper- 

 ature, which was thermostatically controlled ex- 

 cept that cooling was not available in the sum- 

 mer. Weekly mean temperatures were 24 ±: 2°C 

 for both tanks. 



Fish were fed daily according to a fixed sched- 

 ule (Table 1). Uneaten food, feces, and other 

 detritus were siphoned out daily, and two-thirds 

 of the water mass was changed each week, using 

 tap water brought to tank temperature. Char- 

 coal and fiber in the filters were changed once 

 weekly. 



Populations were counted and weighed every 

 2 months. Since T. macrocephala is a mouth 

 breeder, it was not desirable to handle the fish 



Table 1. — Amounts (grams) of food placed in each tank. 



* Although these numbers may appear to be reversed, 

 they are as used. The oxygen concentrations show that 

 they provided approximately equal aeration. 



more often. Aronson (1949) stated that the 

 mean spawning interval for the species was 15 

 days, so the counting interval was about four 

 brood intervals. Exploitation was performed at 

 the time of counting by removing the tenth or 

 fourth fish for 10% and 25% exploitation rates. 

 Weighing was done by draining fish in a net, 

 placing them in a weighed container with water, 

 weighing the container with fish, and subtracting 

 the tare. 



Exploitation started in December 1967, with 

 42 fish weighing 1,148 g in S and 45 fish weighing 

 1,167 g in L (Table 2). These fish were either 

 survivors or descendants of a shipment of 10 

 adults and 50 young fish received on 10 Febru- 

 ary 1966 by air from the Honolulu Biological 

 Laboratory of the Bureau of Commercial Fish- 

 eries (now National Marine Fisheries Service). 



Although numbers of fish declined at the 10% 

 bimonthly exploitation rate (Figure 1), biomass 

 remained relatively constant (Figure 2) and no 

 evidence of recruitment was observed. To de- 

 termine if recruitment would occur at a lower 

 biomass, the exploitation rate was increased to 

 25% at month 16 and was continued at that rate 

 until the end of the experiment. It will be shown 

 below that recruitment did occur under the high- 

 er exploitation rate. 



RESULTS 

 VITAL PROCESSES 



Significant results from exploitation experi- 

 ments require that recruitment and growth occur 

 during the course of the experiment. It is also 

 desirable to know whether natural (nonfishing) 

 mortality was occurring. Evidence relating to 

 these vital processes can be obtained by exam- 



694 



