Boggs: Bioenergetics and growth of Engrauhs mordax 



561 



Estimates of actual growth (in length), as corrected 

 for mortality (Table 2), appeared similar to those for 

 1- to 2-year-old anchovy in nature based on otolith 

 increments (Spratt 1975), but were lower than those 

 measured by Hunter and Leong (1981) in the labora- 

 tory. No increase in length was indicated during the 

 low-ration, fast-speed treatment, so no correction was 

 made for mortality. The most rapid length increase 

 (0.069 mm/day) was estimated for the group with the 

 smallest mean size (Table 2). This group underwent the 

 low-ration, slow-speed treatment yet appeared to have 

 a faster rate of growth (in length) than did the high- 

 ration, slow-speed group. This might suggest that the 

 increase in length was size-dependent and somewhat 

 independent of the net energy surplus. 



Mortality was highest during the high-ration treat- 

 ments and lowest in the low-ration and fasting treat- 

 ments, especially at slow speed (Table 1). Fish bumped 

 into the gratings between flume sections during 

 feeding, which may have led to a high rate of injuries 

 resulting in death. The size bias in mortality might have 

 resulted from large fish being less often injured by the 

 gratings. However, the treatment with the smallest 

 sized fish (low-ration, slow-speed) had one of the lowest 

 mortality rates. 



Dry mass 



Estimates of fish dry mass for each sample were ob- 

 tained from mass-on-length regressions (Table 3) by 

 using length estimates corrected for mortality in three 

 feeding treatments (Table 2) and assuming no length 

 change in the other treatments. In all but one case, the 

 final dry mass estimates were significantly greater 

 (high- and low-ration treatments) or lower (zero-ration 

 treatments) than the initial estimates (<-test, P<0.05). 

 In the zero-ration treatments, the loss of dry mass was 

 more than twice as rapid in the fast-speed treatments 

 as it was in the slow-speed treatments (Table 3). Dry 

 mass increased about four times more rapidly during 

 the high-ration treatments than during the low-ration 

 treatments, and dry mass increased more than twice 

 as rapidly during slow-speed treatments as during fast- 

 speed treatments at both rations. Dry mass did not 

 change significantly between the initial and final 

 estimates when fish were fed the low ration and swam 

 at the fast speed, so growth was assumed to be zero 

 (Table 3). 



In three feeding treatments (Table 3) the increases 

 in mass were dependent on estimated increases in 

 length. Although realistic, the corrected length in- 

 creases (Table 2) may not have been accurate. There- 

 fore, conservative estimates of mass increases in the 

 feeding treatments were calculated by assuming that 

 there was no increase in length (Table 3). A signifi- 



cant length-specific increase over initial dry mass was 

 found in the slow-speed, high-ration treatment. 



Caloric growth and 



gross conversion efficiency 



Fat content (Table 3) increased substantially during the 

 high-ration treatments, wherein about 70% of the in- 

 crease in dry mass was fat. Fat increased about twice 

 as much in the high-ration, slow-speed treatment as in 

 the high-ration, fast-speed treatment. No strong trends 

 in percent fat were found in the low-ration treatments. 



The average percent fat of each sample was used to 

 calculate mass of fat and fat-free dry mass per fish. 

 Then the total fish energy content for each sample was 

 calculated (Table 3) by assuming heat of combustion 

 values equal to 29kcal/g for fat-free dry mass and 

 9.227kcal/g for fat (Hunter and Leong 1981). 



The trends in total energy content from all of the 

 treatments (Fig. 4) were consistent with a simple 

 energy budget in which more food energy taken in (R), 

 or less energy required for metabolism (Q), results in 

 more energy available for reproduction (S), growth (G), 

 and fat storage (F): 



R = Q + X + I + S + G + F cal. 



(5) 



Components of the energy budget that were not in- 

 dividually estimated were excretion (X) and digestive 

 losses (I). No spawning occurred during the experi- 

 ments, although gonad mass changed. In the follow- 

 ing analyses, changes in gonad and somatic mass were 

 combined as overall growth (B) and fat storage (F). 

 Metabolism was separated into maintenance metab- 

 olism (Q M ) and the metabolism associated with ex- 

 tracting usable energy from food (SDA). Ninety per- 

 cent of the energy losses during starvation were 

 assumed to be respired (Qm), with excretion account- 

 ing for the remainder. This is consistent with estimates 

 of endogenous excretion given by Durbin and Durbin 

 (1981, 1983). The "metabolizable fraction" of fasting 

 energy losses has been assumed to be as little as 80% 

 (Brett 1973) for energy derived from body protein. In 

 the present study, about one-third to one-half of the 

 calories lost during fasting were derived from protein. 

 Energy losses during fasting were greatest during 

 the fast-speed treatment (28.2cal  g fish wet mass -1 • 

 day -1 )** and lowest during the slow-speed treatment 

 (17.6calg" 1 day" 1 , Fig. 4). Growth (in calg 1 - 

 day -1 ) was greater at both speeds when rations (R) 

 were high than when they were low (Fig. 4). At both 



** Hereafter the unit of mass (g ') refers to fish wet mass. 



