80 



60 



UJ 



o 



UJ 



o 



CO 

 UJ 



en 

 o 



3 



40 



20 



OL 



1 



.02 .04 .06 .08 .10 

 QUININE SULPHATE {/xg -mr') 



Figure 4. — Fluorescence of various concentrations of 

 quinine sulfate standards read at 10 x gain (lower pair 

 of lines) and 30x gain (upper pair) in 1985 (lower line of 

 each pair) and 1986 (upper of each pair). 



polyethylene tubs augmented with f/2 phyto- 

 plankton nutrients (May 1971; Theilacker and 

 McMaster 1971). One spring and one summer ex- 

 periment were completed at ambient tempera- 

 tures (spring = 17.5°-19°C, summer = 21°-23°C), 

 with one of the "high food" containers in the sec- 

 ond experiment kept at 26°C with an aquarium 

 immersion heater. 



We were unable to maintain absolute high and 

 low food concentrations because of oscillations in 

 the supply of food organisms. Thus, the high con- 

 centration was kept at 3 times the low concentra- 

 tion, although absolute amounts varied. The 

 mean initial food concentrations in the spring ex- 

 periment were 50 jxg C/L (= "low") and 154 (xg 

 C/L (= "high"); in summer, 122 and 394 ^JLg C/L, 

 respectively. Tubs were censused every 2 days 

 through each experiment to determine how much 

 food was uneaten and how many larval fish had 

 died (estimated by counting and removing 

 corpses), and to add fresh food. In the low food 

 containers, it was not unusual to find little un- 

 eaten food, particularly as the larvae grew. Once 

 a week the tubs were emptied and the remaining 

 larvae counted directly. It was clear from this 

 direct census that all dead larvae were not ac- 



counted for by searching for corpses every 2 days 

 because of cannibalism, necrophagy, or decay. 

 Every 4 days a known number of larvae was re- 

 moved; their lengths were measured and they 

 were frozen for future analyses. 



We could not estimate larval ingestion pre- 

 cisely because of the uncertainty in how many 

 fish were alive through a 2-d interval. This prob- 

 lem was exacerbated as the larvae within each 

 tub diverged in size, so that variance in individ- 

 ual ingestion increased. Although 3 times more 

 food was offered in the high food containers, these 

 larvae actually ingested about twice the amount 

 of food as did those in the low food containers 

 (Table 2). This difference was due to better sur- 

 vival in the high food containers, which affected 

 the ratio between available ration and number of 

 larvae. To compensate for this, we routinely har- 

 vested more animals from the high food contain- 

 ers than from the low food containers. 



Freeze-dried animals or tissues were weighed 

 on a Cahn electrobalance. Protein was deter- 

 mined by a method of Dorsey et al. (1977) on an 

 aliquot of tissue homogenized in cold 1 M NaCl. 

 DNA was measured by an ethidium bromide tech- 

 nique (Bentle et al. 1981, as modified by M. S. 

 Lowrey). Basic measures of size — dry weight, 

 protein, and DNA — were strongly and linearly 

 correlated (Fig. 5), so that comparing lipofuscin 

 with any of these measures would give similar 

 patterns. 



Lipofuscin accumulated as the larval fish grew 

 (Fig. 6), but at quite different rates for the 3 spe- 

 cies, grunion accumulating most rapidly (relative 

 to gain in weight) and seabass much the slowest. 



Table 2. — Estimated average ingestion, size, composition, and 

 growth efficiency for 20-day-old, laboratory-reared larval California 

 grunion. Compare with Figure 5. 



'Calculated from measured carbon In Brachionus and Anemia. 

 ^Estimated from literature values. 



411 



