While testing the species of algae, we decided to 

 include different food levels. If P. lutherii were 

 toxic, then its effects may increase with concentra- 

 tion of the algae given to the larvae. Another 

 source of toxic substances could be the algal 

 metabolites which accumulate in the algal cul- 

 tures. In order to test this, we used two different 

 sources of P. lutherii, a young culture and an old 

 one. Bayne (1965) had observed a slightly better 

 growth of M, edulis larvae when fed P. lutherii 

 from a 4-day-old culture compared with those fed a 

 13-day-old culture. 



Methods 



Adult mussels were stimulated to spawn by 

 raising the water temperature from an ambient of 

 15° C to 22°-24° C. The eggs and sperm from five 

 females and seven males were pooled to give a 

 heterogeneous population of larvae. After 2 days 

 the larvae were placed in the various treatment 

 combinations. In the experiment there were five 

 combinations of algae: a) Isochrysis galbana 

 alone; b) /. galbana plus Thalassiosira pseudo- 

 nana (added after 1 wk); c) /. galbana and P. 

 lutherii throughout, plus T. pseudonana after 1 

 wk; d) a young culture of P. lutherii harvested 

 4-7 days after innoculation, and e) an old cul- 

 ture of P. lutherii harvested 14-20 days after in- 

 noculation. In the mixed algae treatments the 

 two or three species were added in equal pro- 

 portion by cell number. 



There were three feeding protocols used. Cell 

 concentrations were increased gradually over the 

 first week of growth, and although the cell con- 

 centrations changed in each protocol they will be 

 referred to as "levels" here for simplicity. The food 

 levels used were: 1 ) 10,000 cells/ml throughout the 

 experiment; 2) 10.000 cells/ml from day 2 today 4, 

 15,000 cells/ml from day 4 to day 6, and 20,000 

 cells/ml for the rest of the experiment; and 3) 

 .50,000 cells/ml from day 2 to day 4, 100,000 cells/ 

 ml from day 4 to day 6, and 500,000 cells/ml for the 

 rest of the experiment. 



Table L — Analysis of variance on size of My til us I'diilis larvae 

 as related to food treatment. Analysis performed on mean larval 

 length for 6 replicates per treatment combination 



Source of variation 



Mean square 



Food level 



Food type 



Food level  food type 



Residual 



2,477 2 



5,952 9 



307 1 



100-3 



24 7-- 

 59.35-' 

 3 06-- 



There were 6 replications in 1 1 beakers at each 

 of the food type-food level combinations. All beak- 

 ers were held at 15° C. The initial density of the 

 larvae at day 2 was 20 larvae/ml. All beakers were 

 sampled when the larvae were 16 days old and up 

 to 10 larvae were measured from each beaker. 



Results 



The main source of variation in the larval 

 lengths at day 16 was due to the food type, with a 

 smaller but significant portion attributable to the 

 food level and the interaction of these two effects 

 (Table 1). The largest source of variation among 

 the types of food was the difference between the 

 larvae fed only P. lutherii and those fed the other 

 food types (Figure 1). There was slightly better 

 growth with the young P. lutherii as food at the 



200- 



z 

 o 



u 

 u 



i 

 z 



o 



z 



180 



1 60 



140 



120 



I I I 



FOOD LEVEL 



■•P-0 01, 



FlOURE 1,— Mean size of mussel larvae at 16 days when grown at 

 three different algal food levels and on five combinations of algal 

 types: a) Isochrysis galbana only; b)/, galbana and Thalassiosira 

 pseudonana; c) /, galbana, T. pseudonana, and Pavlova lutheni; 

 d) young P, /u(/i(?ni:ande)old P lulheni. Means are based on 10 

 animals from each of 6 replicates at each of the treatment combi- 

 nations. See text for description of food levels used. 



716 



