FISHERY BULLETIN: VOL. 84, NO. 4 



using standard methods outlined by Fitch (1951), 

 Jensen (1965), and Collins and Spratt (1969). Each 

 pair of otoliths was read independently by me and 

 another reader, neither knowing the identity of the 

 fish. Our age estimates were in agreement in 80% 

 of the examinations. When differences in ring count 

 occurred, a joint reevaluation was made. 



Total length versus age comparisons were 

 graphed, and regression analysis and F-tests were 

 performed on the log-log transformations. Mean 

 sizes of male and female kelpfish in age classes 

 where differences appeared to occur were tested for 

 significance using i-tests and 2-way ANOVA. Sep- 

 arate regression equations were also calculated for 

 males and females, and ANCOVA was performed 

 to determine whether the distributions were signifi- 

 cantly different (Sokal and Rohlf 1981). 



Seasonal population structure was estimated from 

 collection data taken from February 1981 through 

 January 1983. Kelpfish were grouped in six size 

 classes. Distribution of kelpfish in size classes was 

 analyzed for significant seasonal variations using 

 contingency tables and G-tests (Sokal and Rohlf 

 1981). 



Larval Rearing 



Nine giant kelpfish nests were collected, four in 

 spring 1980 and five in spring 1982, off Santa Cata- 

 lina Island. Both parents of the eggs were collected 

 in three cases when spawning was observed. In six 

 cases, only the male parents, which were guarding 

 the nests, were collected. Eggs were also laid in the 

 laboratory on five separate occasions, but did not 

 hatch normally, apparently because of inadequate 

 dispersion in the nests. 



Algal nests containing eggs were suspended from 

 a glass rod connected to an electric stirring device, 

 simulating wave motion in shallow subtidal habitats 

 (Fig. 3A). This method substantially decreased 

 bacterial and fungal attacks. Parents were not kept 

 with the eggs, as both males and females were 

 sometimes found to eat eggs in the laboratory. Nests 

 were placed in aerated 190 L plastic containers 

 cooled in 1 m deep aquaria of running seawater. 

 Filtered seawater in the containers was replaced 

 every few days. Several eggs were removed daily 

 for examination of development. 



Newly hatched larvae were isolated in lightly 

 aerated 76 L brown plastic containers bathed in 

 large aquaria. Kelpfish larvae were fed laboratory- 

 raised Brachionus plicatilis (marine rotifers) within 

 24 h after hatching. Brachionus plicatilis were cul- 

 tured in high densities of the green flagellate, 



Tetraselmis tetrahele, which was grown in a 

 nutrient-rich medium under constant light, follow- 

 ing methods developed for feeding northern anchovy 

 larvae (Theilacker and McMaster 1971). Brachionus 

 plicatilis, ranging from 0.01 x 0.02 mm to 0.07 x 

 0.20 mm in size, were maintained in the larval kelp- 

 fish containers at concentrations of 10-40/mL. At 

 age 1 wk, kelpfish larvae were changed from closed 

 to open containers of filtered and aerated running 

 seawater, having two 20 x 30 cm panels of 100 ^m 

 mesh. 



After age 2 wk, kelpfish larvae were also fed wild 

 plankton, which primarily contained various devel- 

 opmental stages of the copepod Acartia sp. (92% 

 wet weight) and some barnacle nauplii and cyprid 

 larvae (7% wet weight). Wild plankton were col- 

 lected using a submersible pump attached to a float 

 off the laboratory pier. A light was suspended over 

 the pump and the system connected to an electrical 

 timer. Plankton were filtered through a 335 pm 

 mesh bag into a 190 L plastic container. The con- 

 tainer had a removable inner 100 pm mesh lining 

 and a spillover pipe, retaining only appropriate-sized 

 plankton between the two filter bags (Fig. 4). Best 

 copepod catches were obtained from dusk to 2 h 

 after sunset. Running filtered seawater and an 

 aerator were used to maintain temperature and 

 oxygen levels in the collecting container until the 

 fish larvae were fed the following morning. Den- 

 sities averaged 1-3/mL, which have been shown to 

 support high survival rates in laboratory rearing of 

 other fish larvae (Houde 1973; Hunter 1981). 



When plankton catches were low, giant kelpfish 

 diet was supplemented with cultured Artemia salina 

 (brine shrimp) nauplii. Brachionus plicatilis were 

 discontinued after age 3 wk and plankton continued 

 until age 3 mo. After age 2 mo, diet was supple- 

 mented with frozen adult brine shrimp, Tetramin 2 

 commercial flake food, and live mysids captured 

 from net tows in kelpbeds. 



Ten larvae were removed every 2 d during the 

 first 2 wk of development for measurement and 

 description. After this period, 10 larvae were ex- 

 amined weekly until 2 mo had elapsed. All measure- 

 ments were made on fresh material. Drawings of 

 several stages of larval development were made 

 using a camera lucida and a dissecting microscope. 



Gut contents of three specimens from each weekly 

 sample through age 4 wk were analyzed. While 

 viewing with a dissecting microscope, guts were 

 dissected away from the body and food particles 



2 Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service. NOAA. 



812 



