FISHERY BULLETIN: VOL. 72. NO. 2 



100,000 



10,000 



cr 



a! 1,000 



(T 

 liJ 



m 



s 



100 



n — I — I — I — I — I — I — I — I — I — I — I — I — I — I — r 

 EGGS 



Q 



J F M A M J J 

 1970 



-!••*•* 



JI 



ASONDJ FMAMJ 

 1971 



J LLl 



J FMAMJ 

 1970 



D J F M A M J 

 1971 



Figure 14. — Seasonal variation in abundance of Pleurobrachia eggs and larvae from 8 March 1970 to 2 

 June 197 1 at station 5 located 1.6 km offshore at Scripps Institution. The solid line connects the mean of 

 the replicate samples for each sampling date. The vertical bar represents the range of values for the 

 replicate samples, and solid symbols indicate that the range is less than the size of the symbol; open 

 symbols without any vertical bar are single observations. 



represent the offshore movement of the high 

 ctenophore abundance while field sampling was 

 taking place. Note the secondary peak in abun- 

 dance which was 25 km from shore. Deviations 

 from an exponential decay function are perhaps 

 the result of coastal water eddies (Figure 7), 

 which move offshore with their surface-living 

 species and give rise to offshore aggregations in 

 moderate abundance. An important question to 

 resolve is whether or not the expatriated or ad- 

 vected aggregations are able to survive, grow, 

 and reproduce as well in offshore areas as they do 

 in the coastal waters. 



Seasonal variations in abundance of P. bachei 

 postlarvae, larvae, and eggs on semilogarithmic 

 plot indicated different patterns between these 

 stages in life history (Figures 13 and 14). Recall 

 that egg and larval abundances are based on sam- 

 ples from station 5 only, whereas those of postlar- 

 vae are based on the replicate tows of three to five 

 stations. Note the wide range for the median in 

 abundance of postlarvae (Figure 13), which usu- 

 ally was on the order of one-fifth to five times the 

 overall median. In several instances the values 

 from different stations were closer to each other 

 than they were to their respective replicate sam- 

 ple at the same station. This indicates that spatial 

 heterogeneity within a station on the scale of 100 

 m was often as large as the spatial plus temporal 

 heterogeneity between stations on the scale of 8 



km apart in distance and 1 h apart in time to 32 

 km apart in distance and 5 h apart in time. The 

 95% confidence limits for the mean of replicate 

 samples at one station is the mean multiplied and 

 divided by 2.15 (determined by two-way analysis 

 of variance estimate of the mean square error 

 using 90 pairs of replicate samples at five stations 

 from 18 June 1970 to 2 June 1971). The 95% 

 confidence limits for the mean of all stations at one 

 sampling date is the mean multiplied and divided 

 by 6.23 (also determined by the two-way ANOVA 

 referred to above). The ratio of the 95% confidence 

 limits for the mean of all stations at one sampling 

 date to that for the mean of replicate samples at 

 one station is 8.4 (the ratio is equal to 6.23^/2.152). 

 This latter residual variability is comprised of 

 time-dependent physical variations plus spatial 

 variation and is 1.8 times larger than replicate 

 sample error (1.8 = 8.4/2. 15^). The seasonal pat- 

 tern of postlarvae showed high abundances in 

 May to October 1970, low values in Novem- 

 ber-January and moderate densities in Feb- 

 ruary-June 1971. Note that median abundances 

 in March-June 1970 were one or two orders 

 of magnitude higher than the same interval in 

 1971. The seasonal distribution of larval abun- 

 dance was 180^ out of phase with that of postlar- 

 vae for most of 1970 (Figure 14). In 1970 larval 

 numbers were low during the summer maximum 

 of postlarvae and highest in November when post- 



310 



