240 



Fishery Bulletin 91 12), 1993 



modal, it is clear that this distribution would always 

 be bi- or polymodal. With two recruitment events/yr 

 (Fig. 3B), the distribution again is polymodal and would 

 always be so. With ten recruitment events/yr (Fig. 3C), 

 the modes begin to disappear but the distribution still 

 is weakly polymodal with modes at 0-1, 4-5, and 8-9 

 size units. With 100 recruitment events/yr (Fig. 3D), 

 the distribution is unimodal. The general shapes are 

 similar to the distributions in Fig. 2, with slopes that 

 initially are negative (Fig. 3A) switching to positive 

 (Fig. 3C, 3D). A combination of pulsed recruitment, 

 coupled with a decaying exponential growth pattern 

 and low mortality, can lead to size distributions with a 

 wide range of shapes. A panoply of size-distribution 

 shapes can be produced with identical values for Z, K, 

 and S. using different frequencies of recruitment. 



riod of about 3 wk starting in late March 1990, and 

 over a longer period in spring 1991 (Fig. 4). Timing of 

 settlement was the same for both S. purpuratus and 

 S. franciscanus. A few S. purpuratus settled in June 

 1990, but in terms of influencing the structure of a 

 size-frequency distribution, settlement in 1990 can be 

 considered a single event of short duration. Settlement 

 in 1991 began in late February and continued into 

 early June. The important point is that sea urchin 

 settlement at Scripps Pier was seasonal. Settlement 

 at other sites in California, as well as inside and out- 

 side kelp beds, all showed seasonal settlement (Ebert 

 et al, in prep.). 



Discussion 



Observed settlement of 



S. franciscanus and 5. purpuratus 



Barry & Tegner (1990) used their model specifically to 

 address bimodal size structure in red sea urchins 

 S. franciscanus. We disagree with their assumption of 

 continuous recruitment and base this on observed 

 settlement data. 



Starting in late February 1990, we deployed settle- 

 ment collectors at a number of sites along the Califor- 

 nia coast. Wood-handled scrub brushes (model #0115 

 National Brush Co., Aurora ID were used to evaluate 

 temporal and spatial variability in settlement. Brushes 

 were attached as pairs to a line with two pairs per 

 line. The bottom pair of brushes was suspended lm 

 from the bottom, and the second pair was attached 

 ~20cm farther up the rope. Brushes were tended on a 

 weekly basis at sites within the California Bight and 

 in northern California. One of the 

 sites in southern California was off 

 the end of the pier at Scripps Insti- 

 tution of Oceanography, La Jolla. 



Following weekly collection, 

 brushes were placed in a sonic 

 cleaner with seawater for ~3 min to 

 remove animals. Newly-metamor- 

 phosed sea urchins have a diam- 

 eter of -500 m; thus, following soni- 

 cation, the water and sediment in 

 the sonicator were strained through 

 436 m Nitex. Material retained by 

 the screen was then examined us- 

 ing a dissecting microscope, and 

 newly-settled sea urchins were iden- 

 tified and counted. 



Settlement at Scripps Pier in San 

 Diego County was confined to a pe- 



Bi- or polymodal size distributions and pulsed recruit- 

 ment are common in the literature. We examined 69 

 papers that included larval distribution, settlement, 

 or recruitment information for fish, molluscs, anne- 

 lids, bryozoans, crustaceans, and echinoderms. Out of 

 216 species, only 8 could be considered to have con- 

 tinuous recruitment, and of these only five spider crab 

 species (Hines 1982) appeared to have constant re- 

 cruitment; that is, the same number/mo at all seasons. 

 About 98% of the species failed to meet the assump- 

 tion of constant and continuous recruitment made by 

 Barry & Tegner (1990). 



As shown by our simulations, pulsed recruitment 

 produced bimodal distributions that were not transi- 

 tory in the sense that distributions showed bimodality 

 at all times between recruitment events. However, were 

 a population characterized by pulsed recruitment, sam- 

 pling could be done in such a manner that the relative 

 magnitude of Z and K could be deduced from a simple 



20 



10 



4-i ° 



St rongy locen trot us 

 purpuratus 3 



00 



o 



St rongy locen trot us 

 froncisconus 



MAMJJASONDJFMAMJ 

 1990 1991 



MAMJJASONDJFMAMJ 

 1990 1991 



Figure 4 



Settlement of purple and red sea urchins, Strongylocentrotus purpuratus and S. 

 franciscanus, on eight scrub brushes: four suspended lm from bottom and four at 

 1.2m off bottom at Scripps Pier, Scripps Inst. Oceanogr., La Jolla CA (32°52'N). Solid 

 circles indicate animals not identified to species. 



