Steimie' Population, growth, and production of Echinarachnius parnna 



187 



Mean P:B ratios for E. parma, although generally 

 declining with the age of the cohort (Fig. 9), are com- 

 parable to the range of P:B ratios determined for other 

 echinoderms (Table 3). Because of the variable relation- 

 ship of P:B ratios to the (1) specific population age 

 structure (Fig. 9), (2) segment of the life span, or (3) 

 other factors, the significance of this comparison is 

 uncertain. This problem has attracted the attention of 

 other authors who have attempted to improve the 

 utility of interspecific comparisons of P:B ratios by 

 scaling them to either temperature (Parsons et al. 

 1977), body mass at maturity (Banse and Mosher 1980), 

 or life span (Winberg 1971, Zaika 1972). 



The results could be affected by a number of sources 

 of error or variability that commonly affect similar 

 studies and are usually difficult to assess, e.g., sam- 

 pling error. We did not have samples near the end of 

 the presumed winter-spring peak of larval settlement 

 to obtain accurate estimates of the time and density 

 of larval recruitment. This could affect the growth 

 curves or production estimates. The production of any 

 missed recruits during the few weeks or months before 

 they were sampled would probably be minor because 

 of their small size and high mortality rate. It could 

 influence the cohort P:B-to-age curve (Fig. 9) and par- 

 tially explain the initial low ratio and rise. Later, semi- 

 annual sampling also made it difficult to determine if 



E. parma spawns more than once per year. There are 

 suggestions of substantial recruitment outside of the 

 approximate peak periods (or delayed recruitment) in 

 some of the histograms, especially for the Georges 

 Bank station (Fig. 5). 



Another source of error may be systematic, e.g., an- 

 nual mean production or P:B values do not adequately 

 consider a cyclic annual growth and production pat- 

 terns. These appear to have maximum values in the 

 warmer months, based on the study of Cocanour (1969), 

 when specific rates would be much higher than the 

 mean annual values. 



The final source of error is statistical, especially con- 

 sidering the small sampling sizes available to estimate 

 many of the variables, e.g., the size mode of each 

 cohort. The relatively long time series of data exam- 

 ined, however, allowed a better definition of cohort size 

 modes and growth than is normally possible by at- 

 tempts to intei-pret a limited number of collections over 

 a short time and reduced the necessity for alternate 

 aging estimates. 



