700 



Fishery Bulletin 101(3) 



density. However, we examined the probability of detecting 

 a percentage change in bay-wide egg density over time as a 

 function of the number of beaches sampled by using the 1999 

 bay-wide egg density as the initial value in the time series. 



Results 



When the objective is to monitor egg density within a seg- 

 ment of beach, a sample size of 40 sediment cores is suf- 

 ficient for detecting substantial changes in egg density in 

 the top 20 cm of sediment, but >40 cores would be needed to 

 monitor egg density in the top 5 cm of sediment. Distribu- 

 tions of egg densities were skewed right with median densi- 

 ties of 3 and 275 eggs per core for shallow and deep cores, 

 respectively (Table 1 ). A sample size of 40 cores resulted in 

 a CV of 0.26 for a median density of eggs 0-20 cm deep (Fig. 

 2B). In contrast, about 100 cores would need to be sampled 

 to bring the CV down to 0.3 when .sampling shallow sedi- 

 ment and when egg density was at the median (Fig. 2A). A 

 CV of 0.3 corresponds to a 75% chance of detecting a 509f 

 decline in egg density over 5 years (Fig. 3A) and an 80% 

 chance of detecting a 40% decline over 10 years (Fig. SB). 

 A sample size of 60 shallow cores would result in CV of 0.4 

 for median egg density (Fig. 2A), which would be sufficient 

 for monitoring over 10 years, but not over 5 years. A CV 

 of 0.4 would lead to a better than 85% chance of detecting 

 a 50% decline in density over 10 years (Fig. 3B). Precision 

 and power would improve when sampling higher densities 

 of eggs (Fig. 2). 



At most beaches, observed egg densities within a 100-m 

 segment of beach were not representative of egg densities 

 throughout a larger beach. On only two of the six New Jer- 

 sey beaches examined (South Cape Shore Lab and Reeds) 

 did the observed egg density fall within the interquartile 

 range of beach-wide densities (Fig. 4). On three beaches 

 the observed egg density was greater than all predicted 

 densities, and on one beach observed egg density was less 

 than all predicted densities. 



With egg density at the 1999 level and sampling at 16 

 beaches (i.e. eight beaches per state) distributed throughout 

 the bay, the CV for densities of eggs in 0-20 cm of sediment 

 was 0.26 in May and 0.29 in June (Fig. 5). For densities of 

 shallow eggs, the CV was 0.33 for egg densities in May and 

 0.43 in June. Variability in egg densities among beaches was 

 greater for sampling in June 1999 than in May 1999. 



Discussion 



Eggs in shallow sediment (0-5 cm) consistently yielded 

 lower densities and higher variability than eggs in deep 

 sediment (0-20 cm). A sample size of 40 sediment cores was 

 sufficient for estimating and monitoring density of eggs 

 0-20 cm deep within a 100-m beach segment. However, a 

 larger sample size (260 sediment cores) would be needed 

 for estimating and monitoring density of eggs 0-5 cm deep 

 within a segment of beach. 



Because egg density in a 100-m segment of beach is not 

 necessarily representative of the larger surrounding beach. 



