120 
Fishery Bulletin 107(2) 
and shell fragments without attached live oysters or 
boxes) (Powell et al., 2008); therefore, one possible ex- 
planation for the relationship between broodstock and 
recruitment is that adult abundance increased settle- 
ment success by providing a principal source of clean 
shell. Two avenues of evidence support this idea. First, 
a recruitment-enhancement program initiated in 2005 
strongly indicated that Delaware Bay is not larvae-lim- 
ited, even at low population abundance levels (unpubl. 
data, first author). Clean shell planted at the appropri- 
ate time consistently sustains a settlement rate 5 to 10 
times that for native shell. Second, the relationship be- 
tween adult numbers and recruitment held for the bay 
overall, even though the numbers of animals in various 
regions of the bay varied relatively independently, and 
independently of numbers for the bay as a whole. The 
broodstock-recruitment relationship was nearly identi- 
cal for two key bay areas, the medium-mortality and 
high-mortality beds (Fig. 11), despite widely and inde- 
pendently varying abundances over the time series (Fig. 
5). Different trajectories would have been expected if 
recruitment rate depended upon a stock-wide abundance 
with trends divergent from local peregrinations. 
Broodstock and mortality Epizootics, here defined as 
bay-wide disease-induced mortality events affecting 
greater than 20% of the stock, occurred in about half 
of the years since 1989 (Figs. 4 and 9), but with much 
lower frequency in prior years. Deaths in nonepizootic 
years affected on average around 10% of the stock. 
All but one of the epizootics occurred at abundances 
between 1.5 x 10 9 and 4 x 10 9 . The single outlier occurred 
at just over 10 x 10 9 animals; this is the 1985 MSX epizo- 
otic event that terminated the high-abundance period of 
the 1970s. The remaining events included the relatively 
few MSX epizootics of the 1950s and 1960s and the 
more frequent Dermo epizootics of the 1990s and 2000s. 
The distribution of data points in the four quadrants 
based on information in Figure 9 was 9, 17, 17, and 10 
in quadrants 1, 2, 3, and 4, respectively (Table 4). This 
distribution is unlikely to occur by chance, but barely 
so: P<0.10, P-0.10; P-0.10; P>0.10, for quadrants 1-4, 
respectively (binomial test: p = 0.25, q- 0.75). Note that 
the use of the median mortality of 0.127 to define highl- 
and low-mortality quadrant groups yields a number 
of nonepizootic years in the same quadrants as the 
epizootic years (those with mortalities exceeding 0.20). 
Thus, the high-mortality quadrants include years when 
mortalities were not extraordinarily high. Note also that 
high-abundance years, those with abundance exceeding 
the median of 2.64 xlO 9 , include a few epizootic years 
with abundances near the median. That is, the use of 
