122 
Fishery Bulletin 107(2) 
average freshwater inflows for much of the 1970s, which 
limited the influence of MSX. However, the fact that 
high abundance continued for at least five years after 
freshwater inflows subsided to more normal conditions 
circa 1979, and the depensation in the abundance-mor- 
tality relationship, would indicate that high abundance 
may reduce the probability of epizootics. This possibility 
has been treated theoretically by Powell et al. (1996), 
who showed that simulated oyster populations under- 
going significant increases in abundance were very 
unlikely to also generate Dermo epizootics. Simulations 
indicate that the oyster population can expand more 
rapidly than Dermo can expand and intensify, when 
the number of recruits is high (Fig. 7). Alternatively, or 
perhaps as an abetting process, the number of infective 
elements in the water column may be reduced below 
the level needed to generate an infective dose because 
of the volume of water filtered by the population at 
high abundance. An infective dose is hypothesized for 
MSX (Ford et al., 1999; Powell et al., 1999), and some 
evidence supports dose-dependency in Dermo (Bushek 
et al., 1997). However, insufficient information on the 
interaction of disease with oyster populations at high 
abundance is available to definitively decipher the rela- 
tionship between parasite and host at high abundance 
because oyster populations at high abundance are now 
rare or nonexistent for study. 
Interpretation and application of the compensatory 
and depensatory portions of the mortality curve de- 
scribed by Equation 6 (Fig. 9) come with a number of 
important caveats. 1) The probability of occurrence of an 
epizootic has increased since 1990 with the replacement 
of MSX by Dermo as the primary disease that produces 
mortality. An increase in frequency may be expected 
because of the greater tolerance of the parasite for low 
salinity (e.g., Ford, 1985; Powell et al., 1996; Ford et 
al., 1999; Ragone Calvo et al., 2001). Thus the ambit 
of oyster population dynamics may be more restricted 
by Dermo than by MSX. 2) The time series contains no 
high-abundance years since the replacement of MSX by 
Dermo circa 1990. Whether a return to high abundance 
is precluded by Dermo is unknown, but the difference 
in transmission dynamics between the two parasites 
(e.g., Ford and Tripp, 1996) and the expanded environ- 
mental range of Dermo in comparison to MSX would 
indicate that this may be the case. 3) Environmental 
conditions have not been constant over the 54 years, 
and environmental change significantly influences the 
chief agents of increased mortality, MSX and Dermo, as 
well as the autocorrelational dynamics of the epizootic 
process (e.g., Soniat et al., 1998). The mortality curve 
integrates environmental and biological dynamics. 4) 
The rise in winter temperature since the 1970s, that 
accelerated after 1990 (Scavia et al., 2002; Nixon et 
al., 2004), may have modified the interaction of disease 
with oyster population dynamics (e.g., Ford, 1996; Cook 
et al.; 1998, see also Hofmann et al., 1995; Powell et 
al., 1996), decreasing the applicability of the pre-1990 
portion of the time series. 5) As abundance declines, a 
greater proportion of the oyster population is found on 
Table 6 
One-year transition probabilities, as well as the fre- 
quency of occurrences of the eastern oyster ( Crassostrea 
virginica) population in each quadrant over the 54-yr 
time series were calculated from the Delaware Bay oyster 
recruitment-mortality distribution (Fig. 12). Median 
recruitment was 1.53x10® and the median mortality 
fraction was 0.127. Arrows indicate trajectories between 
quadrants. Quadrants are defined in Figure 10. 
Quadrant 
1 
2 
3 
4 
1 — > 
0.308 
0.385 
0.077 
0.231 
2 
0.231 
0.385 
0.077 
0.308 
3 — > 
0.143 
0.071 
0.714 
0.071 
4 
0.231 
0.231 
0.154 
0.385 
Frequency of 
occurrence 
0.241 
0.259 
0.259 
0.241 
Number of years 
13 
14 
14 
13 
the medium-mortality beds (Powell et al., 2008). As a 
consequence, the probability of an epizootic begins to 
decline at abundances somewhere above 1 x 10 9 animals. 
Insufficient data are available to determine the trajec- 
tory for extrapolating this curve to lower abundances; 
therefore considerable uncertainty exists regarding the 
implementation of the abundance-mortality curve for 
abundances below 0.8 xlO 9 . 
Mortality and recruitment Both MSX and Dermo reduce 
the energy budget of a host (e.g., Hofmann et al., 1995; 
Ford et al., 1999) and, as a consequence, may reduce 
fecundity. Some empirical evidence exists that disease 
reduces the fecundity of individual oysters (Mackin, 
1953; Barber et al., 1988; Ford and Figueras, 1988; 
Barber, 1996; Paynter, 1996; Dittman et al., 2001). 
One expectation is that fecundity may drop during 
epizootic years. No overall pattern is found between 
recruitment and box-count mortality in Delaware Bay 
(Ford and Figueras, 1988); however, the four massive 
settlement events with spat numbers above 1.5 xlO 10 
occurred during years when box-count mortality was 
very low, quadrant 3, (Fig. 12). Whether this coincidence 
is an independent outcome of two processes responding 
to common environmental and population forces, or 
whether it documents a causative connection, cannot 
yet be determined. 
Data points in the recruitment and box-count mor- 
tality distribution fell into quadrants 1-4 with a 
frequency of 13, 14, 14, and 13 years, respectively 
(Table 6). Such a distribution is expected by chance. 
Cases of high recruitment occur equally often with low 
and high mortality. Despite the seeming randomness of 
the relationship, mean first-passage times are far from 
equivalent across all transitions (Table 7). The high- 
recruitment+low-mortality state is reached from the 
other three quadrants about three times less frequently 
