Growth and mortality of dermo-disease-free juvenile oysters (Crassostrea virginica) at three salinity regimes in an enzootic area of Chesapeake Bay.

ABSTRACT Specific-pathogen-free (SPF) oysters (Crassostrea virginica) were set on shell and deployed on three oyster bars along a salinity gradient in the Patuxent River, MD, to determine growth, time to initial infection by Perkinsus marinus (causative agent for dermo disease), and subsequent mortalities. Initial deployment was in September 2000, during the second year of a 4-y drought. The salinity gradient experienced by these oysters during the drought was 4[per thousand] to 6[per thousand] above normal and provided ideal conditions for proliferation and dissemination of P. marinus. Oysters grew well during the first year, but mortalities rose rapidly during the second year, and reached 97% to 98% at all sites by the end of the second year. Although mean P. marinus body burdens reached levels of 2.3 x [10.sup.7] cells x [g.sup.-1] oyster tissue, MSX disease was also detected at both the lower and midriver sites in 2002, and was probably responsible for some mortalities at those sites. Due to extensive mortalities of the first deployment, we began to monitor a second group of oysters that had been deployed in May 2002. These oysters experienced a first year of extreme drought, followed by a freshet year during 2003. By the end of the study in November 2003, the overall growth of these oysters (64 mm) was similar to that observed for the first deployment (63 mm), but mortalities were much lower (maximum 58% at the downriver site and 0% at the upriver site). Perkinsus marinus infections in feral oysters, which were the probable source of infectious cells for SPF oysters, showed differences among sites before the drought and during the 2003 freshet year, but no site differences were detected during the 2000-02 drought years. Condition Indices (CI) differed between harvest and spawning seasons, and was greater at the upriver site than at the others during the drought. No site differences were detected for CI during the freshet. Results of this study showed that use of SPF juvenile oysters may not reduce infection and subsequent mortality by either dermo or MSX diseases when deployed during multiyear drought conditions, and the use of SPF seed in most areas when such conditions exist will not improve chances of survival. SPF juvenile oysters are more likely to survive and grow to market size if subjected to just one year of drought than oysters subjected to two years of drought, but their ultimate success will remain a function of the salinity regime they experience.

KEY WORDS: Crassostrea virginica, specific-pathogen-free oysters, SPF, Perkinsus marinus, dermo disease, Haplosporidium nelsoni, MSX disease, mortality, growth

INTRODUCTION

Annual landings of oysters (Crassostrea virginica Gmelin 1791) in Maryland averaged 2.5 million bushels (bu) from the mid 1930s to the mid 1980s, but decreased to 73,000 bu during 1993-94 (Judy 1999). In 2000, the Chesapeake 2000 Agreement established a goal to increase Chesapeake Bay oyster biomass 10-fold by 2010. However, with a harvest in Maryland of 56,000 bu for the 2002-2003 season, and only 27,000 bu for 2003-2004 (C. Judy, MDNR, pers. comm.), oyster populations have not made the anticipated progress toward that goal. These low harvests, however, occurred during times of drought and high disease pressure, and many of the oysters that might have been harvested died before reaching market size. The 2004-2005 harvest improved to 72,000 bu because 2003 and 2004 were wet years, and the impacts of diseases were reduced.

Oyster disease mortality varies with salinity, and can be very low during wet years when salinities are low, but it can increase rapidly when drought conditions prevail. Dermo disease in the Patuxent River is routinely detected in late-summer samples collected from the mouth of the river to the upriver extent of oyster populations, although infection intensities and mortalities generally decrease in an upriver direction along the decreasing salinity gradient. Small juvenile C. virginica are less likely to become infected by P. marinus than adults (Ray 1954, Andrews & Hewatt 1957, Burreson & Ragone Calvo 1996), but even the youngest juvenile oyster age classes have been shown to acquire and rapidly die from dermo disease when conditions favor disease transmission and pathogenesis (Ray 1954, Andrews & Hewatt 1957, McCollough et al. 2007).

Although the obligately halophilic protozoan oyster pathogen Haplosporidium nelsoni (MSX) had not been detected in Patuxent River oysters since 1992, it was detected at low levels in midriver oysters during 1999 and 2000 (MDNR 2001), coincident with increasing salinity during four consecutive drought years (1999-2002). Because of the wide salinity range tolerated by P. marinus, and the high salinity requirement of H. nelsoni, similar oyster disease distributions occur in other mesohaline Chesapeake Bay areas as well.

The establishment of Oyster Recovery Areas (ORAs) by the Maryland Oyster Roundtable Oyster Recovery Action Plan (1993) was an effort to manage around disease by designating zones in the Chester, Choptank, Magothy, Nanticoke, Patuxent, and Severn rivers where harvesting was limited, and only disease-free oysters could be planted. Most of these ORA tributaries were once major producers. The Patuxent, for example, had no fewer than 27 oyster packing companies along its shores early in the 20th century, and the 1972 to 1973 harvest exceeded 243,000 bu (Johnson 1988). As recently as 1986, the Patuxent produced 96,000 bu (MDNR 1994), but landings during the 1990s were generally insignificant (MDNR 2001). A rebound was experienced during the 2005-2006 season, however, with landings in excess of 16,000 bu (C. Judy, MDNR, pers. comm.).

A typical ORA consists of up to three regulatory zones. Only oyster seed certified as free of both P. marinus and H. nelsoni (specific pathogen-free: SPF) can be planted in the lower-salinity, upriver zones; whereas the highest salinity zone can receive unrestricted seed. By seeding ORA upper tributary zones only with SPF oysters, excluding other infected shell-stocks, and limiting oyster harvests, it was hypothesized that disease prevalences and mortalities would progressively decrease, as healthy and prolific oyster populations were reestablished (Calvo et al. 1996). Because the availability of SPF seed from natural reproduction is always in doubt, there is increased demand for hatchery-produced SPF seed oysters for planting in upper ORA zones. However, simply planting SPF seed in zones that require it may not prevent subsequent P. marinus infections and mortalities, if resident feral oyster populations harbor and transmit dermo disease to the newly planted SPF seed. Whether dermo disease is transmitted to planted SPF seed, and whether those infections result in subsequent mortalities is projected to depend on environmental water temperatures and salinities that regulate P. marinus proliferation and pathology among infected oysters (Burreson & Ragone Calvo 1996). Because of the limited availability of certified SPF seed oysters, it is imperative that they be deployed in ORAs by strategies that maximize their resource enhancement return. Unlike P. marinus, however, transmission of H. nelsoni is not dependent on the presence of infected oysters. Infective H. nelsoni particles can be widely distributed and move rapidly into upbay areas when salinities are favorable.

In efforts to supplement oyster populations, the Maryland Department of Natural Resources (MDNR) has traditionally moved seed oysters from prime spat producing areas in Eastern Bay, Harris and Broad creeks, Little Choptank River, Honga River and Tangier Sound, where spat counts are often 1,000-3,000 [bu.sup.-1] of shell, into areas where recruitment is low (900 million were moved in 1997; C. Judy, MDNR, pets. comm.). Oyster recovery areas typically have low numbers of recruiting spat (<50 [bu.sup.-1]), but are generally good growing areas. Prior to the implementation of the Oyster Recovery Action Plan, areas now located in ORAs often received much of this transplanted natural seed. Unfortunately, natural seed is often already infected with P. marinus once it grows large enough for transplanting, and thus currently cannot be moved into the regulated upper reaches of ORAs.

The Oyster Recovery Partnership (ORP) is the lead regional organization initiating, coordinating and managing oyster restoration efforts in Maryland. It has focused efforts on planting the best oysters produced by the University of Maryland's Horn Point Hatchery in the best possible areas, to maximize resource enhancement benefits. From 2000-2006, the ORP planted more than 930 million disease-free spat-on-shell on sanctuaries, managed reserves and commercial harvest bars (C. Frentz, Oyster Recovery Partnership, pers. comm.). The sanctuaries and reserves are intended to be left alone until they are capable of producing significant numbers of larvae, with the hope that they will be a source of recruiting spat for nearby areas. Regulated harvests from reserves may also supplement the commercial oyster fishery.

This study was designed to empirically determine the seasonal dynamics of dermo disease transmission to SPF seed oysters planted at three salinity regimes in enzootic estuarine waters, and to determine the effects of environmental parameters and feral oyster disease status on disease transmission rates and subsequent SPF seed oyster survival and growth. It was intended to be conducted at salinities near the long-term average for this portion of the Patuxent, which range from about 8[per thousand] upriver to 14[per thousand] downriver. However, because initial deployments were made in the middle of a 4-y drought, salinities ranged as high as 14[per thousand]Too upriver and >18[per thousand] downriver. These conditions resulted in elevated disease levels and extensive oyster mortalities before the end of the second year and the premature termination of this phase of the study. Although not part of our original design, a second deployment of oysters was made during the second year to compensate for the early loss of the first deployment. This paper reports on the survival and growth of these two deployments. Detailed dynamics of dermo disease acquisition by these and several other deployments of SPF oyster spat are reported separately by McCollough et al. (2007).

METHODS

Laboratory and Field Methods

This study began in mid-2000 with SPF oysters set on sterilized shell in artificial seawater. Approximately one-half million eyed C. virginica larvae were obtained from the University of Maryland Horn Point Hatchery, and were introduced for setting into two sterilized, 2.4- x 0.6-m fiberglass raceways. Each raceway contained 1.5 bu of aged oyster shells (approximately 750) in 175 L of water at a temperature of 27[degrees]C to 28[degrees]C and a salinity of 15[per thousand] to 16[per thousand]. Resulting spat were reared in artificial seawater at 12[per thousand] to 13[per thousand] salinity and fed a mixture of Isochrysis sp., Tetraselmis sp., Thalassiosira sp., and Pavlova sp.

After reaching a mean shell height of 3 mm, spat were deployed in September 2000 at each of three sites along a salinity gradient from Holland Point (HP) in the upper river near Benedict (river km 30.7), to Gatton (GA) in the middle river near Broomes Island (river km 15.8), to Town Creek (TC) in the lower river near Solomons (river km 4.3; Fig. 1). Thus the oysters at TC and GA were 3.4 km closer to each other than those at GA and HP.

Additional larvae were set on shell and spat deployed again in the spring and fall 2001, 2002, and spring 2003 for a total of six deployments. All deployments are described by McCollough et al. (2007), but only fall 2000 and spring 2002 deployments were followed for growth and mortality as described here.

Oysters were contained in stainless-steel-mesh or plastic-mesh (ADPI) trays supported 0.1 m off-bottom on platforms constructed of 10-cm (4-in) PVC pipe that was drilled with 25-mm holes for rapid flooding and drainage when submerged or retrieved. These platforms kept the oysters slightly off-bottom to reduce sedimentation but still close to feral oyster populations. Each platform was secured to the main anchor system by a ground line that traveled vertically along a 6-mm stainless-steel cable that connected the main mooring buoy to a 227-kg anchor. The platform was rigged with a bridle and lift-line, which allowed the platform and trays to be lifted into the boat for de-fouling, oyster size and mortality determinations, and collection of disease diagnostics samples.

[FIGURE 1 OMITTED]

At each of the three sites, a Hydrolab Datasonde 3 Multiparameter Water Quality Datalogger was deployed to monitor temperature and salinity at 2-h intervals. These instruments recorded date, time, temperature (accuracy [+ or -] 0.15[degrees]C) and salinity (accuracy [+ or -] 0.2[per thousand]). Because temperature and salinity are generally considered to be two of the most important driving forces behind the number of animals infected and the severity of P. marinus infections (Andrews & Hewatt 1957, Soniat 1985, Soniat & Gauthier 1989, Chu & La Peyre 1993, Burreson & Ragone Calvo 1996), it was critical that temperature and salinity data were collected in a manner that allowed valid correlations between these variables and the responses of the oysters. The Hydrolab near the Gatton site had been in place since 1995; the others were deployed in 2000.

Oysters were examined monthly. One stainless-steel tray at each site was initially deployed with 100 SPF oysters; the shell height of each live oyster was measured to the nearest millimeter to estimate growth, and new boxes (articulated empty shells) were counted to determine interval mortality before removal. From additional plastic trays containing similar oysters, 30 individuals were selected to assay for P. marinus infections by the sensitive alternative Ray's fluid thioglycollate medium (ARFTM) whole body burden method (La Peyre et al. 2003, Bushek et al. 1994). These oysters were transported to the Cooperative Oxford Laboratory for processing and analyses (McCollough et al. 2007).

Thirty feral oysters from adjacent natural bars were dredged simultaneously with tray collections and examined using the RFTM rectal tissue assay (Ray 1952, Ray 1966) to estimate infection prevalences and relative intensities of local infection pressures emanating from pathogen reservoir oyster populations. Infection intensity was rated on a modified, categorical Mackin scale of 0-7 (Gieseker 2001). Prevalence was calculated as the percentage of oysters infected, and mean infection intensity ("weighted prevalence" of historic authors) was calculated as the average categorical infection intensity of all oysters examined (Bushek et al. 1994). At least annually, both experimental and feral oysters from experimental sites were examined histologically for H. nelsoni infections.

Ten feral oysters from each sample were also analyzed for Condition Index (CI), using the methods of Lawrence and Scott (1982) as modified by Abbe and Albright (2003). To determine CI, whole oysters were cleaned of biota, scrubbed, blotted dry and weighed. They were then shucked, and the soft tissues were dried. Valves were damp-dried and weighed. Whole oyster weight, less empty valve weight, yields shell cavity content weight (g), which closely estimates the internal cavity volume ([cm.sup.3]). CI was then determined by the formula:

CI = Dry Meat Weight (g) x 100/Shell Cavity Capacity (g)

Statistical Methods

Environmental data were analyzed on two time scales. The availability of long-term temperature, salinity, and feral oyster disease levels delineated maximum ranges for all variables, and allowed the results of this study to be viewed within an historical context. Long-term temperature and salinity data were available only for the midriver site where a data-logger had been located since 1995. Other data-loggers were deployed at the up-and downriver sites in 2000, to collect additional environmental data during the study period. These data enabled comparisons of seasonal salinity and temperature differences among sites and years, with up to 372 readings per month.

Analyses during the study period were performed on a seasonal basis, using daily mean values calculated from raw data collected every 2 h by the data-loggers. Because of the water temperature patterns in the Maryland Chesapeake Bay, seasons were defined as follows: winter was December 1 through March 31, and summer was June 1 through September 30. Significant differences in mean temperatures and salinities between seasons, and among sites within a season, were determined by ANOVA and subsequent Duncan's pairwise comparison tests. Because threshold effects on dermo disease transmission for temperature and salinity have been documented (Dungan & Hamilton 1995, Ford 1996), the number of days in each season that daily mean water temperature was either <5[degrees]C or >25[degrees]C were also tracked, along with the mean salinities experienced during those periods. Chi-square analysis was used to test for differences in the number of days beyond a threshold, across years and among sites.

Chi square tests were used to test for differences in cumulative mortalities among sites for the drought and for the freshet periods. For the first deployment, the total number that had died by August 2002 was used; and for the second deployment, the total number that had died by November 2003 was used. Monthly interval proportional mortalities were correlated by simple linear regression with the interval means for P. marinus body burden of experimental oysters, to determine the strength of the relationship.

The original study design was to follow growth and mortality of one deployment of oysters (Fall 2000), under different salinity conditions in the river. However, because of the premature deaths of the oysters in the first deployment, a second deployment was used (Spring 2002) so that the study of growth and mortality could continue. Differences in mean shell heights of oysters in each experimental cohort at deployment and termination were determined by comparison of the confidence interval around the difference of means ([alpha] = 0.05). Because we did not follow individual oysters, we were unable to use ANOVA or t-tests to test whether means among groups were equal. Because we had initial mean sizes and final mean sizes, we used the "difference of means" test to compare differences of means rather than the means themselves. Overall growth between sites and overall growth between years were also compared. Because the study of the first deployment ended in July 2002, comparisons of overall growth between years were made between the two September July time periods. Overall growth between deployments was also compared. By deleting the no-growth months of January and February, the number of months for each deployment was equalized at 16.

Perkinsus marinus infection pressures from feral oysters were inferred and estimated as mean infection intensities among all local feral oysters examined during any month. Single-factor ANOVA and Duncan's pairwise comparison tests were used to test for differences among years and sites, using the infection intensity means for August, September, and October: when P. marinus infection intensities and dispersal abundances generally peak in Chesapeake Bay waters (Ragone Calvo et al. 2003).

Condition index was determined to quantify changes in meat condition over the course of the study. Because CI varies seasonally, two separate seasons were analyzed. The harvest season CI was calculated as the mean condition during October through March, and the spawning season C1 was calculated as the mean condition during July through September. Single-factor ANOVA and Duncan's pairwise comparison tests were used to test for differences among years and sites.

RESULTS

Salinity

Most of this study was conducted under drought conditions. Comparison with long-term data at the midriver Gatton site allowed temperature and salinity at all study sites to be put into historical context back to 1995. Salinity comparisons showed that weekly mean salinities at all sites exceeded those of historical seasonal salinities at GA throughout the study period until the spring freshet of 2003 (Fig. 2). From 1995 through 2003, mean summer salinities at the midriver GA site ranged from 8.9-17.6[per thousand]. Following the relatively fresh summer of 2000, when mean salinity was 10.6[per thousand] at GA, summer and winter mean salinities increased steadily at all sites during the next 72 mo, until the spring freshet of 2003 finally reversed that trend. The mean salinity during Summer 2003 at GA (9.2[[per thousand]) was the second lowest in this record, consistent with the fact that precipitation in the Washington DC area, which includes the Patuxent River drainage, was the second highest on record and river flows into the Chesapeake Bay were third highest (USGS 2003, Blankenship 2004).

During the 1999-2002 drought, salinities at all experimental sites were higher than the averages since 1995. Salinities experienced at the upriver site were more typical of normal salinities for the midriver site (Fig. 2). Mean salinity at HP was 10.8[per thousand] in 2001 and 12.8[per thousand] in 2002, but decreased to 5.9[per thousand] in 2003. The drought caused significant changes in salinity in time and space. Salinities often decline during spring as a result of increased precipitation and snowmelt, but if the declines do not reach the normally low levels, these spring salinities may set the stage for summer epizootic dermo disease patterns. During our study the mean spring salinities were above normal and were similar to the annual means for those entire years. Mean spring salinities during 2000-2002 increased from 9.4[per thousand] to 11.8[per thousand] to 14.7[per thousand], respectively, before declining to 8.9[per thousand] during the Spring 2003 freshet. Each of these spring salinities was significantly different from the others (P < 0.001).

Water Temperature

Results of winter water temperature analyses indicated that temperatures were similar across all years and sites, with the exception of Winter 2001-2002, which was significantly warmer (7.4[degrees]C) than 2000-2001 (3.8[degrees]C) or 2002-2003 (3.3[degrees]C; P < 0.001: Fig. 3). The low temperatures of 2000-2001 were similar to those of 2002-2003, but the duration of low temperatures was longer in 2003 (Fig. 3). No meaningful station differences were detected during any of the winter periods. Although there were apparent differences in the number of days with water temperatures below 5[degrees]C between years (Table 1), chi-square analysis showed none to be significant (P = 0.850).

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Results of summer water temperature analyses indicated that temperatures were similar across years, but different among sites. Comparison of mean temperatures among sites during the four summers of the study showed statistically significant differences in all years (P < 0.001-0.050), but the range of mean summer temperatures (25.2[degrees]C to 27.9[degrees]C) was so small that this difference was probably not biologically significant. Chi-square analysis of the number of days with water temperatures above 25[degrees]C showed no significant difference among years (P = 0.350).

Oyster Mortality

Although disease pressure was already high at all sites during the first year, and SPF spat began to acquire infections rapidly after their September 2000 deployment (McCollough et al. 2007), early infection prevalences and parasite body burdens remained low until September 2001, and no cumulative oyster mortalities exceeded 5% (Fig. 4). From November 2000 to August 2001, P. marinus infections were detected at HP, GA and TC during 8, 10, and 8 mo, respectively, but mean prevalences were only 11%, 22%, and 20% with mean body burdens of <l, 2.5 and 13 cells x [g.sup.-1] wet tissue, respectively.

By October 2001, cumulative mortalities at HP and GA remained below 10%, but rose to 33% at TC (Fig. 4). By January 2002, mortalities were 14% and 22% at HP and GA, respectively, but were 53% at TC. Survivorship leveled through the winter of 2002, but oysters began to die again in May at both GA and TC, and in June at HP. By July 2002, cumulative mortalities reached 89% and 92% at GA and TC, respectively, although just 32% at HP. In August 2002, cumulative mortalities at GA and TC reached 98 and 97%, respectively, and 60% at HP. By October, mortality at HP also reached 97%, ending the growth and mortality study for this deployment.

For the second deployment of experimental SPF oysters in May 2002, mortalities began at GA and TC during the high-salinity fall of 2002, and continued until March 2003 at GA (cumulative mortality = 35%) and May at TC (cumulative mortality = 58%), respectively, after which they ceased at all stations (Fig. 4), coincident with falling salinities (Fig. 2) resulting from persistently high freshwater inflows. Oysters deployed at the upriver HP site showed no mortalities during the 18 mo that they were monitored postdeployment (Fig. 4).

Among experimental oysters deployed during Fall 2000, mean P. marinus parasite body burdens peaked in July 2002 at both TC and HP at 0.9 x [10.sup.6] cells x [g.sup.-1] and 8.1 x [10.sup.6] cells x [g.sup.-1] tissue, respectively, and in August at GA at 2.3 x [10.sup.7] cells x [g.sup.-1] tissue (Fig. 5). Mortalities began to accelerate at GA and TC before the rapid rise in P. marinus body burdens. However, because MSX disease was active at GA (7% in January and 83% in May 2002) and TC (45% in January and 63% in May 2002), it was almost certainly responsible for some early-summer and later mortalities at GA and TC. At HP, however, the rapid rise in P. marinus body burdens occurred before mortalities accelerated, and because H. nelsoni infections were not detected there, HP mortalities were probably the result of P. marinus pathology alone.

Mean P. marinus body burdens in tissues of experimental oysters were correlated with interval percent mortalities for each site (Fig. 6). The two were positively correlated, but the strength of the correlation coefficient was modest ([r.sup.2] = 0.490) because some mortalities occurred when mean P. marinus body burdens were well below levels of 5 x [10.sup.5] to 1 x [10.sup.6] P. marinus cells x [g.sup.-1] tissue that are believed to be lethal (Choi et al. 1989, Bushek et al. 1994, La Peyre et al. 2003). Most of this variance was caused by mortalities at the TC site, where MSX disease was most active, but H. nelsoni infections were also detected in oysters at the GA site. At both locations, MSX disease mortalities almost certainly weakened the derived correlation between experimental oyster mortality and P. marinus infection intensities, because those effects could not be differentiated in our analysis.

[FIGURE 4 OMITTED]

When confounding MSX disease effects were eliminated by considering only mortalities and P. marinus burdens of oysters from the H. nelsoni-free HP site, clear threshold effects were seen. As P. marinus body burdens increased in experimental oysters, interval (monthly) mortalities increased slightly, but remained below 10% through 21 mo of exposure (Fig. 5, Holland Point). When body burdens decreased with falling water temperatures, mortalities generally ceased. As water temperatures rose during June of the second year (2002). mean P. marinus body burdens rapidly increased toward 1 x [10.sup.6] cells x [g.sup.-1] tissue, and mortalities increased in parallel with mean body burdens. When mean P. marinus body burdens remained below 1 x [10.sup.6] cells x [g.sup.-1] tissue (vertical line in Fig. 7) interval mortalities were negligible, and remained below 10% (horizontal line in Fig. 7), but once mean body burdens exceeded 1 x [10.sup.6] cells x [g.sup.-1] interval mortalities exceeded 70% (Fig. 7) and cumulative mortalities increased rapidly to more than 90% (Fig. 4).

Oyster Growth

From September 2000 to July 2002, mean shell height of the oysters increased 67 mm at GA, 63 mm at TC and 58 mm at HP. A statistically significant difference was detected between HP and GA ([alpha] = 0.05), but TC was no different than either (Fig. 8). Linear growth rates were highest early in the study when the oysters were smaller, but they were also high from May through November 2001. Oysters continued to grow after that, but more slowly as oysters became larger and more heavily infected by P. marinus. By October 2001, mean P. marinus body burdens reached 3.48 x [10.sup.5] cells x [g.sup.-1] tissue at HP, 2.27 x [10.sup.5] cells x [g.sup.-1] at GA and 1.32 x [10.sup.5] cells x [g.sup.-1] at TC. Little growth occurred at any site from January to March 2002 (Fig. 8) because of low temperatures (feeding generally ceases at 6-7[degrees]C; Galtsoff 1964). No growth was observed during April 2001, after a normally cold winter, but oysters did grow during April 2002, after an unusually warm winter.

Growth of oysters deployed in May of 2002 was determined until November 2003 (Fig. 8). Because salinities were high during 2002 and disease pressure remained heavy, the oysters at TC grew slowly at first. The greatest initial growth of this year class was at HP, where oysters increased their mean shell height by 49 mm during their first 6 mo. Oysters at GA increased 42 mm, but those at TC averaged only 31 mm.

This pattern changed during 2003, when reduced salinities lowered disease pressures, and downriver growth was greater than at HP, where salinity was sometimes marginal for growth (Fig. 3). Crassostrea virginica generally feeds and grows well at salinities above 7.5[per thousand]; but between 5[per thousand] and 7.5[per thousand], feeding is often sporadic and growth slows (Loosanoff 1953). At salinities below 5[per thousand], growth slows even more and may cease altogether. By the end of the second deployment's sampling period in November 2003, sizes at all sites were much more similar than they had been a year earlier (Fig. 8). Mean shell height was 66 mm at HP, 71 mm at GA, and 64 mm at TC, resulting in mean increases of 63, 67 and 61 mm, respectively. Analysis by differences of means showed no differences in overall growth among sites.

Infection Pressures from Feral Oysters

Dermo disease dynamics among young experimental SPF oysters was largely influenced by environmental conditions during this study. The same was also true for adjacent feral oysters. During the drought of September 2000 through 2002, mean infection intensities of monthly feral oyster samples at HP, GA and TC reached levels as high as 4.20, 4.77, and 4.70, respectively (Fig. 9). During 2003, after the drought ended, the highest mean infection intensities detected at the same sites were 2.20, 2.97 and 2.43, respectively (Fig. 9). From January 2001 to December 2002, overall mean intensities at HP, GA and TC were 2.49, 2.65, and 2.08, respectively; during 2003 these averages decreased to 0.76, 1.27 and 1.06. During the drought years of 2000-2002 there were no differences in mean P. marinus infection intensities of feral oysters among years at any site (all P > 0.05). There were also no differences among sites for any year during drought years or the freshet year of 2003 (all P > 0.05). There were, however, differences between the P. marinus infection intensity means for 2003 and the drought years at HP (1.33 versus 3.69; P = 0.0001), at GA (2.59 versus 4.13; P = 0.002) and at TC (1.91 versus 3.45; P = 0.027).

[FIGURE 5 OMITTED]

Condition Index

Mean condition indices (CI) among feral oysters were highest at HP during 2001 and 2002 (8.1 and 9.6, respectively), and lowest at TC (6.7 and 6.4). CI decreased at both GA and TC from 2001 to 2002, but increased at HP (Fig. 9). During 2003, the CI improved at TC, but not at HP or GA. CI is generally lowest when P. marinus intensities are highest, and CI is generally highest when P. marinus infection intensities are lowest, because infection intensity is positively correlated with temperature due to pathogen proliferation, and CI is negatively correlated with temperature because of summer spawning. Inversely correlated lows for summer CI and highs for dermo disease intensity during 2000-2002 are apparent (Fig. 9). However, during 2003, the summer minima for CI were less pronounced than during previous years, when dermo disease intensities were higher.

[FIGURE 6 OMITTED]

Whereas lower CI may reflect pathological effects, spawning activity almost always results in lower CI as well. Therefore, CI data were analyzed by harvest season (October through March) and spawning season (July through September). Seasonal means during the drought and freshet years are shown in Table 2. The harvest CI at HP during the drought was significantly higher (9.3; P < 0.001) than at GA (7.5) or TC (7.0). There were no differences among sites during the freshet year of 2003 (P = 0.09). When harvest CI at the various sites were compared among drought years there were no differences at HP or TC, but harvest CI at GA during 2002-2003 (5.6) was significantly lower (P < 0.001) than during the two previous years (8.0 and 8.6).

[FIGURE 7 OMITTED]

Because CI was higher at HP than elsewhere during the drought years, there was no difference between the mean harvest CI for drought years and the freshet year (P = 0.80). Significant differences were, however, seen at the more downriver sites. There was a difference between the mean value for drought years (7.0) and the 2003 freshet year (10.5) at TC (P < 0.001). CI during drought years at GA (7.5) also differed from that of the 2003 freshet year (9.3; P = 0.001), but this was because of unusually low CI during the 2002-2003 harvest seasons (mean 5.6).

Spawning season CI were consistently lower than harvest season CI, but there were no differences among drought years at any site (P = 0.39, 0.16 and 0.62 for HP, GA and TC, respectively). Neither were there differences among sites during drought years (P = 0.37), with means at HP, GA and TC of 6.1, 5.5 and 5.8, respectively. During the 2003 freshet year, there were also no differences in CI among sites (P = 0.18).

DISCUSSION

The long-term data set at the midriver site clearly showed that salinities during the first two years of this study (September 2000 to September 2002) were above normal and especially so during the second year. Temperature data indicated that 2002 was the warmest of the years during which salinity was peaking. Warm water temperatures and high salinities provide ideal conditions for the proliferation and dissemination of both P. marinus and H. nelsoni.

Increased dermo disease levels are promoted by warm winters and high summer water temperatures (Ford 1996), and by salinities above 10[per thousand] (Krantz & Jordan 1996), with synergistic effects by both parameters (Chu et al. 1994; Soniat 1996), which are maximized when both are elevated. Small elevations in summer temperatures in the Patuxent River may appear to have little effect on oysters, but these slight increases may result in more days above 25[degrees]C, a critical threshold temperature for lethal pathology among P. marinus-infected oysters (Andrews 1996). Because P. marinus proliferates maximally around 28[degrees]C (Dungan & Hamilton 1995), and is inhibited at 4[degrees]C (Chu & Greene 1989), oysters may show lower mortalities when there are many winter days with water temperatures <5[degrees]C and few summer days >25[degrees]C, especially during times when salinities below 9[per thousand] synergistically depress P. marinus-caused pathological changes (Ragone Calvo & Burreson 1994). When salinity exceeds 12[per thousand] for an extended time (as during this 4-y drought), mortalities are likely to be heavy, regardless of number of days with water temperatures <5[degrees]C or >25[degrees]C (Burreson & Ragone Calvo 1996). Water temperature and salinity at the midriver site, and P. marinus infection intensities of feral oysters from 10-11 bars along the length of the river (Table 1) suggest that in the mesohaline Patuxent River, low water temperatures and/or low salinities over prolonged time periods reduce the severity of P. marinus infections, and probably result in lower mortality of feral oysters on natural bars (Ragone Calvo & Burreson 1994). During the drought of 1999-2002, mean P. marinus infection intensities were near or above 4.0 for four successive years, but decreased to 2.6 during 2003, when salinities fell below average levels (Table 1).

Infected oysters on natural bars were the apparent proximate source of infectious P. marinus cells transmitted to SPF experimental oysters deployed during this study. Natural bars at each of the three deployment locations have been sampled annually since 1995, and monthly since 1997 or 1998, depending on location. Mean P. marinus infection intensity among these feral oysters showed a regular seasonal cycle (Fig. 9) similar to that seen in Virginia (Burreson & Ragone Calvo 1996), with infection intensity generally declining to annual minima during April or May. In 1997 cool temperatures and a low mean spring salinity kept infection intensities relatively low (Table 1). Unlike 1997, however, salinities during 2000-2002 were above average, and dermo disease was detected every month in feral oysters at each experimental site. During this extended drought, monthly infection intensity means of feral oysters on the upriver HP bar were often as high as those at the down-river TC site, and sometimes higher.

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

That feral oyster infection intensities are greatly affected by salinity is further exemplified by the data from 1997 through 2003. During 1997-1999, monthly mean infection intensities of oysters on natural bars were lower upriver than down-river (Fig. 9), a pattern that changed during the 1999-2002 drought. During 1997-1999, mean infection intensity at the low-salinity HP site was 1.08 and was significantly lower than the 2.44 at the higher-salinity TC site (P < 0.01). During 2000-2002, mean infection intensities among feral oysters at HP and TC were both 2.51. The annual maximum infection intensity means during late-summers at HP increased from <2.0 in 1997 to >4.0 by 1999 and remained high through 2002 (Fig. 9). Also during that time, the annual minimum mean spring infection intensities at HP increased from <0.2 in 1997 and 1998 to 1.4 in 2002. In contrast, the annual maximum infection intensities of late summer at TC increased from <2.5 in 1997 to >5.0 by 1999, but then decreased to <4.0 by 2002 (Fig. 9). Minimum spring infection intensities were <1.0 at TC during the entire 6-y period. With increasing salinity, all sites in the river became more favorable for P. marinus proliferation, and reached a point where all stations had similar infection pressures from uniformly high infection intensity levels in all adjacent feral oyster populations. During 2003, as salinities returned to levels more comparable to those of 1997-1998, the first change occurred in the pattern of dermo disease among feral oyster populations in 5 y, as infection intensities returned to levels seen during that earlier time. Infection levels in feral oysters at HP were lower than at GA, and in May there was no detectable dermo disease there (Fig. 9). When dermo disease peaked for 2003 (September), the mean infection intensity among HP feral oysters was about half that of 2002.

Low P. marinus infection prevalences and intensities were generally correlated with low mortalities among feral oysters, but as these measures of infection pressure increased, reflecting intensifying P. marinus infections, so did mortalities. The same pattern was seen in the experimental oysters at all sites. Body burdens in experimental oysters of the first deployment peaked in July 2002 at HP and TC and in August at GA (at a level well above the other sites). Body burdens might have reached even higher levels in September at GA and TC, but there were too few survivors left to examine. Although the salinity-limited MSX disease extended upstream during 2002, and was probably responsible for some of the mortalities among experimental oysters at GA and TC, it is difficult to estimate the proportion of mortality attributed primarily to each disease when both are present, and may co-occur in individual oysters. Early-season mortalities at GA and TC that began in May 2002 and continued through that summer, are more typical of MSX disease than of dermo disease (Ragone Calvo et al. 2003), which generally kills most heavily during August to October, after summer water temperatures have reached 25[degrees]C. However, there is considerable overlap in the seasons during which these two diseases can cause mortality.

Although substantial mortalities occurred at times when mean P. marinus body burdens were less than 1 x [10.sup.6] cells x [g.sup.-1] tissue, some individuals in sampled populations probably harbored infections with intensities above that lethal level. With a mean of 0.5 x [10.sup.6], or even 0.1 x [10.sup.6] cells x [g.sup.-1] a population of 30 oysters could contain several individuals with infection intensities exceeding 1 x [10.sup.6] cells x [g.sup.-1] tissue. Once the sample means are above 1 x [10.sup.6] cells x [g.sup.-l] as shown in Figure 6 and Figure 7, there is a greater likelihood that many individuals will harbor lethal body burdens and die.

Perkinsus marinus abundances in tissues of oysters within a given sample show a clear trend toward steeply increasing mortality rates as mean infection intensities approach and exceed 1 x [10.sup.6] P. marinus cells x [g.sup.-1] tissue. This trend is best shown at HP (where H. nelsoni infections were not detected) by the time series of body burden and mortality (Fig. 5-Holland Point) and the regression of the same data (Fig. 7). These results are consistent with the lethal P. marinus body burden of 1 x [10.sup.6] cells x [g.sup.-1] tissue reported for Gulf of Mexico oysters (La Peyre et al. 2003), and with minimum lethal body burdens of 0.1 1.0 x [10.sup.6] P. marinus cells x [g.sup.-1] tissue inferred from regressions of body burdens on categorical Mackin-scale (Mackin 1962) infection intensities (Choi et al. 1989, Bushek et al. 1994).

For the second deployment of experimental SPF oysters in May 2002, mortalities began at GA and TC in Fall 2002 and continued until March and May 2003, respectively, after which they ceased (Fig. 4). These mortalities may have resulted in largest part from MSX disease effects rather than from dermo disease, because the oysters were only a few months old when they began to die. In September 2002 the prevalence of MSX at TC was 13% for ferals and 55% for second-deployment experimental oysters; at GA the prevalence was 10% and 35%, respectively. Although dermo disease prevalences in experimental oysters at both GA and TC ranged up to 90% at this time, mean P. marinus body burdens were below 100 cells x [g.sup.-1] tissue, which is too low to cause epizootic mortalities. Similar seasonal disease and mortality results are reported among experimental oysters concomitantly impacted by both P. marinus and H. nelsoni infections in the York River, VA (Ragone Calvo et al. 2003). Once the carryover mortalities from 2002 ended among experimental oysters, and after the long cold winter with persistent salinities below 10[per thousand], mortalities ended in Spring 2003. Although MSX disease was detected among oysters in the study area during 2002, and was detected again in January 2003 samples (33% at TC and 13% at GA), it was not detected in May 2003 or after that. Salinities below 10[per thousand] have been shown to inhibit MSX, especially when temperatures are above 20[degrees]C (Ford 1985). Low winter temperatures and extended low salinities were also unfavorable to the proliferation and dissemination of P. marinus, resulting in low infection prevalences in samples, and sublethal body burdens in individual experimental oysters.

Differences in growth of experimental SPF oysters among sites appeared to be less affected by disease during the first deployment than during the second. Growth of oysters deployed in September 2000 was similar at all sites until the second year when growth at GA began to show clear differences from growth at HP during a time when disease pressure was high at all sites (Fig. 8). However, during the second deployment, early disease pressure from a combination of dermo and MSX diseases was higher downriver than upriver, as indicated by mortalities (Fig. 4). By November 2002, HP oysters had reached 52 mm in mean shell height compared with only 34 mm at TC and 45 mm at GA (Fig. 8). A year later, after disease pressures subsided, the mean shell height of GA oysters had reached 71 mm, compared with 66 mm and 64 mm at HP and TC, respectively. Growth of experimental oysters at HP during 2003 was slowed by suboptimal salinity during much of the time (Fig. 2). From November 2002 to November 2003, the mean shell height of oysters at TC increased 30 mm, whereas oysters at HP increased only 14 mm.

Condition indices of oysters in the Chesapeake Bay also display a cyclical pattern (Haven 1962, Abbe & Sanders 1988, Riedel et al. 1998). Highest CI occur in late fall and winter, as oysters store metabolic glycogen reserves, and lowest levels occur in late summer after spawning is completed (Fig. 9). However, low CI may occur at any time of year, possibly indicating a disease problem or unfavorable environmental conditions (Haven 1962, Scott & Middaugh 1978, Scott & Vernberg 1979, Scott & Lawrence 1982, Abbe & Sanders 1988). Numerous studies have shown that infection by P. marinus and/or H. nelsoni can reduce filtration rates and affect both oyster growth and condition index (Andrews 1961, Newell 1985, Barber et al. 1988, Crosby & Roberts 1990, Gauthier et al. 1990, Paynter & Burreson 1991, Kennedy 1996, Paynter 1996), although increased infection intensity may not always result in reduced CI (Newell et al. 1994). P. marinus may also have a negative effect on oyster reproduction depending on the time of year (Dittman et al. 2001).

Monthly CI of feral oysters has been determined since 1997 at TC, since 1998 at GA and since 1986 at HP (Abbe & Sanders 1988, Riedel et al. 1998). Even though mean intensities of dermo disease were high at all sites from 1999 through 2002, CI was generally higher at HP than at TC, and also generally lower during the first two years of this study than in 1997 (Fig. 9), when environmental conditions were better. With improving environmental conditions in 2003, CI also improved in many cases.

That CI was highest upriver was not surprising, because of salinity and infection pressure, although high, were below those at downriver sites. Of interest, however, was the fact that CI decreased at GA and TC from 2001 to 2002 but increased at HP, indicating that survivors on the HP natural oyster bar were in relatively good condition at a time when nearby tray-held experimental oysters were suffering near total mortality. The rapid decline in CI at HP from May (13.0) to August 2002 (6.0) may have resulted from a combination of spawning activity and disease. However, the fact that condition began to improve in September, and reached 10.3 by November 2002 (Fig. 9) indicated that dermo disease did not dominate physiological conditions of larger older oysters. If it had, CI may have continued to decline as in 1986-1988 (Abbe & Sanders 1988) during a similar 4-y drought. High mortalities among SPF oysters were not unexpected, considering the environmental and related disease conditions that prevailed, but they were higher than anticipated. Lower mortalities and improved CI among feral oysters at HP, however, contrasted with mortalities of the smaller experimental SPF oysters, suggesting possibly some improved tolerance to P. marinus by these older oysters.

Management Implications

The Maryland Oyster Roundtable (1993) developed a series of strategies to reduce disease pressure and mortality in upper ORA zones through the use of SPF oysters, the exclusion of infected shellstocks and limited oyster harvesting (Calvo et al. 1996). One of the goals of our study was to determine if the use of SPF seed oysters in an upper ORA could actually result in such a reduction in disease and mortality. However, under the high-salinity conditions experienced during 2000-2002, survival and disease performances of SPF seed oysters were poor. Oysters became infected early, although infection prevalences and intensities (body burdens) were low during the first year. During the second year, infection prevalences and intensities were high, and most of the oysters died. Although oysters at the upriver (lower-salinity) site lived 2 mo longer than those at the two downriver sites, the additional time did not allow those oysters to grow much larger. There may have been some minimal ecological benefit to this marginally increased survival time, but it is uncertain how great that benefit might have been.

Although feral oysters showed lower levels of dermo disease in late 2003, it is likely that when higher salinity conditions return, as they are sure to do at some time, the same levels of disease detected during 2000-2002 are likely to recur. The numbers of oysters used in this experiment were small (only a few thousand); had they been in the millions, as in the design proposed by the Maryland Oyster Roundtable, the rate of dermo disease acquisition and mortality may have been lower. In some cases, large plantings of disease-free seed oysters appear to have become infected more slowly and developed lighter infections than natural seed (Paynter et al. 2002). However, large-scale plantings of hatchery-produced SPF oysters in high salinity areas are probably not worth the cost and effort. Planting them in typically low salinity areas during times of high salinity may not be a good investment either, unless large-scale plantings fare better than those of our small-scale experiments. However, there may be little alternative, because planting natural seed oysters collected from seed-producing areas in the Chesapeake Bay may only result in continued introduction of diseases into areas where they did not occur historically, and that practice has been discouraged. It also seems likely, even if dermo disease were to virtually disappear in the upper river and SPF oysters were planted exclusively, that when conditions worsened, many of these SPF oysters would rapidly acquire infections from distant endemic sources and die (McCollough et al. 2007).

Alternative solutions to the oyster disease problem could come from several directions. One is for C. virginica to develop some degree of tolerance to disease on its own (which could take a long time). Selective oyster breeding and controlled hatchery production may generate and propagate effective disease-tolerant oyster strains. Another solution may lie in development and use of triploid or other rapid-growth oysters that may reach harvest size before diseases kill them, but applications for some of these more expensive shellstocks may be limited to off-bottom aquaculture applications with reasonable capital recovery probabilities. Off-bottom growth can be two or three times greater than that of oysters on bottom, and there are some small-scale growers in Maryland (and elsewhere) who have produced market size oysters (76 mm) in 12-18 mo. These genetic alternatives are designed primarily for the fishery or for aquaculture, because neither is conducive to long-term oyster population restoration efforts unless a truly disease-resistant self-sustaining oyster is developed and deployed in large numbers. A fourth alternative is the possible introduction of a disease-resistant nonnative oyster such as C. ariakensis, which has shown promise in field trials in Virginia (Calvo et al. 2000, Erskine 2005). However, introductions of exotic species and their associates may face inevitable uncertainties regarding diverse negative potential impacts (National Research Council 2004).

ACKNOWLEDGMENTS

The authors express their gratitude for the efforts of Brian Albright in the initial design and development of this study, his construction of much of the equipment used, and his field sampling (including serving as research vessel captain) and lab work during the first year of this study. Brian passed away in October 2001, following surgery that was complicated by the events of 9/11 in Washington, DC. In earned recognition, the authors award him senior authorship of this paper, for without his conceptual inspiration and leadership, the study would not have been conducted, or the paper ever written. The authors also thank William Yates, Jr. and J. Hixson, III for their efforts in the field and Candace Morrell for her assistance with lab analyses since Brian's death. The authors appreciate the efforts of the many volunteers who cruised with us during this three-year investigation. We thank Dr. Mark Homer (Maryland DNR) for insightful statistical advice. They are also especially grateful for the millions of oyster larvae used during this study provided to them by Dr. Don Merrit and Ms. Stephanie Tobash at the University of Maryland's Horn Point Oyster Hatchery. Lastly they thank the two reviewers whose critical insights have improved this manuscript. Funding for this study was provided by the National Sea Grant Oyster Disease Research Program award numbers NA06RG0101 and NA96RG0489 to the Academy of Natural Sciences and the Maryland Department of Natural Resources.

LITERATURE CITED

Abbe, G. R. & B. W. Albright. 2003. An improvement to the determination of meat condition index for the eastern oyster Crassostrea virginica (Gmelin, 1791). J. Shellfish Res. 22:747-752.

Abbe, G. R. & J. G. Sanders. 1988. Rapid decline in oyster condition in the Patuxent River Maryland. J. Shellfish Res. 7:57-59.

Andrews, J. D. 1961. Measurement of shell growth in oysters by weighing in water. Proc. Natl. Shellfish. Assoc. 52:1-11.

Andrews, J. D. 1996. History of Perkinsus marinus, a pathogen of oysters in Chesapeake Bay 1950-1984. J. Shellfish Res. 15:13-16.

Andrews, J. D. & W. G. Hewatt. 1957. Oyster mortality studies in Virginia 11. The fungus disease caused by Dermocystidium marinum in oysters of Chesapeake Bay. Ecol. Monogr. 27:1-26.

Barber, B. J., S. E. Ford & H. H. Haskin. 1988. Effects of the parasite MSX (Haplosporidium nelsoni) on oyster (Crassostrea virginica) energy metabolism. I. Condition index and relative fecundity. J. Shellfish Res. 7:25-31.

Blankenship, K. 2004. 2003 river flows to the Bay third highest on record. Ches. Bay J. 14(1). Alliance for the Chesapeake, Seven Valleys, PA.

Burreson, E. M. & L. M. Ragone Calvo. 1996. Epizootiology of Perkinsus marinus disease of oysters in Chesapeake Bay, with emphasis on data since 1985. J. Shellfish Res. 15:17-34.

Bushek, D., S. E. Ford & S. K. Allen. 1994. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkinsus marinus infections in the eastern oyster Crassostrea virginica. Ann. Rev. Fish Dis. 4:201-217.

Calvo, G. W., R. J. Fagan, K. N. Greenhawk, G. F. Smith & S. J. Jordan. 1996. Spatial distribution and intensity of Perkinsus marinus infections in oyster recovery areas in Maryland. J. Shellfish Res. 15:381-389.

Calvo, G. W., M. W. Luckenbach, S. K. Allen & E. M. Burreson. 2000. A comparative field study of Crassostrea ariakensis and Crassostrea virginica in relation to salinity in Virginia. Spec. Rep. Appl. Mar. Sci. Ocean Engineer. No. 360. Virginia Institute of Marine Science. 19 pp.

Choi, K.-S., E. A. Wilson, D. H. Lewis, E. N. Powell & S. M. Ray. 1989. The energetic cost of Perkinsus marinus parasitism in oysters: Quantification of the thioglycollate method. J. Shellfish Res. 8:125-131.

Chu, F.-L. E. & K. H. Greene. 1989. Effect of temperature and salinity on in vitro culture of the oyster pathogen, Perkinsus marinus (Apicomplexa: Perkinsea). J. Invert. Pathol. 53:260-268.

Chu, F.-L. E., A. K. Volety & G. Constatin. 1994. Synergistic effects of temperature and salinity on the response of oysters (Crassostrea virginica) to the pathogen Perkinsus marinus. (abstract). J. Shellfish Res. 13:293.

Chu, F.-L. E. & J. F. La Peyre. 1993. Perkinsus marinus susceptibility and defense-related activities in eastern oysters Crassostrea virginica: temperature effects. Dis. Aquat. Org. 16:223-234.

Crosby, M. P. & C. F. Roberts. 1990. Seasonal infection intensity cycle of the parasite Perkinsus marinus (and an absence of Haplosporidium spp.) in oysters from a South Carolina salt marsh. Dis. Aquat. Org 9:149-155.

Dittman, D. E., S. E. Ford & D. K. Padilla. 2001. Effects of Perkinsus marinus on reproduction and condition of the eastern oyster, Crassostrea virginica, depend on timing. J. Shellfish Res. 20:1025-1034.

Dungan, C. F. & R. M. Hamilton. 1995. Use of a tetrazolium-based cell proliferation assay to measure effects of in vitro conditions on Perkinsus marinus (Apicomplexa) proliferation. J. Eukaryot. Microbiol. 42:379-388.

Erskine, A. J. 2005. Virginia Seafood Council 2003-2005 triploid C. ariakensis field project. Waterman's Gazette 31:8-9. Maryland Watermen's Association, Annapolis, MD..

Ford, S. E. 1985. Effects of salinity on survival of the MSX parasite Haplosporidium nelsoni (Haskin, Stauber, and Mackin) in oysters. J. Shellfish Res. 5:85-90.

Ford, S. E. 1996. Range extension by the oyster parasite Perkinsus marinus into the northeastern United States: Response to climate change? J. Shellfish Res. 15:45-56.

Galtsoff, P. S. 1964. The American Oyster: Crassostrea virginica Gmelin. US. Fish. Wildl. Serv. Fish. Bull. 64:1-480.

Gauthier, J. D., T. M. Soniat & J. S. Rogers. 1990. A parasitological survey of oysters along salinity gradients in coastal Louisiana. J. World Aquacult. Soc 21:105-115.

Gieseker, C. 2001. Year 2000 Maryland Oyster Disease Status Report. Maryland DNR technical report FS-SCOL-01-1, Annapolis, MD. 27 pp.

Haven, D. 1962. Seasonal cycle of condition index of oysters in the York and Rappahannock Rivers. Proc. Nat. Shellfish. Assoc. 51:42-66.

Johnson, P. J. 1988. Working the water: the commercial fisheries of Maryland's Patuxent River. Calvert Marine Museum and The University Press of Virginia, Charlottesville. 218 pp.

Judy, C. 1999. Oyster resource is on the rise. Waterman's Gazette 26(4): 7 & 25. Maryland Watermen's Association, Annapolis, MD.

Kennedy, V. S. 1996. The ecological role of the eastern oyster, Crassostrea virginica, with remarks on disease. J. Shellfish Res. 15:177-183.

Krantz, G. E. & S. J. Jordan. 1996. Management alternatives for protecting Crassostrea virginica fisheries in Perkinsus marinus enzootic and epizootic areas. J. Shellfish Res. 15:167-176.

La Peyre, M. K., A. D. Nickens, A. K. Volety, G. S. Tolley & J. F. La Peyre. 2003. Environmental significance of freshets in reducing Perkinsus marinus infection in eastern oysters Crassostrea virginica: potential management applications. Mar. Ecol. Prog. Ser. 248:165-176.

Lawrence, D. R. & G. I. Scott. 1982. The determination and use of condition index in oysters. Estuaries 5:23-27.

Loosanoff, V. L. 1953. Behavior of oysters in water of low salinities. Proc. Natl. Shellfish. Assoc. 1952:135-151.

Mackin, J. G. 1962. Oyster disease caused by Dermocystidium marinum and other microorganisms in Louisiana. Univ. of Texas. Publ. Inst. Mar. Sci. 7:133-229.

McCollough, C. B., B. W. Albright, G. R. Abbe, L. S. Barker & C. F. Dungan. 2007. Acquisition and progression of Perkinsus marinus infections by specific-pathogen-free juvenile oysters (Crassostrea virginica Gmelin) in a mesohaline Chesapeake Bay tributary. J. Shellfish Res. 26:465-477.

MDNR (Maryland Department of Natural Resources). 1994. Maryland Chesapeake Bay Oyster Recovery Areas: Summary of technical information, second draft, February 3, 1994. Prepared for the Maryland Oyster Roundtable by Maryland DNR, Annapolis, MD. 22 pp.

MDNR. 2001. Maryland Oyster Population Status Report, 1996-2000 Fall Surveys. Prepared by Maryland Department of Natural Resources, Sarbanes Cooperative Laboratory and Shellfish Program, Annapolis, MD. 8 pp.

Maryland Oyster Roundtable. 1993. Maryland oyster recovery action plan. Maryland Department of Natural Resources, Annapolis, MD. 26 pp.

National Research Council of the National Academies. 2004. Nonnative Oysters in the Chesapeake Bay. Ocean Studies Board of the National Research Council. The National Academies Press, Washington, DC. 325 pp.

Newell, R. I. E. 1985. Physiological effects of the MSX parasite, Haplosporidium nelsoni (Haskin, Stauber & Mackin) on the American oyster Crassostrea virginica (Gmelin). J. Shellfish Res. 5:91-95.

Newell, R. I. E., K. T. Paynter & E. M. Burreson. 1994. Physiological effects of protozoan parasitism on the eastern oyster Crassostrea virginica: feeding and metabolism. (abstract). J. Shellfish Res. 13:294.

Paynter, K. T. 1996. The effects of Perkinsus marinus infection on physiological processes in the eastern oyster, Crassostrea virginica. J. Shellfish Res. 15:119-125.

Paynter, K. T. & E. M. Burreson. 1991. Effects of Perkinsus marinus infection in the eastern oyster, Crassostrea virginica: Il. Disease development and impact on growth rate at different salinities. J. Shellfish Res. 10:425-431.

Paynter, K. T., T. Koles & D. Meritt. 2002. Perkinsus prevalence in oysters produced in the hatchery and planted in the field. (abstract). J. Shellfish Res. 21:375.

Ragone Calvo, L. M. & E. M. Burreson. 1994. Characterization of overwintering infections of Perkinsus marinus (Apicomplexa) in Chesapeake Bay oysters. J. Shellfish Res. 13:123-130.

Ragone Calvo, L. M., C. F. Dungan, B. S. Roberson & E. M. Burreson. 2003. Systematic evaluation of factors controlling Perkinsus marinus transmission dynamics in the lower Chesapeake Bay. Dis. Aquat. Org. 56:75-86.

Riedel, G. F., G. R. Abbe & J. G. Sanders. 1998. Temporal and spatial variations of trace metal concentrations in oysters from the Patuxent River, Maryland. Estuaries 21:423-434.

Ray, S. M. 1952. A culture technique for diagnosis of infections with Dermocystidium marinum Mackin, Owen and Collier in oysters. Science 116:360-361.

Ray, S. M. 1954. Studies on the occurrence of Dermocystidium marinum in young oysters. Proc. Natl. Shellfish. Assoc. 44:80-92.

Ray, S. M. 1966. A review of the culture methods for detecting Dermocystidium marinum, with suggested modifications and precautions. Proc. Natl. Shellfish. Assoc. 54:55-59.

Scott, G. I. & D. R. Lawrence. 1982. The American oyster as a coastal zone pollution monitor: A pilot study. Estuaries 5:27-33.

Scott, G. I. & D. P. Middaugh. 1978. Seasonal chronic toxicity of chlorination to the American oyster, Crassostrea virginica. In: Ann Arbor Science Publ., Inc., Water Chlorination. Vol. 2. Environmental impact and health effects. Ann Arbor Science Publ., Inc., Ann Arbor, MI. pp. 311-328.

Scott, G. I. & W. B. Vernberg. 1979. Co-occurring chlorine produced oxidants in seawater and their effect on the growth, survival, and physiology of the American oyster, Crassostrea virginica (Gmelin). Evidence for synergistic effects with seasonal temperature stress. In: W. B. Vernberg, F. Thurberg, A. Calabrese & F. J. Vernberg, editors. Marine pollution: functional responses. New York: Academic Press. pp. 415-435.

Soniat, T. M. 1985. Changes in levels of infection of oysters by Perkinsus marinus, with special reference to the interaction of temperature and salinity upon parasitism. N. E. Gulf Sci. 7:171-174.

Soniat, T. M. 1996. Epizootiology of Perkinsus marinus disease of eastern oysters in the Gulf of Mexico. J. Shellfish Res. 15: 35-43.

Soniat, T. M. & J. D. Gauthier. 1989. The prevalence and intensity of Perkinsus marinus from the mid northern Gulf of Mexico, with comments on the relationship of the oyster parasite to temperature and salinity. Tulane Stud. Zool. Bot 27:21-27.

United States Geological Survey. 2003. Flow to Chesapeake Bay in water year 2003 second highest since 1937. US Dept. Interior, USGS, MD-DE-DC Dist., Baltimore, MD. 5pp. http://md.water. usgs.gov/.

BRIAN W. ALBRIGHT, (1) GEORGE R. ABBE, (1) * CAROL B. MCCOLLOUGH, (2) LINDA S. BARKER (3) AND CHRISTOPHER F. DUNGAN (2)

(1) Academy of Natural Sciences Estuarine Research Center, 10545 Mackall Road, St. Leonard, Maryland 20685; (2) Maryland Department of Natural Resources, Cooperative Oxford Laboratory, 904 S. Morris Street, Oxford Maryland 21654; (3) Maryland Department of Natural Resources, 580 Taylor Avenue, Annapolis, Maryland 21401

* Corresponding author: E-mail:[e-mail]

TABLE 1.
The number of days that water temperature was < 5[degrees]C or
> 25[degrees]C and the mean salinity ([per thousand]) during those
periods from 1995-2003 at the midriver site (GA). The number of days
each year with water temperatures < 5[degrees]C may include days from
December of the previous year. Also shown are mean P. marinus
infection intensities of all oyster bars sampled (usually 10-11 bars)
in the Patuxent River in September each year.

                    1995          1996          1997

Temperature      Days   Sal    Days   Sal    Days   Sal

< 5[degrees]C      41   13.7     74   13.0     39    6.6
> 25[degrees]C     94   16.0     95    8.9     78   15.0

Mean Infection      2.81          2.17          1.30
  Intensity

                    1998          1999          2000

Temperature      Days   Sal    Days   Sal    Days   Sal

< 5[degrees]C       3   14.6     20   16.4     29   13.6
> 25[degrees]C     92   15.1     77   14.8     86   10.6

Mean Infection      3.35          4.55          4.28
  Intensity

                    2001          2002          2003

Temperature      Days   Sal    Days   Sal    Days   Sal

< 5[degrees]C      47   14.5     19   17.6     87   13.3
> 25[degrees]C     62   13.7    102   14.6     77    9.2

Mean Infection      4.06          3.97          2.62
  Intensity

TABLE 2.
Mean condition indices of feral oysters at three sites during
harvest and spawning seasons for the drought and freshet years.

        Holland Point   Gatton   Town Creek

                    Harvest Season

Drought      9.3         7.5        7.0
Freshet      9.0         9.3        10.5

                    Spawning Season

Drought      6.1         5.5        5.8
Freshet      7.7         6.0        7.5