Performance of oysters selected for dermo resistance compared to wild oysters in northern Gulf of Mexico estuaries

The performance of the progeny of eastern oysters Crassostrea virginica from Louisia na selected for resistance to dermo, caused by Perkinsus marinus (referred to as ‘OBOY’) and of wild oysters collected from Louisiana (Calcasieu Lake) and Alabama (Cedar Point, Perdido Pass), USA, estuaries was compared for their potential use in aquaculture. Seed oysters from each stock were deployed in September 2011 at 2 dermo-endemic sites, Dauphin Island and Sandy Bay, Alabama, using an adjustable longline system, and their survival and shell heights were monitored bimonthly. P. marinus infection intensity and condition index were measured at deployment and in March, July and September 2012. The OBOY stock showed lower mortality than the unselected stocks (Cedar Point, Perdido Pass, Calcasieu Lake) at Dauphin Island, and both Louisiana stocks had lower mortality than the Alabama stocks at Sandy Bay, a slightly more saline site. Mortality increased in summer, especially between July and September, concomitant with increasing P. marinus infection intensities at the higher temperatures and favorable salinities. At the higher salinity site, both Louisiana stocks had lower P. marinus infection intensities than the Perdido Pass stock, the stock with the highest percentage of oysters with moderate and heavy infection and cumulative mortality. The OBOY stock reached greater mean shell height than Calcasieu Lake and Perdido Pass stocks. Condition index of the oyster stocks decreased by more than half between March and July following expected spawning. Differences in stock performance highlight the importance of stock selection for aquaculture in dermo-endemic estuaries of the northern Gulf of Mexico.


INTRODUCTION
Off-bottom oyster aquaculture is a nascent and fastexpanding industry in the northern Gulf of Mexico (Maxwell et al. 2008, Walton et al. 2013a).Since 2010, at least 20 commercial oyster farms have begun operation in Louisiana, Alabama and Florida, USA, collectively, with interest in Mississippi where regulations are being developed to allow this type of oyster farming (Petrolia et al. 2017).Off-bottom oyster aquaculture could play an important role in stabilizing production and increasing the sales of premium oysters from Gulf States.
In northern Gulf of Mexico estuaries, most traditional eastern oyster Crassostrea virginica production occurs on-bottom between salinities of 5 and 15 because of excessive mortality due to dermo -a disease caused by the pro tistan parasite Perkinsus marinus and directly transmitted from one oyster to another through the water -and predation from southern oyster drill snails Stramonita haemastoma at salinities >15 (Breithaupt & Dugas 1979, Craig et al. 1989, La Peyre et al. 2016).Off-bottom oyster culture virtually eliminates predation, can improve growth and shell shape and provides opportunities to control fouling and to farm vast areas of previously unsuitable bottom (e.g.mud) and waters at higher salinity where oyster growth is greatest.Moreover, using hatchery-produced seed enables selection of brood stocks for increased resistance to dermo and which are adapted to local environmental conditions (Frank-Lawale et al. 2014, Proestou et al. 2016, Leonhardt et al. 2017).
The need to develop stocks of locally adapted eastern oysters that are resistant to disease has long been recognized (Haskin & Ford 1979, Matthiessen et al. 1990).Previous research studies in the northeastern and mid-Atlantic regions have successfully bred stocks of eastern oysters with increased resistance to the effects of Haplosporidium nelsoni (MSX) (Haskin & Ford 1979, Matthiessen et al. 1990) as well as to the effects of both MSX and dermo (Ragone Calvo et al. 2003a) and Roseovarius oyster disease (Davis & Barber 1999).Following these findings, selective breeding programs were organized and have significantly contributed to the expansion and success of oyster aquaculture in those regions (Guo et al. 2008, Frank-Lawale et al. 2014, Proestou et al. 2016).
In the northern Gulf of Mexico, a Louisiana eastern oyster stock named 'OBOY' has been selectively bred for dermo resistance since 1999 (Stickler et al. 2001), as P. marinus is the only parasite causing significant mortalities in this region.The objective of the current study was to compare the performance of this locally selected stock to wild oysters from Louisiana and Alabama and to determine the true potential of the OBOY stock for use in the rapidly expanding numbers of oyster farms in Alabama estuaries.Specifically, the mortality, growth, P. marinus infection intensity, and condition index of the progeny of wild oysters collected in Louisiana and Alabama, and of the OBOY stock, were compared in a common garden experimental design at 2 dermo-endemic Alabama sites.The wild oysters were collected from waters of intermediate to high salinities where dermo is typically endemic.
The Louisiana stocks were spawned at the Louisiana Sea Grant Oyster Hatchery in Grand Isle in May 2011 to produce a F0 generation of Calcasieu Lake and a F4 generation of the OBOY stock.Each stock (2 separate spawning events) was naturally spawned (150 oysters per stock) and the resulting larvae reared using methods similar to Dupuy et al. (1977).Pediveliger (~280 µm) larvae were set on micro-cultch material (~500 µm ground oyster shell) to produce single oyster spats that were shipped to the Auburn University Shellfish Laboratory (AUSL), Dauphin Island, Alabama, for rearing.
Spat from all 4 oyster stocks were reared in individual flow-through upwellers at a density of ~5000 oysters per upweller at AUSL.Once the majority of oysters were retained on a 6 mm grading screen, all stocks were placed in 4.5 mm mesh oyster bags and reared in off-bottom floating cages (OysterGro®) in Sandy Bay (Fig. 1), Alabama, until experimental deployment.

Experimental design
In September 2011, 4 ALS aquaculture bags of each stock containing 75 oysters per bag were deployed at each study site, for a total of 300 oysters per stock and site.ALS bags were fully enclosed to prevent the risk of predation mortality and suspended beneath the water surface.Since predation was largely removed, mortality could be more readily attributed to stressful abiotic conditions and P. marinus infection.At both sites, oyster growth (shell height) and mortality (counts of live/dead) data were collected every other month, starting in November 2011 and ending in November 2012, for a total of 7 sampling periods.Condition index and P. marinus infection intensity data were collected at the time of deployment to establish pre-deployment baselines for each stock (N = 15 oysters per stock × 4 stocks, total of 60) and in March, July, and September 2012 (N = 15 oysters × 4 stocks × 2 sites, total of 120 per sampling time).

Water quality
Daily mean water salinity and temperature data from September 2011 to November 2012 were obtained from the Dauphin Island Sea Lab station and Point Aux Chenes monitoring stations.Mean water temperature and salinity for each sampling interval were calculated for use as predictor variables in multiple linear regression analyses to examine the relationships between water quality and interval mortality or growth of oysters.

Mortality
At each sampling time the numbers of live and dead oysters in each bag were recorded and the dead oysters were discarded.Interval mortality and cumulative mortality were calculated as described by Ragone Calvo et al. (2003b).

Growth
Shell height, the greatest distance between hinge and growth edge, was measured with digital calipers (Absolute Coolant Proof Calipers, series 500, resolution 0.01 mm, Mituyoto America Corporation) in 100 oysters selected haphazardly per stock, site, and sampling time (25 oysters per bag, 4 bags).Monthly interval growth rate was calculated as the increment in mean shell height between 2 consecutive sampling times divided by the number of days between samplings and normalized to a 30 d period.Mean growth rate was calculated using the mean shell height of each bag (N = 4).

Condition index
Condition index was calculated as the ratio of dry tissue weight to the whole oyster weight minus its shell weight (i.e.filled cavity weight) multiplied by 100 using a variation of Hopkins' (1949) formula as recommended by Lawrence & Scott (1982).For each oyster, a 10 ml aliquot of oyster tissue homogenate that had been prepared to determine P. marinus infection intensity, was dried at 65°C for 48 h, and the dry weight for the whole oyster was calculated based on the total volume of homogenized tissue as described by La Peyre et al. (2003).

Statistical analyses
All statistical analyses were done using SigmaStat version 3.5 (Systat Software).Results for daily salinity and temperature from continuous data recorders were analyzed with a 1-factor (site) ANOVA followed by Tukey's multiple comparison procedure.Interval salinity and temperature were compared with Kruskal-Wallis 1-factor (sampling interval) ANOVA on ranks followed by Dunn's multiple comparison procedure.Cumulative mortalities (%) at the time oysters reached commercial size and at the end of the study were compared using a series of chisquare analyses to determine differences between stocks at each site.Interval mortality (%) at each sampling interval was analyzed with a 2-factor (stock, site) ANOVA followed by Tukey's multiple comparison procedure.Interval mortality (%) of each stock was analyzed with a 2-factor (site, sampling interval) ANOVA followed by Tukey's multiple comparison procedure.Interval mortality data of all stocks and sites were pooled together and analyzed using a multiple linear regression with interval temperature and interval salinity as predictor variables.Shell height was compared with a 2-factor (stock, site) ANOVA at time of deployment (September 2011) and at the end of the study (November 2012) followed by Tukey's multiple comparison procedure.Shell height was also compared by a 1-factor (stock) ANOVA at the time oysters reached commercial size at each site.Interval growth rate (mm mo −1 ) of each stock was analyzed with a 2-factor (site, sampling interval) ANOVA followed by Tukey's multiple comparison procedure.Interval growth rate data of all stocks and sites were pooled together and analyzed using a multiple linear regression with interval temperature, interval salinity and interval initial shell height as predictor variables.P. marinus body burden and CI were compared at the beginning of the study with 1-factor (stock) ANOVA and data collected in March, July and September were compared with a 2-factor (stock, sampling interval) ANOVA followed by Tukey's multiple comparison procedure.The number of oysters with heavy and moderate infections in the 4 stocks at each site was also compared with a Chi-square test.To achieve normality and homogeneity of variance, shell height and P. marinus body burden were log transformed.All data are reported as mean ± SD.

Water quality
Mean daily salinity was significantly higher at Sandy Bay (i.e.Point Aux Chenes, 23.4 ± 4.5, range 5.3−31.4)than at Dauphin Island (21.3 ± 4.9, range 7.6−31.7).Daily salinity at both sites throughout the year tended to be >18 except for the January−March interval.(Fig. 2).Interval salinity at Dauphin Island was significantly lower than at Sandy Bay during the November−January, January−March, and July−September intervals (Fig. 3).
Temperature followed expected seasonal trends and ranges for this region (Fig. 2), and differences between both sites were not found (Dauphin Island: 22.4 ± 5.7°C, range 9.5−31.3°C;Sandy Bay: 23.1 ± 5.7°C, range 10.3−31.8°C).The lowest interval temperatures were during the November−January and January−March intervals while the highest interval temperatures were during the May−July and July− September intervals (Fig. 3).

Mortality
At the end of the study, significant differences in cumulative mortality were found between stocks at Sandy Bay (p < 0.001) and Dauphin Island (p < 0.001).At Sandy Bay, the cumulative mortalities of both Louisiana stocks (Calcasieu Lake: 13%, OBOY: 11%) were significantly lower than the cu mulative mortalities of either Alabama stocks (Perdido Pass: 29%, Cedar Point: 20%), and Cedar Point stock had significantly lower cumulative mortality than Perdido Pass stock (Fig. 4).At Dauphin Island, the cumulative mortality of the OBOY stock (9%) was significantly lower and the cumulative mortality of Perdido Pass stock (26%) was significantly higher than the cumulative mortalities of either Cedar Point (18%) or Calcasieu Lake (17%) stocks.
In March 2012, when oysters reached commercial size at Sandy Bay, no difference in cumulative mortality between stocks could be shown (0.7−1.0% cumulative mortality, Fig. 4).At the time point when oysters reached commercial size at Dauphin Island in May 2012, cumulative mortality was significantly higher (p = 0.001) in Perdido Pass stock (6.2%)than in Cedar Point (1.1%) and OBOY (1.1%) stocks (Fig. 4).
Interval mortality (%) at each sampling interval was only significantly different between stocks during the March− May (p < 0.05) and July−September (p < 0.0001) intervals and differences between sites were not found.Interval mortality of the Perdido Pass stock was higher than the OBOY stock during the March−May interval, and than the OBOY, Calcasieu Lake and Cedar Point stocks in the July−September in terval.The interval mortality of Cedar Point stock was also higher than of OBOY stock in the July− September interval.When analyzed by stock, interval mortality was only significantly different between intervals (p < 0.001) and dif ferences between sites were not found (Fig. 5).Overall mortalities of 4 mo old oysters deployed in September 2011 was very low (< 5%) during fall and winter intervals and increased significantly in summer intervals at both sites (Fig. 5).
Multiple linear regression analysis was used to determine the relationship between interval mortality and the potential predictors of interval temperature and interval salinity.Interval temperature (p < 0.001) but not interval salinity (p = 0.944) contributed significantly and produced a multiple regression model with interval mortality = −7.054+ (0.413 × interval temperature) + (0.009 × interval salinity) and an R 2 of 0.397.

Growth
At the time of deployment, in September 2011, no differences in shell height were found between stocks (in mm; OBOY: 42.7 ± 6.0, Calcasieu Lake: 45.4 ± 6.3, Cedar Point: 42.5 ± 5.2, Perdido Pass: 43.7 ± 6.3) or sites (in mm; Sandy Bay: 43.3 ± 6.3, Dauphin Island: 43.8 ± 5.8).Stocks reached commercial size (75 mm) in March 2012 at Sandy Bay and May 2012 at Dauphin Island, when the oysters were 10 and 12 mo old, respectively (Fig. 6).In March 2012, OBOY stock (85.8± 10.9 mm) was significantly larger than Calcasieu Lake (80.Interval growth rate of each stock was significantly affected by the interaction of site and sampling interval (OBOY, Cedar Point, Perdido Pass: site × sampling interval interaction; p < 0.001, Calcasieu Lake: site × sampling interval interaction, p = 0.017; Fig. 7).Interval growth rates in OBOY stock were the highest (> 6 mm mo −1 ) in the September−November 2011 and September−November 2012 intervals and the lowest in the July−September interval.In the other stocks, there was only 1 high growth rate interval (> 6 mm mo −1 ) at the beginning of the study in the September−November 2011 interval, and the lowest growth rates were also in the July−September interval.Interval oyster growth rate was significantly higher at Sandy Bay than at Dauphin Island during the September−November 2011 interval and tended to be also higher in the January-March interval, while it tended to be lower than at Dauphin Island in the period between March and September 2012 (Fig. 7).
Multiple linear regression analysis was used to determine the relationship between interval growth rate and the potential predictors of interval temperature, interval salinity and initial shell height at the beginning of each interval.Interval temperature (p < 0.056) and initial shell height (p < 0.001), but not interval salinity (p = 0.337), contributed significantly and produced a multiple regression model with interval growth rate = 7.969 + (0.120 × interval temperature) + (0.0895 × interval salinity) − (0.117 × initial shell height) and an R 2 of 0.474.

P. marinus infection intensity
No significant differences in P. marinus infection intensity between stocks were detected at the time of deployment (in log 10 parasites g −1 wet tissue; OBOY: 1.7 ± 1.4, Calcasieu Lake: 1.0 ± 1.2, Cedar Point: 1.5 ± 1.2, Perdido Pass: 1.3 ± 1.3).At Sandy Bay, there were significant effects of sampling times (p < 0.001) and stocks (p = 0.001) on P. marinus infection intensities after deployment.Infection intensities increased significantly from March (1.5 ± 1.5) to July (3.7 ± 1.6) and again from July to September (4.5 ± 1.5).Infection intensities of Calcasieu Lake (2.7 ± 1.9) and OBOY stocks (2.9 ± 1.8) were significantly lower than Perdido Pass stock (3.8 ± 2.0).Calcasieu Lake stock also had significantly lower infection intensity than Cedar Point stock (3.6 ± 2.1).Moreover, the com- Perkinsus marinus infection and with light (<10 4 parasites g −1 wet tissue), moderate (1 × 10 4 to 5 × 10 5 parasites g −1 wet tissue) and heavy (> 5 × 10 5 parasites g −1 wet tissue) infections sampled in March, July, and September 2012 at the Sandy Bay and Dauphin Island study sites bined percentage of oysters with moderate and heavy infections was significantly lower in the Louisiana stocks, OBOY (p = 0.035) and Calcasieu Lake (p = 0.011) stocks, than in the Perdido Pass stock.At Dauphin Island, P. marinus infection intensity was only significantly affected by sampling time (p = 0.047), with higher infection intensities in September (4.5 ± 1.6) compared with March (1.5 ± 1.0).No differences in the combined percentage of oysters with moderate and heavy infections could be shown between stocks at Dauphin Island (Fig. 8).

Condition index
No differences in mean condition indices between stocks were found at the time of deployment (OBOY: 16.9 ± 1.8, Calcasieu Lake: 15.1 ± 2.8, Cedar Point: 17.2 ± 3.1, Perdido Pass: 15.7 ± 1.8).After deployment, there was a significant effect of stock (p = 0.036) and sampling time (p < 0.001) in Dauphin Island.Spe cifically, Perdido Pass (10.5 ± 5.0) stock had a sig nificantly greater condition index than Calcasieu Lake (9.5 ± 5.0) stock, and condition index in March (16.5 ± 1.5) was significantly greater than in July (7.0 ± 1.5) and September (7.0 ± 1.9).Oyster condition index in Sandy Bay was only affected by sampling time, with a significant decrease, by more than half, in condition index of all stocks from March (15.7 ± 2.0) to July (5.8 ± 1.3) and September (7.1 ± 2.3).

DISCUSSION
The mortality, growth, Perkinsus marinus infection intensity and condition index of 1 oyster stock selected for dermo resistance in Louisiana and 3 unselected stocks, 1 from Louisiana and 2 from Alabama estuaries, were compared at 2 Alabama sites.The Louisiana stock selected for dermo resistance showed significantly lower mortality than the unselected stocks over the course of the study at Dauphin Island, and both Louisiana stocks had lower mortality than the Alabama stocks at Sandy Bay, a slightly more saline site.Mortality peaked in summer concomitant with increasing P. marinus infection intensities and higher temperatures.The selected stock reached greater mean shell height than the unselected Louisiana stock and the Alabama oyster stock with the highest mortality and P. marinus infection intensity (i.e.Perdido Pass stock).The shell height of oysters at the higher salinity site was slightly but significantly greater than at the lower salinity site.
Growth rates decreased during intervals with higher temperatures and as shell heights at the beginning of each interval increased.Condition indices of the oyster stocks decreased by more than half between March and July, following expected oyster spawning, and stayed depressed throughout the summer.Differences in stock performance especially in dermorelated mortality highlight the importance of stock selection for intensive aquaculture in dermo-endemic estuaries of the northern Gulf of Mexico.
Less than 10% of the OBOY stock oysters selectively bred for dermo-resistance died by the end of the study compared to > 25% of the worst-performing Alabama stock oysters (i.e.Perdido Pass).The greater oyster mortality in the worst-performing stock was associated with higher P. marinus infection intensities, similarly described in previous studies (Ray 1954, Mackin 1962).Specifically, in Sandy Bay, Perdido Pass stock oysters had significantly higher infection intensity and a greater percentage of moderately and heavily infected oysters with P. marinus than Louisiana stock oysters.In Dauphin Island the performance of Perdido Pass stock oysters followed similar trends but significant differences with the other stocks were not found.Mean daily salinity at Sandy Bay (23.4) was significantly greater than at Dauphin Island (21.3); still, differences were small and average salinities at both sites were well within the range favorable for parasite propagation (Chu et al. 1993, Dungan & Hamilton 1995, La Peyre et al. 2006).The salinities prevalent at Dauphin Island between the end of January and the end of March, however, were much lower than at Sandy Bay and lower than the optimum conditions for P. marinus proliferation (>15).Those salinities impeded the proliferation of P. marinus, as indicated by finding all oysters in March lightly or not-infected at Dauphin Island while some Alabama stock oysters (7−20%) had moderate and heavy infections at Sandy Bay.
The highest susceptibility to dermo of Perdido Pass stock oysters was unexpected as this population inhabits a high-salinity area and was anticipated to have developed some dermo resistance.For all the stocks, but especially for Perdido Pass stock, mortality peaked in summer concomitant with increasing P. marinus infection intensities (7−27% heavily infected oysters) at the higher temperatures.High temperatures are well known to promote the propagation of this parasite in vitro and in vivo in laboratories studies (Chu & La Peyre 1993, Dungan & Hamilton 1995, La Peyre et al. 2008).In field studies, increased infection intensities of P. marinus in eastern oysters have consistently been recorded when temperatures exceed 20°C and at salinity >12−15, leading to peak mortalities following summer high temperatures (Ray 1954, Mackin 1962, Ragone Calvo et al. 2003b).
The unselected Louisiana oyster stock, consisting of the progeny of wild oysters collected from Calcasieu Lake public oyster grounds, showed lower mortalities than the Perdido Pass stock at Dauphin Island and both Alabama stocks at the Sandy Bay site.The progeny of Calcasieu Lake oysters also showed the lowest mortality when compared to the progeny of oysters collected on oyster public grounds from other Louisiana estuaries (Stickler et al. 2001, Leonhardt et al. 2017).This is the reason why oysters from Oyster Bayou on the southern edge of Calcasieu Lake were used as the founding brood stock for the OBOY line.Oysters from this location are exposed to higher salinities than oysters of other Louisiana public grounds, which may have favored selection for increased survival upon P. marinus in fection.Differences in disease susceptibility be tween oyster populations along both the Atlantic and Gulf of Mexico coasts have been reported and should continue to serve as the basis for brood stock development for local use (Bushek & Allen 1996, Stickler et al. 2001, Brown et al. 2005, Frank-Lawale et al. 2014).
At the end of the study, the selected OBOY stock which showed the lowest mortality also reached the highest shell height while the Perdido Bay stock with the highest mortality had the lowest shell height.As with mortality, the mean P. marinus infection intensities and percentage of oysters with moderate and heavy infection intensities which were greatest in Perdido Pass stock oysters than in OBOY stock oysters might explain the lower shell height; increase in P. marinus infection intensities can lead to a decrease in oyster growth rates (Menzel & Hopkins 1955, Paynter & Burreson 1991).Oyster growth rates were the lowest during the July−September interval and might be due to increased P. marinus infection intensities and the elevated temperatures, which often exceeded 30°C.High temperatures may have contributed directly to the decrease in growth rate be cause of reduction or cessation of pumping and shell closure as reported in past studies (Collier 1954, Loosanoff 1958).Interestingly, temperatures tended to be higher and oyster growth rate lower at Sandy Bay compared to Dauphin Island during the July− September interval.Overall, however, temperature and initial shell height were the major predictors of growth rates in this and previous studies (reviewed in Kraeuter et al. 2007).
The differences in shell heights and in salinities between sites were small but significant.Although salinity is a major factor controlling growth rates (Kraeuter et al. 2007), the favorable salinity (generally ≥20) at both sites, along with relatively small differences in interval salinities between sites ex plain why salinity was not identified as a predictor of growth rates under our study's environmental conditions.While salinities at both sites were > 20 during most intervals, it is interesting to note that when salinity was lower (~15) at Dauphin Island compared to Sandy Bay (~20) during the January−March interval, the oyster growth rates at Dauphin Island were also significantly lower than at Sandy Bay.Other factors such as phytoplankton quantity and quality or suspended sediment, which were not measured, may have also had some differential impact on oyster growth at both sites.
The condition index is a prime indicator of how well an oyster uses the shell cavity available for somatic and gonadal tissue growth and reflects physiological or nutritive status (Haven 1961, Rainer & Mann 1992).Oyster condition index is often used to estimate meat quality and yield (Lawrence & Scott 1982).The only significant difference found in condition index was between Perdido Pass and Calcasieu Lake stock oysters, but it is unlikely that the small difference impacted oyster meat quality.Quantitative comparison of oyster meat quality between stocks will be needed to confirm this assertion (Zhang et al. 2016) The most significant differences between oyster stocks in our study were their mortalities, specifically between July and September during peak P. marinus infection intensities.It is therefore recommended that local growers preferably use stocks with greater dermo resistance at sites with comparable or higher salinity regime or harvest oysters prior to this time.In our current study, oysters could have been harvested before summer because they were spawned early (i.e.May) and reached market size within 12 mo, but this strategy may not always work in years where local environmental conditions may not be as favorable for growth as exemplified in an earlier study (Walton et al. 2013b), or where harvest is not allowed due to environmental conditions (e.g.prolonged closures due to rainfall or harmful algal blooms).Better characterization or understanding of the processes involved in disease resistance to dermo may lead in the future to the development of specific markers for use in candidate gene or marker-assisted selection (Cancela et al. 2010, La Peyre et al. 2010).This is important for the development of off-bottom aquaculture at high salinity, an environmental condition that is well known to increase eastern oyster growth rate but also dermo-related mortalities (Ragone Calvo et al. 2003b, Kraeuter et al. 2007, Bushek et al. 2012).

171Fig. 1 .
Fig. 1.Mobile Bay estuary, Alabama, USA, with Dauphin Island and Sandy Bay study sites (stars) where oysters were deployed.Water salinity and temperature recorders were located at the Dauphin Island Sea Lab station located ~150 m from our Dauphin Island study site, and at (black dots) Point Aux Chenes (PC) and Alabama Department of Public Health station 176 (AD)

Fig. 2 .
Fig. 2. Daily water temperature and salinity from September 2011 to November 2012 from continuous recorders at Dauphin Island Sea Lab and Point Aux Chenes stations (see Fig. 1)

Fig. 4 .
Fig. 4. Cumulative mortality of Louisiana eastern oysters from the OBOY (OB) and Calcasieu Lake (CL) stocks, and Alabama oysters from the Cedar Point (CP) and Perdido Pass (PP) stocks at the Sandy Bay and Dauphin Island study sites.Stocks that do not share a letter have cumulative mortalities that are significantly different at the end of the study (p < 0.05)

Fig. 6 .
Fig. 6.Mean (± SD) shell heights of Louisiana eastern oysters from the OBOY (OB) and Calcasieu Lake (CL) stocks, and Alabama oysters from the Cedar Point (CP) and Perdido Pass (PP) stocks at the Sandy Bay and Dauphin Island study sites.Dashed lines at 75 mm represent harvest threshold shell height