Fig 1.
Stocking locations (green triangles, ▲) and current distribution of non-native blue catfish in Chesapeake Bay.
Blue dots (●) correspond to additional locations where blue catfish were collected from the mainstem of the Chesapeake Bay during 2018 and 2019 (total fish collected = 63). The Chesapeake and Delaware Canal (C&D Canal) connecting the Chesapeake Bay with the Delaware Bay is also shown. Note that blue catfish have not been recorded from the Delaware Bay yet. Inset shows the location of the Chesapeake Bay in relation to Virginia (VA) and Maryland (MD). Figure available in color online.
Fig 2.
Vertically-averaged salinity (in psu) in the Chesapeake Bay typically encountered in spring (April) and fall (October) during average (2012), dry (2009) and wet (2011) years.
Maps are based on the model developed by Du and Shen (2015).
Fig 3.
Survival of blue catfish over time after gradual transfer from freshwater to one of three salinity treatments during the pilot experiment.
Figure available in color online.
Fig 4.
Bottom salinity and fork length of invasive blue catfish captured from the Rappahannock, York and James river subestuaries of the Chesapeake Bay during 1975–2017 by the VIMS juvenile fish trawl survey.
The quantile regression splines for salinity ≤ 15 psu are shown for three quantiles (ε = {0.01, 0.1, 0.5}). Note that the y axis of the histogram is truncated for the York and James rivers; the percent of blue catfish captured at salinities < 1 psu in the York (47.8%) and James (46.1%) rivers are indicated in the figure.
Table 1.
Parameter estimates for the stratified Cox proportional hazards model fit to the time-to death data from the toxicity test on blue catfish.
CL = 95% confidence limit.
Fig 5.
Median time-to-death (h) for male and female blue catfish at various salinities predicted by the stratified Cox proportional hazards model.
Numbers at the end of the lines represent salinities in practical salinity units (psu). Figure available in color online.
Table 2.
Bayesian Information Criterion (BIC), ΔBIC, number of parameters and BIC weight for Firth logistic regression models fitted to describe the 72-hour mortality of blue catfish exposed to various salinities.
The two most parsimonious models, highlighted in bold, were averaged to determine the final model. FL = Fork length.
Table 3.
Parameter estimates for the most parsimonious Firth logistic regression model to describe the 72-hour mortality of blue catfish exposed to various salinities.
CL = 95% confidence limit.
Fig 6.
Predicted survival of blue catfish based on Firth Logistic regression fit to data from the 72-hour salinity tolerance experiment.
The point and the bar correspond to the predicted salinity at 50% mortality (LC50) and the corresponding 95% confidence interval based on the logistic regression. Numbers along the line represent the minimum, median and maximum length (mm) of blue catfish used in this study.
Fig 7.
Spatially explicit probabilities of survival (72-hour) for a 224 mm blue catfish (median length in the salinity tolerance experiment) throughout the Chesapeake Bay based on vertically-averaged salinities in spring (April) and fall (October) during average (2012), dry (2009) and wet (2011) years.
Number at the top left corner of each panel denotes the percent area of the Chesapeake Bay where predicted probability of survival for blue catfish was greater than 0.8. Note that the probability of survival was nearly 1 in the Chesapeake and Delaware Canal (C&D Canal).
Fig 8.
Spatially explicit probabilities of survival for a 224 mm blue catfish throughout the Chesapeake Bay based on surface salinities.
See Fig 7 for additional details.