Spatial population dynamics of eastern oyster in the Chesapeake Bay, Maryland

https://doi.org/10.1016/j.fishres.2020.105854Get rights and content

Abstract

Incorporating spatial information can improve estimates from stock assessment models when there are differences in population processes (e.g., natural mortality) among areas. Population dynamics of the eastern oyster Crassostrea virginica vary spatially within the Chesapeake Bay, Maryland, and our objective was to better characterize oyster population dynamics by estimating changes in natural mortality, fishing mortality, abundance, and recruitment over time and space. We developed statistical stage-structured models for 36 regions and fit the models to fishery dependent and independent data sources. Regional patterns in population dynamics emerged that would have been lost in a spatially aggregated approach. Regions that were closer together tended to have similar patterns in natural mortality, exploitation rates, abundance, and recruitment over time. We were able to estimate time-varying natural mortality because ancillary data on the number of dead individuals were incorporated into the population dynamics model. This approach to estimating time-varying natural mortality may be more widely applicable to species where dead individuals are observed in routine surveys.

Introduction

Incorporating spatial information into stock assessments has become increasingly common (Punt, 2019), but spatial variation in natural mortality has received little attention. Spatial models can provide less biased, more precise estimates and better information for management than non-spatial models when there are differences in population processes (e.g., natural and fishing mortality, growth, recruitment) among areas (e.g., Fu and Fanning, 2004; Goethel et al., 2011; Ying et al., 2011; Guan et al., 2013; Punt et al., 2015; Vincent et al., 2017). Spatially varying natural mortality rates may lead to similar biases in stock assessment models as spatial variation in fishing mortality rates (Cao et al., 2020).

In Chesapeake Bay, the eastern oyster Crassostrea virginica historically supported one of the largest oyster fisheries in the world; however, oyster harvest and abundance have declined substantially during the past 130 years (Rothschild et al., 1994; Wilberg et al., 2011). A variety of factors, such as overfishing, disease, and habitat loss, have been linked to the decline of oysters in Chesapeake Bay (Rothschild et al., 1994; Wilberg et al., 2011; Damiano and Wilberg, 2019). In addition to natural mortality, recruitment, growth, and fishing mortality also vary spatially in Chesapeake Bay, Maryland (Jordan et al., 2002; Damiano and Wilberg, 2019). Differences in recruitment among locations could be due to available habitat for oyster settlement, spatial patterns of adult abundance, environmental factors (e.g., hydrodynamics), and larval oyster swimming behavior (Kimmel and Newell, 2007; North et al., 2008). Once larvae settle, growth rates may differ among locations due to environmental and biological factors such as salinity and food availability (Jordan et al., 2002). Exploitation rate varies spatially due to regulations and unequal distribution of effort by commercial fishers (Damiano and Wilberg, 2019). Lastly, persistent differences in exploitation, recruitment, and natural mortality rates among areas can cause spatial patterns in population dynamics because oysters are sessile as juveniles and adults.

Natural mortality is one of the primary factors that affects the abundance of adult eastern oysters in Chesapeake Bay, Maryland and previous research has documented interannual variation and spatial differences in natural mortality rates within Maryland (Vølstad et al., 2008). The dominant source of natural mortality is two diseases: MSX, which is caused by the single-celled parasite Haplosporidium nelsoni and dermo, which is caused by the protozoan parasite Perkinsus marinus (Ewart and Ford, 1993). Areas of higher salinity typically have higher mortality due to disease than areas of lower salinity (Vølstad et al., 2008). Disease-related mortality also varies over time due to changes in disease prevalence and intensity, which is related to spatial patterns in salinity (Tarnowski, 2007). Although occurring less frequently and typically at a smaller spatial scale, high freshwater discharge from rivers can also cause spatial differences in mortality of oysters due to exposure to fresh or low salinity water (Tarnowski, 2019). Because natural mortality is known to be an important driver of oyster dynamics in Maryland (Vølstad et al., 2008) our goal was to develop a modeling framework that allowed for the estimation of time- and spatially-varying natural mortality.

The spatial variability in population dynamics of oysters in the Chesapeake Bay, Maryland, suggests that treating the population as a single well-mixed unit may not be appropriate. Our objectives were to characterize spatial patterns in natural mortality, fishing mortality, abundance, and recruitment of oysters over time in Chesapeake Bay, Maryland. Given the history of disease mortality events for oysters in Maryland (Vølstad et al., 2008), we were especially interested in accounting for natural mortality that varied over time and space. We implemented stage-structured models that estimated time-varying natural mortality for 36 regions, which was the finest spatial scale that could be supported by the available data.

Section snippets

Model overview

We implemented statistical, stage-structured models for oysters within 36 regions in the Maryland portion of Chesapeake Bay (Fig. 1). The remaining five regions (grey regions in Fig. 1) were not modeled due to a lack of survey data. The models estimated abundance within five stages during 1999–2017. The model year started on October 1, the beginning of the oyster fishing season in Maryland. The model year also coincides with the start of the fall dredge survey, which is one of the

Model fits

Fits of the individual region-specific models to all data sources were acceptable overall with the fishery-dependent data generally fit less well than the fishery-independent data (Figs. S1–S36). The models fit the fall survey indices for all regions across all oyster stages with few outliers and little patterning among the residuals. Additionally, the models fit the index of recent natural mortality relatively well in all regions. Fits to the fishing mortality rate time series were generally

Discussion

The dynamics of oysters in Maryland showed substantial differences regionally such that a spatially aggregated modeling approach would have missed important regional differences in natural mortality, exploitation rates, abundance, and recruitment. For example, northern regions generally had a declining trend in abundance of market oysters during the early 2000s. This was due to a combination of one of the largest recruitment events of the previous 40 years that directly preceded the time period

CRediT authorship contribution statement

Marvin M. Mace: Conceptualization, Methodology, Software, Formal analysis, Data curation, Writing - original draft, Writing - review & editing, Visualization. Kathryn L. Doering: Conceptualization, Methodology, Software, Formal analysis, Writing - review & editing, Visualization. Michael J. Wilberg: Conceptualization, Methodology, Software, Formal analysis, Resources, Writing - review & editing, Funding acquisition, Supervision, Project administration. Amy Larimer: Conceptualization,

Declaration of Competing Interest

The authors declare no conflict of interest.

Acknowledgements

We thank the Maryland Department of Natural Resources (MD DNR) and the National Science Foundation under grant number OCE-1427019 for providing funding for this project. We acknowledge the MD DNR Oyster Stock Assessment Team for helpful comments and suggestions. We also thank the numerous individuals at MD DNR who contributed to the collection and maintenance of data we used in our modeling efforts. Two anonymous reviewers provided thoughtful comments that helped improve this manuscript.

References (60)

  • J. Cao et al.

    Spatiotemporal index standardization improves the stock assessment of northern shrimp in the Gulf of Maine

    Can. J. Fish. Aquat. Sci.

    (2017)
  • J. Cao et al.

    A novel spatiotemporal stock assessment framework to better address fine-scale species distributions: development and simulation testing

    Fish Fish.

    (2020)
  • A.-L. Chai et al.

    Evaluation of oyster sampling efficiency of patent tongs and an oyster dredge

    N. Am. J. Fish. Manag.

    (1992)
  • J.F. Christmas et al.

    Oyster shell disarticulation in three Chesapeake Bay tributaries

    J. Shellfish Res.

    (1997)
  • K.L. Doering et al.

    Patterns in oyster natural mortality in Chesapeake Bay, Maryland using a Bayesian model

    Fish. Res.

    (2021)
  • J.W. Ewart et al.

    History and Impact of MSX and Dermo Diseases on Oyster Stocks in the Northeast Region

    (1993)
  • J.C. Field et al.

    Does MPA mean “major problem for assessments”? Considering the consequences of place-based management systems

    Fish Fish.

    (2006)
  • S.E. Ford et al.

    Estimating mortality in natural assemblages of oysters

    Estuaries Coasts

    (2006)
  • R.I.C.C. Francis

    Data weighting in statistical fisheries stock assessment models

    Can. J. Fish. Aquat. Sci.

    (2011)
  • C. Fu et al.

    Spatial considerations in the management of Atlantic cod off Nova Scotia, Canada

    N. Am. J. Fish. Manag.

    (2004)
  • D.R. Goethel et al.

    Incorporating spatial structure in stock assessment: movement modeling in marine fish population dynamics

    Rev. Fish. Sci.

    (2011)
  • W. Guan et al.

    Impacts of population and fishery spatial structures on fishery stock assessment

    Can. J. Fish. Aquat. Sci.

    (2013)
  • S.J. Jordan et al.

    Oyster biomass, abundance, and harvest in northern Chesapeake Bay: trends and forecasts

    J. Shellfish Res.

    (2002)
  • D.G. Kimmel et al.

    The influence of climate variation on eastern oyster (Crassostrea virginica) juvenile abundance in Chesapeake Bay

    Limnol. Oceanogr.

    (2007)
  • D.G. Kimmel et al.

    Long-term (1939 to 2008) spatial patterns in juvenile eastern oyster (Crassostrea virginica, Gmelin 1791) abundance in the Maryland portion of Chesapeake Bay

    J. Shellfish Res.

    (2012)
  • D.G. Kimmel et al.

    The relationship between interannual climate variability and juvenile eastern oyster abundance at a regional scale in Chesapeake Bay

    N. Am. J. Fish. Manag.

    (2014)
  • K. Kristensen et al.

    TMB: automatic differentiation and Laplace approximation

    J. Stat. Softw.

    (2016)
  • M.K. La Peyre et al.

    Analysis of environmental factors influencing salinity patterns, oyster growth, and mortality in lower Breton Sound estuary, Louisiana, using 20 years of data

    J. Coast. Res.

    (2016)
  • P. Legendre et al.

    Numerical Ecology

    (1998)
  • P.H. Leslie et al.

    An attempt to determine the absolute number of rats on a given area

    J. Anim. Ecol.

    (1939)
  • Cited by (0)

    View full text