Understanding the ecological processes that drive patterns in fishery selectivity and survey catchability

Being able to confidently count organisms is fundamental to estimating the dynamics of populations and to making inferences about what influences those dynamics. However, an accurate census is difficult to achieve because an entire population is almost never fully detectable. The population size of aquatic organisms can be particularly problematic to estimate because the environment is not conducive to human observation, and, particularly in the ocean, the entire habitat of the population in question can never be fully sampled. When we attempt to count the numbers of fish or shellfish in a population, for example to help manage a fishery or to increase our understanding of how a population is responding to shifts in climate, this impediment almost always results in at least some portion of the population being unobserved and unquantified. Having some information about the size and composition of this unobserved demographic is fundamental to population ecology and is particularly essential when the relative contribution of the unobserved demographic to the true population changes in space and time. Advancements in fisheries population dynamics and stock assessment science have resulted in the generally accepted application of two parameters that aid in estimating the unobserved portion of the population from the observed one. The first, selectivity, defines the proportion of a given demographic group, available to, and retained by, survey or fishing gear once it comes into contact with it. The second, catchability, is defined as the proportion of the population caught by a single unit of fishing or survey effort. Despite the influence these two parameters have on our perception of what controls fish and shellfish population dynamics, we still don’t fully understand the underlying processes that influence them for many managed fisheries. Using fishery and population survey data from two mid-Atlantic fisheries, summer flounder and Eastern oysters, I use the following four chapters to 1.) identify patterns in selectivity and catchability and understand the underlying ecological processes that drive them, and 2.) propose how we might better utilize this information to assess and manage these and other fisheries. In Chapter 1 I use summer flounder data collected from commercial and recreational landings and a stock assessment trawl survey to evaluate the selectivity of the survey and fishing gear for different demographic groups. Some interesting patterns were identified, particularly that selectivity for female summer flounder is higher in the recreational fishery than in both the commercial fishery or the stock assessment trawl survey. This pattern suggested a highly female-biased recreational catch and that male and female summer flounder separate in space and time. In Chapter 2 I explored, given the size- and sex-specific selectivity patterns and catch composition identified in Chapter 1, whether management actions could be taken to achieve a more sex-balanced harvest in the recreational fishery. I evaluated whether a series of slot limits, size regulations that require landed fish be between some minimum and maximum size, have the potential to simultaneously reduce mortality on large, fecund, females while maintaining or reducing total fishing mortality. The patterns in summer flounder fishery catch composition and selectivity and the prescribed management actions identified and discussed in Chapters 1 and 2 should contribute significantly to our understanding of the life history, particularly relative to sex-specific habitat use, of summer flounder, and will likely be relevant to the stock assessment and management of this commercially and recreationally important fishery going forward. In Chapter 3 I used a set of field experiments to derive empirical estimates of catchability for a survey dredge used in the stock assessment of eastern oysters in Delaware Bay. I identified an along-bay gradient in catchability that appeared to be driven by changes in oyster density. This density-dependent catchability lead to catch-per-unit-effort of the survey dredge being hyperstable at low oyster density, making catch-per-unit-effort (CPUE) an unreliable proxy for abundance at low oyster density. In Chapter 4 I asked the question, given that evidence from Chapter 3 suggests a fixed catchability coefficient is not appropriate for estimating true density from survey CPUE, how do three alternative models perform in estimating the true density in the sampled area. In the first model, I corrected CPUE by applying spatially-explicit catchability coefficients, as opposed to a constant, that account for the along-bay gradient in density. In the second, CPUE was corrected for by estimating catchability in situ for each tow using a logistic model fit to catch composition and tow covariates. For the third model, CPUE data were ignored entirely and a model that accounted for the proportion of the sample composition that was made up of oysters was applied to estimate oyster density in situ for each tow. The simplest model, which ignored both catchability of the survey gear and CPUE, and relied only on an estimate of the portion of the catch that was made up of oysters, performed best in estimating the true density in the sampled area. The density-dependence in catchability identified for an oyster survey dredge in Chapter 3 is an important finding because it adds to a growing body of literature that density-dependent catchability, a phenomenon traditionally attributed to an interaction between fish and fishermen behavior, may be a common problem in standardized stock assessment survey data as well. In addition, Chapter 4 strongly suggests that when catchability varies at fine spatial and temporal scales, raw catch components may more accurately reflect the true density in the sampled area than an index derived from catch-per-unit-effort and catchability. Both findings should have application to how reef growing populations of organisms, and oysters in particular, are assessed and managed going forward.Ph.D.Includes bibliographical referencesby Jason M. Morso