Reevaluation of Eastern Oyster Dredge Efficiency in Survey Mode: Application in Stock Assessment

Abstract

Dredge efficiency measurements were conducted in 2003 by comparing paired dredge and diver samples taken on eastern oyster Crassostrea virginica beds in the New Jersey waters of Delaware Bay. We evaluated whether ancillary data collected during the tow could be used to estimate dredge efficiency rather than periodically conducting costly field experiments. With the exception of market-size eastern oysters, dredge efficiency did not vary between the time when it was last measured (2000) and 2003. Dredge efficiencies were lower upbay in both studies. Capture efficiency was significantly greater for live eastern oysters than for boxes (dead, articulated valves) and significantly greater for boxes than for cultch (shell clumps, shells, and shell fragments without attached live eastern oysters or boxes) in both studies. In the single significant exception, dredge efficiency was significantly lower for market-size eastern oysters in 2003. Survey quantification of eastern oyster abundance by calibrating dredge performance requires recognition that dredge efficiencies are spatially variable and that this spatial variability is nonrandom. An analysis that focused on tow-based variables and variables describing fishing effort during the preceding 3 years revealed a suite of variables providing substantive information about the between-tow variation in dredge efficiency. These variables described aspects of catch volume, industry effort, and cultch availability. Each increased with increasing dredge efficiency. Selected subsets of these variables used in multiple-regression equations with data from an independent data set from the 2003 stock survey reproduced the estuarine gradient in dredge efficiency as well as the gradient in dredge efficiency from the highest values (for live eastern oysters) to the lowest values (for cultch). In only one case (market-size animals) did the efficiencies predicted from survey data using the multiple-regression equations diverge significantly from direct measurements. These analyses provide some confidence that estimates of dredge efficiency can be attained from tow-specific data.