Precision of Striped Bass Fry Density Estimates

Many fisheries for striped bass Morone saxatilis in the United States are supported by stocking hatchery‐reared fish. Fingerlings are usually obtained by spawning adults in captivity, collecting fry in large (25–100‐L) collection containers (vats), moving them to rearing facilities (e.g., earthen ponds and raceways), and subsequently harvesting the fish that remain after some predetermined period. The number of fry stocked is estimated, and the accuracy and precision are seldom known or reported. The effect of treating fry stocking estimates as “error free, complete counts” on survival and production estimates is, therefore, also unknown. The objective of this study was to document and determine the statistical precision of the procedures used at Texas Parks and Wildlife Department's Dundee Fish Hatchery through the 1990s (standard procedures still in use at many hatcheries) for estimating fry densities in vats. The precision of fry density estimates was directly correlated with the mean density and sample volume. Such estimates could be improved by reducing the water volume (especially when densities are low) in vats and using a standard sample volume (at least 15 mL) instead of adjusting the sample volume based on visual inspection of fry in vats. The number of samples could be reduced from 10 to 8 without substantially affecting precision. Further, the practice of removing the highest and lowest sample values should be discontinued. The resulting precision of subsequent fingerling survival and production statistics could benefit from these changes without a substantial increase in data analysis. Received August 5, 2010; accepted October 4, 2010