Effects of underestimating catch and effort on surplus production models

Fisheries can be managed based on surplus production models when only catch and effort data are available. However, reported catch and effort may not equal the true values. We studied the effects of jointly underestimated catch and effort on surplus production model parameter estimates (e.g., MSY, Bmsy and Fmsy) as well as estimates of key ratios (e.g., F/Fmsy). We used ASPIC to examine various scenarios of underreporting for three example fisheries, North Atlantic swordfish, northern pike in Minnesota and queen conch in the Turks and Caicos Islands. With constant underestimation of catch and effort throughout time, MSY, Bmsy and Bnext are all underestimated by the same percentage, while Flast and the ratios, F/Fmsy and B/Bmsy, are not affected. As a result, harvest regulations can be set based on fishing mortality and the ratios. That is, when one thinks the harvest is MSY with F = Fmsy, one is achieving MSY and Fmsy even though the catch is actually larger than it is thought to be. However, increasing or decreasing trends in underreporting of catch and effort over time lead to errors in the parameter and ratio estimates whose direction is case-specific and whose magnitude can be high or low. Each fishery model responded differently to the simulated scenarios, which may be a result of different exploitation histories or the quality of the fit of the production model to the data. The wide range of outcomes observed may be due to the fact that underestimation of catch and effort can lead to a gain or reduction in data contrast. Simulations of a variety of possible scenarios similar to the methods in this study should be conducted if catch and effort are believed to be underestimated to determine how the surplus production model responds.