Assessment of reader accuracy and recommendations to reduce subjectivity in age estimation

Historically, authors have stressed the importance of validating methods to estimate fish age. However, interpretation of annuli involves an element of subjectivity and individual readers using valid techniques can incorrectly estimate fish age. To assess the error resulting from this subjectivity, biologists and technicians (N = 36) who routinely estimate fish age for the Texas Parks and Wildlife Department's Inland Fisheries Division were asked to estimate the ages of 50 known-age largemouth bass (Micropterus salmoides) using a validated otolith technique. Accuracy of individual readers was highly variable and differed among young (ages 0–3; 72–100%) and old (ages 4–14; 36–100%) fish. Most readers who aged less than 90% of the largemouth bass correctly also produced biased age estimates. Ages for young fish were as likely to be underestimated as overestimated, while ages for old fish were underestimated. Known-age reference structures allowed readers to assess accuracy and identify biases. Thus, systematic errors observed in this study should be correctable. These results indicate that valid age estimation techniques are not always correctly applied. Fisheries agencies need to verify the accuracy of age estimates produced by their readers. Because of the limited availability of known-age fish, agencies should collaborate to acquire reference structures. Technologies such as digital imaging and the Internet could allow such collections to be shared.