Using tree-ring crossdating techniques to validate annual growth increments in long-lived fishes

Abstract

We applied crossdating, a dendrochronology (tree-ring analysis) age validation technique, to growth increment widths of 50 Sebastes diploproa otoliths ranging from 30 to 84 years in age. Synchronous growth patterns were matched by the following: (i) checking the dates of conspicuously narrow growth increments for agreement among samples and (ii) statistically verifying that growth patterns correlated among samples. To statistically verify pattern matching, we fit each time series of otolith measurements with a spline, and all measurements were divided by the values predicted by the curve. This standardized each time series to a mean of 1, removing the effects of age on growth and homogenizing variance. Each time series was then correlated with the average growth patterns of all other series, yielding an average correlation coefficient (r) of 0.53. Average growth of all 50 samples was significantly correlated with an upwelling index (r = 0.40, p = 0.002), the Pacific Decadal Oscillation (r = –0.29, p = 0.007), and the Northern Oscillation Index (r = 0.51, p = 0.0001), corroborating accuracy. We believe this approach to age validation will be applicable to a wide range of long-lived marine and freshwater species.