AGE ESTIMATION IN CALCIFIED CALCAREOUS STRUCTURES; PRELIMINARY FINDINGS OF AN INTER-LABORATORY COMPARISON

,SUMMARY Age estimates for albacore tuna are derived using a number of methods including the enumeration of growth increments in spines, scales and otoliths. Inconsistencies in interpretation between structures and between laboratories may introduce error into the age-data available for stock assessment. A comparative age estimation exercise was carried out to assess bias and precision of age estimates between readers and between structures. Six scientists from four laboratories, with varying levels of experience in albacore age estimation participated in the study. Images of the otoliths, scales and spines from 84 albacore tuna collected from the North east Atlantic and the Mediterranean were distributed to each reader. When the age estimates of an experienced reader were compared with those of an inexperienced reader who they had trained, moderate levels of precision were observed (mean % coefficient of variation 11.2, 7.7 and 6.5 for otoliths, scales and spines respectively). However, the wider comparison across laboratories yielded poor precision (% coefficient of variation 35.7, 25.7 and 24.0 for otoliths, scales and spines respectively). Linear bias was detected between readers for all structures, with readers systematically over- or underestimating age relative to each other. Analysis of variance indicates that inter-reader precision is higher for Mediterranean albacore than for Atlantic albacore. Agreement of age estimates derived using different structures from the same fish was low and scales appeared to provide lower estimates of age than spines. The results highlight the need for further intercalibration of albacore tuna ageing methodologies. RESUME