| Summary: | The 1993 Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy agreed to carry out a mackerel otolith exchange programme to salve the problems found in age readings. A total of 398 pairs of otoliths was exchanged, 200 of which came from the western area (ICES Divisions VIa and VIIbc) and 198 were from the southern area (ICES Divisions VIIIc and IXa) , and ten readers from different countries took part in the exchange. The camparison of otolith readings of the two areas was carried out separately. The results show an unacceptable variability in mackerel age determination between the readers from different countries. The average general agreement between readers was low in the two areas, 47% for the western area and 51% for the southern area. Agreement between readers varies between 19% and 71% for the western area and between 4% and 81% for the southern area. Standard deviations by age groups increase considerably for age groups >5 years for the western area sample and for ages >3 years for the southern area sample. The age bias plots of each reader against the modal age show great variability in the samples from the two areas, above all in older ages. The age bias plots for all readers combined show that bias increases with age, and that older fish tend to be underestimated. The results of Wilcoxon's test to analyze bias between readers shows that bias exists between most readers in both samples. The notched box and whisker plots show differences in the interquartile range, average, median and symmetry in the different results of the readers. Considering the importance of age readings in assessment, new investigations are necessary to increase agreement, and to standardize ageings.
|
|---|