International Journal of Aquaculture, 2026, Vol.16, No.5, 354-361 http://www.aquapublisher.com/index.php/ija 358 performance. Pre-standardised data sets were used to estimate allometric coefficients and species-specific reference mean standard lengths for the species, prior to cross-validation, and these were not re-estimated during each iteration of the LOO, making the LOO results an internal assessment of the consistency of the classifications in the present data set, and not an independent estimate of the predictive performance for new specimens. The 100% classification rate reported does not therefore provide an independent estimate of the predictive performance of the set for unknown specimens: the size adjustment parameters would have to be estimated independently of the 100% classification rate. 2.8 Meristic analysis A total of nine meristic characters were counted - dorsal spines, dorsal soft rays, anal spines, anal soft rays, scales in the lateral line, pectoral rays, pelvic rays, caudal rays, and gill rakers. Dorsal and anal soft rays were counted separately from the corresponding spines. A standard procedure was followed when counting gill rakers on the first gill arch of all specimens. Pelvic rays were recorded as a separate meristic character. Meristic characters were compared among Lutjanus agennes, Pseudotolithus senegalensis and Polydactylus quadrifilis using the Kruskal-Wallis H test. Significant overall differences were found and pairwise comparisons were then made for the three possible species-pair comparisons, with Bonferroni-adjusted significance levels. Characters with no within-species variation were described but not tested inferentially. Homogeneous groups were identified for characters having a high pairwise difference, and the values sharing a letter were assumed not to be significantly different while values with different letters were significantly different at p<0.05. 3Results 3.1 Species-specific allometric relationships Allometric regressions showed strong relationships between standard length and the measured morphometric characters in all three species. All regression models were statistically significant (p <0.001). Allometric coefficients for Lutjanus agennes ranged from 0.840 to 1.002, with R² values ranging from 0.853 to 0.998. For Pseudotolithus senegalensis, the coefficients ranged from 0.624 for eye diameter to 1.056 for pre-anal length, with R² values ranging from 0.744 to 0.984. For Polydactylus quadrifilis, the coefficients ranged from 0.987 to 1.012, with R² values ranging from 0.968 to 0.999. The species-specific allometric coefficients, 95% confidence intervals, and coefficients of determination for the measured morphometric characters are presented in Table 1.
RkJQdWJsaXNoZXIy MjQ4ODYzNA==