Introduction
Fish shape is highly relevant in aquaculture since it acts as a driver of consumer decision- making. Shape of the flatfish Senegalese sole is characterized by an elliptic and lanceolate midbody, long caudal and dorsal fin sizes and short jaws with a scarce number of reference landmarks. Hence, outline-based methods such as the E lliptic Fourier descriptors analysis (EFA) seem to be highly useful as predictor of these body curved shapes. Hence, EFA could be useful to assess shape variation and evaluate genetic components in sole . In this species, it is estimated that approximately 12-15% of farmed soles show external morphological alterations that impedes their commercialization (de Azevedo et al. , 2019). Guerrero et al., 2021 reported moderate-high heritabilities for body ellipticity as inferred from standard length and body height, h owever, these traits do not capture complete whole-body variation. Thus, this study aims to evaluate whole-body variation using EFA. Phenotypic information was analyzed and compared with a new set of distance-based and ellipse fitting estimator traits . G enetic estimates for these traits were determined.
Materials and methods
A total of 2,271 soles from in eight evaluation batches (EB) were individually photographed at harvest , and the weight and length were in situ recorded. A piece of caudal fin (CF) was preserved for genotyping. Body contours were individually processed in Fiji 2.1.0/1.53c and images in binary format were imported into the package SHAPE to calculate the Elliptic Fourier descriptors (EFDs). A principal component analysis (PCA) of the EFDs was carried out to extract information for symmetrical (PCs) and asymmetrical (PCa ) features of body shape. Coefficients of EFDs were calculated at ± two standard deviations and s cores of PCs and PCa were used in subsequent genetic analysis.
Additionally, some well-established and new ellipse descriptors based on morphometric distance and ellipse fitting parameters were calculated: standard length (SL), maximum body height (MBH) to caudal peduncle height (CPH) ratio , ellipticity (ELL; (SL-MBH)/(SL + MBH)) , whole-body perimeter (PER), the aspect ratio (AR; SL/MBH) as distance-based trait and the aspect ratio from the fitted ellipse (ARe), and solidity (SOL) . ANOVA analysis were carried out to test statistical differences associated with gender, EB and the presence of amoebic nodules (AN). Genetic parameters were estimated by REML as implemented in the package WOMBAT with gender, EB and AN as fixed factor.
Results
PCA identified f our significant PCs for EFDs that explained 94.0% of shape symmetric variation and 68.6% of the total variation , and four PCa that explained 90.7% of asymmetric variation and only 24.5% of total variation. PC1s and PC1a represented 52.8 and 58.7% of symmetric and asymmetric variation, respectively.
Statistical analysis of PCs and PCa indicated that males were statistically more elliptic (lower PC1s and higher PC3s) with wider CFs ( as revealed by lower PC2s and higher PC3s) than females. Also, males tended to have obtuse angles in the CF (PC3a) with the head orientated toward the abdominal cavity (PC4a). Statistical differences in AN were highly influence by fish weight. Significant differences among EBs were found for all components . Statistical analysis of ellipse descriptors indicated that f emales were 12.7% heavier and 2.2% longer than males. Moreover, males were 2.4% more elliptic than females (AR, ELL and ARe ) with a 1.6% lower MBH/CPH, but no differences in PER.
Heritability estimates were very high for PC1s (0.702), moderate for PC3s and PC4a (0.390-0.400) and low for PC2s, PC4s and PC1a-PC3a (0.039-0.168). Heritabilities f or MBH/CPH , ELL, ARe and AR were high or very high (0.513- 0.804) and moderate for SOL (0.273) and PER (0.432). Genetic correlations were very high between ELL, ARe and AR (³ 0.90) and between PER, weight and SL (³ 0.93). ELL, ARe and AR were strongly correlated with PC1s and PC4a (³ -0.873 and ³ 0.705, respectively). ELL and AR showed a high genetic correlation with PC3s (³ 0.574). SOL was highly and negatively correlated with PC2s (0.916) and PC3a had the highest genetic correlation with ARe (-0.583).
Discussion
Shape quality is very important during fish commercialization since consumers preferably buy products whose external appearance resemble wild fish
(Calanche et al., 2020)
. Senegalese sole’s lanceolate body shape is considered a high-quality shape standard and major attention had been paid to mid-body ellipticity as the main trait to select for shape quality (Guerrero-Cozar et al ., 2021). In our shape model, 20 harmonics were sufficient to capture main symmetric and asymmetric variation through the tested population. This study demonstrated that EFA and ellipse fitting estimators are effective and powerful traits to assess shape quality in Senegalese sole.
C aptured variation of EFD was primarily related to the adjustment of whole and midbody to an ellipse and CF variation, and secondaril y to head orientation. Very high heritabilities for body ellipse fitting traits indicated a high potential to improve it via selective breeding and suitable to be included with growth traits (negatively correlated) to avoid an excess of roundness bodies after some selection rounds. Moderate-high heritabilities were also found for traits related to head and CF peduncle (SOL or MBH/CPH). The latter could be relevant to select specifically CF functionalities and fish swimming capabilities. Heritabilities for CF-related traits were low, however, d ue to the influence of CF on consumer’s perception and its use as a heath indicator, indirect selection using other correlated traits would be more efficient in genetic gain.
All the information here provided is highly useful not only for designing breeding programs but also for the improvement of operational methods to preserve shape quality and monitor health and welfare in the aquaculture industry.
Acknowledgments
This work is funded by MCIU/AEI/FEDER, UE, code RTA2017-00054-C03-01, project EU H2020 research grant agreement 817992 ERANET-BLUEBIO COFUND “Bestbrood” code PCI2020-111994 funded by CDTI, MCIN/AEI/10.13039/501100011033 and EU “NextGenerationEU”/PRTR. PGL is granted with a predoctoral scholarship funded by AEI.
References
de Azevedo et al. 2019. Veterinary Pathology, 56.
Guerrero-Cózar et al. 2021. Animals, 11.
Calanche et al. 2020. Reviews in Aquaculture, 12, 805-821.