Image-based morphological characterization of the erythrocytic shape of green turtle (Chelonia mydas) for its potential use as an environmental biomarker.

Environmental pollutants increase genotoxic damage in green sea turtles (Chelonia mydas). Erythrocytic nuclear abnormalities (ENA) have been identified using the micronucleus (MN) test, a biomarker of DNA damage. The MN test is time-consuming, and counting depends on the observer, who requires training to avoid false positives. This study aimed to evaluate the utility of image-based morphometric measurements of erythrocyte nuclear shape to assess differences among green turtles from the Mexican Caribbean inhabiting four feeding sites with varying levels of anthropogenic pressure. An semi-automated measurement of 350 nuclei per animal (n = 60) per study site (n = 15) was performed, optimizing the traditional, time-consuming process of 1000-2000 cell counts. Eight metrics of the nuclei morphology were obtained and correlated with the frequency of ENA. A non-linear regression analysis was performed at each study site to infer the focus distance (FD) term from the aspect ratio (AR) and to evaluate deviations from the ellipse in cases with nuclear abnormalities. The erythrocyte nuclei's minor axis, AR, FD, circularity, and roundness were correlated with the number of ENA. The green turtles captured in the most highly anthropized sites (i.e., Akumal and Xcalak) showed the highest frequency of ENA, the smallest area, and the nuclear form most deviated from the elliptical nuclear erythrocytic shape expected. The proposed methodology enabled differentiation of sea turtles' habitat conditions. Results from this study agree with previous reports of the highest concentrations of organic persistent pollutants, genotoxic damage, and alterations in oxidative stress in green turtles inhabiting urbanized areas of the Mexican Caribbean.