Comparisons of stable isotope composition among tissues of green turtles

Rationale Ecologists often need to make choices about what body parts (tissues or organs) of an animal to sample. The decision is typically guided by the need to treat animals as humanely as possible, as well as the information that different body parts can provide. When using stable isotopes, decisions are also influenced by whether specimens would require preservation, and whether they have properties (such as high lipid concentrations) that would influence measurements. Sometimes we cannot use a preferred tissue (for example, because of ethical or logistical constraints), and in such cases an ability to reliably predict stable isotope composition for one tissue from data yielded by another would be useful. Methods In this study we analysed multiple tissues (skin, whole blood, red blood cells, plasma and nail) from green turtles (Chelonia mydas) to evaluate variation in C:N ratios, and test hypotheses about the intercept and slope of regressions of stable carbon and nitrogen isotope compositions among tissues. Results Regression models revealed that linear relationships were present for most comparisons, except those involving the δ13C of skin, and the slopes (β1) of most regressions were different from unity. The C:N ratios of skin were significantly higher and more variable than those of other tissues. The δ13C and δ15N of nail were highly correlated with those of the whole blood, red blood cells and plasma. Nail and red blood cells showed low variation in C:N. Conclusions The patterns in slopes of regressions indicate that comparisons of measurements yielded by different tissues of wild animals are complicated by the fact that the tissues are unlikely to be in isotopic equilibrium with their diet. Of the tissues used in this study, nail is simple to collect, requires minimal disturbance to the animal and no special preservation; these traits should make it attractive to turtle ecologists, but more information is needed on aspects such as growth rates.