Influences of ontogeny and latitude on the osteohistology of Apalone spinifera.

Abstract

All paleohistological observations should be contextualized with modern taxa, and although the importance of variables like latitude, body size, and age are broadly acknowledged for all animals, quantification of the importance of these variables has not been undertaken for many groups, including soft-shelled turtles. In an effort to both better understand the model taxon itself and further contextualize fossil taxa, we herein systematically sample the osteohistology of a modern soft shelled turtle from various latitudes and across ontogeny. Apalone spinifera is a large soft shelled turtle with a wide native range in the United States, making it a perfect organism to study how ontogeny and latitude impact trionychid growth dynamics. Specimens with Carapacial Lengths ranging from 6 to 24 cm were procured from across 12° of latitude within the Mississippi river basin. Using standard petrographic thin sectioning techniques, slides were prepared from the mid-diaphysis of the femur, the hyo-hyploplastral bridge, and halfway point of the 4th costal. A. spinifera shows histology that is remarkably intact and unremodeled, but otherwise resembles that of other trionychids, suggesting that they can effectively serve as a modern analog for ancient soft-shelled turtles. Costal width was a robust indicator of Carapacial Length, demonstrating that fragmentary specimens can still provide reliable Carapacial Length estimations, provided that the full width of the costal is preserved. Carapacial Length and growth mark counts were correlative (R² = 0.736), but they are not interchangeable, suggesting that this measure of body size is a decent approximation of age. There was notable variance between retrocalculated GM Estimations, femur GM Counts, costal GM Counts, hypoplastral GM Counts, and hyoplastral GM Counts, demonstrating the importance of a standardized plane of section. The highest number of GMs were typically present in the plastron, but no element reached retrocalculated growth mark (GM) estimates, and variance was larger with advanced ontogeny. Turtles at higher latitudes do grow modestly slower than turtles at low latitudes (y = -21.056x + 41.396) but this relationship is weak (R² = 0.2843), suggesting that it may be negligible when comparing specimens from across multiple localities.