CHARACTERIZING THE PERFORMANCE OF MULTIPLE TESTING MODALITIES TO DETECT EMYDOMYCES TESTAVORANS IN BLANDING'S TURTLES (EMYDOIDEA BLANDINGII)

Emydomyces testavorans is a recently described fungus associated with significant shell disease in chelonians, including Illinois state-endangered Blanding's turtles (Emydoidea blandingii). Characterizing the epidemiology of this pathogen requires accurate detection; however, E. testavorans is challenging to detect and optimal antemortem sampling strategies have not been identified. In total, 90 juvenile Blanding's turtles from two populations naturally infected with E. testavorans were used to determine the detection performance of different testing approaches and establish optimal sampling strategies in the context of commonly encountered clinical and management scenarios. Turtles received comprehensive physical examinations and computed tomography scans to document grossly apparent and/or radiographic shell lesions. Multiple sample types were tested for E. testavorans via quantitative PCR (qPCR), including shell swabs, combined cloacal-oral swabs, combined cloacal-oral-shell swabs (COSSs) swabs, and tank water samples. Latent class modeling was used to determine the sensitivity and specificity of each detection method in the absence of a gold standard test. Using multiple tests often increases information available to decision-makers; therefore, sensitivity, specificity, and positive and negative predictive values were calculated for two E. testavorans detection methods interpreted either in series or in parallel. Latent class modeling demonstrated that COSS and water qPCRs were highly sensitive (98.0 and 94.0%, respectively) and adequately specific (74.9 and 88.3%, respectively) for E. testavorans; thus, these testing methods are recommended for routine surveillance where use of multiple methods is not possible. In wild and head-started Blanding's turtle populations where false negatives are costly, qPCR testing of two COSS samples interpreted in parallel maximized sensitivity (sensitivity=100%, specificity=56.1%) and is recommended for E. testavorans surveillance when multiple tests are available. This study provides objective measures to guide effective E. testavorans testing and inform future epidemiologic studies, support chelonian medicine, and empower conservation managers, ultimately safeguarding turtle health in managed care and wild settings.