Skip to main content
Dryad

Novel disease state model finds most juvenile green turtles develop and recover from fibropapillomatosis

Cite this dataset

Mansfield, Katherine; Kelley, Jacob; Savage, Anna; Kelley, Kayla (2021). Novel disease state model finds most juvenile green turtles develop and recover from fibropapillomatosis [Dataset]. Dryad. https://doi.org/10.5061/dryad.ttdz08kzn

Abstract

Fibropapillomatosis (FP) is a sea turtle disease characterized by benign tumor development on skin, eyes, and/or internal organs. It primarily affects juvenile green turtles (Chelonia mydas) in coastal foraging sites. The Indian River Lagoon (IRL), Florida, USA, is a coastal green turtle foraging site where the observed FP annual rate averaged 49% between 1983 and 2018. FP is not a major cause of sea turtle mortality and most individuals fully recover; however, the overall dynamics of this disease are poorly understood because prior disease history is unknown for individuals without FP at capture time, and future disease outcome is unknown for individuals with FP at capture time. To better evaluate FP dynamics for green turtles in the IRL, we developed a hierarchical model for predicting disease state change. We used data from 4,149 captures of 3,700 individual green turtles captured in the IRL. The hierarchical disease state model contained two levels: level one modeled whether an individual would develop FP, and level two modeled disease state progression, including states for pre-FP affliction, active FP affliction, and full recovery from FP. From the hierarchical model, we estimated 99.8% (95% credibility intervals 99.1-100%) of juvenile green turtles in the IRL developed FP, indicating that nearly every individual in the IRL is affected by this disease. The model also suggested that turtles quickly developed FP upon recruitment to the IRL and then recovered at different rates, with most completely recovering before emigrating from the IRL as they mature. This is the first analysis of long-term sea turtle data suggesting nearly every turtle in an aggregation both develops and recovers from FP.

Methods

Please refer to publication

Usage notes

The R code is uploaded to a text file. The Excel file with FP data includes the following column headers:

EID = assigned identifyer

FP = a binary list of whether a turtle had the tumor disease FP or not (0=no, 1=yes)

FP_YESNO = whether a turtle had the tumor disease FP or not (Yes or No)

Year = year that the sample was collected

SCL = Straight Carapace Length of the turtle sampled (in centimeters)

SCLbinned = the size bin that the turtle was included in based on 5 cm incrimental bins of SCL

Season = the northern hemisphere season in which the turtle was sampled (Spring, Summer, Fall, or Winter)

Columns with seasons broken out = binary list of 0 (not present) or 1 (present) for each season (for model purposes)

y0, y1, y2 = presence/absence recaptures 0 - 2 years after the first encounter (0 = not present; 1 = present).

Funding

Florida Sea Turtle License Plate Grants Program