Multi-epoch species distribution modeling of ticks in Arkansas: Expanded ecological niche and the vector-borne disease landscape
This research quantifies and maps the spatiotemporal variability of ticks across Arkansas, particularly in the context of environmental change. Boolean logic-based ecological niche suitability maps and species distribution models were generated at 250 m spatial resolution across a 16-year range (2009–2024), divided into one four-year training epoch and two six-year modeled epochs (2009–2014 and 2019–2024). The niche maps were used to generate a pseudo-absence grid based on biophysical characteristics that was then paired with tick presence data sampled from across the state between 2015 and 2018 to train a Random Forest model. The Random Forest species distribution model displayed an average probability difference of + 13.1% across Arkansas for the more recent epoch (2019–2024), as compared to the earlier epoch (2009–2014). K-Fold cross validation was performed to evaluate the model’s performance (PR-AUC = 88.6%; ROC-AUC = 87.8%) and feature importance testing showed that evapotranspiration was a key predictor of tick presence.
Authors
- Ashley P. G. Dowling (ORCID: https://orcid.org/0000-0002-1890-1689)
- Joseph P. Messina (ORCID: https://orcid.org/0000-0002-1978-0778)
- B. G. Peter (ORCID: https://orcid.org/0000-0002-5724-4482)
- Dirk Haselow
- Dalton W. Williams
Institutions
- University of Alabama (US)
- University of Arkansas at Fayetteville (US)
- University of Arkansas for Medical Sciences (US)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-18
- DOI
- https://doi.org/10.21203/rs.3.rs-11747762/v1
- Primary Topic
- Vector-borne infectious diseases
- Type
- preprint