Abstract
Background Zoonotic malaria caused by
Plasmodium knowlesi
poses a growing threat to malaria elimination. In Thailand, Ranong Province reports the highest incidence nationally, yet the role of long-tailed macaques (
Macaca fascicularis
) as reservoir hosts across heterogeneous landscapes remains poorly characterized. Methods We describe a prospective, multidisciplinary study on Ko Chang using a spatial grid-cell framework, with dominant land-use type assigned per cell. We will produce a fine-scale land-use classification by combining satellite data with ground-truth observations, distinguishing easily confused land-use types, such as plantations and forest, and addressing the misclassification common in conventional datasets. Audio recording devices and camera traps will be deployed across the landscape and over time to measure macaque and, human presence across land-use classes. Household questionnaires will characterise human exposure. Existing
P. knowlesi
case records (2013–2024) and entomological data (2011–2022) will be integrated within a unified spatial framework to estimate land-use-specific exposure. Satellite imagery will be classified into land-use types using supervised learning model. Macaque detections will be summarised as a relative abundance index and modelled in an occupancy framework to estimate the probability of site use. Human distribution and movement will be derived from interviews, camera-trap and acoustic detections. Macaque and human abundance, with secondary entomological data, will be assigned to land-use classes as inputs to a population-attributable risk (PAR) model. These land-use-explicit abundance estimates will then be combined with zoonotic malaria prevalence from the literature as inputs to a transmission-dynamics model. Expected outcome is a fine-scale land-use map of Ko Chang; land-use-explicit estimates of macaque and human site use and relative abundance; land-use-specific estimates of population-attributable risk and transmission parameters. Conclusion This study will generate fine-scale, ground-validated landscape and reservoir-host–vector data, improving understanding of how macaque movement, human exposure, and land-use change drive spillover risk, and informing targeted interventions.