Influence of protected areas on malaria prevalence in Sub-Saharan Africa. (October 2015)

• Record Type: Journal Article
• Authors: Taber, Eric D.; Smithwick, Erica A.H.
• Is Part Of: Applied geography. Volume 64(2015:Oct.)
• Page Start: 35
• Page End: 45
• Publication Date: 2015-10
• Subjects: Malaria -- Plasmodium falciparum -- Random forests -- Epidemiology -- Geospatial health -- Infectious disease; Geography -- Periodicals; Human geography -- Periodicals; Human ecology -- Periodicals; 910
• DOI: 10.1016/j.apgeog.2015.09.001
• Languages: English
• ISSNs: 0143-6228

Abstract

Abstract: Introduction: Despite exponential growth in the number and extent of protected areas globally, their role within disease dynamics remains unclear. Protected areas shape many biophysical and social factors related to malaria prevalence such as land use-land cover, biodiversity, socioeconomic conditions, and human behavior. This work examines the extent to which protected areas influence Plasmodium falciparum malaria prevalence within surrounding human populations throughout Sub-Saharan Africa.. Methods: Using malaria prevalence data from 2008 to 2012, we tested for differences in mean malaria prevalence at survey locations according to the IUCN classification of the nearest protected area. We also used dual logistic regression and Random Forest approaches to model malaria burdens at survey locations using a variety of known biophysical determinants of malaria in addition to protected area related covariates. Results: We found that malaria prevalence differed by IUCN class, with survey locations near IUCN classes Ia and III exhibiting significantly higher prevalence values than all other classes. Additionally, distance to the nearest protected area emerged as an important predictor of malaria prevalence in the logistic regression model, with lower malaria prevalence at locations closer to protected areas (OR: 1.14; 95% CI: 1.10–1.17). Distance was of moderate importance in the Random Forest models; however, the relationship between distance and prevalence was nonlinear.. Conclusions: We show that, at a continental scale, malaria prevalence was lower for populations closer to protected areas in Africa, compared to farther away. However, we also found evidence of spatially complex relationships, both around individual protected areas, and across protected areas at this geographic extent, reinforcing the need for additional, small-scale case studies. Ultimately, by showing a link between protected areas and disease presence, this work helps improve understanding of the complex, multiscalar drivers of malaria.. Highlights: We examined the relationship between malaria prevalence and protected areas. Malaria prevalence differs by the IUCN class of the nearest protected area. Malaria prevalence was lower at locations closer to protected areas. Protected areas shape malaria prevalence at multiple scales.