Landscape connectivity for yellow fever: a proxy approach to detect displacement and circulation through environmental corridors and interfaces

Abstract

Background: Yellow fever is an arboviral haemorrhagic disease transmitted by mosquitoes in the tropical regions of Africa and South America. WHO estimates that there are 200 000 severe cases and 30 000 deaths worldwide annually. In Brazil, the wild cycle of yellow fever (also termed urban cycle) occurs through transmission between non-human primates with infected mosquitoes of the Haemagogus and Sabethes genus as intermediaries. Serological analysis of deceased primates is the main monitoring method; however, in many cases, these primates die in remote zones, making the geographical monitoring of yellow fever occurrences difficult. Even so, zoonotic surveillance authorities have observed geographical patterns in yellow fever circulation, especially associated with the ecological connectivity between forest fragments with certain sizes and characteristics. However, several potential displacement corridors connecting geographical points with positive cases remain unknown. Unexpectedly, other routes apparently favourable to primate circulation appear to act as a barrier to yellow fever against displacement. The determination of landscape and ecological characteristics acting on this primate (and viral) circulation is essential for prediction and mapping of displacement corridors and susceptible zones, favouring decision making in planning preventive and proactive actions to combat yellow fever. Methods: With use of field data for yellow fever-positive primate distribution, we applied landscape connectivity mapping, combining circuit theory and random walk approaches to determine the most environmentally suitable routes or corridors for the displacement of infected primates among positive cases. Based on the extensive experience of zoonotic surveillance authorities in the field, we established a set of relevant variables (eg, forest cover, path sizes, topography, and climate) to generate the resistance surface (ie, a variable that gives ascending values denoting the level of difficulty of an organism to move across the landscape) for São Paulo state, Brazil. Findings: The preliminary results show a series of potential displacement routes and geographical barriers, providing unprecedented information on environmental affinities and limits for wild yellow fever displacement. These detected routes and points of high susceptibility for yellow fever can be used for field monitoring geographical prioritisation. Interpretation: Based on our final results, authorities will be able to define priority sites for primate serological sampling, and determine unsampled and susceptible remote locations where they need to survey for positive or negative cases, filling geographical gaps, and improving the understanding of the wild cycle of yellow fever in Brazil. Funding: Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP; 2019/12988-7 and 2018/14389-0).