Modelling habitat suitability for occurrence of human tick-borne encephalitis (TBE) cases in Finland

The numbers of reported human tick-borne encephalitis (TBE) cases in Europe have increased in several endemic regions (including Finland) in recent decades, indicative of an increasing threat to public health. As such, it is important to identify the regions at risk and the most influential factors...

Full description

Bibliographic Details
Published in:Ticks and Tick-borne Diseases
Main Authors: Uusitalo, Ruut, Siljander, Mika, Dub, Timothee, Sane, Jussi, Sormunen, Jani J., Pellikka, Petri, Vapalahti, Olli
Other Authors: Department of Geosciences and Geography, Department of Virology, Veterinary Biosciences, University of Helsinki, Earth Change Observation Laboratory (ECHOLAB), Helsinki Institute of Sustainability Science (profit unit at BY-TDK), Institute for Atmospheric and Earth System Research (INAR), Helsinki One Health (HOH), HUSLAB, Veterinary Microbiology and Epidemiology, Olli Pekka Vapalahti / Principal Investigator, Viral Zoonosis Research Unit, University Management, Helsinki University Hospital Area
Format: Article in Journal/Newspaper
Language:English
Published: 2020
Subjects:
GIS
Online Access:http://hdl.handle.net/10138/318325
Description
Summary:The numbers of reported human tick-borne encephalitis (TBE) cases in Europe have increased in several endemic regions (including Finland) in recent decades, indicative of an increasing threat to public health. As such, it is important to identify the regions at risk and the most influential factors associated with TBE distributions, particularly in understudied regions. This study aimed to identify the risk areas of TBE transmission in two different datasets based on human TBE disease cases from 2007 to 2011 (n = 86) and 2012-2017 (n = 244). We also examined which factors best explain the presence of human TBE cases. We used ensemble modelling to determine the relationship of TBE occurrence with environmental, ecological, and anthropogenic factors in Finland. Geospatial data including these variables were acquired from several open data sources and satellite and aerial imagery and, were processed in GIS software. Biomod2, an ensemble platform designed for species dis-tribution modelling, was used to generate ensemble models in R. The proportion of built-up areas, field, forest, and snow-covered land in November, people working in the primary sector, human population density, mean precipitation in April and July, and densities of European hares, white-tailed deer, and raccoon dogs best es-timated distribution of human TBE disease cases in the two datasets. Random forest and generalized boosted regression models performed with a very good to excellent predictive power (ROC = 0.89-0.96) in both time periods. Based on the predictive maps, high-risk areas for TBE transmission were located in the coastal regions in Southern and Western Finland (including the angstrom land Islands), several municipalities in Central and Eastern Finland, and coastal municipalities in Southern Lapland. To explore potential changes in TBE distributions in future climate, we used bioclimatic factors with current and future climate forecast data to reveal possible future hotspot areas. Based on the future forecasts, a slightly wider geographical extent of TBE risk was introduced in the angstrom land Islands and Southern, Western and Northern Finland, even though the risk itself was not increased. Our results are the first steps towards TBE-risk area mapping in current and future climate in Finland. Peer reviewed