Agricultural land use patterns and malaria vector abundance; A case study of the spatial relationships between agricultural land use and mosquitoes of the Anopheles gambiae complex and the Anopheles funestus complex on Rusinga Island, Kenya

Organisator Laboratory of Geo-information Science and Remote Sensing

do 5 juni 2014 12:30 tot 13:00

Locatie Gaia, building number 101
Droevendaalsesteeg 3
6708 PB Wageningen
+31 317 48 16 00
Zaal/kamer 1

By Annemieke Mulder

The SolarMal project aims to reduce the malaria vector population of Rusinga Island, Kenya. Within this project there is information about the spatial and temporal component of those vectors, but analyses with those data where never performed. While those components are important determinants of vector abundance and spread. Therefore, the objective of this study was to study the spatial relationship between specific land use patterns and malaria vector abundance for the two crop growing seasons (March and October) separately in the SolarMal study area, Rusinga Island. The first step was to perform land use classifications for both seasons on basis of information about the different crop types occurring on the island. Furthermore data gathered in the field for two study areas in October 2013 and February/March 2014 about the different land use classes was used for this. The classifications were validated by means of an error matrix via which the total accuracy of the different classifications was calculated. The relational analysis by means of ordinary least squares (OLS) was performed in ArcGIS 10.1. The final products are the two agricultural land use classification maps, validation results and the coefficients of determination for the OLS analysis of the relationship between vector abundance and land use patterns.

The validation results for the land use classifications were < 30% in terms of total accuracy. Therefore, the separate crop type classes were merged, the classifications were repeated and the validation was performed again. From the classifications, it became clear that Rusinga Island is heterogeneous and the agricultural fields are mainly located along the lake shore. The hill tops are covered with shrubs and trees and slopes are mainly bare. The accuracy of the new classifications increased to 57% for the whole of Rusinga Island and to 55% and 44% for study areas 1 and 2 respectively for March and 35% for Rusinga Island as a whole and 36% and 43% for study areas 1 and 2 respectively for October. After the OLS analyses were performed, it became clear that there is no relationship between any land use pattern and vector abundance. Other studies indicated that there is a relationship between land use and larvae of the malaria vector, which can be studied in the future by use of the same models. Furthermore, a combination of land use patterns (percentages) near houses could possibly lead to finding a relationship between those land use type combinations and vector abundance.