Kyalo, Richard: Challenges and opportunities of using ecological and remote sensing variables for crop pest and disease mapping. - Bonn, 2019. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc:
author = {{Richard Kyalo}},
title = {Challenges and opportunities of using ecological and remote sensing variables for crop pest and disease mapping},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2019,
month = jul,

note = {Crop pest and diseases are responsible for major economic losses in the agricultural systems in Africa resulting in food insecurity. Potential yield losses for major crops across Africa are mainly caused by pests and diseases. Total losses have been estimated at 70% with approximately 30% caused by inefficient crop protection practices. With newly emerging crop pests and disease, monitoring plant health and detecting pathogens early is essential to reduce disease spread and to facilitate effective management practices. While many pest and diseases can be acquired from another host or via the environment, the majority are transmitted by biological vectors. Thus, vector ecology can serve an indirect explanation of disease cycles, outbreaks, and prevalence. Hence, better understanding of the vector niche and the dependence of pest and disease processes on their specific spatial and ecological contexts is therefore required for better management and control.
While research in disease ecology has revealed important life history of hosts with the surrounding environment, other aspects need to be explored to better understand vector transmission and control strategies. For instance, choosing appropriate farming practices have proved to be an alternative to the use of synthetic pesticides. For instance, intercropping can serve as a buffer against the spread of plant pests and pathogens by attracting pests away from their host plant and also increasing the distance between plants of the same species, making it more exigent for the pest to target the main crop. Many studies have explored the potential applications of geospatial technology in disease ecology. However, pest and disease mapping in crops is rather crudely done thus far, using Spatial Distribution Models (SDM) on a regional scale.
Previous research has explored climatic data to model habitat suitability and the distribution of different crop pests and diseases. However, there are limitation to using climate data since it ignores the dispersal and competition from other factors which determines the distribution of vectors transmitting the disease, thus resulting in model over prediction. For instance, vegetation patterns and heterogeneity at the landscape level has been identified to play a key role in influencing the vector-host-pathogen transmission, including vector distribution, abundance and diversity at large. Such variables can be extracted from remote sensing dataset with high accuracy over a large extent. The use of remotely sensed variables in modeling crop pest and disease has proved to increase the accuracy and precision of the models by reducing over fitting as compared to when only climatic data which are interpolated over large areas thus disregarding landscape heterogeneity.When used, remotely sensed predictors may capture subtle variances in the vegetation characteristic or in the phenology linked with the niche of the vector transmitting the disease which cannot be explained by climatic variables. Subsequently, the full potential of remote sensing applications to detect changes in habitat condition of species remains uncharted. This study aims at exploring the potential behind developing a framework which integrates both ecological and remotely sensed dataset with a robust mapping/modelling approach with aim of developing an integrated pest management approach for pest and disease affecting both annual and perrennial crops and whom currently there is no cure or existing germplasm to control further spread across sub Saharan Africa.},

url = {}

The following license files are associated with this item: