Análisis espacio-temporal de plagas urbanas

Abstract

Introduction: Urban pests pose significant hazards for public health and the environment, causing considerable economic losses and extra workload for municipal services. Effective control of urban pests is crucial to preserve urban ecosystem health. There are two main approaches to pest control, one is reactive and the other one is proactive. Reactive measures take place to control an established pest problem, whereas proactive measures focus on preventing pest problems from occurring, as well as on early warning. For the purposes of prevention, it is essential to understand the underlying ecological dynamics of pest-environment systems in urban areas, for which analytical models are scarce. With a focus on urban rats in the city of Madrid, Spain, the present thesis aims at bridging this gap by modelling urban rat infestations over space and time, providing tools and relevant information to enhance future urban pest control initiatives. Methodology: A list of putative environmental factors that could favour rat infestations in urban areas was provided through semi-structured surveys and brainstorming sessions with a sample of municipal pest control professionals of Madrid. This information formed the basis for subsequent quantitative analyses conducted to determine and characterize the risk of urban rat infestations over space and time. A total of 10,956 rat sightings reported in the city of Madrid between 2002 and 2008 was analyzed using: 1) GAM models to identify the city hot spots of rat activity and the associated environmental risk factors; 2) GLM models to identify temporal patterns of rat sightings and the associated weather correlates. As spatial and temporal dynamics do work together, combined space-time analyses were also conducted using three different approaches: 1) time-split spatial models to examine the temporal trends (intensity and localization) of rat hot spots; 2) the inhomogeneous K-function to explore space-time clustering of rat infestations; and 3) mechanistic modelling of rat sightings in time and space. Results: Several environmental risk factors of urban rat infestation were identified, these were the age of building, the distance to the nearest market, to the nearest green area, to the nearest water source and to the nearest cat feeding station, human density and the size of the nearest green area. Significant weather correlates of urban rat infestation were temperature and relative humidity (positively correlated) and precipitation (negatively correlated). According to the inhomogeneous K-function analysis, rat sightings tend to occur in clusters with 270 meter radius and to last for 10 days. The mechanistic model simulated rats sightings in space and time. Conclusions: Environmental risk factors, weather correlates and hot spots of rat infestations in space and time were identified, providing relevant information for targeting environmental health massages and control efforts more effectively. The combined application of GIS and statistical modeling proved useful in supporting and generating hypotheses about spatiotemporal dynamics of urban pests.