Efficiency of rodent control in upland rice and potential of forecasting chronic rodent infestation in Northern Laos

摘要

This study followed up on results of an integrated rodent management program that was implemented from 2010 to 2011 in 18 treatment and 18 control villages in three provinces of Northern Laos. The program's impact on reducing rodent damage to upland rice was reported previously. Here, we focused on the efficiency with which upland villagers applied the proposed rodent control methods (snap traps, biological control using the protozoan parasite Sarcocystis singaporensis, community hunts) by comparing control effort (inputs) to an output of more than 73,000 rodents (mainly black rats) culled in Houphanh (HP) province. We also hypothesized that rice yields of a given crop year (2009) could predict rat damage in the three provinces in following years (2010, 2011). A survey of flowering of bamboo and fagacean 'nut' trees in 2011 was used to check for a potential influence of flowering events on rodent infestation. Using regression analysis we observed that efficiency (cumulative culls method(-1)) of snap traps and biological control decreased significantly with increasing field size of upland rice fields, while the opposite trend was apparent with regard to paddy fields. Numbers of rodents culled by hunting and trapping increased with rising numbers of hunting villagers, but culls household(-1) declined with increasing village size. We developed multiple regression models that predicted rodent culls by the paddy/upland rice area ratio and rodent control effort and explained >90% of the variation. Rat damage to wet season rice in 2010 and 2011 (and rodent culls ha(-1)) increased with increasing village rice yields of 2009, while the treatments' regression line was positioned at a 5.5% lower rat damage level, in parallel to the controls. Integration of the observed relationships using the General Linear Model (GLM) allowed predicting rat damage under different treatment and seasonal scenarios. Surprisingly, 99.8% of the variation in rodent culls ha(-1) in HP could be explained by a combination of rodent control effort, rice productivity, and flowering events. Herewith, we provide predictive and explanatory models that could be useful for future rodent management in the uplands. The observed relationships are discussed in view of potential forecasting of chronic rodent infestation and, ultimately, outbreaks. (C) 2017 Elsevier Ltd. All rights reserved.