Can late summer Landsat data be used for locating Asian migratory locust, Locusta migratoria migratoria , oviposition sites in the Amudarya River delta, Uzbekistan?

Existing survey methods for assessing the Asian migratory locust, Locusta migratoria migratoria L. (Orthoptera: Acrididae), infestation risk in the Amudarya River delta, Uzbekistan, are largely constrained by economic resources and site accessibility. The surveys are restricted to a few easily accessible areas, which leads to a misinterpretation of the threat of locust infestation. This often results in indiscriminate blanket treatments of vast areas of wetlands with broad-spectrum insecticides, which may adversely impact non-target fauna and flora. In order to minimize the bias during surveys, one approach would be to allocate the sampling locations based on the distribution of the primary food and shelter plant of the locusts, the common reed, Phragmites australis (Cav.) Trin. ex Steud (Poaceae). In this study, we evaluated the utility of satellite-based remotely sensed data (Landsat TM) acquired in August 2006 to characterize reed distribution in the delta and identify potential locust oviposition sites. The overall accuracy of the Landsat data to map land cover classes in the delta was 84%. The Landsat TM data identified 90% of the reeds, but it was less useful in identifying areas where other vegetations (shrubs and grasses) were mixed with reeds. During the following summer field survey in June 2007, we identified 37 sites that were infested with early-instar locusts. The low migration capacity of young nymphs in dense reed vegetation allowed us to presume that these sites were used for oviposition in the previous summer. Twenty-eight (74%) of these 37 sites had reeds in the previous year. Results from these studies demonstrate that reed distribution maps derived from satellite data could be used for targeting locust egg-pod survey locations, in order to minimize sampling bias while predicting locust infestation risks for the following season.