Evaluating the spatial distribution of Dociostaurus maroccanus egg pods using different sampling designs.

In its gregarious phase the locust Dociostaurus maroccanus (Thunberg) (Orthoptera Acrididae) has periodically caused significant yield losses in many Mediterranean and Asian countries, and alarm in the general public. Population outbreaks in recent years have frequently required the application of control measures, based on those that have low environmental impact, which are only possible with a sound knowledge of locust bio-ethology and ecology. Our research was aimed at studying the spatial distribution of D. maroccanus egg pods in two Apulian egg bed areas in southern Italy, thus contributing to the rationalization of control methods. The distribution of D. maroccanus egg pods was investigated using a geostatistical approach. Three sampling designs (called A, B and C), characterized by different mesh and clod sizes, were compared to evaluate their effectiveness and affordability. In both egg bed areas, the variogram models were asymptotic with a small nugget effect, and indicated an aggregated distribution of egg pods. Contour maps showed that design A, based on a larger mesh and clod size, was characterized by few hot spots and an extended zone of “low density” egg pods, while design B, involving a smaller mesh and clod size, showed a more structured distribution, with various hot spots alternating with zero level zones. Finally, design C, based on a larger mesh size and smaller clods, showed a single extended hot spot surrounded by a large area without egg pods. Moreover, because of the larger amount of soil to be examined, design A was about 2.6- and 10.9-fold more time consuming than designs B and C, respectively. Our data showed that sampling designs providing smaller and denser samples should be preferred over designs with fewer and larger samples when information on both the distribution and density of egg pods is needed.