|
Australian Plague Locust, Chortoicetes terminifera (Walker), can rapidly increase in population size in the remote interior of eastern Australia under favorable habitat conditions and cause severe agricultural damage. To minimize losses, earlydetection of locust outbreaks is essential to the implementation of preventive control. Quantitative measurement of locust habitat suitability is critical for improving the efficiency of ground and aerial surveys, and providing vital information for locust population forecasting. Here, routine locust survey by the Australian Plague Locust Commission during 2003 and 2011 is investigated in relation to the habitat greenness derived from the fortnightly 250 m composites of Normalized Difference Vegetation Index (NDVI), and the rainfall amount from the weekly 5 km grids of modelled precipitation, using the spatial analysis and statistics of ESRI ArcGIS. The sighting dates of high-density locust nymphs (band and sub-band) were assigned into 5 groups corresponding to the nymphal development stages, and the fortnightly NDVI values and weekly rainfall totals for the locust locations were extracted for the previous 13 weeks. The averaged NDVI values for locust habitats showed a slight increase of 0.04-0.13 from initially 0.23-0.29 within 4-7 weeks before 2nd-5th instar bands and sub-bands were sighted. The median values of NDVI increase were on an equivalence scale of 0.05-0.15 from the background of 0.21-0.26; the increments were equal to 12-37% in the historical range from 13-22% and equal to 38-59% from the 11-18% of seasonal maxima, which indicated by normalized NDVI anomalies that the majority of high-density nymphs had all experienced a period of better than average conditions in both historical and seasonal perspectives. However, 5th-instar bands and sub-bands were consistently found in slightly dried habitats, while 1st-instar bands were mostly seen in much green areas but on the trend of dry-off. The time-series of habitat greenness for 1st-instar bands illustrated a very different pattern from the others, which could have resulted from the limited dataset mainly from the winter rain zone. Significant single rainfall of 25-30 mm was required to trigger the locust breeding sequence, and in excess of 40-50 mm total rainfall for locusts to survive the entire nymphal period. These findings will improve the understanding of locust plague mechanisms related to habitat condition, potentially provide practical means to monitor locust habitat conditions remotely and improve the underlying basis for locust survey and population management in Australia.
|
