History of outbreaks of the Australian plague locust, Chortoicetes terminifera (WALK.) between 1933 and 1959, and analysis of the influence of rainfall on these outbreaks

Abstract

An account of the history of outbreaks of the Australian plague locust, Chortoicetes terminifera, is presented for the period 1933–1959 from published information and records maintained by the New South Wales Department of Agriculture. Population fluctuations during this period have been examined in relation to rainfall for seven of the major and intermediate outbreak areas in New South Wales. It has been shown that rainfall during the months August to December is important in its influence on locust numbers the following season, whereas that occurring between January and April is apparently not. Subnormal "early season" rainfall favours an increase in the locust population of an area in the next season, while above-normal rainfall favours a locust decrease in the following season. In each of the outbreak areas selected there was a significant negative correlation between the current first locust generation of the season (measured in terms of an arbitrary scale or index) and preceding early season rainfall; and the estimated regression lines due to the two variables in each area were apparently parallel, although not coincident. A multiple regression analysis was carried out incorporating the preceding autumn locust generation as an extra independent variable. When the Culgoa-Barwon outbreak area was eliminated from the analysis, it was found that the regression equations for each of the six other areas could be considered identical, even to the constant term, any differences between areas being attributable to differences in the values of the independent variables. A single multiple regression equation based on the data for the six areas combined was calculated, and it is suggested that this can be used to forecast approximately the size of the spring locust generation in time to plan appropriate control measures. Difficulties inherent in the data are pointed out and the disturbing influence of invasion of one area by migrating swarms from another discussed. It is believed that more refined methods of estimating locust numbers and greater knowledge of locust behaviour in relation to long-term rainfall influences would allow predictions of much greater accuracy to be made, based on the correlations established in the present paper. The relative significance of outbreak areas situated mainly in the Central Division of New South Wales and those in far western parts of New South Wales and Queensland is discussed. It is suggested that two of the most serious outbreaks on record were the result of long-range migratory flights from the latter areas. The importance of these areas as a source of outbreaks affecting South Australia, Victoria, and southern New South Wales is also emphasized, and a tentative explanation of the role of early season rainfall in these arid regions is given.