The Desert Locust has been known as an important crop pest for at least 3,000 years, ever since swarms of this species constituted one of the 'plagues' of Egypt in biblical times. It is reasonable to suppose that plagues of have occurred at intervals throughout this period. Documentary evidence assembled at the Anti-Locust Research Centre, London, shows (Fig. 1) that probably x major plague periods have occurred in the 77 years nee 1890*.

Fig. 1.-Fluctuations of the Desert Locust plagues, 1857-1966.

The alternation of plague and recession is a characteristic of all locust species and the amount of damage caused varies correspondingly. Thus, in the case of the Desert Locust during recent times, damage to agriculture valued at several millions of pounds was caused in one or two years, followed by a few years when damage still occurred but was much less; this period of reduced damage was in turn followed by a period of recession when swarming populations virtually disappeared and no damage occurred. It is the very great damage that can occur in the few years of peak infestation that makes this locust the most feared pest in its area of influence.

Locusts can exist in swarms as adults and in bands when in the nymphal (hopper) stage. Both are mobile, gregariously-behaving units containing large numbers of individuals; a swarm covering one square mile, for example, may contain 100 million insects (Roy, Gerbier t al. 1965). In the Desert Locust, hopper bands can move over a distance of several miles, and swarms over hundreds, even thousands, of miles, in the course of one generation. Each locust eats roughly its own weight of fresh vegetation a day (Davey 1954) and the daily consumption of one square mile of a Desert Locust swarm might well be as much as 200 tons of fresh plant matter.

The problem of damage becomes much more serious and the remedy much less simple as attention narrows from the international scale, through the individual countries affected, to the basic producing unit of the farm and to the farmer and his dependants. Of the total agricultural production of the Desert Locust invasion area, about 70% is in the form of subsistence farming; that is, the produce is grown largely to feed and clothe the grower and his dependants. Only the relatively small and irregular surplus is marketed and therefore the major portion of the production never enters the monetary system. Although the general trend is away from subsistence economies in all countries of the world, the process is a slow one,, dependent upon the production of marketable surpluses, the proximity of markets and expanding sources of demand, usually urban industrial populations. Any event that militates against this trend puts a restraint upon economic development in those areas of the world which most require continuous and rapid economic growth. The Desert Locust is, at times, a serious restraining factor upon local economic development. Locust attach may have a disastrous effect on the economy at the time of attack, and the risk of attack, even if it occurs rarely, may inhibit long-term development planning.

Two obvious remedies, financial aid and gifts of food from foreign governments, are both unsatisfactory. Financial aid to affected areas is only a palliative and, indeed, may create new sociological and economic problems in some areas. Supplying food is also unsatisfactory, partly because it creates political and monetary problems and also because the distribution of the free food is often both difficult and costly. For example, when the U.S.A. gave Ethiopia 20,000 tons of grain to replace some of the massive losses experienced in 1958 as a result of Desert Locust damage, the internal distribution of it cost the Ethiopian Government US$400,000 (FAO 1959b).

The only practical alternative at present is to attempt to control this insect by international control schemes. These are expensive since they involve the extensive use of aircraft and ground vehicles for both reconnaissance and actual control operations, and the use of costly chemical insecticides in large quantities. The cost of control measures against the Desert Locust during its most recent plague period was estimated to be of the order of £10 million per annum in the worst years, but the great advances that are being made in the techniques, tactics and strategy of control on the international scale can be expected to reduce average costs greatly.

Since a number of locust control schemes are now internationally organised, the question of the relative vulnerability of crops and countries to locust damage is frequently raised, with a view to apportioning the financial contributions of affected countries to supranational organisations in a logical and objective manner. Ideally such questions would be answered by reference to past damage records for each and all of the countries liable to infestation by the Desert Locust, but unfortunately locust damage to crops, though of frequent occurrence, is infrequently and irregularly reported and, even when reported, is rarely described in sufficient detail for any estimate to be made of its effect upon the yield. It is, therefore, impossible to build up a comprehensive picture of past damage from the available records.

A scale of the vulnerability of crop areas to Desert Locust attack can, however, be achieved by the utilisation of existing data concerning the frequency of locust infestation per unit geographical area for each month of the year, combined with crop-distribution data for the same areas, thus obtaining an index of the overlap in time and space of swarming Desert Locusts and crop areas. The purpose of this Memoir is to describe a method of quantifying the vulnerability of crop areas to Desert Locust attack by means of an index termed the 'Crop Vulnerability Index' (abbreviated to CVI), and to apply this method to all the more important crop areas affected.

Analysis of individual reports of the occurrence of damage proved totally inadequate as a means of establishing the relative vulnerability of constituent areas within the general Desert Locust invasion area, owing to the extremely uneven and inconsistent reporting of damage. Some countries at some times assiduously recorded damage, whereas others, unfortunately the majority, rarely if ever reported it even though circumstances, such as the presence of large swarming locust populations in areas known to contain extensive areas of cultivation, indicated that damage must have occurred. Reports of damage have also declined numerically in recent years; this is at least partly due to the increased use of aircraft for detecting swarms, it being usually impossible to detect crop damage from the air. However, the compilation and analysis of individual reports of damage have proved useful in assessing the range of crop plants attacked, the feeding preferences between crop species and varieties, and the relative incidence of damage attributable to the different locust stages (Bullen 1966).

The occurrence of locust swarms and hopper bands is much better reported. For 40 years such information has been accumulated in London, and for the last 24 years the Anti-Locust Research Centre (A-LRC) has been receiving and mapping it on a routine basis for most of the countries liable to infestation by the Desert Locust. This information comes from a variety of sources but by far the most important and consistent is that from the various national and supranational organisations responsible for the control of this pest.

The number of reports of swarms and hopper bands from a unit area over a given period of time could have been used as a measure of liability to crop damage, but this too suffers from the same deficiencies of differential under- and over-reporting as does the recording of damage incidence, although to a much smaller degree. Therefore the concept of frequency of infestation was used as a measure of liability to locust attack in the manner in which it has already been used in previous analyses of the distribution of the Desert Locust in * time and space (e.g. Waloff & Conners 1964) produced by the A-LRC.

The word frequency is used here to mean the number of months (or, in the case of annual-frequency maps, years) in which a geographical area has experienced one or more swarms or hopper bands over a period of 25 years. Thus the maximum possible frequency in all cases is 25. This attempts to overcome the basic unevenness of the reporting of individual swarms and bands.

Maps are available, prepared by the A-LRC, showing the frequency with which Desert Locust swarms and hopper bands have occurred in unit geographical areas for each month of the year during the 25-year period 1939-63 (see Maps 1 and 2 as examples). They are, as yet, only published, in a simplified form, in The Locust Handbook (A-LRC 1966). Maps are also available showing the annual frequency of infestation by hopper bands and by swarms; these show the number of years out of 25 in which each constituent geographic unit has experienced infestation by either hopper bands or swarms. A simplified version of the annual swarm-frequency map is shown in Map 3. The geographic units used are termed degree squares since they have sides equal to 1° of latitude and longitude.

Map 1.-Frequency of Desert Locust swarms, 1939-1963. April. (Number of years in which swarms were recorded during the 25-year period.

Map 2.-Frequency of Desert Locust swarms, 1939-1963. September. (Number of years in which swarms were recorded during the 25-year period.)

Map 3.-Cultivated area (indicated by dots) and Desert Locust swarm frequency. 1939-63

The geographic units termed degree-squares are not square and vary in area with latitude (the area of a degree-square varying from 3,061,800 acres at the equator to 2,258,200 acres at 43°N, the northern limit of the Desert Locust invasion area during the period 1939-63); nevertheless they are used as reference units for the calculation of Crop Vulnerability Indices because the locust frequency information is based on them. Locust frequency values are, to a certain extent, area-dependent, but errors in the calculation of the CVI due to variation in the area of a degree-square are small compared with other sources of error.

While the frequency method was chosen to reduce the effect of unevenness of reporting, in certain regions it is unsatisfactory. This is most evident in the desert areas of North Africa and Arabia where a number of degree squares have either suspiciously low frequency values or no records of ever having been infested even though there is circumstantial evidence of infestation. This is due to lack of observers in these areas rather than absence of locusts. However, for the purposes of the calculation of the overlapping distribution of swarming locusts and crop areas, this is not a serious problem since these areas have little or no cultivation and consequently low or zero CVIs on this basis alone.

Underlying any concept of an infested area is the assumption that, within the unit area used any part might have suffered attack. That is, attacks are expected to be essentially random. Certainly the part of a degree-square actually covered by swarms or hopper bands in any one month is only a small fraction of the total area. Recurrent features of local meteorology and constant features of local topography might render some sections within the area more liable to attack than others, though there is very little evidence to suggest that this is so. However, this would seriously affect the validity of the CVI only if, as a result, locusts were found predominantly in parts of the degree-square under crops.

In the absence of published maps showing in sufficient detail the quantitative distribution of every major crop grown in the Desert Locust invasion area, comprehensive up-to-date maps had to be constructed. From published and unpublished agricultural statistics referring to a year as close to 1960 as possible, the distribution of each of the most important crops was plotted on maps of the general invasion area, originally on a scale of approximately 1: 10 million, using a dot method of presentation. The sources of data used are shown in Appendix II. For some areas data were not available and crop areas and distribution had to be inferred from population data and a knowledge of the relative local importance of the various crops. This process is described in detail in Appendix I.

The unit area of cultivation represented by one dot was chosen to be 10,000 acres (about 4,000 hectares) for most of the crops whose distributions were plotted. Exceptions to this rule were certain crops of high value per unit area, such as citrus fruits and dates, for which a smaller acreage unit was used (either 2,500 acres or 1,000 acres). The general plotting unit of 10,000 acres was chosen because it allowed dots to be independently distinguishable even in the densest areas of cultivation likely to be encountered by the Desert Locust: for example, the states of Bihar end West Bengal in India, and East Pakistan. It proved, however, to be too large a unit for the representation of the small areas of cultivation of the oases in the desert areas of Africa and Arabia, and it was not very satisfactory for plotting the cultivation pattern of dispersed, low-density cultivation in north-central Africa, but these two types of area contribute little to the overall total cultivation of the gross invasion area of the locust. The cultivated areas of some of the oases mentioned above, although not attaining the standard of 10,000 acres, were nevertheless given this status so that the damage potential in these areas could be represented; this leads to slight over-estimation of the potential damage in these areas but the effect on the overall evaluation is negligible.

Since the locust-frequency data are already mapped on a degree-square basis, it follows that the areas devoted to different crops must be estimated similarly. Unfortunately crop distributions are known only for administrative units (Map 29) which vary from roughly the size of a single degree-square or smaller, as in parts of India and in Israel, up to tens of degree-squares as in Nigeria and Sudan. The distribution of the dots, representing the crop acreages, within the boundaries of the administrative subdivisions, was made subjectively after a careful consideration of the local variations in climate, irrigation, relief and population distribution. Some degree of error in position is inevitable in plotting but this is likely to be small enough not to affect significantly the grouping of the dot distribution on a degree-square basis, which is necessary in order to render it comparable with the locust-frequency data. A distribution map was prepared on these lines for every crop dealt with in this Memoir, and two such maps are published as examples (Map 4a for wheat and Map 17a for citrus). The others are not published, because the great reduction involved causes the dots to become invisible or indistinguishable, but all of them are deposited in the archives of the Anti-Locust Research Centre, London, where they are available for reference when required.

Map 29.-National and sub-area boundaries used for preparing the crop-distribution maps, of which Maps 4a and 17a are examples (Names and boundaries are those that were current at the time for which data were available.)

go to next section of this publication