Last update: April 2, 2012 12:28:05 PM E-mail Print


Karoo Caterpillar Research


J.D. Mohr,

Plant Protection Research Institute,

c/o Agricultural Research Institute,

Middelburg, Cape.




Because the Karoo caterpillar Loxostege frustalis Zeller competes with stock for food it has been considered a pasture pest for many years. The pioneer entomologist C P Lounsbury addressed the Zwart Ruggens Farmers' Association in Aberdeen on this subject before the turn of the century (Lounsbury 1897).

Owing to the extensive nature of Karoo farming, the relatively low monetary value of the veld and the sporadic nature of the pest, entomologists have always felt that biological control and/or the manipulation of management practices offer the only likely solution to the problem. It is these two approaches and the thinking behind current research directives, which are outlined below, essentially as a summary of the ideas contained in a review by Annecke and Moran (1977).



Between 1942 and 1952, twelve species of insect parasitoids (that is, parasites which invariably kill their hosts in the course of their development), which normally attack an insect, which is closely related to the Karoo caterpillar, were imported from North America and Canada. Local entomologists intended to supplement the apparently sparse indigenous natural enemy complex associated with the Karoo caterpillar. One of the imported species, Chelonus texanus Cresson, was mass-reared in Pretoria and more than 8,5 million were released, mainly at 491 sites in the Karoo and Orange Free State.

During and after this project a detailed study by S.J.S. Marais revealed that L. frusta/is has a complex of 19 native parasitoids (Marais, 1955). Twelve are primary parasitoids and kill the Karoo caterpillar itself; the rest are secondary and attack the primary parasitoids, so that they indirectly affect the caterpillar population. New information has resulted in slight modifications to this interaction but still, several of the primary and secondary parasitoids, as well as a hitherto unrecorded fungus disease, play an important role in determining Karoo caterpillar numbers. One of the primary parasitoids is Chelonus curvimaculatus Cameron. Initially it was mistaken for the imported C. texanus, which it resembles, and led to the erroneous belief that C. texanus had become established in the Karoo.

None of the imported parasitoids established and, in sharp contrast to many highly successful biological control programmes in South Africa, this initial project against the Karoo caterpillar was unsuccessful. Annecke and Moran identified the lessons to be learned and insisted on a different approach when they wrote that "In retrospect, it is clear that pre-introduction studies on L. frusta/is and its parasitoid complex were completely inadequate even for a preliminary hit or miss attempt at biological control. An almost complete lack of understanding of the ecology of the target insect, incorrect reports of recoveries of the introduced parasitoids and political pressure, maintained the impetus in this expensive and abortive programme for over a decade. The lesson is obvious, but even now, in the late 1970's, our knowledge of L. frustalis population dynamics remains scanty".

These views show a need for a fundamental study of Karoo caterpillar biology, including an extension of Marais' work to determine the value of native natural enemies. This does not imply that further importations will necessarily follow. In fact, Annecke and Moran were sceptical in this regard when they noted that "All in all, we regard it as improbable that parasitoids of related species of Loxostege elsewhere in the world will be found that will improve, even marginally, the natural regulation of Karoo caterpillar populations that already exist in the Karoo". The point is that one must know how effective the local natural enemies are in order to evaluate the feasibility of further introductions.



In their review Annecke and Moran proposed an important hypothesis concerning the influence of Karoo encroachment on the Karoo caterpillar. They wrote that "The vegetation in the Karoo has been subjected to fundamental changes over the past 150 years and these changes are of paramount importance to the modern status of the Karoo caterpillar as a pasture pest. The vegetational changes are the consequences of colonization of the Karoo by man and his animals. Over very large areas the sweet grassveld climax has given way to bushy desert intrusions. Two kinds of change have taken place (Acocks 1975), namely, an expanded distribution of the Karoo bushes, some of which are pioneers of deterioration, and - perhaps more important - an increased density of these species. Many of the Karoo bushes now replace the sweet grasses as the major source of food for sheep. During the past 150 years the numbers of sheep and other stock rose, the desert veld types advanced northeastwards, the grasses gave way to woody Karoo bushes, and erosion tended to stabilize the botanical deterioration. Meanwhile, insects that feed on Karoo bushes must have gradually assumed increasing importance to man and his animals. Among these, and certainly the most spectacular, is the Karoo caterpillar. . .".



The foregoing highlights several very practical considerations. If the size of Karoo caterpillar populations is directly proportional to the number of host Karoo bushes, a doubling, say, of the plant numbers will be accompanied by a twofold increase in the caterpillar population. In other words, the caterpillars per plant will be constant at different plant densities (A in fig. 1). However, the hypothesis suggests an increase in the numbers of caterpillars per plant at higher plant densities (B in fig 1). In this case more than a twofold increase will result from a doubling of the plant population. Such a relationship is very common in nature and the hypothesis is founded on established ecological principles. This augurs well for future attempts to curb the caterpillar, for the reverse would also be true. Following the above arguments, if the plant population is halved, the reduction in the pest population - and the damage caused by it – will also be more than halved. The chances are that a threshold density of hosts exists below which pest populations are unable to "escape" and lead to the periodic severe outbreaks which are characteristic of the Karoo caterpillar. Obviously it is this threshold density which needs to be determined.

An equally practical issue was raised by the above authors and further emphasized in quoting J P H Acocks: " . . . the sheep population reached a maximum as a result of fencing (the veld) into camps with water, on the recommendation of the Drought Investigation Commission. Because fencing without proper management of grazing finally finished off the grass constituent of the veld and made the farmer almost entirely dependent on the inferior 'rusperkos' bush (i.e. food plants utilized by Karoo caterpillars) which partly replaced the grass, the caterpillar became a national problem". It is so that grass (and some bushes) cannot be utilized by the insect and an increase in the number of non-host plants, provided they are palatable to stock, will lessen the impact of outbreaks. Such a situation is illustrated in fig. 2 where, in the event of a Karoo caterpillar outbreak, alternative grazing would be available in the camp with more grass.



The food-value and successional status of the various Karoo plants, and the merits of the different grazing practices, constitute a stimulating and contentious subject. From an entomological point of view it is essential that the plants - especially the species of Pentzia, the principal and extremely abundant hosts of the caterpillar - should be evaluated, not only in relation to the caterpillar, but also their overall agricultural and ecological desirability. Only then will it be possible to measure the impact of the pest in real economic terms and to formulate practical suggestions for control with due regard to conservation and the economic viability of farms.

In those areas where "erosion (has) tended to stabilize the botanical deterioration" vegetational changes to ameliorate the caterpillar problem will be a slow and difficult process. However, on the fringes of the Karoo, where desert encroachment is still only a threat, the position must be viewed in a different light. Here the Karoo caterpillar is a potential problem which should not be overlooked. Vegetational deterioration may very easily and quickly - almost suddenly - lead agriculturalists to elevate what they now regard as weeds, to "Good Karoo".

This article is not intended to imply that panaceas for veld improvement or Karoo caterpillar control exist, or to allocate a special status to the insect as a factor in determining optimum veld composition. But perhaps it is a hitherto unrecognised - and rather influential reason to reconsider our current grazing practices which stabilize the deteriorated Karoo. Neither is it intended to create an impression that only weeds are attacked or that good plants are immune. On the contrary, the reverse is often true, with degrees of susceptibility in both desirable and undesirable groups of plants.

The closing remarks in the Annecke and Moran review effectively summarizes the situation: "Essentially the L. frusta/is problem should be viewed as man-made and alleviation may as readily be achieved by pasture scientists as by entomologists. J P H Acocks (in litt. 1976) adds substance to this view". . . if the perennial climax grasses were more plentiful, as they should be, the sheep would need to make less use of Pentzia incana and no use of P. globosa. In the face of competition from such grasses, these bushes would also become very much less common, which would increase the grazing value of the veld because they would to some extent be replaced by other bushes more palatable to sheep and more compatible with grass. These bushes are the now-scarce ones that were grazed out along with the grass and associated herbaceous plants. Some, at least, of them are palatable to the caterpillar too but it would be as well to note which, if any, are not attacked by it. This information will be needed if a serious attempt is ever made to restore the original vegetation, for the reason that reseeding to the climax bushes will be just as necessary as reseeding to the climax grasses. It should be noted that Karoo without grass is completely unnatural, each and every type of Karoo having its own quota of perennial grasses and herbs capable of filling the spaces between and under the bushes. One way and another it does appear that if the attempt to destroy the caterpillar were to succeed, it would merely tend to have the effect of perpetuating the ridiculous methods of grazing that have destroyed the grass, the favourite food of the sheep, and replaced it with the favourite food of the caterpillar. It seems clear that pasture management and sound entomology should go hand-in-hand in attempts at solution of the Karoo caterpillar problem and that these studies should be well advanced before any renewed attempts at biological control of L. frustalis are even contemplated".



The immediate objectives are to assess the influence of indigenous natural enemies and host plants on the Karoo caterpillar. Larval cases (sometimes incorrectly called cocoons) are collected from the soil and analysed to obtain population estimates of the host and natural enemies which kill the fully-grown larvae.


The natural enemies

(i) It is now known which natural enemies cause the most mortality. "Natural regulation", however, also depends upon a higher percentage kill at greater host densities - on the same principle as the relationship between the caterpillar and its host plants - and the natural enemies are being evaluated in this context. The indications are that some of the organisms are quite effective and continued research along these lines is justified.

(ii) The interactions of the parasitoids with the host, and each other, occupy much of the research effort. No less than 28 associations have been confirmed and there are undoubtedly more. In this respect the hyperparasitoids, because of their considerable influence on the primary parasitoids are clearly important; at least two of the hyperparasitoids are very abundant. Through the study of less common species it may be possible to identify shortcomings in the natural enemy complex, as well as characteristics which should be avoided in candidate biological control agents. Collectively the natural enemies associated with the pest are a major component of the general biological investigation which is the essence of any pre-introduction study. Also, familiarity with all the species is the only safeguard against the disastrous consequences of misidentifications.

(iii) Inadequate information about the survival of other stages of the caterpillar is still a serious limitation. Factors which affect mating, oviposition and the survival of young larvae must be quantitatively assessed before any single factor can be measured against overall survival.


The host plants

(i) Studies of the host plants of L. frustalis revolve around a few species of Pentzia. Their suitability as hosts, and sheer numbers, dictates that they are the most important food source. Counts of the caterpillar are made at different plant densities (to test the Annecke - Moran hypothesis), and in contrasting vegetation types such as that illustrated in fig. 2. The concept of critical threshold densities, below which larvae should be unable to successfully move between plants - as suggested by the eminent ecologist T R E Southwood - has not been examined, but is logically a high priority.

(ii) A sound evaluation of the principal host plants, from both pastoral and entomological viewpoints, is essential if current grazing practices are to be objectively reviewed. The responsibility of entomologists, however, extends to an appraisal of the alternative host plants. As pointed out by Acocks, improved pasture must be less vulnerable to the caterpillar, preferably with a greater stock carrying capacity. This work is just beginning: several species of plants have been sampled in the field and are included in laboratory feeding trials. As already implied, this promises to be a fruitful area for research. For example, the plants on which the caterpillar now mostly occurs are not necessarily the preferred hosts - it is arguable that the insect has "switched" to the most abundant food supply.

(iii) A feed-back from pasture scientists will ensure that the right plants are selected for study by entomologists who, in turn, will then be in a position to make a useful contribution towards defining optimum veld composition. Ideally the matter should be viewed in the light of the botanical succession, an immense undertaking in itself. This will enable everybody to strive for the more desirable climactic condition and, hopefully, ease some of the controversy which surrounds our ideas on the usefulness of many of the Karoo bushes and grasses.



I am indebted to dr D P Annecke (Plant Protection Research Institute, Pretoria), Prof V C Moran (Rhodes University, Grahamstown), Mr M W Pretorius (Plant Protection Research Institute, Middelburg) and Dr P W Roux (Agricultural Research Institute of the Karoo Region, Middelburg) for their constructive comments on the manuscript.



ACOCKS, J.P.H. 1975. Veld types of South Africa. Mem. bot. Surv. S. Afr. 40: 1-128.

ANNECKE, D.P. & V.C. MORAN. 1977. Critical reviews of biological pest control in South Africa. 1. The Karoo caterpillar, Loxostege frustalis Zeller (Lepidoptera: Pyralidae). J. ent. Soc. sth. Afr. 40: 127-145.

LOUNSBURY, C.P., 1897. Report of the Government Entomologist for the year 1896. Cape of Good Hope. Department of Agriculture. Government Printer, Cape Town.

MARAIS, S.J.S. 1955. The biological control project against the Karroo caterpillar, Loxostege frustalis Zeller, in South Africa. (Undated, probably 1955). Unpublished report in two copies (1 held at Plant Protection Research Institute, Pretoria, the other at Grootfontein, Middelburg). 142pp.



Karoo Agric 1 (4), 16-20