Last update: March 26, 2012 02:45:58 PM E-mail Print






1 Plant Protection Research Institute, C/o Grootfontein College of Agriculture, Private Bag X529, Middelburg, 5900

2 Directorate of Biometric and Datametric Services, Private Bag X640, Pretoria, 0001

3 Plant Protection Research Institute, Private Bag X134, Pretoria, 0001



In 1932 the prickly pear moth, Cactoblastis cactorum (Berg) was introduced into South Africa and released between 1933 and 1944 for the biological control of prickly pear, Opuntiaficus-indica (Pettey, 1948). Unfortunately this insect not only attacks prickly pear, but also infests other opuntias, such as the spineless cultivars presently used as drought-resistant fodder plants and those cultivated for their fruit (Annecke et al. 1976).

Between 1969 and 1972, officers of the Plant Protection Research Institute screened several insecticides for the control of C. cactorum on spineless cactus. Eventually deltamethrin, methidathion and carbaryl were registered for this purpose (Burger, 1972; Bot et al. 1985). Recently, however, complaints have been received that carbaryl, the usual insecticide used for the control of the prickly pear moth, does not always provide adequate control, especially in home gardens where it is the only chemical approved for use. It was considered necessary to re-examine the chemical control of C. cacforum on spineless cactus. The feasibility of using systemic insecticides was also investigated as Zimmermann (1982) showed that systemic herbicides injected into the stems of O. ficus-indica were translocated throughout the plants.



Materials and methods

Five field experiments were carried out from 1983 to 1985 at the Grootfontein College of Agriculture, Middelburg, Cape, in an established cactus plantation containing the cultivars Chico, Fusicaulis, Monterrey and Robusta. Field-collected C. cactorum egg sticks were used in the experiments, care being taken to select typical egg sticks with approximately 60 to 80 eggs. For every experiment 15 plants were used per treatment and per untreated control. A few days prior to treatment, an egg stick was attached to each of five randomly selected cladodes (leaf pads) on each plant.


In the first experiment, the egg sticks were glued to the cladodes but the loss of egg sticks was substantial. Handmade, wax paper quills (Fig. 1), each containing an egg stick, were used in all subsequent experiments. These quills, similar to those used by Pettey (1948), were tubular, about 90 mm long and 10 mm in diameter. One end of the quill was open and the other closed. Near the middle and at one side of the quill, two circular holes were made to allow the first-instar larvae to emerge. After the egg sticks were inserted, the open ends of the quills were folded to keep out rain and debris (Fig. 1). The quills were pinned to the cladodes, with the circular holes against the cladodes. To prevent error, all field deposited egg sticks were removed from the test plants for the duration of an experiment.

Because quills were used, the ovicidal effects of the insecticides applied could not be determined.


Details of the insecticidal treatments applied are given in Table 1. Because of the various times taken for the eggs to hatch and for the larvae to penetrate the cladodes, the intervals between treatment and assessment were not constant. The insecticidal cover sprays were applied with a rucksack sprayer, care being taken not to contaminate adjoining treated plants. Ten cm3 of spray mix was sufficient to cover approximately 10 plants. Injection of monocrotophos into the main stems of each plant was done by driving a sharp iron peg, 15 mm in diameter, about 150 mm into the stem. The monocrotophos solution was then injected into the hole with a calibrated sheep-dosing gun. The number of main stems per cactus plant varied from two to four.

In assessing treatment effects, an egg stick was regarded as a unit and the following ratings were used:

  1. if the eggs hatched, but the larvae failed to penetrate the epidermis of the cladode, the unit was considered dead;
  2. if the eggs hatched and the larvae penetrated the cladode and were found feeding inside, the unit was considered alive.

Egg stick losses caused by dehydration or removal by ants were recorded but not used in the assessment.


Translocation of monocrotophos

A preliminary study on the translocation of monocrotophos was undertaken. Seven plants of similar size were injected with 5 cm3 40 % monocrotophos per main stem. An untreated plant acted as control. Two half cladodes were taken from the terminal and middle parts of each plant one, two, four and 12 weeks after treatment. The half cladodes were cut into small pieces and homogenized. Two 100 g sub samples were taken from the needed for high levels of protection may be somewhat homogenate, the monocrotophos extracted and the amount determined by gas chromatography (Zweig, 1972). For samples taken one, two and four weeks after treatment, the amount of monocrotophos was determined for each plant separately. For samples taken 12 weeks after treatment a composite sample of all seven plants was analysed.

A pilot trial was also carried out to determine the effect of monocrotophos on first instar C. cactorum larvae feeding on cladodes from the terminal and middle parts of the plants. On 14 November 1984, five plants were injected with 5 cm3 40 % monocrotophos per main stem. An egg stick in a quill was attached to each of five randomly selected cladodes at the terminal and middle parts of each plant. This procedure was repeated for five untreated control plants. Rainfall during the experiment was recorded and mortality assessments were made from 2 to 4January1985.


Statistical analyses


Experiment 1 was laid out sequentially but the remaining four experiments were laid out in randomized block designs (Van Ark, 1981) with the blocking factor being similar-sized and healthy plants situated relatively close together. For all experiments, 2 x 2 chi-squared tests were used for multiple comparisons of treatments within an experiment and also for certain treatments between experiments. For these tests the test level of a = 0,05 was divided by the number of comparisons made (Type 1 error protection).

For experiments 4 and 5 (Table 1) probit analyses (Finney, 1971) were also carried out to obtain the expected doses needed for high levels of protection. Unfortunately, only four doses could be used for each line and no doses resulting in protection of less than 50 % were applied. The probit lines may therefore be too flat (with too small regression coefficients) and the expected doses needed for high levels of protection may be somewhat excessive. Therefore, for cypermethrin (Experiment 5) a probit line was also calculated using the protection obtained for only 0,001; 0,002 and 0,003 % of the insecticide.

All analyses were carried out on a Burroughs B7900 computer, using appropriate biometric programs from the library of the Directorate of Biometric and Datametric Services.



The results of the five experiments are given in Table 2. The protection rates are ranked from lowest to highest, with the protection in the controls indicating the natural protection for all treatments. The protection rates were corrected with Abbott's formula (Finney, 1971) and these percentages protection are also given in Table 2.

The results of the probit analyses for Experiments 4 and 5 are given in Table 3, while the results of the translocation experiments are presented in Fig. 2 and Table 4.



From Table 2 it is clear that the prothiophos treatment did not prevent larval penetration into the cladodes. In Experiment 1 the registered rate of 0,149 % carbaryl (Bot et al., 1985) gave 96 % protection against larval penetration. However, in Experiment 2, only 72 % protection was obtained. A possible explanation for this significant (P ≤ 0,05) discrepancy is the difference in rainfall during the two experiments. Although more rain fell during Experiment 1 (109,1 mm) than Experiment 2 (87,3 mm), the first substantial rainfall (25,5 mm) occurred a fortnight after treatment in Experiment 1. However, the first substantial rainfall (12,0 mm) during Experiment 2 occurred only six days after treatment, probably washing most of the carbaryl off the plants before all the eggs had hatched. This would also explain the occasional complaints that carbaryl does not provide adequate protection against C. cactorum infestations. The efficacy of the very large dose of cypermethrin (0,02 %) was not affected by 12 mm of rain after application.

Cypermethrin cover sprays at rates of 0,004 % and higher resulted in 97 to 100 % protection. Considering the protection obtained for only 0,001; 0,002 and 0,003 % cypermethrin, 99 % protection can be expected from a treatment rate of 0,00424 % and 99,9 % protection from 0,0076 % (Table 3). However, such treatments will probably not kill the larvae already inside the cladodes, and it would be necessary to treat the plants before the eggs hatch. C. cactorum has two discrete generations per year. The moths start flying during February and October and the first eggs hatch during early March and early November (Pettey, 1948; Annecke et al., 1976). The residual effect of cypermethrin on first-instar larvae must still be determined, but it seems feasible that if treatments are timed correctly, a single application of cypermethrin during these periods should provide adequate protection against C. cactorum.

A cover spray of 9,024 % monocrotophos resulted in only 52 % protection, but higher doses may be more effective and cost-effective when compared to cypermethrin (see cost comparisons). Absorption of monocrotophos by the cladodes subsequent to spraying, must still be investigated. Stem injection of monocrotophos also failed to control C. cactorum adequately. As much as 638 cm3 40 % monocrotophos per main stem will probably be required for 99 % control (Table 3). The same total dose may also be needed if multiple injections of smaller amounts are made into each main stem. Chemical analyses showed that monocrotophos is translocated to the terminal cladodes (Fig. 2) and the possibility exists that other highly systemic insecticides may well be effective. Fig. 2 shows considerable and unexplained variations in the translocation rates between plants even though they were all about the same size, and this may explain the poor protection. The effect of monocrotophos on older larvae is also unknown. However, higher doses would probably be required to kill mature larvae, thus making stem injection of monocrotophos uneconomical.

From Table 4 it is evident that monocrotophos stem injections at 5 cm3 40 % gave somewhat better protection than the corresponding dose in Experiment 5. The difference between these protection rates was, however, non-significant at P ≤0,05. The higher protection rate may have been due to improved translocation as a result of 43,7 mm rain recorded after the injections of monocrotophos compared to the minimal rainfall recorded during Experiment 5.

It is doubtful whether stem injection of monocrotophos is feasible for controlling C. cactorum larvae, except perhaps when few plants must be treated (e.g. in home gardens), and the stems are easily accessible. However, other systemic insecticides may be more effective.

It is noteworthy that the natural protection against attack of C. cactorum larvae on the cladodes from the middle section of the plants was slightly higher (although non-significant at P ≤0,05), than on the terminal cladodes. This may be caused by the thicker and less penetrable epidermis of older cladodes. Further investigation of this possible natural control may be of interest, especially to improve experimental procedures.


Cost comparisons

The cost of an insecticidal treatment depends mainly on the density of the spineless cactus plants in a plantation. According to Burger (1972) 1 000 plants are usually found per hectare. If 1 cm3 of spray mix is sufficient to cover spray an averaged sized plant (2-2,5 m high and 2 m in diameter), the cost (at July, 1985 prices) of a cypermethrin application at a rate of 0,00424 % would be R26,71 per hectare. Applying the registered rate of 0,149 % carbaryl would cost R44 per hectare, nearly double that for cypermethrin. If it is assumed that a monocrotophos cover spray at a rate of 0,08 % may provide adequate control, the cost would be R30 per hectare.

The data in Table 3 suggest that 99 % protection against first instar larvae could be obtained by injecting as much as 638 cm3 40 0/0 monocrotophos into each of four main stems of the plants. Not only would it be impossible to inject this large amount of insecticide into a cactus plant, but the cost (R38 280 per hectare) would be prohibitive).

It is concluded that cover spraying with cypermenthrin is a promising control measure. The effectiveness of monocrotophos cover sprays at rates of 0,06 to 0,1 % must still be determined.



Mr A. J. du Plessis of the Grootfontein College of Agriculture is thanked for photographic services. The continued interest and advice of Dr. J. D. Mohr and Dr H. G. Zimmermann of the Plant Protection Research Institute is greatly appreciated. Dr D. P. Keetch of the same Institute is thanked for reading the manuscript and making helpful suggestions.



ANNECKE, D. P., BURGER, W. A. & COETZEE, H., 1976. Pest status of Cactoblastis cactorum (Berg) (Lepidoptera: Phycitidae) and Dactylopius opuntiae (Cockerell) (Coccoidea: Dactylopiidae) in spineless cactus plantations in South Africa. Journal of the Entomological Society of southern Africa 39: 111-116.

BOT, J., SMIT, C. J. B., SWEET, SEUGNET & HOLLINGS NORA, 1985. A guide to the use of pesticides and fungicides in the Republic of South Africa. Government Printer, Pretoria, 312 pp.

BURGER, W. A., 1972. Control cactoblastis and cochineal. Farming in South Africa 48 (12): 6-8.

FINNEY, D. J., 1971. Probit analysis. Cambridge Univ. Press, London, 333 pp.

PETTEY, F. W., 1948. The biological control of prickly pears in South Africa. Science Bulletin Department of Agriculture Union of South Africa 271, 163pp.

VAN ARK, H., 1981. Eenvoudige biometriese tegnieke en proefontwerpe met spesiale verwysing na entomologiese navorsing. Wetenskaplike Pamflet Departement van Landbou en Visserye Republiek van Suid-Afrika 396, 117, pp.

ZIMMERMANN, H. G., 1982. Control of prickly pear. Weeds Bl. 1/1982,FarminginSouthAfrica.2pp.

ZWEIG, G., 1972 (Ed.). Analytical methods for pesticides and plant growth regulators. Vol. VI: 293-298. Academic Press, New York.



Phytophylactica 18 (3)