Julie Stanton, Department of Primary Industries, Indooroopilly, Tony Pattison and Stewart Lindsay, Centre for Wet Tropics Agriculture, South Johnstone.

Abstract

Existing banana crops in north Queensland were used to determine the economic and action thresholds. At a crop value of $25,000/ha/year and with nematicide application costs of $1,600-1,900/ha/year, the yield loss to Radopholus similis must be 6-9% to warrant nematicide application. By sampling plants just after bunching, we found that 6-9% yield loss is caused when the disease index is 15-20. To prevent this loss, nematicide must be applied to that pseudostem as it is developing, ie. in the previous bunching cycle. We found that disease index increases about 4.5 in nine months, which is the length of the bunching cycle in north Queensland. Therefore, the action threshold for nematicide application is a disease index of about 10-15.

Introduction

Nematicide application for control of burrowing nematode on bananas costs $1,600-1,900/ha/year. At the present crop value of $25,000/ha/year, this equals about 6-9% of crop value. If growers do not know the severity of their nematode problem, they cannot make informed decisions on whether or not to apply nematicide. To warrant nematicide application, the yield loss to be prevented should be at least 6-9%.

The disease index (Broadley 1979a) of a crop is determined by taking root samples from the bunching pseudostem. Sampling is best done at bract fall which is the physiological stage at which no new roots are produced but before older roots have senesced. Because of the asynchrony of bunching in north Queensland, this may be done every 3-4 months. To obtain an estimate of the nematode status of a crop with 80% accuracy, twenty samples are taken uniformly throughout the crop (Stanton et al. unpublished). Roots are split lengthwise and rated for the proportion of cortex occupied by R. similis lesions and disease index calculated (Broadley 1979a).

Estimates of populations of R. similis which cause yield loss to banana is very variable worldwide. In West Africa, 1000 R. similis/100 g roots are considered to cause serious yield loss while 20,000 R. similis/100 g roots are required to cause similar loss in central America (Gowen 1995). However, Gowen and Quénéhervé (1990) consider that 2000 R. similis/100 g roots are a potential cause of yield loss in commercial cultivars.

In north Queensland, previous work has shown a response to nematicide with 500 R. similis/100 g roots (Broadley 1979b). In nematicide trials, no response to nematicide was seen when the disease index (Broadley 1979a) was about 25. However, there was a response when the disease index was about 40 (Schipke & Ramsey 1994).

The aim of this study was to determine the action threshold of the nematode. To achieve this, it was necessary to determine:

the Economic Threshold, i.e. the nematode severity at bract fall which causes yield loss equal to the cost of controlling the nematodes.

the Action Threshold; nematicide application at this stage will prevent the nematode severity reaching the economic threshold during the following bunching cycle. To determine the action threshold, we need to know the rate of increase of nematode severity from year to year.

Population dynamics of burrowing nematode was studied in two ways.

Mixed growth stage sampling used small areas in a crop which were used to collect information on changes in disease index. However, they consisted of plants at various growth stages so were not useful for collecting information on plant growth.

Bract fall stage sampling used whole crops so that there were always a number of plants at bract fall growth stage to determine the relationship between plant growth and nematode severity. These sites were not useful for determining changes in nematode populations because nematicide was applied to these sites by growers.

Materials and methods

Fourteen sites in existing crops in north Queensland were chosen to represent the range of soil management types in the region. These were used for mixed growth stage sampling and/or bract fall stage sampling as indicated below.

Mixed growth stage sampling Three adjacent rows of 20 mats were marked in each crop and used throughout the project. Where possible, these were maintained free of nematicide. In crops where nematicide was applied, two rows were kept free of nematicide to allow assessment of the effect of nematicide.

Every three months from February 1994 to August 1996, four plants in each row were sampled to determine the disease index of roots, no. nematodes/100 g root and no. nematodes/200 g soil.

Bract fall stage sampling Six crops were sampled every three months from March 1995 to determine the disease index of roots, no. nematodes/100 g root, plant girth at 75 cm and estimated no. fingers/bunch (Table 1). The last two characters were used to estimate potential bunch weight (Table 1). Only plants at the bract fall stage, i.e. just before bagging bunches, were sampled.

Table 1. Estimates of potential bunch weight at bract fall (J. Daniells unpublished)

Mixed growth stage sampling There was no consistent seasonal effect on disease index. Nor was there a relationship between population dynamics and soil type (data not shown). Comparison of disease index over time showed that disease index in the most heavily infested crops increased from an average of 15 to 22 in the 30 months of monitoring. However, the range of change in disease index over the 30-month sampling period was -27 to +29. Of those farms where an increase was observed, the mean increase was 14 over 30 months which is about 0.5 per month and about 4.5 in nine months, the average bunch cycling time in north Queensland.

Bract fall stage sampling Most pairs of variables were significantly correlated (Table 2).

Calculating economic and action thresholds The correlation between disease index and bunch weight estimated on no. fingers was greater than that between disease index and bunch weight estimated on girth so the former was used to calculate economic threshold and action threshold (Table 3) using the equations in Table 2.

To calculate economic threshold, it is necessary to determined the disease index when yield loss equals the cost of nematode control. At current prices, this is about 6-9% yield loss. Table 3 shows the calculation of economic threshold for six crops using no. of fingers to estimate bunch weight.

Therefore, when yield loss is 6-9%, disease index (i.e. economic threshold) is approximately 15 - 20. Using data obtained in the mixed stage sampling which showed that disease index increased 4.5 in nine month, the action threshold in north Queensland occurs when the disease index is about 10 - 15, i.e. 4.5 less than the economic threshold.

Table 2. Equations relating variables measured in bract fall stage sampling.

Discussion

There was great variability in the rate of change in disease index over time. In some crops, the disease index increased, in some it decreased while there was little change in others. This variability was not obviously due to soil type or nematicide use. Differences in crop management affect plant growth and, therefore, reproduction of the nematode and this could contribute to the variability. In addition, the sampling technique is only 80% accurate.

A moderately conservative approach to recommendations of nematicide application would be based on an action threshold derived from the average change in disease index on crops where an increase in disease index was found, i.e. about 0.5 index point/month. However, it is likely that growers could finetune the action threshold for individual crops by monitoring the crop over several seasons and using that information to predict how quickly the economic threshold will be reached.

Equations developed in this study show that, when the disease index was 10, 15 and 20, root populations of R. similis were about 1800, 2700 and 3600, respectively, when all growth stages were assessed. When only plants at bract fall were assessed, the equivalent populations were approximately double. These values are consistent with those estimated to cause yield loss in other parts of the world (Gowen and Quénéhervé 1990).

Broadley, R.A. (1979a) - A simple method for estimating banana root rot. Australasian Plant Pathology 8: 24-25.

Broadley, R.A. (1979b) - Non-volatile nematicides for control of burrowing nematode in banana plantations in north Queensland. Australasian Plant Pathology 19: 626-630.

Gowen, S.R. (1995) - Bananas and plantains. Chapman & Hall, London.

Gowen, S.R. and Quénéhervé, P. (1990) - Nematode parasites of bananas, plantains and abaca. In Plant parasitic nematodes in subtropical and tropical agriculture. (eds. Luc, M., Sikora, R.A. and Bridge, J.) CAB International, Wallingford.