Much thought was given to how best to evaluate the incidence of banana weevil larvae and damage resulting from their feeding. I felt that the existing "Percentage Coefficient Index" (PCI) systems, of VILLARDEBO (1973) and MITCHELL (1978), had an inherent limitation because they permit only the evaluation of tunnelling by scoring the damage seen around the periphery of a banana corm.

STOVER & SIMMONDS (1987) cited evidence which showed that the number of fruits in the bunch was closely coordinated with the area of the third to eighth last leaves. Other work had shown that the banana plant produces leaves in excess of its needs to produce fruit, as some 20% or more of leaf area could be removed without significant loss in fruit weight. A banana plant will produce some 41 leaves prior to the emergence of the aerial stem which will bear the flower. The growth of aerial tissue, initially leaves and finally the stem proper, is central, so, the earliest leaves are on the outside. The early leaves have a short lifespan, 30-40 days only, but the later leaves survive for 125-165 days. At flowering, some 375 days after emergence of the first leaf, 12-16 leaves are surviving and at fruit maturity, at an age of around 470 days, the last 8-9 leaves remain. The sheaths of leaves remain well after leaf death, so, at any one time, a transverse cut of the pseudostem will show some 15 leaf sheaths. The outer leaf sheaths are crescentic, each covering around half the circumference of the pseudostem. Generally four leaves overlap to make up the full circumferential layer of about 12 cm in thickness at the junction of the pseudostem and the corm. Alternative evidence shows, by deduction, that at flowering there are some 7-8 leaf sheaths exterior to the all important final eight leaves which will provide the nutrient for the maturing fruit.

From the foregoing it can be seen that at the stem-corm junction of a 30 cm diameter stem the outer half, 7.5 cm thick around the perimeter, is made up of leaves which have no importance for production and yield. It is the central 15 cm diameter core that is of direct importance and damage to it is what will cause crop loss. Because it is at the stem-corm junction that the PCI assessments are normally made, it seemed fair to conclude that such techniques cannot reveal the amount of damage to the central, all-important core area. The usefulness of the PCI technique, therefore, is restricted to giving an indication of larval population levels, if, that is, numbers of peripheral feeding tunnels do reflect larval numbers. Even that assumes that the tunnels do not simply meander around the periphery. It may be that erosion of the outher layers can affect the mechanical strength of the pseudostem and thus render the plant more liable to stem snapping under stress from, for instance, wind or heavy bunches.

Thus, I decided to endeavour to design a gauge that would reflect the perceived greater importance of the central core area of the corm. It was found that the clones found in Kagera banana shambas, when in good condition, had a diameter of around 30 cm and so the starting point for the gauge is a circle of 30 cm diameter which is subdivided into five subcircles each of equal area. The total area is 706.86 cm² therefore, each subcircle has to have an area of 141.37cm² and the radii of the five subcircles are 15 cm, 13.4 cm, 11.6 cm, 9.5 cm and 6.7 cm respectively. An arbitrary score of 20 points is allocated to each subcircle but, to allow for the probability that the more central the damage the greater its significance to fruit production, the nearer the centre the subcircle is the fewer the number of its subdivisions for scoring purposes. So, the subdivisions are - the innermost 6.7 cm radius subcircle has 4 segments scoring 5 points; the next 9.5 cm radius subcircle has 8 segments scoring 2.5 points; the next 11.6 cm radius subcircle has 12 segments scoring 1.6 points; the next 13.4 cm radius subcircle has 16 segments scoring 1.3 points; the outer 15 cm radius subcircle has 20 segments scoring 1 point. The gauge can be made of a transparent material, such as thin acrylic sheet.

After harvest, given that pre-harvest sampling is not likely to be possible in on-farm trials, the pseudostem should be cut transversely at the stem-corm junction. The gauge is then laid flat on the top of the stump. Penetration of any segment on the gauge should be scored once only; two tunnels in a segment, for instance, do not double the score. If all segments are penetrated a score of 100 would be given. Thus, using this scoring system, damage to a 30 cm diameter stem can be scored directly as a percentage. As it is a direct scoring system with no requirement for subjective assessment, there should be minimal variation attributable to individual technicians. The system will work for smaller plants, for example, a 13.4 cm diameter stem could be scored using the gauge and a percentage figure derived from the score out of a maximum of 80. Below a stem diameter of 13.4 cm with a percentage from a maximum of 60, the stem is not very likely to bear fruit.

Example of gauge in use. Score=40 % of maximum Composite simulation

First tests of the gauge showed its use to be extremely simple and it was hoped that, as data became available, comparison could be made between the PCI, which can be determined with this gauge from the number of segments of the outermost subcircle which have larval tunnels, and the damage as evaluated using the gauge. The intention was that as the plants in the various trials came to maturity, the relation between damage, scored using the gauge, and bunch weight would be examined and compared with the relation between the PCI and bunch weight.

Although the opportunity to fully test the gauge in Kagera did not materialise, the doctorate research studies by my protegé, Daniel Rukazambuga, included an assessment of methods of evaluating weevil damage (RUKAZAMBUGA, 1996), see Postscript.

NEXT

href="\bananas2000\gauge.htm"