Discussion

The bumper crop storage systems described provide alternatives to traditional systems: bag warehouses are replaced by flexible silos or cover and plinth and for bulk grain, aerated concrete/steel silos are replaced by various types of bunker. The bumper crop stores are usually uniquely suitable for one location; what works in Turkey may not be suitable for Zimbabwe and vice versa. (Most of the systems described were managed by marketing boards but could equally well be privately managed). Where choice of site with history of rodent activity is unavoidable, anti-rodent measures are essential.

Bumper storage methods should not be evaluated in isolation because they are part of an existing marketing system whether they are in bag or in bulk. The latter is usually more capital intensive and requires a technically skilled type of management, while bag storage requires constant but relatively unskilled attention. Analysis of bumper storage systems raises questions for management.

· Can local skills and resources be mobilized as in Burma and India? These resources are often quicker to procure locally and also save foreign exchange. If local resources are inadequate, flexible silos provide a tested alternative for bag storage.
· If, as in Zimbabwe, bags must be imported, are bunker stores practical, economic alternatives to bag storage?
· If bunkers are economic, what aeration/moisture management will they require? Mechanical aeration is needed for bunkers in parts of the United States when moisture content is more than 12%. This question of moisture management requires immediate research and development if grain losses and connected delays in removing damaged grain are to be avoided.
· If bunker storage is adopted, has the agency the necessary engineering expertise to manufacture solutions to local problems like the Australian 'Lobster'? Total dependence on outside skills complicates maintenance and prevents innovation.
· Bulk storage may be adopted because of a coincidence of circumstances. In Zimbabwe there was need for reserve stocks in a year which simultaneously produced a bumper harvest and in which bag prices doubled. In these circumstances a sense of urgency may dominate evaluation. For example annual costs may be calculated and employed for evaluation purposes, but bumper storage may occur infrequently and use of capital costs may be more appropriate because the store is employed for only one season. Morley (1988) has recognized this point and has advocated the hire of tarpaulins rather than purchase where possible (see below).
· When the need for bumper crop storage is intermittent there remains the problem of caring for the equipment and sheeting when it is unused and therefore vulnerable to damage (O'Dowd and Kenneford, 1983). It has been suggested that hiring tarpaulins weekly, where practicable, ensures their quick return and low wear and tear (Morley, 1988).
· When bumper storage is intermittent, purchase of high capital cost handling and bulk transport equipment may be uneconomic. If therefore a bag handling system is considered, is sufficient labour available? Alternatively if bunker/bulk storage is adopted will the decision to handle in bulk rather than in bag cause problems locally?

In an attempt to resolve some of these questions and provide a guide to bumper store selection a performance profile may be used. The performance profile is a concept derived from a management interview technique (Ansty, 1987). In this instance the profile is divided into three parts covering different sectors of performance.

1 Resources, capital costs, management and local resources.

2 Operations include five standard operations as a guide; these can include relocation, and speed of procurement can include speed of obtaining sacks.

3 Quality control includes ease/effectiveness of pest control, aeration and inspection required. In addition, ease of storm proofing is included because this embraces features such as the likelihood of strong winds and storms bringing rain into contact with the grain.

Management specify the level of resources, the ease of operations and the degree of quality control required. This step requires management to have detailed knowledge of local requirements. For example, a local requirement might be for a system with average (medium) cost, below average technical management, ample local labour (resource), fast loading, good pest control, but below average storm proofing (see Figure 2). When these have been pencilled in, the characteristics of the alternative systems are marked with hatching over the requirement profile, when they can be compared. In Figure 3, CAP storage has lower capital cost/tonne, lower management demands than required but uses local resources (labour and materials). All operations are more than up to requirement but CAP is under requirements for pest control, aeration, inspection and storm proofing. Flexible silos (see Figure 4) cost more, demand more technical management than CAP, but use fewer local resources. They are easier to site/erect than required, though slower to load than required.

Figure 2: Field performance profile local requirement

Figure 3: Field performance profile, cover and plinth storage

Figure 4: Field performance profile, flexible silo storage

The silos are superior on pest control and are more storm proof, but are below requirements on aeration and inspection.

The field performance profile enables any number of relevant factors to be considered. Final choice could be assisted with cost-benefit analysis and depends on the local priorities. In this example, where cost is not limiting, flexible silos have the edge with more effective pest control, but both systems have aeration problems. The profile is a searching technique which is deliberately qualitative rather than quantitative, and seeks to reveal qualitative aspects in system comparison which will then have quantitative effects in cost-benefit analysis. It requires a sound assessment of the attributes of a system and is therefore best applied with established techniques.