Contents - Next

Since 1977, FAO has given priority to the prevention of post-harvest losses, particularly through action to reduce losses at the farm and village level. A serious constraint for developing countries in organizing and implementing post-harvest loss prevention programmes is a shortage of trained national staff.

The FAO Action Programme for the Prevention of Food Losses has undertaken a training programme during which several workshops were organized to train African technical officers to reduce post-harvest losses. The workshops covered various aspects of crop storage and processing, including storage pests and their control, loss assessment, drying, storage, grain processing and socioeconomic implications.

The material in this manual has been tested during training courses, and has also been reviewed on the basis of experience gained. It is now being published as part of the FAO Training Series.

I trust that this manual will help to provide practical training for those responsible for the prevention of post-harvest food losses in developing countries.

D.F.R. Bommer
Assistant Director-General
Agriculture Department


This manual presents material from a wide range of disciplines associated with the prevention of food losses in, particularly, cereals, pulses, roots and tubers. It is directed at field staff, project supervisors and extension personnel involved in food-loss prevention programmes.

The manual should serve as a single-volume reference work on food-loss prevention during storage. While the approach is technical in the sections on drying, processing and loss assessment, the economic and social aspects of food losses are also considered.

It is hoped that participants in training workshops will find that the manual meets their basic needs, and that it will be supplemented by detailed worksheets covering special topics (especially for practical work), and by handouts on subjects of particular local importance.

1. Introduction

1.1 Definitions

It is important to understand the principles of sampling and measurement. Only then may the procedures defined here be adapted to suit local conditions, with reasonable confidence that measurements will be valid. The terms defined in the following list are commonly used in dealing with post-harvest food losses.

Post-harvest. The period between maturity of the crop and the time of its final consumption.

Food loss. Any change in the availability, edibility, wholesomeness or quality of food that reduces its value to humans.

Direct loss. Loss by spillage or consumption by insects, rodents and birds.

Indirect loss. Loss caused by a lowering of quality leading to rejection as food. This type of loss may be locally defined and related to custom.

Losses of crop product. Crop products may be lost from the food chain at any or all of the periods between planting and preparation for immediate consumption. Three general periods have been identified.

(a) Pre-harvest losses occur before the harvesting process begins and may be due to such factors as insects, weeds or diseases afflicting the crop.

(b) Harvest losses occur during the harvesting process and may be due, for example, to shattering and shedding of the grain from the ears to the ground.

(c) Post-harvest losses occur during the post-harvest period.

Post-production losses. Losses consisting of the combined harvest losses and post-harvest losses.

It is always difficult to distinguish clearly between the arbitrarily defined stages from production to consumption. The maturing/drying/processing periods will often overlap during the post-harvest period, as, for example, in the fielddrying of maize after it has reached maturity. There is nothing to be gained by defining rigid boundaries and making artificial distinctions between overlapping stages. It may be preferable to relate losses to a process or operation rather than to a definite period.

Food. Those commodities which people normally eat: the weight of wholesome edible material, measured on a moisture-free basis, that would normally be consumed by humans. Inedible portions of the crop, such as stalks, hulls and leaves, are not food. Crops for consumption by animals are not considered food. Post-harvest loss assessments are generally made on the basis of dry-matter changes. Normally, no attention is paid to nutritional or financial losses.

Grain loss. The loss in weight, occurring over a specified period and expressed on a moisture-free basis, of grain which would otherwise have been available as human food.

Moisture content (mc). The quantity of free water in a specified material. Materials of organic origin are defined for scientific purposes as consisting of dry matter and water. Loss of moisture during drying is not a food loss. Moisture content is expressed either as a decimal ratio or as a percentage in one of two ways.

(a) Wet basis (wb). The moisture content is defined as the ratio of the weight of water to the total weight of dry matter and water. This is the most commonly used method in agriculture.

(b) Dry basis (db). The moisture content is defined as a ratio of the weight of water to the dry-matter weight. This method is normally used in scientific laboratory work.

In agriculture it is traditional to use wet-basis moisture contents. Where moisture contents are expressed without an indication of wb or db, it may be assumed that the moisture contents are on a wet basis.

Decimal numerical values may be transformed from one basis to the other by the following relationships:

It should be noted that the values given by these formulas differ for the same sample.


1.2 Storage

Farmers produce crop products. Some of these require some processing before becoming suitable as food for humans. Crop products become available during different short periods of the year, but people wish to consume the food steadily throughout the year. Some form of storage is therefore required.

The storage requirements of crops show wide variation. For durables, such as cereal grains, the requirements are comparatively simple; while for perishable crops, such as fruit or vegetables, the cost of providing long-term storage is very high. Such difficulties may be overcome either by lengthening the production season of the perishables, or by partially or completely processing them into a more concentrated form.

Figure 1.1 Traditional storage structures


1.3 Storage requirements

The crop product must be stored so that:

(a) the quality does not deteriorate during the storage period;

(b) the quantity in storage is not unintentionally reduced;

(c) it is secure against pests, diseases and physical loss; and

(d) it is accessible at the time and in the quantity required.

The main crop products which may require storage facilities are:

Some processing may be necessary for the perishable products. Special care and specialized structures may be necessary for semi-durables, such as yams and sweet potatoes, before they can be stored successfully. The cost of processing and the cost of storage are important considerations in planning storage strategy.

Durable crop products are relatively easy to store, compared with the other two categories.


1.4 Agents causing deterioration of stored produce

The main agents causing deterioration of stored produce are:

Fungi. The most important type of microorganism causing or supporting crop deterioration. Although belonging to the plant kingdom, fungi possess no chlorophyll and are therefore unable to manufacture their own food by photosynthesis. Thus, they live on other living bodies as parasites; or on inactively alive or dead bodies as saprophytes. Parasitic fungi may cause diseases in the host body, while saprophytic fungi degrade or destroy the body on which they feed. Saprophytic fungi are more important in relation to stored durable crops.

Bacteria. Not generally a problem with dry-stored durables. They may, however, invade already damaged portions of the crop product during storage, and multiply.

Insects. Many species of insect are found in stored crop products, but only a few cause damage and loss. Some may even be beneficial because they attack other insect pests. It is important to be able to identify accurately the main insect species in order to assess their effect on the stored product and to devise the necessary control measures.

Figure 1.2 Insect pests

Figure 1.3 Rodent damage

Rodents. Rodents prefer not to live in grain stores because there is no drinkingwater. Although they can subsist without freely available water, the climate in the store is too dry for them to multiply rapidly unless they can leave the store to find water and return easily. Rodents consume grains and damage sacks and building structures, but they contaminate much greater quantities of grains with urine and droppings than they consume. They are controlled by poisoning and by preventing their access to the stored commodities.

Birds. Like rodents, birds consume some grain but also contaminate a greater quantity with their droppings. Losses caused by birds are avoided by preventing their access to the stored commodities.

Metabolic activity. The crop product is living material, and its normal chemical reactions produce heat and chemical by-products. Heat is also generated by insects, mites and microorganisms which, if present in large numbers, may lead to a significant rise in the temperature of the stored product.


1.5 Controlling the agents causing deterioration

The agents causing deterioration (with the exception of a few anaerobic species) require moisture, oxygen and an equable temperature in order to multiply and thereby damage the product.

Such agents are controlled by keeping one or more of these factors at levels which prevent (or at least deter) their growth or by measures such as the application of insecticides, or fungicides (e.g. propionic acid).

Figure 1.4 Metabolic activity

Figure 1.5 Reduction of moisture in cereals

1.5.1 Reduction of moisture content. The rate of metabolic activity is significantly reduced in most cereals if the grain moisture content is reduced to 14 percent; below 8 percent, metabolic activity practically ceases. Drying is therefore a standard treatment for wet cereal crops before storage. Drying requires energy to evaporate the moisture, and air movement to remove the resultant water vapour. The energy may be derived from burning fossil fuel or wood, or from solar energy, as in sun-drying. It may also be derived from ambient air that is not fully saturated with vapour (as in the crib drying of ear maize). Air movement may arise through convection currents caused by very small temperature differences, by a general air movement such as wind or a breeze, or by artificial means such as a fan. Drying processes are well documented and results can be reliably predicted.

1.5.2 Reducing oxygen. Bulk grain may be stored in airtight containers to exclude oxygen. If the grain is wet (17-20 percent me), metabolic activity soon exhausts the initial oxygen supply and the grain will not deteriorate in feeding quality. The germ, however, is destroyed and anaerobic fermentation may lead to unacceptable taints. Such grain is only used for animal feeding. If the grain is dry (12-13 percent me), it may be stored for several years, with careful management. In controlled (modified)-atmosphere storage, nitrogen or carbon dioxide is often used to replace the original air when the container is first loaded.

1.5.3 Controlling temperature. Levels of insect activity and general metabolic activity rise with increasing temperatures up to 42C. The maintenance of low temperatures in the bulk grain mass by using modern refrigeration techniques has been used successfully to control deterioration and maintain viability of the stored grain. The method is used in specialized fields such as seed storage and storage of grain for brewing. Equipment and running costs are high.

1.5.4 Chemical control. The bulk grain is treated with an organic acid or with gaseous ammonia. This sterilizes the grain and kills the germ, and generally leaves an odour disagreeable to humans in the grain, which is then used for stock feed. Insecticide and fumigant treatments may also be considered as chemical control methods.

2. Measurement

2.1 Introduction

Observation by an experienced person can often identify a problem and suggest a possible solution. It is extremely difficult, however, for even an experienced observer to predict precisely the value of the proposed solution in terms of cost and potential benefits. Measurements, made correctly, allow the problem and possible solution to be quantified. They also enable other workers in the field to benefit from the results.


2.2 Units of measurement

Primary measurements are mass, length, temperature and time; all other units are derived from these.

Derived units are those used in everyday measurements and calculations. The more important of these included in this manual are:


Parameter Unit Symbol Other unite
Weight kilogram kg tonne (1 000 kg)
quintal (100 kg)
Time second s minute (min)
hour (fur)
Distance metre m kilometre (km)
Temperature degree Celsius C degree Fahrenheit (F)
Area square metre m2 hectare (ha)
Volume cubic metre m3 litre (l)
Density kilograms/cubic metre kg/m3 grams/millilitre (g/ml)
Force newton N  
Pressure pascal Pa newton/square metre (N/m2)


2.3 Repeatability and precision

The degree of precision required in measurement depends on the general magnitude of the measurement and its purpose. To refer to a 100-km car journey in terms of the nearest metre would be pointless; yet units of 1 mm would be too large to describe the size of a grain mite. Errors in measurement arise from the imprecision of instruments and from the repeatability with which they can indicate the same measurements.

When making general measurements or calculations, it is usually sufficient to use four significant figures and to round off the answer to three significant figures; this should be remembered when using electronic calculators which display ten digits!


2.4 Measurement of moisture content

The moisture content of stored crop products is probably the most important single factor involved in providing safe storage. The moisture content of grain is particularly critical. Safe moisture levels for different types of storage cannot be specified because of the large number of other factors involved , such as: grain type and variety; degree of contamination with foreign matter; degree of damage; quantity in store; and provision for aeration. The wetter the grain, however, the greater the risk of loss.

It is therefore important to know the moisture content of the grain when it enters the store and how this changes during storage. In working stores (as opposed to experimental stores) it is not normally necessary to know moisture content with great accuracy; this is difficult anyway because of the problems of sampling. The store manager, however, must have some idea of the moisture content-to within 1 percent is usually sufficient-during storage.

Methods of measuring moisture content fall into two categories: direct measurements and indirect measurements.

Direct measurements of moisture content. For direct measurements the sample is divided into subsamples and each subsample is treated in succession. The moisture content is determined either by weighing a subsample, then removing the water and reweighing the dried sample (the difference in weight being equal to the water initially present); or by collecting and weighing the water given off. The most common method is the first; the water is removed from the sample by heating it in an oven under controlled conditions. The moisture contents of the subsamples are then averaged to give the moisture content of the original sample.

Oven-drying method. The material is weighed, dried in an oven at a specified temperature for a specified time and reweighed. The loss in weight is assumed to be the water in the original material. To avoid mistakes, a standard form is used to record the measured weights.

Accuracy of weighing. Weighing should be accurate to 1 part in 1 000 and the moisture content of subsamples expressed to three significant figures. The size of the subsample required depends on the type of weighing device, the oven space and sample containers available. A typical sample size would be 50 g of grain in a 75-mm diameter x 20-mm deep tin, weighed to 0.01 g.

The drying oven. The drying oven must have fan-assisted ventilation, and its temperature must be adjustable and controllable to + 1C over the range 95135C.

Time and temperature of drying. Many combinations are used, depending on the environment and whether the grain is whole or ground. The four most commonly used combinations are as follows:

(a) 2 hours at 130C (ground grain)

(b) 16 hours at 105C (ground grain)

(c) 72 hours at 100C (whole grain)

(d) 16 hours at 130C (whole grain)

Methods (c) and (d) are generally used with whole grain because the process of grinding the grain sample may release moisture before the sample is weighed, thus leading to inaccuracies.

If the grain has more than 25-percent mcwb, two-stage drying is recommended: the first stage at 95C for the time required to reduce the moisture level to about 18-percent mcwb, and the second for 16 hours at 130"C.

Indirect measurements of moisture content. Indirect methods for determining moisture content measure a property of the grain that is itself related to moisture content. The two most common properties measured are the electrical resistance and the dielectric constant of a grain sample which has been loaded into a measuring cell according to a standard procedure. This ensures that uniform quantities of grain are used and that uniform pressure is exerted on the sample.

The instruments designed to use with these methods present the moisture content as a reading on a scale, or can be used with a conversion scale. On some instruments a correction is necessary to compensate for temperatures outside the calibration range.

There are many commercially marketed meters using these properties. It is essential that the manufacturer's instructions be followed precisely and the instruments recalibrated against oven-dried samples at least once every year.

Another property of grain used for estimating the moisture content is the equilibrium humidity of the hair surrounding the grains. The hair hygrometer responds to relative humidity changes and can be calibrated in terms of grain moisture content. This method is less accurate than the others, and several hours may be necessary for the instrument to reach equilibrium.

The salt test. This is a very simple method for testing the suitability of grain for storage. Dry common salt (non-iodized) is mixed with the grain sample in a glass jar and shaken. The equilibrium relative humidity of dry salt is 75 percent at ambient temperatures. The equilibrium moisture content of grain at 75-percent relative humidity is about 15 percent. So, if the salt in the grain sample adheres to the walls of the glass, it has absorbed moisture from the air which must therefore have been at a relative humidity greater than 75 percent. This means that the grain had a moisture content greater than 15-percent mcwb, and is unsuitable for storage in bulk. The method is not precise, but it costs little and is simple to carry out.

Figure 2.1 Salt test

Figure 2.2 Biting grain to determine storage suitability

Other methods of determining fitness of grain for storage. The experienced person, without access to moisture meters, is able to obtain some idea of the suitability of grain for storage by the sight, feel and hardness of the individual kernels. Although this method is unreliable for determining moisture content accurately, it can prove useful in helping to decide whether grain may be stored with relative safety.


2.5 Sampling for loss assessment

2.5.1 Background. It is difficult to learn of the conditions in a total situation by direct measurement. This is because of the sheer size of the task involved and because the very process of taking a measurement probably changes what is being studied and measured, so that it is no longer true of the real situation. It is a more common practice to carry out tests on samples. If the results of these sample tests, however, are to be applicable to the situation as a whole, the sample itself must be truly representative.

For example, if information is required about the moisture content of a consignment of 100 tonnes of maize (the total situation), of uniform quality and uniform moisture content (a condition not possible in practice), the moisture content of any 1-g sample from any portion of the 100 million g in the 100 tonnes would be the same. The only errors which might arise would be from human factors or from defects in the measuring instruments. Sampling would present no problem; all samples, however selected, would contain maize of the same quality and the same moisture content.

In practice, however, any batch of 100 tonnes of maize would, initially, have variations of quality and of moisture content throughout its bulk, and even between individual grains. The variations themselves will change with time; insects and moulds will attack different parts of the maize. Moreover, heating, which produces "hot spots" and consequently more rapid deterioration in quality and changes in moisture content, will occur in local pockets.

Assessment of grain loss depends on being able to make accurate measurements on a representative sample. However accurately the characteristics of a sample in the laboratory may be determined, the results will be of little value if the sample was not representative of the original material. It is also true, of course, that however well a sample may represent the original, the final result will only be as valid as the accuracy of the instruments used and the competence of the operators. In practice, an acceptable degree of accuracy obtained at reasonable cost must be the goal.

2.5.2 Sampling. Taking samples that truly represent the original material is not easy. Samples may contain errors or bias. Thus, if the best-looking stack of grain in the field, the one nearest the home or the one that the farmer selects is chosen, bias will be present. The same will be true if samples are always taken near the granary entrance or where the grain looks good or bad. On the other hand, efforts to avoid bias may result in overcorrection. In avoiding bags within easy reach, those that are most difficult to reach may always be chosen. The solution is to remove the choice from the individual, and rely upon a table of random numbers. This is the basis of probability sampling, and the samples so obtained are termed probability samples. The advantages of a probability sampling plan are as follows:

(a) the degree of error due to sampling can be predicted;

(b) the number of samples required to obtain the desired accuracy in sampling can also be predicted; and

(c) the sample so obtained is certain to be representative of the original.

Figure 2.3 Box divider for sampling

2.5.3 Unit of observation. The observational unit is the container, location or process from which a sample is taken and used to determine the loss. It is the smallest division or unit in which grain is held. It may consist of stacks in a field, small silos or granaries on a farm, or woven baskets-but a single basket rather than all a farmer's storage baskets; or individual bags rather than the whole warehouse. The value of the entire survey will depend on the accuracy with which the loss is determined for each observation unit.

To make sampling easier, the observational unit should be as small as possible. This makes it feasible to mix all the grain thoroughly and produce a representative sample by coning and quartering or by using a sample divider. Such a procedure may apply where the grain is in baskets or in stacks in the field. In silos or granaries, however, it may not be possible; and unless the operation is done with skill, the sample may contain a systematic error which cannot be removed by later calculation or analysis.

When a container is sampled as a unit, the assumption is that the defect, contamination or other characteristic to be identified is uniformly (or at least randomly) distributed within the unit. In practice this is usually not the case. For example, insects or mites, mouldy kernels, rodent depredation and insectdamaged kernels are more usually found in pockets or layers within the bulk.

With limitations of time and money and often with traditional cultural considerations, the best course is to design a method of sampling so that the grain will be as representative as possible of the undamaged material and the defects throughout the layers or pockets.

In any study the investigator must report what was done and why, so that the significance of the data can be understood by those who will use them.

2.5.4 Taking samples. Most quotations of loss assessment are based on a weighin/weigh-out system. For example, in a batch process all observational units are weighed after initial samples have been taken, and all observational units are reweighed after completion of the batch process and before final samples are taken.

Figure 2.4 Sack sampling spears

Figure 2.5 Inadequacies of sack sampling spears (where black dots represent individual insects). A: large populations can be underestimated; B: small populations can be overemphasized

In many cases, however, observational units must be selected, using random numbers. The general procedures given in the previous section on loss assessment would then apply.


2.6 Farm- and village-level grain flow pattern

The flow of grain from the field to the eventual consumer is often complex; it may be compared with the flow of water through a system of pipes. Losses of water can occur anywhere in such a system. It is important to establish the magnitude of individual water leaks so that the largest can be reduced.

Similarly, grain flows from producer to consumer and is subject to leaks or losses on the way. It is important to obtain a relative perspective in order to visualize the importance of the total grain lost in flowing from producer to consumer, compare this with the quantity lost at a particular point, and measure this loss as a proportion of the grain passing that point.

In any programme of loss assessment it is necessary to obtain as much local information on grain flows as possible: how and when the grain moves from harvest to consumer; the routes it follows and its holding patterns; and where and how processing is carried out. Each district or community area has a marketing system for food grains. It is essential to establish the flow lines and the quantities they carry, so that priority points may be established for the observation and measurement of losses.

3. Storage pests

3.1 Post-harvest microbiology

3.1.1 Fungi. The simplest structure of a fungus consists of a thread (or hypha) which grows inside the host material. Several hyphae produce a mat known as a mycelium. The asexual reproductive structures known as sporangiophores grow out of this mycelium and extend beyond the surface of the substrate, or host material. At the end of these sporangiophores is the sac (or sporangium) containing the individual spores. Sexual reproductive structures are less frequently observed. The classes of fungi most important in crop storage are the moulds or microscopic fungi, whose optimum temperature for development is above 20C.

In order to multiply, fungi require water, oxygen and a suitable temperature. They also require food materials from the substrate; these are dissolved before absorption into the mycelium. Unless precautions are taken, stored crop products form an ideal substrate for fungus growth.

From an ecological point of view fungi may be divided into field and storage fungi.

Field fungi, such as Alternaria, Fusarium, Cladosporium and Helminthosporium, invade the seeds before harvest. These fungi develop only on seeds with a high moisture content (22-25 percent) and will die out under correct storage conditions.

Storage fungi, mainly Aspergillus and Penicillium species, develop on seeds with 12- to 18-percent moisture content.

3.1.2. Some important post-harvest fungi. There are many varieties of postharvest fungi. Some of the most important include the following:

Aspergillus flavus Grows on, and causes deterioration of, proteins, starches and oils; in particular, it reduces the quality of oil. Some strains produce the poisonous toxin aflatoxin, particularly in inadequately dried oil-seeds and cereals.

A. niger. Similar to A. flavus but the toxin produced is not so dangerous. The spore heads are black.

A. glaucus group. A very common group of moulds able to grow on substrates with very low moisture and high sugar content; usually the primary invaders of stored crop products.

Penicillium spp. Commonly associated with fruit rots. The mycelium is bluishgreen and may be aerial or embedded in the substrate; widely dispersed.

Botryodiplodia spp. Attack fruits or seeds in the field and deterioration continues in storage. The mycelium is black in B. theabromae; spores are produced in enclosed pycnidia (cavities) at the surface of the substrate.

Fusarium spp. A widely occurring species, as a fungus associated with storage rots, and as a pathogen causing blights and blasts of cereals and sugar cane. May survive in the seed and may continue growing during storage. Some species produce toxins on stored maize that has not been dried to a safe moisture content. Two species belonging to the Phycomycetes are also common on stored products.

Rhizopus otrhijus. Very widespread and reproduces sexually in characteristic sporangia on many crops, but is not a primary invader.

Mucor pusillus. A fungus associated with spoilage and decay. Strongly thermophilic; for example, it can survive the high temperatures of fermenting cocoa.

3.1.3 Controlling fungal growth on stored products. The primary cause of fungal deterioration of stored products is the presence of excessive water. This leads to relative humidities exceeding 70 percent in the air within the bulk of the stored product. Many fungal mycelia will develop at this relative humidity, leading to an increase in biological activity and increased deterioration.

The 70-percent level of relative humidity (rh) (see Section 6.3) has come to be regarded as a "safe" level; the moisture content of commodities in equilibrium with this rh indicates the upper limit of moisture content for safe storage (see Table 1).

TABLE 1. Values of moisture content equilibrium at 70 percent relative humidity and 27C

Commodity mcwb Commodity mcwb
Maize 13.5 Groundnuts (shelled) 7.0
Wheat 13.5 Cottonseed 10.0
Sorghum 16.0 Cocoa beans 7.0
Millet 16.0 Copra 5.8
Paddy 14.0 Palm kernels 5.7
Cowpeas 13.5 Gari (yellow) 13.6
Beans 15.0 Gari (white) 12.7

1 Determined horn prolonged exposure to controlled atmosphere-conoitions which do not always apply to stored products.

Contents - Next