The APHLIS website displays postharvest loss estimates (PHLs) for the cereal crops of Sub-Saharan Africa, for individual countries and for their provinces. The PHL calculator that makes these estimates can also be downloaded as an Excel spreadsheet where the default values can be changed to users’ preferences. To learn about all the facilities of APHLIS and how losses are calculated click here to see the guide to Understanding APHLIS.

Estimates of postharvest losses are important data for policy makers, for food security staff making cereal supply estimates and for agricultural practitioners proposing, or actually managing, interventions to reduce postharvest losses. Before the introduction of APHLIS, the origin and justification of PHLS figures were not well founded. APHLIS was established to generate figures for the PHLs of cereal crops in a fully transparent manner and in a way that can enable the updating of PHL estimates as new data become available. The postharvest losses are presented as maps and as tables, to see the tables click on ‘Losses tables’. There are three different types of tables which show -

  • % cumulative weight loss from production,
  • the actual weight of grain lost from production, and
  • the loss density, which is the weight loss per unit area (tonne/km2) of the province concerned. This enables users to assess the relative importance of losses between different provinces.
The % cumulative losses from production and the loss density can also be displayed as maps such as the one shown below, which may easily be inserted into reports. For a quick view of the latest maps click ‘Quick overview of Losses’ for more detailed interactive mapping click ‘Interactive maps’.

There are five basic components of the system.

  • A database that hold keys agricultural data by country and by province for most of the countries of Sub-Saharan Africa.
  • A network of local experts – the APHLIS Network – that supplies agricultural data and verifies PHL estimates – see this using ‘APHLIS Network’ button in the left hand column.
  • A PHL calculator that determines a cumulative % weight loss (PHL) using the information in the database and a PHL profile (the typical PHLs incurred at each step in the postharvest chain) for the specific conditions in question. The estimated losses are displayed by crop type in tables for the region, country and province. These are accessed from the ‘Losses Tables’ button on the left hand column.
  • A mapping tool that displays the losses for cereal type by province as well as other data such as the climate types (by Köppen code) for the provinces in the study area and the incidence of larger grain borer.
  • A downloadable version of the PHL calculator in which the user can define most of the default values.

What are the PHLs

Postharvest losses occur during harvesting and handling due to grain shattering, due to spillage during transport and also result from biodeterioration at all steps in the postharvest chain including storage. The principle agents of biodeterioration are moulds, insects, rodents and birds. A variety of insects pests, like the larger grain borer shown here on maize grain, are one cause of postharvest losses. They start their attack in the mature crop in the field and carry over into storage.

Larger grain borer (Prostephanus truncatus) a serious pest of stored maize

Each step in the postharvest chain may have its own associated weight losses, an example is shown below with a typical weight loss range for each step, but it is important to remember that losses vary considerably according to the cereal crop in question, the agro-climatic conditions, the scale of farming, from season to season and from year to year. APHLIS takes into account some of this variation depending on the amount of data submitted by the APHLIS network.

Typical ranges of weight losses for various links in the postharvest chain

The APHLIS Network

APHLIS network of local experts is responsible for submitting some of the data required for making the loss calculations. The experts have their own access to the APHLIS database and enter production figures and data on the seasonal factors that affect losses. Where possible they are also encouraged to submit data on losses.

Understanding loss figures

On APHLIS, the estimated PHLs are the reduction in weight of ready to consume grain incurred during harvesting operations, farm storage, transport and market storage and specifically -

  • Reflect losses of quantity (weight loss); quality change is only relevant if food is no longer fit for human consumption.
  • Include grain spilt during handling or given to farmers’ animals to consume as such grain is no longer available for human consumption, even if the animals are a component of the farmers’ livelihoods.
  • Are from this year’s production, cereals from the previous season(s) are carryover stocks and dealt with separately.
  • Are reported for national and sub-national units (provinces) which follow political rather than agro-climatic boundaries
  • Do not take into account any cereal processing losses.

The APHLIS loss figures are ‘best estimates’ of the losses that occur across provinces. The actual losses suffered by individual farmers are likely to vary considerably from one locality to another, both between seasons and between years. As provinces normally cover large geographical areas, provincial loss estimates are generalisations. If estimates are required for specific location then users should work with the downloadable calculator where the APHLIS defaults can be changed to recently observed loss values for the location concerned.

APHLIS loss estimates only take into account losses of weight but quality losses may also be significant and often result in farmers selling their grain for less. Quality losses are difficult to take into account in a system like APHLIS but practitioners working on projects to reduce postharvest losses should endeavour to measure them. More information on this is given in the APHLIS Qualitative Losses report.

APHLIS also offers advice on how to collect new weight loss data and data on the seasonal factors that affect losses. In the ‘Collecting new loss data’ you will find a manual offering stepwise instructions to postharvest scientists who wish to adopt a rapid, systematic approach to generating new loss estimates and an interview form that can be used in face-to-face or telephone interviews with agricultural specialists who can provide seasonal data for the areas of a country where they have specialist knowledge.

For more information about the postharvest situation and losses in certain countries, consult the ‘Country postharvest narratives’. These vary in the amount of information available and initially detailed accounts are available for just a few countries. They put the APHLIS loss estimates into a national context and where possible consider the validity of the APHLIS estimates.

Warnings about the PHL estimates

When users of APHLIS request estimates for the PHL of a particular crop, in a particular province, then a pre-determined set of appropriate PHL values for each step in the PH chain is inserted into the PHL calculator. This set of figures is called a ‘PHL profile’. One problem faced in seeking to provide PH profiles is that for many provinces there are no specific PHL data. It is therefore inevitable that many different provinces will have to share the same data. This was achieved by clustering together the provinces of many countries that are basically similar with respect to the factors that influence PHLs; the most convenient method of doing this is based on climate classification. Climates have been classified into one of three types, tropical savannah/forest, arid/desert or warm temperate. For each crop there is a PHL profile for each climate, so with seven crops there is a total of 21 profiles. In assembling PH profiles it is necessary to create a generalised loss figures for each step in the PH chain. These generalised figures were obtained from all estimates available by 1) removing outliers, 2) avoiding the use of ‘of questionnaire/guesstimate data where there is sufficient measured loss estimates, and 3) averaging what data remained. However, in many cases the PHL profiles are very generalised, i.e. are not composed of loss figures from the same cereal or from the same climatic area in question. This is indicated in the downloadable PHL calculator which records each individual figure in a PHL profile as being ‘same’ or ‘other’ and estimates as being ‘measured’ or ‘questionnaire/guesstimates’. Generally speaking, loss estimates from the model can be regarded as more reliable when they are from the ‘same’ situations and are ‘measured’. As the PHL profiles used by the PHL calculator are intended to be specific to the crops and region for which they were developed, the calculator should not be used for the estimation of losses for other crops, farming systems or regions. PHL estimates are based on the best data available, but this is not necessarily very accurate. If more up-to-date data are to hand then these could be used by downloading the spreadsheet version of the PHL calculator and altering the default values.

Feedback on PHL estimates and contribution of new PHL data

Contact us to give feedback on using the system. We are keen to receive well documented loss figures to add to the database, to learn about successes or failures in using the system and suggestions for improvements.


The European Commission and its agents accept no responsibility for the accuracy or otherwise of data or estimates obtained from APHLIS and these are provided entirely at the users’ risk.