1. JRC EUROPEAN COMMISSSION Components of APHLIS and how postharvest losses are calculated Composants des APHLIS et des pertes post- récolte comment sont calculées

2. What we will cover  What is APHLIS  How the PHL calculator works  The kinds of figures APHLIS produces  How we assess the quality of the loss estimates  Introduce the downloadable calculator

3. What is APHLIS? APHLIS is a unique service. It provides estimates of postharvest losses of cereal grains in sub-Saharan Africa.  It is based on a network of local experts who submit data and verify loss estimates  It gives loss estimates by cereal, by country and by province  Loss estimates are updated annually  The method and the data used to derive losses are displayed so the system is fully transparent, and  Better loss data can easily be added to the system so loss estimation can improve over time

4. East and Southern Africa June 2008 West and Central Africa April 2012 Components of APHLIS to supply data and verify PHL estimates Network of local experts

5. PHL database  Stores key data Production/yield Rainfall Climatic extremes etc  Is accessible by network  Annual up-dates

6. PHL calculator PHL database  Estimates cumulative weight loss from production  Uses figures for loss from literature and from network  Network verifies loss estimates

7. PHL calculator http://www.phlosses.net The web site Displays PHL estimates and key data PHLs by crop country and provinces Key agric. data GIS maps of PHLs and other data PHL tables Data tables

8. Downloads Users’ Guide Calculator spreadsheet  Allows users to enter own figures PHLs by crop country and provinces Key agric. data GIS maps of PHLs and other data PHL tables Data tables

9. How the PHL calculator works The PHL calculator determines a cumulative weight loss from production using loss figures for each link in the postharvest chain. A set of losses figures for the links of the postharvest chain is called a PHL profile Harvesting/field drying6.4 Drying4.0 Shelling/threshing1.2 Winnowing- Transport to store2.3 Storage5.3 Transport to market1.0 Market storage4.0 Example of a PHL profile for maize grain Figures taken from the literature or contributed by network experts Exemple d'un profil de PHL pour le grain de maïs Comment la calculatrice de PHL fonctionne

10. PHL Calculator contd PHL profiles are specific for  Climate type (A – tropical, B - arid/desert, C – warm temperate)  Crop type (different cereals)  Scale of farming (subsistence/commercial) Climate typeACBBA Crop typeMaize SorghumMilletRice Scale of farmingSmallLargeSmall Harvesting/field drying6.42.04.93.54.3 Drying4.03.5--- Shelling/threshing1.22.34.02.52.6 Winnowing----2.5 Transport to store2.31.92.12.51.3 Storage5.32.12.21.11.2 Transport to market1.0 Market storage4.0 Five examples of PHL profiles

11. PHL Calculator contd The PHL profile values are modified according to – 1.Wet/damp weather at harvest 2.Length of storage period (0-3, 4-6, >6 months) 3.Larger grain borer infestation (for maize only) … and the PHL calculation takes into account – 4.The number of harvests annually (1, 2 or 3) 5.Amount of crop marketed or retained in farm storage NB PHL values are affected much more by the application of modifiers than by the initial selection of the PHL profile.

12. The website Two ways to get PHL estimates  Consult the tables and/or maps on the website for losses by region, country or province  Download the PHL Calculator spreadsheet to enter user-preferred values for losses at a user defined geographical scale Postharvest Losses Information System Losses estimates Losses maps (interactive) Literature Downloads PHL Network About us Contacts Links Production Yield Larger grain borer Average farm size Home

13. Loss tables Regional losses for all cereals and by cereal type Click Estimated Postharvest Losses (%) 2003 - 2009

14. Loss tables by cereal type and country Click Estimated Postharvest Losses (%) 2003 - 2009

15. Loss tables by cereal type and province Click on one of these figures to get details of the loss calculation Estimated Postharvest Losses (%) 2003 - 2009

16. Details of the loss calculation. 1. Production data by farm type and losses over seasons Calculation matrix documenting the PH loss calculation quality of data sources and references to sources Country: Malawi Province: Area under National Administration Climate: Humid subtropical (Cwa) Year: 2007 Crop: Maize Production Annual production and losses Grain remaining Lost grain tonne Seasonal production and losses % SeasonFarm typeProduction (t) Remaining (%) Losses (t)Production (%)Remaining (t) Losses (%)

17. Details of the loss calculation 2. Factors modifying the PHL profile Rain at harvest – increases loss at harvest time. Larger Grain Borer – LGB attack doubles farm storage losses. Marketed at harvest % - divides the harvest between what is stored on farm and what is sent to market. Storage duration - loss increases with longer storage periods. Marketed at harvest (%) Rain at harvest Storage duration (months) Larger grain borer no data yes no data 20 PHL (%) calculation PHL (%) Calculation: Season: 1 Farm Type: small

18. Details of the loss calculation 3. The PHL profile and loss increments Stages Harvesting/field drying Platform drying Threshing and shelling Winnowing Transport to farm Farm storage Transport to market Market storage PH profile (adjusted) Remaining grainLoss increment Total 78.921.1 93.66.4 89.83.74 88.71.11.2 88.70- 86.72.12.3 78.97.89 78.901 04 27.9 Season 2 – smallholder only no grain marketed, all remains on farm.

19. Datum not a measured estimate Data overall specific to maize Details of the loss calculation 4. Quality of the data in the PH profile and references to data sources 1 0 Datum not specific to maize 0 Data overall not measured 0 The reference to Boxall 1998 StagesLoss figureReference CerealClimateFarm typeMethod References and individual loss figures % for small farms Origin of figure 6.4 5.0 9.5 5.8 9.9 2.0 Harvesting/field drying

20. The PHLs are also displayed on maps The PHLs are also displayed on maps APHLlS

21. There are also maps of LGB by year APhLlS Locations where Larger Grain Borer (Prostephanus truncatus) was considered to be a significant pest in 2007

22. Conclusions In the initial stages, APHLIS may or may not produce loss figures that are different from those currently in use and if they are different there will be no solid evidence that they will be more accurate. However, the new system generates estimates for PHLs of cereal grains that are -  Transparent in the way they are calculated  Contributed (in part) and verified by local experts  Updated annually with the latest production figures  Based on the primary national unit (i.e. province)  Upgradeable as more (reliable) loss data become available