The participants to the Expert Consultation prepared a number of conclusions and recommendations, which were discussed and consolidated as follows:
a. Root crops are gaining importance as food and cash crops. In their cultivation and processing (particularly cassava) women are heavily involved.
b. Root crops have advantage in terms of food security due to their tolerance to drought (no critical growing period, contrary to cereal crops).
c. Cassava is the first root crop in terms of spread and quantity in Africa & S.E. Asia and yam is the second root crop in Africa.
d. National Food Balance Sheets are to include all foods - not just dominant foods.
For the sake of convenient use and reference, the recommendations, made by the Expert consultation, are grouped into three main categories: a) general/agronomical, b) statistical/technical, c) data analysis and utilization, and one final recommendation.
a) Recommendations on general and agronomical aspects:
1. Adopt Latin names in addition to local/common names when referring to crops.
2. All forms of the products available should be taken into consideration (including leaf and/or stem consumption as for cassava, sweetpotato, coco yam and taro).
3. Stakeholders (agronomists, economists, statisticians) should learn about each other’s concerns for better estimation of production.
4. Countries and donors need to invest more resources in root and tuber crops statistics; this is imperative to global Root Crop Initiative and in particular to the Global Cassava Development Strategy.
5. Emphasize coordination among all statisticians, agronomists and economists involved in agricultural statistical work.
6. For promoting and attracting support for agricultural statistics it is important to clearly define capital and recurrent costs.
b) Recommendations on statistical and technical aspects:
7. To reduce costs at acceptable precision, stratification based on cropping and production patterns should be used in statistical surveys.
8. Stratification should be based on administrative units at the least or agro-ecological units/zone at best.
9. Crop specific survey designs provide more accurate statistics than multi-purpose designs; they should be implemented where possible.
10. To estimate production, distinction should be made between harvesting patterns:
For single harvested crops (e.g.: yam, sweetpotato, Irish potato, taro) estimation of production can be obtained by multiplying area and yield (estimated through crop-cutting).
For continuous harvested crops (such as cassava in Central Africa), the production should be estimated directly by periodical measurements of harvested quantities.
11. Providing data as a range rather than a single/specific number (e.g. coefficient of variation (CV), Standard Deviation (SD), or margins).
12. Cross-price elasticity between competing crops can affect the production of root and tubers and this should be taken into account in forecasting where applicable.
13. It is advisable to use a large sample size for yam production estimation where heterogeneity of the cropping system is high, as is often the case.
14. Improve production data using farm records from the farmer where applicable.
15. Use local measuring units (e.g. sacks, basins, etc.) and conversion factors where standard weights are not applicable.
16. For large sampling populations the number of farms sampled is more important than the sampling fraction.
17. Knowledge of measure of dispersion is important for determining optimum sample size for surveys.
18. Consider alternative (but statistically valid) approaches when working with limited resources.
19. Conduct more studies on the use of GPS for area estimation.
c) Recommendations on data analysis and utilization:
21. Units used for quantifying food into the balance sheet should be standardized for easier comparison among country using such standards as calories, proteins and fat.
22. Studies should be undertaken to develop country/location specific coefficients for quantifying root crops in NFBS
23. Quantify post-harvest losses and adjust figures appropriately for food availability particularly for root crops.
24. Explanatory notes should accompany data for appropriate use and understanding of National Food Balance Sheets.
25. Determine extent of commodity cross-substitution in consumption acceptable to the population.
26. Simple econometric models may not always be adequate; rather, the use of multi-equation simulations should be encouraged for production forecasting and planning
Final recommendation: The Expert Consultation set up an Informal Committee to follow-up on the present initiative on Root and Tuber Crop Statistics, to be composed of seven people:
a. Philippe Vernier
b. M. O. Akoroda
c. Eric Tollens
d. N. B. Lutaladio
e. Kajonwan Itharattana
f. B. P. Chaura
g. Naman Keita - Secretary.
The meeting proposed the following Terms of Reference for continuous follow-up of the activities discussed in this Expert Consultation. The Informal Committee shall:
Monitor progress with respect to the plan of Action and the studies to be carried out;
Facilitate the process of developing proposals for specific activities regarding root crops production estimation;
Coordinate efforts in the development of activities on R&T statistics improvement on a global level to avoid duplication and promote economies of scale;
Advocacy for root crop statistics improvement in on-going and future initiative and programmes;
Share R&T crop statistical methodology across Regions by networking;
Review and propose and advise FAO on specific activities to be undertaken in order to provide information on new initiatives related to R&T statistics.
