4. Special Panel Discussions

Two parallel panel groups were formed as follows:

Panel Group on Roots/Cassava

Facilitator: Prof. E. Tollens
Reporter: Mr. B. P. Chaura
Members: N. Keita; L. M. Machirovi; E. Mpyisi; F. Ngopya; N. Ngonde;
N.B. Lutaladio; Chalit Amnuay; Kajonwan Itharattana; V. Sandifolo

Panel Group on Tubers (yam, sweet potatoes, cocoyam)

Facilitator: Dr. S. Doumbia
Reporter: Ph. Vernier
Members: M. O. Akoroda; M. Nyoni; J. Q. Rijks; H. Som

4.1 The Group on Roots discussed and proposed the following

4.1.1 Concepts, Definitions, Standards

Depending on the main use of the production, the following classification and related concepts and methods of data collection is proposed:

Main purpose of production	Number of harvests	Mixture	Concept of production	Region	Method
Total or mainly for market	One	Pure	Production harvested	Asia; emerging Africa Latin America	Interview + crop cuts at harvest time
Mainly food reserve; mixed (commercial and subsistence)	Several	Mainly mixture with cassava as dominant crop	Production harvested + potential	Africa + part of Latin America	Interview (recording harvest in local units) + a few crop cuts (for potential production)
Mainly subsistence, remote areas or land locked	Several + partial	Mixture	Production harvested + potential	Africa + part of Latin America	Interview (recording harvest in local units) + a few crop cuts (for potential production)

Note: Measure area for crop cutting fields (P = A * Y)

Components of Production:

In the perspective of food security it is important to estimate production of leafs as well as of the roots.
Production is to be reported in weight of clean fresh product.
Measurements and data collection can be done in several ways e.g. fresh, dry, etc.

Conversion factors:

This should be done in fresh equivalent and should be area specific.
To be determined through a specific survey with indication of mean and coefficient of variance (CV).
It is required that all countries produce conversion factors for all products from the different zones of the country (product to clean fresh roots).

4.1.2 Data collection / survey design

a) Stratification variables

1. Administrative Units

2. Agro-ecological Zones

3. Market access/ population density, remoteness

4. Urban/rural

5. Farming systems (sole/mixed - area with cassava dominant/minor crop, see FAO classification 2001)

6. Use of GIS Maps.

All the above factors depend on country situation.

b) Crop cutting

Optimum number and size to be studied (40 plants ´ 4 replications or 5 by 5 meters as an indication).

c) Interviews

- Monthly visits to farmers to be validated by additional trials (optimal effective cost and frequency of visits

- Calibration of local units (weight/volume to Kg, and area to ha).

4.1.3 Post -Harvest Crop Losses

- What is loss?

Loss is weight loss due to diseases, pests and rodents, transport/ handling

Quality loss due to time lag before getting food to consumption, pests may lead to reduced weight and price.

- Not enough information
- Tend to be overestimated
- Vary seasonally
- Competing foods or type of product
- Need for further studies.

4.1.4 Follow-up action

a) Advocacy for improvement of RTC/Cassava statistics given the importance in food security for countries in terms of global cassava strategy. Better Cassava statistics is an imperative.

b) We recommend that, where relevant, to conduct crop specific surveys be conducted for root and tuber crops.

c) Need to elaborate specific methodologies for Cassava statistics to reduce under- and/or over- estimation.

d) Need for further studies:

- on crop losses throughout the food/market chain
- optimum size for crop cutting plots
- use of GPS for area measurement
- use of crop growth physiological models
- econometric models for forecasting and policy planning
- optimal/ cost-effective frequency of visits to farmers/recall in interviews
- conversion factors
- leaf statistics (weight, frequency of harvest, price of produce, effect of leaf picking on root yield).

4.1.5 Action Plan

a) Suggested pilot countries for studies and/or programmes:

- DRC (Multi-donor projects)
- Nigeria (IFAD project)
- Cameroon (IFAD project)
- Malawi (USAID-project)
- Mozambique (USAID, ROCKEFELLER)
- Thailand (JICA)
- Brazil (CLAYUCA, EMBRAPA).

b) Global Research Programme on Cassava Statistics (3 years)

It is recommended and encouraged that there be a Global Research Programme on Cassava Statistics involving FAO, National and International Research Institutes and Networks. These organizations need to work together and with national governments to improve statistics.

FAO should coordinate this global effort given its mandate and comparative advantage. This consortium should seek funding for this programme to carry out this objective.

4.2 The Group on Tubers discussed and proposed the following

4.2.1 Concepts and Definitions

a) Plot (parcelle)

Single dominant variety
Not necessarily homogeneous

b) Field (champ)

Group of plots (1 or more).

Fields can be located in different parts of farm.4.2.2 What should be measured?

The variables that should be measured can de differentiated by objective:

For Research - Yield is to be measured,
For Production Statistics - Production is required.

4.2.3 Which Production to Measure?

The various aspects of production can be differentiated as follows:

Actually harvested (in the form of clean fresh tubers)
In case of double harvest (the 2^d is for seed, 3 to 4 months after first harvest).

For total production:	- Plot by plot
	- Aggregate
For sample production:	- Crop-cut

For stock estimation (reserve portion of harvested crop): Stock can be in a fresh form or processed form.

4.2.4 Stratification for Collecting Data

a) By Administrative /Agro-ecological zones:

- Administrative unit
- Soil type
- Rainfall (total and pattern).

b) By Production System:

- Intensification level
- Production purpose: Consumption, Commercial, Mixed objectives.

4.2.5 Area Estimation

Sample of farms
Measure/estimate the whole farm if small enough or sample within a farm (parcelle/plot)
Estimate area from mounds/mound density.

4.2.6 Yield Estimation

Sampling from crop cuts within the strata
Number of crop cuts to determine according to variability within the fields
Optimal size of crop cut area (number of mounds recommended is 40 plants x 4 replications).

4.2.7 Production estimation

Multiply estimated area by estimated yield.

In the case of mixed cropping: P= yield per plant x plant density observed.

4.2.8 Post Harvest Crop Losses

Post harvest crop losses are affected by, or varies by:

Food chain level (producers/traders)
Producers/Farmers (sprouting, evaporation, rot, rodents, etc.)
Traders
Variety
Storage conditions
Processing forms (e.g. insect attacks on chips...)
Agro-ecological zones or regions, etc.

Further studies are required to assess actual loss rates.

4.2.9 Where are we?

Current state of knowledge
Where to find information?
Agricultural projects, NGOs, extension, research
What can we find?
Production, climate, yield, monographs
What can we learn?
Past methodologies, use baseline surveys to reduce costs.

4.2.10 Action Plan

Situation in a few countries (consultation)
State of knowledge
What still needs to be done
Regional project (3 years)
Validate available results relating to estimation procedures in other countries.
Address remaining problems.