The real adoption of CA in the Lake Alaotra area after 10 years of diffusion. eric penot
1. THE REAL ADOPTION OF
CONSERVATION AGRICULTURE
(CA) IN THE LAKE ALAOTRA
AREA AFTER 10 YEARS OF
DIFFUSION
MADAGASCAR
E. Penot, CIRAD, UMR innovation/URSIA-SCRID
Joana Fabre, IRC Montpellier
R Domas, BRL/Madagascar
2. Introduction and objective
Extension of Conservation Agriculture in Lake Alaotra started slowly in
1998
Large scale development project BV-Lac/AFD, from 2003 to 2013,
farming system and watershed approach.
The GSDM, regrouping Research and Development CA operators,
compile statistics on CA adoption from projects database.
These statistics, area and number of adoptants, however, combine
different types of activity and systems, which are not all CA activities.
A need of a method to correctly assess cropped areas with real CA
practices (FAO definition). Clearly identify what is CA and what is not…
Objective: to avoid to “mythicize” CA and provide reliable data
Implemented within the Pampa project (AFD funding).
3. Methodology:
distinction between real CA cropping
patterns and other cropping systems
Identification of what can be considered as a
real CA farmer’s plot.
On Upland hills = tanety
Low land in upland conditions but access to
water in dry season = baiboho
Lowland rice with Poor Water Management
or control PWMLR (RMME in french)
6. challenge of a change in paradigm
for farmers.
Definition of CA
The approach is based on partnership, farming
system analysis, and modelling for a Decision
Support Systems (DSS) project orientation.
The Madagascar case : the Lake Alaotra
case study.
7. The non CA plots to be removed from the
current statistics
On tanety and baiboho
pastures, fodder crops, improved fallow, re-forestation, re-greening with
covercrops, cover-crops as preliminary improved fallow (potential CA)
first year (Y0) or CA “introductive” year with tillage : the first real year
in CA is the year 2 (Y1)
On lowlands : poor water control rice systems PWMLR
Given the very poor quality data on poor water control rice systems
(the 2/3 of rice fields in the lake : 70 000 ha!), we decided that only 10%
of the areas currently monitored by the BV-Lac project were under CA
(after expert and plot database analysis)
8. Official CA data
2500
Number
2000 of farmers
1500 Area
Surface
Number of adopting
1000
500
0
2003-04 2004-05 2005-06 2006-07 2007-08 2008-09 2009-10
9. The cohort analysis
Assessment of the abandonment rate, per year, by plot age.
According to the cohort analysis of databases, a recount of
real CA has been done.
Removing all plots in the first year (Y0) and the non CA
plots,
The method of Cohort analysis reveals the abandon % and the
difficulty of a sustainable adoption at a medium term
Provide a better assessment of how cropped area under CA
was evolving year by year
10. Rate of abandon according to CA plot age
80%
70%
60%
50%
40%
30%
20%
10%
0%
Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7
Rate of abandonment according to the age of plot
11.
12. The main results:
The real CA area is assessed at 419 ha in 2010
(estimated at 450 ha in 2011)
The number of adopting farmers is between 600
and 1000
- Abandon rate decrease from 60 % in Y0 to 10 % in
Y5 BUT increase the year 6 : some plots are tilled.
Our hypothesis:
- a poor mulch leading to future weed problem
- Soil compaction : do the covercrop plays its role ?
13. One main reason for adoption: yield stability on
mid term
Average yields on CA systems with upland rice in 2009/2010 BRL
Evolution des rendements en fonction de l'ancienneté en SCV des parcelles
4500
4000
Yield
3500
Arachide Stylosanthes
3000
Arachide En culture pure
Rendements (kg / ha)
Maïs Dolique
2500 Maïs Mucuna
Maïs Niébé
Maïs Stylosanthes
2000 Maïs Vigna umbellata
Pois de terre Stylosanthes
Riz Stylosanthes
1500
Riz En culture pure
1000
500
0 Plot Age
0 1 2 3 4 5
Ancienneté en SCV des parcelles
14. CA systems evolution
From imported biomass based with local Bozaka (Aristida spp) in 2003
- to the current systems based on covercrops such as Dolichos, Stylosanthes,
Bracharia, Vetch and Crotalaria mainly associated with Rice, Maize,
cowpeas and cassava (see poster).
Causes of abandon
abandon rate is between 40 and 60% after Y0 (opportunistic farmers)
with a clear decreasing trend over time (30 to 10 % from Y1 to Y5),
Other sources of abandon after Y1;
land status for 34 % (share cropping or renting)
difficulties to master and develop technically the system occurs for only
22 %
lack of cash to invest in the CA system for 15 %
15. What is an “adopting farmer”
The adoption process of CA is slow (between
3 and 6 years) and complex.
Requires training, knowledge acquisition,
understanding, know-how and then practices
Adopting farmer =
more than one plot in Y1 (CA areaincreases in
the farm).
CA plots from Y1 to Y8
16. Conclusion
The first 3 years of CA implementation (learning,
transforming, appropriation and then know-how) are
key years in adoption and innovation requiring labor,
investment, skills and willingness to move, from a
paradigm (tillage), to a new method (no tillage
coupled with long term vision and strategy).
If intensification with mineral fertilizers generally
provides an immediate response, this is not the case
with CA practices: profits are not always
immediately seen or even recorded
17. Due to crop rotation, the impact on production
stability may only be seen after a minimum of 3
years.
Yields do not increase significantly trough CA
practices (3 % per year) but intensification do.
Y0 and Y1 can be therefore considered as an
“experimental phase of CA”,
During the following years Y2, Y3 and Y4 and
beyond, abandonment decreased drastically.
18. The difficulty of identifying “opportunistic farmers”
withdrawing from the end of the first year is
jeopardizing development efforts on CA.
CA = a paradigm change in terms of practices
AND
A move from short term “mining” agriculture
to mid term CA strategies
Separate as well impact of CA from impact of
intensification….
19. Thanks for your attention
Rice and vetch Upland Rice
in baiboho after Stylosanthes
In tanety