Transcript Future rice production in Madagascar

Future rice production
in Madagascar
Where and how much ?
Robert Hijmans & Alice Laborte
International Rice Research Institute
Background
• Madagascar’s population is largely rural, very poor &
dependent on agriculture
• Rice is the main crop and staple food & cassava is also
important.
• Madagascar is a net rice importer. Government aims to
double rice production between 2005 and 2012
• Production (kg) = Area (ha) * Yield (kg / ha)
• Area expansion (habitat loss?) or intensification (pollution?)
2 modeling approaches
• Regression: Where would area expansion
take place?
• Simulation: Climate change effect on yield
Where is rice grown ?
BD50
IDFN1
Regression model
Rice area as function of:
a)
b)
c)
d)
biophysical factors
socioeconomic factors
biophysical + socioeconomic factors
bio. + soc. factors (2 regions: W, CE)
Biophysical variables
Socio-economic variables
Model
Variable
Unit
Intercept
1a
17.039**
b2
0.364**
c
3
12.469**
Elevation
km
-0.679**
-0.490**
Slope
deg
-0.055**
-0.063**
Slope – focal mean
deg
-0.147**
-0.114**
Sqr(Slope)
deg2
0.005**
0.005**
Rainfall
m
0.757**
0.677**
Mean temperature
°C
-1.255**
-0.851**
Sqr(Temperature)
°C2
0.022**
0.013**
Distance to roads
km
-0.121**
-0.114**
Distance to towns
km
-0.094**
-0.091**
Population density
person km-2
0.009**
0.002**
Population density – change
%
Income
million francs
-0.002**
1.458**
2.165**
D2
0.09
0.11
0.16
Area under ROC
0.70
0.73
0.76
26,930
26,382
24,928
AIC
**Significant at 0%, * 5%
Model d
Variable
Unit
Intercept
Elevation
km
Sqr(Elevation)
West
AS, MA,
TL
Center-East
AV, FI, TM
8.722**
-14.417**
-3.304**
-4.306**
2.007**
1.295**
Slope
deg
-0.071**
Slope – focal mean
deg
-0.125**
-0.150**
Sqr(Slope)
deg2
0.003**
0.005**
Flow accumulation – focal mean
-0.012**
Rainfall
m
1.817**
Mean temperature
°C
-0.328**
Sqr(Temperature)
°C2
Distance to roads
km
-0.150**
-0.125**
Distance to towns
km
-0.097**
-0.107**
Population density
person km-2
0.009**
0.002**
Population density – change
%
-0.035**
0.011**
Income
million francs
2.393**
-0.075**
5.139**
D2
0.18
0.20
Area under ROC
0.78
0.79
**Significant at 0%, *1%
Biophysical variables
Socio-economic variables
Biophysical &
socioeconmic variables
Two regions
Biophysical variables
Socio-economic variables
Biophysical &
socioeconmic variables
Two regions
Forest loss model
Business as usual:
Forest area reduced by
25% in 2032
50% by 2077.
Rice growth simulation model
Baseline
Climate: 1960-1990 (CO2 = 340 ppm)
Cultivar: IR72
Model: Oryza
Pot. yld.
(t / ha)
January
April
July
October
< 2.5
2.5 - 5
5 - 7.5 7.5 - 10 > 10
November planting of IR64, irrigated
2080
A2a, Hadcm3
Current conditions
< 2.5
2.5 - 5
5 - 7.5 7.5 - 10 > 10
IR64, rainfed
Implications?
Part of the climate change effects on yield potential can be easily
dealt with (length of growing season)
Current yields are very low; Need for intensification; opportunities
are not much affected.
Water supply, double cropping?
Resilience to extreme events (flooding, drought)
More rice in the central highlands? Or diversification to high value
crops?
• Intensification:
– less pressure on the land;
– more pollution, where would that be a threat?
• Extensification:
– Not necessarily in the forest areas
– Rehabilitation of irrigation schemes
– Need for alternative livelihoods