Addressing soil fertility and food security issues with fertilizer trees in Malawi
-
Upload
world-agroforestry-centre-icraf -
Category
Environment
-
view
194 -
download
0
Transcript of Addressing soil fertility and food security issues with fertilizer trees in Malawi
Addressing soil fertility and food security issues with fertilizer trees in
Malawi
Richard Coe, Joyce Njoloma, Fergus Sinclair, Bruce Sosola, Isaac Nyoka
14 April, 2015 BICC
Beating Famine Southern Africa Conference
Introduction
• Agroforestry technologies provide
the services and production functions oftrees
• Contribute to solutions on the declining soilphysical, chemical, biological characteristicsand qualities.
impacts on soil health
• affect the soil productivity
• subsequent food security
Introduction
o Agroforestry for Food Security Project (AFSP) started in 2007
o Aim of bringing benefits of agroforestry to large numbers of farms in Malawi,
o Through development of input channels, policy and capacity.
o Followed conventional model
o a theoretical basis
o agricultural experimentation by researchers,
o adaptive research ..similar results from elsewhere and small pilot projects with farmers
Methodology
• Study period 2011/12 and 2012/13 cropping seasons,
• Questionnaire administered to the household heads
• A 6m x 6m quadrat was marked in the agroforestry and non-agroforestry plots.
• Maize yield was measured in the quadrats.
Promoted trees soil fertility technologies
Interplanting maize with
Faidherbia albida
Gliricidia sepium
Annual undersowing with
Tephrosia vogelii
Cajanus cajan (pigeon pea).
Rotational with
Sesbania sesban
Data analysis• Data analysis was based on the difference in maize
yield of the two quadrats
• Yield difference = yield on agroforestry system plot –yield on sole maize plot in the same season.
• Statistical analysis used descriptive statistics and regression analysis of yield difference on plot and farm-level explanatory variables.
Data analysis
• When the two quadrats (with and without
agroforestry) measured on the same farm had
different levels of an explanatory variable, then
linear mixed models that included a random
effect for farms was done.
• All statistical analyses were done using R (R
Core Team 2014).
Results and Discussion
Cropping system Mean maize yield (t ha-1
)
2011/12 season 2012/13 season
Gliricidia sepium 3.62 3.62
Faidherbia albida 3.06 3.13
Cajanus cajan 3.12 2.97
Tephrosia vogelii 3.09 3.35
Sesbania sesban 3.10 ----
Control fields (no
trees)
2.73 2.78
Table 1. Overall mean yields of maize grain in
agroforestry and matching sole crop plots.
Figure 1: Distribution of effects ofintercropping gliricidia on maize. Our data(‘farms’) compared with other publishedresults. See text for details
Means effects quoted in these three papers by (Akinnifesi et al. 2006; Akinnifesi et al. 2007; Akinnifesi et al. 2010). These are from experiments on maize and gliricidia in Malawi.
A survey similar to this one (Akinnifesi et al. 2009).
Results of a meta-analysis compiling data from all over Africa and other tree species managed in a similar way to the gliricidia (Sileshi et al. 2008).
Fertilizer on control
yes no
Fertilizer on
gliricidia
yes 0.64 (51) 2.46 (3)
no 0.04 (21) 1.31 (9)
Table 2. Effects of gliricidia. Mean maize
yield difference (gliricidia - sole) (tha-1) for
plots with and without fertilizer. Number of
pairs in parentheses.
Difference, d, between maize
with gliricidia and sole maize
CaseElevatio
n (m)
Trees per
plot
Fertilizer
applied mean
Probability
d>2
Probability
d<0
1 500 10 Both 1.3 0.32 0.22
2 1000 10 Both 0.3 0.15 0.42
3 1500 10 Both -0.6 0.05 0.65
4 500 30 Both 1.9 0.47 0.12
5 1000 30 Both 1.0 0.26 0.28
6 1500 30 Both 0.0 0.11 0.50
7 500 10 Sole only 0.6 0.19 0.36
8 1000 10 Sole only -0.3 0.07 0.58
9 1500 10 Sole only -1.3 0.02 0.78
10 500 30 Sole only 1.2 0.32 0.22
11 1000 30 Sole only 0.3 0.15 0.43
12 1500 30 Sole only -0.6 0.05 0.65
Best 500 40 Neither 2.4 0.60 0.07
All contexts sampled 0.6 0.18 0.33
Table 3. Predicted mean increments of maize yield (tha-1)
with inclusion of gliricidia for difference cases, along
with the chance of increases >2 tha-1
and <0.
Figure 2: High within field spatial variability
Homestead: high fertility
Out field: low fertility
Out field: medium fertility
Variation sources
Conclusion
• The crop yield performance of theagroforestry options in Malawi studied hereare highly variable.
• The variable performance of new croppingoptions is probably the norm rather thansomething unusual.
• The variability can be divided into two types:
– That explained by contextual factors
– Remaining unpredictable variation.
Conclusion
Acknowledgement
Field team Led by ICRAF Southern Africa Programme Malawi
District offices of Malawi’s Ministry of Agriculture and FoodSecurity in Kasungu, Neno, Ntchisi, and Salima districts,
Landscape management for environmental services,biodiversity conservation and livelihoods Science Domain ofICRAF
Funding support for the study was received the Governmentof Ireland and Science domain Theme on Land Management