1. Drivers of on farm tree diversity contribute
to climate change mitigation, adaptation and
resilience
By: Musana S. Bernard, Ndayamabaje J.D., Ndoli A., Mukuralinda A.,
Safari D., Mbonigaba J.J.
2. Outline
• Problem statement
• Land cover changes facts
• Forest degradation risk factors
• How on farm tree diversity impacts climate resilience
• Methodology
• Species present in the study area
• Results of the Regression analysis
• Discussion
• Relations between feeding systems, feeds and feeds costs
• Conclusion and recommendation
3. Problem statement
• Considering the last 100 years landscapes history, an impressive
biodiversity erosion in tree stratum has dramatically change the availability
of high value wood, wild fruits, medicinal plants and cultural plants.
• Fortunately since 1986 a consciousness of endangered species has develop
a philosophy of Agroforestry that can be considered at some extend as a
transition toward recovering some of the important products and services
(Konig, 1992; Kalinganire, 1996; Oksanen, 1997; Dixon,2003)
• AF have moved slowly due to poor understand socio-economic drivers and
institutionals barriers to adoption (Mukuralinda et al.,2015; Bucagu et
al.,2013)
4. 4
Land cover changes facts
Careful interpretation of biomass based indicators (NDVI is important)
• Loss in NDVI (biomass) in forest area is degradation/deforestation
• Gain in NDVI (biomass) in savanna ecosystems could be again degradation due to bush
encroachment
5. Forest degradation risk factors
Meta-analysis of Geist & Lambin 2004: “Our results show that desertification is driven by
a limited suite of recurrent core variables”
6. How on farm tree diversity impacts climate
resilience
• Tree diversity as indicator of tree adoption
• Tree diversity and agro-ecology
• Diversity of trees and sustainability of wood production
• Diversity of trees and soil biota diversity (Barrios,2014)
• Diversity of trees and overall agri-system diversity (Barrios,2014)
• Biodiversity and agriculture
7. Methodology
• 145 farmers have been interview (10-30 farmers/sectors) in 2 Districts
• The relationship of key household characteristic with the management of
trees and Diversity of trees in the
• farm has been investigated from household level information. 42
regressions have been conducted using linear model (Gaussian model) and
quasi-binomial regression using the generalized linear model following the
algorithm of the glm package in R. The details of the regression and the
pseudo – R2 and the R syntax used are presented in appendix of the report.
To simplify the Interpretation results are presented in 3 group regressions:
• Regression Analysis of Tree Diversity
• Regression Analysis of Tree management
• Regression analysis for Tree management and Diversity
11. Species present in the study area
11%
9%
7%
5%
5%
5%
4%
4%
4%
2%
Mangifera Indica
Grevillea robusta
Persea americana
Citrus SP
Carica Papaya
Eucalyptus SP
Markhamia Lutea
Citrus Lemon
Citrus Sinensis
Calliandra
• 90% of the top tree
species in the east are
exotic species
• Markhamia lutea is the
first indigenous species
present in farmer’s land
12. Results of the Regression analysis
• The 3 regression were constructed from logistic regression that has a
dependent or independent variables numerical (continuous) or
categorical variables (discrete variables).
• The first type having quantitative dependent variable, it has been
presented in tabular format and discussed. The covariate type of
response and qualitative response are too complex to assist decision
making because it needs definitions of Dummy variables and complex
transformation.
13. R2(1) Intercept Variables
Coeff T value P value Coeff T value P value
SUBCOUNTY 0.08
6.01639 13.782 <2e-16
***
Nyamata -2.91639 -2.507 0.0133 *
Rweru -3.01639 -2.478 0.0144 *
security
5.1296 15.448 <2e-16
***
Yes 1.6620 2.134 0.0347 *
month_secured -Qtt 0.00
5.15901 7.477 6.98e-12
***
0.04161 0.451 0.655ns
tenure 0.04 ns ns
Training 0.17
3.5098 7.86 8.45e-13
***
Yes 2.9796 5.372 3.08e-07 ***
Distancet-Qtt na
6.22082 13.319 <2e-16
***
0.07178 1.221 0.228(ns)
Livestock 0.04
3.88 5.67 7.62e-08
***
Yes 1.8867 2.508 0.0133 *
Income source 0.47 5.00 10.701 <2.00E-
16***
(income,1,11)* 2 3.495 0.000484***
(income,1,2)* 4.667 7.737 1.57E-14***
(income,1,2,5)* 1 1.888 0.059225.
(income,1,3,6)* 4.667 7.737 1.57E-14***
(income,1,6)* -1.333 -2.21 0.02718*
(income,6,11)* -3 -5.243 1.74E-07***
Results of the Regression analysis (2)
15. - Most of farmers interviewed have
income that is below 300,000 RWF;
- The number of income does not
explain the diversity but diversity
above 10 tree species are seen where
farmers have more than one source of
income
• Income is not clearly related diversity
Discussion
16. Discussion(2)
• e.g. Nyamata Revenue and Diversity not correlated:
Despite high revenue Nyamata has a reduced diversity
in general
• There is trend of number of income has a diversity
17. • Location and diversity indicates that in Nyamata and Rweru sectors,
there is a reduction of almost 50% of tree diversity.
• It has also been observed that for area with 5month of food security
the diversity reaches 6 species and for 10 month it reaches 7 species.
• Land tenure was found not significantly correlated with tree diversity
• Training on Natural regeneration was significantly correlated to
diversity (3 species than the non-trained)
• Development of livestock may induce increase of diversity
18. • More than 50% of the variability of on farm tree diversity could be
explained by the details of income sources and average annual
income.
• Among farmers who rely on tree products : Farmers who have
addition income such as wages, salaries or casual labor tend to
reduce the tree diversity; while those who depend more on their
products from their farms (food products) owned the highest tree
diversity.
• The group with the highest diversity has at least 6 species on average.
Farmers with livestock have 48% more species diversity compared to
those who does not have livestock.
20. Conclusion and Recommendations:
• More detailed typology of
farmer is needed to achieved
both increase tree coverage and
tree diversity (location-
accessibility, income source,
income, livestock activities and
feed availability)
• More Training is need to natural
regenerate trees on farm
• Diversifying on-farm income
through integrated crop-
livestock systems would increase
the adoption of more tree
species.
• Specialization of tree value chain
is crucial to increase diversity on
farm particularly for farmer are
more interested on tree
products and services