Exploring linkages between livelihood assets and smallholders food security in rural Mali, West Africa

Transcription

1 Exploring linkages between livelihood assets and smallholders food security in rural Mali, West Africa S. N Danikou, R.S. Vodouhe, M. Bellon, A. Sidibe, and H. Coulibaly Bioversity International, Benin Office & University of Abomey-Calavi, Benin ndanikou@gmail.com Quantitative Methods for Integrated Food and Nutrition Security Measurements Lessons to be learned! Brussels, Belgium, November 2017

2 Outline Background Research question Study area Material and methods Results and discussion Conclusion

3 Background Food insecurity, a global concern Globally, absolute hunger is expected to decline below 8% of total population by 2030, But some groups such as SSA will remain disproportionally behind HFS exacerbated by climate variability and conflicts FAO (2014, 2017), UNDP (2014)

4 Background Figure 1. Global food insecurity index, Mali

5 Background Also, agricultural homogenization affects diets + resilience of food systems Considerable disconnect between people s diets and local food sources Reduction of hunger requires integrated approach Household Food Security = f{livelihood assets, } Remans (2015)

6 Background Conceptual framework External economic and social resources and context Livelihood assets (inputs) Management strategies Food security Sufficient, adequate, acceptable, certain, and sustainable Availability Accessibility Utilization of food Stability Dietary intake Adapted from Wolfe and Frongillo (2001), Gross et al. (2000), and FAO (2009) Nutritional status (outcomes) Physical well-being (e.g. child growth and death) Figure 2. Conceptual framework for the study

7 Question Which components of the livelihood assets are key to smallholder households food security?

8 Study area Sikasso region, Koutiala, Mali Figure 3. Location of Mali in Africa

9 Study area Source: RGHP 4, Mali (2012) Sikasso region, Koutiala district, Mali Figure 4. Poverty incidence & food insecurity per region, Source: RGHP 4, Mali (2012)

10 Material and methods A baseline survey in three sites A gradient of remoteness Differential access to livelihood assets HH questionnaires + 6 FGDs Random sampling: 180 HHs enrolled Prior-informed consent before enrolment

11 Material and methods Table 1. Characteristics of research sites N Variables N Goutjina Farakoro Kani 2 Remoteness: Distance to Koutiala, the nearest main town (Km) Access to public health services (1=low, 2=medium, 3=high) Access to private health services (1=low, 2=medium, 3=high) Existence of primary school (1=yes ; 0=no) Existence secondary school (1=yes ; 0=no) Existence of market (1=yes ; 0=no) Road quality (3=Practicable in all seasons; 2=Practicable for few months in year; 1=Impracticable in all seasons) 9 Access to clean water (1=low, 2=medium, 3=high) Access to extension services (1=low, 2=medium, 3=high) Access to credit (1=low, 2=medium, 3=high) Access to adults education services (1=low, 2=medium, 3=high) Agricultural infrastructure development (1=low, 2=medium, 3=high) Freshwater and hydrography (ponds, streams, and rivers) (1=low, =medium, 3=high) 15 Agricultural land (1=small, 2=medium, 3=large) Community pasture land (1=small, 2=medium, 3=large) 2 3 3

12 Material and methods Table 2. Livelihood assets and elements measured in the community HHs Variables Elements measured Response variable Food security index 14 short term, food-based coping strategies Natural capital* Land and freshwater resources Wild plants and animals, including aquatic resources and insects Explanatory variables Financial capital Human capital Social capital Infrastructural or Physical capital Income from crops and livestock, feeds Employment Housing/habitat Market participation Household goods House and farm water holding facilities Other farm equipment Household size Education Age Family labour Number of spouses in house Ethnic background and role in community Networking Road quality Schools Health systems Market infrastructure and institutional support Participation in development programmes Clean water Agricultural infrastructures

13 Material and methods Table 3. Sheet used to capture household coping strategies and their frequencies Less severe options Moderately severe options Most severe options Maxwell (1996), Hoddinott (1999)

14 Material and methods FSI = σ n i=1 CSRi Freq i HFIA categories - based on the tertiles of the FSI Wealth categories - DHS-based method Simpson and Shannon s diversity indices at HH level Alpha and beta diversity within and between community

15 Material and methods Kruskall-Wallis, to test effect of individual HH assets on HFS Mixed-effects models fitted by maximum likelihood, to test differential use of livelihood assets and effect on HFS RandomForest & Conditional inference models (party, Ctree) to identify components of livelihood assets with most significant effects on HFS

16 Results and discussion Livelihood assets and HFS Table 4. main HH characteristics (N=180) Parameters (average per household) N Goutjina Farakoro Kani Agricultural land 8.35± ± ±9.78 Household size 14.70± ± ±18.21 Family labour 5.25± ± ±7.35 Wealth categories (DHS index - based) Poorest (frequency, % in brackets) 21(35.00) 19(31.67) 20(33.33) Middle (frequency, % in brackets) 28(46.67) 18(30.00) 14(23.33) Richest (frequency, % in brackets) 11(18.33) 23(38.33) 26(43.33) Total 60(100) 60(100) 60(100)

17 Results and discussion Livelihood assets and HFS wealth p = 0.48 Figure 6. Average household wealth index per HFIA category and per village

18 Results and discussion Livelihood assets and HFS - overall significance Table 5. Most influencing capital assets on HFS, based on recursive partitioning tree models Capital assets Most important indicators Partitions Average insecurity index Human capital Risk attitude of the head of household Natural capital Richness of wild food plants managed in the wet season (WFPWS) Financial capital Income from non-agricultural employment (INAE) Social capital Physical/ infrastructural capital Social group membership (SGM), number Responsibilities in the community (RC), number Existence of water dams in village (EWDV) Existence of open well in village (EOWV) Low risk taking High risk taking WFPWS 3 WFPWS >3 INAE = No INAE = Yes SGM> SGM 8 & RC 4 SGM 8 & RC > EWDV=Yes EWDV=No & EOWV=Yes EWDV=No & EOWV=No p-value p<0.001 p=0.02 p=0.007 p<0.001 p<0.001

19 Results and discussion Livelihood assets and HFS overall sig. tree models Figure 7. Result of the conditional inference with all capitals showing the most important factors for HFS

20 Results and discussion Livelihood assets description biodiversity Figure 8. Estimated total species richness, Simpson and Shannon s diversity indices of useful biological resources in Koutiala, based on incidence data

21 Conclusions Household food security is under a complex set of factors Most influential factors: Dev. Programmes, ABD, vital agric. infrastructure, access to extra-agric. Income The research provides a combination quantitative data analysis techniques

22 Implications Livestock services should be enhanced in the study sites Investment in the conservation & increased access to ABD (domestication), to increase availability Panel data needed to monitor the role of the identified livelihood assets on HFS

23 Acknowledgments The people of survey villages in Koutiala for their participation in the study and for sharing their information

24 Thank you, any question? Sognigbe N Danikou, PhD Bioversity International, Benin Office & University of Abomey-Calavi, Benin ndanikou@gmail.com