adoption and intensity of adoption of conservation farming practices in zambia
DESCRIPTION
www.fao.org/climatechange/epic Full paper: http://www.fao.org/docrep/017/aq288e/aq288e.pdf This presentation outlines an analysis of the determinants and the intensity of adoption of two components (i.e. the use of zero tillage and planting basins) of Conservation Farming in Zambia. We find a strong and robust relationship between the district level variation in historical rainfall during the growing season and adoption as well as the intensity of adoption of these practices in Zambia. This finding suggests that farmers are using these practices as a strategy to mitigate the risk of rainfall variability, providing evidence – albeit indirectly – of a synergy between these practices and adaptation to climate variability.TRANSCRIPT
Aslihan Arslan (Co-authors: Nancy McCarthy, Leslie Lipper, Solomon Asfaw and Andrea Cattaneo)
Global Science Conference on Climate Smart Agriculture
March 20-22, UC Davis
Adoption and Intensity of Adoption of Conservation Farming Practices in Zambia
• Background on CA and CF in Zambia • Data sources • Methodology • Adoption & Disadoption • Determinants of adoption and its
intensity • Conclusions
Outline
Background on CA
• Conservation agriculture (CA) aims to sustainably improve farm productivity, profits and food security by combining (FAO 2012): 1. Minimum mechanical soil disturbance 2. Permanent organic soil cover 3. Crop rotation
• Born out of ecological & economic hardships in the US in ’30s • Popular during the oil crisis in ’70s • Mainly large commercial farms in Brazil, South Africa &
Zimbabwe • Promoted by many in SSA as a solution to soil degradation & low
productivity
CA in Zambia
• Promoted to smallholders in ZM as Conservation Farming (CF): 1. Reduced tillage (<15 % of the area) 2. Precise permanent planting basins/ripping of soil with
a ‘Magoye ripper’ 3. Leaving of crop residues on the field 4. Rotation of cereals with legumes 5. Dry season land preparation (CFU, 2007)
• MoAL adopted as priority in 1999: ZNFU, GART, CFU • Int’l support: SIDA, Norad, FAO, World Bank, WFP,
EU, IFAD…
Mpika
Solwezi
Sesheke
Kaoma
Serenje
Kalabo
Chama
Mkushi
Mumbwa
Kasempa
Lukulu
Chinsali
Mwinilunga
Kalomo
Senanga
MufumbweZambezi
Lundazi
Kaputa
Kazungula
Isoka
Kabompo
Mansa
Mongu
Mbala
Nyimba
Itezhi-Tezhi
Shangombo
Samfya
Kasama
Chibombo
Chongwe
Mungwi
Kapiri Mposhi
Luwingu
Mporokoso
Petauke
Kafue
Choma
Lufwanyama
Mpongwe
Mpulungu
Chipata
Mwense
Kawambwa
Milenge
MonzeMazabuka
Mambwe
Chilubi
Namwala
Katete
Chavuma Masaiti
Chiengi
Nakonde
Gwembe
Luangwa
Siavonga
Sina
zong
we
Nchelenge
Chadiza
Kabwe
Livingstone
Lusaka Urban
Chililabombwe
MufuliraChingola
Kalulushi Kitwe
Luanshya
Ndola
200 0 200 400 Kilometers
N
EW
S
Agro-Ecological Regions
District boundary
KEY
Source: Soil Survey, Mt. Makulu ChilangaDecember 2002
Scale 1: 2,500,000
RegionsI
IIa
IIb
III
LEGEND
Classic Barriers to Adoption
• Risky new technology • Credit const. • Time lag • Labor const. • Seed market const. • Agro-ecological const. • Tenure security • Opportunity costs of residues
Barriers for CA
• Farm size in Africa & education in North America (Knowler & Bradshaw, ’07)
• Lack of infrastructure, existing livestock mgmt norms, imperfect input & credit, land tenure (Nkala et al.’11)
• Zambia: Opportunity cost of crop residue, land and labor constraints , distance to markets, extension (Umar et al.’11; Baudron et al.’07, Chomba ’04; Haggblade&Tembo,’03)
• BUT: Most studies are subject to small samples, selection bias or both
Data Sources
• Rural Incomes and Livelihoods Surveys 2004 &
2008 (MAFF & FSRP/IAPRI)
• Historical Rainfall Estimates (NOAA-CPC)
• Soil Nutrient Availability (Harmonized World
Soil Database)
Shifting Rainy Season Onset
9
Map here
Tillage & Crop Management
Practices 2004 2008 Hand hoeing 0.60 0.44*** Planting basins 0.03 0.02*** Zero tillage 0.11 0.03*** Ploughing 0.29 0.31* Ripping 0.02 0.01*** Ridging/bunding 0.23 0.41*** Crop residue left in the field 0.74 n.a. CF Practices Analyzed Min. Disturbance (P. basins/zero tillage) 0.14 0.05*** Rotation (diff crops for 3 years) 0.57 0.56
Adoption & Dis-adoption (0/1)
2004 No Yes Total 2004 No Yes TotalNo # 3,498 165 3,663 No # 1,071 755 1,826
% 95.5 4.5 100 % 58.7 41.3 100
Yes # 505 19 524 Yes # 822 1,539 2,361% 96.4 3.6 100 % 34.8 65.2 100
2008
National transition matrix, Minimum Soil Disturbance (MSD)
National transition matrix, Crop Rotation (CR)
2008
Adoption Intensity (area share)
Adoption intensity by land size
2004 2008 2004 2008<=1.5 0.41 0.64 0.47 0.571.5 - 2.5 0.28 0.48 0.43 0.522.5 - 5 0.24 0.37 0.40 0.465-20 0.18 0.17 0.35 0.35> 20 0.03 0.06 0.09 0.13
Land (ha)MSD Intensity CR Intensity
Empirical Approach
1. Decision to Adopt:
Conditional Maximum Likelihood (CMLE) Probit model (Chamberlain, ’80)
2. Intensity of Adoption: Correlated Random Effects Tobit (Wooldridge, ’02) & Pooled Fractional Probit (Papke&Wooldridge, ’08)
*it it it iC X u vβ= + +
*it it it iS X u vβ= + +
*
* *
*
0 0
0 1
1 1
it
it it it
it
if SS S if S
if S
≤
= < < ≥
Determinants of Adoption
Variables MSD CR# Adults (age>=15) 0.005 0.029*Education (average) 0.025 0.033**Ag-wealth index 0.093 0.086**# Oxen owned -0.058 0.050***ASP district dummy -0.048 0.094*Moderate soil constraint -0.123 0.109*Rain onset delay 0.860** 0.861***Received MSD/CR extension (% SEA) 1.533*** 0.724***RFE CV (1996-2011) 8.140*** -0.3492008 dummy -0.663*** -0.075*Number of obs. 8,208 8,208
Determinants of Adoption Intensity Variables MSD CREducation (average) 0.021 0.014**Dependency ratio 0.017 0.006Land per capita -0.01 -0.025***Ag-wealth index 0.05 0.034***# Oxen owned -0.03 0.012**ASP district dummy -0.033 0.028Moderate soil const. -0.106* 0.062***Severe soil const. -0.034 0.083***Rain onset delay 0.664** 0.325***Received MSD/CRextension (% SEA) 1.058*** 0.196***RFE CV (1996-2011) 6.264*** 0.1532008 dummy -0.418*** 0.031**Number of obs. 8,208 8,208
Summary of Findings
MSD CR MSD CRSocio-economic variables (labor, educ, ag wealth) + +Soil constraints - +Delay in the onset of rains + + + +Extension coverage + + + +Historical rainfall variability + +
Adoption Intensity
Conclusions • Simple cross-sectional analyses of adoption &
barriers fail to capture the real determinants • High levels of dis-adoption of CF practices in ZM • CF seems suitable only under certain agro-
ecological conditions • Suggestive evidence of adaptation benefits to
highly variable & delayed rainfall • Extension coverage is critical, but effects of
subsidized inputs/incentives need to be understood
THANK YOU!