- 1. Farm typologies and resilience:Household diversity seen
asalternative system statesPablo Tittonell1,21Centre de coopration
Internationale en Recherche Agronomique pour le
DveloppementMontpellier, France2Tropical Resource Ecology Program,
University of Zimbabwe, HarareFarming Systems Design September 27,
2011Brisbane, Australia
2. IntroductionFarm typologies(i) Aim at categorising diversity
in livelihood strategies and/or levels ofhousehold resource
endowment;(ii) Used in a diversity of applications (research,
policy, monitoring and evaluation, econometrics, etc.);(iii) Should
respond to the objectives of the study/ intervention;(iv) Often
used as the basis for scaling-up/ scaling out-approaches;(v)
Different methods are used to categorise household diversity:
statisticalclustering, participatory rankings, expert knowledge,
etc.StructuralBased on resources and asset levels Functional
Livelihood strategies and household dynamics 3. Structural
typologiesSmallholder households in NE ZimbabweFarm type Farm size
# Livestock# Scotch Maize yield (ha)carts (t ha-1)Poor 20 < 0.7
0 None 0.2 1.0 Clustering (e.g. multi-dimensional scaling)Medium0.7
1.22411.0 1.2 40Rich> 1.2 4 - 2222.0 3.5 50%
similarityWithin-group similarity (%) 60 80100 Farm samples 4.
Functional typologies Resource endowment (allocation pattern)
Production orientation (subsistence, markets) Livelihood strategy
(e.g., access to non-/ off-farm income) Household structure
(position on farm development cycle) Household dynamics (where do
they come from/ go?)Dependence on off-farm income Hypothesis: -+ +
+ Different household types may be Type 2 as alternate states of
theRich seen farms same system (in this case, the smallholder rural
livelihood system) + Type 1 +Resource endowmentResource
endowmentfarms Market orientation This may allow: Food s elfType 3
su Medium Understanding the nature and resilience of poverty
trapsyffic ien cfarms Analysing possible shifts (or not) between
household types in response to Type 4 farms - e.g. poverty
alleviation measures, market or climatic scenarios, etc. - Type
5Poor farms - - 5. Underlying assumptions about household
diversityFarm productivitySmallholder farming systemsA) No
alternate regimesarmsB) Two alternate regimesB B fState of capital
stock (fast variable)tock State of capital stock (fast variable)- L
ives Syst em stateI- Hysteresis ThresholdA A+Destocking+I
Stockingte ISyst em sta Underlying (controlling) variable
Underlying (controlling) variable Assumptions: Assumptions:
Policies and developmentMoving form A to B may not be so
interventions may impact on thefarms easy; these are two
alternative right driving variables to move -lives tock system
regimes; interventions need gradually from A to BNonto provoke a
jump (hysteresis) A threshold may be thereDiscontinuity,
irreversibilityResources 6. Livelihoods aspirations and strategies
of the poor Dorward (2009) People aspire to maintain their current
welfare and to advance it Expanding their existing activities
and/or moving into new activities Hanging In: assets are held and
activities engaged in to maintain livelihoodlevels (adverse
socio-economic circumstances) Stepping Up current activities
engaged in, with investments in assets to expand these activities,
to increase production and income to improve livelihoods Stepping
Out activities engaged in to accumulate assets which in time can
provide a launch pad for moving into different activities e.g.
accumulation of livestock as savings to finance childrens education
7. Western Kenya1000 inhabitants per Km2 8. Heterogeneity and
landscape dynamicsFarm developmental cycle (Forbes,
1949)ResourcesMaturityh De wtocli gr ne ndanta d Maintaining
&endismreproducing sosh resources; blilut taion productionEs
Expanding Formerlyfamily & household that has been subdivided
as the children married a single may
exceedSub-dividingconsumptionland resources Time (life cycle)
Tittonell et al., Ag Sys 2007 9. A functional typology for East
African highland systemsType 1 Type 3MKT LVSTKFOODMKT CSHCNSHOME
OFF-FA RM OEWealthier householdsMid-class to poor
householdsCSHWOODLVSTKType
2ResourceHOMECSHallocationCNSWOODstrategiesMKT LVSTK Type 4MKT
LVSTKCNSCNSFOODHOME FOODHOMEOFF-FARM WOOD WOOD Type 5 Cash MKT FOOD
Labour CNS HOME OFF-FARM Nutrients WOODCSHTittonell et al., AGEE
2005a,b 10. Functional farm types and system statesPerformance
(well-being)T2 T1 Stepping out PStepping up T3 PT4 Hanging in T5R
RResources (natural, social, human) 11. 100 40Total
housIncomeIndicators of resources and performance Total household
income (kSh yr )-1120 300Household type 1 Income per capita (kSh yr
)5020 -1Household type 2250 100Household type 3Household type
420080 Household type 5 0 0 300120 01 2 3 4 50.0 0.4Total household
income (kSh yr )15060System state II -1Income per capita (kSh yr
)-16250 1.0100 2 t ha-140 Food production per capita (t
dm)100Stepping outFood production (t dm farm-1)1 us$ day-1 50 520
200 0.880 0 1 t ha-100 4 1 23450.0 0.4 0.81.2 150 0.6606-12 t ha120
1.0Food production per capita (t dm)3Household type 1 Food
production (t dm farm-1) 100 40 come per capita (kSh yr )5
-10.8Household type 2100-10.4 1 t haHousehold type 34 2500.620
Household type 43800.2 Household type 510.40 02 System state 10
I602 3450.000.00.2101 2 Cropping land (ha) 3 450.0 0.4
Self-sufficiency 6-11.0 Cropping landt (ha)40 2 ha (t dm)0 0.0
Landrm-1) 01 23 4 50.0 0.4 0.81.2 5Cropping land (ha) Land:labour
ratio1 us$ day-1 0.8 20 12. Testing across a wider range of
systems(Market orientation) % % area allocated to cash crops(Market
orientation) area allocated to cash cropsMeru S. Meru
S.ViableMbalef ar Mbalemsiz e sDesirable effectof
intensificationAgroecological potentialPopulation (People per Km2)
MiMbeere 0-2nim Mbeere 3 - 10 umfar 11 - 20 Vihiga mMbalesizTororo
21 - 50 Vihigaes 51 - 100Siaya Tororo 101 - 200 Siaya Meru S. 201 -
500 501 - 1000 Vihiga >1000 % area under fallow% area under
fallow(Traditional management)Siaya(Traditional
management)TororoMbeere Markopport etunitiesion lat y pu t Po ensi
Tittonell et al., AgSys 2010d TSBF, 2007 13. Concluding remarks
There is ample potential to bring down to earth the attractive
concepts around resilience thinking for use in the context of
farming systems research A promising entry point: Farm typologies
seen as alternative states of a given rural livelihood system This
challenges a few assumptions: the existence of thresholds,
continuity, reversibility, and the use of classical socio-economic
indicators to cluster similar groups out of large household surveys
Poverty traps become evident: improving livelihoods (i.e.
facilitating a shift upwards) does not necessarily imply more
resources (e.g. agricultural inputs, livestock or more efficient
technologies) Will this always work? Where not, why? More research
is [email protected] for your attention