Combining land restoration and livelihoods - examples from Niger
Session 3.1 small farm diversification strategies
1. March 12, 2014
Small farm diversification strategies by coffee farmers
around Mount Kenya in Kenya
Sammy Carsan
Aldo Stroebel, Frank Place and Ramni Jamnadass
Word Congress on Agroforestry, 11th February, 2014.
DELHI, INDIA
2. March 12, 2014
1) Background: Evolution of Kenya’s coffee smallholder sector
2) Study Context: Why coffee agro-forests?
3) Statement of objectives & hypotheses
4) Methods: sampling approach
a) Coffee farms typologies
6) Results & Discussions
Conclusions
Presentation outline
3. March 12, 2014
• Kenya’s coffee in the past estimated to offer livelihood support to over
5 million people directly and indirectly
• Developments summed during and after the ICA market regime
• Living standards, incomes, food security in coffee growing negatively
• Uncertainties in international market and loss of coffee productivity
have affected overall coffee profitability
• Are smallholders ready for incentives to shift from traditional cash
crop systems? Is AF tree cultivation a good incentive?
• How will resource availability and market drivers influence
investments on cash crops
Evolution of Kenya’s smallholder coffee
4. March 12, 2014
Compared to robustas the arabica price differential (premium) is about US$ C 60 (s.d
34.63, n = 129) per pound in the last ten years.
Mean price for Columbian Milds, New York Composite and Robusta coffee from
2000 to 2010
Source: ICO Statistics, 2010
5. March 12, 2014
Kenya’s coffee exports fell by over 50% between 2000 and September, 2010;
world market share declined from 3.1% in 1986 to 0.6% by 2006 (ICO, 2010).
Columbian mild coffee exports by: Columbia, Kenya and Tanzania
Source: ICO statistics, 2010
6. March 12, 2014
• ‘Shaded’ coffee as opposed to open ‘sun’ coffee’-a more sustainable production approach (Mas and
Diestch, 2004). Coffee AF systems act as reservoirs of indigenous tree species (Perfecto et al., 2005).
Trees yield complimentary products e.g. fruits, timber and firewood which diversify, diet and
stabilize farmer incomes
• Peeters (2003) :coffee shaded with any density of Cordia alliodora has better benefit-cost ratio than
un-shaded estates although yields were lower. Simplifying these systems was disadvantageous even
if coffee production increases.
• Structurally complex habitats support more diverse fauna (Garcia et al., 2009).
• Coffee AF seen as an approaches to build alliances between ecologically sustainable agriculture and
conservation efforts in protected areas. Trees contribute ecological services similar to those
provided by forest e.g. soil protection, nutrient cycling, water retention and carbon capture
(Chazdon et al., 2009).
• Farmers benefit culturally by maintaining biological diversity that ensure productivity (Lengkeek et
al. 2005).
• Genetic diversity helps farmers to manage their inputs in more efficient ways- e.g. a mix of fast
growing and slow growing timber grown for different markets; fruit species with different fruiting
phenology to contribute to HH food security (Dawson et al., 2009)
Coffee agro-forests and sustainability?
7. March 12, 2014
• To assess how changes in coffee production influences small farm
intensification/diversification strategies
Hypothesis:
• Smallholders with stagnated or decreasing coffee production (yields,
density of bushes) support higher enterprise intensification rates with
maize, banana, livestock and trees.
Study objectives
8. March 12, 2014
• Cross-sectional survey in three
coffee districts of Mt Kenya (Meru ,
Embu & Kirinyaga)
• The zones are comparable on coffee
and other crops production practices
and largely representative of
smallholder coffee systems in Kenya
• The regions have strong farmer
organization by cooperatives and
societies
Research Methods
9. March 12, 2014
• 10 Farmer Cooperative Societies
(FCS) covering UM1,2 & 3 in Meru,
Embu & Kirinyaga Counties selected
• Membership used to draw a random
set of farms categorized as either
“increasing”, “decreasing” or
“stable”
• Cherry deliveries for last 5 years
(2004-2008) considered
• 5 farmers were picked per category
selecting 15 farms per FCS, later 2
farms were selected giving 6 farms
per FCS.
• 60 farms were selected per category
comprising 180 farms for the entire
survey done from June- August 2009
Sampling strategy
10. March 12, 2014
• SES data collected: family & land size, labour and non labour costs, distance to markets, coffee
size (bushes & cherry), tree inventory, maize & banana value, livestock units (TLU) & daily milk
value, fertilizer and manure use, crop diversity, tree basal area, avocado trees
• Simple descriptive statistics such as means(s.d) were used to profile household and farm
characteristics
• Intensification (value/Ha) verified by: coffee bushes, cherry value, no of trees, maize, banana
outputs per hectare ad inputs such as fertilizer, manure labour and non labour costs per hectare
• Regression analysis: pairwise correlations and GLMs used to assess effects of coffee production
‘trends’ on farm intensification
• Response variates were transformed by natural log or square root (after Shapiro Normality tests)
• Farm types: increasing, decreasing and stable were used as explanatory variates. Farm size
effects were also tested
• Fisher’s least significance difference (LSD) was used to compare differences between farm
categories at P < 0.05 and P < 0.1 level)
• Farm crop diversity was calculated using Simpson Index of Diversification (SID)
Farm and HH assessments
11. March 12, 2014
Farm tree inventories methods
• All trees ≥5 cm DBH measured
• Tree basal area (tree cross-sectional area
measured at breast height) calculated
• Local/common names of trees recorded
from local farmer consultations
• All trees were identified to species level
according to Beentje (1994) or Maundu
and Tengnäs (2005).
12. March 12, 2014
Results- Farm crop diversity
• Farm crop richness (SID) was 7.1 (s.d=1.71)
• Farm with lowest diversity had 2 crop types and one with highest had 12
• The ten most prevalent crops (maize, beans, banana, avocado, macadamia,
mango, beans, papaw, irish potatoes and khat) on farms represent 88% of
all crop types present on all farms.
• Stable farms had higher counts of livestock, banana and avocado;
increasing farms had fewer bananas but higher livestock presence while the
decreasing ones had smaller enterprise count overall.
• Other results showed that crop diversity was strongly and positively
associated with farm size but negatively related to fertilizer and TLU ha-1;
and maize and banana value ha-1 even though analyzed data was
inconclusive with the later.
13. March 12, 2014
Coffee and trees on farms…?
• 75% (156) farms cultivate 250-750 bushes ha-1
• 61%(110) produce 1000-2000 kg cherry ha-1 yr-1
• 41% (75) farms, tree density : 100-200 trees ha-1
• 30% (54) farms: TBA class of 1.1-1.9 m2 ha-1
• 22% (40) farms: TBA class of 2.0-2.9 m2 ha-1
• 35% (66) farms : TBA class of 3-5 m2 ha-1
• Average tree volume : 36.31 (31.1-41.5 ) m3 ha-1
14. March 12, 2014
Findings: functional coffee farm typology
2101 kg yr-1 (s.d = 1380)
1032 kg yr-1 (s.d = 907)
688 kg yr-1 (s.d = 709).
• Farm categories were statistically different (ANOVA test, P < 0.001)
• Increasing farms had more bushes than stable (17% , P < 0.05) and decreasing
ones (5.5% P < 0.05). Decreasing farms had higher bushes than stable (12.3%,
P>0.05)
15. March 12, 2014
Farms descriptive
variables
Coffee farm categories and means (s.d) for all variables
‘Stable’ ‘Decreasing’ ‘Increasing’ Rank
Maize value ha-1
17843(17955) 19098(19138) 14769(13594) Dec>Stab>Inc
Banana value ha-1
48463(78453) 57806(105745) 28416(35768) Dec>Stab>Inc
Coffee age years 35.23(12.96) 37.7(12.43) 33.7(11.97) Dec>Stab>Inc
Avocado trees ha-1
5.956(8.41) 7.580(15.77) 4.064(4.123) Dec>Stab>Inc
AFT Vol. ha-1
30.4(20.4) 34.4(38.1) 32.4(21.6) Dec>Incr>Stab
Family size 4.9(2.5) 4.8(2.2) 5.3(2.2) Inc>Stab>Dec
Land size (ha) 1.3(1.3) 1.1(0.84) 1.4(0.97) Inc>Stab>Dec
Cherry/bush (Kg) 4.4(3.2) 3.84(3.7) 4.63(2.53) Inc>Stab>Dec
Cherry value ha-1
61180(58526) 47220(49604) 69587(54482) Inc>Stab>Dec
Milk value day-1
118(119.4) 94(112.4) 152.3(143.8) Inc>Stab>Dec
Coffee bushes ha-1
499.7(379.1) 569.6(694.5) 601.6(497.4) Inc>Dec>Stab
TLU Ha-1
5.38(5.5) 4.26(4.9) 3.87(2.9) Stab>Dec>Inc
AFT ha-1
229(223) 202(152) 182(99) Stab>Dec>Inc
Macadamia trees ha-1
18.04(27.6) 16.89(19.81) 13.52(18.02) Stab>Dec>Inc
Mango trees ha-1
10.54(24.53) 7.57(8.678) 5.65 (9.68) Stab>Dec>Inc
Farms and households characteristics
16. March 12, 2014
Farms input sizes
• Fertilizer use ha-1 increases significantly (P < 0.05), with the coffee bushes, maize value,
manure, TLU; but is negatively related to farms crops diversity and distance to market.
• Manure application ha-1 was positively associated with coffee bushes, TLU and banana value.
Manure application was negatively related to farm crop diversity.
• Labour expenditure were positively correlated to fertilizer, manure, coffee bushes and
negatively related to market distance and crop diversity.
• Fertilizer and manure application were negatively correlated to farm sizes (P < 0.05).
17. March 12, 2014
General linear regressions coefficient for various ‘intensification’ variables
against farm categories and farm size
LSDs: Increasing vs Stable (*); Increasing Vs Decreasing (+); Decreasing Vs Stable (#). Significant
difference level: P<0.01(*** +++ ###); P< 0.05(** ++ ##); P<0.1 (*+ #)
18. March 12, 2014
Small coffee farms intensification
• Comparisons: cherry value, milk value, avocado trees, fertilizer, labour, non-
labour costs were used to gauge levels of farm intensification.
• Cherry value by the increasing farms was significantly bigger (P < 0.05) than the
decreasing ones but not for the stable ones. While value for the decreasing farms
was significantly lower (P < 0.1) compared to the stable farms.
• Daily milk value by the increasing farms was significantly bigger (P < 0.05) than
the decreasing (38.3%) and stable (22.5%) ones. The stable farms milk value was
not significantly higher (P>0.1) compared to the decreasing (20%) ones. Other
results confirmed that milk outputs could be increased in bigger farms.
• Surprisingly, results showed that maize value increases significantly with farms
TBA, no. of coffee bushes and banana value, but negatively related to farm and
household size.
• The no. of avocado tree ha-1 were significantly higher (P < 0.05) in the decreasing
(46%) and stable (32%) compared to the increasing ones. Findings further
showed that avocado intensification is significantly reduced in bigger farm sizes.
• Finally, findings revealed that cherry value and coffee bushes ha-1 was significantly
(P < 0.001) smaller in bigger farm sizes.
19. March 12, 2014
• Annual labour costs were significantly higher (P < 0.1) for increasing farms
compared to the stable (30.2%) and the decreasing ones (24.4%). Labour
costs for the decreasing farms were not significantly higher (P>0.01) than the
stable ones (7.7%).
• Other non-labour costs such as pesticides and seeds for the increasing farms
were significantly bigger (P < 0.05) compared to the stable (14.2%) and
decreasing ones (12.9%). Non labour costs were not significantly (P > 0.1)
higher in the decreasing (1.4 %) farms compared to the stable ones.
• Fertilizer use was significantly higher (P < 0.05) for the increasing farms
compared to the stable (33.9%) and the decreasing ones (27.9%). Fertilizer
use on decreasing farms was not significantly higher compared to the stable
(8.4%) ones.
Small coffee farms intensification
20. March 12, 2014
Discussion points
• Findings showing that coffee is significantly reduced in bigger farm sizes
suggests that input affordability, negatively related to farm size, could limit
future plans on coffee expansion.
• Farms decreasing or stable in coffee production are characterized by more
trees, banana, and maize while those with increasing coffee have bigger dairy
milk enterprises
• Smaller land sizes accelerates farm intensification with trees, coffee, banana,
maize and livestock enterprises while crop diversification is significantly
enhanced by bigger land sizes.
• Fertilizer use increases significantly with the no. of coffee stems, maize value
and amount of manure applied; however distance to markets have a negative
correlation with fertilizer use
• Manure use is positively associated with labour costs, coffee bushes, TLU and
banana value but significantly reduced with higher farm crop diversity
• Labour costs can significantly affect fertilizer & manure use implying higher
costs could seriously influence farm nutrient status
21. March 12, 2014
• Farmer intensification/diversification interests with lower input but higher
return crops such as avocado shown to be significantly bigger in
decreasing and stable farms compared to the increasing ones.
• Farmers that can’t afford high costs of maintaining coffee production opt
for crops that provide food security (maize, banana). They also have
higher levels of livestock units and trees to serve as self- insurance for the
household.
• Measures such as the no. of avocado trees and milk output (dairy
activities) per farm were useful indicators to discern farm productivity
differences rather than total tree and livestock counts per farm
• It was clear that inputs such as fertilizer are tremendously reduced with
declining coffee production raising concerns that small coffee farms could
decline to low productivity systems.
• These trends would probably change with a high profitability shift in coffee.
It is therefore critical that policy makers are aware of these trends in order
to support enterprises that are attractive to farmers.
Concluding remarks