in Kenya

Document Sample
in Kenya Powered By Docstoc
					2.9 Environmental Impact of Smallholder Dairy Technology Options
in Kenya

An analysis of the environmental impact of smallholder dairy technology followed the economic analysis.
Environmental impacts were assessed at regional and watershed levels. Several environmental indicators
were monitored and reported as part of the simulation process. Two of the important ones were water
runoff and erosion.

In the simulations for Kenya, these indicators were extracted for each of the simulations conducted on the
selected representative farms. The area-weighted mean water runoff and soil erosion loss for each province
delineated for agricultural sector analysis was computed.


2.9.1 Methodology

The Erosion Productivity Impact Calculator (EPIC) was used to assess the runoff and soil erosion impacts.
Simulation were set up for representative households that reflected crop rotations and practices for each of
the ecological zones. For these environmental impact calculations, results from the same simulations were
used in all districts when a zone crossed more than one political district. In retrospect, a more refined
approach should conduct simulations for each zone/district combination.

EPIC is a hydrologically based model capable of growing multiple crops in rotation as impacted by soil
moisture, fertility, management practice, and crop production coefficients. The model output shows crop
yields, biomass production, hay, and environmental responses such as runoff, nutrient loss and carbon and
pesticide loading.

In this study, erosion and runoff were targeted since limited fertilizer and pesticides were used on these
farms.

The runoff and erosions simulations were extracted from each of the 20-year simulations by crop rotation
and ecological zone. To expand these representative household simulations to the Kenya political units
(province), each ecological zone was assigned a weight (fraction) for each of the six provinces. If a crop
was not reported in an ecological zone, a zero was assigned to the weight for that zone. For example, coffee
was not reported in the wheat zone of the Rift Valley province so a zero value was reported for the
weighted value of coffee. After these assignments were made, the zone area-weighted yields for each crop
were recalculated for each district. The number of hectares for each crop in each political district was
obtained from the equilibrium values from ASM model. No attempt was made in this analysis to reallocate
weighted yields used for each ecological zone within a political district among scenarios. Only the total
change in area used by each crop rotation for each scenario among political districts was considered.

A land use category of idle land was added to make district totals of cropland equal in all scenarios. Since
no simulations were made to estimate erosion on unused (idle) cropland, the erosion and runoff rates from
native grass were used as proxies for these coefficients.

2.9.2 Results

This environmental impact analysis indicates that impacts due to the land allocations between districts among
the scenario are environmentally neutral for runoff and erosion. Many simplifying assumptions were made in
this analysis. One should note that technologies can be used to impact the environment. For example, this
analysis assumed fixed distribution of crop rotations across soils and slope. With the increases in areas of
native grasses, opportunities exist for development of policies to encourage the planting of the grasses on the
highly erosive soil types and the lands with steep slopes. This would definitely reduce erosion. Individual
simulations indicated erosion reductions of up to 40% when moving from crops like maize to native grasses
on the same fields. For example, in the coffee zone, the erosion estimate for maize was 11.6 vs. 6.3 mt/ha
for native grasses. With further sub-divisions of the land base, these differences in land allocation and
resulting environmental implications could be quantified.

Although there were wide variations in the simulated values of the respective land uses (e.g. annual erosion
rates from 20.1mt/ha in maize in the wheat zone vs. 2.6 mt/ha for native grass in the coastal zone), there was
no significant environmental impact across any of the political districts for the four technology scenarios
considered. The changes in land uses associated with the technology packages were environmentally neutral
in Kenya at the household level. Table 2.9.2.1 shows the relative weights assigned to each of the zones by
the six political districts. These values are the fraction of total zonal land area falling in each province. These
weights were adjusted in the individual crop calculations as described above. Table 2.9.2.2 provides the
average runoff by zone and district weighted by the respective crop rotation areas. Table 2.9.2.3 gives the
erosion rates using the same procedure. Both the runoff and erosion tables report the values for the “Im-
proved Dairy Current Adoption” scenario. Values for the Traditional Dairy and the Improved Dairy Full
Adoption Scenarios are not reported here because they show only very minor changes from the Current
Adoption scenario. Only one of the cell values in either table varies more than 3% from the Current Adop-
tion values. The only exception was reported in the horticultural zone of the Western Province where
erosion dropped from 8.6 mt/ha with the current adoption of existing technology to 5.5 mt/ha for both the
traditional and full adoption technology scenarios (not shown in the tables). This is traceable to the changes
in crop areas dedicated to groundnuts and sorghum in the Current Adoption scenario coming from and
returning to the idle land category in the Traditional and full adoption technology scenarios respectively.
However, this environmental zone only accounts for 6% of the Western District’s total area.


2.10 River Basin Level Environmental Impacts-The Sondu River

2.10.1 Selection of Watershed

The final analysis in the Kenyan case study was to assess the environmental impact smallholder dairies were
having on the Sondu River watershed. To assess changes at a watershed scale, we linked the projected land

      Table 2.9.2.1 Weights used in Runoff and Erosion Calculation by Ecological
      Zones and Kenya Districts
      District      Wheat     Tea     Sheep     Horticulture   Coffee   Coast     Total
      Central        0.13     0.10      0.45            0.32                       1.00
      Coast                                             0.07             0.93      1.00
      Eastern        0.04     0.02       0.03           0.91                       1.00
      Nyanza                  0.23                      0.32     0.45              1.00
      Rift Valley    0.15     0.18       0.16           0.37     0.14              1.00
      Western                 0.24       0.10           0.06     0.60              1.00