Agriculture Reference
In-Depth Information
strategy (Alwang and Siegel, 1999). This is especially true for many FHHs, further reduc-
ing the time they can dedicate to their own plots of land (Simtowe, 2010; Takane, 2009).
Interestingly, there is evidence that in addition to providing short-term livelihood ben-
efits, legumes help reduce the risk of very low yields. Sirrine, Shennan, Snapp, et al. (2010)
measured risk of low yields as 75% lower confidence limits (LCLs), which is measured
as a one-in-four chance of yields falling below this level. Systems receiving fertilizer had
higher LCLs than those with no fertilizer, but notably, they also found that the maize/
legume intercrops had consistently higher LCLs than the equivalent nonlegume controls.
That is, in both the presence and absence of fertilizer, legumes reduced the risk of low
yields. Similar patterns of relative risk have been found in two other studies in Malawi.
Kamanga et al. (2009) also looked at risk using LCLs and found that addition of fertil-
izer reduced risk, but that maize intercropped with pigeon pea had the lowest risk of any
system. In another study, spatial variability in yields was consistently higher in unfertil-
ized than in fertilized maize (Snapp et al., 2010), and yet superior yield stability as mea-
sured by coefficients of variation was observed in maize/shrubby grain legume rotations.
Greater yield stability with legumes also translates into more stability of returns since
using legumes either avoids (if no fertilizer is used) or reduces the vulnerability of farmers
to fluctuations in fertilizer pricing (Snapp et al., 2010).
Overall, from livelihood, maize yield, and risk perspectives, we found that relay
intercropping pigeon pea with maize offered the most sustainable and low-risk, low-cost
option for the poorest farmers to improve production and food supply. If fertilizer is avail-
able at low cost, even limited amounts could improve the pigeon pea system for these
farmers. In contrast, the wealthier farmers have many more options. They are well posi-
tioned to benefit from the fertilizer and legumes due to their flexible marketing strategies,
higher-quality landholdings, access to labor, and the ability to afford inputs. Based on
net returns, under D1 the full fertilizer, followed by
S. sesban
plus full fertilizer, were the
most promising systems, whereas under D2 the fully fertilized maize gave the highest
returns. When the reduced risk associated with legumes is taken into account, as well
as their soil improvement potential, any of the legumes together with fertilizer are good
options for wealthier farmers. Again, pigeon pea with full fertilizer may be the best option
given the value placed on it by women and the potential for women to gain control over
any money generated from pigeon pea sales—a point that was raised in our focus groups,
where women spoke of the importance of having income that they could use to purchase
household necessities.
In terms of long-term soil quality improvement, we were unable to detect significant
changes in soil percentage C or percentage N over time under any of the systems tested, so
we cannot use these indicators to differentiate among the systems. More nuanced indica-
tors of nutrient cycling and SOM dynamics were needed to better assess legume impacts
on soil fertility. One challenge is coming up with indicators, especially biological indicators
that are low cost and feasible, to measure at multiple field sites distant from laboratory or
cooling/freezing facilities, as in this case. Yet, it is critical to collect this kind of data under
realistic field conditions and not only at experiment stations, where legume productivity
tends to be much higher (Mafongoya et al., 2006). Moebius-Clune et al. (2011) argued for a
suite of relatively low-cost soil quality indicators that reflect a combination of soil physical,
chemical, and biological properties, each linked to important ecosystem processes. These
include a range of macro- and micronutrient levels, water-stable aggregates, available water
capacity, penetrometer resistance, and biologically active soil carbon measured with a very
dilute potassium permanganate method and a handheld colorimeter. Penetrometer resis-
tance was measured in the field, and the remaining measurements were made using sieved
Search WWH ::
Custom Search