Agroecological Intensification of Smallholder Farming - The Oxford Topic of Food, Politics, and Society - page 113

the parasitic weed Striga and stemborers) as well as water and nutrient deficiencies. An
intercropped legume ( Desmodium uncinatum , which is considered a dangerous inva-
sive weed in some countries) is dramatically antagonistic to Striga , while a border crop
of Napier grass reduces the reproductive success of the stemborer. The economic viabil-
ity of the approach is linked to the importance of the cut-and-carry livestock indus-
try in the area, as well as to the importance of maize (De Groote et al. 2010; Khan et al.
2008). The value of the Napier grass component is high when (a) fodder is needed and
(b) stemborers are a problem (De Groote et al. 2010). The value of Desmodium is high
when Striga is a threat to maize production. In addition, Desmodium can also supple-
ment Napier grass as a fodder. Local adaptation that maintains the key elements but var-
ies detail is needed; this is already happening with extension to other cereals in western
Kenya. The push-pull system's stability has been challenged by a disease of Napier grass
that threatens not only the push-pull system, but also the viability of the crop-livestock
system of the region. It might be anticipated that pests of Desmondium will become a
problem if planting becomes more widespread. A greater diversity of fodder crops is
clearly needed, and alternatives to provide the functions of Desmodium would be par-
ticularly useful.
According to Oerke (2005), weeds alone have the potential to cause 34 percent losses
in major crops worldwide. IPM strategies for weeds include improving herbicide use
efficiency (minimizing chemical use), developing biological and mechanical methods
(alternative curatives), and developing cultural or ecological methods (Bastiaans et al.
2008). Crop competitiveness can be increased by transplanting, seed priming, targeting
fertilizers to crop rows, and breeding for traits like early vigor and allelopathy. Systems
approaches have shifted the work of weed ecologists from a focus on the effects of weed
competition on crop growth to a focus on influencing weed population dynamics and
the longer-term development of weed populations (Kropff 2001). Approaches to reduc-
ing weed seed production include the use of weed-suppressive crop varieties, particu-
larly those that compete effectively for light through vigorous early growth or those with
allelopathic properties (Belz 2007).
While much of the literature on polycultures considers plot-level issues, it is acknowl-
edged that higher scales are important as well. The concept of “ecoagriculture” involves
the importance of managing agricultural landscapes in such a way as to conserve bio-
diversity, including wild species (Scherr & McNeely 2008). The diversity of nonculti-
vated crops can benefit farmers, including through the availability of harvestable foods.
Noncultivated areas in agricultural landscapes can provide refugia for natural enemies
of crop pests. Spatially explicit modeling revealed the importance of non-crop habitats
surrounding agricultural plots and their spatial arrangements relative to crop fields
(Bianchi et al. 2010), underlining the relevance of information about natural enemies'
behavioral traits (e.g., ability to disperse, tendency to aggregate). Although aggregation
of crop plots usually favors pest populations, the opposite was found for the Andean
potato weevil: clustering potato fields is both traditional and effective in controlling this
key pest (Parsa et al. 2011). A meta-analysis showed that the effects of landscape com-
plexity on pest pressure were variable: in 45 percent of studies, landscape complexity