Knowledge, science, and practice in ecological weed management: farmer–extensionist–scientist interactions - Ecological Management of Agricultural Weeds - page 129

A recent evaluation identified specific strengths in the PFI approach (Harp,
1996). Scientists and farmers developed a common language based on the
mutual understanding of each others' constraints and opportunities.
Scientists learned more about farmers' research needs, while farmers had a
channel to influence the university research agenda. For university research
and extension programs, the PFI network guaranteed accelerated diffusion of
results.Participating farmers from PFI developed leadership skills by organiz-
ing a farmer-managed research program to identify lower-cost cropping prac-
tices with reduced environmental impact.
Several difficulties were also identified (Harp, 1996). University staff cited
problems with colleagues and job tenure from on-farm work. The trials from
individual farms provided only site-specific results that were difficult to
publish in scientific journals. The adopted procedure of standardized treat-
ments for multifarm trials conflicted with PFI philosophy that prioritized
individual farmer decisions about treatments.Farmers also found that oppor-
tunity costs of data collection and trial management were high because these
activities made little immediate contribution to farm profits.
From 1987 to 1994, PFI farmers conducted 394 trials, including 78 on
weed management. Fifty-one trials on maize and soybean demonstrated that
ridge tillage without herbicides suffered no yield reductions and had lower
production costs (Harp, 1996). These results and others concerning fertilizer
reductions provided assurance for farmers contemplating input reduction.
However, a more diverse participatory learning process that included field
monitoring, group analysis of farmer planning and decision-making, and
reviews of weed patterns could have promoted more extensive farmer-scien-
tist collaboration.This broader range of co-learning activities might also have
allowed the active participation of farmers beyond those who were motivated
to run replicated trials. Improved management of the spatial and temporal
variability of weeds requires that more farmers document weed numbers and
distribution over multiple seasons and then use these data to discuss their cri-
teria for decisionmaking.
Ground cover in coffee: farmers observe weeds and propose
management alternatives
In Central America coffee is often produced on sloped land, either
under shade with low inputs or in open sun with pesticides and high levels
of fertilizers (Rice & Ward, 1996). Yields vary from 200 to 2700 kg of green
beans ha 1 , depending on soils, climates, farm size, and farmer resources.The
diversity of growing conditions and the layout of coffee fields create spatial
and temporal variability in weeds, insect pests, and diseases.Improving yields