ILEIA Newsletter Vol. 11 No. 3 p. 22

Selective weeding: ground cover and soil conservation in coffee in Nicaragua

Charles Staver, Amilcar Aguilar, Victor Aguilar and Santiago Somarriba

Farmers readily recognise differences in the growth habit of weeds. They know which weeds are perennial or annuals. They have observed how roots grow and how different weeds came back after weeding. They also have the practical experience that different weeds cause different damage to crops.

For some smallholder coffee farmers weeds are not a problem, because they interplant fruit trees, coffee, bananas and plantains and other short and long-cycle crops in a small mixed orchard, often an extension of the dooryard garden. A second and more common category of small producers in Central America grows coffee in slightly larger fields in association with shade trees and sometimes fruit trees and bananas. These farmers, in spite of recognising different weed types, generally treat all non-crop green matter as weeds to be controlled.

Weed control for this type of smallholder coffee farmers in Central America consists at the minimum of a complete machete weeding several times a year. Weeds are cut back nearly to ground level when they reach 20-40 cm. Frequently two weeks or so after each machete weeding, a cheap herbicide like paraquat or 2,4-D is applied. When farmers have more ready access to credit and receive more technical assistance, they use herbicide mixtures, including pre-emergent, which keep the soil surface free of weeds for a longer period after weeds have been slashed back.

Under this approach to weed control, coffee fields alternate between high weed levels and bare soil. Right before weeds are slashed, they often are shading lower coffee branches and interfering with root growth. The regular use of herbicides leads to longer periods of bare soil and increased risk of erosion. In addition, certain hard-to-control weeds recover quickly from the commonly used weed control practices. These weeds tend to increase as the same weed control measures are repeated year after year. Vines, grasses and creeping and erect perennial broadleafs, all of which sprout vigorously after slashing or a burndown herbicide, are among farmers’ problem weeds.

Selective management

An alternative approach to weed management in perennial crops is being developed in Nicaragua based on differences in growth habit among different weeds and farmers’ abilities to distinguish these differences. Two parallel objectives are proposed. First, instead of practices which indiscriminately control the general green biomass, selective practices are directed at the most problematic and least desired weed types in a specific field. These generally include vines, grasses and tall growing perennial broadleafs, although particular farmers may have their own ideas.

Second, weed control practices should not eliminate weeds which serve as natural living ground cover. Such weeds are usually creeping, low-growing plants with shallow root systems. They tend to form a complete ground cover reducing other weed growth without competing with coffee plants. Selective weed management with these two objectives over time should lead to fewer aggressive weeds, more complete and continuous soil cover and lower weed control costs.

Over the past five years we had carried out experiments in small plots to look at the biological feasibility of selective weed management and in large plots to measure costs, effects on coffee growth and yield, and interactions with other pest problems. At the same time we have also developed participatory training and action research approaches for ground cover management in coffee. This work is part of a larger project on pest management with small farmers in Nicaragua who produce coffee, tomato, or plantains/bananas.

Biologically feasible

To validate the ecological concept that the species of weeds in a field change in response to selective pressures, different weed control practices were applied in small interrow plots in a young coffee field. When the study began in June 1991 the field was dominated by weeds damaging to coffee, primarily annual broadleafs, with significant perennial broadleafs and some grasses. Natural cover weeds were present in very small amounts. The balance between damaging and cover weeds changed in response to different control measures during two years (Figure 1).

Slashing weeds with a machete or the use of a mixture of herbicides after slashing (Figures 1a and 1b) maintained high levels of damaging weeds and very low amounts of natural cover weeds. The use of 2,4-D after slashing weeds reduced damaging broadleaf weeds with a corresponding increase in natural cover grass weeds like the one shown in the photo. Figure 1c shows that each year the balance between damaging and cover weeds shifted, although even in June 1993 damaging weeds were more prevalent again.

Flexible management was the only treatment in June 1993 with cover weeds outgrowing damaging weeds. In this treatment slashing and herbicides were used selectively to promote cover weeds and to eliminate damaging weeds. Practices used varied as the type of weed in the plot changed from weeding to weeding. By the third year, herbicide use in the flexible treatment declined to 1.5 l/ha, compared with 4-7 l/ha in the other two herbicide treatments.

Less costs, more information

Three different approaches to weed control were compared in large plots starting in 1992. In conventional non-selective control all weeds were eliminated twice a year with slashing followed by the application of paraquat, 2,4-D and a pre-emergent, a common and longstanding practice in Nicaragua. In selective management, with slashing and herbicides, spot applications with selective herbicides and selective machete chopping were used to reduce the growth and seed production of damaging weeds.

In selective management with only machete use the stubble height was varied to favour certain weed types and reduce others. In both selective management approaches growth and seed production of natural cover weeds were conserved. The coffee rows in all three treatments were kept free of weeds.

As Table 1 shows, natural cover weeds were abundant in the selectively managed plots, while the conventional approach had virtually no cover weeds. Selective management was comparable in cost with the conventional non-selective control and used less herbicide. In selective management with machete coffee yield was reduced, because, even though natural ground cover weeds predominated, total weed biomass was too high.

Currently we are trying to reduce total weed biomass and still conserve the natural cover weeds in the selective slashing-only management. Increasing the width of the weed-free row area, more frequent weed control, chopping of cover weeds, and increased shade could help reduce the problem of reduced yields. The selective approaches require more frequent, shorter-duration weed control activities and a better decision-making ability. They could be characterised as information-intensive.

Farmers learn selective management

Participatory training methods have been developed to assist farmers in organising their practical observations and experience with different weeds into an experimental ground cover management plan for their coffee fields. In a first workshop farmers begin by roaming a coffee field and bringing back different damaging and cover weeds. They discuss the characteristics of each weed and how the different weeds might be grouped by their growth habit and potential damage to coffee plants. A simple matrix lay-out on a white sheet has ten columns for weed types and other ground covers.

The matrix has three rows: height of weed growth in relation to the coffee plant, advantages and disadvantages in terms of competition and soil protection, and weed reproduction mechanisms (sprout and/or seeds). Symbols and graphics like red and green buttons for disadvantages and advantages are used instead of lettering to fill out the matrix.

The following seven groups of weeds are commonly formed: grasses, sedges, annual broadleafs, perennial broadleafs, vines, broadleaf cover weeds and grass cover weeds. Farmers may decide to form more groups. Certain weed groups may also be absent from a field. Farmers are also encouraged to observe other cover in their coffee fields like leaf litter and shade. These are incorporated into the matrix to demonstrate that ground cover weeds, leaf litter and shade trees offer advantages like soil protection without disadvantages like competition with coffee plants.

Small groups of 2-3 farmers then walk the field systematically observing the type of weed or ground cover in an imaginary point the size of a pencil point at their shoe tip. Each group adds up their results which are then summed up for the whole field in a large group session. The workshop closes with a discussion on how to change weed control practices to reduce the more damaging weeds and to promote cover weeds.

A set of cards with drawings of different weed types helps the group visualise how certain practices favour certain weeds and disfavour other weed types. Cards can be partially covered to illustrate slashing to different stubble heights or turned over to symbolise elimination by hand or with herbicides. Cards with weeds in flower are used to emphasize the importance of seed production in future weed presence. Cards are moved laterally to illustrate one weed type increasing at the expense of another. The photo shows the card layout for discussion about the effect of shade on different weeds.

Action research

The coffee field where the workshop is done becomes an action research plot with nearby fields for comparison. Practices are discussed and the group agrees on a set of actions to carry out before the next meeting, such as planting shade, pulling up vines by their roots, or not slashing cover weeds. The same field is also used for other workshops on coffee insect and disease problems.

The group of farmers discuss how changes in management might affect their different pest problems. As they prepare to return to their homes, they are encouraged to try the sampling in their own fields and return for follow-up meetings to discuss what they found and what changes they decided to make in their weed management.

The goal of the series of workshops organised around a group of farmers, action research plots and systematic observation and data collection is to increase each farmer’s capacity to make better resource management decisions. This capacity includes how to use their local resources more effectively, how to manage year-to-year and field-to-field variability and how to evaluate possible new practices or modifications learned from other farmers, extension agents or agrochemical field sales representatives.

Charles Staver and Amilcar Aguilar, CATIE IPM Project, Apartado P-116, Managua, Nicaragua.
Victor Aguilar, National Agrarian University, Apartado 453, Managua, Nicaragua.
Santiago Somarriba, Centro Experimental de Cafe, Masatepe, Nicaragua.

Table 1

Fig. 1

Back to Top