Pasture development in the uplands

The big push to open up new cattle pastures in Amazonia started in the late 1960s and was targeted at forest areas. Attracted by generous fiscal incentives promulgated in 1967, corporations started investing up to half their taxes in development projects in Amazonia. A "grass rush" of major proportions quickly ensued in eastern and southern Amazonia (Sternberg 1973). Some 15 million ha of forest, mostly in southern Pará, and northern Mato Grosso, have been felled and planted to a variety of grasses since the mid-1960s.

First-generation grasses on recently formed pastures in the uplands were mainly guinea grass (Panicum maximum), Brachiaria (Brachiaria decumbens), and, in drier areas, jaraguá (Hyparrhenia rufa). Within about five years, however, weeds, resprouting trees, and compacted soils depressed the productivity of many of these new pastures. The carrying capacity of neglected pastures often fell below 1 head/ha, compared with 1.5-2 head/ha on newer, bettermanaged pastures. At Fazenda Boi Branco near Paragominas, Pará, for example, the carrying capacity of a 12-year-old Panicum maximum pasture had dropped to 0.5 head/ha until it was restored with Brachiaria brizantha cv. Marandu in 1990.

Table 7.1 Pasture grass turnover in Amazonia in response to weed and pest problems, among other factors

Note: All species are still cultivated in the Brazilian Amazon, but first-generation species are now much less common.

Pests and germ-plasm turnover

Spittlebugs (Deios spp.) have triggered a turnover of grasses in artificial pastures in Amazonia. In the mid-1970s, spittlebug attacks, especially by Deios incomplete, exacerbated weed problems in many Brachiaria (Brachiaria decumbens) pastures (fig. 7.2). Known as cigarrinha in Brazil, the small bugs can spread and multiply quickly. Spittlebugs feed on the grass sap, thereby withering the pasture and allowing weeds to proliferate (Penny and Arias 1982: 65).

Spittlebugs forced ranchers to seek alternative grasses in order to recoup lost productivity. The stage was set for a second generation of pasture grasses (table 7.1). In 1976, many ranchers turned to quicuio da Amazonia (Brachiaria humidicola), a fast-growing grass with moderate resistance to pests. An African grass, as are all introduced pasture grasses in Amazonia, quicuio has also been widely used for erosion control along roads, railways, and electrical transmission lines. By 1982, however, some pastures of B. humidicola were being severely attacked by Spittlebugs in the Paragominas area, possibly because of the pronounced dry season, which weakens the plants.

In wetter areas of Amazonia, such as near Belém Brachiaria humidicola and to a lesser extent B. decumbens still resist spittlebugs. Belém and much of the adjacent Bragantina zone receive around 3,000 mm of rain annually, compared with about 1,700 mm in the Paragominas and Manaus areas where spittlebugs have wreaked havoc. At Fazenda Itaqui, 54 km east of Belém on the BR 316 highway, a 20-year-old pasture of B. humidicola is still productive because it is fenced, thereby allowing rotation of cattle-grazing, and is weeded and fertilized periodically (fig. 7.3). Also in the Bragantina zone, a 20-year-old pasture of B. decumbens cleared from second growth is still grazed because the heavy rainfall depresses spittlebug populations and because the pasture is fenced and weeded; this pasture, located at Fazenda São Judas Tadeu, 19 km from São Miguel do Guama, has never been fertilized.

Many ranchers have been planting braquiarao (Brachiaria brizantha cv. Marandu) since 1983, in part to escape problems of spittle-bugs. Also known as brizantão, B. brizantha is more vigorous than B. humidicola, provides better ground cover to suppress weeds, and currently resists spittlebugs. Although more demanding of soil nutrients and physical conditions than first-cycle Brachiaria species, brizantão is rapidly replacing first- and second-generation pasture grasses in many parts of the Brazilian Amazon (fig. 7.4). Third-generation pastures are generally under more intensive management.

Weeds and degraded pastures

Close to half of the artificial pastures in Amazonia are degraded (Hecht 1985; Serrão and Toledo 1988). Degradation generally refers to weed infestation, which in turn can be facilitated by pests or deteriorating soil conditions. The proliferation of weeds can be a symptom of soil exhaustion or compaction, but not always. Pastures with numerous termite mounds appear to be in poor condition. The infestation of volunteer plants can thus be a consequence of soil depletion and compaction, overgrazing, underutilization, or insufficient weeding. Weed infestation is a sure sign of poor pasture management.

Weeds originate from seeds dropped by birds, other animals, and the wind, and from imported seed. Stump resprouting in relatively new pastures can also quickly shade out grass. Weeds can take over relatively fertile pasture if they are not checked. Overstocking or understocking can also favour the emergence of unwanted plants in pastures. Manual clearing is the most common method of controlling weeds on an annual basis; herbicides are generally too expensive.

Dozens of unwanted plants arise in pastures in Amazonia. The composition of weedy communities varies markedly within an area and between different parts of the Amazon. A few weeds stand out as being especially troublesome. In the Manaus area, vassoura de botão (Borreria sp.) is a pernicious rubiaceous weed, while around Paragominas, matapasto (Cassia spp.), milkweed (Asclepias spp.), and species of Verbena infest some pastures. Along the PA 256 road linking Paragominas and Tomé-Açu, lacre (Vismia guianensis) forms virtually pure stands in some degraded pastures. In the Marabá, area, babaçu palm (Attalea speciosa) or assar-peixe (Vernonia sp.) typically dominate poorly managed pastures. Some pasture weeds are an important source of food for butterflies and bees, which may in turn provide pollination services to other plants, including economic ones. The complex interactions of weed populations and pasture management warrant further study.

Pastures sown with various species of Brachiaria are prone to infestation with several grass weeds. Seed from the Brachiaria grasses is collected from the ground, whereas seed from Panicum maximum is gathered from the mature spikes, thus reducing the chances of weed contamination (Nepstad, Uhl, and Serrão 1991). Pastures of Brachiaria can be established by cuttings, but this procedure is labour intensive and can pose problems for ranchers, even though the minimum wage is only US$60/month. Furthermore, workers are often scarce in pioneer areas. At least one vigorous pasture weed, capim navalha (Paspalum virgatum), is thought to have entered Amazonia with Brachiaria planting seed (M. Simao, pers. comm.). Although P. virgatum is also a grass, it is much less palatable to cattle. Sapé (Imperata brasiliensis), the ecological equivalent of the notorious alang-alang grass (Imperata cylindrica) of South-East Asia, may have been introduced to some parts of Amazonia after 1970 in seed of Brachiaria humidicola. Both Paspalum virgatum and lmperata hrasiliensis are indigenous to Amazonia; a striking feature of weeds in both pastures and fields of Amazonia is that most are native to lowland South America.

Some ranchers are aware that not all invading or resprouting plants in pastures are a nuisance. Fazenda Boi Branco near Paragominas is allowing some forest trees to resprout in pastures if they have timber value. Some leguminous pasture "weeds," such as Mimosa sensitiva and matapasto, enrich the soil with nitrogen. Still other volunteer plants in pastures are highly nutritious and are eaten by cattle (Camarão et al. 1990).