ASHS 2015 Annual Conference

Amazonian soils have recently been exposed to erosion through construction, extractive industries, and agriculture. Due to low fertility, heavy rainfall, and repeated disruption, revegetation of these sites can be challenging. The Amazonian Highway BR-319 from Manaus to Porto Velho was opened in 1973. Over much of its length, the construction of the highway exposed subsoil and parent material that failed to support spontaneous vegetation and the parent material was extremely vulnerable to erosion. At the initiative of the Brazilian national highway department, then DNER, native and introduced grasses and legumes were evaluated for their ability to thrive under the local conditions with a minimum of fertilization and liming and without maintenance. The trials were part of a larger DNER supported revegetation research program conducted by the IRI Research Institute, Matão, São Paulo, from 1975-1978. In one trial a total of 209 lines of 149 species were planted. Of these only 57 lines were considered usefully vigorous for maintaining soil stability after three years. Only 35 lines persisted after eight years. After 37 years of intermittent grazing and other disturbances, only 7 species remained. A second, replicated trial using 15 grasses and 9 legumes was planted in 1976 using a Latin square experimental design with three replicates. After two and a half years, 13 grasses and 3 legumes were considered useful for revegetation, and after eight years 10 grasses and 3 legumes persisted. After 36 years with little to no further disturbance, only three of the planted species, two grasses and one legume, could be found at the site, and only Urochloa brizantha (palisade grass, 'Brachiaria brizantha') covered more than its original planted area. Both experimental sites retained vegetation cover and have been succeeded by mixtures of plant species shaped by the management of the sites. Given the observations over the past 37 years, the criteria for desirable revegetation species should be reconsidered based on the ability of the plants to vigorously protect the soil and then give way to other species through natural plant succession. There are few signs of erosion in the planted areas, and soil formation is occurring at both trial sites. Meanwhile, extreme rates of erosion have occurred outside the boundaries of the planted areas.