/ Practically no information exists on the impact of human trampling on tropical rain forest vegetation. We studied three trails with varying periods of use and recovery in a tropical rain forest in Costa Rica. Human impact on trailside plants was curvilinearly related to use, as found by other workers in temperate zone vegetation. Recovery in a period of two years and eight months had been rapid, and herbs and seedlings were more abundant along the recovering trail than in undisturbed forest. The results imply that a shifting mosaic of trails, analogous to the mosaic created by light gaps, may be the best management technique to minimize the impact of human visitors in tropical rain forests.Although the literature on the impact of truman trampling on the vegetation and soils along trails is substantial (Kuss and Graefe 1985, Kuss 1986, Cole and Schreiner 1981, none of the hundreds of studies deal with tropical rain forests. Nearly all the published papers were done in the temperate zone, especially in the western United States. Of 348 papers on the subject reviewed by Cole and Schreiner in 1981, every single one came from a north-temperate-zone country, and since then only two have been added from tropical or subtropical regions: the study of Liddle and Thyer (1986) in subtropical eucalyptus forest in Australia, and the work of Jim (1987) on the parks of Hong Kong.With the development of ecotourism leading to in-

O cultivo do café Conilon no extremo sul da Bahia vem se expandindo consideravelmente, estimando-se que existam mais de 20 milhões de cafeeiros plantados em solos de tabuleiros coesos. A presença de horizontes coesos (adensados) no perfil do solo afeta as relações entre drenagem, teor de água disponível, aeração, temperatura, penetração radicular e absorção de nutrientes, com reflexos negativos na produção agrícola. Desta maneira, objetivouse estudar os efeitos do preparo do solo sob diferentes regimes hídricos num Argissolo Amarelo Distrocoeso e os efeitos na cultura do café Conilon, no município de Eunápolis, BA. Os tratamentos foram: 1-Subsolador com uma haste (atingindo de 60 a 70 cm de profundidade) na linha; 2-Grade aradora + sulcador + subsolador com três hastes (atingindo 30 a 40 cm de profundidade) na linha; 3-Grade aradora + sulcador na linha; e 4-Subsolador com três hastes na linha. As plantas nestes tratamentos foram submetidos a dois regimes hídricos (com e sem irrigação). As análises dos dados de crescimento do café Conilon mostram que houve maior desenvolvimento no diâmetro do caule das plantas (p < 0,05) com o emprego do subsolador de três hastes comparado ao uso do subsolador de uma haste no preparo do solo na área irrigada. Também, observou-se maior crescimento nas copas das plantas com a utilização da grade aradora + sulcador + subsolador com três hastes, em comparação com o uso do subsolador de uma haste no manejo do solo na área irrigada. Na fase produtiva, a análise dos dados não evidenciou efeitos significativos (p < 0,05) entre os diferentes preparos do solo, nem entre a área irrigada e a não irrigada.Palavras-chave: café, solo coeso, subsolagem, irrigação, física e química do solo.Key words: coffee, cohesive soil, subsoiling, irrigation, soil physics and chemistry.Effects of different soil preparations and watering on edaphic attributes, growth and production of Coffea canephora cv. conilon in Argissolo Amarelo Distrocoeso (ultisol). The cultivation of Conilon coffee in South Bahia, Brazil is expanding considerably, and it is estimated that more than 20 million coffee trees are planted in cohesive soils at the tableland Brazilian ecosystem. The presence of cohesive horizons (hardened) in the soil profile affects the relationships among drainage, water availability, aeration, temperature, root penetration and absorption of nutrients with negative reflexes on the agricultural production. This work was aimed to study the effects of soil preparation under different hydric regimes in an Argissolo Amarelo Distrocoeso (cohesive Ultisol) and the effects on coffee cultivation, in the municipality of Eunápolis, Bahia, Brazil. The treatments were: 1 -Subsoiler with one leg (reaching from 60 to 70 cm depth) in the line; 2 -Disc plow harrow + furrow + subsoiler with three legs (reaching 30 to 40 cm depth) in the line; 3 -Disc plow harrow + furrow in the line; 4 -Subsoiler with three legs in the line. Plants in these treatments were submitted to two watering regimes (with and without irrigation). The ...