Abstract This study aims to determine the effect of nursery production time and container size on the quality standards of production of Simarouba amara seedlings based on nursery and field performance. The work was conducted in two phases: evaluation of nursery growth and field performance. A completely randomized design was used in both phases. Three container sizes (55, 180 and 280cm³) and four nursery production times (105, 125, 135 and 145 days) were tested. Height, collar diameter and survival were evaluated. S. amara seedlings can be produced in a 180 cm³ container with a maximum nursery production time of 125 days. The quality standard for seedlings corresponds to collar diameter ≥ 4.2 and height ≥ 9.1 cm.

Este trabalho procurou detectar a influência de efeitos de borda, espaciais e variáveis do substrato sobre as variações estruturais da comunidade arbórea de um pequeno (4,0 ha) fragmento florestal situado em Lavras, MG. Árvores com diâmetro à altura do peito (DAP) > 5 cm foram identificadas e medidas em 29 parcelas de 20 ×20 m, distribuídas sistematicamente por todo o fragmento. Variáveis relacionadas ao substrato foram obtidas para cada parcela a partir da topografia, análises laboratoriais de amostras do solo e medidas da compactação do solo. Um 'fator borda' foi calculado por parcela a partir da morfometria do fragmento. Variáveis espaciais foram derivadas das coordenadas cartesianas do centro das parcelas. Uma análise de correspondência canônica da distribuição da abundância das espécies indicou correlações significativas com algumas variáveis das matrizes ambiental e espacial. As primeiras foram classe de drenagem, cota média e saturação por bases, que explicaram 22,4% da variância total. As espaciais, x e x2*y, explicaram 12,7%, mas os dois conjuntos compartilharam, indistintamente, 4,6% da variância total. Portanto, descontadas as variações estocásticas e não explicadas (69,5%), a distribuição das espécies arbóreas no fragmento é primariamente influenciada pela distribuição espacial, status nutricional e regime de água dos solos.

The present study aimed at detecting the influence of both edge effects and space substrate-related variables on the structural variations of the tree community in a small (4.0 ha) forest fragment in Lavras, Minas Gerais state, SE Brazil. Trees with diameter at breast height (dbh) > 5 cm were identified and measured in 29 (20 ×20 m), systematically distributed plots over the fragment. Substrate-related variables were obtained for each plot from the topography, laboratory analyses of soil samples and soil compaction measures. An 'edge factor' was calculated per plot from the fragment morphometry. Spatial variables were derived from the Cartesian coordinates of plot center. A canonical correspondence analysis of the species distribution abundance indicated significant correlations with some variables of the environmental and spatial matrices. The first ones were drainage classes, mean elevation, and saturation of bases, which explained 22.4% of the total variance. The spatial variables, x and x2*y, explained 12.7%, but the two sets shared, indistinctly, 4.6% of the total variance. Therefore, apart from the unexplained and stochastic variations (69.5%), the species distribution on the fragment is primarily influenced by spatial proximity, soil nutritional status and ground water regime.