In order to identify the effect of P. arachnoideum, we studied 11 native tree species commonly used in reforestation projects. Bioassays were conducted in laboratory to evaluate the effect of bracken leachate on the germination and morphology of seedlings. Juveniles of some of these species were planted in two adjacent but contrasting areas in relation to the dominance of P. arachnoideum. The evaluation of growth and survivorship was performed after six and twelve months. This study reveals that for some pioneer and secondary trees P. arachnoideum leachate exerted an inhibitory effect on seed germination and seedling morphology. Field experiments revealed that pioneers are apparently more resistant to P. arachnoideum leachate than secondary species.Keywords: invasive species, Pteridium arachnoideum, tropical trees, bracken, germination, seedlings. Interferência de Pteridium arachnoideum (Kaulf.) Maxon. (Dennstaedtiaceae) no estabelecimento de árvores tropicais ResumoNa tentativa de identificar o efeito de P. arachnoideum, estudamos 11 espécies nativas comumente utilizadas em projetos de reflorestamento. Bioensaios foram conduzidos em laboratório para avaliar o efeito do extrato aquoso do samambaião sobre a germinação e morfologia de plântulas. Plantas jovens de algumas destas espécies foram plantadas em duas áreas adjacentes, porém contrastantes em relação à dominância de P. arachnoideum. A avaliação do crescimento e sobrevivência foi realizada após seis e doze meses. Este estudo revela que, para algumas espécies pioneiras e secundárias tropicais, P. arachnoideum exerce um efeito inibidor sobre a germinação de sementes e a morfologia de plântulas. Experimentos de campo mostram que as espécies pioneiras são aparentemente mais resistentes ao P. arachnoideum do que as espécies secundárias.

Studies of allometry are important in explaining effects of fire and herbivory, for estimating biomass in forests, and so on. There has been extensive research on plant allometry in temperate and tropical forests, showing that plant architecture often adjusts to the elastic similarity model, but not in Brazilian savannas (cerrado). We studied allometry of Dalbergia miscolobium, Diospyros hispida, Erythroxylum suberosum, Miconia albicans, M. ligustroides, Schefflera vinosa, and Xylopia aromatica in a cerrado sensu stricto area that was affected by a fire in August 2006. We expected that the study species would not adjust to any of the allometric models considered common for forest species ("constant stress", "elastic similarity", and "geometric growth"), and that there would be differences in allometry in burnt and unburnt patches. We sampled two species in 60 5 × 5 m contiguous plots placed in three transects, and five species in 100 5 × 5 m contiguous plots placed in five transects, where we measured the diameters at soil level (DSL) and the heights of all shoots. We used standardized major axis regressions on log-transformed data. The regression slope between the height and DSL was higher than 1.0 (p< 0.05) for four species, showing a greater height than would be expected under geometric growth, not predicted by theoretical models. We found significant differences (p < 0.05) in regression slopes and/or correlation coefficients between burnt and unburnt plots for five species, indicating that fire may influence plant allometry in the Brazilian cerrado, and that such a response is highly variable between species.

Introduced species can establish self-replacing populations in introduction sites despite climatic differences between these sites and the native range. Schizolobium parahyba (Fabaceae) is a fast-growing tree from the Atlantic Forest in Brazil and is invasive in semideciduous forests. We studied the allometry and population structure of S. parahyba within its native and invaded ranges. We expected to find differences in allometric relationships and ontogenetic stage sizes between forest types. We measured individuals within native and invaded ranges and analysed height : diameter allometry using standardised major axis regression. We used the chi-squared test and permutational multivariate analysis of variance to compare the frequencies of individuals of each ontogenetic stage between native and invasive populations. We found that allometry differed between forest types for seedlings, but only marginally for adults. The population structure differed between ranges and suggests active recruitment of juvenile and subadult plants in invaded sites, where the species is naturalising.