Search

1. The effects of proximity to an exposed forest edge on plants in Amazonian lowland rainforest were assessed by examining carbon isotopic composition ([variation].sup.13]C) of canopy leaves and of understorey leaves and air. 2. Exposed canopy leaves of Licania heteromorpha and Scleronema micranthum, were collected (a) in continuous forest, (b) within 100 m of a 5-year-old edge, and (c) in isolated 1-ha forest fragments. Leaves of the understorey species, Duguetia aff. flagellaris and Astrocaryum sociale, were sampled at intervals along transects from a 5-year-old edge towards the forest interior and in control areas. Leaves of D. flagellaris were also collected from young gaps, gap edges, old gaps, undisturbed understorey and from secondary vegetation. [variation].sup.13]C and [CO.sub.2] concentration of unde air were measured along the study transects. The proportion of the transect area occupied by gaps was estimated. 3. The isotopic composition of the carbon in canopy leaves did not differ among the three types of sampling areas. However, D. flagellaris had significantly higher [variation].sup.13]C within 60 m of the edge than in undisturbed control areas. [variation].sup.13]C air was elevated up to 40m from the forest edge (-9.3%, cf. -10.5% in control areas). This difference accounted for 87% of th elevation of foliar [variation].sup.13]C in D. flagellaris near the edge. 4. Gaps occupied almost 90% of the transect area within 60 m of the edge as compared to c. 14% in the control areas. Foliar [variation].sup.13]C of D. flagellaris growing in gaps was significantly less negative than in undisturbed understorey, old gaps and gap edges, which had similar values. 5. The frequency of gaps near the edge is probably responsible for the extent of the elevation in foliar [variation].sup.13]C . While the exact nature of edge effects in tropical forest remains complex, it is clear that at least the outer 60m of a forest patch must be considered 'edge-affected'. Key-words: carbon isotope discrimination, [CO.sub.2], gaps, plant--water relations, tropical forest fragments

In a montane rain forest at [is nearly equal] 2500 m in Venezuela we fertilized plots with N, with P, or with N and P for 4.5 yr. We measured trunk growth in circumference in control plots and in those fertilized with N and P, and measured litterfall in all four treatments. There were five 12 x 12 m plots per treatment. N was applied at 225 kg.[ha.sup.-1.yr.sup.-1] and P at 75 [kg.hr.sup.-1.yr.sup.-1] for the first 2 yr, and subsequently N was applied at 150 [kg.hr.sup.-1.yr.sup.-1] and P at 50 [kg.ha.sup.-1.yr.sup.-1]. Rates of trunk growth in N and P fertilized plots were roughly double those in control plots (1.4 vs. 0.6 mm/yr). Total litterfall average 4.3 [Mg.ha.sup.-1.yr.sup.-1] in the control plot. In the 4th yr after fertilization was significantly higher in plots fertilized with N + P (1.50 [g.m.sup.-2.d.sup.-1]) than in control plots (1.00 [g.m.sup.-2.d.sup.-1). There were no significant differences in rates of litterfall between treatments during the first 3 yr after the first fertilization. The changes of rates litterfall with time were significantly different between treatments. Litterfall in control fell by 9% per year, that in P-fertilized plots fell by 5% per year, and litterfall in N- and in N + P-fertilized plots remained approximately constant. Nitrogen concentrations in litterfall were unaffected by fertilization with N or with P. Phosphorus concentrations were significantly increased by P fertilization. We conclude that N is more limiting than P in this forest. Key words: fertilizers; forest; litterfall; nitrogen phosphorus; tropical; trunk growth.