Rock outcrop communities usually receive very little attention from scientists and environmentalists. We examined the vegetation occurring in eight gneiss-granite rock outcrops at Rio de Janeiro State (Brazilian Atlantic coast) which exists in natural associations on soil islands. A total of 86 vascular plant species, belonging to 30 families, was found on 347 soil islands. Bromeliaceae, Asteraceae and Velloziaceae species were the most frequent plants, many of them endemic to these habitats. Ordination and cluster analyses using species frequency on each site made evident some major distinctions related to local influences, most probably the proximity to the sea. Each outcrop presented high values of the Shannon-Wiener index of species diversity. Species richness was very dependent on the total area, and high beta diversity was observed amongst sites. Similarities with the South American and African rock-outcrop communities were found. Despite their uniqueness as habitats, their possession of several endemic species and the fragility of the ecosystem involved, Brazilian rock outcrops are not protected by specific environmental legislation and we propose urgent actions for their protection.

Activities involving fauna monitoring are usually limited by the lack of resources; therefore, the choice of a proper and efficient methodology is fundamental to maximize the cost-benefit ratio. Both direct and indirect methods can be used to survey mammals, but the latter are preferred due to the difficulty to come in sight of and/or to capture the individuals, besides being cheaper. We compared the performance of two methods to survey medium and large-sized mammal: track plot recording and camera trapping, and their costs were assessed. At Jataí Ecological Station (S21°31'15"-W47°34'42"-Brazil) we installed ten camera traps along a dirt road directly in front of ten track plots, and monitored them for 10 days. We cleaned the plots, adjusted the cameras, and noted down the recorded species daily. Records taken by both methods showed they sample the local richness in different ways (Wilcoxon, T= 231; p;;0.01). The track plot method performed better on registering individuals whereas camera trapping provided records which permitted more accurate species identification. The type of infra-red sensor camera used showed a strong bias towards individual body mass (R 2 =0.70; p= 0.017), and the variable expenses of this method in a 10-day survey were estimated about 2.04 times higher compared to track plot method; however, in a long run camera trapping becomes cheaper than track plot recording. Concluding, track plot recording is good enough for quick surveys under a limited budget, and camera trapping is best for precise species identification and the investigation of species details, performing better for large animals. When used together, these methods can be complementary.

The effects of the contents and chemical composition of the foliar epicuticular waxes of species from the caatinga (Aspidosperma pyrifolium, Capparis yco, Maytenus rigida and Ziziphus joazeiro) and cerrado (Aristolochia esperanzae, Didymopanax vinosum, Strychnos pseudoquina and Tocoyena formosa) were evaluated as to the resistance to water loss by means of an experimental device constructed for this purpose. In general, the waxes of the caatinga species investigated were more efficient against water loss than cerrado species. Increase of the thickness of the waxy deposits from 40 to 90µ g.cm −2 had no significant effect on the resistance to water loss. The chemistry of the wax constituents was shown to be an important factor to determine the degree of resistance to evaporation. n-Alkanes and alcoholic triterpenes were the most efficient barriers, while hentriacontan-16-one (a ketone) and ursolic acid (an acid triterpene) revealed low efficiency. The higher efficiency of the waxes of the leaves from caatinga species (mainly those of C. yco and Z. joazeiro) is probably accounted for the predominance of n-alkanes in their composition. The lower efficiency of the waxes of A. pyrifolium (caatinga), T. formosa and A. esperanzae (both species from the cerrado) is probably a consequence of the predominance of triterpenoids in the waxes of the two former species and hentriacontan-16-one in the latter.