Technische Universität Berlin
Land-use effects on plant biodiversity and carbon cycling in seasonally dry tropical forests in north-eastern Brazil
Abstract
dc:description.abstractOne of the most threatened ecosystems of Brazil is the so-called ‘Caatinga’. It is the seasonally dry tropical forest located in north-eastern Brazil. Due to the semi-arid climate, livestock husbandry (mainly goats) is the main economic sector. However, besides high stocking densities free roaming animals aggravate the degradation of this ecosystem. In areas near the Itaparica Reservoir, irrigation farming is conducted with perennial plants such as coconut, banana, guava, and mango and annual crops such as onion and melon being the most common crop species. Here, typical problems such as a high usage of agrochemicals and an improper irrigation, often leading to salinisation of the soils, occur. The present thesis analyses the effect of different land-use practices and intensities on the plant diversity and organic carbon stocks in an area of Caatinga seasonally dry tropical forest. The data assessment was carried out on 45 permanent study plots in the Caatinga dry forests (Itacuruba and Floresta) and on 21 agricultural areas in the irrigation schemes Apolônio Sales and Icó-Mandantes (Petrolândia). In the dry forests, the grazing intensity was quantified for each plot. The diversity of perennial plant species was recorded on a 20 m × 20 m plot (abundance). Herbaceous species were assessed on four subplots (each 2 m × 2 m) using the Braun-Blanquet cover-abundance scale. Further, the aboveground carbon stocks of woody plants (subplot, 10 m × 10 m) and herbaceous species (two subplots, each 2 m × 2 m) were assessed. For analysing soil organic carbon stocks, three samples of the topsoil and samples of two soil profiles were taken in each Caatinga plot. On the agricultural fields, two study plots (each 5 m × 5 m) were established. On these plots, vegetation relevés of weed species were carried out (cover-abundance scale according to Braun-Blanquet). Further, crop plants were measured to estimate the biomass per crop plant. Soil organic carbon stocks were determined by analysing two samples of the topsoil per study plot. The study results reveal that the present land use has detrimental effects on the plant diversity and organic carbon stocks of the Caatinga dry forests. In total, 140 plant taxa (including morphospecies) were found within the dry forest study plots. The species richness and the Shannon and Simpson indices of the woody vegetation were significantly higher on plots with no/low grazing intensity compared to the heavily grazed plots. The number of herbaceous species was also significantly higher on plots with no/low grazing. Aboveground carbon stocks were generally low (about 16 Mg ha-1) and the aboveground carbon stocks of herbaceous species were significantly reduced by grazing. Further, the recruitment of tree and shrub species (excluding the most abundant tree species) was significantly negatively affected by grazing. Moreover, high grazing pressure significantly reduced soil organic carbon stocks in the topsoil (7.02 Mg ha-1 vs. 3.39 Mg ha 1). The arable land showed a greater potential to sequester carbon (20.79 Mg ha-1) and a high weed diversity (77 species). Yet, the species assemblage showed only few similarities with the Caatinga dry forest plots. However, the data on crop yield imply a mostly significant positive relationship between crop yield and weed diversity. The present study presents the first comprehensive overview on the impact of the current land use on the vegetation and carbon stocks in an area of Caatinga in north-eastern Brazil. These novel insights are of crucial importance for the development of sustainable land management options. Based on the findings of this study, a set of recommendations was developed and incorporated into a guidance manual.
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Schulz, Katharina
Rights
- Licence dc:rights.uri
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- http://dx.doi.org/10.14279/depositonce-10483
- OAI identifier oai:identifier
- oai:depositonce.tu-berlin.de:11303/11594