Fitoecologia e avaliação de técnicas de restauração na transição savana-floresta decidual sobre gradiente topográfico na Serra da Bodoquena

Abstract

The Brazilian deciduous tropical forests are restricted to increasingly smaller fragments, threatened by human activities, and many of their peculiarities are still little known. On the western border of the legal boundary between the Cerrado and the Pantanal, the Serra da Bodoquena National Park (PARNA) protects an area of 77,021.58 ha distributed in two elongated fragments in the north-south direction and, of this total, 91% corresponds to the Submontane Deciduous Seasonal Forest type, with a connection between the forest edges and the savannas (Cerrado). The substrate of limestone rocks occurs in a large part of the terrain, especially in the topographic elevations, between 300 m and 800 m altitude. All these characteristics make this vegetation even more special, and, for this region, this implies great ecological, social, and economic relevance. Within the PARNA, the area known as "sink" or "headwaters" of the Perdido River has areas with brachiaria pastures, abandoned after the implementation of legal protection, and have shown insufficient resilience for natural regeneration to occur, requiring intervention for ecological restoration. This thesis was proposed with the objectives of knowing the composition and structure of the shrub and tree community of this deciduous forest on an altitudinal gradient of karst relief, surveying its ecological characteristics, and evaluating two low-cost restoration techniques on this pasture, aiming to meet the needs of the region. The research generated three articles, which are organized as follows for better systematization of the information and knowledge produced here. The raw data underwent statistical treatments, uni and multivariate, to test our hypotheses. In the first paper, the composition, structure, and diversity of the shrub-woody component of a 1.02 ha plot of this forest was surveyed and evaluated, following the altitudinal gradient. To understand the effects of topography on the vegetation, three positions on the relief were sampled, one in the highest part (top), one in the intermediate part (slope), and another in the lower part (base). As a result, in the total area 2,385 individuals belonging to 70 species, 49 genera and 29 families were sampled. Density increased towards the lower part of the slope. Diversity and equability were higher at the top, followed by the bottom. Diameters were concentrated in the first class, up to 13.1 cm, and the average height was 5.25 m, with a predominance of understory. Some species sampled had not yet been recorded for the region. The exploratory ordination analysis revealed greater distance of the hilltop plots and species compared to the plots and species of the two hillside areas. It was verified at this stage that topography influenced the observed floristic-structural heterogeneity. In the second paper, we followed the dynamics of emergences in the seed bank in beds with and without topsoil transposition of the sampled forest, following the same three topographic positions, under two degrees of shading and full sun. In the total area, 1877 individuals of 82 species and 45 families were observed and identified. The topographic position generated great floristic heterogeneity even though the areas were very close, with similarity of only 20% of the base with the other areas. Diversity was high in all topographic ranges, with H' 3.45 and J' 0.78 in the total area. The number of recruitments was higher on the slope. Shading, especially at the 50% level, positively influenced abundance and species richness in all areas. The addition of topsoil allowed for a 68.3% increase in species, compared to the soil bank present in the pasture, and was responsible for 89% of seedling abundance. There was ecological heterogeneity along the gradient, and shrub/ tree (67.3%), pioneer (58.6%) and zoochoric (53.9%) species predominated overall. The technique of topsoil transposition showed promise for forest restoration over brachiaria grass pasture, especially if associated with shading and if the characteristics of the species are respected as to their environmental requirements. The third paper evaluated the potential use of 15 tree species from the savanna-deciduous forest transition in direct seeding and the consequences of interactions between abiotic (topography, shading, topsoil and observation time) and biotic (potential for plant emergence and establishment) variables in six treatments, under three positions on the relief. Additionally, the effects of precipitation and herbivory on the dynamics were analyzed. For planting in rows, 720 seeds of each species were used. Emergence rates ranged from 0.5% to 50.5%, and total mortality at the end of the evaluation was 45%. Observation time and shading influenced emergence, but the addition of topsoil in the beds did not show important results at this stage of development. Precipitation had a positive correlation with emergence, while mortality had a positive correlation with herbivory. The species had varied behavior in front of the treatments, but with intra-taxonomic similarities. Topography did not significantly affect the composition and abundance of the set of species analyzed. Environmental preferences and herbivory served as ecological filters for the resulting community. The results of this work reinforce the premise of high variation in the biodiversity of tropical forests and savannas as a consequence of the factors that affect them. Ecological restoration associated with the knowledge of biodiversity is an important and necessary tool to reestablish the ecological processes of disturbed areas, because it assists in the resumption of ecosystem services and, subsequently, in the conservation and maintenance of environmental balance, benefiting, including, human productive activities.