CHAPTER 3 - VEGETATION STRUCTURE AS AN INDICATOR FOR LAND-COVER DYNAMICS ASSESSMENT IN THE AMAZON
3.6. Trends in research of tropical forest secondary succession
The findings and topics exposed in this chapter illustrate the importance of studying secondary succession in tropical environments. The research also identified many research initiatives regarding the integration of vegetation structure data and the use of satellite images in monitoring land-cover dynamics in the Amazon.
The rapid and aggressive regrowth of secondary vegetation in tropical areas has already been discussed by a number of researchers. The Amazon colonization produces widespread deforestation but also a mosaic of secondary successional vegetation stages. Different regeneration patterns occur depending on land management following deforestation. Although it is clear that secondary vegetation will not preserve the total biodiversity of mature forests, it is also clear that it plays an important role in the Amazonian landscape structure and function (Smith et al. 1997).
Perhaps one important question not covered by this chapter is the role of species composition within the different stages of regrowth. Studies have shown that disturbance from slash-and-burn agriculture affects species composition much more than stand structure and biomass (Uhl 1987). Vieira et al. (1996) have also pointed out that even after 40 years of recovery, richness is less than half of a primary forest. Although such an issue is of central relevance to the maintenance of local and regional biodiversity, we are far from being able to differentiate distinct tropical forest communities based on species composition when using satellite data. Current applications can only recognize different structural patterns and processes.
Research regarding these latter applications includes many new approaches. In terms of the availability of new sensors and data, optical and microwave data provide complementary information about land use and forest fragmentation. Besides overcoming the problem of cloud cover, the use of low frequency radar systems and its integration with other spatial and spectral data is promising for land-cover mapping in the Amazon (Rignot et al. 1997, Saatchi et al. 1997, Yanasse et al. 1997).
A second way of improving the extraction of earth surface feature information for LULC classifications in the Amazon is the use of state-of-the-art techniques for image processing and classification. Among others, spectral mixture analysis (Schweik 1995, Adams et al. 1995), spatial-spectral classifiers (Foody et al. 1996), spectral indices of canopy brightness (Steininger 1996) and GIS-informed classifications (Hinton 1996, Batistella 2000) are among the main trends to improve monitoring of secondary succession.
Recent questions have arisen about the importance of successional land covers to carbon sequestration (Fearnside and Guimarães 1996). Other studies have addressed the process of degradation of Amazonian forests (Nepstad et al. Flames 1999, Vieira et al. 1993). Activities such as selective logging have been responsible for the impoverishment of forests. Logging and fire increase forest vulnerability to future burning, 'potentially doubling net carbon emissions from regional land use during severe El Niño episodes' (Nepstad et al. Large-scale 1999).
The ecological functions of secondary forests at local and regional scales have just recently been investigated. Besides maintaining one-third of the native species, their role in carbon sequestration seems to be even more important (Vieira et al. 1996). Furthermore, successional vegetation re-evaporates an important part of the rainfall input in spite of the marked seasonal distribution of rainfall (Holscher et al. 1997). Possible regional climatic changes due to deforestation may be less severe in areas where woody secondary vegetation plays an important role in land cover.
At the local scale, the stages of secondary succession are directly associated with cycles of production and abandonment. Slash and burn clearing, cropping, and fallowing correspond to different phases within these cycles and depend on decision-making processes among the landowners. In this sense, monitoring the outcomes in terms of vegetation structure and LULC may also provide information about actions being taken by farmers. The study of LULC dynamics, and its fundamental importance to the understanding of general patterns of landscape transformation, is the goal of the next chapter.
