Índice de área foliar em caatinga preservada: avaliação in-situ e através da resposta espectral da vegetação

Abstract

Studies about the Caatinga biome constitute a challenge for the Brazilian scientific knowledge, for various reasons, including the fact that the Caatinga is limited to the national territory. The biome has great physiological variation due to drought characteristics of semiarid region. Therefore, the knowledge of this variation may be important, for example, the understanding of its hydrological processes, since the vegetation has direct participation in the hydrological cycle. An important parameter of the vegetation to be studied is the leaf area index. This index is important for the understanding of biological and physical processes of vegetation on the Earth surface and can be evaluated by direct or indirect methods. Among the direct methods, the most suitable for deciduous vegetation, such as Caatinga, is the litter collection method. Among the indirect methods, we highlight the spectral vegetation indices derived from remote sensing, which have become an important tool to help environmental monitoring. The aim of this study was to evaluate the leaf area index, in a Caatinga preserved area, measured in situ by litter collection; and to assess its correlation with spectral vegetation indices. These ratios derived from images of OLI sensor Landsat8 satellite were NDVI, SAVI, EVI, EVI2 and ARVI. The study was conducted at the Aiuaba Experimental Basin - AEB with 12 km² area, whose predominant vegetation is the preserved Caatinga. AEB was divided into three soil-vegetation associations (SVAs). The results show that it is possible to estimate the leaf area index in the Caatinga, using the litter method. The average Caatinga LAI ranged between 1.8 and 3.5 m2.m-2 in the three SAVs. These values are influenced by the soil type and the annual distribution of rainfall. The high LAI reduction rates occurred between May and July 2014 with a reduction of 90%, 89% and 95% between June and July 2015 with a reduction of 63%, 62% and 72% in SAV1, SVA2 and SVA3, respectively. It was possible to estimate the LAI from spectral vegetation indices, since the statistical coefficients used to calibrate and validate the models showed good values. The good results are possibly due to the fact the Caatinga does not reach large values of LAI, making it impossible to saturate the IEVs, indicating the feasibility of the use of remote sensing for monitoring IAF in the Caatinga.