ANÁLISE MULTIFRACTAL DE PLANTAS-DANINHASEM SISTEMA DE SEMEADURA DIRETA

Abstract

Weeds show high spatial variability, exhibiting infestation behavior in patches or patches, with heterogeneity at scales that can be evaluated through multifractal analysis. Thus, the objective was to evaluate the spatial variability of weeds through multifractal analysis in direct seeding system. The specific objectives were: to investigate the dominance of weeds in different climatic seasons; evaluate the dynamics of relationships between weeds and crops; characterize the ecology (richness and abundance) of weed species in a direct seeding area, through multifractal analysis. Sampling was carried out at 1,015 sampling points in a regular 5 × 5 m mesh (2.38 ha), and at 1,071 sampling points in a regular 10 × 10 m mesh, in an experimental plot with direct seeding in Campinas (São Paulo, Brazil) and Mata Roma (Maranhão, Brazil), respectively. In the municipality of Campinas, a predominance of Raphanus raphanistrum was identified in the winter harvest, and of Commelina ssp. in summer cultivation. The uniqueness spectrum showed greater asymmetry for Raphanus raphanistrum and Commelina ssp., in relation to the category of other weeds. In the municipality of Mata Roma, the weed Commelina benghanlensis L. was the species that presented the greatest heterogeneity mea singularity spectrum (AI = 2.324). Thus, in this study it was demonstrated that weeds present high spatial variability, with behavior in patches or patches and heterogeneity at different scales, which can be evaluated through multifractal analysis. Furthermore, the present study demonstrates the importance of multifractal analysis to understand the spatial variability of weeds, providing valuable insights for the efficient management of these species in areas conducted under direct seeding systems.