Variabilidade espacial do banco de sementes de plantas daninhas em área de cana-de-açúcar

Abstract

The phytosociological survey added with geostatistical techniques and network analysis can contribute to monitoring the seed bank, revealing the pattern of weed seed dispersion in sugarcane crops and influencing decision-making. Therefore, the hypothesis is that the distribution and/or dispersion of weed seeds present spatial variability as it is influenced by external factors through soil management resulting from agricultural mechanization in the period of sugarcane reform, the relief by its slope and by the área runnof. The objective of this work was to characterize the spatial variability of the distribution of weed seeds in two areas, respectively, upper slope and half slope, through a phytosociological survey where the parameters of similarity, frequency and seed density indices were evaluated of weeds in sugarcane cultivation area. The experiment was conducted in the city of Guariba, São Paulo, at Usina São Martinho, the soil is classified as eutroferric red latosol, with active production for more than 30 years, presenting a slight slope in its landscape. To characterize the spatial variability, a geostatistical analysis was performed by demarcating 55 points in each part of the total study area, which was divided into two, the upper slope area and the half slope area. At each point, ten sub-samples were extracted, totaling 550 sub-samples per area. With the data obtained, geostatistical analyzes were used with the representation of the variogram in the graphic description of the relationship between the distances the áreas observed under study and the average of these values between the áreas in the visualization of spatialization of the seed bank. The results showed that there is spatial dependence in the distribution of the seed bank for the community that presented the maximum reach with values of 28.29 m, on the upper slope and 13.53 m, on the middle slope, and that within this limit they presente continuous distribution in the formation from the weed seed bank, therefore, after these limits indicate that the values are independent from each other with no autocorrelation. The spatial pattern maps show, through the phytosociological survey and the geostatistical analysis, the seed distribution behavior along the areas. The use of schematic representation with the analysis of graphs and networks, through geostatistical analyzes of the sampled points, made it possible, through density, connectivity and proximity of the points, to observe the behavior of seed distribution in the areas under study, upper slope and half slope, revealing that the seed bank is in compliance with the soil surface flow. The use of geostatistical techniques, the phytosociological study and the analysis of networks complement each other and guide each other to identify weed communities and the distribution pattern along the sugarcane area.