Distribuição longitudinal e população de plantas de soja

Abstract

This work aims to evaluate the quality of soybean sowing in a commercial area without soil disturbance, by means of the evaluation of the population of plants and the longitudinal distribution of plants. The work was conducted in a commercial area, located in the municipality of Naviraí (MS). To evaluate the test, was collected the number and distance between seedlings emerged in two meters, in an area of 18.1 hectares (ha), totalizing 93 georeferenced sample points. Posteriorly, was inferred the “stand of plants”, “normal” spacings, “double” and “faulted”. The data were submitted to descriptive analysis and geostatistics. The population of soybean plants at sowing presented spatial variability of 7.84%, adjustment parameter considered within the recommended. The longitudinal distribution presented 67.98% below 50% of “normal” spacings, 81.28% above 20% of “faulted” spacings and 95.07% above 20% of “double” spacings, being classified as not satisfactory, through the recommended adjustment parameters. The coefficient of variation (CV) was classified as low for “stand of plants”, and for the “faulted”, “double” and “normal” spacings were classified as medium. With the geostatistical analysis, the “stand of plants” and the “double” spacing was adjusted to the exponential model, since the “faulted” and “normal” spacing was adjusted to the spherical model. The coefficient of determination (R2) shows that 88, 82, 76 and 55% of the “stand of plants”, “faulted”, “double” and “normal”, respectively, are explained by spatial variability, being that the error is due to the parameters not measured. Most of the distribution of “normal” spacings was below 50% and only in some points managed to reach close to 70%. There were no out of control points for the “normal” distribution, however the longitudinal distribution is not satisfactory, because it didn’t meet the minimum requirement for a pneumatic seeder (90%).