Precisão da estimativa da massa de forragem com discos medidores em pastagem natural

Abstract

The objective of this work was to evaluate precision of forage mass estimatings (FM) in natural pasture (NP) with discs. The treatments were three different discs areas, respectively 0.1; 0.2 and 0.3 m2 and each one combined with three weights of discs 5, 10 and 15 kg/m2. The experimental design was Completely Randomized Design with 50 replications, in an 3 x 3 factorial arrangement (3 disc areas x 3 weights of disc). The obtained results were submitted to regression analysis between height of disc and FM determined in each date of evaluation, area and weight of disc. From the mathematical models it were obtained the coefficient of residual variation (CV). Later, were adopted the analysis of variance method in Complete Blocks Design with seven replications of an factorial experiment (3 disc areas x 3 weights of disc) of periods. The relationship between three areas of discs associated with three weights and the CV of the measures gotten with discs were quadratic and positive. The disc area of 0.1 m2 and 5 weight kg/m2 presented smaller CV of the readings obtained with disc in the evaluated periods. With the increase the disc area, increased the CV for the weights of 5 and 10 kg/m2. The relations between the three weights and areas of discs and the CV were linear and positive. The smaller weight of disc of 5 kg/m2, associated with the area 0.1 m2 presented smaller CV. Weights added of 5 and 10 kg/m2 on the same disc area increased the CV in the evaluated periods. The relation between times of evaluation and the CV were linear and positive. The disc of smaller area, 0.1 m2 and 5 weight kg/m2, presented smaller CV for MF estimate in NP, therefore the most indicated to evaluate of MF in NP. The advance of the time of evaluation increased the CV of MF estimate in NP.