A quick and accurate sampling method for determining yield in peach orchards could lead to better crop management decisions, more accurate insurance claim adjustment, and reduced expenses for the insurance industry. Given that sample size depends exclusively on the variability of the trees on the orchard, it is necessary to have a quick and objective way of assessing this variability. The aim of this study was to use remote sensing to detect the spatial variability within peach orchards and classify trees into homogeneous zones that constitute sampling strata to decrease sample size. Five mature peach orchards with different degrees of spatial variability were used. A regular grid of trees was established on each orchard, their trunk cross-sectional area (TCSA) was measured, and yield was measured as number of fruits/tree on the central tree of each one of them. Red Vegetation Index (RVI) was calculated from aerial images with 0.25 m·pixel−1
resolution, and used, either alone or in combination with TCSA, to delineate sampling strata using cluster fuzzy k-means
. Completely randomized (CRS) and stratified samplings were compared through 10,000 iterations, and the Minimum Sample Size required to obtain estimates of actual production for three quality levels of sampling was calculated in each case. The images allowed accurate determination of the number of trees, allowing a proper application of completely randomized sampling designs. Tree size and the canopy density estimated by means of multispectral indices are complementary parameters suitable for orchard stratification, decreasing the sample size required to determine fruit count up to 20–35% compared to completely randomized samples.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited