Design and Evaluation of a Yield Monitoring System for Pistachios

Abstract

Information on the yield of individual trees within an orchard can be used to improve management practices. To achieve this goal, an automated yield monitoring system has been developed that is capable of correctly identifying trees in a commercial orchard and measuring yields of pistachio trees with small errors. Two commercial pistachio catch-frame harvesters were retrofitted and operated at approximately 75% of commercial harvesting field capacity. The system consists of a weighing bucket that measures individual tree yields in the transition between harvested trees. When the subsequent tree is shaken, nuts are weighed, dumped into the harvester's internal conveyor, and an embedded computer automatically records tree identification and yield. A DGPS-based automated row identification system was developed and tested throughout the 2006-2008 seasons. Inside dense orchard canopies, where a DGPS signal cannot be received with reliable accuracy, an algorithm was developed using real-time sensed odometric information to localize trees. System performance was tested in a commercial orchard by measuring yield and location of more than 70,000 trees over a six-year period. Tree identification and yield data were collected and mapped. Calibrations and validations of the weighing system resulted in weighing prediction standard errors of approximately 0.9 kg, while DGPS row identification was 100% accurate in over 480 estimates, and individual tree identification was highly accurate over 70,000 observations. The ability to measure individual tree yield provides previously unattainable information on tree productivity, facilitates precision field management, and enables the investigation of the mechanisms underlying fruit production and alternate bearing. The data recorded by the use of this system have been used in experimental studies to improve the productivity and management of pistachio orchards, investigate alternate bearing behavior, and provide the basis for developing efficient nutrient management protocols.