By Rhiann Jakubowski
A few weeks back, Luca Brillante gave a glimpse into the world of grapevine pruning in California, where the sun is shining, precipitation is rare, and t-shirts are prevalent. From 2,800 miles away, pruning season in the Lake Erie region looks (and feels) a little different. No matter the contrasts in climate, collecting pruning weight samples in both regions aims to achieve the same goal – crop load mapping.
Throughout the 2017 growing season, the Efficient Vineyard group in the Lake Erie region carried out several variable rate management trials, including early-season shoot thinning and mid-season crop thinning, as ways to manipulate crop load. While weekly berry sampling for berry weight, color, and juice soluble solids monitored fruit development during the season, taking pruning weight measurements in the dormant season allowed us to measure and map crop load, or the Ravaz Index (crop yield/pruning weight), across those experimental blocks.
The images included are from one of the mid-season fruit thinning vineyards in Westfield, NY. For this experiment, NDVI (Normalized Difference Vegetation Index) was taken four times pre-verasion using CropCircle proximal canopy sensors. The NDVI data was layered with soil EC (electronic conductivity) measurements to create three distinct vineyard management classes. Samples for crop estimation and berry sampling/pruning weight sampling were stratified throughout the three classes. At 30 days post-bloom, the grower collected crop estimation samples from the six sample locations, creating a predicted yield within each class. From those yield estimates, the grower decided to variable rate fruit thin this vineyard by targeting to remove two tons per acre (harvest weight) from each of the three classes. The harvester was equipped with variable rate technology, which allowed the grower to upload a prescription map, such as the one shown in Figure 1, into a field computer controlling the shake rate of the harvest rods. The harvester was also equipped with a grape yield monitor in order to collect the weight of the fruit removed during thinning.
Crop load maps were generated by translating the NDVI data into pruning weight based on in-field calibration measurements and calculating spatial crop load using yield monitor data collected at harvest. Figure 2 shows a map of crop load in 2017 if mid-season fruit thinning had not been performed. Without thinning, only 31.89% of the vineyard block would be balanced to slightly overcropped, while the remaining 68.11% of the vineyard would have been moderately to severely overcropped. Figure 3 presents the actual crop load of the vineyard in 2017, revealing 75.67% of the vineyard in the balanced to slightly overcropped range and 24.33% in the moderately to severely overcropped range. However, as Figure 3 shows, most of the severely overcropped areas of this vineyard reside in the control rows where no fruit was removed during thinning.
We know from previous published research that overcropping vines will lead to a loss of vine size and detrimental effects on fruit quality. In 2018, we will continue to collect weekly berry samples to monitor the effects of fruit thinning in year one on fruit quality and vine size in year two. Also planned for 2018 is collection of pruning weight samples in this dormant season followed by crop adjustment to achieve balanced crop load in the coming growing season using variable rate management.
Watch the video below to see a close up side-by-side look at the difference between ground speed and beater speed during crop estimation and fruit thinning.