Greetings Fodder Folks, In early April I participated in the "National Viticulture and Enology Extension Leadership Conference" (NVEELC) which was held at the UC Oakville Station in the lovely Napa Valley.  The conference is an opportunity for viticulture extension folks from around the country to share industry news and ideas for working together, and includes a day of professional development. 44 extension-related viticulture and enology professionals participated in the 2 day conference.  

The meeting highlight for me was a tour co-organized by Rhonda Smith, UCCE-

Sonoma County Viticulture Advisor, and our recently hired UCCE Viticulture Specialist Dr. Kaan Kurtural. Rhonda gave the group an introduction to the E&J Gallo Frei Ranch, where the Kurtural team is collaborating with Gallo on a unique project using "proximal sensing" to manage variability in vineyards. Kaan's specialty is precision vineyard management, and a part of that is integrating sensors to detect vineyard variability-and then developing a means to address that variability to give a desired outcome, such as a uniform crop. 

Dr. Luca Brillante, a post-doctoral scientist in the Kurtural lab, helped to demonstrate the proximal sensing project he is co-leading.  The trial site has soil variability that follows a slope (like many foothill vineyards also do), resulting in non-uniform available soil moisture and, following, non-uniform fruit maturation and yield. Gallo now maps all of it's vineyard site soils using magnetic resonance-measuring up to 75 different soil variability aspects (water holding capacity, pH, etc.).  Using the mapped soil variability information as a base, the team employs an NDVI sensor that picks up the amount and tone of canopy "greenness"and a rapidly firing camera that counts the number of

Luca Brillante demonstrates the ATV mounted NDVI and GPS-camera sensors.

berries and clusters. Using all of this information, coupled with crop modeling, the team is creating a map of developing cluster zones for harvest.  The goal is an even crop load, and since the grower doesn't want to drop any crop, crop load will be managed by manipulating the cover crop using chemical mowing at different times.  

With all of the UAV (i.e. drone) and sensor craze lately, NDVI mapping may not seem like such a big deal. But Kurtural warned of the lack of "ground truthing" with NDVI flyover information-a critical component of understanding NDVI generated data.  Another catch is that the sensors need to be "trained", that is, the scale for which the data they are generating needs to be calibrated to be meaningful.  For the camera that is generating crop load information, occlusions are another hurtle for the Kurtural team. Counting clusters is not a good enough predictor of yield, rather,

number of berries per m of row explains 92% of the yield curve.  The Team performs various degrees of leaf removal and runs the camera across the fruiting zone to train it to pick up berry data.

Additionally, the Kurtural lab is studying the genetic pathway for anthocyanin (color) development in berries. With this information, the timing of applied water stress can be targeted for key times in color development. This is also important for the table grape industry, where similar studies are being conducted. 

The wave of the future? Clearly, "smart" farming is nearly here. UCCE is on the cutting edge of technology to manage our precious resources, save labor, and produce a premium crop.