Precision Viticulture Symposium Launches the 2021 ASEV Conference
June 24, 2021
The American Society for Enology and Viticulture (ASEV) partnered with the National Grape Research Alliance (NGRA) to hold a virtual one-day symposium on precision viticulture on June 21, 2021. Moderated by Patty Skinkis, president of ASEV, and Donnell Brown, president of NGRA, the symposium took advantage of being held online to include speakers from Australia, France, New Zealand, California and six other states.
Dr. Rob Bramley, senior principal research scientist at Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia (and considered to be the “father precision viticulture”), noted in his keynote presentation on “Precision Viticulture and Beyond – Digital Approaches to Winegrowing” that precision viticulture is not new, as the first winegrape yield maps were produced in 1999. There has been considerable research and development since then that has tended to focus on the development of AgTech, rather than the application of AgTech to the production system as a whole or thinking about variability within the production system. He commented that as a consequence, we have a lot of solutions looking for problems, instead of a greater focus on the benefit/cost that the research had been looking at and how that might be adopted.
“The basic premise of precision viticulture is that land is variable,” Bramley commented. It can be flat, undulating or sloping. Some land variation is more subtle: for example, the soil profile can vary in water retention or in subsoil composition. Even though many vineyards are managed uniformly, that is often “a sub-optimal strategy,” as part of a vineyard may not perform as well as another that is more profitable, and the difference is in the soil.
Precision viticulture is based first on observation, with a primary source of information being yield maps, as well as remotely sensed imaging, digital elevation models, high resolution soil mapping, soil and tissue testing, and crop assessment. That data must then be evaluated, interpreted, and analyzed in order to come up with a targeted management plan. Variable rate application of fertilizer, irrigation water, mulch, sprays, selective harvesting, can be part of that plan. The whole process needs to be repeated over several years to verify that the data produced is accurate.
Australians have not had much interest in the variable rate input, according to Bramley. Europeans have used varying sprays and rates, while in California, Gallo has been a leader in variable irrigation. The cost for infrastructure to apply different amounts of water can be high and may deter some vineyard owners from installing variable rate systems. Australians have also been concerned about the costs for pumping water.
Variable rate outputs such as selective harvesting and product streaming is “a key area of opportunity across a range of production scales,” Bramley said. If the market opportunity exists, selective harvesting can be very profitable, especially when a product can be promoted as distinctive.
If growers harvest high quality blocks of fruit with similar blocks, they can get the benefit of selling those grapes at a higher price, and the winemaker can sell the resulting wine for more money. He gave one example where a vineyard with a 66-ton potential yield was mapped by remote imaging, which was validated by ground truthing. The mapping information divided the vineyard into four different quality levels for the grapes. Instead of selling the 66 tons of grapes at $1,600 for a total of $105,600, the vineyard was separated into four quality levels: 15 tons at $1,300/ton, 31 tons at $1,600/ton, 5 tons at $3,500/ton. and 15 tons for $5,000. The total the grower received rose to $161,600, while the winery had their wine value increase by $470,000.
Bramley thinks that there is no doubt that terroir is a fact, not fiction. He looked at the spatial variability of one compound, rotundone (which makes some cool climate Shiraz wines taste peppery). They sampled a large number of vines and generated maps of the concentration of rotundone over three different vintages spread over a four-year time period. The patterns of variation in this compound were quite stable in time and seemed to be closely related to soil properties, including slope, particularly the aspect of that slope as expressed in degrees from North. They could then identify areas of higher rotundone concentration associated with soils of low molecular soil conductivity and a larger orientation away from North.
The Shiraz from the areas with high levels of rotundone sells for $120/bottle, while the wine from areas with lesser amounts sells for $30/bottle. Examples such as this one suggest that there is nothing mythological about terroir. There is an opportunity in precision viticulture to understand of terroir and use that understanding to improve the management of the wine production system.
In Australia, fruit is graded pre-harvest and assigned to a product stream. Payment is made on the basis of assessed grade, often with a later adjustment depending on the ultimate end-use. In California’s Central Valley, a price per ton is paid based on variety and location. If some grapes get allocated to a higher grade at the winery, there is no benefit to the grower. Bramley asked, “Are both grape growers and winemakers in California missing an opportunity?”
Bramley went on to identify some key enablers for profitable digital viiiculture, including:
Prediction of yield, especially early in the season, which helps with planning and management of harvest, as well as winery and market logistics. Various yield prediction models have varying utility, and scale of estimation is a key issue. Other sensor-based solutions are under development.
- Yield monitoring, while available on harvesters, is not widely adopted.
- More work needs to be done on the prediction of fruit quality and composition.
- Sensing of fruit quality and composition is a major research and development need by growers.
- Simple tools for data analysis have been somewhat ignored.
Wine is a highly differentiated product, in Bramley’s opinion, and as a result, selective harvesting is an appropriate area of focus. However, the key aspect underlying all of this is data, and if a winery isn’t “set up to collect and make use of data, then I’d suggest you’re behind the eight-ball.”
He noted that digital opportunities are not confined to the within-vineyard scale, but also can be applied to properties and regions. There can be a wide variation in elevation and season degree days within a 66-hectare property. The difference in degree days can affect the date when one is ready for harvest and another is not, sometimes by as much as six or even 11 days.
In Australia, researchers have done work on a regional basis to see, for example, if yield estimation in one place could inform an estimate in another. They now have regional-scale yield maps (kg/meter) based on several years of data. In spite of the seasons having very different conditions and therefore different yields, the patterns of variation were very stable across those five seasons. Researchers then clustered the data together and identify some areas of characteristic performance and the patterns of different data were very stable and an example of regional terroir.
In closing, Bramley posed a question: “Suppose a winemaker wins a trophy for a wine made from a 20 ton parcel harvested from within a block on a night when 100 tons were harvested using more than one machine, and that the prize was awarded because the judges considered the wine to ‘reflect the terroir of the region.’ What should the winemaker do next year? How should they ensure that the gravelly and silty soils are reflected in the same way in next year’s potentially prize winning parcel (and do they need to do that?), given that the precise location of the prize winning parcel was not recorded and there is a 30% difference in predicted yield between the two years?”
He suggested that if grower or winemaker is not adopting any of the digital technologies that he has spoken about, he or she couldn’t answer those questions. However, if they are adopting precision viticulture technologies, they could have a pretty good chance of answering all of them and therefore get more control over your production system.
Bramley noted that increasingly, agriculture is a data-driven activity, and the wine sector (and other grape sectors) already collect a lot of data. His final question was: Are we making the best use of that data?
The symposium is followed by the 2021 ASEV National Conference on June 22-24. Because the conference is being held virtually, there will be speakers from California and seven other states as well as from Australia, Canada, China, and Germany.