Big Data and the Productivity Challenge for Wine Grapes Nick Dokoozlian Agricultural Outlook Forum February 2016 0
Big Data and the Productivity Challenge for Wine Grapes Outline Current production challenges Lessons learned from annual crops How will we utilize Big Data to meet our challenges? Measure Model Manage Summary 1
The Productivity Challenge for Wine Grapes Suitable land, labor and water for agriculture are becoming more scarce and expensive Need to increase grape supply without increasing production area and environmental impact Must increase both yield and quality simultaneously Similar challenges are faced by nearly all agricultural commodities worldwide 2
How are annual crops addressing these challenges? Dramatic increases in the productivity of agronomic crops have been achieved during the past century via: Genetics traditional breeding and genomics Improved agronomic practices and resource management Application of remote sensing and other technologies 3
How are perennial crops different in their approach? Progress has been much slower in wine grapes and other perennial crops: Critical mass limited acres = limited attention despite farm-gate value Genetics research, breeding cycle and market tradition Production cycle and innovation adoption Yield quality relationships 4
Integrated Objective systems of Today s are required Meetingfor improving productivity and quality Germplasm Improvement Systems Biology Precision Agriculture Clonal selection Cultivar and rootstock improvement via traditional breeding Pest/disease resistance Elucidate the regulation of key yield and fruit quality pathways Functional genomics linking genes to key traits Characterize the parameters regulating vine productivity and quality Model key relationships Variable rate management 5
Integrated Objective systems of Today s are required Meetingfor improving productivity and quality Germplasm Improvement Systems Biology Precision Agriculture Clonal selection Cultivar and rootstock improvement via traditional breeding Pest/disease resistance Elucidate the regulation of key yield and fruit quality pathways Functional genomics linking genes to key traits Characterize the parameters regulating vine productivity and quality Model key relationships Variable rate management 6
The Future of Grape Growing MEASURE Automated sensors measuring intrafield variability crop load, canopy size, irrigation requirements Measures used to construct geospatial maps of key relationships MANAGE Information used to spatially alter cultural practices MODEL 7
Historical sensors are site specific
Historical soil measures X X X 9
Sensors provide high density soil information 10
High Resolution Maps - EM Sensor
Characterizing Yield Variability 12
13
Why does variability matter? Cabernet Sauvignon 9.2 tons/acre 22.7 tons/ha 14
0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 15 10 5 0 Why does variability matter? Colony 2A Cabernet Sauvignon / 32.1 acres Mean yield = 9.2 tons per acre Yield (tons/acre) 15 Block acreage (%)
Block acreage (%) 20 15 10 5 0 What Why does is the variability size of the matter? prize? Colony 2A Cabernet Sauvignon / 32.1 acres Mean yield = 9.2 tons per acre 40% of vines producing below mean block yield Block improvement opportunity = 30% yield increase 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 Yield (tons/acre) 16
Block acreage (%) 20 15 10 5 0 What Why does is the variability size of the matter? prize? Colony 2A Cabernet Sauvignon / 32.1 acres Mean yield per acre = 9.2 Annual tons increase in 40% of vines producing below mean block yield Block improvement opportunity = 30% yield increase 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 Yield (tons/acre) revenue = $900/acre Estimated cost = $100/acre 30% yield increase without planting additional acreage Capital avoidance/acre Land - $50,000 Establishment - $35,000 17
What is the size of the prize for Big Data? 20 15 10 5 Colony 2A Cabernet Sauvignon / 32.1 acres Mean yield = 9.2 tons per acre 20% of vines produce quality below district average 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 Block acreage (%) Yield (tons/acre) 18
What is the size of the prize for Big Data? Block acreage (%) 20 15 10 0 Colony 2A Cabernet Sauvignon / 32.1 acres Mean yield = 9.2 tons per acre Additional revenue based on fruit quality improvement = 5$2,200/acre at farm gate Additional cost = $500/acre 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 Yield (tons/acre) 20% of vines produce quality below district average 19
Proximal Sensor Integrated systems - analytics Remote Sensor - Proximal Sensor Proximal Sensor Plant available water in soil Vegetation Index (NDVI) Yield Fruit Quality 20
Modeling Yield and Fruit Quality Data with Soil Parameters Significant Correlations with Fruit Yield Parameter Correlation (r 2 ) Subsurface K + Soil rooting depth Subsurface ph Subsurface P Subsurface organic matter Subsurface K/Mg ratio 0.903 0.774 0.805 0.805 0.882 0.890 Significant Correlations with Fruit Quality Parameter Correlation (r 2 ) Soil rooting depth Surface CA Subsurface CA / Mg ratio Surface CEC 0.673 0.506 0.510 0.554 21
Variable rate management 22
Contribution to total explained variance variance (%) Relative importance of soil parameters 100% 80% 60% 40% 20% 0% to block yield whc variability 5% Anions 11% Soil compaction 13% Textural class 16% ph Water holding capacity 0.46 0.46 0.46 0.46 0.46 0.46 0.46 0.46 46% 46% 23
Vine water use is variable based on canopy size Sap Flow (g H 2 O/hour) 120 100 80 60 40 Sap Flow - Colony 2A - 2012 28 gallons per vine per week 17 gallons per vine per week High Vigor Low Vigor Row direction 20 0 00:00 04:00 08:00 12:00 16:00 20:00 00:00 Time 24
Variable Rate Drip Irrigation Changes in canopy vigor (NDVI) Each square = 30m x 30m LANDSAT Pixel Before variable rate irrigation After variable rate irrigation Pixel level management based on canopy size 25
Impact of Precision Irrigation 2012 Block Yield 8.1 tons/ac 2015 Block Yield 10.2 tons/ac Yield improved 20%; Water use efficiency improved 30%
Summary and future challenges Sensor technology has advanced real-time, high density data collection Geospatial analytics for characterizing vineyard variables environment, growth, yield and quality Our ability to measure exceeds our ability to interpret Understand what is important and actionable Large gaps exist in variable rate application technologies for geospatial management Example: Variable rate drip irrigation Research collaboration (USDA - ARS) and industry partnership are essential to advance 27