Title: Precision Agriculture in Australian Viticulture
1Precision Agriculture in Australian Viticulture
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
2Outline
- Australian Viticulture
- Precision Viticulture - why?
- Yield Monitoring
- Spatial Variability
- Examples of Variability in Vineyards
- The Future.
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
3Introduction
- Viticulture began in Australia with the first
fleet - Today there are gt4000 growers cropping 100,000
ha of vineyards in Australia - They produced almost 1.2 million tonnes of
crushed grapes in 1999 (over double that produced
in 1991)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
4Export and Domestic sales of Australian Wine
1984-1998
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
5Australian Winegrape Production
State Tonnage
South Australia 448, 561
52.4 Victoria 208, 406 24.3 New South
Wales 174, 468 20.4
Ranked 11th in the World in production Major
varieties are Chardonnay, Shiraz and Cabernet
Sauvignon
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
6Why Precision Viticulture?
- Horticultural crops often drip irrigated
- good water management potential
- Decreased influence of water on growth, yield and
quality leads to increase influence of other
factors - soil nutrients
- management decisions
- environmental variables
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
7Why Precision Viticulture?
- Horticultural industries - high profitability -
intensively managed - very quality orientated - Viticulture all these plus - dominated by large
producers - growth industry - highly mechanized
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
8Why not so far?
- To date research has focused on broadacre
cropping - wheat, cotton and sugarcane - Horticultural and root crops lagged due to
absence of effective yield monitors - Diversification of PA into other agricultural
industries
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
9Current Research
- Yield Monitoring and Sensors
- Dr Graham Moore - Uni. Of Melbourne
- Dr Francis Sevila - ENSAM Uni in Montpellier
- Dr Shrini Upadhyaya - UC Davis
- Industry - HarvestMaster, AB Electronics
- Agronomic interpretation
- NASA - remote sensing to predict quality
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
10Research Site
- Richmond Grove Vineyard, Cowra NSW
- Operated by Orlando-Wyndham
- 700 acres of vines including Chardonnay,
Semillon, Pinot Noir, Cabernet Sauvignon and
Cabernet Franc - Vineyard self contained with own machinery and
small laboratory
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
11Yield Monitoring
G 60K Pull type Grape Harvester
http//www.gregoire.com.au
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
12Types of Grape Yield Monitors
- Load cells - conventional horticultural crops,
measures delta weigh of crop - Ultra-sonic sensor - measures change in height of
grapes and interpolates volume
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
13How Grape Yield Monitors WorkLoad Cells
- Off load ramp suspended in a cradle
- Load cells bear weight of the ramp
- Weigh bins
- Change in mass monitored
- Error source from vibrations
Load cells
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
14How Grape Yield Monitors WorkUltra-sonic sensor
- Sensor Mounted above the off load ramp
- Height above ramp measured by pulse
- Change in height leads to a change in pulse time
- Volume interpolated from ?height
Sensors
Grapes
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
15Yield Monitors - are they reliable?
- What sort of accuracy/precision do they give?
- Where are the error sources and how can they be
improved? - How does the monitor react under different
situations?
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
16Yield Monitor Performance
- Ultra-sonic tested
- Accuracy was between -5.6 and 11.5
- Skewed towards over estimation of yield
-6
0
6
12
-3
3
9
Percentage error
Mean 3.1 Maximum 11.5 Minimum -5.6
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
17Raw yield at Richmond Grove Vineyard Cowra, 1999
400m
N
7
8
9
10
11
Yield (Mgha-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
18Raw Yield Map of Block 25
100 m
7
10
Yield (Mgha-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
19South 25 Raw Yield Description
Mean 8.8 Std Dev. 3.27 Median 8.7 25
Quantile 6.5 75 Quantile 10.9
Yield (Mgha-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
20Yield Variogram for South 25
10
- Range 12m
- Small compared to other crops
- Noisy data
- High spatial variability
9
8
Gamma
7
6
5
10
15
20
25
Distance (m)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
21South 25 Kriged Yield Map, Richmond Grove 1999
100 m
N
7
8
9
10
11
Yield (Mg/ha)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
22Kriged Yield Map, 1999
400m
N
7
8
9
10
11
12
Yield (Mgha-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
23Spatial Variability and Management Zones
- Soil ECa
- Yield indicators - bud, bunch counts
- An example only
- Look at one block (D3-4) of grapes
- relationship between yield, soil ECa, bud and
bunch counts
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
24Apparent Soil Electrolytic Conductivity (ECa)
- Real time sensing using Veris 3100
- Soil moisture, salt and clay content
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
25Soil ECa Map of Richmond Grove, May 1999
400m
N
25
50
75
Soil ECa Quartiles
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
26Kriged Soil ECa map of D3-4, May 1999
100m
N
10
12.5
15
Soil ECa (mSm-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
27Variogram for Soil ECa
Gamma
Distance (m)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
28Bud Counts vs. Yield in Block D3-4
250
- Low yield associated with low bud counts
- High has no association with bud counts
200
Bud Count
150
100
4
6
8
10
12
14
Yield (Mg/ha)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
29Bunch Counts vs. Yield in Block D3-4
- No relationship between yield and bunch
220
180
Bunch Count
140
100
4
6
8
10
12
14
Yield (Mg/ha)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
30Variograms for Bud and Bunch Counts
900
700
700
Gamma bunch counts
Gamma bud counts
500
500
300
300
50
200
50
200
100
150
100
150
Distance (m)
Distance (m)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
31Kriged Yield Map of D3-4, 1999
100m
N
7.5
9
10.5
12
Yield (Mgha-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
32Irrigation plan for D3-4
100m
N
7.5
9
10.5
12
Yield (Mgha-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
33Kriged Soil ECa map of D3-4, May 1999
100m
N
10
12.5
15
Soil ECa (mSm-1)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
34Potential Management zones based on Soil ECa and
Yield map, 1999
Zone Yield ECa (MG/ha) (mS/m)
N
Red 7.2 10.6 Green 11.0
11.1 Blue 12.1 15.4
100m
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
35The Future...
- Immediate
- determine the potential for Precision Viticulture
in Australia - is there enough spatial variability
- can we manage this variability
- Long term
- application to table grapes other horticultural
crops - derivation of DSS (eg AUSVIT) and farmer friendly
technology
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
36Our Research Goals
- Assessment of Yield Monitors
- Build an extensive vineyard database
- Yield maps
- Geo-referencing vines for crop counts
- Intensive site directed soil sampling
- Remote sensing
- Moisture - Veris calibration and intensive
monitoring in crops
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
37The Quality Issue
- Value adding has big potential in viticulture
- Two issues
- Real time sensing of grape quality
- Prediction of wine quality in the vineyard
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
38Acknowledgements
- Orlando-Wyndham Pty Ltd
- Mark Sims - Assistant Manager at Richmond Grove
- Ron Campbell and HarvestMaster U.S. Pty Ltd
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
39A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
40Making Precision Viticulture Work
- Yield monitoring
- Quantifying spatial variation and interactions
between yield determining factors - Defining management zones and modelling
interactions - Tailoring agronomic practices to suit management
zones
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
41Clusters Of Bud vs. Yield
Cluster Bud Yield (Count)
(MG/ha) Red 159 7.4 Green
160 11.4 Blue 224 12.8
250
200
Bud Count
150
- When Green cluster removed the linear fit is
improved
100
4
6
8
10
12
14
Yield (Mg/ha)
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
42Spatial Location of Clusters
- Blue Management Zone- high yield, high bud
count and high ECa - Red management zone - low yield, Bud count and
ECa - Green Management Zone- High yield but low ECa
and Bud count
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
43Spatial Variability in Vineyards
- Yield is spatially variable
- Other vineyard parameters
- soil properties
- physical
- chemical
- yield (quantity) indicators
- bud, bunch counts
- remote sensing
- quality indicators
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e
44S25 - elevation yield and Eca and bud/bnch counts
A u s t r a l i a n C e n t r e f o r P r e c
i s i o n A g r i c u l t u r e