Management Zones - University of Nebraska–Lincoln€¦ · Rotated soybean No-till 24 Study 2...
Transcript of Management Zones - University of Nebraska–Lincoln€¦ · Rotated soybean No-till 24 Study 2...
11/5/2019
© 2019 University of Nebraska –Lincoln1
Management ZonesJoe Luck, Nathan Mueller, and Laura Thompson
Contact Info
Nathan Mueller Laura Thompson1206 W 23rd StFremont, NE 68025-2504Phone: 402-727-2775Email: [email protected]: @croptechcafeBlog: http://croptechcafe.org/
1705 Stone StFalls City, NE 68355Phone: 402-245-2224Email: [email protected]: @agtechlauraTwitter: @onfarmresearch
Joe Luck
200 L.W. Chase HallLincoln, NE 68583Phone: 402-472-1488Email: [email protected]: @joeluck_unl
Group Discussion
Questions
1. What information should I use as a basis for creating different management zones?
2. How do I create a standard/reproducible procedure for delineating unique management zones?
3. How many unique management zones should a field be divided into?
4. How to I validate the quality of my management zones after creation?
Fridgen et al. 2004. Management Zone Analyst (MZA): Software for Subfield Management Zone Delineation. Agron. J. 96:100-108
Various Data Layers – Question 1
2012 Corn Yield Map
Drought
2013 Soybean Aerial Imagery
Flash Drought
Various Data Layers – Question 1
1979 Soil Map IDW Deep EC Map
11/5/2019
© 2019 University of Nebraska –Lincoln2
Various Data Layers – Question 1
Historic 1941 Aerial Image Deep Soil EC Data Points
What would your management zones look like for this field?
Management Zone Analyst –Questions 2 and 3
Yield
Electrical Conductivity
Aerial Imagery
Overview – Management Zones
• History of Management Zones (MZs)
• Challenges
• Peer-Reviewed Research Results
• Nebraska On-Farm Research Examples
• Platforms
• Summary
History of Management Zones
Management Zones
Under the pivot and dryland pivot corners – 2 management zones
11/5/2019
© 2019 University of Nebraska –Lincoln3
History of Management Zones – 1980s
• Concepts tested in the late 80s early 90s• Differed from farming by soil series in
the early 80s• Variability was significant within soil
series boundaries, needs to be determined by grid or transect sampling
• “Each management zone should ideally represent portions of the field that are relatively similar and homogeneous in soil fertility status so that a different uniform fertilizer recommendation can be made for each zone”
Mulla, D, and R. Khosla. 2015. Historical evolution and recent advances in precision farming. Ch. 1. In: (R. Lal, and B.A. Stewart, eds.), Soil Specific Farming: Precision Agriculture. Adv. Soil Sci. Taylor and Francis Publ., Boca Raton, FL.
Soil mapping units and yield map
History of Management Zones – 1990s and early 2000s
• Late 90s: “Are sub-regions of a field that express a homogeneous combination of yield-limiting factors for which a single crop input is appropriate”
• Common approaches for delineating MZs:• Soil texture and OM• Electrical conductivity mapping• Remoting sensing
• Less commonly used were yield mapping and elevation
Mulla, D, and R. Khosla. 2015. Historical evolution and recent advances in precision farming. Ch. 1. In: (R. Lal, and B.A. Stewart, eds.), Soil Specific Farming: Precision Agriculture. Adv. Soil Sci. Taylor and Francis Publ., Boca Raton, FL.
Source: EC163 Site-specific Nitrogen Management for Irrigated Corn
History of Management Zones – 2000s
• Yield history and multi-year analysis utilized more frequent to create management zones
• Management Zone Analyst (2004)• Software program that could take
numerous types of data layers to help delineate zones
• Standardized procedure
• Allowed user to determine number of zones
Fridgen et al. 2004. Management Zone Analyst (MZA): Software for Subfield Management Zone Delineation. Agron. J. 96:100-108
History of Management Zones – Current
• Complexity: Realization that the ideal set of zones for a field is not static, but changes year to year
• Scaling up: Realization for the need to create management zones by farm or region for within fields, not a unique set of management zones for each field.
• Need for modeling: Interest in moving past classification outcomes such as management zones to models that represent the soil as a continuum.
Challenges
11/5/2019
© 2019 University of Nebraska –Lincoln4
Challenges with various data sources
Yield Spatial Variability
• Measures yield difference among subfields within a field in a given year
Yield Temporal Variability
• Measures yield difference among fields subfields across years
• Often greater than spatial variability
What Yield Maps Should I Use?
Corn only
Include available years
Use multi-year average
Corn only
All except extremes
Use multi-year average
Soybeans only
All except extremes
Use multi-year average
Corn and Soybeans
Yield index
Run for specific scenarios and years (wet, average, dry)
Corn and Soybeans
Include all years
Use Relative Yield/Index
Soybeans only
Include available years
Use multi-year average
Challenges with various data source
• Yield• Plant integrate all
environmental factors
• Input or management interaction across the field
• Spatial and temporal variability are a real change
Bunselmeyer and Lauer. 2015. Using Corn and Soybean Yield History to Predict Subfield Yield Response. Agron. J. 107:558-562
Study by Bunselmeyer and Lauer (2015)
Rotation Tillage Time to ConsistentPatterns
Continuous corn Conventional till 4
Continuous corn No-till 20
Rotated Corn Conventional till 20
Rotated Corn No-till 20
Continuous soybean Conventional till 18
Continuous soybean No-till 18
Rotated soybean Conventional till 10
Rotated soybean No-till 24
Challenges with various data source
• Soil Electrical Conductivity• Significant relationship for 9 out of
13 site-years
• Climate, crop type, and specific field information needed to explain shape of the interaction
Kitchen et al., 1999. Soil Electrical Conductivity as a Crop Productivity Measure for Clayplan Soils. J. Prod. Agric. 12:607-617
How many sets of management zones do we need and assumptions often made
• Management zones for• For each crop (corn and soybeans
separate or together)• Seeding rates• Hybrid/variety placement• Seed treatment placement• Irrigation• Fertilizer• Pests: Weeds, insect, diseases
• Assumptions • We can easily create management
zones for whole-farm datasets• Once management zones are create,
they don’t need to change• We capture a majority of the spatial
and temporal variability with management zones
• Some constant factors on yield over time (i.e. soil texture)
• Response (yield and return) to each input by management zone is known
• No input by management zone interactions
Peer Reviewed
Publications
11/5/2019
© 2019 University of Nebraska –Lincoln5
Using Corn and Soybean Yield History to Predict Subfield Yield Response
• Authors• Hayley A. Bunselmeyer and
Joseph G. Lauer
• From University of Wisconsin-Madison
• Published in 2015
• Agronomy Journal 107:558-562
• Situation• Yield history often used to
delineate management zones in a field.
• Questions remain about the type and amount of data required to classify subfields into management zones
• How much years of yield history is needed to identify a consistent yield pattern within a subfield?
Materials and Methods
• Study 1• Plano silt loam near Arlington,
Wisconsin• 1987-2012 - 26 years
• Crop Rotations• Continuous corn• Rotation corn-soybean• Continuous soybeans
• Tillage• Conventional till
• Fall chisel plow• 2x spring field cultivator
• No-till
• Study 2• Plano silt loam near Arlington,
Wisconsin• 2003-2012 - 10 years
• Crop Rotations• Continuous corn• Rotation corn-soybean
• Tillage• Conventional till
• Fall chisel plow• 2x spring field cultivator
• No-till
Results and Discussion
• Definition of Consistent Pattern: Data subtracted one 2-yr cycle at a time until the pattern of high and low yielding subfields were still the same as the full data set.
Study 1 (1987-2012) – 26 years
Rotation Tillage Time to ConsistentPatterns
Continuous corn Conventional till 4
Continuous corn No-till 20
Rotated Corn Conventional till 20
Rotated Corn No-till 20
Continuous soybean Conventional till 18
Continuous soybean No-till 18
Rotated soybean Conventional till 10
Rotated soybean No-till 24
Study 2 (Analysis 2003-2012) - 10 years
Rotation Tillage Time to ConsistentPatterns
Continuous corn Conventional till 6
Continuous corn No-till 10
Rotated Corn Conventional till 6
Rotated Corn No-till 8
Rotated soybean Conventional till 6
Rotated soybean No-till 10
Results and Discussion
• It can take 4 to 20 years even with long-term management practices for consistent yield patterns to persist.
• Can corn yield be used to predict soybean yield and vice versa?• 87% of the subfields had the same trend for both corn and soybean
yield
• 9% of the subfields ranked low for corn yield, but high for soybean yield
• 4% of the subfields ranked high for corn yield, but low for soybean yield
Conclusions
• Many years of yield needed to define consistent patterns of subfield yields
• Subfield yield temporal variability is greater than spatial yield variability
• Classifying subfields into management zones for input prescriptions remains challenging
Peer Reviewed
Publications
11/5/2019
© 2019 University of Nebraska –Lincoln6
Combining Spatial and Temporal Corn Silage Yield Variability for Management Zone Development
• Authors• Tulsi P. Kharel et al.
• From Cornell University
• Published in 2019
• Agronomy Journal 111:1-9
• Situation• Yield data from multiple years
are standardized and relative yield values are used to development management zones
• However, fields and areas within fields can vary both spatially and temporally
Materials and Methods
• Study• Six farms in New York
• Data from 2015, 2016, and 2017
• 78 fields
• Corn silage yield
• Used USDA Yield Editor and SMS Advanced from AgLeader
Quadrant 1High YieldTemporal Stability
Quadrant 4Low YieldTemporal Stability
Quadrant 3Low YieldTemporal Instability
Quadrant 2High YieldTemporal Instability
Quadrant 1High YieldTemporal Stability
Quadrant 4Low YieldTemporal Stability
Quadrant 3Low YieldTemporal Instability
Quadrant 2High YieldTemporal Instability
Results and Discussion
• Stables zones easier to manage as yield can be predicted in advance with greater accuracy
• 56 to 70% of the field
Quadrant 1High YieldTemporal Stability
Quadrant 4Low YieldTemporal Stability
Quadrant 3Low YieldTemporal Instability
Quadrant 2High YieldTemporal Instability
• Potential for increasing production efficiency with within-season precision ag management
• 44 to 30% of the field
• Need to determine yield-limiting factors and accurately predict crop response to management under various weather scenarios
• Need to define the appropriate number of zone per farm to address spatial and temporal variability
Conclusions
• Spatial and temporal variability were not correlated (Field 2)
• Each substantial enough to warrant the development of stability-based yield zones that minimized input or maximized output
• A whole-farm quadrant-based approach eliminates problems with single-field and relative yield-based management zones
• No BMPs on the number years needed
Peer Reviewed
Publications
Validating a Digital Soil Map with Corn Yield Data for Precision Agriculture Decision Support
• Authors• C. W. Bobryk, N. Kitchen, S.
Drummond et al.
• From USDA-ARS and University of Missouri
• Published in 2016
• Agronomy Journal 108:957-965
• Situation• Managing variability in crop
production is complex due to interactions in GxExM• Genetic potential (G)
• Biosphysical environment (E)
• Management activity (M)
• Need for improved digital soil maps for precision ag management decisions
11/5/2019
© 2019 University of Nebraska –Lincoln7
Materials and Methods
• Study• NE, MN IA, and IN• Data from 2010, 2011, and 2012• 409 fields• Corn yield
• This new digital soil map (DSM) called environmental response unit (ERU) was the product of a public-private collaboration between the USDA-ARS, University of Missouri, and Dupont Pioneer
Materials and Methods
• Study• SSURGO
enhanced with LIDAR data
• 16.5 x 16.5 ft mapping resolution
• Eliminated ERS smaller than 0.5 acres
bu/ac 31-111111-159159-190190-223223-286
Results and Discussion
• 57% improvement within ERU over SSURGO
• The yield variance reduction differed from year to year and state to stat, but the greatest variance reduction was in 2012.
25%
16%
Conclusions
• Study used corn yield maps from in combination with a variance reduction algorithm to validate a high resolution ERU DSM
• The ERU DSM provided a great reduction of corn yield variance than SSURGO
• The process was able to classify soil and landscape characteristics in a way that will allow for improved precision ag management applications
Nebraska On-Farm Research Examples
Management Zones for Variable Rate Seeding
• Planter technology allows growers to change seeding rates within a field.
• Goal: Optimize yield/economics. Put more seeds where we can push to get a higher yield response, put less seeds where the soil/fertility cannot support it.
• Companies offer ready to use seeding rate prescriptions
11/5/2019
© 2019 University of Nebraska –Lincoln8
Variable Rate Seeding
• UNL Extension/On-Farm Research Recommendation: TEST IT.
2. Seeding rates evaluated with blocks of each rate multiple time in each zone. Implemented with VR seeding capabilities on the monitor/planter.
1. Management zones for seeding rate by Helena tool
Management Zones for Variable Rate Seeding Studies (multi-year yield history)
County (Year) Treatment Yield (bu/ac) Net Return ($/ac)
Lancaster (2013) 28,000 158B -
Variable Rate (24, 28, 32K) 160A -
Lancaster (2014) 28,000 204A 714.00
Variable Rate (24, 28, 32K) 203A 710.00
Lancaster (2015) 29,000 200A 730.00
Variable Rate (24, 29, 34K) 200A 730.00
Jefferson (2014) 23,500 129A 379.69
Variable Rate (17, 21, 24, 27K) 131A 387.31
Bushels per acre corrected to 15.5% moistureValues with the same letter are not significantly different at the 90% confidence level
Lancaster County (2013)
Source: https://cropwatch.unl.edu/farmresearch/articlearchives/vr-seed-experiments
Resources
Corn seeding rate optimization in Iowa, USA (Licht, Lnssen, and Elmore, 2016): https://link.springer.com/article/10.1007/s11119-016-9464-7
CropWatch Variable Rate Seeding Considerations: https://cropwatch.unl.edu/farmresearch/articlearchives/vr-seed-experiments
Multi-Hybrid
• Goal: place hybrids that have some trait that will make them perform better in the area they will do best.
• Examples: • Place hybrids with greater drought
tolerance where soils have lower water holding capacity
• Used treated seed for SDS prone areas within a field and non-treated seed in non-SDS areas
• Ability to place different hybrids in different portions of the field to best match genetics and environment
Multi-Hybrid
• Research at UNL – Stevens, et al. 2018
• 10, full field scale corn research sites• Drought tolerant hybrid (defensive) and
offensive hybrid
• Many sites showed that one hybrid was better for the entire field
• Where there were differences, the zone delineation was not stable from year to year
11/5/2019
© 2019 University of Nebraska –Lincoln9
2016 and 2017 multi-hybrid planting and management zones
Yield
Electrical Conductivity
Aerial Imagery
Multi-hybrid planting 2016 & 2017 in Dodge County
Treatment
Defensive
Hybrid
Offensive
Hybrid P-Value
2016
Defensive Zone 230 0.7020
Offensive Zone 242 0.1521
2017
Defensive Zone 212 0.2228
Offensive Zone 236 0.1156
2016 2017 2016-17
Defensive Hybrids Offensive Hybrid
What should have been the management zones Multi-Hybrid
• Research at UNL – Stevens, et al. 2018
• 4 soybean seed treatment sites
• Goal – place ILeVO seed treatment for sudden death syndrome in portions of the field where disease is generally more prevalent (generally soils that are more frequently ponded).
Multi-Hybrid
Figure 1. Zone prescription for soybean treated with
standard treatment (dark grey) and ILeVO (light grey).
Multi-HybridTreatment Standard Treatment + ILeVO® Standard Treatment P-Value
Yield (bu/ac) †
SDS Zone 67 A* 57 B 0.0003
Standard Zone 70 A 70 A 0.849
P-Value 0.9631 0.9494
Marginal Net Return ($/ac)‡
SDS Zone 608.19 529.75
Standard Zone 637.60 651.57
• The drastic yield difference between the ILeVO and standard treatment in the SDS zone resulted in a $79 advantage for using the ILeVO treatment.
• Considering the size of the SDS zone (around 50 acres), the additional return by using the ILeVO treatment would equal around $4,000 for the field.
• If the additional cost of a multi-hybrid planter is around $20,000, the technology could be paid off in around five soybean growing seasons in this field.
11/5/2019
© 2019 University of Nebraska –Lincoln10
Platforms
Management Zone Analyst
Yield
Electrical Conductivity
Aerial Imagery
Granular
• Formerly called Encirca
• We review the publication earlier on this process
Climate FieldView™ Nitrogen AdvisorClimate Cooporation, acquired by Monsanto 2013; Monsanto acquired by Bayer 2018
Adapt-NDeveloped by Cornell University, acquired by Yara for scaling Field Variability and Management
• Companies are taking advantage of “data mining” techniques to offer advice to farmers
11/5/2019
© 2019 University of Nebraska –Lincoln11
Numerous GIS Platforms
AgLeader SMSMapShots AgStudioJohn Deere ApexFarmWorks
Summary
• What information should I use as a basis for creating different management zones?
• How do I create a standard/reproducible procedure for delineating unique management zones?
• How many unique management zones should a field be divided into?
• How to I validate the quality of my management zones after creation?
Fridgen et al. 2004. Management Zone Analyst (MZA): Software for Subfield Management Zone Delineation. Agron. J. 96:100-108