Relationships Among Soil Properties, Crop Yield, Protein, and Response to Nitrogen Fertilizer Application in an Undulating Landscape in South Central Saskatchewan Elliott Hildebrand and Jeff Schoenau ASA, CSSA and SSSA Long Beach, CA November, 2014 Dept. Soil Science University of Saskatchewan
Context 4 Rs of Nitrogen (N) Management Goal to increase N-use efficiency (NUE) Global NUE for cereal production < 50% http://www.ipni.net/article/ipni-3323
Context Precision Agriculture GPS guidance Sectional control Variable Rate (VR) N Fertilizer Application Concept: Match N rates to varying production potentials in a field Does it work?
Context In theory VR N should: Improve yield and N-use efficiency However, improvements at the farm level have been difficult to document Reflects knowledge gaps of: Temporal and spatial variation in soil properties controlling yield Environmental controls How this affects response to applied N (Cassman et al., 2002)
Context In practice: Many farmers apply the same fertilizer rate across a whole field regardless of variability in yield potential Why? Efficient means are needed to create a variable application map Cost to ID, sample and predict crop response in separate zones Uncertainty surrounding benefits to be achieved Challenge: ID efficient reliable mechanisms to make VR map
Research Questions What is the effect of soil properties on crop yield and protein in a typical landscape in southern SK? Will protein concentration of crops help delineate effective fertilizer management zones?
Research Question Yield Can establish how much N it takes to produce a target yield Protein Reflects balance of N to other yield limitations
Yield vs Protein Curve Spring Wheat (adapted from Engel et al., 1999)
Study Objectives Summer 2012 Typical southern Saskatchewan undulating farm field Determine relationships between: Crop yield Crop protein Soil landscape properties Salinity Organic matter ph Soil nutrients
Study Objectives Summer 2013 Use these relationships to: 1) develop VR N prescription 2) compare performance to constant rate Side by side comparison
SW 31-20-03 W3 (2012) PEAS CANOLA WHEAT
SW 31-20-03 W3 (2013) WHEAT WHEAT CANOLA
Variable N Rates Strategy: Increase N rates where previous crop had low or medium protein Decrease N rates where previous crop had high protein
Wheat on Canola Stubble Fertilizer Rates Transect 1 Control 50 kg N ha -1 (109 kg 46-0-0 ha -1 ) 9m Variable Rate 40 (87) 40 (87) 70 (152) 0 70 (152) 60 (130) 40 (87) 60 (130) 9m Transect Point WC1 WC2 WC3 WC4 WC5 WC6 WC7 WC8 2012 Canola HP MY LP HY MP LY LP LY LP MY MP MY HP LY MP MY Transect 2 Variable Rate 40 (87) 70 (152) 0 70 (152) 60 (130) 60 (130) 60 (130) 0 9m Control 50 kg N ha -1 (109 kg 46-0-0 ha -1 ) 9m Transect Point WC9 WC10 WC11 WC12 WC13 WC14 WC15 WC16 2012 Canola HP LY LP MY LP LY LP MY MP HY MP HY MP HY LP LY
Wheat on Pea Stubble Transect 1 Control N Rate Varied N Rates N July 3/2013
Relationship of Canola Yield and Protein to Soil Properties Canola: Soil Relationships 2012 2013 2013 Organic carbon (%) (0-30 cm) Base Year Variable N Constant N Protei Yield Protein Yield n Yield Protein ns r=0.65* ns ns r=0.57* ns EC (ds m -1 ) (0-30 cm) 2013 Soil Moisture Spring (30-60 cm) ns ns r=0.51* ns ns ns ns ns ns ns ns r=-0.50* * indicates significant correlation at p < 0.05
Relationship of Wheat Yield and Protein to Soil Properties Wheat: Soil Relationships 2012 2013 Base Year Wheat on Canola Stubble Wheat on Pea Stubble Variable Constant Variable Constant Yield Protein Yield Protein Yield Protein Yield Protein Yield Protein Organic Carbon (%) (0-30 cm) 0.74* ns 0.51* ns ns ns ns ns ns ns EC (ds m -1 ) (0-30cm) ns 0.51* ns ns ns ns -0.52* ns -0.51* ns 2013 Soil Moisture Spring (30-60 cm) ns ns ns ns ns -0.60* ns ns ns * indicates significant correlation at p < 0.05 ns
Yield (kg ha -1 ) 5000 4500 2013 Crop Yield (kg ha -1 ) No significant differences at p<0.05 4000 3500 3000 2789 2662 2661 3012 3410 3180 2500 2000 1500 1000 500 Canola on Wheat Wheat on Canola Wheat on Peas Varied Control Constant
Protein (%) 18 2013 Crop Protein (%) 17 16 16.0 16.1 No significant differences at p<0.05 15 14 14.6 14.5 13 12 11 12.9 12.9 10 Canola on Wheat Wheat on Canola Wheat on Peas Varied Control Constant
Did protein help create N zones? Increasing N rates: Generally a positive yield and protein response where previous crop was low or medium protein Low and medium protein good indicator that more N was required Decreasing N rates: Generally a negative yield and protein response where previous crop was high protein High protein a less reliable indicator that N could be reduced
Conclusions Surface soil organic carbon content was sometimes positively related to yield and protein of canola and wheat Reflects variation in OM and N supply from mineralization Varied from year to year and field to field Surface soil electrical conductivity (salinity) was positively related to protein content, often negatively with yield Greater soil moisture at depth = lower protein
Conclusions Average yield and protein across the landscape in varied N rate and constant N rate were similar Since similar total amounts of N fertilizer were used in each, no difference in economic return Same results for each crop Prescription approach needs refining? What can be improved? Be careful about reducing N rates in a VR prescription!
Thank You! Questions?
Supplemental Slides
SW 31-20-03 W3 High ground Bathtub ring salinity Old rail line Saline depression
Precipitation Rainfall (mm) Month 2011 2012 2013 April 3 26 6 May 38 116 29 June 11 109 82 July 52 37 54 Aug 53 26 60 Sept 6 4 42 Oct 27 0 0 Total (mm) 190 318 273 Total (inches) 7.6 12.7 10.9
Context Current methods to create VR maps include: Soil electrical conductivity (EC) maps Satellite imagery Elevation maps Yield maps Soil surveys All reveal variability in many different aspects
Soil Association Kh4:l- Ad5:l Ga3:St2 (Ayres et al., 1985)
Harvest 2013 Results Wheat on Pea Stubble Transect 1 Control N Rate Varied N Rates N August 23/2013
Wheat on Canola Stubble Control Varied N Rates
Acknowledgements Case New Holland Jim Henry Joel Gervais Ross Welford Software
Why no significant differences? Mean Impact of Variable Rate N Strategy versus Constant N Rate on Yield and Protein VR Strategy Canola on Wheat Wheat on Canola Wheat on Peas Decrease N Rate Yield (kg ha -1 ) -99-336 -67 Protein (%) -1.2-0.5 0.1
Variable Rate Seeding 2014 VR Strategy Increase N rate where previous crop Low protein Medium protein Keep N rate constant where previous crop High protein Harvest completed; anticipating results!
Protein Sensor
Seeding Outfit 2013
Relationship of EC to Soil N Canola on Wheat Field Area 1 EC 0-30cm (µs cm -1 ) EC 30-60cm (µs cm -1 ) Wheat on Canola Field Area 2 EC 0-30cm (µs cm -1 ) EC 30-60cm (µs cm -1 ) Wheat on Peas Field Area 3 EC 0-30cm (µs cm -1 ) EC 30-60cm (µs cm -1 ) Extractable NO - 3 - N AEM NO - 3 - N ------------------------------ Depth (cm) --------------------------------- 0-30 30-60 0-30 30-60 0-30 30-60 ---Spring 2013--- ------------------Fall 2012---------------------- 0.55* (0.02) 0.68 (0.004) 0.52 (0.05) 0.57 (0.02) 0.53 (0.04) -0.51 (0.05) 0.52 (0.04) -0.69 (0.003) -0.51 (0.04) -0.71 (0.002)
Soil Association Kh4:l- Ad5:l Ga3:St2 (Ayres et al., 1985)