Assessment of the Use of Geographically Weighted Regression for Analysis of Large On-Farm Experiments and Implications for Practical Application

On-farm experimentation (OFE) is a farmer-centric process that can enhance the adoption of digital agriculture technologies and improve farm profitability and sustainability. Farmers work with consultants or researchers to design and implement experiments using their own machinery to test management practices at the field or farm scale. Analysis of data from OFE is challenging because of the large spatial variation influenced by spatial autocorrelation that is not due to the treatment being tested and is often much larger than treatment effects. In addition, the relationship between treatment and yield response may also vary spatially. We investigate the use of geographically weighted regression (GWR) for analysis of data from large on-farm experiments. GWR estimates local regressions, where data are weighted by distance from the site using a distance-decay kernel. It is a simple approach that can be easily explained to farmers and their agronomic advisors. We use simulated data to test the ability of GWR to separate yield variation due to treatment from any underlying spatial variation in yield that is not due to treatment; show that GWR kernel bandwidth can be based on experimental design to accurately separate the underlying spatial variability from treatment effects; and demonstrate a step-wise model selection approach to determine when the response to treatment is global across the experiment or locally varying. We demonstrate our recommended approach on two large-scale experiments conducted on farms in Western Australia to investigate grain yield response to potassium fertiliser. We discuss the implications of our results for routine practical application to OFE and conclude that GWR has potential for wide application in a semi-automated manner to analyse OFE data, improve farm decision-making, and enhance the adoption of digital technologies.

[1]  J. Cock,et al.  From Observation to Information: Data-Driven Understanding of on Farm Yield Variation , 2016, PloS one.

[2]  N. Martin,et al.  Spatial variability of crop responses to agronomic inputs in on-farm precision experimentation , 2020, Precision Agriculture.

[4]  P. Kyveryga.  On‐Farm Research: Experimental Approaches, Analytical Frameworks, Case Studies, and Impact , 2019, Agronomy Journal.

[5]  Martin Charlton,et al.  GWmodel: An R Package for Exploring Spatial Heterogeneity Using Geographically Weighted Models , 2013, 1306.0413.

[6]  Ludwig Theuvsen,et al.  Adoption of precision agriculture technologies by German crop farmers , 2016, Precision Agriculture.

[7]  Adrian Baddeley,et al.  Novel approach to the analysis of spatially-varying treatment effects in on-farm experiments , 2020 .

[8]  T. Kutter,et al.  The role of communication and co-operation in the adoption of precision farming , 2011, Precision Agriculture.

[9]  S. Fotheringham,et al.  Geographically Weighted Regression , 1998 .

[10]  Dayton M. Lambert,et al.  A Comparison of Four Spatial Regression Models for Yield Monitor Data: A Case Study from Argentina , 2004, Precision Agriculture.

[11]  Michael Robertson,et al.  Within-field variability of wheat yield and economic implications for spatially variable nutrient management , 2008 .

[12]  I. Goldringer,et al.  Influence of experimental design on decentralized, on-farm evaluation of populations: a simulation study , 2019, Euphytica.

[13]  Peter Carberry,et al.  Farmers use intuition to reinvent analytic decision support for managing seasonal climatic variability , 2012 .

[14]  Craig L. Dobbins,et al.  Spatial analysis of yield monitor data: case studies of on-farm trials and farm management decision making , 2008, Precision Agriculture.

[15]  Luc Anselin,et al.  A Spatial Econometric Approach to the Economics of Site‐Specific Nitrogen Management in Corn Production , 2004 .

[16]  B. Marchant,et al.  Agronōmics: transforming crop science through digital technologies , 2017 .

[17]  D. Pannell.  Economic perspectives on nitrogen in farming systems: managing trade-offs between production, risk and the environment , 2017 .

[18]  R. Bell,et al.  Simulating wheat growth response to potassium availability under field conditions in sandy soils. II. Effect of subsurface potassium on grain yield response to potassium fertiliser , 2015 .

[19]  Joe D. Luck,et al.  The Data‐Intensive Farm Management Project: Changing Agronomic Research Through On‐Farm Precision Experimentation , 2019, Agronomy Journal.

[20]  J. Melkonian,et al.  Adapt-N Outperforms Grower-Selected Nitrogen Rates in Northeast and Midwestern United States Strip Trials , 2016 .

[21]  J. Cock,et al.  Crop management based on field observations: case studies in sugarcane and coffee , 2011 .

[22]  Brett Whelan,et al.  A ‘small strip’ approach to empirically determining management class yield response functions and calculating the potential financial ‘net wastage’ associated with whole-field uniform-rate fertiliser application , 2012 .

[23]  R. G. V. Bramley,et al.  Adoption of variable rate fertiliser application in the Australian grains industry: status, issues and prospects , 2011, Precision Agriculture.

[24]  J. Bouma,et al.  Future Directions of Precision Agriculture , 2005, Precision Agriculture.

[25]  M. Charlton,et al.  Some Notes on Parametric Significance Tests for Geographically Weighted Regression , 1999 .

[26]  R. G. V. Bramley,et al.  Farmer attitudes to the use of sensors and automation in fertilizer decision-making: nitrogen fertilization in the Australian grains sector , 2018, Precision Agriculture.

[27]  R. M. Lark,et al.  The geostatistical analysis of experiments at the landscape-scale , 2006 .

[28]  R. G. V. Bramley,et al.  Enhancing the value of field experimentation through whole-of-block designs , 2010, Precision Agriculture.

[29]  A. Stewart Fotheringham,et al.  Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity , 2010 .

[30]  D. Bullock,et al.  Quadratic and quadratic-plus-plateau models for predicting optimal nitrogen rate of corn: a comparison , 1994 .

[31]  Brian R. Cullis,et al.  Accounting for natural and extraneous variation in the analysis of field experiments , 1997 .

[32]  H. Piepho,et al.  More, Larger, Simpler: How Comparable Are On‐Farm and On‐Station Trials for Cultivar Evaluation? , 2018 .

[33]  S. Cook,et al.  A decision support system for mapping the site-specific potassium requirement of wheat in the field , 2001 .

[34]  S. E. Cook,et al.  Field-Scale Experiments for Site-Specific Crop Management. Part II: A Geostatistical Analysis , 2004, Precision Agriculture.

[35]  Hans-Peter Piepho,et al.  Statistical aspects of on-farm experimentation , 2011 .

[36]  M. Bell,et al.  What soil information do crop advisors use to develop nitrogen fertilizer recommendations for grain growers in New South Wales, Australia? , 2019, Soil Use and Management.

[37]  Yee Leung,et al.  Statistical Tests for Spatial Nonstationarity Based on the Geographically Weighted Regression Model , 2000 .

[38]  Jane Elith,et al.  blockCV: an R package for generating spatially or environmentally separated folds for k-fold cross-validation of species distribution models , 2018, bioRxiv.

[39]  G Philip Robertson,et al.  Field-scale experiments reveal persistent yield gaps in low-input and organic cropping systems , 2017, Proceedings of the National Academy of Sciences.

[40]  W. Tobler.  A Computer Movie Simulating Urban Growth in the Detroit Region , 1970 .

[41]  L. Zhang,et al.  Comparison of bandwidth selection in application of geographically weighted regression : a case study , 2008 .

[42]  Peter Adams,et al.  Improving pathways to adoption: Putting the right P's in precision agriculture , 2008 .

[43]  R. M. Lark,et al.  A Method to Investigate Within‐Field Variation of the Response of Combinable Crops to an Input , 2003 .

[44]  R. Bramley,et al.  A Simple Method for the Analysis of On‐Farm Strip Trials , 2012 .

[45]  K. Smith,et al.  Efficiencies of nitrogen fertilizers for winter cereal production, with implications for greenhouse gas intensities of grain , 2012, The Journal of Agricultural Science.

[46]  M. Stone.  An Asymptotic Equivalence of Choice of Model by Cross‐Validation and Akaike's Criterion , 1977 .

[47]  A. Páez,et al.  A General Framework for Estimation and Inference of Geographically Weighted Regression Models: 1. Location-Specific Kernel Bandwidths and a Test for Locational Heterogeneity , 2002 .

[48]  R. M. Lark,et al.  A landscape-scale experiment on the changes in available potassium over a winter wheat cropping season , 2007 .

[49]  A. Fotheringham,et al.  The Multiple Testing Issue in Geographically Weighted Regression , 2016 .

[50]  R. Bramley,et al.  Precision agriculture — opportunities, benefits and pitfalls of site-specific crop management in Australia , 1998 .

[51]  B. Brorsen,et al.  Crop Input Response Functions with Stochastic Plateaus , 2008 .

[52]  Giacomo Carli,et al.  6 th International Conference on Information and Communication Technologies in Agriculture, Food and Environment (HAICTA 2013) Drivers of Precision Agriculture Technologies Adoption: A Literature Review , 2013 .

[53]  G. Marshall,et al.  Understanding and promoting adoption of conservation practices by rural landholders , 2006 .

[54]  Kai-Tai Fang,et al.  A Note on the Mixed Geographically Weighted Regression Model , 2004 .

[55]  H. Piepho,et al.  Beyond Latin Squares: A Brief Tour of Row‐Column Designs , 2015 .

[56]  S. Welham,et al.  Establishing the precision and robustness of farmers’ crop experiments , 2019, Field Crops Research.

[57]  Martin Charlton,et al.  The GWmodel R package: further topics for exploring spatial heterogeneity using geographically weighted models , 2013, Geo spatial Inf. Sci..