Centre for Sustainable Cropping

DEXi-CSC

Interpreting DEXi Evaluations

DEXi-CSC input attributes

DEXi-CSC heat map

DEXi-CSC radar plots

DEXi is a computer program for creating multi-attribute decision models (MADMs) that:

allows interactive development of qualitative multi-attribute decision models
allows comparisons of differing attribute options
can be used to support complex decision-making tasks, i.e. selection of best option from a range of possible options
is used to create hierarchical, multi-attribute models from the decomposition of problems into smaller, less complex sub-problems that may be solved more easily

MADMs, like DEXi, are well suited to use in assessing crop production systems that are reliant on complex networks of interactions and trade-offs between biotic, abiotic and economic components.

DEXi-CSC has been developed to assess systems such as the CSC's integrated cropping system directly against standard commercial practice over a six year croping sequence.

The effect of cropping treatment was assessed according to the responses of a suite of indicators which were used to parameterise a qualitative multi-attribute model. Scenarios were run to test whether the integrated cropping system achieved greater levels of overall sustainability relative to standard commercial practice.

This case study demonstrates the value of a whole-systems approach using qualitative multi-attribute modelling for the assessment of existing cropping systems and for predicting the likely impact of new management interventions on arable sustainability.

DEXi-CSC is comprised of 97 input variables representing the basic biotic, abiotic and financial indicators of the arable cropping system (see DEXi-CSC input attributes). Input variables are aggregated into a hierarchical tree with a total of 332 aggregate variables. Overall sustainability, the top level, is the aggregate of two, second level branches Environmental sustainability and Economic sustainability. Environmental sustainability is the aggregate of three, third level branches (Biodiversity, Losses and Resource use). Economic sustainability is the aggregate of two, third level branches (Viability and Real profitability).

These branches reflect the main goals of the CSC platform:

maintainence of yield whilst enhancing biodiversity (in-field and field margin dicot weeds that benefit invertebrates)
increasing resource use efficiency (reduce reliance on non-renewables)
minimise losses via leaching, run-off, erosion and greenhouse gas emissions

1. Average market price	50. Price of crop seed
2. Average price of fuel	51. Price of desiccants
3. Baling fuel	52. Price of fungicide
4. Carting fuel	53. Price of FYM
5. Connectivity of semi-natural habitats	54. Price of herbicide
6. Crop competition	55. Price of insecticide
7. Compost fuel	56. Price of irrigation
8. Conditions for spray drift	57. Price of lime
9. Conditions for volatilisation	58. Price of mineral K fertiliser
10. Cover crop fuel	59. Price of mineral N fertiliser
11. Crop cover	60. Price of mineral P fertiliser
12. Crop fuel	61. Price of molluscicide
13. Crop rotation diversity/intensity	62. Price of seed treatment
14. Crop type	63. Price of soil sterilants
15. Deep inversion tillage fuel	64. Price of straw
16. Desiccants AI/rate	65. Price of trace elements
17. Desiccant fuel	66. Price of undersow seed
18. Environmental subsidies	67. Product quality
19. Field margin floral diversity	68. Production risk
20. Field margin structure	69. Proportion of gross margin due to main crop
21. Field properties	70. Rate compost
22. Fungicide AI/rate	71. Rate FYM
23. Fungicide fuel	72. Rate lime
24. FYM fuel	73. Rate mineral K fertiliser
25. Harvesting fuel	74. Rate mineral N fertiliser
26. Herbicide AI/rate	75. Rate mineral P fertiliser
27. Herbicide fuel	76. Rate straw
28. Insecticide AI/rate	77. Rate trace elements
29. Insecticide fuel	78. Rate undersow
30. Inversion tillage fuel	79. Ratio semi-natural: cultivated land
31. Irrigation availability	80. Requirement for agricultural equipment
32. Irrigation fuel	81. Seed cover crop
33. Irrigation requirement	82. Seed rate crop
34. Labour hourly wage	83. Seed treatment AI/rate
35. Lime fuel	84. Seed undersow
36. Mineral fertilisers fuel	85. Soil cover at pesticide application
37. Molluscicide AI/rate	86. Soil sterilants AI/rate
38. Molluscicide fuel	87. Soil sterilant fuel
39. N soil residual	88. Soil type
40. Non-environmental subsidies	89. Straw chopping fuel
41. Non-inversion tillage fuel	90. Straw selling price
42. Number of hours	91. Straw yield
43. P soil residual	92. Stubble/straw management
44. Pathogen pressure	93. Tillage intensity
45. Pest pressure	94. Trace elements fuel
46. Post-plant herbicide AI/rate	95. Undersow fuel
47. Precipitation	96. Weather conditions
48. Price of compost	97. Weed management strategy
49. Price of cover crop seed

The scales used for each input and aggregate function are available here.

An interactive Java implemntation of the DEXi-CSC model can be accessed here.

Viewing and interpreting several evaluations of a DEXi tree with over 300 aggregate functions can be challenging. To aid interpretation we use two types of visual representation of the output i.e. heat maps and radar plots.

These heat maps show the top seven layers of evaluation output for each of the six crops in the CSC the rotation. Input data is aggregated across six years of the first crop rotation. As a contrast conventional potato performs poorly whilst Integrated bean performs well.

Spring barley

Winter barley

Winter oilseed rape

Winter wheat

Conventional

Integrated

These radar plots show selected attributes from layers three and four of evaluation output for each of the six crops in the CSC the rotation. Input data is aggregated across six years of the first crop rotation.

Bean

Potato

Spring barley