An originally Danish DSS for IWM, which was designed and constructed by IPM Consult in 2014, as a 4th generation tool in terms of weed modelling and IT basis, was used to construct operational DSS for IWM in Spain, Germany and Denmark for use in wheat and maize.
Based on field reports on weed infestations, a tool named ‘Consultation’ in this DSS will run through the following 3 calculation steps:
- Quantification of needs for weed control on the weed species level, by combining a threshold model with a model, which provides efficacy targets, when threshold values have been exceeded, which increase gradually with increased weed density (with the aim of keeping weeds on a sufficiently low level)
- Identification of herbicide products and connected dose rates plus measures for mechanical control, which meet efficacy targets
- With a set-off in dose rates calculated in step 2: Optimization of herbicide 2-4 way tank-mixtures minimize farmer’s cost or some index values
In the tool ‘Consultation’, 2 aspects of herbicide resistance management have also been included:
- Control of already resistant biotypes of weeds, by use of relatively high efficacy targets and dose-response calculations in Step 2, which provide very low efficacy against herbicide ‘mode-of-action’ (MOAs), where resistance has already been detected. Automatically, this will force ‘Consultation’ to seek alternative herbicides/MOAs, which can meet the efficacy targets
- Options for the user of DSS-IWM to filter out solutions from step 3, which contain risky MOAs, and which the farmer knows that was already used in the previous crop(s)
By reversing calculations in Step 1-3, additional tools have been developed to assist farmers and also their advisors to achieve various overviews of expected efficacy of herbicides/mechanical control for combinations of:
- Weed species (compulsory)
- Classes of weed size (optional)
- Classes of temperatures (optional)
- Classes of drought stress (optional)
where efficacy of herbicides/mechanical control is evaluated for:
- Single herbicides/mechanical methods as selected by the user
- 2-4 way herbicide tank-mixtures, as selected by the user
- Total list of herbicides/mechanical options, in a crop as selected by the user
As these calculations are run by a joint mathematical model, integrity between output from the different tools is automatically ensured.
Step 1 was customized by setting up firstly, threshold values and connected values of efficacy targets, which experts in JKI, UdL and IPMC considered to be relatively safe for farmers. When done, systematic reductions were made on these initial efficacy targets. Subsequently, these different versions were all tested in field experiments to identify versions, which demonstrated suitable balances between achieved weed control and potentials for saving input of herbicides/mechanical control.
Step 2 was customized by collecting efficacy data on the single herbicide products against single weed species. On the crop level, JKI, UdL and IPMC decided whether to include also specific adjustments for optional parameters, i.e. classes of weed size, temperatures and drought stress.
Step 3 was customized by identifying 2-4 way herbicide tank-mixtures, which shall be available for optimization, i.e. minimization of cost, TFI or similar.
The following numbers of combinations of crops, weeds and chemical/mechanical options for control have been included in DSS-IWM:
Germany: 2 crops (winter wheat and maize), 87 weed species incl. 16 herbicide resistant biotypes, 163 herbicide products incl. generic products, 469 2-3 way herbicide tank-mixtures, available to cost-optimization, 2 methods for mechanical control (only demonstration of principle).
Spain: 6 crops (wheat plus 5 additional cereals and maize), 41 weed species incl. 17 herbicide resistant biotypes, 121 herbicide products incl. generic products, 1023 2-3 herbicide tank-mixtures, available to cost-optimization, 2 methods for mechanical control (only demonstration of principle).
Denmark: 33 crops (cereals, grasses, clovers, rape seeds, maize and beets), 119 weed species incl. 15 herbicide resistant biotypes, 202 herbicide products incl. generic products, 829 2-4 way herbicide tank-mixtures, available to cost- and TFI-optimization, 6 methods for mechanical control (full-functional).
Based on a completely new DSS source-code and integrated database structure, which was designed and constructed before the DSS-IWM project, IPMC contributed to customize new facilities in the Danish version of DSS-IWM regarding:
- new and higher classes of weed density, which include both very inferior and very extreme classes
- a connected new model, which combines a ‘threshold model’ with a new equation, which connect weed density with ‘target efficacy’
- new and higher classes of weed growth stages, where use of herbicides is no longer possible, while use of mechanical control is still possible
- a connected new equation, which connect classes of weed size with herbicide dose-response calculations
- a new equation, which connect classes of temperatures (on the day of herbicide application) with herbicide dose-response calculations
- customization and integration in of measures to prevent/delay new cases of herbicide resistance
- customization of measures for mechanical weed control
In close collaboration with JKI, IPMC contributed also intensively to transform efficacy data into estimates of parameters in equations used in herbicide dose-response calculations. With both JKI and UdL, IPMC contributed to enter all materials required to run DSS-IWM also in Spain and Germany. This included set-up of national, initially empty versions of the integrated DSS-IWM database, and there after collection and gradual integration all details, including also complete user-interfaces in Spanish and German languages, respectively.
Field validation trials in maize and winter wheat were carried out to test the applicability of the new DSS-IWM. In summary, the results show:
In Spain: Field trial results confirm that the obtained efficacies are equal or slightly higher than the predicted by the DSS-IWM. In addition, herbicide use is lowered compared to the standard treatment proposed by the advisors as a mean of 30%. So, the objectives of the project have been achieved.
In Germany: Field trail results confirm that the objectives of the DSS-IWM have been achieved. The average efficacy in both maize and winter wheat according to treatment suggestions of the DSS program was slightly lower than according to the local standard treatments, but at many sites it exceeded 90%. A small residual weed infestation may be accepted and does not interfere with the yield. The treatment frequency index (TFI) in the DSS variants was reduced by up to 50%, resulting in cost savings for herbicides of 50% to 60%. The program is therefore suitable for supporting the ecological and economical objectives of weed control within the framework of Integrated Weed Control.
In Denmark: Unfortunately, planned and subcontracted field validation experiments, was cancelled, because the contractor explained that he was unable to follow the instructions in the joint experimental protocol. Despite the cancellation, IPMC had already worked for 1-½ year to prepare this work, which included contributions to joint experimental protocol, public procurement, sub-contracting and huge efforts to clarify methods, where methods for measuring efficacy turned out to be a key problem to the contractor. This also affected plans for publication of results, the DSS-IWM consortium decided to execute a replacement experiment, where an imagined, joint field report on weed infestation was used as input to the Danish, German and Spanish version of DSS-IWM, respectively, where results in terms of choice of herbicide products, dose rates, cost, expected efficacy, etc. were analysed. A research paper with these results was submitted to an international scientific journal (Rydahl et al., 2019).
However, in the period 1989-2006, >2,000 field validation trials have been executed in Denmark with predecessor versions of DSS-IWM in different crops, with different facilities included in calculations and on 2-3 levels of efficacy targets. Results from DSS versions that were officially released show that no cases of yield losses have occurred, residual weed infestations were max. 10-15% total ground cover at harvest time and the input of herbicides were reduced by about 40% in cereals and about 20% in other crops.
Compared to the predecessor systems, DSS-IWM in Denmark has intendedly and systematically been designed to be relatively safer to end-users. This was achieved by invention and integration in DSS-IWM of new equations, which have the systematic effect that herbicide dose rates and connected, expected levels of efficacy against weeds will be slightly increased, as compared to equations used on the Danish predecessor DSS. Consequently, IPMC concluded that the relatively few additional field experiments, which were planned in the DSS-IWM project, would not disturb plans for release in Denmark.
Figure 1: Predicted efficacies for Lolium rigidum susceptible and resistant compared to the field results for different treatments in a field demonstration for ADAMA, Spain
Table 1: Mean efficacies, treatment frequency indices and prices in experiments with maize, assessment ca. 4 weeks after application, Germany, DSS 1 (safe), DSS 2 (risky)