Decision Support System For Weed Management Research Articles

Towards an integrated weed management decision support system: A simulation model for weed-crop competition and control

A mathematical simulation model for the multi-annual assessment of Integrated Weed Management strategies is presented. The model allows the simulation of the competitive interaction between annual weeds and crops. For weed development, the following processes are represented: (i) demographic dynamics on a daily basis considering the numeric composition of the different phenological states, (ii) intra and inter specific competition, (iii) seed production and (v) the effect of different control methods. For the crop development, the computed variables are: (i) Leaf Area Index, (ii) competitive effect over the weed, (iii) the expected yield as a function of weed competition. The model was developed in close collaboration with agricultural technicians and extensionists aiming to target a decision-making related audience. Results are provided for a wild oat (Avena fatua) – winter wheat (Triticum aestivum) / malting barley (Hordeum distichum) rotation system, typical of the Semiarid Pampean Region of Argentina. Multi-annual scenarios were generated to evaluate the effect of different management strategies against common herbicide-based practices in susceptible and resistant weed populations. Parts of the model were validated with independent experimental data. Finally, future improvements of the model and some guidelines towards the development of a long-term Decision Support System for weed management are provided.

RIM: Anatomy of a Weed Management Decision Support System for Adaptation and Wider Application

RIM, or “Ryegrass Integrated Management,” is a model-based software allowing users to conveniently test and compare the long-term performance and profitability of numerous ryegrass control options used in Australian cropping systems. As a user-friendly decision support system that can be used by farmers, advisers, and industry professionals, RIM can aid the delivery of key recommendations among the agricultural community for broadacre cropping systems threatened by herbicide resistance. This paper provides advanced users and future developers with the keys to modify the latest version of RIM in order to facilitate future updates, modifications, and adaptations to other situations. The various components of RIM are mapped and explained, and the key principles underlying the construction of the model are explained. The implementation of RIM into a Microsoft Excel® software format is also documented, with details on how user inputs are coded and parameterized. An overview of the biological, agronomic, and economic components of the model is provided, with emphasis on the ryegrass biological characteristics most critical for its effective management. The extreme variability of these parameters and the subsequent limits of RIM are discussed. The necessary compromises were achieved by emphasizing the primary end-use of the program as a decision support system for farmers and advisors.

Modeling seed dormancy release and germination for predicting Avena fatua L. field emergence: A genetic algorithm approach

Avena fatua is a cosmopolite weed species which produces severe yield losses in small-grain production systems in temperate and semiarid climates. In the semiarid region of Argentina, A. fatua field emergence patterns show great year-to-year variability mainly due to the effect of highly unpredictable precipitation regimes as well as a complex seedbank dormancy behavior regulated by both, genetic and environmental factors. Previously developed models for the same agroecological system based on Non-Linear Regression techniques (NLR) and Artificial Neural Networks (ANN) were either unable to accurately predict field emergence or lacked explanatory power. The main objective of the present work is to develop a simple (i.e. parsimonious) model for A. fatua field emergence prediction for the semiarid region of Argentina based on the disaggregation of the dormancy release phase from the germination/pre-emergence growth processes, using easy accessible soil microclimate derived indices as input variables and observed cumulative field emergence data as output variable. The parsimony and predictive capability of the newly developed model were compared with NLR and ANN approaches developed by the same authors for the same agroecological system. Specifically, dormancy release was modeled as a logistic function of an after-ripening thermal-time index while germination/pre-emergence growth was represented by a logistic distribution of hydrothermal-time accumulation. A total of 528 input/output data pairs corresponding to 11 years of data collection were used in this study. Due to its implementation simplicity and good convergence features, a Genetic Algorithm (GA) was adopted to solve the resulting optimization problem consisting on the minimization of the Mean Square Error (MSE) between training data and experimentally obtained field emergence data. The newly developed GA based approach resulted in a significantly more parsimonious model (BIC=−1.54) compared to ANN (BIC=−1.36) and NLR (BIC=−1.32) models. Model evaluation with independent data also showed a better predictive capacity of the GA approach (RMSE=0.07) compared to NLR (RMSE=0.19) and ANN (RMSE=0.11) alternatives. These outcomes suggest the potential applicability of the proposed predictive tool in weed management decision support systems design.

A comparative study between non-linear regression and artificial neural network approaches for modelling wild oat (Avena fatua) field emergence

SUMMARYNon-linear regression (NLR) techniques are used widely to fit weed field emergence patterns to soil microclimatic indices using S-type functions. Artificial neural networks (ANNs) present interesting and alternative features for such modelling purposes. In the present work, a univariate hydrothermal-time based Weibull model and a bivariate (hydro-time and thermal-time) ANN were developed to study wild oat emergence under non-moisture restriction conditions using data from different locations worldwide. Results indicated a higher accuracy of the neural network in comparison with the NLR approach due to the improved descriptive capacity of thermal-time and the hydro-time as independent explanatory variables. The bivariate ANN model outperformed the conventional Weibull approach, in terms of RMSE of the test set, by 70·8%. These outcomes suggest the potential applicability of the proposed modelling approach in the design of weed management decision support systems.

Modeling Avena fatua seedling emergence dynamics: An artificial neural network approach

Avena fatua is an invasive weed of the semiarid region of Argentina. Seedling emergence patterns are very irregular along the season showing a great year-to-year variability mainly due to a highly unpredictable precipitation regime. Non-linear regression techniques are usually unable to accurately predict field emergence under such environmental conditions. Artificial Neural Networks (ANNs) are known for their capacity to describe highly non-linear relationships among variables thus showing a high potential applicability in ecological systems. The objectives of the present work were to develop different ANN models for A. fatua seedling emergence prediction and to compare their predictive capability against non-linear regression techniques. Classical hydrothermal-time indices were used as input variable for the development of univariate models, while thermal-time and hydro-time were used as independent input variables for developing bivariate models. The accumulated proportion of seedling emergence was the output variable in all cases. A total of 528 input/output data pairs corresponding to 11years of data collection were used in this study. Obtained results indicate a higher accuracy and generalization performance of the optimal ANN model in comparison to non-linear regression approaches. It is also demonstrated that the use of thermal-time and hydro-time as independent explanatory variables in ANN models yields better prediction than using combined hydrothermal-time indices in classical NLR models. The best obtained ANN model outperformed in 43.3% the best NLR model in terms of RMSE of the test set. Moreover, the best obtained ANN predicted accumulated emergence within the first 50% of total emergence 48.3% better in average than the best developed NLR model. These outcomes suggest the potential applicability of the proposed modeling approach in weed management decision support systems design.

Predicted Soybean Yield Loss As Affected by Emergence Time of Mixed-Species Weed Communities

Potential crop yield loss due to early-season weed competition is an important risk associated with postemergence weed management programs. WeedSOFT is a weed management decision support system that has the potential to greatly reduce such risk. Previous research has shown that weed emergence time can greatly affect the accuracy of corn yield loss predictions by WeedSOFT, but our understanding of its predictive accuracy for soybean yield loss as affected by weed emergence time is limited. We conducted experiments at several sites across the Midwestern United States to assess accuracy of WeedSOFT predictions of soybean yield loss associated with mixed-species weed communities established at emergence (VE), cotyledon (VC), first-node (V1), or third-node (V3) soybean. Weed communities across research sites consisted mostly of annual grass species and moderately competitive annual broadleaf species. Soybean yield loss occurred in seven of nine site-years for weed communities established at VE soybean, four site-years for weed communities established at VC soybean, and one site-year for weed communities established at V1 soybean. No soybean yield loss was associated with weed communities established at the V3 stage. Nonlinear regression analyses of predicted and observed soybean yield data pooled over site-years showed that predicted yields were less than observed yields at all soybean growth stages, indicating overestimation of soybean yield loss. Pearson correlation analyses indicated that yield loss functions overestimated the competitive ability of high densities of giant and yellow foxtail with soybean, indicating that adjustments to competitive index values or yield loss function parameters for these species may improve soybean yield loss prediction accuracy and increase the usefulness of WeedSOFT as a weed management decision support system.

A model to simulate yield losses in winter wheat caused by weeds, for use in a weed management decision support system

The ‘within-season’ module of the Weed Manager decision support system (DSS) predicts the effect of twelve UK arable weeds on winter wheat yields and profitability. The model and decision algorithm that underpin the DSS are described and their performance discussed. The model comprises: (i) seedling germination and emergence, (ii) early growth, (iii) phenological development, (iv) herbicide and cultivation effects and (v) crop yield loss. Crop and weed emergence are predicted from the timing and method of cultivation, species biology, and the weather. Wheat and weeds compete for resources, and yield losses are predicted from their relative leaf area at canopy closure. Herbicides and cultural control methods reduce weed green area index, improving crop yield. A decision algorithm identifies economically successful weed management strategies based on model output. The output of the Weed Manager model and decision algorithm was extensively validated by experts, who confirmed the predicted responses to herbicide application were sufficiently accurate for practical use. Limited independent data were also used in the validation. The development of the module required integrating novel and existing approaches for simulating weed seedling establishment, plant development and decision algorithm design. Combining these within Weed Manager created a framework suitable for commercial use.

Using stochastic dynamic programming to support weed management decisions over a rotation

Benjamin LR, Milne AE, Parsons DJ, Cussans J & Lutman PJW (2009). Using stochastic dynamic programming to support weed management decisions over a rotation. Weed Research49, 207–216.SummaryThis study describes a model that predicts the impact of weed management on the population dynamics of arable weeds over a rotation and presents the economic consequences. A stochastic dynamic programming optimisation is applied to the model to identify the management strategy that maximises gross margin over the rotation. The model and dynamic programme were developed for the weed management decision support system –‘Weed Manager’. Users can investigate the effect of management practices (crop, sowing time, weed control and cultivation practices) on their most important weeds over the rotation or use the dynamic programme to evaluate the best theoretical weed management strategy. Examples of the output are given in this paper, along with discussion on their validation. Through this study, we demonstrate how biological models can (i) be integrated into a decision framework and (ii) deliver valuable weed management guidance to users.

The Ecology of Crop-Weed Interactions

Summary Understanding the ecology of crop-weed interactions includes a wide array of population and community ecology theory. The first thorough application of theory in Weed Ecology was predicting weed impacts on crops. The generality of crop yield response to weeds has been well supported with a wealth of empirical studies, however, inclusion of these models into weed management decision support systems (DSS) has been less fruitful. Parameterization of the empirical models has indicated extreme variability over space and time, even in experimental plots. Thus, further development and implementation of ecologically based weed management must depend on understanding sources of crop response variation, understanding complexity of interactions among the many sources of variation, and the realization that forecasting crop response will always have some level of uncertainty. Ecological theory may help us understand how to best construct conceptual and predictive models of crop-weed communities to accomplish our applied goals.

WeedSOFT®: a weed management decision support system

WeedSOFT® is a decision support system that was developed to help farmers and consultants in Nebraska with the selection of optimal weed management strategies. WeedSOFT® evolved from HERB, a bioeconomic model for soybean that was developed in North Carolina. The program is composed of four independent modules, namely, ADVISOR, EnviroFX, MapVIEW, and WeedVIEW. ADVISOR helps the user select a treatment based on maximum yield or maximum net gain. EnviroFX and MapVIEW provide environmentally relevant herbicide information and county soil maps that indicate vulnerability to groundwater contamination. WeedVIEW is a visual library of color images and line drawings of 46 common weed species. Over 500 farmers and consultants in Nebraska and adjacent states use WeedSOFT®. As a result of the current regionalization effort, the user base is expected to increase rapidly during the next 2 or 3 yr. This article explains the algorithms implemented in the current version of WeedSOFT®.