In late 2000, the European Union adopted the Water Framework Directive (WFD) and funded a series of research and innovation projects to support its implementation. One of these was the MULINO project (MULti-sectoral, INtegrated and Operational Decision Support System for Sustainable Use of Water Resources at the Catchment Scale). Its main product was a decision support system (mDSS) tool designed to help water managers make choices related to WFD implementation in a participatory manner. After the end of MULINO, a long sequence of research projects allowed for the maintenance and continuous development of its tool, which has been applied for more than 20 years in various contexts related to environmental and integrated management. This experience and an analysis of the literature allow us to draw some general conclusions regarding DSS tools for water management and their role in our societies. Lessons learned are proposed, from the need to frame tools within sound methodological frameworks for the management of decision processes, supporting instead of substituting decision-makers in their roles, to the trade-offs that appear between ease of use and specificity on one side and flexibility and reusability on the other. The specific strengths attributed to mDSS include the provision of an interface based on a simplified and understandable conceptual framework that facilitates communication with interested parties, the flexibility and ability to approach a wide variety of decisional issues, the relatively simple and understandable decision rules provided by the tool, and the simplified connections with other software environments. This paper presents the current version of the software and reports on the experience of its development and use over more than two decades; it also identifies the way forward.

A computer-based model was constructed to assess enrichment materials (EMats) for intensively-farmed weaned, growing and fattening pigs on a scale from 0 to 10. This model, called RICHPIG, was constructed in order to support the further implementation of EC Directive 2001/93/EC, which states that “pigs must have permanent access to a sufficient quantity of material to enable proper investigation and manipulation activities”. This paper describes the underlying conceptual framework for assessing EMats and explains the concepts, procedures and calculation rules used for semantic modelling. A (parsimonious) weighted average calculation rule was used to calculate enrichment scores from assessment criteria scores (which specify welfare relevant material properties of EMats) and weighting factors (WFs, which specify the relative importance of the assessment criteria). In total, 30 assessment criteria were identified and classified as object design criteria (eg novelty and accessibility), behavioural elements (eg nose, root, chew), biological functions (explore and forage), manipulations (ie object-directed behaviours), other (non-manipulative) consequences (eg aggression and stress) and object performance criteria (eg changeability/destructibility and hygiene). WFs were calculated from a systematic analysis of 573 scientific statements collected in the database, using 11 so-called weighting categories (Wcat, ie scientific paradigms to assess welfare such as the study of natural behaviour, consumer demand studies and stress-physiology) to assign Wcat level scores (which indicate the intensity, duration and incidence of a welfare impact) to the assessment criteria. The main advantages of the RICHPIG model are that it is based explicitly on available scientific information, that it has an explicitly formulated conceptual framework, is transparent, disputable, upgradeable, robust and reasonably in accordance with expert opinion. Major scope for improvements exist in the form of the need for further upgrading with new knowledge, empirical validation and (further) implementation in political decision-making processes.

Animals have various behavioural and physiological needs that are important for welfare. Fulfilment of these needs depends on the quality of housing, management and animal characteristics. The objective of this study was to develop a model to assign welfare scores to husbandry systems for dairy cattle, based on scientific results, and thereby supporting the design of new, welfare-friendly systems. COWEL is a computer-based decision support system that contains attributes regarding housing and management conditions. These attributes are technical specifications that contain various technical units called levels. These levels are ranked from best-to-worst regarding welfare, based on scientific information about animal-based parameters. This information, inserted in the model as statements, was weighted depending on the impact it has on welfare by using weighting categories. Thereafter, a weighting factor was calculated for each attribute which determines how important an attribute is for welfare. The COWEL model contains 2,343 statements on dairy cattle welfare from 476 sources found during a literature survey. The model was applied to four husbandry systems, namely a tie-stall, cubicle housing, a straw yard and a pasture-based system. The welfare scores, calculated by COWEL for these husbandry systems, correspond with the general opinion about these systems. A tie-stall receives a low and a pasture-based system a high welfare score: 211 and 271, respectively. A husbandry system can receive a maximum of 313 on the welfare scale of COWEL. We conclude that COWEL can be used to rank husbandry systems on a welfare scale, and may be a useful tool to develop new, sustainable and welfare-friendly systems for dairy cattle.

Climate change is one of the most significant and pressing issues faced by humanity; it frequently results in major natural disasters, such as catastrophic floods, which require the establishment of effective management policies by local and national authorities. These policies involve complex multistep decision-making processes that require combined assessment of various sources of data by different stakeholders. Even though an abundance of data is being collected to monitor climate change and estimate its consequences on the society, the environment, and the economy, policy-making is still largely based on intuition rather than evidence due to lack of a structured approach for modeling the decision-making process and considering the appropriate use of data in every step of the process. The goal of this work is to introduce a novel decision support system that can guide policy makers through a structured data-driven decision-making process aiming to create policies for flood risk management. The proposed system is a multifacet platform that guides policy makers through five phases—inform, advise, monitor, evaluate, and revise—of the policy cycle. For each phase, different dashboards provide relevant information regarding the environmental, social, and economic conditions. To demonstrate the potential of the proposed system, we use it to assess a flood protection policy in the city of Vicenza, Italy. The results reveal the benefits and challenges of the proposed decision support tool for public administrations involved in flood risk management.

In areas where water is scarce, the use of regulated deficit irrigation, combined with decision support system tools, may decrease the impact of agriculture on natural water resources, as well as on energy consumption, thereby improving the profitability of farms. With this aim, the SUBSTOR-Potato model (incorporated in the DSSAT Program) was evaluated with a 2-year field test (2011 and 2012) conducted in a semi-arid area (Albacete, Spain) applying four irrigation levels (120, 100, 80 and 60% of irrigation requirements). Subsequently, the model was used for simulating the potato yield under several deficit irrigation strategies (ISs) during 30 years of a semi-arid climate (1988–2017) and determining the most profitable option. The considered ISs were deemed those most suitable from the yield and water productivity point of view by some authors. The model performance for tuber yield was satisfactory with an index of agreement >0.91 and errors between 0.71 and 3.06 × 103 kg/ha. The ISs simulated with SUBSTOR-Potato showed that slight deficit irrigation (5–10%) may increase the water productivity and profitability of the farms. Moreover, tuber formation (from onset of tuber initiation to harvest) was shown to be the most sensitive stage, therefore it is highly recommended to avoid deficit during this stage, which would cause a large reduction in yield (around 8 t/ha, depending on the level of deficit suffered by the crop).

The construction of future technological systems in work domains that do not yet exist, known as the envisioned world problem, is an increasingly important topic for designers, particularly given the rapid rate of technological advancement in the modern era. This paper first discusses the theoretical underpinnings of using cognitive work analysis (CWA) for developing a decision support system (DSS) situated within the envisioned world problem and recasts the problem as pathway-dependent processes. Using this pathway-dependent framework, each stage of the envisioning process is described to reveal how human factors experts can link existing work domains to envisioned instances. Finally, a case study example of the envisioning process that incorporates CWA modelling is demonstrated as it pertains to the advancement of the human spaceflight domain. As a result, this paper provides a unified treatment of the envisioned world problem with an end-to-end example of one approach to designing future technologies for future work domains.

This paper will present a dynamic Variable Rate Irrigation System developed by the University of Georgia. The system consists of the EZZone management zone delineation tool, the UGA Smart Sensor Array (UGA SSA) and an irrigation scheduling decision support tool. An experiment was conducted in 2015 and 2016 in two different peanut fields to evaluate the performance of using the UGA SSA to dynamically schedule Variable Rate Irrigation (VRI). For comparison reasons strips were designed within the fields. These strips were irrigated according to either UGA SSA or Irrigator Pro recommendations. The results showed that Irrigator Pro is a conservative irrigation method which results in high yields. On the other hand the UGA SSA recommendations worked very well with the VRI system and in both years it recommended an average of 25% less irrigation water than the Irrigator Pro.

This paper describes the design, realization, and application of a custom temperature sensor devoted to the monitoring of the temperature differential between the leaf and the air. This difference is strictly related to the plant water stress and can be used as an input information for an intelligent and flexible irrigation system. A wireless temperature sensor network can be thought as a decision support system used to start irrigation when effectively needed by the cultivation, thus saving water, pump fuel oil, and preventing plant illness caused by over-watering.

Field trials were established in 2005 and continued in 2006 to evaluate a conventional broadcast herbicide sprayer compared to a variable spray (sensor-activated) weed-sensing sprayer (WSS). The computer-based Herbicide Application Decision Support System (WebHADSS™) was used to determine a portion of the herbicides applied (based on herbicide efficacy and economics). Weed control, herbicide usage, crop yield, and net returns were compared across treatments. The broadcast applications were usually the most effective at controlling weeds. A PPI herbicide did not always improve weed control compared to treatments in which no PPI herbicide was applied. Variable treatments used less herbicide than the broadcast system in both years. Cotton lint yields in broadcast applications were similar to the weed-free check in both years of the study. Variable treatments often provided equivalent net returns (gross yield revenue less weed control cost) to the broadcast treatments. Although herbicide savings were observed in the variable treatments when compared to a broadcast system, a reduction in weed control was observed, indicating the need for future improvements of this system. A site-specific weed management program used in conjunction with WebHADSS™ may have potential in cotton production systems in the Texas Southern High Plains where weed densities are low.

Weed management decisions based on weed threshold models offer the opportunity to reduce herbicide use by allowing the possibility of forgoing treatment or lowering rates. Weed thresholds based on a relative leaf-cover model were tested during a 4-yr period at two locations. Two 1.62-ha fields, planted to conventional and glyphosate-resistant corn (2004, 2005, 2007) or soybean (2006), were divided in 900 m2 sections. Herbicides were applied postemergence to each of these sections with either variable rates based on weed thresholds, or constant full rates. Variable herbicide rates included: no application, half rate, or full rate. Relative weed cover values of 0.2 and 0.4 (corn) or 0.1 and 0.3 (soybean) served as thresholds for incremental rates. Digital images were used to evaluate the relative weed cover. Weed density was assessed before and after herbicide application. Weed seed production was estimated for two species in 2004 and 2005. No difference in crop yield, relative weed cover, weed density, or weed seed production was observed between conventional and glyphosate-resistant cropping systems. During the first year, herbicide use reduction was obtained (−85.4%) with marginal crop yield loss (5 to 15%). In the subsequent 3 yr, preherbicide weed densities increased and concomitant increases in relative weed cover values did not allow more than a 10% overall reduction in herbicide use. This threshold model designed to maintain crop yields within a given year did not allow significant reduction in herbicide use during the following 3 yr. Residual weed populations most likely replenished the seed bank to levels that allowed weed densities to increase afterward. Increased weed density over time in plots treated with full rates of herbicide every year also indicated that a single postemergence herbicide treatment was not sufficient to contain weed populations at low levels every year in this corn–soybean rotation.

WeedSOFT is a state-of-the-art decision support system for weed management in the north central region of the United States, but its accuracy to predict corn yield loss associated with later-emerging weed communities has not been adequately assessed. We conducted experiments in 2004 and 2005 to compare observed and predicted corn yield related to four establishment times of mixed-species weed communities for validation of competitive index modifier (CIM) values in WeedSOFT. Weed communities were established at VE, V2, V4, and V6 corn (emergence, second-leaf, fourth-leaf, and sixth-leaf stages, respectively), and consisted largely of annual grass and moderately competitive annual broadleaf species. Compared to weed-free corn, yield loss occurred in each of seven site-years for weed communities established at VE corn, but in only one site-year for communities established at V2 corn. No corn yield loss was associated with weed communities established at V4 or V6 corn. For communities established at VE corn, predicted corn yield differed from observed yield in all but one site year, with predicted yield less than observed yield in three site-years, and greater than observed yield in two site-years; however, nonlinear regression analyses of yield data pooled over site-years showed that fitted values were similar between predicted and observed yield. For communities established at V2 and V4 corn, predicted yield was less than observed yield in six and five site-years, respectively. For communities established at V6 corn, predicted yield was less than observed yield in three of six site-years, but was similar to observed yield in three of six site-years. These results indicated that the CIM values in WeedSOFT tended to overestimate the competitiveness of late-emerging weed communities. Corn yield data from a pooled analysis of all site-years were used to generate a revised set of growth stage CIM values, which improved the accuracy of predicted corn yield. These results should improve weed management decisions and reduce the need for herbicide applications to late-emerging weeds.

Seed production from weeds that are missed by herbicide application can affect future weed populations and management decisions. It may be possible to expand the utility of computerized weed management decision aids to include an estimate of weed seed production resulting from selected treatments based on crop yield potential. Field studies were conducted in soybean near Columbia, MO, to determine whether weed control recommendations based on crop yield potential from a computerized weed management decision aid influence weed seed production in two soybean row spacings. At approximately 28 d after planting, weed densities and heights were entered into WeedSOFT® to generate a list of treatments ranked by predicted crop yields. Treatments included: (1) highest predicted crop yield in a glyphosate-resistant system, (2) highest predicted crop yield in a nonglyphosate-resistant system, (3) a 10% yield reduction, (4) a 20% yield reduction, and (5) an untreated control. These treatments were applied to soybean grown in 38- and 76-cm rows. Treatments that provided 90% or higher control of an individual species at 22 d after treatment usually produced less seed than untreated checks. Weed seed production based on early-season herbicide efficacy showed a linear relationship and was relatively predictable (r2 ≥ 0.52) for the predominant weed species. For less dominant weed species, weed seed production was not strongly correlated (r2 ≤ 0.27) to early-season herbicide efficacy but apparently influenced by control of other weed species. Narrow row spacing reduced giant foxtail biomass both years but did not reduce common ragweed and ivyleaf morningglory biomass. Narrow rows did not decrease giant foxtail, common ragweed, and ivyleaf morningglory seed production.

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.

Approximately 50% of the genetically modified herbicide-resistant corn hectares in the United States are treated only with POST-applied herbicides for weed management. Although a high degree of efficacy can be obtained with POST-applied herbicides, delayed timing of application may result in substantial corn yield loss. Our goal was to characterize on-farm corn–weed communities prior to POST herbicide application and estimate potential corn-yield loss associated with early-season corn–weed competition. In 2008 and 2009, field surveys were conducted across 95 site-years in southern Wisconsin and recorded weed species, density, and height in addition to crop height, growth stage, and row spacing. WeedSOFT® was used to predict corn yield loss. Common lambsquarters, velvetleaf, dandelion, common ragweed, and Amaranthus species were the five most abundant broadleaf weed species across site-years, present in 92, 86, 59, 45, and 44% of all fields, respectively, at mean densities of 19, 3, 3, 4, and 3 plants m−2, respectively. Mean plant heights among these species were 17 cm or less. Grass and sedge species occurred in 96% of fields at a mean density of 25 plants m−2 and height of 7 cm. The mean and median of total weed density across site-years were 96 and 52 plants m−2, with heights of 14 and 13 cm, respectively. Mean predicted corn yield loss was 4.5% with a mean economic loss of $62 ha−1. However, predicted yield loss was greater than 5% on one-third of the site-years, with a maximum of 26%. These results indicate that delayed application of POST herbicides has led to corn yield loss due to early-season weed-crop competition on a substantial number of fields across southern Wisconsin, and suggest that management tactics need to be improved to protect corn yield potential fully.

Glyphosate applied POST can provide a high level of efficacy on many weed species in soybean, but delayed application beyond optimal weed growth stages might fail to fully protect yield potential. Further, we do not have a good understanding of the extent to which delayed glyphosate application and its associated yield loss is occurring on-farm. Our goal was to characterize on-farm weed communities in glyphosate-resistant soybean just prior to glyphosate application and estimate potential yield loss associated with early-season soybean-weed competition. In field surveys conducted across 64 site-yr in southern Wisconsin in 2008 and 2009, common lambsquarters, velvetleaf, dandelion, Polygonum spp., and Amaranthus spp. were the five most abundant broadleaf weed species across site-years, present in 92, 69, 64, 42, and 50% of all fields, respectively, at average densities of 14, 5, 5, 14, and 10 plants m−2, respectively. Average height of these species was 21 cm or less at or near the time of glyphosate application. Grass and sedge species occurred in 95% of fields at an average density of 41 plants m−2 and height of 21 cm. The mean and median values of total weed density across site-years were 101 and 41 plants m−2, with heights of 19 and 17 cm, respectively. Recommended height for treatment is 15 cm. Glyphosate application occurred on average at V3 to V4 soybean growth stage, which is later than V2 soybean typically targeted to protect yield. Average yield loss predicted by WeedSOFT® was 5% with a mean economic loss of $47 ha−1. Predicted yield loss was greater than 5% on one-fourth of the site-years, all of which were treated at V4 soybean or later. The maximum predicted yield loss was 27%. These results suggest that glyphosate was applied at weed height and soybean growth stages that were greater than optimal to protect yield in many fields across southern Wisconsin. A soil-residual herbicide applied PRE, or a more timely POST application of glyphosate would alleviate the majority of these losses.

Unique long-term historical emergence records were used to assess the association between weed seedling emergence and various elements of meteorological data. These elements included both temperature-based and rainfall-related variables in the 7-d periods before and during which emergence occurred. Five weed species (Stellaria media, Chenopodium album, Capsella bursa-pastoris, Matricaria perforata, and Veronica hederifolia) with contrasting emergence patterns were studied in disturbed soil. Logistic regression analysis was used to identify meteorological variables of interest and allowed their relative importance to be assessed and ranked. Logistic regression was further used to associate probabilities of emergence with observed levels of important individual meteorological elements. This approach enabled prediction of the probability of emergence following given meteorological conditions and hence an assessment of the risk of omitting weed control measures. Predictions were made based on single meteorological variables and compared with observed data. Results indicated that temperature was the dominant factor in predicting emergence. Soil moisture, while also important, was a secondary factor only becoming important once the species-specific temperature requirement had been satisfied. The potential for further development of the model is discussed.

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®.

On-farm trials were conducted in 16 North Carolina peanut fields to obtain estimates of scouting times and quality of herbicide recommendations for different weed scouting methods. The fields were monitored for weed species and population density using four scouting methods: windshield (estimate made from the edge of the field), whole-field (estimate based on walk through the field), range (weed densities rated on 1–5 scale at six locations in the field), and counts (weeds estimated by counting at six locations in the field). The herbicide application decision support system (HADSS) was used to determine theoretical net return over herbicide investment and yield loss ($ and %) for each treatment in each field. Three scouts estimated average weed population densities using each scouting method. These values were entered into HADSS to obtain treatment recommendations. Independently collected count data from all three scouts were combined to determine the optimal treatment in each field and the relative ranking of each available treatment. When using the whole-field method, scouts observed a greater number of weed species than when using the other methods. The windshield, whole-field, and range scouting methods tended to overestimate density slightly at low densities and underestimate density substantially at high densities, compared to the count method. The windshield method required the least amount of time to complete (6 min per field), but also resulted in the greatest average loss. Even for this method, recommendations had theoretical net returns within 10% of the return for the optimal treatment 80% of the time. The count method appears to have less economic risk than the windshield, whole-field, and range scouting methods.

A field study was conducted at two sites in southwestern Ontario, Canada, from 2002 to 2005 to assess the cumulative effects of different weed management strategies applied to the same plots over time in a corn–soybean rotation. These strategies included a POST broad-spectrum herbicide applied at recommended or reduced rates and use of a computer decision aid to select the POST herbicide having the greatest net return (economic threshold approach) based on weed populations before treatment or seed-bank estimates. Four years after study initiation, increased weed populations were observed only in the nontreated control. The reduced rate and economic threshold approaches did not lead to increased weed seed banks compared with the recommended rate of a broad-spectrum herbicide. Reductions in herbicide rate were more effective than the use of economic thresholds for reducing the risk of environmental impact in corn and soybeans. Weed populations in this study were generally above threshold levels, and herbicides selected by the computer decision aid tended to be older, less-expensive herbicides with high use rates. Recommendations by the computer decision aid were generally more effective and more profitable when based on weed populations before treatment than when based on seed-bank estimates.

RIM, or “Ryegrass Integrated Management,” is a user-friendly weed management software that integrates long-term economics. As a model-based decision support system, RIM enables users to easily build 10-year cropping scenarios and evaluate the impacts of management choices on annual rigid ryegrass populations and long-term profitability. Best used in a workshop format to enable learning through interactions, RIM can provide insights for the sustainable management of ryegrass through “what-if” scenarios in regions facing herbicide resistance issues. The upgrade of RIM is presented, with changes justified from an end-user perspective. The implementation of the model in a new, intuitive software format is presented, as well as the revision, update, and documentation of over 40 management options. Enterprises, establishment systems, and control options were redefined to represent current practices, with the notable inclusion of customizable herbicide options and techniques for weed seed control at harvest. Several examples of how RIM can be used with farmers to demonstrate the benefits of adopting recommended practices for managing or delaying the onset of herbicide resistance are presented. Originally designed for the dryland broadacre systems of the Australian southern grainbelt, RIM's underlying modeling was restructured to facilitate future updates and adaptation to other weed species and cropping regions.