Crop Protection Decisions Research Articles

Weather Variables for Within-Vineyard Awareness of Botrytis Risk

Background and Aims. Botrytis bunch rot (BBR) in cool temperate climates is a key constraint to the consistent supply of grapes to winery specifications. BBR severities have been correlated with specific environmental conditions; however, data-driven applications intended to support crop protection decisions are restricted in access and/or of unknown commercial value. The aims of this study were to evaluate variables providing within-vineyard awareness of BBR risk in Riesling vines. Methods and Results. Descriptors of BBR epidemics from eight site years, 2009–2014, were developed for vine areas of ∼0.5 ha within two regions of Tasmania with different climates. Two variables using the daily Bacchus index, from crop stage E-L 19 to E-L 31 or 34, accounted for >80% of the variance in the final mean BBR severity. A BBR risk index (BBR-I), incorporating the mean daily Bacchus index from E-L 19 to E-L 31 and the median daily vapour pressure deficit of air at 15 : 00 during the late-season interval, accounted for up to 99.5% of the variance in the final mean BBR severity. The late-season interval (days) or median daily RH (%) at 15:00 in the same period accounted for 86.4 or 83.3% of the variance. Spatial variability of BBR severity mapped in 4.8 ha of Sauvignon Blanc in 2018-19 confirmed the need to apply BBR risk indicators at an appropriate spatial scale. Conclusions. Environmental variables with biological relevance served as indicators of BBR risk at the study sites and have the potential to discriminate BBR risk among production regions in Tasmania. Significance of the Study. Study findings are expected to support the development of applications that raise awareness of BBR risk at an appropriate spatial scale for in-season adaption of crop protection, diagnosis of crop protection efficacy, and/or site selection decisions. Accompanying formulae with sample data in Microsoft® Excel will support transitions to automated data analyses.

A Deep-Learning-Based Detection Approach for the Identification of Insect Species of Economic Importance.

Artificial Intelligence (AI) and automation are fostering more sustainable and effective solutions for a wide spectrum of agricultural problems. Pest management is a major challenge for crop production that can benefit from machine learning techniques to detect and monitor specific pests and diseases. Traditional monitoring is labor intensive, time demanding, and expensive, while machine learning paradigms may support cost-effective crop protection decisions. However, previous studies mainly relied on morphological images of stationary or immobilized animals. Other features related to living animals behaving in the environment (e.g., walking trajectories, different postures, etc.) have been overlooked so far. In this study, we developed a detection method based on convolutional neural network (CNN) that can accurately classify in real-time two tephritid species (Ceratitis capitata and Bactrocera oleae) free to move and change their posture. Results showed a successful automatic detection (i.e., precision rate about 93%) in real-time of C. capitata and B. oleae adults using a camera sensor at a fixed height. In addition, the similar shape and movement patterns of the two insects did not interfere with the network precision. The proposed method can be extended to other pest species, needing minimal data pre-processing and similar architecture.

Integrating Experience, Evidence and Expertise in the Crop Protection Decision Process.

Generically, farm-scale crop protection decision making may be characterized as a process beginning with an initial assessment of disease risk followed by the accumulation of evidence related to current risk factors, leading to a risk prediction. What action is then taken depends on the response of the decision owner, taking into account previous experience, advice from trusted sources, alongside policy or legislative constraints on crop protection practice that are intended to mitigate any impacts that may transcend the farm scale. This process has commonalities with decision-making in the strategy of preventive medicine. This article delves into the clinical literature in order to provide a perspective on some recent discussions of shared decision making presented there, discussions that relate to issues also faced in sustainable crop protection.

Innovations in air sampling to detect plant pathogens.

Many innovations in the development and use of air sampling devices have occurred in plant pathology since the first description of the Hirst spore trap. These include improvements in capture efficiency at relatively high air-volume collection rates, methods to enhance the ease of sample processing with downstream diagnostic methods and even full automation of sampling, diagnosis and wireless reporting of results. Other innovations have been to mount air samplers on mobile platforms such as UAVs and ground vehicles to allow sampling at different altitudes and locations in a short space of time to identify potential sources and population structure. Geographical Information Systems and the application to a network of samplers can allow a greater prediction of airborne inoculum and dispersal dynamics. This field of technology is now developing quickly as novel diagnostic methods allow increasingly rapid and accurate quantifications of airborne species and genetic traits. Sampling and interpretation of results, particularly action-thresholds, is improved by understanding components of air dispersal and dilution processes and can add greater precision in the application of crop protection products as part of integrated pest and disease management decisions. The applications of air samplers are likely to increase, with much greater adoption by growers or industry support workers to aid in crop protection decisions. The same devices are likely to improve information available for detection of allergens causing hay fever and asthma or provide valuable metadata for regional plant disease dynamics.

Information graphs for binary predictors.

Binary predictors are used in a wide range of crop protection decision making applications. Such predictors provide a simple analytical apparatus for the formulation of evidence related to risk factors, for use in the process of Bayesian updating of probabilities of crop disease. For diagrammatic interpretation of diagnostic probabilities, the receiver operating characteristic is available. Here, we view binary predictors from the perspective of diagnostic information. After a brief introduction to the basic information theoretic concepts of entropy and expected mutual information, we use an example data set to provide diagrammatic interpretations of expected mutual information, relative entropy, information inaccuracy, information updating and specific information. Our information graphs also illustrate correspondences between diagnostic information and diagnostic probabilities.

The use of data mining to assist crop protection decisions on kiwifruit in New Zealand

Data mining algorithms were used to develop models to forecast the outcome of leafroller pest monitoring decisions on ‘Hayward’ kiwifruit crops in New Zealand. Using industry spray diary and pest monitoring data gathered at an orchard block level for compliance purposes, 80 attributes (independent variables) were created in three categories from the spray diary data: (1) individual insecticide applications applied during 2-week time windows, (2) groups of insecticide applications within time periods prior to or after fruit set and (3) orchard management attributes. Five machine learning algorithms (Decision Tree, Naïve Bayes, Random Forest, AdaBoost, Support Vector Machine) and one statistical method (Logistic regression) (classifiers) were used to develop models to forecast insecticide application decisions for leafroller control, by predicting whether pest monitoring results were above or below a spray threshold. Models to forecast 2011 spraying decisions were trained on 2008 and 2009 data and tested on 2010 data. Forecasts were made for spray and no-spray decisions based upon pre-determined acceptable rates of precision (proportion of correct decisions in test results). Orchard blocks in which a forecast could not be made to a prescribed degree of precision were recommended to be monitored, which is the normal practice. Spray decisions could not be forecast to an acceptable degree of precision, but decisions not to spray were successfully forecast for 49% of the blocks to a precision of 98% (AdaBoost) and 70% of the blocks to a precision of 95% (Naïve Bayes). Models with as few as four attributes gave useful forecasts, and orchard management attributes were the most important determinants of model forecasting accuracy. The potential for this methodology to assist with pest spray forecasting using customised data sets is discussed.

The structure of diagnostic information

Diagnosis is characterized as an exercise in classification, where the task is to assign a crop to a risk group as a basis for evidence-based crop protection decision making. Underlying the process of diagnostic decision making is Bayesian updating of probabilities. Alongside updating of probabilities, assessments of diagnostic information allow further description of the characteristics of diagnostic tests, and of the predictions made on the basis of test outcomes. This is illustrated analytically, graphically (by means of iso-information contour plots and information graphs) and by discussion of example epidemiological scenarios.

Soil heterogeneity at the field scale: a challenge for precision crop protection

Crop protection seldom takes into account soil heterogeneity at the field scale. Yet, variable site characteristics affect the incidence of pests as well as the efficacy and fate of pesticides in soil. This article reviews crucial starting points for incorporating soil information into precision crop protection (PCP). At present, the lack of adequate field maps is a major drawback. Conventional soil analyses are too expensive to capture soil heterogeneity at the field scale with the required spatial resolution. Therefore, we discuss alternative procedures exemplified by our own results concerning (i) minimally and non-invasive sensor techniques for the estimation of soil properties, (ii) the evidence of soil heterogeneity with respect to PCP, and (iii) current possibilities for incorporation of high resolution soil information into crop protection decisions. Soil organic carbon (SOC) and soil texture are extremely interesting for PCP. Their determination with minimally invasive techniques requires the sampling of soils, because the sensors must be used in the laboratory. However, this technique delivers precise information at low cost. We accurately determined SOC in the near-infrared. In the mid-infrared, texture and lime content were also exactly quantified. Non-invasive sensors require less effort. The airborne HyMap sensor was suitable for the detection of variability in SOC at high resolution, thus promising further progress regarding SOC data acquisition from bare soil. The apparent electrical conductivity as measured by an EM38 sensor was shown to be a suitable proxy for soil texture and layering. A survey of arable fields near Bonn (Germany) revealed widespread within-field heterogeneity of texture-related ECa, SOC and other characteristics. Maps of herbicide sorption and application rate were derived from sensor data, showing that optimal herbicide dosage is strongly governed by soil variability. A phytoassay with isoproturon confirmed the reliability of spatially varied herbicide application rates. Mapping areas with an enhanced leaching risk within fields allows them to be kept free of pesticides with related regulatory restrictions. We conclude that the use of information on soil heterogeneity within the concept of PCP is beneficial, both economically and ecologically.

Rapid Visual Estimates of Thrips (Thysanoptera: Thripidae) Densities on Cucumber and Rose Crops

Scouting techniques combining rapid counting methods must be developed to help growers with immediate decision making in integrated pest management programs. We evaluated a method for estimating densities of western flower thrips, Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae), one of the most damaging insect pests of greenhouse cucumber, Cucumis sativus L., and rose, Rosa x hybrida crops in southeastern France. This method is based on abundance classes of thrips observed on sampling units of flowers and foliage during a period of <1 min. Classes were calibrated using actual counts, and precision was improved by introducing additional predictive variables into multivariate nonparametric regression models. Regression models using infestation variables with and without climatic variables significantly increased calibration precision and made possible the accurate description of population dynamics. Rapid visual scouting methods could be combined for surveys of different pests and diseases. When calibrated, they provide growers or technicians with accurate tools guiding crop protection decisions.

Organic pest management decisions: a systems approach to technology adoption

AbstractOrganic farmers make system‐level crop protection decisions that combine complementary insect, disease, nematode, and weed management strategies. Data from a 1997 national survey of U.S. organic farmers were used in a multivariate count data model to identify the farm and regional factors influencing the adoption of practices across the linked pest management categories. The results show that weed management requires the greatest management effort by organic farmers. More intensive information‐seeking and on‐farm experimentation, higher educational attainment, and intensity of commitment to organic farming are positively related to the number of weed control strategies adopted. Predictions of adoption based on this model and customized to farm and region specifications will give information providers lead time to develop technical support for reduced chemical pest management systems.

Using economic incentives for pesticide usage reductions: responsiveness to input taxation and agricultural systems

There is growing interest in using environmental taxes to address the problems of agricultural pesticide contamination, given the potential of economic instruments for higher efficiency compared to regulatory approaches. However, research to date has suggested low producer responsiveness to input price changes. It is important to examine crop protection decisions and the options for adjustment in more detail. A case-study arable farm model is presented to evaluate the implications of new, currently experimental, low-input farming production systems for pesticide policy. If producers adhere to current systems, a pesticides tax at politically acceptable levels introduced as a stand-alone measure would perform poorly. Consistent with a number of studies, the model suggests that pesticide use levels could be reduced significantly while actually increasing farm income levels through conversion to low-input farming. Pesticide taxation cannot be viewed in isolation as a policy tool but should be part of a package of measures, including in particular education and training to encourage and assist farming system change.

Sampling for IPM Decision Making: Where Should We Invest Time and Resources?

ABSTRACT Guides for making crop protection decisions based on assessments of pest abundance or incidence are cornerstones of many integrated pest management systems. Much research has been devoted to developing sample plans for use in these guides. The development of sampling plans has usually focused on collecting information on the sampling distribution of the pest, describing this sampling distribution with a mathematical model, formulating a sample plan, and sometimes, but not always, evaluating the performance of the proposed sample plan. For crop protection decision making, classification of density or incidence is usually more appropriate than estimation. When classification is done, the average outcome of classification (the operating characteristic) is frequently robust to large changes in the sampling distribution, including estimates of the variance of pest counts, and to sample size. In contrast, the critical density, or critical incidence, about which classifications are made, has a large influence on the operating characteristic. We suggest that rather than investing resources in elaborate descriptions of sampling distributions, or in fine-tuning sample size to achieve desired levels of precision, greater emphasis should be placed on characterizing pest densities that signal the need for management action and on designing decision guides that will be adopted by practitioners.

Sampling for Decision Making in Crop Loss Assessment and Pest Management: Introduction

ABSTRACT Data obtained by sampling are crucial for decision making in crop loss assessment and pest management. Such data improve farmers' perceptions of the threat of pests and can, therefore, improve the quality of decision making in the practice of crop protection. The concept of a threshold, representing the dividing line between two alternative courses of action relating to seed or crop health, is an important aspect of crop protection decision making. Sampling is the means by which the required judgment can be guided. Operating characteristic curves are an important tool in the evaluation and comparison of the performance of sampling schemes. Precision integrated pest management, in which the objective is spatially variable pest management within fields, poses new problems for decision makers and statisticians developing sampling methodology in the context of crop protection.

Voluntary Economic and Environmental Risk Tradeoffs in Crop Protection Decisions

An indirect utility model is employed for measuring farmers’ willingness to voluntarily accept yield losses for a reduction in environmental risk by decreasing pesticide use. Results support the hypothesis that farmers have self-described risk perceptions that enable them to make assessments of risk-yield tradeoffs. Policies designed to encourage and assist farmers making voluntary pesticide reductions can result in environmental risk reduction.

Efficacité privée et publique de la gestion du risque phytosanitaire : le rôle de l'information

The analysis of farmers' phytosanitary information use requires the specification of information as an input that is produced and used on the farm. The information input is assumed to consist in a signal that is correlated with (future) pest infestation of the farmers' fields. This signal is received by farmers before they choose their pest applications. The quality of information is measured by the accuracy of the signal in predicting the actual infestation. Using this approach it is easily shown that farmers' attitude toward risk, information quality and pesticide application costs play crucial roles in farmers' crop protection decisions. In order to study farmers' information use decisions we also consider the production cost of this input for farmers. Within this framework, although they appear inefficient from a public standpoint, it is assumed that systematic pesticide applications may be chosen by economically efficient farmers. Showing that the information production cost is likely to be high for a large part of French farmers' population we conclude that this cost is likely to be one of the most determinant factor explaining the current use of systematic pesticide applications by French farmers. Despite their weaknesses, the models developed in this paper provide some guidelines for pesticide use reduction policy design. Within the current European economic context, the use of information relative to pest infestation should result in a decrease in farmers' pesticide uses. This result is, to a large extent, supported by the fact that pesticides appear relatively cheap. As a consequence, the implementation of instruments aimed at the promotion of the use of phytosanitary information could be used as a policy intended to decrease farmers' pesticide applications. However, the promotion of information use can only be viewed as a medium or long term policy. Infortnation use by farmers requires adjustments in quasi-fixed factors (human capital, labour) and/or structural adjustments (creations of public or private firms producing expertise services). It is also shown that farmers' willingness to pay for phytosanitary information is an increasing function of pest application costs. This result implies that farmers' pesticide demand elasticity could be important (in absolute value) in the long run, i.e. due to adjustments in information uses. It also implies that a tax on the price of pesticide would strengthen the effects of policies aimed at reducing the information production costs for farmers.

Monitoring Pest Abundance by Cascading Density Classification

Crop protection decisions often are based on classifying pest abundance with respect to a predetermined threshold. The performance of sampling plans used for this purpose can be compared based on the cost of taking a sample and the cost of misclassifying pest abundance and, thus, making an inappropriate crop protection decision. The probability that a sample will classify a pest population into one category or another can be summarized by the operating characteristic (OC) function. The operating characteristic function shows, for any true mean, the probability that density will be classified as less than the threshold. When sampling plans are used repeatedly to monitor population density over an extended period of time, the probability of classifying populations as being above the nominal threshold increases, and this could result in unnecessary intervention. Sampling plans that categorize density into 3 classes (tripartite classification plans) rather than into 2 classes (dichotomous plans) provide more information and are more effective constituents of monitoring protocols. When tripartite plans are used to monitor population density, the overall probability of never recommending intervention over an entire monitoring period is less influenced by the process of repeated sampling. Two principles can guide selection of tripartite classification sampling plans for use in a monitoring protocol. First, thresholds must often be raised if erroneous decisions to intervene are to be minimized. Second, because monitoring with tripartite classification plans allows more flexibility for correcting (at subsequent sampling times) a wrong decision not to intervene, lower precision of constituent sampling plans (e.g., smaller sample size) often may be acceptable.

Damage and economic injury levels of eriophyid and tetranychid mites on grapes in Czechoslovakia

The relationships between population densities ofCalepitrimerus vitis (Nalepa),Colomerus vitis (Pagenstecher) andTetranychus urticae (Koch), and grape weight and sugar content were studied during 3-year-long trials. In the case ofCalepitrimerus vitis, the influence of population density on leaf area, length of spring shoots and number of grapes in the following year was studied. Significant relationships between population densities of bothCalepitrimerus vitis andT. urticae in late summer and grape weight were determined. The sugar content of the grapes was markedly less influenced by mite infestation. The negative effect ofColomerus vitis on grape weight was not confirmed. Significant relationships for crop-protection decisions are presented.

The Role of Forecasting in Crop Protection Decision Making: an Economic Viewpoint1

In making crop protection decisions, farmers have to answer two questions: what type of control measure(s) should I adopt? How should I apply this (these) measure (s), when, how frequently, etc.? The answer to these questions will depend upon: a) the dimensions of pest attack and the damage it causes, b) farmers' goals, c) the range of protection measures available to them which they can use, and d) the information available, in the form of monitored and forecast information. It is within this context that the role of forecasting is considered. An analysis is made of the possible impact that forecast infomation can have on farmer decision‐making, particularly with regard to the accuracy of the forecast and the time at which it is available. This leads on to the problem of assessing the value of a forecast at a regional level, against which the costs of providing such information can be set. Finally, situations in which forecast information is likely to be of greatest value (and of least value) are identified, and implications are drawn concerning the institutional and other changes that could be taken to allow better use to be made of forecast information.Lorsque les agriculteurs decident de proteger leurs cultures, ils doivent repondre aux deux questions suivantes: quel est le mode d'intervention a choisir et comment, quand et a frequence faut‐il intervenir. La reponse dependra notamment: de I'etendue de I'infestation et de L'importance des deglts; des buts que s'est assignes L'arboriculteur; de la gamme de methodes dont il dispose; des previsions qui lui sont fournies. La prevision est condideree dans ce cadre, puisqu'elle est susceptible d'infuler sur les decisions de L'agriculteur dans la mesure ou elle est etablie avec precision et au bon moment. I1 est donc utile d'evaluer L'inter6t economique d'une prevision regionale en faisant intervenir le cofit de celle‐ci. Enfin, il est fait etat de situations dans lesquelles la prevision est profitable ou non. Suivant les cas, certaines modifications structurelles peuvent s'imposer afin de mieux tirer benefice des previsions.

Analysis of decision making in crop protection

The crop protection decisions taken by a farmer are determined by four main factors: the farmer's objectives, his perception of past attack and the damage it can cause, the control measures available to him, and the decision rules by which he operates. In analysing the way in which these factors interact, the purpose of this paper is to increase our understanding of the crop protection decision problem and to show how an appreciation of the farmer's situation is important in designing successful research and extension programs. The paper is divided into two parts. In the first part, two classes of decision model are illustrated by means of case studies. For the economic threshold model, requiring specific information on pest attack and using conventional economic analysis, the potato cyst eelworm problem is used as an example. The prophylactic decision model, which uses information on the probability distribution of attack and relies more on decision theory techniques, is illustrated by reference to the problem of potato blight. It is recognised, however, that in many real life situations these decision models are not appropriate, the result of inadequate information, limited managerial ability, or extreme risk constraints. In this situation, farmers do not attempt to optimise but search for a crop protection strategy that is satisfactory, one that meets their objectives as well as their operational constraints. With this view of the farmer's situation, the second part of the paper concentrates on the way in which research and extension programs can improve it. Attention is focussed on the three methods of achieving this; by widening the range of control measures available, increasing the perception of relevant information, and by improving decision rules. In concluding, it is stressed that a successful reduction in a particular constraint to better decision making is only likely to be achieved where the measures taken are designed in the context of those constraints that remain.