Bio-economic Analysis of Grape Leafroll Virus Epidemics in California
Objectives of Proposed Research: 1. Analyze historical data maps of leafroll symptoms to elucidate quantitative parameters summarizing disease progress in time and space at an individual vineyard scale and at a landscape scale. 2. Develop a robust simulation model of disease spread based on the parameters elucidated in objective 1. 3. Validate the simulation analysis by comparing simulated disease dynamics against data collected from surveys of a range of vineyard types and an existing validation data set. 4.Use the simulation model in conjunction with financial data for grape production to explore a range of disease management tactics appropriate to different market sectors. 5.Analyze variation in levels of knowledge and opinion among growers concerning the impact of leafroll disease on grape/wine quality and vine economic life expectancy. 6.Explore the potential for developing area-wide disease management plans by facilitating grower focus groups for leafroll disease. Executive Summary: Progress has been made on six objectives of the project. The overall project is composed of two sub¬projects, the first concerning the basic epidemiology of the disease and its impacts at the individual planting block scale (Objectives 1 – 4), the second concerning the human-disease interaction (Objectives 5 and 6). Leafroll Epidemiology Objective 1: Analyses of historical data sets from the Napa Valley and other locations have revealed a consistency in the spatial and temporal statistical properties of leafroll epidemics that strongly suggests the same dispersal mechanism operates in all leafroll epidemics and that such epidemics will behave in a way that is predictable in general terms. Across four published studies (two from Spain, one from New Zealand, one from Napa) re-analysis of the original data suggest that leafroll disease incidence increased at an average rate of just under 12% of plants per year (mean percentage increase = 11.7%, minimum = 4.4%, maximum = 20.1%). The rate parameter of the logistic growth function estimated from the pooled disease incidence data was estimated to be 0.58. Starting from a completely healthy condition, with increase in disease incidence determined by these parameters it would take 8 years for disease incidence to reach 50% and after 15 years 99% of plants would be diseased. Analysis of the spatial pattern of disease incidence using a quadrat-based approach indicated that leafroll epidemics have an effective sample size of approximately 3 plants. This effective sample size is robust across studies involving epidemics covering different ranges of disease incidence, different sizes of vineyard, and different grape cultivars in different locations. The results indicate that leafroll has a highly patchy spatial pattern and a very short average dispersal distance within vine blocks. The consistency of the relationship is good news for efforts to develop robust simulations of the disease and for making predictions of disease impacts. However, the patchy nature of the disease has negative effects on sampling accuracy, and will make accurate estimates of true disease incidence relatively difficult to obtain at the outset of epidemics. Objective 2 The temporal and spatial statistics estimated from the historical data, together with simple generic rules about relationships between disease incidence measurements at different scales  provide the basic components of the simulation model for generating leafroll development patterns. The simulation model generates blocks of plants in which disease incidence increases according to a logistic growth curve and the pattern of diseased plants in the block is constrained in two ways. Following the analyses of published data the frequency distribution of diseased plants per quadrat in the simulated data is assumed to follow a time-dependent binomial distribution in which the sample size is 3 plants. Objective 3 ? Work is currently in progress to compare the simulation model with validation data collected in the Napa Valley. Objective 4 ? With the format of the simulation established we will be able to attach financial values to disease on a per plant basis and project revenue and costs for infected blocks according to using either generic information on costs or specific information for particular blocks of vines in specific locations. Human-disease interactions Objective 5 & 6 Research on objectives 5 and 6 was combined in a single study based on a Q-method approach. Three data generating workshops were held with key stakeholders in the Napa Valley grape growing and wine-making sectors. At each workshop participants were asked to write answers to a set of open-ended questions concerning the importance of leafroll and the different approaches that can be used to manage it, including cooperative pest management in neighborhood groups. We sorted the responses into subject area groups and generated a set of 47 response statements that summarized viewpoints expressed in the written responses. Email and telephone contacts were used to generate a second set of 37 respondents (with some overlap to the original workshop participants) who were invited to take part in the Q-sort exercise. In each Q-sort interview a respondent was asked to rank each of the statements on a scale from -5 (‘does not reflect my opinion’) to 5 (‘completely represents my opinion’). During the sorting the number of statements allowed in each category was constrained so the respondents were forced to rank the statements according to their own viewpoint. The analysis revealed a generally high level of awareness of the seriousness of the GLRaV among respondents, but a wide variety of views about the most important factors in the disease and how it should be managed. A small group of statements, focusing on the importance, availability and reliability of virus-tested planting material ranked highest of all statements (scoring about 75% of the maximum possible score achievable if all respondents had rated them in maximum accordance with their opinion). The lowest scoring statement concerned the relative importance of irrigation and leafroll to vine health (35% of maximum possible). Statements concerning the value of cooperation for leafroll management tended to score close to the average and did not have particularly high variances. These results indicate neither generally strongly favorable nor negative views about cooperation. Principal components analysis of the responses suggested that while each person had an essentially unique viewpoint there was some grouping among the respondents. We are currently analyzing the data in more detail, but the initial interpretation of the analysis is that the respondents can be classified by their viewpoints as: focused either on the financial aspects of the disease or its technical implications; either having an open-ended view of the problem or seeing it as an issue with a definite end-point; either being essentially strategic in approach or essentially tactical; either being optimistic or fatalistic.