A decade of grapevine red blotch disease epidemiology reveals zonal roguing as novel disease management

Abstract

Red blotch disease, a threat to the grape industry, is caused by grapevine red blotch virus. This work is the first to study epidemiological patterns in a vineyard over the course of a decade, revealing an increase in disease incidence from 3.9% in 2014 to 36.4% in 2023 with rapid virus spread proximal to a transmission hotspot. Logistic and exponential models provided the best fit of spread in areas of high and low disease incidence and aggregation, respectively. An inverse spatial incidence of virus strains 1 and 2 suggested secondary spread mostly from diseased to neighboring vines and virus influx from background sources. Precipitation (3-4 years later) and air temperature (the same or 1 year later) significantly influenced epidemic parameters. Finally, asymptomatic infections contributed to spatial aggregations at increasing lags. These findings were salient for considering zonal roguing, the removal of diseased and surrounding vines, as a disease management option.

Fig. 1. Disease incidence and spatial statistics of a 2-hectare red blotch diseased ‘Cabernet franc’ vineyard in Napa Valley, California.

A Schematic images of the diseased vineyard in which vines exhibiting red blotch disease symptoms for the first time are indicated in red, those symptomatic in prior surveys in pink, and asymptomatic vines in green. Missing or dead vines are in white. Each map represents each year of study from 2014–2023. B Disease incidence (y) from 2014 to 2023. The x-axis represents years of the study, the y-axis represents the proportion of diseased vines over the total number of vines in the vineyard. Regression statistics values are listed. An exponential model best fit the disease epidemic with supporting statistics listed. C Index of dispersion from 2014 to 2023 using 2 × 2, 3 × 3, and 4 × 4 vine quadrats. The x-axis represents years of the study, the y-axis represents the index of dispersion (D) in which D > 1 indicates aggregated disease patterns. D Beta-binomial distribution fitting calculated using 2 × 2, 3 × 3, and 4 × 4 vine quadrats. The x-axis represents years of the study, the y-axis represents the measure of disease heterogeneity, θ. E Spatial Analysis by Distance IndicEs (SADIE) analysis using 2 × 2, 3 × 3, and 4 × 4 vine quadrats. The x-axis represents study year, the y-axis represents the index of aggregation I_a in which I_a > 1 indicates an aggregated pattern.

Fig. 2. Disease incidence, spatiotemporal statistics, and occurrence of asymptomatic infections in five subsections (A-E) of a 2-hectare red blotch diseased ‘Cabernet franc’ vineyard in Napa Valley, California from 2014 to 2023.

A Red blotch disease incidence map of the ‘Cabernet franc’ study vineyard in 2023 with five subsections in which every vine was tested for grapevine red blotch virus (GRBV) by PCR, and strains 1 and 2 by restriction fragment length polymorphism in 2022 and 2023. For subsections, vines infected with GRBV strain 1 shown in blue and strain 2 in pink. Vines testing positive for GRBV and symptomatic are depicted as a closed circle, those positive for GRBV and asymptomatic are shown as an ‘X’. Vines negative for GRBV by PCR and not exhibiting symptoms are shown in green, and open spaces indicate missing or dead vines. B Disease incidence by year for each subsection. Circles are observed values, and lines are predicted incidences using best-fitting models. The model of best fit and supporting statistics associated with symptomatic vines is depicted in each panel. Pairwise-t-tests were conducted to determine if the incidence of GRBV strain 1 compared to strain 2 was statistically significant, p values are provided for each subsection. The total number of infected vines represented by bar plots in 2023 includes symptomatic and asymptomatic infections. The y-axis represents the proportion of diseased vines over the total number of vines in the subsection. Non-linear regression and best-fitting model statistics are listed. C Index of dispersion (D) calculated each year of the study using 2 × 2 vine quadrats. D > 1 indicates aggregated disease patterns. D Aggregation index (θ) of the beta-binomial distribution. θ > 0 indicates an aggregated pattern. E Index of aggregation (I_a) of the Spatial Analysis by Distance IndicEs (SADIE) analysis. I_a > 1 indicates an aggregated pattern. For panels, B–E, vines infected with GRBV strain 1 are in blue, strain 2 in pink, and all infected vines in black. Statistical tests with p values less than the level of significance, α = 0.05, are represented by a closed circle, those greater than the level of significance are an open circle in C, D.

Fig. 3. Spatial relationship amongst the vines neighboring a single diseased vine across all five subsections of the 2-hectare red blotch diseased ‘Cabernet franc’ vineyard in Napa Valley, California.

Join-count analyses were conducted for all grapevine red blotch virus (GRBV)-infected vines (black), vines infected with a GRBV strain 1 (blue) or strain 2 (pink). Z-statistics less than -1.96 (closed circle) indicate vines surrounding an infected vine at the assigned spatial lag are likely to be infected. A lack of spatial aggregation, Z-statistics greater than -1.96, are represented by open circles. The x-axis represents the spatial lag, or the number of neighboring vines surrounding an infected vine, A within the same row, B across rows, or, C in a rook or diagonal pattern. The y-axis represents the Z-statistic of the join-count analysis.

Fig. 4. Pearson’s correlation coefficient depicting the association between weather data and epidemiological metrics with seasonal lag in a diseased ‘Cabernet franc’ vineyard in Napa Valley, California.

Climatic variables shown on the y-axis were correlated to disease incidence (y) and disease rate (dy/dt) found on the x-axis for spring (March–June) and summer (June–September). Results were calculated for each year of the study, 2014–2023 wherein associations with epidemic parameters were determined based on weather data of the same year (lag = 0), and seasonal lags one to four years prior to the reported epidemic parameters (lag = −1 and −4, respectively). Seasonal lags of −1, −2, −3, and −4 relied on weather data reported in 2013–2022, 2012–2021, 2011–2020, and 2010–2019, respectively. Red and blue colors represent a negative and positive correlation, respectively, with higher values of r represented by darker shades. Correlation coefficient values are presented if significant following the F distribution (p ≤ 0.05).

Fig. 5. Proposed zonal roguing in a red blotch diseased ‘Cabernet franc’ vineyard in Napa Valley, California.

A Red blotch disease incidence map in 2023 with five overlayed subsections in which every vine was assessed visually for symptoms and tested for grapevine red blotch (GRBV) by PCR and used for join-count analysis. B Proposed zonal roguing based on disease incidence and aggregation. When incidence is less than 20%, and vines are aggregated as in subsection A, roguing the diseased vine (shown in red) and six surrounding vines is recommended, two on each side within a row and one from each side across the row. When disease incidence is less than 20% and not aggregated, as observed in subsections B and C, roguing of eight vines in addition to the diseased vine is recommended: one vine on each side of the diseased vine within row, across row, and diagonal. When incidence is between 21 and 50% and aggregated, roguing the diseased vine and 24 surrounding vines is recommended: three vines on each side of the diseased vine within a row, across a row and diagonal. When incidence is >51%, roguing the diseased vine and 48 surrounding vines is recommended: six vines on each side of the diseased vine within a row, across a row, and diagonal. Despite these data being statistically supported, maintaining vineyards with incidence >30% is not recommended, based on data shown previously by bioeconomic studies.