Plant Disease Models and Forecasting: Changes in Principles and Applications over the Last 50 Years.

This review gives a perspective of selected advances made since the middle of the 20th century in plant disease modeling, and the associated increase in the number of models published during that time frame. This progress can be mainly attributed to advances in (i) sensors and automatic environmental data collection technology, (ii) instrumentation and methods for studying botanical epidemiology, and (iii) data analytics and computer science. We review the evolution of techniques for developing data-based (empirical) models and process-based (mechanistic) models using the wheat rusts as a case study. We also describe the increased importance of knowledge about biological processes for plant disease modeling by using apple scab as a second case study. For both wheat rusts and apple scab, we describe how the models have evolved over the last 50 years by considering certain milestones that have been achieved in disease modeling. Finally, we describe how plant disease models are used as part of a multi-modeling approach to develop decision-making tools in the application of integrated pest management.

Development and Validation of a Mechanistic Model That Predicts Infection by Diaporthe ampelina, the Causal Agent of Phomopsis Cane and Leaf Spot of Grapevines.

Phomopsis cane and leaf spot (PCLS), known in Europe as "excoriose," is an important fungal disease of grapevines caused by Diaporthe spp., and most often by Diaporthe ampelina (synonym Phomopsis viticola). PCLS is re-emerging worldwide, likely due to climate change, changes in the management of downy mildew from calendar- to risk-based criteria that eliminate early-season (unnecessary) sprays, and the progressive reduction in the application of broad-spectrum fungicides. In this study, a mechanistic model for D. ampelina infection was developed based on published information. The model accounts for the following processes: (i) overwintering and maturation of pycnidia on affected canes; (ii) dispersal of alpha conidia to shoots and leaves; (iii) infection; and (iv) onset of disease symptoms. The model uses weather and host phenology to predict infection periods and disease progress during the season. Model output was validated against 11 independent PCLS epidemics that occurred in Italy (4 vineyards in 2019 and 2020) and Montenegro (3 vineyards in 2020). The model accurately predicted PCLS disease progress, with a concordance correlation coefficient (CCC) = 0.925 between observed and predicted data. A ROC analysis (AUROC>0.7) confirmed the ability of the model to predict the infection periods leading to an increase in PCLS severity in the field, indicating that growers could use the model to perform risk-based fungicide applications.

A Systematic Map of the Research on Disease Modelling for Agricultural Crops Worldwide.

In this work, we developed a systematic map to identify and catalogue the literature pertaining to disease modelling for agricultural crops worldwide. Searches were performed in 2021 in the Web of Science and Scopus for papers reporting any type of disease model for 103 crops. In total, 768 papers were retrieved, and their descriptive metadata were extracted. The number of papers found increased from the mid-1900s to 2020, and most of the studies were from North America and Europe. More disease models were retrieved for wheat, potatoes, grapes, and apples than for other crops; the number of papers was more affected by the crop's economic value than by its cultivated area. The systematic map revealed an underrepresentation of disease models for maize and rice, which is not justified by either the crop economic value or by disease impact. Most of the models were developed to understand the pathosystem, and fewer were developed for tactical disease management, strategic planning, or scenario analysis. The systematic map highlights a variety of knowledge gaps and suggests questions that warrant further research.

Dynamics of Diaporthe ampelina Conidia Released from Grape Canes that Overwintered in the Vineyard.

Phomopsis cane and leaf spot (PCLS) is an important disease of grapevines that is mainly caused by Diaporthe ampelina. Dispersal dynamics of D. ampelina spores were investigated in two vineyards, one in northern Italy and one in Montenegro, by using spore samplers that collected α- and ß-conidia from rain water running off from PCLS-affected canes. The canes were collected from each vineyard, deployed, and overwintered in the corresponding vineyards. In each of three years (2016, 2017, and 2018), conidial dispersal was investigated during one (Montenegro) or two (Italy) growing seasons following the deployment of the PCLS-affected canes. In the first growing season following cane deployment in both vineyards, α-conidia were mostly found in runoff water after grapevine bud break, especially in April and May, and ß-conidia were regularly found in numbers comparable to those of α-conidia, most frequently from June to September. In Italy, high numbers of α- and ß-conidia were also collected during the second growing season following cane deployment. The dispersal dynamics of α-conidia over time were described by a Gompertz equation using hydrothermal time (i.e., the accumulated effect of temperature on the maturation rate of pycnidia on days in which the number of hours of wetness was ≥6 or 9 h), with R2 and concordance correlation coefficient >0.9. Rain (≥0.2 mm) was a good predictor of conidial dispersal, with an overall accuracy of 0.97. These results increase our understanding of D. ampelina spore dispersal and should be integrated into warning systems for PCLS management.

Temporal Dispersal Patterns of Phaeomoniella chlamydospora, Causal Agent of Petri Disease and Esca, in Vineyards.

Although the fungus Phaeomoniella chlamydospora is the most commonly detected causal agent of Petri disease and esca, two important fungal grapevine trunk diseases, little is known about the dispersal patterns of P. chlamydospora inoculum. In this work, we studied the dispersal of P. chlamydospora airborne inoculum from 2016 to 2018 in two viticultural areas of eastern (Ontinyent) and northern (Logroño) Spain. The vineyards were monitored weekly from November to April using microscope slide traps, and P. chlamydospora was detected and quantified by a specific real-time quantitative (qPCR) method set up in this work. The method was found to be sensitive, and a good correlation was observed between numbers of P. chlamydospora conidia (counted by microscope) and DNA copy numbers (quantified by qPCR). We consistently detected DNA of P. chlamydospora at both locations and in all seasons but in different quantities. In most cases, DNA was first detected in the last half of November, and most of the DNA was detected from December to early April. When rain was used as a predictor of P. chlamydospora DNA detection in traps, false-negative detections were observed, but these involved only 4% of the total. The dispersal pattern of P. chlamydospora DNA over time was best described (R2 = 0.765 and concordance correlation coefficient = 0.870) by a Gompertz equation, with time expressed as hydrothermal time (a physiological time accounting for the effects of temperature and rain). This equation could be used to predict periods with a high risk of dispersal of P. chlamydospora.

Consideration of Latent Infections Improves the Prediction of Botrytis Bunch Rot Severity in Vineyards.

The current study validated a mechanistic model for Botrytis cinerea on grapevine with data from 23 independent Botrytis bunch rot (BBR) epidemics (combinations of vineyards × year) that occurred between 1997 and 2018 in Italy, France, and Spain. The model was operated for each vineyard by using weather data and vine growth stages to anticipate, at any day of the vine-growing season, the disease severity (DS) at harvest (severe, DS ≥ 15%; intermediate, 5 < DS < 15%; and mild, DS ≤ 5%). To determine the ability of the model to account for latent infections, postharvest incubation assays were also conducted using mature berries without symptoms or signs of BBR. The model correctly classified the severity of 15 of 23 epidemics (65% of epidemics) when the classification was based on field assessments of BBR severity; when the model was operated to include BBR severity after incubation assays, its ability to correctly predict BBR severity increased from 65% to >87%. This result showed that the model correctly accounts for latent infections, which is important because latent infections can substantially increase DS. The model was sensitive and specific, with the false-positive and false-negative proportion of model predictions equal to 0.24 and 0, respectively. Therefore, the model may be considered a reliable tool for decision-making for BBR control in vineyards.

Reduction of Botrytis cinerea Colonization of and Sporulation on Bunch Trash.

Botrytis bunch rot (BBR) of grapevine, caused by Botrytis cinerea, is commonly managed by fungicide (FUN) sprays at flowering (A), at prebunch closure (B), at veraison (C), and before harvest. Applications at A, B, and C are recommended to reduce B. cinerea colonization of bunch trash and the production of conidia during berry ripening. The effects of these applications were previously evaluated as reductions in BBR severity at harvest rather than as reductions in bunch trash colonization and sporulation by B. cinerea. This study investigated the effects of FUNs (a commercial mixture of fludioxonil and cyprodonil), biological control agents (BCAs; Aureobasium pullulans and Trichoderma atroviride), and botanicals (BOTs; a commercial mixture of eugenol, geraniol, and thymol) applied at different timings (A, B, C, or ABC) compared with a nontreated control (NT) on B. cinerea bunch trash colonization and sporulation in vineyards. The ability of B. cinerea to colonize the bunch trash (as indicated by B. cinerea DNA content) and sporulate (as indicated by the number of conidia produced under optimal laboratory conditions) was highly variable, and this variability was higher between years (2015 to 2018) than among the three vineyards and three sampling times (i.e., 1 week after applications at A, B, and C). B. cinerea sporulation on bunch trash was significantly lower in plots treated with FUN than in NT in only 3 of 18 cases (3 vineyards × 2 years × 3 sampling times). FUN applications, however, significantly reduced B. cinerea colonization of bunch trash compared with NT; for colonization, BCA efficacy was similar to that of FUN, but BOT efficacy was variable. For all products, colonization reduction was the same with application at A versus ABC, meaning that the effect of an early season application lasted from flowering to 1 week after veraison. These results indicate that the early season control of B. cinerea is important to reduce the saprophytic colonization of bunch trash, especially when the risk of BBR is high.

A network meta-analysis provides new insight into fungicide scheduling for the control of Botrytis cinerea in vineyards.

BACKGROUND: Control of Botrytis bunch rot (BBR) is currently based on the application of fungicides at four timings corresponding to specific growth stages of vines: end of flowering (A), pre-bunch closure (B), veraison (C) and before harvest (D). The current research provides a network meta-analysis of 116 studies conducted between 1963 and 2016 in nine countries, in which 14 strategies (based on combinations of 1, 2, 3, or 4 sprays applied in A, B, C, and/or D) were compared. RESULTS: When a one-spray strategy was applied, BBR control was more effective with sprays applied in A, C, or D than B. With a two-spray strategy, strategy AC provided similar control as strategy BC; strategy CD also provided good control. For a 3-spray strategy, the best disease control was consistently obtained with strategy ACD. Four-spray strategy ABCD provided the best control but often involved needless sprays so that the routine application of four sprays is not justified. CONCLUSIONS: Spraying at timing A seems to be very important for achieving efficient and flexible disease control. Flexibility is reduced by spraying at timing B rather than A. © 2018 Society of Chemical Industry.

Pest categorisation of Apiosporina morbosa.

The Panel on Plant Health performed a pest categorisation of the fungus Apiosporina morbosa, the causal agent of black knot, for the EU. The identity of the pest is well established and reliable methods exist for its detection/identification. The pest is listed in Annex IIAI of Directive 2000/29/EC and is not known to occur in the EU. Apiosporina morbosa is present in Alaska, Canada, Mexico and the continental states of the USA. The major hosts of A. morbosa are Prunus domestica and Prunus cerasus; the host status of other Prunus species and hybrids is uncertain because of contradictory reports or lack of information. The pest could potentially enter the EU on host plants for planting and plant parts originating in infested third countries. Wood of Prunus spp. is also a pathway of entry, but of minor importance. The current pest distribution and climate matching suggest that the pest could establish and spread in the EU wherever the hosts are grown. In the infested areas, the pest causes girdling of twigs and occasionally of larger branches, whereas trees with multiple infections loose vigour, bloom poorly, and become unproductive, stunted and susceptible to winter injury and infection by other pathogens. The presence of black knots makes trees unsuitable for timber production. It is expected that the pest introduction and spread in the EU would impact host production. Uncertainty exists on whether the agricultural practices and chemical control methods applied in the EU could prevent the establishment and spread of A. morbosa. A. morbosa meets all the criteria assessed by EFSA for consideration as potential Union quarantine pest. As the pest is not known to occur in the EU, this criterion to consider it as Union regulated non-quarantine pest is not met.

Biology and Epidemiology of Venturia Species Affecting Fruit Crops: A Review.

The fungal genus Venturia Sacc. (anamorph Fusicladium Bonord.) includes plant pathogens that cause substantial economic damage to fruit crops worldwide. Although Venturia inaequalis is considered a model species in plant pathology, other Venturia spp. also cause scab on other fruit trees. Relative to the substantial research that has been conducted on V. inaequalis and apple scab, little research has been conducted on Venturia spp. affecting other fruit trees. In this review, the main characteristics of plant-pathogenic species of Venturia are discussed with special attention to V. inaequalis affecting apple, V. pyrina affecting European pear, V. nashicola affecting Asian pear, V. carpophila affecting peach and almond, Fusicladium oleagineum affecting olive, F. effusum affecting pecan, and F. eriobotryae affecting loquat. This review has two main objectives: (i) to identify the main gaps in our knowledge regarding the biology and epidemiology of Venturia spp. affecting fruit trees; and (ii) to identify similarities and differences among these Venturia spp. in order to improve disease management. A thorough review has been conducted of studies regarding the phylogenetic relationships, host ranges, biologies, and epidemiologies of Venturia spp. A multiple correspondence analysis (CA) has also been performed on the main epidemiological components of these Venturia spp. CA separated the Venturia spp. into two main groups, according to their epidemiological behavior: the first group included V. inaequalis, V. pyrina, V. nashicola, and V. carpophila, the second F. oleagineum and F. eriobotryae, with F. effusum having an intermediate position. This review shows that Venturia spp. affecting fruit trees are highly host-specific, and that important gaps in understanding the life cycle exist for some species, including V. pyrina; gaps include pseudothecia formation, ascospore and conidia germination, and mycelial growth. Considering the epidemiological information reviewed, this paper shows that the use of Mills tables to predict infection periods should be avoided for Venturia spp. other than V. inaequalis.

Pest categorisation of Venturia nashicola.

The Panel on Plant Health performed a pest categorisation of Venturia nashicola, the causal agent of Asian pear scab, for the European Union (EU). The pathogen is a well-defined, distinguishable fungal species affecting Pyrus pyrifolia var. culta, P. ussuriensis and P. bretschneideri in Asian countries. P. communis (European pear) is not a host of V. nashicola, but the host status of other Pyrus species is unclear. V. nashicola is not known to occur in the EU. It is listed in Annex IIAI of Directive 2000/29/EC. The pathogen could potentially enter the EU on host plants for planting and fruit originated in infested countries. There are no climatic factors limiting the potential establishment and spread of the pathogen in the EU, as its epidemiology is similar to those of Venturia inaequalis (apple scab) and Venturia pyrina (European pear scab), which are well-established in the EU. The hosts are present in the EU, but no data were found on their abundance and distribution. In the infested areas, V. nashicola causes premature leaf and fruit drop and fruit distortion resulting in considerable yield/quality losses. The introduction of the pathogen into the EU could cause yield/quality losses and environmental consequences because of the additional fungicide sprays for disease control. Cultural practices and chemical measures applied in the infested areas reduce the inoculum sources but they cannot eliminate the pathogen. Phytosanitary measures are available to mitigate the risk of introduction and spread of the pathogen in the EU. All criteria assessed by EFSA for consideration as a potential Union quarantine pest are met. As V. nashicola is not known to occur in the EU, this criterion assessed by EFSA to consider it as a Union regulated non-quarantine pest is not met.

Pest categorisation of Pseudocercospora angolensis.

The Panel on Plant Health performed a pest categorisation of Cercospora angolensis, the fungus responsible for Pseudocercospora fruit and leaf spot of citrus, for all territories except of the Union territories defined in Article 1 point 3 of Regulation (EU) 2016/2031. C. angolensis is listed in Annex IIAI of Directive 2000/29/EC and is not known to be present in the EU. The pathogen, which has recently been reclassified as Pseudocercospora angolensis, is a well-defined, distinguishable fungal species affecting all cultivated Citrus spp. and Fortunella japonica plants. It is currently distributed in sub-Saharan Africa (altitudes 80-1,800 m) and Yemen. Although the epidemiology of P. angolensis is not well understood, infection is favoured by warm temperatures and humidity. The current distribution of the pathogen and climate matching suggests that it might not be well adapted to Mediterranean climates. However, the pathogen is also present in arid areas of Yemen and can infect young fruit with short wetness durations. Uncertainty exists on whether and at which extent the irrigation applied to EU citrus orchards can make the microclimate favourable for P. angolensis. There are no eco-climatic factors limiting the potential spread of the pathogen in the EU. Long-distance spread occurs by wind-disseminated conidia and movement of infected plants for planting and fruit. Short-distance spread occurs via water splash and/or wind-driven rain. In the infested areas, the disease causes premature abscission of young leaves and fruit resulting in yield losses up to 50-100%. Cultural practices and chemical measures applied in the infested areas reduce inoculum but they cannot eliminate the pathogen. All criteria assessed by EFSA for consideration as a potential Union quarantine pest are met. As P. angolensis is not known to occur in the EU, this criterion assessed by EFSA to consider it as a Union regulated non-quarantine pest is not met.

Accurate prediction of black rot epidemics in vineyards using a weather-driven disease model.

BACKGROUND: Grapevine black rot caused by Guignardia bidwellii is a serious threat in vineyards, especially in areas with cool and humid springs. A mechanistic, weather-driven model was recently developed for the detailed prediction of black rot epidemics. The aim of this work was to evaluate the model by comparison with observed disease development in leaves and clusters in a vineyard in north Italy from 2013 to 2015. RESULTS: The model accurately predicted disease onset. The probability of predicting new infections that did not occur (i.e. unjustified alarms) was ≤0.180, while the probability of missing actual infections was 0.175 for leaves and 0.263 for clusters. In 78% of these false negative predictions, the difference between expected and actual disease onset was ±2 days; therefore, only one infection period was actually missed by the model. The model slightly overestimated disease severity (mainly on leaves) when the observed disease severity was >0.6. CONCLUSION: The model was highly accurate and robust in predicting the infection periods and dynamics of black rot epidemics. The model can be used for scheduling fungicide sprays in vineyards. © 2016 Society of Chemical Industry.

A Mechanistic Model of Botrytis cinerea on Grapevines That Includes Weather, Vine Growth Stage, and the Main Infection Pathways.

A mechanistic model for Botrytis cinerea on grapevine was developed. The model, which accounts for conidia production on various inoculum sources and for multiple infection pathways, considers two infection periods. During the first period ("inflorescences clearly visible" to "berries groat-sized"), the model calculates: i) infection severity on inflorescences and young clusters caused by conidia (SEV1). During the second period ("majority of berries touching" to "berries ripe for harvest"), the model calculates: ii) infection severity of ripening berries by conidia (SEV2); and iii) severity of berry-to-berry infection caused by mycelium (SEV3). The model was validated in 21 epidemics (vineyard × year combinations) between 2009 and 2014 in Italy and France. A discriminant function analysis (DFA) was used to: i) evaluate the ability of the model to predict mild, intermediate, and severe epidemics; and ii) assess how SEV1, SEV2, and SEV3 contribute to epidemics. The model correctly classified the severity of 17 of 21 epidemics. Results from DFA were also used to calculate the daily probabilities that an ongoing epidemic would be mild, intermediate, or severe. SEV1 was the most influential variable in discriminating between mild and intermediate epidemics, whereas SEV2 and SEV3 were relevant for discriminating between intermediate and severe epidemics. The model represents an improvement of previous B. cinerea models in viticulture and could be useful for making decisions about Botrytis bunch rot control.

Development and validation of a weather-based model for predicting infection of loquat fruit by Fusicladium eriobotryae.

A mechanistic, dynamic model was developed to predict infection of loquat fruit by conidia of Fusicladium eriobotryae, the causal agent of loquat scab. The model simulates scab infection periods and their severity through the sub-processes of spore dispersal, infection, and latency (i.e., the state variables); change from one state to the following one depends on environmental conditions and on processes described by mathematical equations. Equations were developed using published data on F. eriobotryae mycelium growth, conidial germination, infection, and conidial dispersion pattern. The model was then validated by comparing model output with three independent data sets. The model accurately predicts the occurrence and severity of infection periods as well as the progress of loquat scab incidence on fruit (with concordance correlation coefficients >0.95). Model output agreed with expert assessment of the disease severity in seven loquat-growing seasons. Use of the model for scheduling fungicide applications in loquat orchards may help optimise scab management and reduce fungicide applications.