There are various theories about the occurrence of oak decline, but what most researchers agree on is that one factor can’t be the only cause of this phenomenon, instead it is influenced by several factors that occur simultaneously or frequently (Crampton et al., 2020; Pourhashemi & Sadeghi, 2020; Finch et al., 2021). The multidimensionality of the disease has made it difficult to manage. Studies have shown that biotic factors such as oomycetes (Grunwald et al., 2012; Hyun & Choi, 2014; Forrestel et al., 2015), fungi (Costa et al., 2020, 2022; Denman et al., 2022), bacteria (Crampton et al., 2020; Gathercole et al., 2021; Denman et al., 2022), nematodes (Maleita et al., 2015; Pedram et al., 2018; Ahmadi et al., 2019), insects (Thomas, 2008; Brown et al., 2015; Haavik et al., 2015; Denman et al., 2022), viruses (Bandte et al., 2020), and hemiparasitic plants (e.g., mistletoe) (Dolezal et al., 2010, 2016), as well as abiotic factors such as drought, frost damage, soil waterlogging, cultural practices, air pollution, excess of nitrogen, and fires (Thomas et al., 2002; Thomas, 2008; Gentilesca et al., 2017; Denman et al., 2022; Machacova et al., 2022), are involved in this disease. As the oak phenology shifts due to global warming, abiotic factors are becoming increasingly important (Machacova et al., 2022). Climate change is causing the initiation of cambial activity and leaf development earlier than several decades ago. This results in a higher risk of spring defoliation caused by late spring frost in the case of early oaks (Dantec et al., 2015; Puchałka et al., 2017). In the conceptual model for the interaction of abiotic and biotic factors crucial in the emergence of oak decline, we can consider three key factors: insect defoliation, summer drought with heat waves, and winter and/or spring frost (Thomas et al., 2002; Machacova et al., 2022).

The disease decline spiral (Fig. 1) addresses both the phenomenon of multidimensionality and the role of each factor in the disease cycle (Manion, 1981). By moving to the center of the spiral, apart from abiotic factors, an increase in the number of biotic factors involved in the disease becomes apparent. While abiotic factors can initiate the disease cycle and contribute to its continuation, biotic factors tend to play a more significant role in the later stages of the disease. The spiral emphasizes the interdependence of various factors in the disease cycle and underscores the need for a comprehensive approach to disease diagnosis and management. By understanding the role of each factor in the spiral, researchers and managers can develop strategies to break the cycle and prevent further tree decline. This may involve implementing measures such as proper tree care, monitoring for pests and diseases, and promoting healthy ecosystems (Thomas et al., 2002; Denman et al., 2014; Bendixsen et al., 2015; Haavik et al., 2015). Denman et al. (2022) suggest an update to the decline disease spiral model (Fig. 2), emphasizing the passage of time, which was implied in the original model but requires an annotation for clarity. They note that the original model did not include a microbial component in the predisposing phase, despite the crucial and under researched role microorganisms can play in predisposition, considering both primary and secondary pathogens, as well as saprophytic and beneficial organisms. Non-lethal pathogens and insects can also play a significant role in predisposing trees through direct and indirect effects, such as feeder root necrosis and loss of ectomycorrhizal relationships, predisposing trees in an initial phase.

The oak decline can be related with various symptoms depending on the environmental conditions and the tree species affected. Decline symptoms may include leaf yellowing, wilting, thinning of the crown, bark and cambium tissue necrosis, trunk cankers, and infestation by wood-boring beetles. The disease can lead to dieback of branches and twigs, reduced tree growth and vigor, and acorn production. Infected trees may also become more susceptible to other pests and diseases, and in severe cases, the disease can result in the tree’s death. In fact, oak decline is a general term that refers to several health issues affecting oak trees. Due to the complexity of each type of decline and the need for effective management strategies, scientists have worked to identify the key components involved in these distinct diseases and establish proof of cause and effect. The rate of disease development and symptomatology are among the metrics used to differentiate types of decline (Denman et al., 2022).

According to scientific literature, oak decline can be classified into various types:

a) Chronic oak decline (COD): It is characterized by a slow development of symptoms and disease, which can take several decades with emphasis on failing root health as a primary causal factor of the decline. The disease is primarily identified by thinning crowns, fine twig shedding, and stubby small branch ends. In advanced stages, the dieback of scaffold branches can occur. The mortality rate is low and some trees can recover partially or completely (Denman & Webber, 2009; Lonsdale, 2015; Gagen et al., 2019).

b) Sudden oak death: In most cases, due to biotic factors (usually pathogens), affected trees can die in a very short time. An example is a sudden oak death that occurred on the coast of California in the mid-1990s due to the phytopathogenic oomycete Phytophthora ramorum (Rizzo et al., 2002; Grunwald et al., 2012; Forrestel et al., 2015).

c) Acute oak decline (AOD): It is a recently identified decline disease affecting native oak species in the United Kingdom (Quercus robur and Q. petraea). It has also been reported in other European countries as well as the Middle East (Iran) on other species such as Q. ilex, Q. castaneifolia, Q. pyrenaica, Q. fabri, Q. brantii, Q. suber, Q. Cerris, Q. pubescens, Q. rubra (Moradi‐Amirabad et al., 2019; Ruffner et al., 2020; Fernandes et al., 2022; Pernek et al., 2022). Acute oak decline is characterized by episodes of rapid decline over 5-10 years, associated with high levels of tree mortality although sometimes trees may stabilise and even recover. There are four primary characteristics that define the disease: weeping patches that are vertically aligned on the trunk of oak trees, cracks between bark plates that seep dark fluid, necrosis of the inner bark, and the presence of larval galleries of the oak buprestid, Agrilus biguttatus, on the phloem-sapwood interface in over 90 percent of cases. Given the clear and distinctive symptoms associated with acute oak decline, it is hypothesized that it is a distinctive, identifiable condition within the broader oak decline syndrome (Denman et al., 2014, 2022; Brown et al., 2015; Crampton et al., 2020).

The first verified reports of oak decline date back to the 18th and 19th centuries in Central Europe and Northeastern United States of America (USA), as recorded in the literature. From the 1900s to the 1970s, more cases of oak decline were reported in these regions and in adjacent areas such as Western and Northeastern Europe and Central USA. During this time, deciduous oaks, particularly those from the Quercus and Lobatae groups, were the most affected by the decline. Starting from the 1980s, oak decline episodes were frequently reported across a wide range of northern hemisphere forests, affecting both deciduous species from the Cerris group and evergreen species such as Q. ilex from the Ilex group. Oak decline records emerged in Southern Europe, North Africa, Western USA, Mexico, Colombia, China, and Japan (Thomas, 2008; Gil-Pelegrin et al., 2008; Rodriguez-Calcerrada et al., 2017).

According to scientific sources, oak decline is currently affecting 39 countries worldwide. These countries are listed in Table S1 [suppl] and include Ukraine, Croatia, Colombia, Italy, the Netherlands, France, Germany, Bulgaria, Portugal, Austria, Switzerland, the Baltic States, Moldavia, Poland, Sweden, Spain, Mexico, Turkey, Japan, Belgium, Canada, Romania, Greece, the Czech Republic, UK, Algeria, China, Hungary, Iran, Latvia, Slovakia, Finland, South Korea, Russia, Slovenia, Serbia, USA, Tunisia, and Morocco (Fig. 3). As shown in Table S1, the framework used to present the distribution, factors, and host species is the same as that employed by Gottschalk & Wargo (1997). However, this paper provides new official records from around the world on previously unrecorded countries, biotic and abiotic agents, and hosts for oak decline. Some of these records were not mentioned in detail or were not included in Gottschalk & Wargo’s study from 1997.

Researchers use some indexes to evaluate plant diseases. This helps the management and epidemiology of the diseases, including topics such as determining the number of losses, assessing the disease threshold for control, understanding the effectiveness or ineffectiveness of some treatments for the disease, and studying the resistance and tolerance levels of cultivars. Therefore, methods of assessing the incidence and severity of the disease are important issues. If they are not properly assessed, subsequent measures in its management may also fail. Assessments can vary depending on the host plant, the type of disease, and even the region of occurrence of the disease, and the same version of the assessment cannot be used for all diseases (Campbell & Neher, 1994; Chiang et al., 2017; Finch et al., 2021).

Numerous visual phenotypic descriptors aid in classifying declining oak trees. When severe decline occurs, symptoms such as low foliage density and multiple dead branches are noticeable. However, some symptoms develop slowly, particularly in cases of chronic decline, which can make it difficult to obtain a complete classification of healthy versus declining trees. Because the symptoms are evaluated visually, the results of the disease assessment will also be variable because of the several factors contributing to the oak decline and surveyor bias. Furthermore, differences in tree age and size at different geographical locations can make it difficult to objectively compare health status between sites. These factors must be standardized to reduce potentially distorting comparisons of tree health status (Pontius & Hallett, 2014; Denman et al., 2014, 2017; Ahmadi et al., 2019).

Recently, due to the importance of disease assessment, some researchers have made good efforts to achieve a valid disease assessment in oak decline (Finch et al., 2021). They analyzed the issues surrounding oak decline and then presented the phenotypic decline index (PDI) and the decline acuteness index (DAI). An important feature of the above study is that phenotypic descriptors are easily measurable in the field and based on well-established tree condition assessment protocols. In addition, the machine learning approach used to extract these indexes could be applied to long-established oak health monitoring programs to help sensitivity in diagnosing overall trends in tree health. The collected phenotypic measurements (36 descriptors) from 174 trees were checked from nine locations across England and included healthy, AOD, COD, and AOD trees in remission. PDI and DAI indexes derived from unsupervised random forest machine learning models, trained using the collected phenotypic information. PDI is an index to measure decline severity and DAI is an index to measure differentiating between chronically and acutely declining oak trees (Table 1). The results showed that two descriptors, crown condition, and trees size, contributed positively to the PDI. Trees with smaller crowns (less foliage and smaller canopy) in poor condition had greater PDI values. Descriptors including tree stature and the presence of stem bleeding, contributed a lot to the DAI, allowing differentiation between trees with AOD and COD syndromes. AOD trees had relatively larger stature and the presence of stem bleeding while COD trees had a small stature and stem bleeding was absent. PDI and DAI indexes can be shown as a 2-dimensional continuum under which the spectrum of decline severity and type can be assessed (Fig. 4) (Finch et al., 2021). It is crucial to note in this section that the research by Finch et al. (2021) primarily focuses on AOD, and PDI and DAI are only validated for Q. petraea and Q. robur, which are significantly different from other Quercus species. Consequently, these phenotypic indexes must be adapted and validated for various Quercus species.

Remote sensing data can provide valuable information for the assessment of oak decline, as it can capture detailed information on vegetation properties and structure at a large scale. Different remote sensing technologies, such as hyperspectral imaging, LiDAR, and satellite imagery, can be used to detect changes in vegetation properties related to oak decline. Remote sensing data can be analyzed using various techniques, such as machine learning algorithms, to identify complex patterns and relationships related to oak health and decline. These techniques can effectively process large volumes of data and provide valuable information for informed management decision-making. There are challenges and limitations to the use of remote sensing technologies for assessing oak decline. For example, the accuracy of the classification may be affected by factors such as atmospheric conditions, sensor calibration, and variations in vegetation properties. Nevertheless, remote sensing offers a valuable tool for assessing oak decline at a detailed level and detecting early signs of decline. By providing more precise information on vegetation properties, it can help to inform targeted management interventions aimed at mitigating the impact of the decline (Pontius et al., 2020; Liu et al., 2020).

Since oak decline is influenced by a range of environmental factors, and numerous parameters are involved in this disease, identifying these factors and selecting an appropriate modeling approach are key challenges in this field. Accurately determining the factors that contribute to oak decline can be challenging due to the complexity of the disease and the numerous variables that can be involved. Therefore, selecting an appropriate modeling method is essential for accurately representing the complex relationships between environmental factors and oak decline (Ahmadi et al., 2014). Since the 1990s, artificial intelligence (AI) has been utilized to model processes with high complexity. Artificial neural networks (ANNs) is a notable example which have proven to be effective in modeling complex systems due to their ability to identify patterns and relationships in large datasets. ANNs consist of interconnected nodes, called “neurons”, organized into layers. The input layer receives data, which is then passed through one or more hidden layers that perform computations, before reaching the output layer. During the training process, ANNs learn to adjust the weights of the connections between neurons in order to optimize their performance on a specific task. Once the ANN has been trained, it can be used to make predictions or classifications on new input data (Rosa, 2013; Aggarwal, 2018).

In the context of oak decline, ANNs can be used to analyze large and complex datasets that include environmental and biological variables. These variables can include factors such as climate data, soil conditions, tree age and species, as well as biotic factors such as pest and disease presence. By training ANNs on these datasets, researchers can identify patterns and relationships that are not easily discernible through traditional statistical methods as in Gutierrez-Giron et al. (2019) in Mediterranean forests predicting the severity and distribution of oak decline under future climate scenario, Guo et al. (2019) in the Loess Plateau of China and Martin-Benito et al. (2017) to improve the accuracy of oak decline risk assessment in Mediterranean woodlands and to identify the most important factors contributing to oak decline. Zhao & Zhang (2018) were able to accurately predict the risk of oak decline using ANNs and logistic regression. Li et al. (2020) developed multi-scale ANN models to predict oak decline through the use of environmental and biological variables at different spatial scales, including landscape, patch, and tree levels. This study highlights the potential of ANNs to incorporate information at multiple spatial scales in the study of oak decline.

There are potential troubleshooting challenges in the use of AI methods for oak decline assessment. One of the main challenges is the need for large amounts of high-quality data to train the ANN models effectively. In addition, the results of ANN models can be difficult to interpret and explain, which can limit their usefulness for informing management decisions. Furthermore, different studies may use different datasets or modeling approaches, which can lead to inconsistencies or contradictions in the results (Zhang et al., 2018). This highlights the importance of considering multiple modeling approaches and comparing their results to identify the most accurate and reliable predictions. With continued research and development, AI methods have the potential to contribute significantly to our understanding of oak decline and inform management decisions to mitigate its impact.

Metagenomics, the study of genetic material recovered directly from environmental samples, has emerged as a powerful tool in understanding the complex microbial communities in forest ecosystems. Over the last decade, there has been a growing interest in the use of metagenomics in studying forest decline and identifying the role of microbial communities in maintaining healthy forest ecosystems (Cardenas et al., 2015; Eaton et al., 2017; Denman et al., 2018b; Venice et al., 2021). Metagenomics has been used to develop new approaches for forest restoration. A study by Bahram et al. (2013) used metagenomics to assess the impact of different forest restoration strategies on the microbial communities in degraded forests, suggesting that microbial diversity is an important factor in the success of forest restoration efforts. This technique has also been used to identify potential bioindicators of soil health and forest decline (Deveau et al., 2018; Duque-Zapata et al., 2023).

Recently, there has been a growing interest in studying the microbiome of oak ecosystems and its potential role in oak decline. The use of metagenomics in oak decline research has often involved comparing the microbial communities associated with healthy trees and diseased trees to identify potential differences and determine which microbial taxa or pathways may be involved in oak decline (Lamichhane & Venturi, 2015; Broberg et al., 2018; Denman et al., 2018b; Pinho et al., 2020). However, it’s worth noting that not all studies have focused solely on comparing the microbiome of healthy and diseased trees. Some studies have also investigated the impact of environmental factors, such as soil type and tree age (Meaden et al., 2016; Denman et al., 2018a), physicochemical properties (Poudel et al., 2020) and land use (Barrios-Masias et al., 2019), on the microbial communities of oak ecosystems, or have examined the potential role of specific microbial taxa or functional genes in oak health and decline (Barcenas-Moreno et al., 2019). In addition to metagenomics, some studies on oak decline are using multi-omic methods such as metatranscriptomics (Barcenas-Moreno et al., 2019; Poudel et al., 2020) and metaproteomics (Poudel et al., 2020) to investigate the functional potential and activity of microbial communities associated with oak ecosystems.

The rhizosphere is indeed the front line of plant defense against pathogens, and it can help plants overcome biotic and abiotic stresses (Mendes et al., 2013; Berg et al., 2014; Munir et al., 2022). Deepening studies on the rhizosphere microbiome of oak trees can be effective in managing oak decline by identifying dominant microbial plant growth promoters and understanding the complex interactions between microbes and oak trees. Studying the rhizosphere microbiome of oak trees can also help identify potential biocontrol agents that can be used to manage oak decline (Fernandez-Gonzalez et al., 2017; Poudel et al., 2020). Overall, these studies highlight the importance of studying the rhizosphere microbiome of oak trees for managing oak decline and promoting oak health. By identifying dominant microbial plant growth promoters and potential biocontrol agents, researchers can develop effective strategies for managing oak decline and mitigating the impact of disease causing pathogens.