Next Article in Journal
Linking Water Policy, Agriculture, and Predator Responses in Hyperarid Landscapes
Previous Article in Journal
Mating Modifies the Survival and Oviposition of Tetranychus merganser Boudreaux (Acari: Tetranychidae) Females on Five Host Plants
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agriculture
Volume 16
Issue 5
10.3390/agriculture16050505
agriculture-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Modeling Current and Future Global Distribution of the Quarantine Pathogen Phytophthora ramorum Using MaxEnt

by
Bingyan Zheng
1,
Sixi Lin
1,
Xiaorui Zhang
1 and
Tingting Dai
1,2,*
1
Co-Innovation Center for Sustainable Forestry in Southern China, Nanjing Forestry University, Nanjing 210037, China
2
Advanced Analysis and Testing Center, Nanjing Forestry University, Nanjing 210037, China
*
Author to whom correspondence should be addressed.
Agriculture 2026, 16(5), 505; https://doi.org/10.3390/agriculture16050505
Submission received: 9 January 2026 / Revised: 13 February 2026 / Accepted: 16 February 2026 / Published: 26 February 2026
(This article belongs to the Section Crop Protection, Diseases, Pests and Weeds)
Browse Figures
Versions Notes

Abstract

Phytophthora ramorum causes Sudden Oak Death, threatening global forests and nurseries. We modelled its current and future global distribution using an optimized MaxEnt model (mean AUC = 0.978). Temperature seasonality was the key driver (29.4%). Currently, suitable habitats (~42.48 million km2) concentrate in mid-latitude coastal regions (e.g., North America, Europe, East Asia). Future projections (SSP126-585) show no systematic contraction but rather significant spatial shifts: suitable areas expand poleward and upward in elevation while contracting in some current regions. High-suitability areas (5.52–6.40 million km2) change by −3% to +12%, with the centroid shifting northwestward, notably into the African Sahel under high emissions. Our findings indicate that, within the framework of the current modeling approach and the emission scenarios applied, climate warming is likely to lead to a net increase in the suitable habitat for this pathogen at the global scale, providing a critical scientific basis for the development of international phytosanitary strategies.

1. Introduction

Forest ecosystems serve as critical reservoirs of global biodiversity, and their integrity plays an irreplaceable role in maintaining ecological service functions and biogeochemical cycles [1,2]. Sudden Oak Death (SOD) and related wilting diseases, caused by the oomycete pathogen Phytophthora ramorum Werres, De Cock & Man in ‘t Veld. It is lethal to specific oak species such as California black oak (Quercus kelloggii) and coast live oak (Quercus agrifolia). However, its impact extends far beyond these trees. The pathogen can also infect a wide range of plants, including rhododendrons (Rhododendron spp.), mountain laurel (Kalmia latifolia), coast redwood (Sequoia sempervirens), camellias (Camellia japonica), and viburnums (Viburnum spp.), causing branch and foliage blight on these hosts and producing large quantities of spores that accelerate the spread of the epidemic [3,4]. First detected in 1995 in coastal California, USA, the pathogen has since established outbreaks in at least 27 countries and regions worldwide—including the United States, Canada, Japan, and several European nations—inflicting catastrophic impacts on native forest ecosystems [5,6]. Its geographic distribution exhibits a significant coupling with maritime climates, with suitable habitats predominantly concentrated in humid coastal zones characterized by abundant precipitation and stable humidity [7]. Temperature thresholds and leaf wetness duration are identified as key ecological factors limiting its population establishment [8]. The pathogen disperses via rain splash, air currents, and contaminated plant propagation materials, while international trade in ornamental plants and movement of nursery stock constitute primary pathways for transcontinental spread [6,9]. In Western Europe, the disease has severely threatened heathland shrubs and Japanese larch plantations, prompting the European Union to list it in the A2 quarantine list and implement emergency control measures [10]. Recent projections using the CLIMEX model suggest that under future global warming scenarios, its suitable range may expand toward higher latitudes, though infection risks could decline in some Mediterranean regions due to reduced summer rainfall [11].
Current management of P. ramorum primarily relies on reactive strategies such as quarantine regulations, surveillance, and removal of infected host plants [12]. The effectiveness of these measures is increasingly challenged by the pathogen’s capacity for long-distance spread via global plant trade and its sensitivity to climate change. Furthermore, modeling the distribution of this pathogen faces unique difficulties: disease occurrence records reflect not only environmental suitability but also host distribution, surveillance intensity, and reporting biases [13,14,15]. Accurate identification of Phytophthora species often requires molecular methods (e.g., qPCR, LAMP) [16]. Although on-site rapid detection technologies have advanced, their large-scale application remains constrained by cost and operational limitations, adding uncertainty to distribution data [17]. Consequently, there is a pressing need to shift from containment strategies based on static historical distribution towards a dynamic risk assessment framework capable of anticipating future distributional changes.
Against the backdrop of intensifying global climate change and increasing vegetation landscape fragmentation, the geographical invasion patterns of plant-pathogenic oomycetes are undergoing profound restructuring. Soil, topography, and climate collectively constitute a regulatory complex that governs the survival and reproduction of organisms, and the epidemic dynamics of pathogens through complex feedback mechanisms [18]. Specifically, soil physicochemical properties—such as organic matter content, pH gradients, pore structure, and water-holding capacity—directly influence zoospore survival rates and chlamydospore germination potential [19,20]. Topographic parameters, including elevation gradients, aspect indices, and surface roughness, not only condition local microclimates but also constrain the suitability of habitats for pathogen colonization by modulating surface runoff pathways and spore dispersal fluxes [21,22]. Climatic variables—particularly temperature variability and precipitation patterns—act as core drivers, regulating phenotypic plasticity, outbreak thresholds, and the transmission potential of the pathogen, as mediated by thermodynamic and hydrodynamic processes [23]. These combined parameter variables interact nonlinearly and generate cascading effects, collectively shaping the complex environmental background of the pathogen’s ecological niche [24].
Species distribution models (SDMs) integrate environmental gradient data with species occurrence records to effectively simulate spatial heterogeneity in suitable habitats, decipher the multidimensional niche profile, and assess colonization risks of invasive species [25]. Among these, the Maximum Entropy (MaxEnt) model, grounded in the niche conservatism hypothesis, has become a mainstream predictive tool in invasion biology due to its robustness in handling small sample sizes and complex environmental relationships [26,27,28]. The MaxEnt algorithm infers potential geographical distribution probabilities based on the principle of maximum entropy and can quantify the relative contribution of environmental variables and identify key limiting factors [29]. Furthermore, MaxEnt supports the construction of quadratic and interaction terms for environmental variables, offering a methodological pathway to decipher the ecological mechanisms that underpin disease distribution patterns [30]. In this study, the ENMTools v5.26 platform was employed to systematically evaluate optimization schemes involving different feature class combinations (FCs) and regularization multipliers (RMs) [31]. The optimal parameter set was selected based on the delta AICc criterion, significantly enhancing model robustness and spatial transferability [32].
This study integrates soil horizon attributes, topographic wetness indices, and bioclimatic variables to systematically analyze the dynamic patterns of the global potential suitable habitat for P. ramorum using the MaxEnt model. Hierarchical partitioning is applied to quantify the independent contributions of environmental factors, clarifying their niche requirements, while simulating range shift trajectories across three Shared Socioeconomic Pathways (SSPs 126, 370, and 585) [33]. The findings will provide theoretical support for establishing cross-border early warning networks regarding tree disease spread. Additionally, by linking environmental filtering mechanisms to pathogen distribution, this study broadens the application of the MaxEnt model in oomycete biogeography and provides a methodological framework for model optimization and validation in forest pathology [34,35].

2. Materials and Methods

2.1. Occurrence Data Acquisition and Preprocessing

Distribution data for Phytophthora ramorum and its primary host plants were obtained from the Global Biodiversity Information Facility (GBIF Occurrence Download. Available online: https://doi.org/10.15468/dl.paw2dy (accessed on 10 October 2025)), the Centre for Agriculture and Bioscience International (CABI; https://www.cabi.org/; accessed on 29 December 2025), and the EPPO Global Database (https://gd.eppo.int; accessed on 30 December 2025). Relevant scientific literature was also consulted [36,37]. A total of 923 occurrence records were compiled for P. ramorum, while host plant records were collected separately. All geographic coordinates (in decimal degrees) from these sources were verified and standardized using Google Maps v4.5.0.
To mitigate the influence of spatial autocorrelation on model performance, spatial filtering was applied specifically to the pathogen occurrence data. Using the “Spatially Rarefy Occurrence Data for SDMs” tool in ArcGIS (version 10.8), the pathogen records were spatially rarefied with a 5 km grid, retaining a maximum of one record per grid cell [38,39]. This process yielded 192 spatially independent occurrence points for P. ramorum (Figure 1A), which were used for subsequent MaxEnt model calibration. These filtered pathogen occurrences are primarily concentrated in North America, Europe, and North Africa, with scattered records in South Asia and Southern Africa. The host plant occurrences, which underwent coordinate verification and standardization but not spatial rarefaction, are mainly located in North America and Europe, with limited records on other continents.

2.2. Acquisition and Preprocessing of Environmental Data

The climate data utilized in this study were obtained from the WorldClim database (version 2.1; http://www.worldclim.org) under the CMIP6 Shared Socioeconomic Pathways (SSPs) framework [40]. Three SSP scenarios (SSP126, SSP370, and SSP585) were selected, covering a historical baseline period (1970–2000) and four future time slices (2021–2040, 2041–2060, 2061–2080, and 2081–2100), using a 2.5 arc-minute resolution grid. A total of 19 bioclimatic variables (Bio01–Bio19) were initially extracted for each scenario. Topographic variables—elevation, slope, and aspect—were derived from a digital elevation model (DEM) using the surface analysis tools in ArcGIS [41]. Slope and aspect were calculated from the elevation data. Soil data were sourced from the Harmonized World Soil Database (version 1.2; https://gaez.fao.org/pages/hwsd, accessed on 30 December 2025), from which 14 topsoil properties were extracted in ArcGIS (Table S1) [42,43]. To mitigate multicollinearity among environmental predictors, a Pearson correlation analysis was performed using ENMtools. Variables with a correlation coefficient absolute value below 0.8 were retained; for pairs exceeding this threshold, one variable was excluded based on its contribution and permutation importance in preliminary model runs [44]. Following this procedure, four bioclimatic variables (Bio4, Bio06, Bio14 and Bio19), two topographic factors (aspect, slope) and five edaphic factors (T_bulk, T_ece, T_gravel, T_silt, and T_usda) were selected for subsequent species distribution modeling (Figure 2A).

2.3. Optimization and Construction of the MaxEnt Model

The predictive accuracy of the MaxEnt model is primarily influenced by two key parameters: the regularization multiplier (RM) and the feature combination (FC) [45,46]. The FC includes five categories of environmental predictors: threshold (T), linear (L), quadratic (Q), hinge (H), and product (P), with default settings of RM = 1 and FC = LQHPT [47]. To mitigate the risk of model overfitting, this study employed the ENMeval package in R version 2.0.5.2 for parameter optimization. The parameter settings were as follows: nine feature combinations (L, H, LQ, QH, QHP, LQP, LQH, LQHP, and LQHPT) were tested; the RM parameter ranged from 0 to 6.0 with an interval of 0.5, resulting in 12 levels [48]. A total of 108 parameter combination models were constructed by permuting these two parameters. Model fit was evaluated using the Akaike Information Criterion (AIC). The model with the smallest ΔAICc value (i.e., the difference between its AICc and the lowest AICc among all models) was selected as the optimal predictive model [49]. A total of 192 occurrence records of the oak decline pathogen and the selected key environmental variables were incorporated into the optimized MaxEnt model. The background points (or pseudo-absence points) used in the model were set to 10,000, randomly sampled across the entire study area. Their ecological meaning is to represent the available environmental background space for the species during the modeling process, rather than confirmed absence locations. Subsequently, the dataset (including species occurrence points and background points) was randomly partitioned into 75% for training and 25% for testing. Model runs were repeated 10 times using the Bootstrap resampling method [50]. Outputs were generated in “Cloglog” format and exported as “.asc” files. Response curves were plotted, and a jackknife procedure was employed to assess the contribution of each predictor to the suitability distribution, thereby quantifying variable importance and validating model performance [51].

2.4. Evaluation of MaxEnt Model Accuracy

The area under the receiver operating characteristic curve (AUC) was used as the metric to assess model performance, with values ranging from 0.0 to 1.0. A higher AUC value signifies superior model discriminative ability. According to widely adopted classification criteria, AUC values are interpreted as follows: AUC thresholds: ≤0.6 (Failed), 0.6–0.7 (Poor), 0.7–0.8 (Fair), 0.8–0.9 (Good), >0.9 (Excellent) [52,53].

2.5. Habitat Suitability Classification and Centroid Shift Analysis

In the ArcGIS platform, the output “.asc” files from the optimized model were processed through spatial data conversion and classification procedures. The files were first converted into raster format (“raster”) using the format conversion tool. Habitat suitability indices were then reclassified by the Jenks Natural Breaks classification method into four categories: unsuitable (0–0.065), slight suitability (0.065–0.195), moderate suitability (0.195–0.365), and high suitability (0.365–1.000) [54]. By overlaying and comparing the potentially suitable habitats under current and future climate scenarios, the distribution shifts of P. ramorum were characterized into three types: retained, expanded, and lost areas. The area of each type was quantified to reflect the spatial contraction or expansion of suitable habitats over time [55]. Additionally, the centroid coordinates of the suitable areas for each period were calculated using SDMToolbox [54]. By analyzing the displacement distance and direction of centroids across time series, the spatial migration dynamics of the pathogen’s potential habitat were elucidated.

3. Result

3.1. Model Parameter Optimization and Performance Evaluation

The MaxEnt model was optimized using the ENMeval package. The results (Figure 2B) indicated that when the feature class (FC) was set to quadratic and hinge (QH) and the regularization multiplier (RM) was 1.5, the corrected Akaike Information Criterion (ΔAICc) reached zero, suggesting that the model under this parameter combination represented the optimal structure. Using this optimal parameter set, the model was run iteratively 10 times, and the receiver operating characteristic (ROC) curve was plotted for accuracy assessment. With a mean AUC of 0.978 (Figure 2C), the model demonstrated excellent discriminative performance, which is significantly higher than the theoretical AUC of a random model (0.5). This demonstrates that the optimized model exhibits strong predictive performance and stability, and it can be reliably applied to subsequent predictions of the potentially suitable habitat distribution for P. ramorum.

3.2. Environmental Drivers Shaping the Distribution of P. ramorum

MaxEnt modeling revealed that, among the 11 environmental variables evaluated, temperature seasonality (Bio04) was the most influential factor shaping the potential geographic distribution of P. ramorum, contributing 29.4% of the total explained variation (Table 1). Secondary contributors included precipitation of the coldest quarter (Bio19, 20.4%), minimum temperature of the coldest month (Bio06, 18.2%), soil gravel content (T_gravel, 7%), and slope (6.7%). Together, these five variables accounted for 81.7% of the cumulative contribution rate. To elucidate the relationships between key environmental predictors and species occurrence probability, jackknife tests and single-variable response curves were further employed (Figure 3). By controlling for the effects of other variables, the suitable ranges corresponding to habitats with a probability threshold > 0.5 were identified as follows: temperature seasonality (Bio04), 523.11–1219.04 (unit: standard deviation × 100); minimum temperature of the coldest month (Bio06), –5.39 to 5.61 °C; and precipitation of the coldest quarter (Bio19), 118.01–434.01 mm. In summary, temperature and precipitation variables exerted decisive influences on the distribution pattern of P. ramorum, whereas soil and topographic factors played relatively limited roles. These findings provide a scientific basis for monitoring, early warning, and regional management strategies targeting this pathogen.
Furthermore, based on the distribution points of the main hosts of Phytophthora ramorum (Figure 1C), we constructed a potential geographical distribution map of its hosts (Figure 1D). The results show that in Europe and America, the potential distribution areas of the hosts are highly consistent with the suitable habitats of P. ramorum; in other regions, the two also exhibit a partially overlapping spatial pattern. This indicates that host distribution is one of the important factors influencing the potential distribution of P. ramorum.

3.3. Simulation of Ecologically Suitable Areas Under Current Climatic Conditions

Under current climatic conditions, the total projected suitable area for P. ramorum worldwide is 4247.64 × 104 km2, exhibiting distinct latitudinal gradients and heterogeneity in coastal versus inland distribution (Figure 1A; Table 2). The overall suitability intensity decreases progressively from humid coastal regions toward arid interior zones, showing high consistency with global records of pathogen occurrence. Highly suitable areas (631.97 × 104 km2, 14.88%) are mainly concentrated in mid-latitude maritime climate zones, including the coastal regions of California and Oregon in the United States, southwestern Europe (northern Spain, western France, and northwestern Italy), and southeastern Oceania (North Island of New Zealand and Tasmania, Australia). Moderately suitable areas (1302.34 × 104 km2, 30.66%) form a transitional belt surrounding the highly suitable zones, encompassing the southeastern United States, the British Isles, southern Scandinavia, the Balkan Peninsula, and Honshu Island in Japan. Low-suitability areas (2313.33 × 104 km2, 54.46%) are widely distributed across arid and semi-arid inland regions as well as subtropical/tropical transitional zones, such as the North American Great Plains, interior Eurasia, eastern Mediterranean regions, southwestern South Africa, and Patagonia in South America.
Under future climate scenarios, the projected distribution of P. ramorum exhibits clear spatial dynamics and varying degrees of habitat change. Across the three SSP scenarios (SSP126, SSP370, and SSP585), the total suitable area fluctuates over time, with net changes ranging from −6.00% to +8.06% relative to the historical baseline. Notably, highly suitable habitats show a consistent declining trend in all scenarios, with losses of up to −15.11% (SSP370, 2061–2080), indicating potential habitat contraction in core maritime zones. In contrast, moderately and poorly suitable areas display more variable patterns, with some periods and scenarios showing expansion, particularly under SSP585, where the poorly suitable area increases by up to +13.77% (2041–2060). These quantitative estimates highlight a progressive shift in suitability structure, with an overall tendency toward habitat degradation in historically optimal regions and potential range expansion into marginally suitable zones.

3.4. Distribution Dynamics of P. ramorum Under Future Climate Scenarios

The global suitable habitat distribution of P. ramorum remains largely consistent across the three future climate scenarios (SSP126, SSP370, and SSP585), yet the area covered by each suitability class exhibits significant and complex dynamic changes (Figure 4; Table 2). Compared to the total suitable area during the historical baseline period (1970–2000; 4247.64 × 104 km2), the projected total suitable area under future scenarios fluctuates between 3992.88 × 104 and 4590.16 × 104 km2, demonstrating a non-monotonic trend. Specifically, under the SSP126 scenario, the total suitable area peaks during 2021–2040 (4512.30 × 104 km2) before declining to 4341.14 × 104 km2 in 2041–2060. Under SSP370, the area reaches its minimum (3992.88 × 104 km2) in 2041–2060, then recovers to 4303.00 × 104 km2 by 2081–2100. In contrast, under SSP585, the area attains its maximum (4590.16 × 104 km2) in 2041–2060. Despite these temporal fluctuations, the proportional distribution among suitability classes maintains a stable hierarchical structure across all periods and scenarios, consistently following the gradient: low suitability > moderate suitability > high suitability. Low-suitability areas constitute the majority, accounting for 54.46–58.52% of the total suitable area, while moderate- and high-suitability areas represent 28.62–31.30% and 12.46–15.12%, respectively. This persistent structure reflects the pathogen’s distinct adaptive thresholds to global environmental conditions. Notably, the magnitude of change in area varies considerably among suitability classes under different climate scenarios. The high-suitability area is minimized (552.44 × 104 km2) under SSP585 during 2041–2060, but peaks (640.29 × 104 km2) under SSP126 during 2021–2040. The moderate-suitability area shows the most pronounced variation, ranging from 1142.76 × 104 km2 under SSP370 (2041–2060) to 1365.59 × 104 km2 under SSP585 (2041–2060). Meanwhile, the low-suitability area reaches its maximum extent (2307.76 × 104 km2) under SSP370 in 2041–2060, comprising 57.80% of the total suitable area under that scenario and period.

3.5. Shifts in the Potential Habitat of P. ramorum Under Future Climate Scenarios

Based on projections under three future climate scenarios (SSP126, SSP370, and SSP585), the spatial configuration of global suitable habitats for P. ramorum exhibited pronounced dynamic stability, with total suitable area fluctuating between 3135.24 × 104 and 3421.83 × 104 km2 (Figure 4). This demonstrates that global warming did not trigger a systematic contraction of its potential distribution range, but rather impelled a complex reorganization of internal habitat structures. Expansion zones manifested conspicuous spatiotemporal heterogeneity, reflecting climate-driven dynamics of distribution fronts. Under SSP126, the expansion area reached 670.97 × 104 km2 during 2061–2080, representing the maximum value across all scenario periods. In the SSP585 scenario, expansion areas attained 745.70 × 104 km2 (2041–2060) and 812.61 × 104 km2 (2081–2100), indicating persistent colonization toward high-latitude and high-altitude regions under high-emission pathways. The expansion trajectory is primarily oriented toward southern Canada in northeastern North America, southern Scandinavia in Europe, and the temperate zones of Asia, evidencing that climatic warming disrupted incumbent temperature-limiting factors and facilitated the progressive conversion of formerly marginal cold regions into novel suitable habitat frontiers. The differential magnitude between contraction and expansion zones unveiled the bidirectional effects of climate change. Under SSP370 (2081–2100), the contraction area peaked at 732.41 × 104 km2, substantially exceeding its corresponding expansion zone (582.99 × 104 km2), implying that certain historically suitable habitats forfeited ecological suitability due to excessive warming or radical precipitation regime shifts. Similarly, under SSP585 (2061–2080), the contraction area (814.33 × 104 km2) markedly surpassed the expansion zone (459.78 × 104 km2), suggesting that extreme warming may precipitate suitability decline in traditional epidemic regions such as Mediterranean climate zones. Comparative analysis across scenarios revealed that expansion zone increments generally outweighed contraction zones. Notably, under SSP585 during 2041–2060 and 2081–2100, expansion zones exceeded contraction zones by 337.21 × 104 km2 and 186.26 × 104 km2, respectively, disclosing a net positive effect of global warming on this quarantine pathogen at the planetary scale. This asymmetric transformation indicates that, notwithstanding suitability loss in some regions due to climatic stress, the rate and magnitude of emerging habitat creation were more pronounced, heralding a potential systematic migration of the geographic focus for future global plant quarantine governance.

3.6. Centroid Shift in the Suitable Habitats of P. ramorum

Under future climate change scenarios, the global potential suitable habitat of Phytophthora ramorum exhibits significant centroid migration and spatial reconfiguration characteristics (Figure 5; Table S2). Currently, the centroid is located in Murzuq City, Murzuq District, Libya (20.731501° E, 21.652051° N). Under the SSP126 scenario, the centroid generally shifts southeastward. From 2021 to 2040, it migrates 63.95 km southeast to Jalu County, Al Kufrah Province, Libya (21.168444° E, 21.252427° N). During 2041–2060, it continues southeastward for 90.83 km within the same province. From 2061 to 2080, it shifts markedly southeastward by 381.58 km to Al Kufrah County, Al Kufrah Province (24.161188° E, 21.519164° N). In the period 2081–2100, it shifts back southwestward by 58.38 km, returning to Murzuq Province. Under the SSP370 scenario, the centroid oscillates within Libya. During 2021–2040, it shifts 137.56 km northwest to Murzuq Province (19.503143° E, 21.848939° N). Subsequently, from 2041 to 2080, it undergoes continuous modest southwestward shifts, with migration distances ranging between 82.54 km and 84.99 km. From 2081 to 2100, it shifts again northwestward by 104.78 km. Under the SSP585 scenario, the centroid migration is most pronounced. During 2021–2040, it shifts 97.29 km northeast to Ghat Province (21.237743° E, 22.508962° N). From 2041 to 2060, it moves 154.5 km northwest to the northern part of the same province. During 2061–2080, it shifts substantially southwestward by 455.54 km to Dirkou Department, Agadez Region, Niger (16.640797° E, 21.062766° N). From 2081 to 2100, it continues northwestward by 290.91 km to Fada, Ennedi Region, Chad (18.40423° E, 22.828064° N). These trends indicate that high-emission scenarios drive a pronounced centroid shift toward the Sahel zone of Africa, whereas under medium- and low-emission scenarios, the centroid remains relatively stable within northern Libya. This reflects the profound influence of different socioeconomic pathways on the geographic distribution of quarantine pathogens.

4. Discussion

4.1. Model Robustness and Climatic Determinism in Pathogen Distribution

The optimized MaxEnt model (AUC = 0.978) demonstrates exceptional discriminatory capacity, substantially exceeding performance metrics typical for forest pathogen SDMs [56,57]. The ΔAICc-driven parameter selection (FC = QH, RM = 1.5) mitigates overfitting risks common with high-dimensional occurrence data, ensuring transferability across climate scenarios [58]. The primacy of temperature seasonality (29.4% contribution) over edaphic variables reflects P. ramorum’s oomycete physiology, where thermal niche conservatism governs overwintering survival and sporangia production [55]. The identified Bio06 threshold (−5.39 to 5.61 °C) precisely aligns with chilling requirements for zoospore release, while Bio19 (118.01–434.01 mm) captures critical winter moisture windows. This climatic determinism challenges microhabitat-centric paradigms, corroborating recent evidence that macroclimatic filtering dominates continental-scale pathogen biogeography [59].

4.2. “Habitat Churning” Versus Poleward Migration Under Climate Change

Contrary to simplistic range-shift hypotheses, our results reveal a “habitat churning” phenomenon where total suitable area fluctuates non-monotonically while maintaining a stable hierarchical structure (Slight > moderately > highly suitability). This asymmetrical reorganization—where contraction zones (e.g., Mediterranean regions, 814.33 × 104 km2 under SSP585) can exceed expansion zones—challenges the climate-velocity framework [60,61]. P. ramorum does not track isotherms continuously; instead, it opportunistically colonizes fragmented, ephemeral climate windows at high latitudes (southern Canada, Scandinavia) while abandoning traditional foci experiencing precipitation regime shifts. This pattern mirrors Fusarium circinatum dynamics [62,63] but diverges in its Sahelward trajectory, revealing species-specific responses to winter precipitation changes rather than annual mean temperature. The net positive habitat creation under SSP585 (+337.21 × 104 km2) suggests global warming may amplify invasion debt, particularly in cold-limited regions where host naivety is high.

4.3. Centroid Migration Toward the Sahel: An Emerging Phytosanitary Frontier

This study found that the distribution centroid of the pathogen has shifted significantly southwestward by approximately 455.54 km. Unlike the poleward migration documented for Heterobasidion annosum [64], this equatorward shift reflects an intensification of winter precipitation in arid subtropical regions, creating transient but viable “hopscotch corridors” [65]. Human-mediated long-distance dispersal via trade routes will increasingly intersect these emerging hotspots, accelerating jump dispersal beyond natural spread rates [66]. The oscillation of the centroid within Libya under the medium-emission scenario (SSP370), in contrast to its sustained migration toward the Sahel under SSP585, underscores that socioeconomic pathways fundamentally reshape the geography of invasion, not merely the magnitude of climate change. It should be noted that centroid analysis reflects the weighted average center of spatial distribution, and its displacement indicates the macro-level trend of overall distribution range changes. However, the centroid itself may be located in areas that are actually unsuitable habitats (such as the Sahel region in this study) [67]. The quarantine significance of this trend lies in revealing that the potential distribution range of the pathogen is undergoing systematic shifts driven by climate change, rather than indicating that the disease has already spread or will inevitably break out at the specific geographic location of the centroid (e.g., the Sahel). Future risk assessments should integrate actual suitable habitats.

4.4. Hierarchical Stability and Adaptive Management Strategies

The persistent low-suitability dominance (54.46–58.52%) across all scenarios reveals P. ramorum as an “opportunistic generalist,” sacrificing optimal habitats for broader marginal occupancy. This paradoxical strategy demands dynamic risk stratification: high-suitability zones (California, Iberia) require containment-focused rapid response, while moderate-suitability frontiers (southern Scandinavia, Asian temperate zones) need enhanced monitoring and host resistance deployment. The concentration of 81.7% explained variance in five variables enables parsimonious early warning systems, particularly where Bio04 and Bio19 approach critical thresholds. Scientific value: our framework transitions SDMs from static maps to process-informed decision tools, integrating niche constraints with spatial reorganization metrics. Future work must mechanistically link these climate thresholds to infection cycles under extreme weather events [68,69] and incorporate host distribution heterogeneity to refine epidemic-risk versus establishment-risk zones.

4.5. Limitations and Forward-Looking Perspectives

While this study reveals the potential distribution trends of pathogens under climate change through model predictions, it also acknowledges inherent limitations. The model is based on the assumption of niche conservatism, which fails to fully incorporate key processes such as pathogen adaptive evolution, dynamic succession of host communities, and anthropogenic dispersal interventions—factors that will fundamentally reshape future actual distribution patterns [70,71]. Consequently, the present findings should be interpreted as projections of ecological suitability under specific scenarios rather than deterministic forecasts, and their implications for public health or ecological risk are therefore contingent upon these uncertain dynamics. Additionally, constrained by data and methodological limitations, current predictions should be regarded more as macro-scale scenario projections rather than precise forecasts. To better assess disease dynamics and potential shifts, future efforts must prioritize the integration of high-resolution, time-series data on host populations, vector distributions, climate variables at relevant scales, and anthropogenic factors such as land-use change and transportation networks [72,73]. Future research should focus on developing a dynamic “climate-host-human activity” coupled framework that integrates high-resolution host distribution data, global logistics pathways, and the micro-level mechanisms of pathogen transmission, thereby advancing assessment methods from static suitability mapping to dynamic risk process simulation. This integrated framework is also highly applicable to studying a broad spectrum of other climate-sensitive pathogens, particularly plant pathogenic fungi (e.g., Fusarium head blight fungus, rice blast fungus) and plant diseases transmitted by insects or other vectors, whose distributional shifts are increasingly concerning [74,75]. This shift is crucial for addressing the phenomenon of “shifts in risk centroids” identified in this study and underscores the need for a more flexible and forward-looking risk management system to adapt to the evolving risk landscape [76].

5. Conclusions

The optimized MaxEnt model (AUC = 0.978) demonstrated robust performance in delineating global habitat suitability for Phytophthora ramorum. Temperature seasonality (29.4% contribution) emerged as the principal environmental driver, followed by precipitation of the coldest quarter (20.4%) and minimum temperature of the coldest month (18.2%), with the top five variables collectively explaining 81.7% of distribution patterns. Under the current climate, suitable habitat encompasses 4247.64 × 104 km2, predominantly in mid-latitude maritime regions. Future projections across SSP126, SSP370, and SSP585 scenarios reveal dynamic stability in total suitable area (3992.88–4590.16 × 104 km2) while exhibiting complex internal reorganization. High-emission pathways drive significant poleward and Sahel expansion, generating net habitat gains that exceed contractions. Centroid migration intensifies under SSP585, shifting toward the Sahel by 2100. These findings underscore climate warming’s dual role in redistributing pathogen risk asymmetrically, expanding threat into novel regions while reducing suitability in some traditional epidemic zones, thereby necessitating proactive adaptation of international plant quarantine strategies and enhanced surveillance in emerging high-risk areas.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/agriculture16050505/s1, Table S1. Environmental factors used in the study and their contribution rates. Table S2. Centroid migration of P. ramorum. Figure S1. Comparative analysis of the potential geographic distribution of P. ramorum under current and future climate scenarios. A: Expansion area of suitable habitat for P. ramorum. B: Contraction area of the suitable range for P. ramorum. Figure S2. Future suitable habitats and spatial changes in P. ramorum under the SSP1-2.6 scenario. Figure S3. Future suitable habitats and spatial changes in P. ramorum under the SSP3-7.0 scenario.

Author Contributions

Formal analysis, B.Z.; Investigation, B.Z., S.L., X.Z. and T.D.; Data curation, B.Z., S.L., X.Z. and T.D.; Writing—original draft, B.Z.; Writing—review and editing, B.Z., S.L., X.Z. and T.D.; Visualization, B.Z.; Project administration, T.D. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (32471873), the China Postdoctoral Science Foundation (2024M751426, 2025T180541), the National Key R&D Program of China (2023YFD1401304), the Natural Science Foundation of Jiangsu Province, China (BK20231291), and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

Data Availability Statement

The occurrence data of Phytophthora ramorum used in this study were obtained from publicly available databases. These datasets were sourced from the Global Biodiversity Information Facility (GBIF Occurrence Download. Available online: https://doi.org/10.15468/dl.paw2dy (accessed on 10 October 2025)), the Centre for Agriculture and Bioscience International (CABI) Compendium (https://www.cabi.org/; accessed on 29 December 2025), and the EPPO Global Database (https://gd.eppo.int/; accessed on 30 December 2025). The environmental data layers used for modeling are also publicly available from their respective sources as cited in the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Chen, S.; Chen, J.; Jiang, C.; Yao, R.T.; Xue, J.; Bai, Y.; Wang, H.; Jiang, C.; Wang, S.; Zhong, Y. Trends in research on forest ecosystem services in the most recent 20 years: A bibliometric analysis. Forests 2022, 13, 1087. [Google Scholar] [CrossRef]
  2. Brouwer, R.; Bongers, F.; Peña-Claros, M.; Zuidema, P.A.; Brancalion, P.; Lohbeck, M.; Guzmán, A.H.; Heinze, A.; Guillemot, J.; Kramer, K. Forest restoration, biodiversity, and ecosystem services. In Restoring Forests and Trees for Sustainable Development; Oxford University Press: New York, NY, USA, 2024; pp. 160–198. [Google Scholar]
  3. EFSA Panel on Plant Health (PLH). Scientific Opinion on the Pest Risk Analysis on Phytophthora ramorum prepared by the FP6 project RAPRA. EFSA J. 2011, 9, 2186. [Google Scholar]
  4. Garbelotto, M.; Hayden, K.J. Sudden oak death: Interactions of the exotic oomycete Phytophthora ramorum with naïve North American hosts. Eukaryot. Cell 2012, 11, 1313–1323. [Google Scholar] [CrossRef] [PubMed]
  5. Pastalka, T.; Abeysekara, N.; Schweigkofler, W. First report of Phytophthora ramorum on Cotoneaster sp. in the USA. Plant Dis. 2023, 108, 231. [Google Scholar] [CrossRef]
  6. Grünwald, N.J.; Goss, E.M.; Press, C.M. Phytophthora ramorum: A pathogen with a remarkably wide host range causing sudden oak death on oaks and ramorum blight on woody ornamentals. Mol. Plant Pathol. 2008, 9, 729–740. [Google Scholar] [CrossRef]
  7. Kartakis, S.; Jung, T.; Kriticos, D.J.; Wesseler, J. Modeling disease expression of Phytophthora ramorum to estimate potential economic impacts in European forests. For. Ecol. Manag. 2026, 601, 123367. [Google Scholar] [CrossRef]
  8. Venette, R.C.; Cohen, S.D. Potential climatic suitability for establishment of Phytophthora ramorum within the contiguous United States. For. Ecol. Manag. 2006, 231, 18–26. [Google Scholar] [CrossRef]
  9. Grünwald, N.J.; Garbelotto, M.; Goss, E.M.; Heungens, K.; Prospero, S. Emergence of the sudden oak death pathogen Phytophthora ramorum. Trends Microbiol. 2012, 20, 131–138. [Google Scholar] [CrossRef]
  10. Webber, J.F. Status of Phytophthora ramorum and P. kernoviae in Europe. In Proceedings of the Sudden Oak Death Third Science Symposium; General Technical Report PSW-GTR-214; Frankel, S.J., Kliejunas, J.T., Palmieri, K.M., Eds.; U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station: Albany, CA, USA, 2008; pp. 19–26. [Google Scholar]
  11. Ireland, K.B.; Hardy, G.E.S.J.; Kriticos, D.J. Combining inferential and deductive approaches to estimate the potential geographical range of the invasive plant pathogen, Phytophthora ramorum. PLoS ONE 2013, 8, e63508. [Google Scholar] [CrossRef]
  12. Liao, T.L.; Li, B.S. Risk Analysis of Phytophthora ramorum Imported into China. J. Southwest For. Coll. 2004, 24, 34–37. [Google Scholar] [CrossRef]
  13. Davidson, J.M.; Patterson, H.A.; Wickland, A.C.; Fichtner, E.J.; Rizzo, D.M. Forest type influences transmission of Phytophthora ramorum in California oak woodlands. Phytopathology 2011, 101, 492–501. [Google Scholar] [CrossRef]
  14. Green, S.; Cooke, D.E.; Dunn, M.; Barwell, L.; Purse, B.; Chapman, D.S.; Valatin, G.; Schlenzig, A.; Barbrook, J.; Pettitt, T. PHYTO-THREATS: Addressing threats to UK forests and woodlands from Phytophthora; identifying risks of spread in trade and methods for mitigation. Forests 2021, 12, 1617. [Google Scholar] [CrossRef]
  15. Bussell, E.; Cunniffe, N. Applying optimal control theory to a spatial simulation model of sudden oak death: Ongoing surveillance protects tanoak while conserving biodiversity. J. R. Soc. Interface 2020, 17, 20190671. [Google Scholar] [CrossRef]
  16. Søndreli, K.L.; Tabima, J.F.; LeBoldus, J.M. Rapid new diagnostic LAMP (loop-mediated isothermal amplification) assays to distinguish among the four lineages of Phytophthora ramorum. Plant Dis. 2023, 107, 3553–3559. [Google Scholar] [CrossRef] [PubMed]
  17. Ahmad, H.; Rajput, N.A.; Atiq, M.; Kachelo, G.A.; Usman, M.; Tariq, H.; Wahab, M. Detection of Phytophthora nicotianae induced citrus gummosis by the loop-mediated isothermal amplification. Pak. J. Bot. 2024, 56, 1–11. [Google Scholar] [CrossRef] [PubMed]
  18. Li, Q.; Shao, H. The Role of Pathogens in Plant Invasion: Accumulation of Local Pathogens Hypothesis. Microb. Ecol. 2024, 87, 178. [Google Scholar] [CrossRef] [PubMed]
  19. Hunter, S.; Waipara, N.; Burns, B.; Scott, P.; Williams, N. Survival and dispersal of Phytophthora multivora zoospores in soil substrates. Australas. Plant Pathol. 2025, 54, 289–297. [Google Scholar] [CrossRef]
  20. Blaker, N.S. Influence of Container Medium pH on Sporangium Formation, Zoospore Release, and Infection of Rhododendron by Phytophthora cinnamomi. Plant Dis. 1983, 67, 259–263. [Google Scholar] [CrossRef]
  21. Thomsen, I.M.; Alsanius, B. Updated pest risk assessment of Phytophthora ramorum in Norway. In Report from the Norwegian Scientific Committee for Food and Environment (VKM); Norwegian Scientific Committee for Food and Environment: Oslo, Norway, 2023. [Google Scholar]
  22. Xu, J.; Zhang, H.; Qiao, Y.; Yuan, H.; Xu, W.; Xia, X. Effects of Microtopography on Neighborhood Diversity and Competition in Subtropical Forests. Plants 2025, 14, 870. [Google Scholar] [CrossRef]
  23. Martens, M. Ring Test for Recombinase Polymerase Amplification (RPA) of Phytophthora ramorum. Master’s Thesis, Cornell University, Ithaca, NY, USA, 2023. [Google Scholar]
  24. Van Poucke, K.; Franceschini, S.; Webber, J.F.; Vercauteren, A.; Turner, J.A.; McCracken, A.R.; Heungens, K.; Brasier, C.M. Discovery of a fourth evolutionary lineage of Phytophthora ramorum: EU2. Fungal Biol. 2012, 116, 1178–1191. [Google Scholar] [CrossRef]
  25. Niu, Z.; Han, D.; Niu, Z.; Roussos, P.A.; Cheng, T.; Xie, J.; Jin, N.; Zhang, D. Assessment of potential habitat suitability for kiwifruit (Actinidia spp) in Shaanxi Province under climate change scenarios. Front. Plant Sci. 2025, 16, 1617802. [Google Scholar] [CrossRef] [PubMed]
  26. Campos, J.C.; Garcia, N.; Alírio, J.; Arenas-Castro, S.S.; Teodoro, A.; Sillero, N. Ecological Niche models using MaxEnt in Google Earth engine: Evaluation, guidelines and recommendations. Ecol. Inform. 2023, 76, 102147. [Google Scholar] [CrossRef]
  27. Lissovsky, A.A.; Dudov, S.V. Species-Distribution Modeling: Advantages and Limitations of Its Application. 2. MaxEnt. Biol. Bull. Rev. 2021, 11, 265–275. [Google Scholar] [CrossRef]
  28. Steven, J.P.; Miroslav, D.; Schapire, R. Maxent Software for Modeling Species Niches and Distributions (Version 3.4.1). Available online: http://biodiversityinformatics.amnh.org/open_source/maxent/ (accessed on 12 January 2017).
  29. Fan, Y.; Zhang, X.; Yang, J.; Yang, J.; Zhang, H.; Yang, B.; Li, X. Prediction of global potential distribution and assessment of habitat suitability for Xanthium spinosum driven by climate change. Front. Plant Sci. 2025, 16, 1690546. [Google Scholar] [CrossRef]
  30. Mu, S.; Zhen, J.; Xi, C.; Wang, L. Suitability Assessment of Pastoral Human Settlements in Xilingol League Based on an Optimized MaxEnt Model. Land 2025, 14, 2052. [Google Scholar] [CrossRef]
  31. Lane, S.L.; Erland, L.A. Differential temperature adaptation mechanisms in the High Arctic-adapted Cerastium regelii Ostenf. and the widespread Stellaria longipes Goldie. bioRxiv 2025. [Google Scholar] [CrossRef]
  32. Xiang, Y.; Li, S.; Yang, Q.; Liu, J.; Liu, Y.; Zhao, L.; Lin, H.; Luo, Y.; Ren, J.; Luo, X. Modeling the Habitat Suitability and Range Shift of Daphniphyllum macropodum in China Under Climate Change Using an Optimized MaxEnt Model. Biology 2025, 14, 1360. [Google Scholar] [CrossRef]
  33. Fan, L.; Mi, C.; Li, J.; Zhang, Y.; Zhang, H.; Zhang, G.; Wang, H. Projecting global shifts in the invasive potential of Bidens pilosa L. under climate change using species distribution models. Front. Plant Sci. 2025, 16, 1580278. [Google Scholar] [CrossRef]
  34. Lin, Z.; Halliday, F.W.; Zhang, P.; Wang, X.; Chen, F.; Shi, A.; Shi, J.; Xiao, Y.; Liu, X. Above-and belowground plant pathogens along elevational gradients: Patterns and potential mechanisms. Oikos 2025, 2025, e10455. [Google Scholar] [CrossRef]
  35. Serrano, M.S.; Romero, M.Á.; Homet, P.; Gómez-Aparicio, L. Climate change impact on the population dynamics of exotic pathogens: The case of the worldwide pathogen Phytophthora cinnamomi. Agric. For. Meteorol. 2022, 322, 109002. [Google Scholar] [CrossRef]
  36. Shamoun, S.F.; Elliott, M. Phytophthora ramorum Werres, de Cock & Man in’t Veld, Sudden Oak Death/Encre des chênes rouges (Peronosporaceae). In Biological Control Programmes in Canada 2013–2023; CABI: Oxfordshire, UK, 2024; pp. 593–602. [Google Scholar]
  37. Brown, J.L. SDMtoolbox: A python-based GIS toolkit for landscape genetic, biogeographic and species distribution model analyses. Methods Ecol. Evol. 2014, 5, 694–700. [Google Scholar] [CrossRef]
  38. Waititu, J.M.; Mundia, C.N.; Sichangi, A.W. Assessing distribution changes of selected native and alien invasive plant species under changing climatic conditions in Nyeri County, Kenya. PLoS ONE 2020, 17, e0275360. [Google Scholar] [CrossRef]
  39. Aparecido, L.E.D.O.; Lorenone, J.A.; Lorenone, P.A.; Torsoni, G.B.; Lima, R.F.D.; Chiquitto, A.G.; Saqui, D.; Oliveira Júnior, G.G.d.; Rolim, G.D.S. Climate Change and Phoma spp. Leaf Spot of Arabica Coffee: A CMIP6 Modeling Approach. Rev. Bras. Meteorol. 2023, 38, e38230041. [Google Scholar] [CrossRef]
  40. Xiong, T.; Wang, S.; Kang, F.; Liu, J.; Qin, Y. The Projected Effects of Climate Change on the Potential Distribution of Planococcus minor Based on Ensemble Species Distribution Models. Agronomy 2025, 15, 1165. [Google Scholar] [CrossRef]
  41. Wang, C.; Wang, J.; Naudiyal, N.; Wu, N.; Cui, X.; Wei, Y.; Chen, Q. Multiple effects of topographic factors on Spatio-temporal variations of vegetation patterns in the three parallel rivers region, Southeast Qinghai-Tibet Plateau. Remote Sens. 2021, 14, 151. [Google Scholar] [CrossRef]
  42. Nachtergaele, F.; Van Velthuizen, H.; Verelst, L.; Batjes, N.; Dijkshoorn, K.; van Engelen, V.; Fischer, G.; Jones, A.; Montanarella, L.; Petri, M. Harmonized world soil database. In Proceedings of the 19th World Congress of Soil Science, Soil Solutions for a Changing World, Brisbane, Australia, 1–6 August 2010; pp. 34–37. [Google Scholar]
  43. Zhou, J.; Tang, Q.; Zong, D.; Hu, X.; Wang, B.; Wang, T. Drivers of species distribution and niche dynamics for ornamental plants originating at different latitudes. Diversity 2023, 15, 877. [Google Scholar] [CrossRef]
  44. Xue, N.; Li, K.; Chen, K.; Li, P.; Ji, X.; Ma, Z.; Ji, W. Predicting climate change impacts on distribution and conservation of critically endangered Picea neoveitchii using MaxEnt. Front. For. Glob. Change 2024, 7, 1472857. [Google Scholar] [CrossRef]
  45. Fang, H.Q.; Zhang, P.F.; Xu, S.W.; Xu, T.; He, B.; Wang, E.; Dong, C.W.; Yang, Q.S. The ecological suitability area of Cirsium lineare (Thunb.) Sch.-Bip. under future climate change in China based on MaxEnt modeling. Ecol. Evol. 2024, 14, e10848. [Google Scholar] [CrossRef]
  46. Zhang, M.; Sun, X.; Miao, Y.; Li, M.; Huang, L. Cordyceps cicadae and Cordyceps gunnii have closer species correlation with Cordyceps sinensis: From the perspective of metabonomic and MaxEnt models. Sci. Rep. 2022, 12, 20469. [Google Scholar] [CrossRef]
  47. Vignoles, A. Guide francophone pour la modélisation de niches écologiques. Biodivers. Inform. 2022, 17, 67–95. [Google Scholar] [CrossRef]
  48. Xue, Y.; Lin, C.; Wang, Y.; Zhang, Y.; Ji, L. Ecological niche complexity of invasive and native cryptic species of the Bemisia tabaci species complex in China. J. Pest Sci. 2022, 95, 1245–1259. [Google Scholar] [CrossRef]
  49. Palagi, E.; Norscia, I. Bonobos protect and console friends and kin. PLoS ONE 2013, 8, e79290. [Google Scholar] [CrossRef] [PubMed]
  50. Kantidakis, G.; Putter, H.; Litière, S.; Fiocco, M. Statistical models versus machine learning for competing risks: Development and validation of prognostic models. BMC Med. Res. Methodol. 2023, 23, 51. [Google Scholar] [CrossRef] [PubMed]
  51. Al Ruheili, A.M.; Boluwade, A.; Al Subhi, A.M. Assessing the impact of climate change on the distribution of lime (16SRII-b) and alfalfa (16srii-d) phytoplasma disease using maxent. Plants 2021, 10, 460. [Google Scholar] [CrossRef]
  52. Multimodality Writing Group for Chronic Coronary Disease; Winchester, D.E.; Maron, D.J.; Blankstein, R.; Chang, I.C.; Kirtane, A.J.; Kwong, R.Y.; Pellikka, P.A.; Prutkin, J.M.; Russell, R. ACC/AHA/ASE/ASNC/ASPC/HFSA/HRS/SCAI/SCCT/SCMR/STS 2023 multimodality appropriate use criteria for the detection and risk assessment of chronic coronary disease. J. Am. Coll. Cardiol. 2023, 81, 2445–2467. [Google Scholar] [CrossRef]
  53. Metheny, N.; Kitson, J.; Graves, A.; Shariq, A.; Wright, K.; Taylor, L.; Cryer, J.; Raimondi, J.; Schipper, S.; Binny, J.M.S. Applied Artificial Intelligence-Capstone Chronicles 2024; University of San Diego: San Diego, CA, USA, 2024. [Google Scholar]
  54. Sun, Z.; Zhao, T.; Hu, S.; Chen, G.; Zhu, W.; Peng, X.; Su, S. Integrating climate change and fine-scale habitat suitability to assess amphibian range shift in Mount Emei, China. Front. Zool. 2025, 22, 16. [Google Scholar] [CrossRef]
  55. Dun, H.F.; MacKay, J.J.; Green, S. Expansion of natural infection of Japanese larch by Phytophthora ramorum shows trends associated with seasonality and climate. Plant Pathol. 2024, 73, 419–430. [Google Scholar] [CrossRef]
  56. Khalaf, S.M.; Alqahtani, M.S.; Selim, Y.A.; Elsayed, K.O.; Bendary, H.A. Spatiotemporal dynamics of Bacillus anthracis under climate change: A machine learning approach. Front. Microbiol. 2025, 16, 1659876. [Google Scholar] [CrossRef]
  57. Ogris, N.; Drenkhan, R.; Vahalík, P.; Cech, T.; Mullett, M.; Tubby, K. The potential global distribution of an emerging forest pathogen, Lecanosticta acicola, under a changing climate. Front. For. Glob. Change 2023, 6, 1221339. [Google Scholar] [CrossRef]
  58. Negi, P.S.; Paul, S.; Lata, S. MaxEnt distribution modelling of Betula utilis D.Don under changing climatic conditions in North Western Himalaya, India. Discov. Geosci. 2025, 3, 214. [Google Scholar] [CrossRef]
  59. Caballol, M.; Redondo, M.Á.; Catalán, N.; Corcobado, T.; Jung, T.; Marçais, B.; Milenković, I.; Nemesio-Gorriz, M.; Stenlid, J.; Oliva, J. Climate acts as an environmental filter to plant pathogens. ISME J. 2024, 18, wrae010. [Google Scholar] [CrossRef] [PubMed]
  60. Ragab, S.H.; Alqurashi, S.I.; Aljameeli, M.M.; Tyshenko, M.G.; Abdelwahab, A.H.; Selim, T.A. Predicting the Global Distribution of Gryllus bimaculatus Under Climate Change: Implications for Biodiversity and Animal Feed Production. Sustainability 2024, 16, 10278. [Google Scholar] [CrossRef]
  61. Cai, C.; Zhang, X.; Zha, J.; Li, J.; Li, J. Predicting climate change impacts on the rare and endangered Horsfieldia tetratepala in China. Forests 2022, 13, 1051. [Google Scholar] [CrossRef]
  62. Nordström, I.; Sherwood, P.; Bohman, B.; Woodward, S.; Peterson, D.L.; Niño-Sánchez, J.; Sánchez-Gómez, T.; Díez, J.J.; Cleary, M. Utilizing volatile organic compounds for early detection of Fusarium circinatum. Sci. Rep. 2022, 12, 21661. [Google Scholar] [CrossRef]
  63. Zhang, X.; Chen, C.; Wang, F.; Dai, T. Predicting the Global Distribution of Fusarium circinatum Using MaxEnt Modeling. Agronomy 2025, 15, 1913. [Google Scholar] [CrossRef]
  64. Shen, S.; Zhang, X.; Jian, S. The Distributional Range Changes of European Heterobasidion Under Future Climate Change. Forests 2024, 15, 1863. [Google Scholar] [CrossRef]
  65. Sturrock, R.N.; Frankel, S.J.; Brown, A.V.; Hennon, P.E.; Kliejunas, J.T.; Lewis, K.J.; Worrall, J.J.; Woods, A.J. Climate change and forest diseases. Plant Pathol. 2011, 60, 133–149. [Google Scholar] [CrossRef]
  66. Runghen, R.; Llopis-Belenguer, C.; Stouffer, D.B. Using network analysis to study and manage human-mediated dispersal of exotic species. Biol. Invasions 2023, 25, 3369–3389. [Google Scholar] [CrossRef]
  67. Liu, B.; Li, L.; Zhang, Z.; Ran, H.; Xing, M. Modeling the Potential Distribution and Future Dynamics of Important Vector Culex tritaeniorhynchus Under Climate Change Scenarios in China. Insects 2025, 16, 382. [Google Scholar] [CrossRef]
  68. Sinha, P.; Chen, X. Potential infection risks of the wheat stripe rust and stem rust pathogens on barberry in Asia and southeastern Europe. Plants 2021, 10, 957. [Google Scholar] [CrossRef]
  69. Jiranek, J.; Miller, I.F.; An, R.; Bruns, E.; Metcalf, C.J.E. Mechanistic models to meet the challenge of climate change in plant–pathogen systems. Philos. Trans. R. Soc. B 2023, 378, 20220017. [Google Scholar] [CrossRef]
  70. Tang, X.; Deng, Y.; He, Z.; Zhou, M.; Yuan, Y.; Zeng, K. Modelling the potential distribution and niche shift of Solenopsis invicta Buren under climate change and invasion process. Front. For. Glob. Change 2025, 8, 1659630. [Google Scholar] [CrossRef]
  71. Lin, M.; Ye, X.; Zhao, Z.; Chen, S.; Liu, B. Comparative Analysis of Habitat Expansion Mechanisms for Four Invasive Amaranthaceae Plants Under Current and Future Climates Using MaxEnt. Plants 2025, 14, 2363. [Google Scholar] [CrossRef]
  72. Padilla, D.G. Modeling the Role of Land-Use Change on the Spread of Infectious Disease; Arizona State University: Tempe, AZ, USA, 2020. [Google Scholar]
  73. Hoffmann, P.; Rechid, D.; Reinhart, V.; Asmus, C.; Davin, E.; Katragkou, E.; De Noblet-Ducoudré, N.; Bechtel, B.; Böhner, J. Long-term high-resolution land-use/land-cover change dataset for regional climate modeling. In AGU Fall Meeting Abstracts; American Geophysical Union: Washington, DC, USA, 2020; p. GC007-0008. [Google Scholar]
  74. Tagyan, A.I.; AlAshaal, S.; Alkhalifah, D.H.M.; Abdelghany, W.R.; Hozzein, W.N. Estimating the climate change-driven global distribution of Fusarium proliferatum and mycotoxin risk assessment under future warming scenarios. Front. For. Glob. Change 2025, 8, 1673494. [Google Scholar] [CrossRef]
  75. Skendžić, S.; Zovko, M.; Živković, I.P.; Lešić, V.; Lemić, D. The impact of climate change on agricultural insect pests. Insects 2021, 12, 440. [Google Scholar] [CrossRef]
  76. Qiao, H.; Peterson, A.T.; Myers, C.E.; Yang, Q.; Saupe, E.E. Ecological niche conservatism spurs diversification in response to climate change. Nat. Ecol. Evol. 2024, 8, 13. [Google Scholar] [CrossRef] [PubMed]
Agriculture 16 00505 g001
Figure 1. Predicted global distribution and climatic suitability of Phytophthora ramorum under current climate conditions. (A) Global known distribution of P. ramorum; (B) Potential climatic suitability of P. ramorum predicted by the optimal MaxEnt model; (C) Global known distribution of its major host plants; (D) Predicted potential distribution of major host plants under a future climate scenario.
Figure 1. Predicted global distribution and climatic suitability of Phytophthora ramorum under current climate conditions. (A) Global known distribution of P. ramorum; (B) Potential climatic suitability of P. ramorum predicted by the optimal MaxEnt model; (C) Global known distribution of its major host plants; (D) Predicted potential distribution of major host plants under a future climate scenario.
Agriculture 16 00505 g001
Agriculture 16 00505 g002
Figure 2. Evaluating the accuracy of maxent models in predicting the potential geographic range of P. ramorum. (A) Correlation analysis and selection of critical climatic variables; (B) Optimization of feature combinations and regularization parameters in the model; (C) Model performance assessment based on the mean AUC value derived from repeated cross-validation.
Figure 2. Evaluating the accuracy of maxent models in predicting the potential geographic range of P. ramorum. (A) Correlation analysis and selection of critical climatic variables; (B) Optimization of feature combinations and regularization parameters in the model; (C) Model performance assessment based on the mean AUC value derived from repeated cross-validation.
Agriculture 16 00505 g002
Agriculture 16 00505 g003
Figure 3. Response Curves of Key Environmental Variables in the MaxEnt Model for P. ramorum. (A) A jackknife test was employed to analyze the contributions of key environmental variables to the geographic distribution of P. ramorum. (B) Bio04, temperature seasonality. (C) Bio06, Minimum temperature of coldest month. (D) Bio19, Precipitation of the coldest quarter.
Figure 3. Response Curves of Key Environmental Variables in the MaxEnt Model for P. ramorum. (A) A jackknife test was employed to analyze the contributions of key environmental variables to the geographic distribution of P. ramorum. (B) Bio04, temperature seasonality. (C) Bio06, Minimum temperature of coldest month. (D) Bio19, Precipitation of the coldest quarter.
Agriculture 16 00505 g003
Agriculture 16 00505 g004
Figure 4. Future suitable habitats and spatial changes in P. ramorum under the SSP5-8.5 scenario.
Figure 4. Future suitable habitats and spatial changes in P. ramorum under the SSP5-8.5 scenario.
Agriculture 16 00505 g004
Agriculture 16 00505 g005
Figure 5. Migration of the Range Centroid for P. ramorum in Response to Divergent Climate Futures.
Figure 5. Migration of the Range Centroid for P. ramorum in Response to Divergent Climate Futures.
Agriculture 16 00505 g005
Table 1. Environmental variables for screening, contribution and displacement.
Table 1. Environmental variables for screening, contribution and displacement.
NumberVariablePercent Contribution (%)Permutation Importance (%)
1Bio0429.423.1
2Bio1920.423.9
3Bio0618.226.9
4T_gravel74.4
5Slpoe61.7
6Aspect5.73.9
7Bio1445.3
8T_bulk3.23.7
9T_usda2.42.1
10T_silt2.43.8
11T_ece1.31.4
Table 2. Predicted suitable habitat area of P. ramorum across time periods.
Table 2. Predicted suitable habitat area of P. ramorum across time periods.
SSP ScenarioTime PeriodTotal AreaHighly SuitableModerately SuitablePoorly Suitable
(104 km2/Change %)(104 km2/Change %)(104 km2/Change %)(104 km2/Change %)
Historical1970–20004247.64/-631.97/-1302.34/-2313.33/-
SSP1262021–20404512.30/+6.23%640.29/+1.31%1412.40/+8.45%2459.61/+6.32%
2041–20604341.14/+2.20%598.00/−5.38%1327.04/+1.90%2416.11/+4.44%
2061–20804418.18/+4.01%553.55/−12.41%1373.00/+5.42%2491.64/+7.71%
2081–21004306.47/+1.38%543.92/−13.94%1314.03/+0.90%2448.52/+5.84%
SSP3702021–20404149.67/−2.31%627.26/−0.75%1258.70/−3.35%2263.72/−2.14%
2041–20603992.88/−6.00%542.36/−14.18%1142.76/−12.26%2307.76/−0.24%
2061–20804111.40/−3.21%536.50/−15.11%1185.32/−8.98%2389.58/+3.30%
2081–21004303.00/+1.30%579.07/−8.37%1317.37/+1.15%2406.57/+4.03%
SSP5852021–20404295.29/+1.12%640.73/+1.39%1296.55/−0.44%2358.01/+1.93%
2041–20604590.16/+8.06%592.88/−6.19%1365.59/+4.86%2631.69/+13.77%
2061–20804539.00/+6.86%594.97/−5.86%1364.83/+4.80%2579.20/+11.49%
2081–21004433.42/+4.37%552.44/−12.59%1286.54/−1.21%2594.44/+12.16%
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zheng, B.; Lin, S.; Zhang, X.; Dai, T. Modeling Current and Future Global Distribution of the Quarantine Pathogen Phytophthora ramorum Using MaxEnt. Agriculture 2026, 16, 505. https://doi.org/10.3390/agriculture16050505

AMA Style

Zheng B, Lin S, Zhang X, Dai T. Modeling Current and Future Global Distribution of the Quarantine Pathogen Phytophthora ramorum Using MaxEnt. Agriculture. 2026; 16(5):505. https://doi.org/10.3390/agriculture16050505

Chicago/Turabian Style

Zheng, Bingyan, Sixi Lin, Xiaorui Zhang, and Tingting Dai. 2026. "Modeling Current and Future Global Distribution of the Quarantine Pathogen Phytophthora ramorum Using MaxEnt" Agriculture 16, no. 5: 505. https://doi.org/10.3390/agriculture16050505

APA Style

Zheng, B., Lin, S., Zhang, X., & Dai, T. (2026). Modeling Current and Future Global Distribution of the Quarantine Pathogen Phytophthora ramorum Using MaxEnt. Agriculture, 16(5), 505. https://doi.org/10.3390/agriculture16050505

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop