Introduction
Climate change is becoming recognised as one of the most significant causes of ecological imbalance, especially in the agricultural sector (Horn, Reference Horn2022; Baek et al., Reference Baek, Lee and Park2024). For insect pests, climate change can alter their distribution, overwintering survival, generation frequency, and phenological synchrony with host plants. Climate change also alters interspecific interactions, increases invasion risks, raises the rate of insect-borne diseases, and makes the biological control process less effective. The synergistic effects of these factors intensify crop damage and economic losses globally (Skendzic et al., Reference Skendzic, Zovko, Živković, Lešić and Lemić2021; Subedi et al., Reference Subedi, Poudel and Aryal2023).
Pochazia shantungensis (Chou & Lu, 1977) (Hemiptera, Fulgoromorpha, Ricaniidae) was first described in Shandong Province, China. This species has now spread from Asia across the globe and into Europe, with populations now established in Korea, Japan, Turkey, France, Germany, the Netherlands, and Italy (Bourgoin et al., Reference Bourgoin, Gros and Stroiński2020; Bras et al., Reference Bras, Lombaert, Kenis, Li, Bernard, Rousselet, Roques and Auger-Rozenberg2022). This species is a polyphagous insect that can attack more than 200 host plants from 81 families and typically causes economic losses to agricultural and forestry production (Matova et al., Reference Matova, Kamutando, Magorokosho, Kutywayo, Gutsa and Labuschagne2020; Pajac Beus et al., Reference Pajac Beus, Lemić, Skendžić, Čirjak and Pajač Živković2024). In South Korea, for instance, P. shantungensis caused significant yield reduction in apple and pear orchards, resulting in annual losses of 25–30 million dollars. In Europe, fruit-producing areas have suffered from yield losses ranging from 15% to 40% per year, leading to substantial economic losses (Johnson et al., Reference Johnson, Bloom, Dunning, Gunter, Boyette and Creamer2019).
Since the first description of P. shantungensis by Chou & Lu in 1977, global temperatures have increased by over 2°C with accelerated warming since 1975 (NASA, 2025). This warming has likely contributed to the spread of P. shantungensis in recent years. Therefore, we employ the MaxEnt model to predict the potential future distribution of P. shantungensis, as it is particularly suitable for predicting species distributions with limited occurrence data and has demonstrated high accuracy in forecasting pest invasion risks under various environmental scenarios. This modelling approach aims to elucidate predicted expansion patterns and identify climatically suitable regions for pest establishment worldwide. The MaxEnt-generated predictions will facilitate the development of proactive management strategies and guide international quarantine efforts to prevent further spread and control this economically significant pest (Ali et al., Reference Ali, Muhammad, Sanda, Huang and Hou2019, Reference Ali, Hou and Tahir2023a, Reference Ali, Noor Shah, Bilal Tahir, Fiaz and Ali2023b; Ali and Xiao, Reference Ali and Xiao2025).
Distribution of P. shantungensis (Chou & Lu, 1977) currently includes: China (Guangxi, Fujian, Henan, Hubei, Shaanxi, Shandong, Zhejiang); Japan (Honshu, Kyushu); Korea (Chungcheongnam-do, Jeollabuk-do, Jeollanam-do, Gyeonggi-do, Jeju-do); France (Alpes-Côte d’Azur, Corsica, Occitanie); Germany (Baden-Württemberg); the Netherlands (Ijsselmonde); Bulgaria (Burgas); Hungary (Baranya County); Italy (Pistoia region); Russia (Sochi, Krasnodar Krai); and Turkey (Istanbul, Kocaeli, Yalova). Currently, this species is not known to be present in the Americas, Africa, and Oceania (DEFRA, 2025), making these continents particularly important focus areas for predictive distribution modelling and risk assessment. The lack of established populations in these regions represents valuable agricultural and ecological resources requiring enhanced border biosecurity and early warning systems. Predicting the potential future distribution in these currently uninfested areas is essential for developing region-specific contingency plans and preventing the substantial economic losses already documented in established pest regions.
Materials and methods
Species distribution data collection
Occurrence records of P. shantungensis were collected from three sources: (1) the Center for Agriculture and Bioscience International (https://www.cabi.org/, accessed on 22 February 2025); (2) the Global Biodiversity Information Facility database (https://www.gbif.org/, accessed on 11 March 2025); and (3) published literature (Rahman et al., Reference Rahman, Kwon, Suh, Youn and Jo2012; Hizal et al., Reference Hizal, Oztemiz and Gjonov2019; Zhang et al., Reference Zhang, Fang, Zhao, Jiang, Stroiński and Qin2022; Kobayashi et al., Reference Kobayashi, Suzuki, Kuwahara, Park, Yamada and Jung2024). The overall number of occurrence records obtained using this method was 490 localities. All locations were converted into geocoordinates using Google Earth (http://ditu.google.cn/, accessed on 15 March 2025).
Occurrence data processing and validation
Distribution records were processed to remove entries lacking geographic coordinates, those with zero-value coordinates, and duplicate records. CoordinateCleaner (version 3.0.1) was employed for automated quality control. We used a spatial filtering tool for each grid cell, with a single distribution point and a resolution of 2.5 arc minutes. After filtering, we got 445 distribution records and placed them on a world map using ArcGIS 10.8.2 (http://www.arcgis.com/) (fig. 1). We downloaded the Globe boundary shapefile from the ‘Resource and Environmental Science Data Platform’ website (https://www.resdc.cn/, accessed 12 March 2025) using the method by Rashid et al. (Reference Rashid, Shi, Rahim, Qasim, Baloch, Bohnett, Yang, Khan and Ahmad2021).
Occurrence records of P. shantungensis globally.
Figure 1 Long description
A world map displays habitat suitability for P. Shantungensis, marked with color gradients from low to high suitability. Study areas are highlighted in red. Insets show detailed views of specific regions, including parts of North America, Europe and Asia. The map includes a compass rose and a scale bar for reference. Geographic coordinates are marked along the edges and major oceans and regions are labeled, such as the Arctic Sea, Pacific Northwest and Indian Ocean Eastern. The map provides a visual representation of potential habitats across different continents.
Bioclimatic variables and future projections
The present study employed 19 bioclimatic variables sourced from WorldClim (http://www.worldclim.org/, accessed on 20 March 2025) at a spatial resolution of 2.5 arc-minutes. These variables represent current climatic conditions, encompassing the maximum, minimum, and average mean values of precipitation and temperature recorded between 1970 and 2000. Furthermore, future climate projections were incorporated to evaluate the potential impact of climate change on species distributions. The future climate data were long-term climatological averages of three periods: 2041–2060, 2061–2080, and 2081–2100. These projections were obtained from the outputs of the BCC-CSM2-MR global climate model (GCM) under four Shared Socioeconomic Pathway (SSP) scenarios: SSP126, SSP245, SSP370, and SSP585. For this study, only one GCM was selected, BCC-CSM2-MR, as it is particularly suitable for the targeted study area. All combinations of the two SSPs (SSP126 and SSP585) were considered, yielding four climate change scenarios. ArcGIS software (version 10.8.2) was utilised to process the data required for the study area globally.
Variable selection
The ENM (Ecological Niche Modeling) tool was used to calculate Pearson correlation coefficients (r) between variables from 1970 to 2000 to eliminate multicollinearity bias (fig. 2A). If the absolute correlation exceeded 0.85, one of the two variables was removed.
(A) Pearson correlation analysis: The horizontal axis denotes the correlation coefficient values. (B) Principal bioclimatic factors.
Figure 2 Long description
(a) A Pearson correlation matrix with colored circles representing correlation coefficients between variables. The horizontal and vertical axes list variables labeled from Bio 1 to Bio 11. The color scale on the right indicates correlation strength, ranging from negative to positive values. (b) A bar chart titled 'Key Bioclimatic Factors' lists factors such as Annual Precipitation and Mean Temperature of Coldest Quarter. Each bar shows a percentage, with Annual Precipitation at 24.1 percent and Wettest Quarter at 3.2 percent. The bars are horizontally aligned with percentages on the right.
After the elimination process, only 10 environmental variables (bio1, bio4, bio5, bio6, bio7, bio8, bio11, bio12, bio13, and bio14) were used in modelling P. shantungensis (fig. 2B). Additionally, species distribution model (SDM) analyses were performed with all variables to assess their contribution percentages using the jackknife test. The factors of annual precipitation, mean temperature of the coldest quarter, precipitation of the driest month, and temperature seasonality were identified as the most significant influences on the distribution of P. shantungensis. The three resampling techniques available in MaxEnt software are bootstrapping, cross-validation, and subsampling. These techniques produce distinct model selection results or their suitability values. The subsampling method was selected for the modelling.
MaxEnt model parameter settings
Using current and future climate scenarios, the potential distribution of P. shantungensis worldwide was predicted using MaxEnt software (version 3.4.4, http://biodiversityinformatics.amnh.org/open-source/maxent/, accessed 18 March 2025). The MaxEnt model was developed utilising environmental variables and distribution data. The model was executed 10 times, employing a maximum of 10,000 background points and 5,000 iterations. Thirty per cent of the distribution data was allocated for model testing using occurrence records. The environmental factors influencing the distribution of P. shantungensis were identified through the jackknife test and response curves (fig. 3).
Jackknife analysis of the significance of environmental variables.
Figure 3 Long description
A bar graph titled 'Jackknife of AUC for P. shantungensis displays the impact of different environmental variables on AUC. The x-axis represents AUC values ranging from 0.75 to 1.00. The y-axis lists environmental variables: bio underscore 1, bio underscore 11, bio underscore 12, bio underscore 13, bio underscore 14, bio underscore 4, bio underscore 5, bio underscore 6, bio underscore 7 and bio underscore 8. Each variable is represented by three bars: 'Without variable' in green, 'With only variable' in blue and 'With all variables' in red. The graph shows varying AUC contributions of each variable, with bio underscore 8 having the highest AUC when used alone.
Classification of suitable habitats
SDM Toolbox 10.9 and ArcGIS 10.8.2 were utilised to analyse the changes in P. shantungensis distribution under several climate scenarios. The distribution results for P. shantungensis were categorised into four groups based on the output of the Equal Training Sensitivity and Specificity (ETSS) threshold: unsuitable environment (0–0.38), poorly suited habitat (0.38–0.69), moderately suitable habitat (0.69–0.87), and very suitable habitat (0.87–1) (table 2). Under SDM Toolbox 10.9, we assessed changes in P. shantungensis distribution by comparing current and future projections, classifying results into four categories: no occupancy, no change, range contraction, or range expansion.
Distribution change estimation
P. shantungensis was analysed using different methods to assess how climate change affects the potential distribution of the species. Potential habitats of species under various climate scenarios and habitat suitability types were determined with the zonal statistics tool in ArcGIS 10.8.2. The current suitable areas were compared with those considered suitable under different future scenarios, both with and without occupied habitats. Binary models and distribution maps were employed to classify habitats as either suitable or unsuitable. Distribution changes between current and future projections were analysed using binary SDM tools within the SDM toolbox.
Methodological approach for SDM
The methodological approach for P. shantungensis used in this study follows that described by Chan et al. (Reference Chan, Leow, Bea, Cheng, Phoong, Hong and Chen2022) and Liu et al. (Reference Liu, Zheng, Guan, Han and Liu2023).
Results
Evaluation of modelling performance and analysis of environmental variable contributions
The test AUC value derived from the optimised model is shown in fig. 4. An AUC of 0.984 for P. shantungensis indicates high model performance in predicting suitable habitat areas for this species. MaxEnt results identified the following bioclimatic variables as significantly influencing suitable habitat for P. shantungensis are bio12, bio11, bio14, bio1, bio4, bio13, bio5, bio6, bio7, and bio8, with respective contribution rates of 24.1%, 19.21%, 17.6%, 10.9%, 9.4%, 5.9%, 5.7%, 3.9%, 3.2%, and 0% (table 1). Notably, bio12, bio11, bio14, bio1, and bio4 have the greatest impact on the suitable area for P. shantungensis, accounting for 81.2% of the cumulative contribution. The permutation importance values for bio12, bio11, bio14, bio1, bio4, bio13, bio5, bio6, bio7, and bio8 are 0.0, 0.1, 10.7, 27.4, 12.3, 6.6, 25.4, 13.0, 4.4, and 0.1, respectively (table 1).
Presents the area under the curve (AUC) for distribution models of P. shantungensis under current climatic conditions.
Figure 4 Long description
A graph showing average sensitivity versus 1 minus specificity for P. shantungensis. The x-axis is labeled '1 - Specificity (Fractional Predicted Area)' and the y-axis is labeled 'Sensitivity (Fraction of Positive Sites)'. The curve represents the model's performance, with a mean AUC of 0.984, indicating high accuracy. The graph includes a line for random prediction. The curve is marked with mean plus one standard deviation and the legend identifies these elements.
Percentage contribution of environmental factors to the suitable habitat of P. shantungensis
Table 1 Long description
The table measures the percentage contribution of various environmental factors to the suitable habitat of P. Shantungensis. Annual Precipitation has the highest contribution at 24.1%, indicating its significant role in habitat suitability. Mean Temperature of Coldest Quarter also plays a major role with a 19.21% contribution. In contrast, Mean Temperature of Wettest Quarter has a negligible impact, contributing 0%. Temperature-related variables like Annual Mean Temperature and Temperature Seasonality contribute moderately, with 10.9% and 9.4% respectively. Precipitation of Driest Month is another key factor, contributing 17.6%. These data points suggest that precipitation factors are generally more influential than temperature factors in determining habitat suitability for P. Shantungensis.
Current habitat suitability and geographic distribution of P. shantungensis
Based on MaxEnt modelling results, habitat suitability for P. shantungensis was classified into four categories based on MaxEnt probability thresholds: unsuitable (0–0.38), marginal (0.38–0.69), moderate (0.69–0.87), and highly suitable (0.87–1.0). Marginally suitable habitats (0.38–0.69) were geographically restricted to North Africa and southeastern Australia, whereas highly suitable areas (0.87–1.0) showed extensive distribution across Asia, Russia, Europe, and parts of the Americas. This probabilistic classification approach enables quantitative assessment of species environment relationships and facilitates mapping of geographic distribution at multiple spatial scales under current climatic conditions. The total suitable area for P. shantungensis globally encompasses approximately 4,386.97 × 10⁴ km2 (100%), comprising highly suitable areas (625.33 × 10⁴ km2, 14.3%), moderately suitable areas (1,285.65 × 10⁴ km2, 29.3%), and marginally suitable areas (2,475.99 × 10⁴ km2, 56.4%) (fig. 5 and table 2).
Potentially suitable global distribution areas for P. shantungensis under current climatic projections.
Figure 5 Long description
A world map illustrating habitat suitability for a species during the current period. The map uses color coding to indicate different suitability levels: unsuitable areas are in gray, low suitable areas in yellow, moderately suitable areas in green and highly suitable areas in red. The map includes a compass rose indicating north, south, east and west. Various regions across continents are marked with these colors, showing the distribution of suitable habitats globally.
Based on the ETSS threshold results, four suitability levels of current distribution regions were defined for P. shantungensis
Table 2 Long description
The table categorizes regions based on their suitability for the distribution of P. Shantungensis, measured in square kilometers. Unsuitable regions cover the largest area at 4,662 square kilometers, followed by marginally suitable regions at 2,476 square kilometers. Moderately suitable regions account for 1,286 square kilometers, while highly suitable regions are the smallest at 625 square kilometers. This data suggests a significant portion of the area is unsuitable, with progressively smaller areas being more suitable. The suitability levels are defined by specific value ranges, indicating a gradation in habitat suitability. These findings can guide conservation efforts and habitat management for P. Shantungensis.
Potential suitability regions change for P. shantungensis under future climate scenarios
MaxEnt projections for three future periods (2041–2060, 2061–2080, and 2081–2100) under both SSP126 and SSP585 scenarios indicate consistent expansion of the suitable distribution area for P. shantungensis. Specifically, by 2041–2060, the potential distribution area of P. shantungensis is anticipated to increase by 8.83% (SSP126) and 9.04% (SSP585), respectively, compared to the current (8.60%) suitable area. For 2061–2080, projected increases are 9.04% (SSP126) and 8.62% (SSP585). By 2081–2100, the largest expansions are projected; 9.41% (SSP126) and 9.32% (SSP585) (fig. 7).
Future projections indicate continued global expansion of suitable habitat for P. shantungensis, with highly suitable regions exhibiting the highest growth rates. By 2081–2100 under SSP126, highly suitable areas are projected to reach 48,100 × 10⁴ km2, a 9.41% increase from current conditions (table 3).
The total suitable area of P. shantungensis under different climate scenarios
Table 3 Long description
The table measures the total suitable area for P. Shantungensis under various climate scenarios from the present to 2100. Currently, the suitable area is 43,900 square kilometers, representing 8.60% of the total area. Under future scenarios, the area increases, with the highest suitable area projected for 2081-2100 under SSP126 at 48,100 square kilometers, or 9.41%. Comparatively, SSP585 scenarios also show an increase, with 46,200 square kilometers (9.04%) in 2041-2060 and 47,600 square kilometers (9.32%) in 2081-2100. The data suggests a trend of increasing suitable area over time, with variations depending on the specific climate scenario. However, the projections should be interpreted with caution due to potential uncertainties in climate modeling.
The geographical distribution of P. shantungensis demonstrates considerable variability in response to alterations in potentially suitable regions under future climate scenarios. Projected changes in suitable habitat distribution vary considerably by continent, with Southeast Asia experiencing the most substantial expansion, followed by Northeast Asia, Northwest Europe, and limited areas in North Africa. Future projections indicate range expansion from Northwest Europe and North America extending into South America, representing new geographic regions at risk of P. shantungensis establishment (table 4; figs. 6 and 8).
The total suitable area of P. shantungensis increases under different climate scenarios compared to the total suitable area under the current climate scenario
Table 4 Long description
The table compares the increase in suitable area for P. Shantungensis under different climate scenarios from 2041 to 2100. SSP126 scenarios consistently show a substantial increase in suitable area, with percentages nearing 90% by 2080 and 2100. In contrast, SSP585 scenarios exhibit much smaller increases, with percentages just above 5% for 2041-2060 and 2061-2080, but rising to 89.1% by 2100. This suggests that SSP126 scenarios may be more favorable for the expansion of suitable habitats for P. Shantungensis over time. However, the dramatic increase in SSP585 by 2100 indicates potential variability or shifts in climate impacts.
Suitability areas by P. shantungensis globally in the future (SSP126).
Figure 6 Long description
Three world maps display the suitability index for SSP 126 over three time periods: 2041-2060, 2061-2080 and 2081-2100. Each map uses color coding to indicate areas as unsuitable, low suitable, moderately suitable and highly suitable. The maps show changes in suitability across continents, with notable variations in regions such as Southeast Asia, Northeast Asia and parts of Europe and North America. A compass rose is present in the top right corner of each map.
Suitability areas by P. shantungensis globally in the future (SSP585).
Figure 7 Long description
Three world maps display the suitability index for SSP 585 across different time periods: 2041 to 2060, 2061 to 2080 and 2081 to 2100. Each map uses color coding to indicate suitability levels: unsuitable, low suitable, moderately suitable and highly suitable. The maps show changes in suitability across various regions globally, with a compass rose indicating north, south, east and west directions. The maps highlight shifts in suitability over time, with notable changes in regions such as North America, Europe and Asia.
Area changes: current vs future climate distribution.
Figure 8 Long description
The bar graph displays habitat suitability across different climate scenarios from current to 2081-2100. The x-axis shows the scenarios: Current, SSP 126 (2041-2060, 2061-2080, 2081-2100) and SSP 585 (2041-2060, 2061-2080, 2081-2100). The y-axis represents area in square kilometers, ranging from 0 to 700,000. Bars are categorized as poorly suitable, moderately suitable and highly suitable. The three pie charts below detail the distribution of suitability for each category. The first pie chart shows poorly suitable areas with values: Current (649087), SSP 126 (2041-2060: 608351, 2061-2080: 652884, 2081-2100: 616357), SSP 585 (2041-2060: 650474, 2061-2080: 659474, 2081-2100: 616357). The second pie chart represents moderately suitable areas with values: Current (354340), SSP 126 (2041-2060: 315885, 2061-2080: 380091, 2081-2100: 332691), SSP 585 (2041-2060: 354756, 2061-2080: 380091, 2081-2100: 332370). The third pie chart shows highly suitable areas with values: Current (1605310), SSP 126 (2041-2060: 156364, 2061-2080: 186120, 2081-2100: 156590), SSP 585 (2041-2060: 141925, 2061-2080: 141925, 2081-2100: 141504).
Discussion
The accuracy of SDMs depends critically on high-quality occurrence data and appropriate software configuration. MaxEnt achieved an average AUC value exceeding 0.98, demonstrating high model performance and reliable predictions of both current and future distribution for this invasive planthopper. The findings for SSP126 and SSP585 suggest that the habitat of P. shantungensis will increase in all future periods from 2041 to 2100. This expansion is largely driven by global warming, which not only extends the insect’s life cycle but also increases its developmental success (DEFRA, 2025). Currently, P. shantungensis predominantly occupies humid subtropical (Cfa) climates, present in 99.9% of suitable areas in Zhejiang Province, China, and 20.2% of the Republic of Korea (Bourgoin et al., Reference Bourgoin, Gros and Stroiński2020). Additional suitable climates include humid subtropical with dry winters (Cwa; 38.5% of the Republic of Korea), humid continental with dry winters (Dwa; 26.3% of the Republic of Korea), and Mediterranean types (Csb, Csa) in parts of France and Turkey (Hizal et al., Reference Hizal, Oztemiz and Gjonov2019; Park and Jung, Reference Park and Jung2021 ). Habitat suitability is primarily determined by temperature variables, including maximum temperature of the warmest month (bio5), mean annual temperature (bio1), minimum temperature of the coldest month (bio11), and annual precipitation (bio14) (Baek et al., Reference Baek, Lee and Park2024). Developmental thresholds for this species (requiring temperatures between 12.0°C and 36.9°C) restrict its geographic range to temperate and subtropical zones, limiting establishment in arid, tropical, and subarctic regions.
Climate matching shows that southeastern North America and parts of South Australia and South America are very likely to be invaded, as is southern and eastern Europe. The establishment in Japan and South Korea is facilitated by climatic conditions similar to its native range, the nearness of these geographical areas, as well as human-facilitated travel via trade (Kim and Kil, Reference Kim and Kil2016). Mediterranean regions (Csa, Csb) in Turkey and parts of Europe show strong climate matching with P. shantungensis occurrence (Hizal et al., Reference Hizal, Oztemiz and Gjonov2019). These regions provide the hot, dry summers required for pupae development and the mild winters necessary for mature males to mate. The absence of P. shantungensis in Inner Mongolia, Xinjiang, and parts of Western Asia is due to unsuitable dry or cold conditions (Zhao et al., Reference Zhao, Wang, Song, Xue, Zhang, Peth and Horn2023; Qiao et al., Reference Qiao, Zhuang, Wang, Xie, Zhang, Shen and Zhou2025).
Climate change may shift the suitability of regions for P. shantungensis, with some marginal areas becoming favourable while current hotspots may become unsuitable due to rising temperatures (Zhang et al., Reference Zhang, Fang, Zhao, Jiang, Stroiński and Qin2022). Effective management requires integrating climate data with host plant distribution (Kim and Kil, Reference Kim and Kil2016; Song et al., Reference Song, Li, Gao, Yan, Zhang, Liu and Zhou2024; Dong et al., Reference Dong, Cai, Dong, Wang and An2025), assessing climate impacts on pest populations and evaluating biotic factors influencing establishment. Most at-risk countries remain unaware of the threat, despite the pest’s rapid spread via trade and its potential to become a major invasive species globally. Therefore, to prevent further expansion of P. shantungensis, the results of the MaxEnt model show that high-risk countries should establish intensive plant quarantine programs and control strategies (Yan et al., Reference Yan, Yang, Zhang, Li, Wang, Liu and Yu2025).
Several important factors not considered in this study may influence P. shantungensis distribution, including: (1) host plant responses to climate change, (2) natural enemies and competitive interactions, (3) human activity patterns and microclimate variation, (4) extreme weather events, and (5) adaptive evolution and phenotypic plasticity. Future research should integrate field monitoring, interaction studies, and high-resolution datasets to improve prediction accuracy and management strategies.
Conclusion
MaxEnt modelling projects that P. shantungensis will expand its suitable habitat range by 2100 under future climate scenarios. The species is projected to expand from its current Asian range into temperate regions of North America, Northwest Europe, South America, and Australia, with only Antarctica remaining climatically unsuitable. The primary bioclimatic drivers of global distribution are annual precipitation (bio12), mean temperature of the coldest quarter (bio11), and precipitation of the driest month (bio14), highlighting the importance of temperature and moisture availability for this invasive agricultural pest. Future research should integrate local-scale environmental variables, identify implementation barriers, and optimise management strategies to develop targeted frameworks in high-risk regions and enhance broader biosecurity objectives.
Author contributions
M.A.I.: Writing – original draft, software, methodology, formal analysis, conceptualisation. D.X. and Z.X.: Review and editing. Q.D.: Supervision, writing – review and editing, funding acquisition, data curation.
Financial support
This work was supported by the National Natural Science Foundation of China (grant numbers 32170475, 31672340, 31750002).
Competing interests
The authors declare no conflict of interest.
Data availability
Data will be made available on request.
