Non-Pessimistic Predictions of the Distributions and Suitability of Metasequoia glyptostroboides under Climate Change Using a Random Forest Model

Distribution and protection of Thesium chinense Turcz. under climate and land use change

Wild medicinal plants are prominent in the field of Traditional Chinese Medicine (TCM), but their availability is being impacted by human activities and ecological degradation in China. To ensure sustainable use of these resources, it is crucial to scientifically plan areas for wild plant cultivation. Thesium chinense, a known plant antibiotic, has been overharvested in recent years, resulting in a sharp reduction in its wild resources. In this study, we employed three atmospheric circulation models and four socio-economic approaches (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5) to investigate the primary environmental factors influencing the distribution of T. chinense. We also examined changes in its suitable area using the Biomod2 package. Additionally, we utilized the PLUS model to project and analyze future land use changes in climate-stable regions for T. chinense. Our planning for wild tending areas of T. chinense was facilitated by the ZONATION software. Over the next century, the climate-stable regions for T. chinense in China is approximately 383.05 × 104 km2, while the natural habitat in this region will progressively decline. Under the current climate conditions, about 65.06% of the habitats in the high suitable areas of T. chinense are not affected by future land use changes in China. Through hotspot analysis, we identified 17 hotspot cities as ideal areas for the wild tending of T. chinense, including 6 core hotspot cities, 6 sub-hotspot cities, and 5 fringe hotspot cities. These findings contribute to a comprehensive research framework for the cultivation planning of T. chinense and other medicinal plants.

Utilising random forests in the modelling of Eragrostis curvula presence and absence in an Australian grassland system

Eragrostis curvula is an agronomically and ecologically undesirable perennial tussock grass dispersed across Australia. The objective of this study is to investigate relationships of ecologically relevant abiotic variables with the presence of E. curvula at a landscape scale in the Snowy Monaro region, Australia. Through vegetation surveys across 21 privately owned properties and freely available ancillary data on E. curvula presence, we used seven predictor variables, including Sentinel 2 NDVI reflectance, topography, distance from roads and watercourses and climate, to predict the presence or absence of E. curvula across its invaded range using a random forest (RF) algorithm. Assessment of performance metrics resulted in a pseudo-R squared of 0.96, a kappa of 0.97 and an R squared for out-of-bag samples of 0.67. Temperature had the largest influence on the model's performance, followed by linear features such as highways and rivers. Highways' high importance in the model may indicate that the presence or absence of E. curvula is related to the density of human transit, thus as a vector of E. curvula propagule dispersal. Further, humans' tendency to reside adjacent to rivers may indicate that E. curvula's presence or absence is related to human density and E. curvula's potential to spread via water courses.

Invasive Trends of Spartina alterniflora in the Southeastern Coast of China and Potential Distributional Impacts on Mangrove Forests

Mangrove forests are one of the most productive and seriously threatened ecosystems in the world. The widespread invasion of Spartina alterniflora has seriously imperiled the security of mangroves as well as coastal mudflat ecosystems. Based on a model evaluation index, we selected RF, GBM, and GLM as a predictive model for building a high-precision ensemble model. We used the species occurrence records combined with bioclimate, sea-land topography, and marine environmental factors to predict the potentially suitable habitats of mangrove forests and the potentially suitable invasive habitats of S. alterniflora in the southeastern coast of China. We then applied the invasion risk index (IRI) to assess the risk that S. alterniflora would invade mangrove forests. The results show that the suitable habitats for mangrove forests are mainly distributed along the coastal provinces of Guangdong, Hainan, and the eastern coast of Guangxi. The suitable invasive habitats for S. alterniflora are mainly distributed along the coast of Zhejiang, Fujian, and relatively less in the southern provinces. The high-risk areas for S. alterniflora invasion of mangrove forests are concentrated in Zhejiang and Fujian. Bioclimate variables are the most important variables affecting the survival and distribution of mangrove forests and S. alterniflora. Among them, temperature is the most important environmental variable determining the large-scale distribution of mangrove forests. Meanwhile, S. alterniflora is more sensitive to precipitation than temperature. Our results can provide scientific insights and references for mangrove forest conservation and control of S. alterniflora.

Prediction of elevated groundwater fluoride across India using multi-model approach: insights on the influence of geologic and environmental factors

Elevated fluoride in groundwater is a severe problem in India due to its extensive occurrence and detrimental health impacts on the large population that thrives on groundwater. Although fluoride is primarily a geogenic pollutant, existing model-based studies lack the amalgamation of the influence of geologic factors, specifically tectonics, for identifying groundwater fluoride distribution. This drawback encourages the present study to investigate the association of the tectonic framework with fluoride in a multi-model approach. We have applied three machine learning models (random forest, boosted regression tree, and logistic regression) to predict elevated groundwater fluoride based on fluoride measurements across India. The random forest model outperformed other models with an accuracy of 93%. Tectonics was found to be one of the most important predictors alongside "depth to water table." Two major areas of high risk identified were the northwest parts and the south-southeast cratonic peninsular region. The random forest model also performed significantly well over the validation dataset. We estimate that nearly 257 million people are exposed to elevated fluoride risk in India. We endeavor that the findings of our study would be an effective tool for identifying the areas at risk of elevated fluoride and also assist in undertaking effective groundwater management strategies.

Predicting the invasive trend of exotic plants in China based on the ensemble model under climate change: A case for three invasive plants of Asteraceae

Ageratina adenophora, Eupatorium odoratum, and Mikania micrantha are three highly destructive invasive plants of Compositae in China. Through the screening of SDMs, random forest (RF), gradient boosting model (GBM), artificial neural network (ANN), and flexible discriminant analysis (FDA) with TSS greater than 0.8 are selected to construct a high-precision ensemble model (EM) as the prediction model. We use specimen sites and environmental variables containing climate, soil, terrain, and human activities to simulate and predict the invasion trend of three invasive weeds in China in current, the 2050s, and the 2070s. Results indicate that the highly invasive risk area of three exotic plants is mostly distributed along the river in the provinces south of 30° N. In the future scenario, the three exotic plants obviously invade northwards Yunnan, Sichuan, Guizhou, Jiangxi and Fujian. Climate is the most important variable that affects the spread of three kinds of alien plant invasions. Temperature and precipitation variables have a similar effect on A. adenophora and E. odoratum, while M. micrantha is more sensitive to temperature. It has been reported that Ipomoea batatas and Vitex negundo can prevent the invasion of three invasive plants. Hence, we also simulate the suitable planting areas for I. batatas and V. negundo. The results show that I. batatas and V. negundo are suitable to be planted in the areas where the three weeds show invasion tendency. In the paper, predicting invasion trends of exotic plants and simulating the planting suitability of crops that can block invasion, to provide a practical significance reference and suggestion for the management, prevention, and control of the invasion of exotic plants in China.