True Skill Statistic Research Articles

Comprehensive study of thunderstorm indices threshold favorable for thunderstorms during monsoon season using WRF–ARW model and ERA5 over India

IntroductionThe current research investigates into the application of various thunderstorm indices to predict severe thunderstorm occurrences during the monsoon season across four distinct regions in India. Methods: The study assesses the prediction model’s efficacy using various skill scores and the Weather Research and Forecasting (WRF) model has been integrated for 30 h with double moment microphysics scheme NSSL-17 which accurately reproduces vertical and meteorological measures.ObjectiveFurthermore, it investigates fifteen thunderstorm indices derived from the ERA5 dataset to identify the most effective index for forecasting severe thunderstorms.ResultsThe results indicate that combining thunderstorm indices with skill scores, such as the Heidke Skill Score and True Skill Statistic, enhances the accuracy of severe thunderstorm predictions in the Indian monsoon season. The accurate predictions rely on determining optimal thresholds for each index. The study emphasizes the importance of using multiple indices rather relying solely on single measure for predicting severe thunderstorms. Advanced indices like the Energy Helicity Index (EHI) and Supercell Composite Parameter (SCP) perform well in forecasting extreme severe thunderstormsdue to their strong reliance on wind shears. The EHI (> 1), and SCP (≥ 3.5), STP (≥ 1.2) along with low SRH at 3 km (100 m2/s2), indicated no evidence of helicity or tornado activity during the event. On the other hand, the CAPE, K Index, and VT Index demonstrate robust predictive capabilities for non-severe category thunderstorms.ConclusionsIntegrating numerous thunderstorm indices improves meteorologists’ forecasts, ensuring public safety.Based on this work, future research can improve severe weather forecasting models’ accuracy and reliability.

Interspecific niche overlap and climatic associations of native Quercus species in the Zagros forests of Iran

In the arid regions surrounding southwestern Asia, oak trees play an important role in the environment by conserving water and preserving soil and supporting wildlife. There are concerns for the health and continued productivity of these forests, and there is a widespread oak decline. In this study, ecological niche modeling (ENM) and multivariate principal component analysis (PCA) were used to determine the environmental variations among Quercus brantii, Q. infectoria, and Q. libani in the Zagros forests of western Iran. Based on the area under the curve (AUC > 0.9) and the true skill statistic (TSS > 0.8), three algorithms, including random forest (RF), maximum entropy (MaxEnt), and maximum picking neural net (MAXnet), were used in an ensemble approach to explore the most important environmental variables within the projected suitable habitat of each species. Environmental variables related to climate, including annual precipitation (bio12) and annual mean temperature (bio1), were among the most important drivers of habitat suitability for oak species. The results illustrated that suitable areas for Q. infectoria and Q. libani often occur in the northern parts of the Zagros, which are humid and cooler than the southern areas. The results suggested that Q. brantii has a higher tolerance for harsh conditions in Zagros and therefore inhabits almost all areas of Zagros, including the southern parts. The climatic overlap between pairs of species was found to be significant based on equivalency and similarity tests, suggesting that the environmental conditions of the two species compared are more similar than would be expected by chance. The gap analysis for the three oak species revealed that only a few protected areas currently exist within their suitable habitats, with most of the highly suitable habitats located on non-protected lands, particularly for Q. libani and Q. infectoria. In addition, deforestation is known to sometimes occur within the boundaries of protected areas. Improving the level of protection by upgrading the current protected areas to national parks can help with conservation.

Modelling spatial distributions of biogenic habitat‐forming taxa to inform marine spatial planning

Abstract Biogenic habitats are foundational habitats for species assemblages and drive a range of ecosystem functions. The Hauraki Gulf/Tiikapa Moana is the most intensively used coastal area in Aotearoa/New Zealand, and decades of commercial fishing, sedimentation and industrialization have degraded biogenic habitats in the Gulf. In response, the marine spatial plan ‘Sea Change’ includes proposals to create new and extend existing marine protected areas (MPAs) and restrict the area open to mobile bottom‐impact fishing methods to conserve and help the recovery of biogenic habitats. To assess the benefits of different spatial planning scenarios for biogenic habitats, information on their spatial distribution is needed, but data limitations are a significant challenge. Here, an approach is detailed that maximized the information extracted from limited species occurrence data, by incorporating expert knowledge to develop and evaluate models of biogenic habitat‐forming taxa. Ensemble habitat suitability models (using boosted regression tree and random forest models) were created for 20 biogenic habitat groups. Using withheld data for validation, area under curve (AUC) scores ranged from 0.58 to 0.95 and true skill statistics (TSS) ranged from 0.37 to 0.84, though three models were further evaluated as insufficient representations of known ecological habitats by expert assessors. Models produced here provide substantially increased information to inform progress on implementation of the Sea Change plan. Two stakeholder processes have been held, resulting in the development of spatial plans for bottom trawl mitigation that are currently under public consultation. Systematic surveys that gather abundance and high‐resolution environmental data should be a priority to improve the utility of predictive models and inform future management processes in the Hauraki Gulf, and funding has been allocated to support additional data collection to determine the effectiveness of spatial management measures in the Hauraki Gulf.

Image Synthesis for Solar Flare Prediction

Solar flare prediction is a topic of interest to many researchers owing to the potential of solar flares to affect various technological systems, both terrestrial and in orbit. In recent years, the forecasting task has become progressively more reliant on data-driven computations and machine-learning algorithms. Although these efforts have improved solar flare predictions, they still falter in doing so for large solar flares, in particular under operational conditions, since large-flare data are very scarce and labeled data are heavily imbalanced. In this work, we seek to address this fundamental issue and present a scheme for generating synthetic magnetograms to reduce the imbalance in the data. Our method consists of (1) synthetic oversampling of line-of-sight magnetograms using Gaussian mixture model representation, followed by (2) a global optimization technique to ensure consistency of both physical features and flare precursors, and (3) the mapping of the generated representations to realistic magnetogram images using deep generative models. We show that these synthetically generated data indeed improve the capacity of solar flare prediction models and that, when tested on such a state-of-the-art model, it significantly enhances its forecasting performance, achieving an F1-score as high as 0.43 ± 0.08 and a true skill statistic of 0.64 ± 0.10 for X-class flares in the 24 hr operational solar flare data split.

Prediction of potential suitable habitats in the 21st century and GAP analysis of priority conservation areas of Chionanthus retusus based on the MaxEnt and Marxan models.

Chionanthus retusus (C. retusus) has a high economic and medicinal value, but in recent years it has been included in the list of China's major protected plants and China's Red List of Biodiversity due to the serious destruction of its wild germplasm resources. Based on 131 sample points of C. retusus, this study simulated potential habitats and spatial changes of C. retusus in the 21st century using the Maxent model combined with the geographic information system ArcGIS, predicted prioritized protected areas by the Marxan model, and assessed current conservation status through GAP analysis. The results showed that (1) when the regularization multiplier was 1.5 and the feature combinations were linear, quadratic, and fragmented, the area under the curve of the subjects in the training and test sets were both above 0.9, the true skill statistic value was 0.80, and the maximum Kappa value was 0.62, meaning that the model had high accuracy; (2) Temperature seasonality, annual precipitation, min temperature for coldest month, and precipitation of wettest month had relatively strong influences on species' ranges. (3) The moderately and optimally suitable habitats of C. retusus were primly located in the areas of southwestern Shanxi, central Hebei, western Henan, Shandong, Shaanxi, Anhui and Hubei; (4) Under different future climate scenarios, the area of each class of suitable habitat will increase for varied amounts compared to the current period, with a general trend of expansion to the south; (5) The C. retusus priority protected areas were mainly located in most of Shandong, southern Liaoning, southwestern Shanxi, western Henan, and central Hebei, and its conservation vacancy area was relatively large compared to its protected area. These results will provide scientific strategies for implementing long-term conservation of C. retusus in China and similar regions under warming conditions in the 21st century.

PRELIMINARY MODEL OF POTENTIAL AREAS FOR RESTORATION FOR THE INTER-ANDEAN VALLEY OF CAUCA RIVER (COLOMBIA, SOUTH AMERICA) BASED ON HABITAT SUITABILITY MODELS

Tropical dry forests (TDF) are highly susceptible to land degradation. The inter-Andean Valley of the Cauca River (IVCR) has the most fragmented Colombian dry forests, and their restoration is essential. Here, potential areas for restoration were identified using a habitat suitability modeling (HSM) approach. TDF vascular plants and bioclimatic predictors were used. Species were selected based on threatened status, endemism, and priority level for conservation. Two sets of predictors were chosen using Variance Inflation Factor (VIF) and Principal Component Analysis (PCA). Then, with a maximum entropy algorithm, PCA and VIF models were projected for the selected species. These models were evaluated via true skill statistics (TSS) and area under the curve (AUC) statistical metrics. Models with good performance (TSS, AUC, standard deviation, variance) were ensembled, and a preliminary model where areas with suitable bioclimatic conditions for the selected species were generated. Results show that nearly 45 % of the IVCR has suitable conditions for the selected species. Although potential conflicts may arise in areas under permanent or semipermanent crops which represent more than 80 % of the IVCR, cropland mosaics, and natural and seminatural land covers might provide alternative solutions to reduce the land-use conflict. The potential areas for restoration identified in this study may provide a comprehensive framework for environmental impact and regional risk assessments related to the current land use and land cover change dynamics. Also, they may provide relevant information for designing landscape restoration programs as an adaptive strategy toward climate change.

Global spatial distribution of Prosopis juliflora -one of the world's worst 100 invasive alien species under changing climate using multiple machine learning models.

Climate change is one of the factors contributing to the spread of invasive alien species. As a result, it is critical to investigate potential invasion dynamics on a global scale in the face of climate change. We used updated occurrence data, bioclimatic variables, and Köppen-Geiger climatic zones to better understand the climatic niche dynamics of Prosopis juliflora L. (Fabaceae). In this study, we first compared several algorithms-MaxEnt, generalized linear model (GLM), artificial neural network (ANN), generalized boosted model (GBM), generalized additive model (GAM), and random forest (RF)-to investigate the relationships between species-environment and climate for mesquite. We identified the global climate niche similarity sites (NSSs) using the coalesce approach. This study focused on the current and future climatic suitability of P. juliflora under two global circulation models (GCMs) and two climatic scenarios, i.e., Representative Concentration Pathways (RCPs), 4.5 and 8.5, for 2050 and 2070, respectively. Sensitivity, specificity, true skill statistic (TSS), kappa coefficient, and correlation were used to evaluate model performance. Among the tested models, the machine learning algorithm random forest (RF) demonstrated the highest accuracy. The vast swaths of currently uninvaded land on multiple continents are ideal habitats for invasion. Approximately 9.65% of the area is highly suitable for the establishment of P. juliflora. Consequently, certain regions in the Americas, Africa, Asia, Europe, and Oceania have become particularly vulnerable to invasion. In relation to RCPs, we identified suitable area changes (expansion, loss, and stability). The findings of this study show that NSSs and RCPs increase the risk of invasion in specific parts of the world. Our findings contribute to a cross-border continental conservation effort to combat P. juliflora expansion into new potential invasion areas.

A New Algorithm Using Support Vector Machines to Detect and Monitor Bloom-Forming Pseudo-nitzschia from OLCI Data

Pseudo-nitzschia spp. blooms are a recurrent problem in many coastal areas globally, imposing some significant threats to the health of humans, ecosystems and the economy. Monitoring programmes have been established, where feasible, to mitigate the impacts caused by Pseudo-nitzschia spp. and other harmful algae blooms. The detection of such blooms from satellite data could really provide timely information on emerging risks but the development of taxa-specific algorithms from available multispectral data is still challenged by coupled optical properties with other taxa and water constituents, availability of ground data and generalisation capabilities of algorithms. Here, we developed a new set of algorithms (PNOI) for the detection and monitoring of Pseudo-nitzschia spp. blooms over the Galician coast (NW Iberian Peninsula) from Sentinel-3 OLCI reflectances using a support vector machine (SVM). Our algorithm was trained and tested with reflectance data from 260 OLCI images and 4607 Pseudo-nitzschia spp. match up data points, of which 2171 were of high quality. The performance of the no bloom/bloom model in the independent test set was robust, showing values of 0.80, 0.72 and 0.79 for the area under the curve (AUC), sensitivity and specificity, respectively. Similar results were obtained by our below detection limit/presence model. We also present different model thresholds based on optimisation of true skill statistic (TSS) and F1-score. PNOI outperforms linear models, while its relationship with in situ chlorophyll-a concentrations is weak, demonstrating a poor correlation with the phytoplankton abundance. We showcase the importance of the PNOI algorithm and OLCI sensor for monitoring the bloom evolution between the weekly ground sampling and during periods of ground data absence, such as due to COVID-19.

Response of plant species to impact of climate change in Hugumbrda Grat-Kahsu forest, Tigray, Ethiopia: Implications for domestication and climate change mitigation

This study aimed to predict distribution and Total Carbon Stock (TCS) dynamics of Acacia abyssinica, Carissa edulis, and Juniperus procera in the Hugumbrda Grat-Kahsu National Forest in current (1970–2000) and future climate scenarios (2021–2100). Bioclimatic, soil, and elevation data were used for modeling using Maxent, with model accuracy evaluated using Area Under the Curve (AUC), Kappa test and True Skill Statistic (TSS). Significant differences were observed in distribution of species between current and future periods under Shared Socioeconomic Pathways (SSPs) of SSP2-4.5 and SSP5-8.5 scenarios. The main contributing predictors of the species distribution were temperature seasonality, altitude, and precipitation of the warmest quarter. All species were projected to shift to higher altitudes in the future. Acacia abyssinica’s current potential distribution (42.9 %) could expand to 77.1–99.2 % (SSP2-4.5) and 63.8–72.9 % (SSP5-8.5). Carissa edulis could extend from 54.2 % to 89.5–100 % (SSP2-4.5) and 77.1–87.9 % (SSP5-8.5). Juniperus procera’s might increase from 63.8 % to 91.8-99.7 % (SSP2-4.5) and 78–88.1 % (SSP5-8.5). The projected future climate is expected to result in an expansion of new suitable areas for all three species. The TCS estimates per km2 were 169 (Acacia abyssinica), 46 (Carissa edulis), and 1381 ton (Juniperus procera). In SSP2-4.5, Acacia abyssinica’s TCS could rise from 25,688 to 59,319 tons, Carissa edulis from 8,832 to 16,284 tons, and Juniperus procera from 312,106 to 487,493 tons. In SSP5-8.5, projections indicated 43,602 tons (Acacia abyssinica), 14,306 tons (Carissa edulis), and 430,872 tons (Juniperus procera). The study concludes by recommending the strategic planting of these species in both current and future suitable areas to enhance ecosystem services and ensure their sustained existence in the face of changing climates.

Assessment of resampling methods on performance of landslide susceptibility predictions using machine learning in Kendari City, Indonesia

Abstract Landslide susceptibility projections that rely on independent models produce biased results. This situation will worsen class balance if working with a small population. This study proposes a landslide susceptibility prediction model based on resampling, cross-validation, bootstrap, and random subsampling approaches, which is integrated with the machine learning model, generalized linear model, support vector machine, random forest, boosted regression trees, classification and regression tree, multivariate adaptive regression splines, mixture discriminate analysis, flexible discriminant analysis, maximum entropy, and maximum likelihood. This methodology was applied in Kendari City, an urban area which faced destructive erosion. Area under the ROC curve (AUC), true skill statistics (TSS), correlation coefficient (COR), normalized mutual information (NMI), and correct classification rate (CCR) were used to evaluate the predictive accuracy of the proposed model. The results show that the resampling algorithm improves the performance of the standalone model. Results also revealed that standalone models had better performance with the CV algorithm compared to the Bt and RS algorithms. The Bt-RF model excels in statistical measures (AUC = 0.97, TSS = 0.97, COR = 0.99, NMI = 0.50, and CCR = 0.93). Given the admirable performance of the proposed models in a moderate scale area, promising results can be expected from these models for other regions.

Uso de ferramenta geoespacial para predição da distribuição de Melanoides tuberculata (Müller, 1774) e alguns gastrópodes límnicos nativos encontrados no Brasil

Abstract: One of the most common environmental impacts that has caused irreversible effects on ecosystems is the introduction of exotic species. In addition to the ecological disasters that can be caused, such as the decline and extinction of native species, this phenomenon can generate significant economic losses and the possibility of the spread of parasites that are transmitted by these exotic species. These processes can be accentuated by climate change, which over time alters the global distribution pattern of the affected species, generating cases of niche overlap with species that share similar niches. In this sense, the knowledge about the areas with possible occurrence these species can direct the monitoring and control measures. In this study, we developed current and future prediction models to identify areas of suitability in the Neotropics for Melanoides tuberculata (Müller, 1774) and nine species of mollusks native to the American continent using the ecological niche modeling tool. In addition, we evaluated the ecological niche overlap between the invasive species and the native species of freshwater mollusks to verify whether the effects of climate change would alter the distribution of these organisms. The following methodological procedures were adopted to prepare the forecasting models: records of occurrence of the mollusks in different databases and search of environmental data for climate conditions in current and future scenarios in WorldClim 2.0 (SSP2-4.5 and SSP2-8.5). Besides, modeling procedures using seven packages of R software, evaluation of the models using the true skill statistic (TSS) metric, construction of maps and quantification and overlapping of ecological niche of the species included in the analysis. The results indicated that several areas of the Neotropics are suitable for the occurrence of M. tuberculata in the current scenarios. Moreover, the suitable areas for its occurrence will probably be expanded in both future scenarios. For native species, there were significant differences in relation to the areas of suitability, with a reduction for some species. Niche similarity tests indicated significant overlap only between M. tuberculata and the planorbid Biomphalaria straminea (Dunker, 1848). We discuss that the expansion of M. tuberculata can have negative consequences, including the reduction of native gastropod species and the spread of trematodes of medical and veterinary importance that this mollusk can transmit.

Spatiotemporal analysis of seasonal trends in land surface temperature within the distribution range of Moringa peregrina (Forssk.) in Southern and Southeastern Iran.

Temperature fluctuations and related factors are among the main causes of climate change. Understanding the temporal and spatial variations in temperature can shed light on how species respond to climate change. Plants generally persist in suitable microclimates in response to environmental change; however, examining long-term temperature variations within a species' range can be challenging using field observations. Thermal remote sensing, on the other hand, provides multi-scale time-series data with good coverage and regularity to overcome the challenges associated with field observations in environmental monitoring. Although changes in land surface temperature (LST) affect climate, hydrological processes, land-atmosphere interactions, and ecological activities, this metric has not received much research attention. This study aimed to analyze changes in habitat suitability and microclimatic conditions for Moringa peregrina. Seasonal changes in LST within the distribution range of the species were also investigated. To this aim, mean seasonal LST was computed in Google Earth Engine using the daily MODIS/006/MYD13A2 product from 2003 to 2023. Subsequently, a binary habitat suitability map was created based on the true skill statistic (TSS). The Mann-Kendall test was used to analyze seasonal LST trends. Major trends in LST were quantified based on the z-score, and compatibility with habitat suitability was evaluated via GAP analysis and the Kappa index. Seasonal temperature trends were evaluated by comparing each season with the following season using binary comparison. Landforms at presence points were regarded as microclimates and the sensitivity of microclimates to LST was evaluated using two methods: Principal component analysis (PCA) was used to quantify seasonal LST heterogeneity and the random forest (RF) approach was used to evaluate the effect of environmental parameters on habitat suitability within microclimates. The Kappa index revealed a weak overlapping between suitable / unsuitable habitat and the surfaces affected by the trend of changes. Moreover, the suitable habitat of Moringa peregrina in spring, autumn and winter is spatially overlapped by areas that have shown an increasing LST trend, and the presence points have not experienced an increasing temperature trend only in the summer. The findings show that the analysis of seasonal trends in LST provides insights into the effect of LST on habitat suitability and the condition of vegetation. The current study clearly shows that seasonal changes have had a significant impact on the distribution and habitat suitability of M. peregrina, particularly during summer and winter. Improved habitat suitability and range expansion were observed throughout the year. The study also highlights the role of landforms in regulating temperature. Landforms such as local ridges with minimal temperature fluctuations and regions near the Oman Sea were identified as potential future habitats due to favorable humidity conditions.

Identifying risk zones and landscape features that affect common leopard depredation.

Human-wildlife conflict (HWC) is a pressing issue worldwide but varies by species over time and place. One of the most prevalent forms of HWC in the mid-hills of Nepal is human-common-leopard conflict (HLC). Leopard attacks, especially in forested areas, can severely impact villagers and their livestock. Information on HLC in the Gorkha district was scarce, thus making it an ideal location to identify high-risk zones and landscape variables associated with such events. Registered cases were collected and reviewed from the Division Forest Office (DFO) during 2019-2021. Claims from DFO records were confirmed with herders and villagers via eight focus group discussions. To enhance modeling success, researchers identified a total of 163 leopard attack locations on livestock, ensuring a minimum distance of at least 100 meters between locations. Using maximum entropy (MaxEnt) and considering 13 environmental variables, we mapped common leopard attack risk zones. True Skill Statistics (TSS) and area under receiver-operator curve (AUC) were used to evaluate and validate the Output. Furthermore, 10 replications, 1,000 maximum iterations, and 1000 background points were employed during modeling. The average AUC value for the model, which was 0.726 ± 0.021, revealed good accuracy. The model performed well, as indicated by a TSS value of 0.61 ± 0.03. Of the total research area (27.92 km2), about 74% was designated as a low-risk area, 19% as a medium-risk area, and 7% as a high-risk area. Of the 13 environmental variables, distance to water (25.2%) was the most significant predictor of risk, followed by distance to road (16.2%) and elevation (10.7%). According to response curves, the risk of common leopard is highest in the areas between 1.5 to 2 km distances from the water sources, followed by the closest distance from a road and an elevation of 700 to 800 m. Results suggest that managers and local governments should employ intervention strategies immediately to safeguard rural livelihoods in high-risk areas. Improvements include better design of livestock corrals, insurance, and total compensation of livestock losses. Settlements near roads and water sources should improve the design and construction of pens and cages to prevent livestock loss. More studies on the characteristics of victims are suggested to enhance understanding of common leopard attacks, in addition to landscape variables. Such information can be helpful in formulating the best management practices.

Predicting Solar Proton Events of Solar Cycles 22–24 Using GOES Proton and Soft-X-Ray Flux Features

Solar energetic particle (SEP) events and their major subclass, solar proton events (SPEs), can have unfavorable consequences on numerous aspects of life and technology, making them one of the most harmful effects of solar activity. Garnering knowledge preceding such events by studying operational data flows is essential for their forecasting. Considering only solar cycle (SC) 24 in our previous study, we found that it may be sufficient to only utilize proton and soft X-ray (SXR) parameters for SPE forecasts. Here, we report a catalog recording ≥10 MeV ≥10 particle flux unit SPEs with their properties, spanning SCs 22–24, using NOAA’s Geostationary Operational Environmental Satellite flux data. We report an additional catalog of daily proton and SXR flux statistics for this period, employing it to test the application of machine learning (ML) on the prediction of SPEs using a support vector machine (SVM) and extreme gradient boosting (XGBoost). We explore the effects of training models with data from one and two SCs, evaluating how transferable a model might be across different time periods. XGBoost proved to be more accurate than SVMs for almost every test considered, while also outperforming operational SWPC NOAA predictions and a persistence forecast. Interestingly, training done with SC 24 produces weaker true skill statistic and Heidke skill scores2, even when paired with SC 22 or SC 23, indicating transferability issues. This work contributes toward validating forecasts using long-spanning data—an understudied area in SEP research that should be considered to verify the cross cycle robustness of ML-driven forecasts.

Effect of environmental and anthropogenic factors on the distribution and co-occurrence of cold-water corals

Cold-water corals (CWCs) are bioengineering species that can increase habitat heterogeneity and improve the deep sea’s biological diversity and ecosystem functioning. Knowledge of their distribution provides a critical baseline for assessing the effect of natural and anthropogenic impacts on these important deep-sea habitats. The aims of this study are: i) provide new data on the spatial distribution of six CWCs species in the Strait of Sicily, ii) describe the principal environmental and anthropogenic variables that play a role in shaping their distribution, iii) identify hotspots in which individuals belonging to the various species co-occur. Presence-only data of six CWCs species, ten environmental variables (depth, slope, rugosity, aspect, flowdir, temperature, salinity, north bottom current, east bottom current, chlorophyll-a), and one variable relating to bottom trawling effort (Automatic Information System – AIS) were used to predict the suitable habitats. We used Maximum Entropy modelling (MaxEnt) approach and used the AUC (area under the receiver operating characteristic curve) and TSS (true skill statistics) to evaluate the model performance. The results showed excellent AUC, TSS and AUC’s standard deviation mean values for all six species. The validation show high predictive performance. MaxEnt identified slope, depth, and rugosity as the most important predictors, showing the highest percentage contribution for all six species considered. Throughout the study area, highlyinterspecific persistent density hotspot of CWCs co-occurrence were discovered, with a total extension of 4.05 km2 where all species co-occur. Although studies on the effect of environmental and anthropogenic factors that impact the distribution of these species of conservation interest remain scarce, the results of this study offer useful guidance for decision-makers to develop necessary conservation measures.

Background selection complexity influences Maxent predictive performance in freshwater systems

Absence data are often lacking for species distribution modeling (SDM) purposes. This necessitates selecting background or pseudo-absence observations that influence SDM performance. Little is understood about how background selection affects SDM prediction in lotic systems. Here we test six background selection methods that implement different combinations of three selection filters concerning 1) sampling biases in species occurrence data, 2) geographic restriction to regions accessible to the species modeled, and 3) species occurrence relative to stream size, a key habitat factor. These six methods are used with Maxent to develop binary presence-absence predictions of 71 freshwater mussel distributions in the Midwestern United States. Prediction accuracy was evaluated with a separate validation presence-absence dataset derived from intensive surveys. Pairwise comparisons of background selection methods across species recorded in the validation dataset revealed significant differences relative to the Area Under Curve (AUC), the similarity between the prediction and observation, and the True Skill Statistic (TSS) metrics. The prediction specificity for those species absent in the validation dataset was also significantly affected by the background selection method. Implementing the sampling bias filter increased prediction similarity with validation data, AUC and TSS for species with validation presences, as well as prediction specificity for species without validation presences. Our results provide much needed insight into how background selection influences presence-background SDM performance in lotic systems. These findings can guide how to leverage available data and biological understanding to produce accurate SDM predictions that prioritize research objectives and goals regardless of study system or habitat.

Assessing Rice Sheath Blight Disease Habitat Suitability at a Regional Scale through Multisource Data Analysis

Extensive occurrence of rice sheath blight has been observed in China in recent years due to agricultural practices and climatic conditions, posing a serious threat to rice production. Assessing habitat suitability for rice sheath blight at a regional scale can provide important information for disease forecasting. In this context, the present study aims to propose a regional-scale habitat suitability evaluation method for rice sheath blight in Yangzhou city using multisource data, including remote sensing data, meteorological data, and disease survey data. By combining the epidemiological characteristics of the crop disease and the Relief-F algorithm, some habitat variables from key stages were selected. The maximum entropy (Maxent) and logistic regression models were adopted and compared in constructing the disease habitat suitability assessment model. The results from the Relief-F algorithm showed that some remote sensing variables in specific temporal phases are particularly crucial for evaluating disease habitat suitability, including the MODIS products of LAI (4–20 August), FPAR (9–25 June), NDVI (12–20 August), and LST (11–27 July). Based on these remote sensing variables and meteorological features, the Maxent model yielded better accuracy than the logistic regression model, with an area under the curve (AUC) value of 0.90, overall accuracy (OA) of 0.75, and a true skill statistics (TSS) value of 0.76. Indeed, the results of the habitat suitability assessment models were consistent with the actual distribution of the disease in the study area, suggesting promising predictive capability. Therefore, it is feasible to utilize remotely sensed and meteorological variables for assessing disease habitat suitability at a regional scale. The proposed method is expected to facilitate prevention and control practices for rice sheath blight disease.

Early diagnosis of candidemia with explainable machine learning on automatically extracted laboratory and microbiological data: results of the AUTO-CAND project

Background Candidemia is associated with a heavy burden of morbidity and mortality in hospitalized patients. The availability of blood culture results could require up to 48–72 h after blood draw; thus, early treatment decisions are made in the absence of a definite diagnosis. Methods In this retrospective study, we assessed the performance of different supervised machine learning algorithms for the early differential diagnosis of candidemia and bacteremia in adult patients on a large dataset automatically extracted within the AUTO-CAND project. Results Overall, 12,483 episodes of candidemia (1275; 10%) or bacteremia (11,208; 90%) were included in the analysis. A random forest classifier achieved the best diagnostic performance for candidemia, with sensitivity 0.98 and specificity 0.65 on the training set (true skill statistic [TSS] = 0.63) and sensitivity 0.74 and specificity 0.57 on the test set (TSS = 0.31). Then, the random classifier was trained in the subgroup of patients with available serum β-D-glucan (BDG) and procalcitonin (PCT) values by exploiting the feature ranking learned in the entire dataset. Although no statistically significant differences were observed from the performance measures obtained by employing BDG and PCT alone, the performance measures of the classifier that included the features selected in the entire dataset, plus BDG and PCT, were the highest in most cases. Conclusions Random forest classifiers trained on large datasets of automatically extracted data have the potential to improve current diagnostic algorithms for candidemia. However, further development through implementation of automatically extracted clinical features may be necessary to achieve crucial improvements.

An Evaluation of Suitable Habitats for Amur Tigers (Panthera tigris altaica) in Northeastern China Based on the Random Forest Model.

Amur tigers are at the top of the food chain and play an important role in maintaining the health of forest ecosystems. Scientific and detailed assessment of the habitat quality of Amur tigers in China is the key to maintaining the forest ecosystem and also addressing the urgent need to protect and restore the wild population of Amur tigers in China. This study used the random forest method to predict the potential habitat of Amur tigers in Heilongjiang and Jilin provinces using animal occurrence sites and a variety of environmental variables. Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. The results showed that the AUC value of the test set was 0.955. The true skill statistic (TSS) value is 0.5924, indicating that the model had good prediction accuracy. Using the optimal threshold determined by the Youden index as the cutoff value, we found that the suitable habitat for Amur tigers in the field was approximately 107,600 km2, accounting for 16.3% of the total study areas. It was mainly distributed in the Sino-Russian border areas in the south of the Laoyeling Mountains at the junction of Jilin and Heilongjiang provinces, the Sino-Russian border areas of Hulin-Raohe in the eastern part of the Wanda Mountains, and the Lesser Khingan Mountain forest region. The habitat suitability of the Greater Khingan Mountain and the plain areas connecting Harbin and Changchun was relatively low. Prey potential richness was the most critical factor driving the distribution of Amur tigers. Compared with their prey, the potential habitats for Amur tigers in Heilongjiang and Jilin provinces were small in total areas, sporadically distributed, and had low continuity and a lack of connectivity between patches. This indicates that some factors may restrict the diffusion of the Amur tiger, whereas the diffusion of ungulates is less restricted. The Amur tigers in this area face a serious threat of habitat fragmentation, suggesting that habitat protection, restoration, and ecological corridor construction should be strengthened to increase population dispersal and exchange. We provide a reference for future population conservation, habitat restoration, construction of ecological migration corridors, and population exchange of Amur tigers.

Modeling post-fire regeneration patterns under different restoration scenarios to improve forest recovery in degraded ecosystems

BackgroundChanges in disturbance regimes triggered by land use and climate change can significantly alter forest ecosystems by modifying the distribution of some species and hindering post-disturbance tree regeneration dynamics. Applied nucleation (AN) could be a valuable active restoration approach for promoting natural recovery in forest ecosystems affected by stand-replacing disturbances as it improves seed availability and microsite conditions. ObjectivesThe study aimed to investigate the potential of AN under different scenarios in a mountain forest ecosystem of the Northwestern Italian Alps dominated by Scots pine (Pinus sylvestris L.). The area was affected by a large stand-replacing fire in 2005 and post-fire salvage logging that amplified ecosystem degradation and dampened natural tree regeneration. MethodsWe assessed the main drivers guiding natural post-fire natural recovery and identified suitable sites for tree regeneration through a machine learning correlative model (Bayesian Additive Regression Tree, BART). Specifically, we used several environmental predictors (e.g., topography, wind direction, and distance from seed trees) to model the occurrence of natural tree regeneration. We predicted the probability of tree regeneration presence at landscape scale under the current situation (fire followed by salvage logging) and a set of AN scenarios characterized by an increase in nuclei density, since distance from seed trees emerged as the most important driver for natural tree regeneration. Starting from the situation 16 years after the fire, we reclassified the prediction raster into a binary map of intervention priority (priority and non-priority patches), using the probability value that maximized the model accuracy (true skill statistic; TSS) as threshold. Patches with scarce pine regeneration were considered as high intervention priority sites for AN. These predictions made it possible to assess the most efficient active management scenario in terms of promoting forest recovery. ConclusionsThe simulations showed the positive effects of AN on natural tree regeneration and the importance of site selection for plantations, proving that AN could be a promising post-fire management technique that can minimize human interventions and their associated economic and ecological costs. To our knowledge, this work is the first AN simulation in a temperate mountain ecosystem. The selection of favorable sites can be further improved by considering fine-scale characteristics through field experiments and cross-scale integration.