An Ecological Paradox: The African Wild Dog (Lycaon Pictus) Is Not Attracted to Water Points When Water Is Scarce in Hwange National Park,Zimbabwe

In dry biomes, spatio-temporal variation in surface water resource stocks is pervasive, with unknown effects on the ranging behaviour of large predators. This study assessed the effect of spatial variation in surface water resources on the ranging behaviour of the African wild dog (Lycaon pictus). We analyzed data for 1992 (dry year with 20 water points) and 2000 (wet year with 30 water points) against presence-only data for five packs of L. pictus in a part of Hwange National Park and adjacent smallholder communal farming areas in western Zimbabwe. Modelling the potential habitat for L. pictus using Maxent with distance from water points (D_w) and Normalized Difference Vegetation Index (NDVI) as predictor variables was successful for 2000 (AUC = 0.793) but not successful for 1992 (AUC = 0.423), with L. pictus probability of occurrence near water points being more for year 2000 than for year 1992. The predicted L. pictus range was wider in 1992 (~13888.1 km²) than in 2000 (~958.4 km²) (Test of Proportions, χ² = 124.52, df = 1, P = 0.00). Using the 2^nd order Multitype Nearest Neighbour Distance Function (Gcross), we also observed significant attraction between L. pictus and water points within only ~1km radius for 1992 but up to ~8km radius for 2000. Our study reinforced the notion that surface water resources attract wild dogs in the savannahs but paradoxically less so when water resources are scarce. In particular, our study furthers current understanding of the effects of changing water availability regimes on the endangered L. pictus, providing evidence that the endangered predator’s home range encroaches into potential ecological traps (i.e., smallholder communal farming areas) when water resources are scarce.     

A Winter Distribution Model for Bicknell’s Thrush (Catharus bicknelli), a Conservation Tool for a Threatened Migratory Songbird

Conservation planning and implementation require identifying pertinent habitats and locations where protection and management may improve viability of targeted species. The winter range of Bicknell’s Thrush (Catharus bicknelli), a threatened Nearctic-Neotropical migratory songbird, is restricted to the Greater Antilles. We analyzed winter records from the mid-1970s to 2009 to quantitatively evaluate winter distribution and habitat selection. Additionally, we conducted targeted surveys in Jamaica (n = 433), Cuba (n = 363), Dominican Republic (n = 1,000), Haiti (n = 131) and Puerto Rico (n = 242) yielding 179 sites with thrush presence. We modeled Bicknell’s Thrush winter habitat selection and distribution in the Greater Antilles in Maxent version 3.3.1. using environmental predictors represented in 30 arc second study area rasters. These included nine landform, land cover and climatic variables that were thought a priori to have potentially high predictive power. We used the average training gain from ten model runs to select the best subset of predictors. Total winter precipitation, aspect and land cover, particularly broadleaf forests, emerged as important variables. A five-variable model that contained land cover, winter precipitation, aspect, slope, and elevation was the most parsimonious and not significantly different than the models with more variables. We used the best fitting model to depict potential winter habitat. Using the 10 percentile threshold (>0.25), we estimated winter habitat to cover 33,170 km², nearly 10% of the study area. The Dominican Republic contained half of all potential habitat (51%), followed by Cuba (15.1%), Jamaica (13.5%), Haiti (10.6%), and Puerto Rico (9.9%). Nearly one-third of the range was found to be in protected areas. By providing the first detailed predictive map of Bicknell’s Thrush winter distribution, our study provides a useful tool to prioritize and direct conservation planning for this and other wet, broadleaf forest specialists in the Greater Antilles.     

Anticipating Knowledge to Inform Species Management: Predicting Spatially Explicit Habitat Suitability of a Colonial Vulture Spreading Its Range

Background

The knowledge of both potential distribution and habitat suitability is fundamental in spreading species to inform in advance management and conservation planning. After a severe decline in the past decades, the griffon vulture (Gyps fulvus) is now spreading its breeding range towards the northwest in Spain and Europe. Because of its key ecological function, anticipated spatial knowledge is required to inform appropriately both vulture and ecosystem management.

Methodology/Findings

Here we used maximum entropy (Maxent) models to determine the habitat suitability of potential and current breeding distribution of the griffon vulture using presence-only data (N = 124 colonies) in north-western Spain. The most relevant ecological factors shaping this habitat suitability were also identified. The resulting model had a high predictive performance and was able to predict species'' historical distribution. 7.5% (∼1,850 km²) of the study area resulted to be suitable breeding habitat, most of which (∼70%) is already occupied by the species. Cliff availability and livestock density, especially of sheep and goats, around 10 km of the colonies were the fundamental factors determining breeding habitat suitability for this species.

Conclusions/Significance

Griffon vultures could still spread 50–60 km towards the west, increasing their breeding range in 1,782 km². According to our results, 7.22% of the area suitable for griffon vulture will be affected by wind farms, so our results could help to better plan wind farm locations. The approach here developed could be useful to inform management of reintroductions and recovery programmes currently being implemented for both the griffon vulture and other threatened vulture species.     

Predicting the Distribution Pattern of Small Carnivores in Response to Environmental Factors in the Western Ghats

Due to their secretive habits, predicting the pattern of spatial distribution of small carnivores has been typically challenging, yet for conservation management it is essential to understand the association between this group of animals and environmental factors. We applied maximum entropy modeling (MaxEnt) to build distribution models and identify environmental predictors including bioclimatic variables, forest and land cover type, topography, vegetation index and anthropogenic variables for six small carnivore species in Mudumalai Tiger Reserve. Species occurrence records were collated from camera-traps and vehicle transects during the years 2010 and 2011. We used the average training gain from forty model runs for each species to select the best set of predictors. The area under the curve (AUC) of the receiver operating characteristic plot (ROC) ranged from 0.81 to 0.93 for the training data and 0.72 to 0.87 for the test data. In habitat models for F. chaus, P. hermaphroditus, and H. smithii “distance to village” and precipitation of the warmest quarter emerged as some of the most important variables. “Distance to village” and aspect were important for V. indica while “distance to village” and precipitation of the coldest quarter were significant for H. vitticollis. “Distance to village”, precipitation of the warmest quarter and land cover were influential variables in the distribution of H. edwardsii. The map of predicted probabilities of occurrence showed potentially suitable habitats accounting for 46 km² of the reserve for F. chaus, 62 km² for V. indica, 30 km² for P. hermaphroditus, 63 km² for H. vitticollis, 45 km² for H. smithii and 28 km² for H. edwardsii. Habitat heterogeneity driven by the east-west climatic gradient was correlated with the spatial distribution of small carnivores. This study exemplifies the usefulness of modeling small carnivore distribution to prioritize and direct conservation planning for habitat specialists in southern India.     

High Resolution Niche Models of Malaria Vectors in Northern Tanzania: A New Capacity to Predict Malaria Risk?

Background

Malaria transmission rates in Africa can vary dramatically over the space of a few kilometres. This spatial heterogeneity reflects variation in vector mosquito habitat and presents an important obstacle to the efficient allocation of malaria control resources. Malaria control is further complicated by combinations of vector species that respond differently to control interventions. Recent modelling innovations make it possible to predict vector distributions and extrapolate malaria risk continentally, but these risk mapping efforts have not yet bridged the spatial gap to guide on-the-ground control efforts.

Methodology/Principal Findings

We used Maximum Entropy with purpose-built, high resolution land cover data and other environmental factors to model the spatial distributions of the three dominant malaria vector species in a 94,000 km² region of east Africa. Remotely sensed land cover was necessary in each vector''s niche model. Seasonality of precipitation and maximum annual temperature also contributed to niche models for Anopheles arabiensis and An. funestus s.l. (AUC 0.989 and 0.991, respectively), but cold season precipitation and elevation were important for An. gambiae s.s. (AUC 0.997). Although these niche models appear highly accurate, the critical test is whether they improve predictions of malaria prevalence in human populations. Vector habitat within 1.5 km of community-based malaria prevalence measurements interacts with elevation to substantially improve predictions of Plasmodium falciparum prevalence in children. The inclusion of the mechanistic link between malaria prevalence and vector habitat greatly improves the precision and accuracy of prevalence predictions (r² = 0.83 including vector habitat, or r² = 0.50 without vector habitat). Predictions including vector habitat are unbiased (observations vs. model predictions of prevalence: slope = 1.02). Using this model, we generate a high resolution map of predicted malaria prevalence throughout the study region.

Conclusions/Significance

The interaction between mosquito niche space and microclimate along elevational gradients indicates worrisome potential for climate and land use changes to exacerbate malaria resurgence in the east African highlands. Nevertheless, it is possible to direct interventions precisely to ameliorate potential impacts.     

Expansion risk of the toxic dinoflagellate Gymnodinium catenatum blooms in Chinese waters under climate change

The paralytic shellfish poison toxin (PST)-producing dinoflagellate, Gymnodinium catenatum, frequently blooms in China, posing a threat to food safety and human health. To understand the drivers of G. catenatum blooms and predict potential habitats for G. catenatum under climate change, samples from occurrence localities and environmental datasets from multiple agencies were aggregated and used to model the habitat suitability of G. catenatum in the China Sea using a maximum entropy model (Maxent). The accumulated variable contributions for the Maxent model were defined to measure the importance of key predictors in the model. The most important environmental variables were distance to the coastline, depth of seawater, and long-term average of the minimum annual temperature. This highlights the main reasons why G. catenatum blooms always occur in coastal waters. Occurrence probabilities higher than 0.66 were defined as habitats with high suitability for shellfish management and aquaculture. Projected habitats with high suitability in Haizhou Bay, coastal waters along the western Taiwan Strait, and Bohai Bay remained stable with increasing temperature by 2100, regardless of the IPCC Representative Concentration Pathways (RCPs). However, those in the China Sea would be reduced overall, leading to a northward movement of the center of integrated habitats. Habitats with a spatial area of >6000 km² in the Bohai Sea, Yellow Sea, and South China Sea and >23,000 km² in the East China Sea would be exposed to high risk under low greenhouse gas emission scenarios (RCP2.6).     

Habitat invasion risk assessment based on Landsat 5 data,exemplified by the shrub Rosa rubiginosa in southern Argentina

Prediction of invasive species spread helps to plan management actions. We performed a risk assessment by quantifying habitat invasibility, predicted the potential distribution of an invasive species using the Maxent modelling program and confirmed patterns using detailed field studies. Our study was conducted in southern Argentina, Patagonia, where large areas are already invaded by the European shrub Rosa rubiginosa. A total of 163 R. rubiginosa locations served as ground truth data, and predictors were obtained from the spaceborne sensor Landsat 5. Based on the Maxent Method (area under the receiver operating characteristic curve 0.8), the habitat invasibility map covered about 5000 km². Our model revealed that R. rubiginosa has the potential to invade 36% of the area along a steep precipitation gradient (target region 600–1400 mm per year). The Tasseled Cap brightness index and the normalized vegetation index explained most of the variance in our model, followed by the Tasseled Cap greenness and wetness indices, which can be interpreted as indicators of disturbance. Highest levels of invasibility were predicted for urban areas, along roads and rivers, on pastures, in Austrocedrus chilensis forests and inside Nothofagus dombeyi forest gaps. Detailed field assessments of rose cover performed in seven habitat types supported these results: rose cover significantly decreased with increasing tree cover (P < 0.01). Our data revealed that the occurrence of R. rubiginosa is not connected to a certain habitat type, but that it thrives in open patches following habitat disturbance. Our approach is a widely applicable, cost‐free remote sensing method that can serve as a risk assessment tool for alien plant species invasion of habitats.     

Non-Sentinel Lymph Node Metastasis Prediction in Breast Cancer with Metastatic Sentinel Lymph Node: Impact of Molecular Subtypes Classification

Introduction

To decipher the interaction between the molecular subtype classification and the probability of a non-sentinel node metastasis in breast cancer patients with a metastatic sentinel lymph-node, we applied two validated predictors (Tenon Score and MSKCC Nomogram) on two large independent datasets.

Materials and Methods

Our datasets consisted of 656 and 574 early-stage breast cancer patients with a metastatic sentinel lymph-node biopsy treated at first by surgery. We applied both predictors on the whole dataset and on each molecular immune-phenotype subgroups. The performances of the two predictors were analyzed in terms of discrimination and calibration. Probability of non-sentinel lymph node metastasis was detailed for each molecular subtype.

Results

Similar results were obtained with both predictors. We showed that the performance in terms of discrimination was as expected in ER Positive HER2 negative subgroup in both datasets (MSKCC AUC Dataset 1 = 0.73 [0.69–0.78], MSKCC AUC Dataset 2 = 0.71 (0.65–0.76), Tenon Score AUC Dataset 1 = 0.7 (0.65–0.75), Tenon Score AUC Dataset 2 = 0.72 (0.66–0.76)). Probability of non-sentinel node metastatic involvement was slightly under-estimated. Contradictory results were obtained in other subgroups (ER negative HER2 negative, HER2 positive subgroups) in both datasets probably due to a small sample size issue. We showed that merging the two datasets shifted the performance close to the ER positive HER2 negative subgroup.

Discussion

We showed that validated predictors like the Tenon Score or the MSKCC nomogram built on heterogeneous population of breast cancer performed equally on the different subgroups analyzed. Our present study re-enforce the idea that performing subgroup analysis of such predictors within less than 200 samples subgroup is at major risk of misleading conclusions.     

The impact of climate change on the distribution of rare and endangered tree Firmiana kwangsiensis using the Maxent modeling

The upsurge in anthropogenic climate change has accelerated the habitat loss and fragmentation of wild animals and plants. The rare and endangered plants are important biodiversity elements. However, the lack of comprehensive and reliable information on the spatial distribution of these organisms has hampered holistic and efficient conservation management measures. We explored the consequences of climate change on the geographical distribution of Firmiana kwangsiensis (Malvaceae), an endangered species, to provide a reference for conservation, introduction, and cultivation of this species in new ecological zones. Modeling of the potential distribution of F. kwangsiensis under the current and two future climate scenarios in maximum entropy was performed based on 30 occurrence records and 27 environmental variables of the plant. We found that precipitation‐associated and temperature‐associated variables limited the potentially suitable habitats for F. kwangsiensis. Our model predicted 259,504 km² of F. kwangsiensis habitat based on 25 percentile thresholds. However, the high suitable habitat for F. kwangsiensis is only about 41,027 km². F. kwangsiensis is most distributed in Guangxi''s protected areas. However, the existing reserves are only 2.7% of the total suitable habitat and 4.2% of the high suitable habitat for the plant, lower than the average protection area in Guangxi (7.2%). This means the current protected areas network is insufficient, underlining the need for alternative conservation mechanisms to protect the plant habitat. Our findings will help identify additional F. kwangsiensis localities and potential habitats and inform the development and implementation of conservation, management, and cultivation practices of such rare tree species.     

Predictive Habitat Modelling as a Tool to Assess the Change in Distribution and Extent of an OSPAR Priority Habitat under an Increased Ocean Temperature Scenario: Consequences for Marine Protected Area Networks and Management

The aims of this study were to determine the extent and distribution of an OSPAR priority habitat under current baseline ocean temperatures; to illustrate the prospect for habitat loss under a changing ocean temperature scenario; and to demonstrate the potential application of predictive habitat mapping in “future-proofing” conservation and biodiversity management.Maxent modelling and GIS environmental envelope analysis of the biogenic bed forming species, Modiolus modiolus was carried out. The Maxent model was tested and validated using 75%/25% training/test occurrence records and validated against two sampling biases (the whole study area and a 20km buffer). The model was compared to the envelope analysis and the area under the receiver operating characteristic curve (Area Under the curve; AUC) was evaluated.The performance of the Maxent model was rated as ‘good’ to ‘excellent’ on all replicated runs and low variation in the runs was recorded from the AUC values. The extent of “most suitable”, “less suitable” and “unsuitable” habitat was calculated for the baseline year (2009) and the projected increased ocean temperature scenarios (2030, 2050, 2080 and 2100). A loss of 100% of “most suitable” habitat was reported by 2080.Maintaining a suitable level of protection of marine habitats/species of conservation importance may require management of the decline and migration rather than maintenance of present extent. Methods applied in this study provide the initial application of a plausible “conservation management tool”.     

Projected impacts of climate change on snow leopard habitat in Qinghai Province,China

Assessing species’ vulnerability to climate change is a prerequisite for developing effective strategies to reduce emerging climate‐related threats. We used the maximum entropy algorithm (MaxEnt model) to assess potential changes in suitable snow leopard (Panthera uncia) habitat in Qinghai Province, China, under a mild climate change scenario. Our results showed that the area of suitable snow leopard habitat in Qinghai Province was 302,821 km² under current conditions and 228,997 km² under the 2050s climatic scenario, with a mean upward shift in elevation of 90 m. At present, nature reserves protect 38.78% of currently suitable habitat and will protect 42.56% of future suitable habitat. Current areas of climate refugia amounted to 212,341 km² and are mainly distributed in the Sanjiangyuan region, Qilian mountains, and surrounding areas. Our results provide valuable information for formulating strategies to meet future conservation challenges brought on by climate stress. We suggest that conservation efforts in Qinghai Province should focus on protecting areas of climate refugia and on maintaining or building corridors when planning for future species management.     

Modeling the potential North American distribution of Russian olive,an invader of riparian ecosystems

Russian olive (Elaeagnus angustifolia L., Elaeagnaceae) has gained notoriety as an invasive tree in the United States (US), particularly owing to its impacts within western riparian ecosystems. In Canada, its potential for range expansion has yet to be assessed, despite alarming infestations in parts of southern British Columbia (BC). Existing niche model predictions are of limited utility because they are restricted to the US, were constructed in the absence of higher latitude records in Canada, and did not consider potentially important soil-related predictors. Here, we address these gaps, and include more than 1400 new occurrence records for Canada, most of which were collected using Google Street View. Our Maxent niche models achieved excellent performance (AUC > 0.9), and identified mean temperature of the coldest quarter and topsoil pH as the first and second-most important predictor variables, respectively, neither of which was included in previously published niche models. High habitat suitability is predicted in areas of western Canada that presently lack occurrence records, including along several major rivers in south-central BC. Our findings should prove valuable to nascent detection and management efforts in western Canada, and also highlight the benefits of basing niche model predictions on occurrence records encompassing as much of the invaded range as possible.     

Habitat suitability of Gymnocladus assamicus - A critically endangered plant of Arunachal Pradesh,India using machine learning and statistical modeling

The present study sought to identify the potential distribution range of critically endangered Gymnocladus assamicus in Arunachal Pradesh based on published data and field collection. We used the Maxent model to estimate the range of distribution and the result was then compared with three other models, i.e., the Generalized Linear Model (GLM), the Bioclim and the Random Forest model to assess the species' habitat suitability. A total of 23 different environmental variables were used, including bioclimatic ones, monthly minimum and maximum temperature, monthly precipitation and elevation data. The Maxent output listed 12 variables explaining 99.9% variation in the model. In comparison, Maxent showed the maximum region under habitat suitability criteria (1884.48 km²), followed by Random Forest (70.73 km²) and Bioclim (11.62 km²) model. Except for the Maxent model, suitable habitats predicted by other models are highly restricted within and across the study species' current distribution range. The average model prediction shows an expanded distribution range for the species up to Tawang which is the closest district of currently known distribution of the species in the state. Thus, the present study recognizes the importance of the geographic range of G. assamicus, a critically endangered species with very limited spatial distribution range and also provides some specific details to explore possible habitats for the species in new areas of potential occurrence in Arunachal Pradesh, India.     

Predicting potential distribution and identifying priority areas for conservation of the lowland tapir (Tapirus terrestris) in Peruvian Amazon

A key conservation biology tool is the information on the geographic distribution of species as well as the variables driving those patterns. Here, we used maximum entropy modeling, MaxEnt, to model the total potential distribution of Tapirus terrestris, classified as “Vulnerable” on the IUCN Red List of Threatened Species. In this study, we recorded 117 occurrence records and considered 18 environmental variables. The total potential distribution area covers 96,055.6 km², meaning 12.3 % of the territory of the Peruvian Amazon, with “high potential” habitat covering 3,891.36 km², “moderate potential” habitat covering 22,849.5 km², and “low potential” habitat covering 69,314.7 km². Natural Protected Areas (NPAs) shelter 32.2 % (30,966.2 km²) of the total potential distribution area of the species, being the Bahuaja Sonene and Manu National Parks, the NPAs with the largest total potential distribution, 8,220.2 km² and 7,619.7 km² respectively. Eventually, 67.8 % (65,089.4 km²) of the total potential distribution were identified without any type of protection category by SINANPE and its complementary categories; therefore, we consider this area as a priority for the conservation of T. terrestris in Peru.     

Comparing Population Patterns to Processes: Abundance and Survival of a Forest Salamander following Habitat Degradation

Habitat degradation resulting from anthropogenic activities poses immediate and prolonged threats to biodiversity, particularly among declining amphibians. Many studies infer amphibian response to habitat degradation by correlating patterns in species occupancy or abundance with environmental effects, often without regard to the demographic processes underlying these patterns. We evaluated how retention of vertical green trees (CANOPY) and coarse woody debris (CWD) influenced terrestrial salamander abundance and apparent survival in recently clearcut forests. Estimated abundance of unmarked salamanders was positively related to CANOPY ( _Canopy  = 0.21 (0.02–1.19; 95% CI), but not CWD ( _CWD  = 0.11 (−0.13–0.35) within 3,600 m² sites, whereas estimated abundance of unmarked salamanders was not related to CANOPY ( _Canopy  = −0.01 (−0.21–0.18) or CWD ( _CWD  = −0.02 (−0.23–0.19) for 9 m² enclosures. In contrast, apparent survival of marked salamanders within our enclosures over 1 month was positively influenced by both CANOPY and CWD retention ( _Canopy  = 0.73 (0.27–1.19; 95% CI) and _CWD  = 1.01 (0.53–1.50). Our results indicate that environmental correlates to abundance are scale dependent reflecting habitat selection processes and organism movements after a habitat disturbance event. Our study also provides a cautionary example of how scientific inference is conditional on the response variable(s), and scale(s) of measure chosen by the investigator, which can have important implications for species conservation and management. Our research highlights the need for joint evaluation of population state variables, such as abundance, and population-level process, such as survival, when assessing anthropogenic impacts on forest biodiversity.     

Predicting seasonal habitat suitability for the critically endangered African wild ass in the Danakil,Ethiopia

The African wild ass (Equus africanus) is the most endangered wild equid in the world and is listed as a Critically Endangered (CR) on the IUCN Red list. Today, only relict populations remain in Ethiopia and Eritrea. The current Ethiopian population persists in the Danakil Desert at a very low density. Wildlife managers need to identify the extent of the remaining suitable habitat and understand human–wildlife interactions for appropriate conservation strategies. This study employed the maximum entropy model (Maxent) to determine suitable habitat and seasonal distribution of African wild ass in the Danakil Desert of Ethiopia. Field surveys were conducted four times annually, twice during the wet season and twice during the dry season, for 2 years. Field data and predictor variables were separated into the dry and wet seasons, and models were generated for each season independently. Distance from water, distance from settlements, herbaceous cover and slope were the best predictors of suitable habitat for both dry and wet seasons. Evaluations of model performances were high with area under the curve (AUC) values of 0.94 and 0.95 for the dry and wet seasons, respectively. Our results will be critical for identifying the available suitable habitat that should be conserved to safeguard this species from extinction.     

Modelling the distribution of a potential invasive tropical fern,Cyclosorus afer in Nigeria

In this study, we predicted the distribution of Cyclosorus afer in Nigeria using the Maximum Entropy (Maxent) technique. We used 95 occurrence points in one State to extrapolate its spread in other States in Nigeria. Three sites of sizes 500 × 500 m² separated by minimum distance of 1,000 m² were sampled in the study area. Seven bioclimatic and elevation variables were selected for the model. Maxent model was run using standard settings with 70% of the occurrence data as training and remaining 30% as test data. The result showed that Maxent performed better than random prediction due to the area under curve for receiver operating characteristics of training (0.990) and test data (0.987) closer to 1. The sensitivity of the Maxent model for occurrence data of C. afer was found to be 100%. The model predicted higher probability of occurrence covering an area of 26019.11 km² in 4 States of Nigeria. Jackknife evaluation of the model revealed that the environmental predictors of C. afer in Nigeria are annual mean temperature, mean temperature of driest quarter, precipitation seasonality, Precipitation of driest quarter and precipitation of coldest quarter. These variables all showed higher gain and contributions to the model.     

松嫩平原苍鹭秋季栖息地选择及适宜性分布

苍鹭（Ardea cinerea）是松嫩平原湿地的常见鸟种,松嫩平原也是苍鹭重要的栖息地。为了了解苍鹭潜在栖息地的适宜性分布,利用GPS/GSM卫星跟踪技术,结合遥感影像和地理信息系统,应用Maxent模型对松嫩平原苍鹭秋季潜在的栖息地进行了评价,并对其适宜性分布进行了分析。结果显示：水源距离和绿度指数是影响松嫩平原苍鹭秋季栖息地适宜性的重要环境变量;松嫩平原内苍鹭适宜栖息地面积为2761.06 km²（占研究区域的1.24%）,主要分布在大庆（756.86 km²,占适宜栖息地面积的27.41%）、白城（537.14 km²,占适宜栖息地面积的19.45%）、齐齐哈尔（439.43 km²,占适宜栖息地面积的15.92%）等地市行政区,以大庆市杜尔伯特蒙古族自治县（429.90 km²,占适宜栖息地面积的15.57%）、白城市镇赉县（334.92 km²,占适宜栖息地面积的12.13%）、大庆市肇源县（185.54 km²,占适宜栖息地面积的6.72%）等县级行政区为主;其中,15.79%的适宜栖息地依次受到莫莫格保护区（10.34%）、扎龙保护区（3.47%）、向海保护区（0.67%）、查干湖保护区（0.54%）、大布苏保护区（0.41%）、乌裕尔河保护区（0.36%）等国家级自然保护区的保护。建议对未受到保护的零星小面积栖息地给与更多关注。     

Ecological modelling for the conservation of Gluta travancorica Bedd. - An endemic tree species of southern Western Ghats,India

Endemic species are highly adapted to grow exclusively in a specific geographical area. The goal of the current study is to determine the probable habitat distribution range of the narrowly endemic species Gluta travancorica. An ecological niche modelling is carried out, using four different models viz., BioClim, MaxEnt, Random Forest and Deep Neural Networks (DNN). A total of 506 G. travancorica cluster locations were surveyed and used for this study with thirty different ecogeographic, edaphic and bioclimatic environmental parameters. After a preliminary investigation using multi-collinearity analysis, soil parameter variables like pH, cation exchange capacity (CEC), silt and clay content are used for final modelling. Factor analysis of ecological niche revealed the soil parameters like pH, CEC, silt and clay content as the key predictors. The result is validated using true skill statistics, sensitivity, specificity, kappa statistic and AUC-ROC. Results of the present study show that DNN have exceptional prediction performance, demonstrated by its AUC score of 0.959. DNN model projected 32.37% (938.18 km²) of the study region to have a ‘highly suitable habitat’, whereas 67.63% (1960.82 km²) was classified as having ‘low habitat suitability’. Besides, back-to-field assessments have also proven DNN's potential in predicting the habitat suitability of G. travancorica. The study results will facilitate the prioritization of conservation and seedling restoration strategies. The forest range explored in this work is a component of one of the most important global biodiversity hotspots, and it has significant implications for regional biodiversity conservation.