Comparing the performance of global, geographically weighted and ecologically weighted species distribution models for Scottish wildcats using GLM and Random Forest predictive modeling

Ecological Modelling, Journal Year: 2024, Volume and Issue: 492, P. 110691 - 110691

Published: April 8, 2024

Species distribution modeling has emerged as a foundational method to predict occurrence and suitability of species in relation environmental variables advance ecological understanding guide conservation planning. Recent research, however, shown that species-environmental relationships habitat model predictions are often nonstationary space, time context. This calls into question approaches assume global, stationary realized niche use predictive describe it. paper explores this issue by comparing the performance models for wildcat hybrid based on (1) global pooled data across individuals, (2) geographically weighted aggregation individual models, (3) ecologically (4) combinations geographical weighting. Our study system included GPS telemetry from 14 hybrids Scotland. We developed both using Generalized Linear Models (GLM) Random Forest machine learning compare these differing algorithms how they analyses. validated predicted four different ways. First, we used independent hold-out collared hybrids. Second, 8 additional previous were not training sample. Third, sightings sent public researchers expert opinion. Fourth, collected camera trap surveys between 2012 – 2021 various sources produce combined dataset showing where wildcats had been detected. results show validation individuals train provides highly biased assessment true other locations, with particular appearing perform exceptionally (and inaccurately) well when same models. Very obtained three sources. Each sets gave result terms best overall model. The average datasets suggested produced potential was an ensemble Model GLM suggests debate over whether which vs is superior or aggregated may be false choice. presented here prediction applies combination all framework.

Language: Английский

---

Modeling the effects of climate change scenarios on the potential distribution of Vespa crabro Linnaeus, 1758 (Hymenoptera: Vespidae) in a Mediterranean biodiversity hotspot

Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103006 - 103006

Published: Jan. 1, 2025

Language: Английский

---

Machine learning methods: Modeling net growth in the Atlantic Forest of Brazil

Ecological Informatics, Journal Year: 2024, Volume and Issue: 81, P. 102564 - 102564

Published: March 18, 2024

Tree growth models are an important and essential part of modeling forest dynamics valuable tools for management planning biodiversity conservation strategies. We applied three different machine learning models, namely Artificial Neural Networks (ANN), Support Vector Machine (SVM) Random Forest (RF) to predict tree at the plot-level in Atlantic Brazil. attributes, land use history, landscape, soil climatic characteristics were used modeling. Recursive Feature Elimination was select best subset predictor variables. found that edaphic, attributes variables shaping Brazilian Forest. Soil acidity most characteristic. The methods efficient. method showed superiority over others Nemenyi test pointed out difference between RF model other techniques greater than calculated critical (CD). can be tool fragments They help understanding biome developing strategies aimed recovering reducing deleterious effects fragmentation.

Language: Английский

---

Host, environment, and anthropogenic factors drive landscape dynamics of an environmentally transmitted pathogen: Sarcoptic mange in the bare‐nosed wombat

Journal of Animal Ecology, Journal Year: 2023, Volume and Issue: 92(9), P. 1786 - 1801

Published: May 23, 2023

Understanding the spatial dynamics and drivers of wildlife pathogens is constrained by sampling logistics, with implications for advancing field landscape epidemiology targeted allocation management resources. However, visually apparent diseases, when combined remote-surveillance distribution modelling technologies, present an opportunity to overcome this landscape-scale problem. Here, we investigated disease, using clinical signs sarcoptic mange (caused Sarcoptes scabiei) in its bare-nosed wombat (BNW; Vombatus ursinus) host. We used 53,089 camera-trap observations from over 3261 locations across 68,401 km2 area Tasmania, Australia, data ensemble species (SDM). investigated: (1) variables predicted drive habitat suitability host; (2) host associated disease (3) environmental conditions at greatest risk occurrence, including some Bass Strait islands where BNW translocations are proposed. showed that Tasmanian landscape, ecosystems therein, nearly ubiquitously suited BNWs. Only high mean annual precipitation reduced In contrast, BNWs were widespread, but heterogeneously distributed landscape. Mange (which environmentally transmitted BNWs) was most likely be observed areas increased suitability, lower precipitation, near sources freshwater topographic roughness minimal (e.g. human modified landscapes, such as farmland intensive land-use areas, shrub grass lands). Thus, a confluence host, anthropogenic appear influence transmission S. scabiei. identified Islands highly suitable mix low pathogen. This study largest assessment any species, advances understanding research illustrates how host-pathogen co-suitability can useful allocating resources

Language: Английский

---

Cross-realm transferability of species distribution models–Species characteristics and prevalence matter more than modelling methods applied

Ecological Modelling, Journal Year: 2024, Volume and Issue: 499, P. 110950 - 110950

Published: Nov. 20, 2024

Language: Английский

---

Identifying gaps in the conservation of small wild cats of Southeast Asia

Biodiversity and Conservation, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 20, 2025

Abstract Southeast Asia hosts more felid species than any other region and, although smaller (< 30 kg) felids have important ecological roles, regional conservation has mainly focused on a few charismatic big cats. Information the ecology and status of small is often lacking or geographically limited. We used empirically derived scale-optimized models for seven in three regions (mainland, Borneo Sumatra) to evaluate effectiveness existing protected areas network preserving suitable habitats, map protection. Finally, we assessed whether are good proxies broader terrestrial biodiversity. On mainland, largest most habitats occurred Northern Forest Complex Myanmar between Eastern Myanmar, Laos Vietnam. In these also highlighted areas. Borneo, central highlands Sabah. Sumatra, strongholds habitat suitability were Barisan Mountains, western extent island, highly concentrated within found that aggregated was correlated strongly vertebrate biodiversity single individually, suggesting multiple an association with high overall Overall, our assessment distribution highlights fundamental importance conservation, given associated large extents forest. Our results clarion call expand extent, improve management, remaining core Asia, work enhance protect connectivity them ensure long-term demographic genetic exchange among region’s wildlife populations.

Language: Английский

---

Predicting Bat Roosts in Bridges using Bayesian Additive Regression Trees

Global Ecology and Conservation, Journal Year: 2025, Volume and Issue: unknown, P. e03551 - e03551

Published: March 1, 2025

Language: Английский

---

Machine learning allows for large-scale habitat prediction of a wide-ranging carnivore across diverse ecoregions

Landscape Ecology, Journal Year: 2024, Volume and Issue: 39(5)

Published: May 15, 2024

Abstract Context Resource selection functions are powerful tools for predicting habitat of animals. Recently, machine-learning methods such as random forest have gained popularity due to their flexibility and strong predictive performance. Objectives We tested two continental-scale, second-order a wide-ranging large carnivore, the mountain lion ( Puma concolor ), support continent-wide conservation management, including estimating abundance, predict suitability recolonizing or reintroduced Methods compared generalized linear model (GLM) using GPS location data from 476 individuals across 20 study sites in western USA Canada remotely-sensed landscape data. internally validated models examined ability correctly classify used available points by calculating area under receiver operating characteristics (AUC). performed leave-one-out (LOO) out-of-sample tests strength on both models. Results Both suggested that lions select steeper slopes, areas closer water, with higher normalized difference vegetation index (NDVI), against variables associated human impact. The (AUC = 0.94) demonstrated can be accurately predicted at continental scales, outperforming traditional GLM 0.68). Our LOO validation provided similar results (x̄ 0.93 x̄ 0.65 GLM). Conclusions found added deeper insights into how individual covariates impacted diverse ecosystems. analyses our unoccupied where local unavailable. thus provides tool discussions relevant management metapopulation abundance.

Language: Английский

---

Beyond Description: Unlocking the Predictive Potential of African Ecology

African Journal of Ecology, Journal Year: 2025, Volume and Issue: 63(3)

Published: April 1, 2025

ABSTRACT Ecology's strength lies in its ability to explain and predict interactions between organisms their environment. However, African ecological research has historically been dominated by descriptive studies, focusing on biodiversity patterns, species distributions, behavioural observations or monitoring of large mammal populations (especially East savannahs). This pattern also traditionally characterised the studies community ecology. While valuable, these often fall short providing predictive insights essential for addressing pressing challenges such as climate change, ecosystem resilience. We advocate a paradigm shift ecology—moving beyond description hypothesis‐driven, research. Community ecology Africa can transcend documentation uncover mechanisms underlying processes integrating methodologies null models, Monte Carlo simulations modelling based upon data mining techniques. Predictive interactions, assembly functions have potential enhance both theoretical applied science, ensuring global relevance. Curriculum reforms statistics methodological training academic institutions will be crucial fostering this transformation. As Journal Ecology seeks champion transition, we urge researchers embrace frameworks that not only document but provide actionable into dynamics. could achieved re‐analysing long‐term sets published several less‐distributed journals, other languages than English. is critical positioning at forefront international discourse, driving impactful conservation management strategies.

Language: Английский

---

Differentially biased sampling strategies reveal the non-stationarity of species distribution models for Indian small felids

Ecological Modelling, Journal Year: 2024, Volume and Issue: 493, P. 110749 - 110749

Published: May 11, 2024

Language: Английский

---