A standard protocol for reporting species distribution models

Ecography, Год журнала: 2020, Номер 43(9), С. 1261 - 1277

Опубликована: Июнь 1, 2020

Species distribution models (SDMs) constitute the most common class of across ecology, evolution and conservation. The advent ready‐to‐use software packages increasing availability digital geoinformation have considerably assisted application SDMs in past decade, greatly enabling their broader use for informing conservation management, quantifying impacts from global change. However, must be fit purpose, with all important aspects development applications properly considered. Despite widespread SDMs, standardisation documentation modelling protocols remain limited, which makes it hard to assess whether steps are appropriate end use. To address these issues, we propose a standard protocol reporting an emphasis on describing how study's objective is achieved through series modeling decisions. We call this ODMAP (Overview, Data, Model, Assessment Prediction) protocol, as its components reflect main involved building other empirically‐based biodiversity models. serves two purposes. First, provides checklist authors, detailing key model analyses, thus represents quick guide generic workflow modern SDMs. Second, introduces structured format documenting communicating models, ensuring transparency reproducibility, facilitating peer review expert evaluation quality, well meta‐analyses. detail elements ODMAP, explain can used different objectives applications, complements efforts store associated metadata define standards. illustrate utility by revisiting nine previously published case studies, provide interactive web‐based facilitate plan advance encouraging further refinement adoption scientific community.

Язык: Английский

---

Predictive performance of presence‐only species distribution models: a benchmark study with reproducible code

Ecological Monographs, Год журнала: 2021, Номер 92(1)

Опубликована: Окт. 8, 2021

Abstract Species distribution modeling (SDM) is widely used in ecology and conservation. Currently, the most available data for SDM are species presence‐only records (available through digital databases). There have been many studies comparing performance of alternative algorithms data. Among these, a 2006 paper from Elith colleagues has particularly influential field, partly because they several novel methods (at time) on global set that included independent presence–absence model evaluation. Since its publication, some further developed new ones emerged. In this paper, we explore patterns predictive across methods, by reanalyzing same (225 six different regions) using updated knowledge practices. We apply well‐established such as generalized additive models MaxEnt, alongside others received attention more recently, including regularized regressions, point‐process weighted random forests, XGBoost, support vector machines, ensemble framework biomod. All use include background samples (a sample environments landscape) fitting. impacts weights presence points introduce ways evaluating fitted to these data, area under precision‐recall gain curve, focusing rank results. find way matters. The top method was an tuned individual models. contrast, ensembles built biomod with default parameters performed no better than single moderate performing Similarly, second forest parameterized deal (contrasted relatively few records), which substantially outperformed other implementations. that, general, nonparametric techniques capability controlling complexity traditional regression MaxEnt boosted trees still among code working examples provided make study fully reproducible.

Язык: Английский

---

Testing whether ensemble modelling is advantageous for maximising predictive performance of species distribution models

Ecography, Год журнала: 2020, Номер 43(4), С. 549 - 558

Опубликована: Янв. 27, 2020

Predictive performance is important to many applications of species distribution models (SDMs). The SDM ‘ensemble’ approach, which combines predictions across different modelling methods, believed improve predictive performance, and used in recent studies. Here, we aim compare the ensemble that individual models, using a large presence–absence dataset eucalypt tree species. To test model divided our into calibration evaluation folds two spatial blocking strategies (checkerboard‐pattern latitudinal slicing). We calibrated cross‐validated all within folds, both repeated random division data (a common approach) blocking. Ensembles were built software package ‘biomod2’, with standard (‘untuned’) settings. Boosted regression (BRT) also fitted same data, tuned according published procedures. then ensembles against their component untuned BRTs. area under receiver‐operating characteristic curve (AUC) log‐likelihood for assessing performance. In tests, performed well, but not consistently better than or BRTs tests. Moreover, choosing best cross‐validation yielded good external blocked proving suited this choice, study, cross‐validation. slice was only possible four species; showed some particularly one, performing ensembles. This study shows no particular benefit over models. It suggests further robust testing required situations where are predict distant places environments.

Язык: Английский

---

A comparison between Ensemble and MaxEnt species distribution modelling approaches for conservation: A case study with Egyptian medicinal plants

Ecological Informatics, Год журнала: 2020, Номер 60, С. 101150 - 101150

Опубликована: Сен. 3, 2020

Язык: Английский

---

Want to model a species niche? A step-by-step guideline on correlative ecological niche modelling

Ecological Modelling, Год журнала: 2021, Номер 456, С. 109671 - 109671

Опубликована: Июль 19, 2021

Язык: Английский

---

ENMTML: An R package for a straightforward construction of complex ecological niche models

Environmental Modelling & Software, Год журнала: 2020, Номер 125, С. 104615 - 104615

Опубликована: Янв. 6, 2020

Язык: Английский

---

Model complexity affects species distribution projections under climate change

Journal of Biogeography, Год журнала: 2019, Номер 47(1), С. 130 - 142

Опубликована: Ноя. 3, 2019

Abstract Aim Statistical species distribution models (SDMs) are the most common tool to predict impact of climate change on biodiversity. They can be tuned fit relationships at various levels complexity (defined here as parameterization complexity, number predictors, and multicollinearity) that may co‐determine whether projections novel climatic conditions useful or misleading. Here, we assessed how model affects performance extrapolations influences ranges under future change. Location Europe. Taxon 34 European tree species. Methods We sampled three replicates predictor sets for all combinations 10 ( n = 3–12) environmental variables (climate, terrain, soil) multicollinearity. used these each four SDM algorithms complexity. The >100,000 resulting fits were then evaluated block cross‐validation projected 2061–2080 considering two emission scenarios. Finally, investigated design with distributional changes. Results Model affected both Fits intermediate performed best, more complex parameterizations associated higher loss current ranges. peaked 10–11 but increasing had no consistent effect projections. Multicollinearity a low distinctly increased Main conclusions SDM‐based assessments should based ensembles projections, varying well besides scenarios models. kept reasonably small classical threshold maximum absolute Pearson correlation 0.7 restricts collinearity‐driven effects in

Язык: Английский

---

Natural hybridization reduces vulnerability to climate change

Nature Climate Change, Год журнала: 2023, Номер unknown

Опубликована: Янв. 30, 2023

Abstract Under climate change, species unable to track their niche via range shifts are largely reliant on genetic variation adapt and persist. Genomic vulnerability predictions used identify populations that lack the necessary variation, particularly at climate-relevant genes. However, hybridization as a source of novel adaptive is typically ignored in genomic studies. We estimated environmental models for closely related rainbowfish ( Melanotaenia spp.) across an elevational gradient Australian Wet Tropics. Hybrid between widespread generalist several narrow endemic exhibited reduced projected climates compared pure endemics. Overlaps introgressed regions were consistent with signal introgression. Our findings highlight often-underappreciated conservation value hybrid indicate introgression may contribute evolutionary rescue ranges.

Язык: Английский

---

Top ten hazards to avoid when modeling species distributions: a didactic guide of assumptions, problems, and recommendations

Ecography, Год журнала: 2024, Номер 2024(4)

Опубликована: Янв. 31, 2024

Species distribution models, also known as ecological niche models or habitat suitability have become commonplace for addressing fundamental and applied biodiversity questions. Although the field has progressed rapidly regarding theory implementation, key assumptions are still frequently violated recommendations inadvertently overlooked. This leads to poor being published used in real‐world applications. In a structured, didactic treatment, we summarize what our view constitute ten most problematic issues, hazards, negatively affecting implementation of correlative approaches species modeling (specifically those that model by comparing environments species' occurrence records with background pseudoabsence sample). For each hazard, state relevant assumptions, detail problems arise when violating them, convey straightforward existing recommendations. We discuss five major outstanding questions active current research. hope this contribution will promote more rigorous these valuable stimulate further advancements.

Язык: Английский

---

Modelling current and future potential distributions of two desert jerboas under climate change in Iran

Ecological Informatics, Год журнала: 2019, Номер 52, С. 7 - 13

Опубликована: Апрель 17, 2019

Язык: Английский

---