The Science of The Total Environment, Год журнала: 2024, Номер 958, С. 178091 - 178091
Опубликована: Дек. 20, 2024
Язык: Английский
Global Ecology and Biogeography, Год журнала: 2023, Номер 33(3), С. 371 - 384
Опубликована: Ноя. 29, 2023
Abstract Aim Ecological forecasting is critical in understanding of ecological responses to climate change and increasingly used mitigation plans. The forecasts from correlative models can be challenged by model complexity, training collinearity, collinearity shift novel conditions predictors that are common during extrapolation. individual effect these four factors has been investigated, but it still unclear how interactively affect forecasting. To fill this gap, we conducted a comprehensive simulation experiment quantify the influence Location Simulated regions. Time Period scenarios. Methods We modelled three response variables commonly following normal, Poisson binomial distributions as function functional relationships represented complexity under levels using generalized linear models. By calculating prediction error 3,780,000 testing scenarios, partitioned its variance shift, predictor novelty their interactions. Results found increased degraded performance, leading up double errors when predictor's range ~22% or correlation r between two changed >~0.8 for combination high interaction relationship. Predictor reduced on suggesting negative them. This pattern was more pronounced collinearity. Main Conclusions accuracy depends Besides consideration parsimonious 0.7 training, our study further recommends threshold <22%–50% depending and/or <0.8 making reliable
Язык: Английский
Процитировано
6
Ecology and Evolution, Год журнала: 2024, Номер 14(6)
Опубликована: Июнь 1, 2024
Abstract In response to the pressing challenges of ongoing biodiversity crisis, protection endangered species and their habitats, as well monitoring invasive are crucial. Habitat suitability modeling (HSM) is often treated silver bullet address these challenges, commonly relying on generic variables sourced from widely available datasets. However, for with high habitat requirements, or habitats within geographic range a species, at coarse level detail may fall short. Consequently, there potential value in considering incorporation more targeted data, which extend beyond readily land cover climate this study, we investigate impact incorporating (specifically tree composition) vertical structure information (derived LiDAR data) HSM outcomes three forest specialist bat ( Barbastella barbastellus , Myotis bechsteinii Plecotus auritus ) Rhineland‐Palatinate, Germany, compared utilized environmental variables, such land‐cover classifications (e.g., Corine Land Cover) Bioclim). The integration enhanced performance models all species. Furthermore, our results showed difference distribution maps that resulted using different levels variables. This underscores importance making effort generate appropriate rather than simply used ones, necessity exercising caution when tool inform conservation strategies spatial planning efforts.
Язык: Английский
Процитировано
2
The Science of The Total Environment, Год журнала: 2024, Номер 953, С. 175794 - 175794
Опубликована: Сен. 3, 2024
Язык: Английский
Процитировано
2
Ecology and Evolution, Год журнала: 2024, Номер 14(10)
Опубликована: Окт. 1, 2024
Species distribution modeling (SDM) is an essential tool in ecology and conservation for predicting species distributions based on presence/absence data environmental variables. The present study aimed to understand the pattern habitat suitability of
Язык: Английский
Процитировано
2
The Science of The Total Environment, Год журнала: 2024, Номер 958, С. 178091 - 178091
Опубликована: Дек. 20, 2024
Язык: Английский
Процитировано
2