Multiscale ecological niche modeling exhibits varying climate change impacts on habitat suitability of Madrean Pine-Oak trees

This resource contains the code and data used in this publication.

Publication Abstract:
Anthropogenic climate change and increasing greenhouse gas emissions are expected to globally impact the biological function, community structure, and spatial distribution of biodiversity. Many existing studies explore the effect of climate change on biodiversity, generally at a single spatial scale. This study explores the potential effects of climate change on the habitat suitability of seven tree species at two distinct spatial scales: the Coronado National Forest (CNF), a local management area, and the Sierra Madre Occidental (SMO), an ecoregional extent. Habitat suitability was determined by extrapolating Ecological Niche Models (ENMs) based on citizen-science tree occurrence records into future climatic conditions using projected 30-year normals for two anthropogenic emissions scenarios through the end of the century. These ENMs, examined at a spatial resolution of 1 km2, are constructed using a mean average ensemble of three commonly used machine learning algorithms. The results show that habitat suitability is expected to decrease for all seven tree species at varying degrees. Results also show that climate-forcing scenario choice appears to be far less important for understanding changes in species habitat suitability than the spatial scale of modeling extent. Additionally, we observed non-linear changes in tree species habitat suitability within the SMO and CNF dependent on forest community type, latitude, and elevational gradient. The paper concludes with a discussion of the necessary steps to verify the estimated alters of these tree species under climate change. Most importantly, provides a framework for characterizing habitat suitability across spatial scales.

本资源包含本出版物所使用的代码与数据集。

出版物摘要：
人为气候变化与持续增长的温室气体排放预计将对全球生物多样性的生物学功能、群落结构及空间分布产生广泛影响。当前多数相关研究多基于单一空间尺度探讨气候变化对生物多样性的影响。本研究针对两种不同空间尺度，探究气候变化对7种树木物种生境适宜性的潜在影响：其一为科罗纳多国家森林（Coronado National Forest, CNF），即地方管理区域；其二为西马德雷山脉（Sierra Madre Occidental, SMO），即生态区域尺度。
本研究基于公民科学获取的树木物种出现记录构建生态位模型（Ecological Niche Models, ENMs），并结合本世纪末两种人为排放情景下的预估30年气候基准值，将模型外推至未来气候情景，以此确定各物种的生境适宜性。这些以1 km²空间分辨率构建的生态位模型，采用3种常用机器学习算法的平均集成方案构建而成。
研究结果显示，7种树木物种的生境适宜性均将出现不同程度的下降；同时相较于模型模拟范围的空间尺度，气候强迫情景的选择对物种生境适宜性变化的影响相对更小。此外，在西马德雷山脉与科罗纳多国家森林区域内，树木物种的生境适宜性会随森林群落类型、纬度与海拔梯度呈现非线性变化。
本文最后讨论了验证这些树木物种在气候变化下生境变化预估结果所需的必要步骤，并尤为重要地提供了一套跨空间尺度表征生境适宜性的研究框架。