Department of Biological Sciences, University of Bergen, Bergen, Norway.
Bjerknes Centre for Climate Research, University of Bergen, Bergen, Norway.
Glob Chang Biol. 2021 May;27(10):2088-2101. doi: 10.1111/gcb.15532. Epub 2021 Feb 20.
Context-dependencies in species' responses to the same climate change frustrate attempts to generalize and make predictions based on experimental and observational approaches in biodiversity science. Here, we propose predictability may be enhanced by explicitly incorporating macroecological context into analyses of species' responses to climate manipulations. We combined vascular plant species' responses to an 8-year, 12-site turf transplant climate change experiment set in southwestern Norway with climate niche data from the observed 151 species. We used the difference between a species' mean climate across their range and climate conditions at the transplant site ("climate differences") to predict colonization probability, extinction probability, and change in abundance of a species at a site. In analyses across species that ignore species-specific patterns, colonization success increased as species' distribution optima were increasingly warmer than the experimental target site. Extinction probability increased as species' distribution optima were increasingly colder than the target site. These patterns were reflected in change in abundance analyses. We found weak responses to increased precipitation in these oceanic climates. Climate differences were better predictors of species' responses to climate manipulations than range size. Interestingly, similar patterns were found when analyses focused on variation in species-specific responses across sites. These results provide an experimental underpinning to observational studies that report thermophilization of communities and suggest that space-for-time substitutions may be valid for predicting species' responses to climate warming, given other conditions are accounted for (e.g., soil nutrients). Finally, we suggest that this method of putting climate change experiments into macroecological context has the potential to generalize and predict species' responses to climate manipulations globally.
在生物多样性科学中，由于同一气候变化对不同物种的影响存在差异，因此尝试基于实验和观测方法进行概括和预测的努力受到了阻碍。在这里，我们提出通过明确将宏观生态背景纳入对物种对气候干预的响应分析中，可以提高可预测性。我们结合了挪威西南部一个为期 8 年、涉及 12 个地点的草坪移植气候变化实验中维管植物物种对气候变化的响应数据，以及观察到的 151 个物种的气候生态位数据。我们使用物种在其分布范围内的平均气候与移植地点的气候条件之间的差异（“气候差异”）来预测物种在一个地点的定植概率、灭绝概率和丰度变化。在忽略物种特异性模式的跨物种分析中，当物种的分布最适温度比实验目标地点更温暖时，定植成功率增加。当物种的分布最适温度比目标地点更冷时，灭绝概率增加。这些模式在丰度变化分析中得到了反映。在这些海洋性气候中，我们发现降水增加的响应较弱。气候差异是预测物种对气候干预响应的更好指标，而不是物种分布范围大小。有趣的是，当分析集中在不同地点的物种特异性响应变化时，也发现了类似的模式。这些结果为报告群落热适应的观测研究提供了实验依据，并表明在考虑其他条件（例如土壤养分）的情况下，时空替代可能是预测物种对气候变暖响应的有效方法。最后，我们认为，将气候变化实验置于宏观生态背景下的这种方法有可能在全球范围内推广和预测物种对气候干预的响应。
