Integrating Species‐Specific Information in Models Improves Regional Projections Under Climate Change. Issue 12 (19th June 2019)

• Record Type: Journal Article
• Authors: Remy, Cécile C.; Krofcheck, Dan J.; Keyser, Alisa R.; Litvak, Marcy E.; Collins, Scott L.; Hurteau, Matthew D.
• Is Part Of: Geophysical research letters. Volume 46:Issue 12(2019)
• Page Start: 6554
• Page End: 6562
• Publication Date: 2019-06-19
• Subjects: LANDIS‐II -- PnET‐succession -- eddy‐covariance -- North America southwest -- carbon -- climate change; Geophysics -- Periodicals; Planets -- Periodicals; Lunar geology -- Periodicals; 550
• DOI: 10.1029/2019GL082762
• Languages: English
• ISSNs: 0094-8276

Abstract

Abstract: Models commonly used to project forest carbon response to climate change reduce biodiversity to a small number of plant functional types or plant functional traits for the sake of computational efficiency at large spatial scales. We simulated the climate sensitivity of the dominant woody vegetation types in New Mexico using both a generalized functional type and a species‐specific model parameterization. Both parameterizations achieve reasonable current carbon uptake rates and aboveground biomass amount at the ecosystem scale. When vegetation types are subjected to increasing temperature and decreasing precipitation, the generalized parameterization differs substantially from the species‐specific parameterization by homogenizing the diversity of adaptations for dealing with higher temperature and drought, leading to divergent responses under changing climate. We recommend integrating species‐specific information, when available, to improve projections of climate change impacts on forested ecosystems to develop robust ecosystem management strategies at regional scales. Plain Language Summary: Vegetation responses to climate change are commonly simulated using models that generalize the characteristics of species and ecosystems to facilitate global–scale modeling efforts. We compared the climate sensitivity of the dominant woody vegetation types in New Mexico using a simplified model parameterization that treated all species the same, regardless of ecosystem type, versus a species‐specific model parameterization. Our results show that a simplified model parameterization can achieve reasonable current carbon uptake rates at the ecosystem scale. However, when subjected to increasing temperature and decreasing precipitation, the generalized parameterization differs substantially from the species‐specific parameterization by homogenizing the diversity of adaptations for dealing with higher temperature and drought. We recommend integrating species‐specific information, when available, to facilitate the development of ecosystem management strategies because management decisions focus on the biology of the species that comprise ecosystems. Key Points: Models using plant functional types or plant functional traits at large spatial scales lead to erroneous biome responses under changing climate Species‐specific model parameterization more accurately captures individual species responses, a key tool to optimize forest management strategies Databases of species‐specific physiologic traits for parameterization should help to improve ecosystem model projections under future climate