Improving collaborations between empiricists and modelers to advance grassland community dynamics in ecosystem models

• Published in: The New phytologist 2020-12, Vol.228 (5), p.1467-1471
• Main Authors: Wilcox, Kevin R., Komatsu, Kimberly J., Avolio, Meghan L.
• Format: Article
• Language: English
• Subjects: community ecology; dynamic global vegetation model (DGVM); Ecology; Ecosystem; ecosystem function; Ecosystem models; Environment models; ENVIRONMENTAL SCIENCES; gap model; Grassland; grassland dynamics; Grasslands; Meetings; process‐based models; statistical models; trait‐based models
• ISSN: 0028-646X; 1469-8137
• DOI: 10.1111/nph.16900

Climate change, increasing atmospheric CO2, and land use change have altered biogeochemical and hydrologic cycles world-wide, with grassland systems being particularly vulnerable to resulting vegetation shifts (Komatsu et al., 2019). Therefore, incorporating plant community dynamics into ecosystem models is critical for accurate forecasting of ecosystem responses to global change (Levine, 2016). Process-based ecosystem models, which simulate the biogeochemical transfers of mass and energy among biota, the subsurface, and atmosphere, require representation of dynamic composition of organisms within ecosystems. For example, these models simulate leaf and plant-level characteristics, such as electron transport rate and allometry of carbon (C) allocation, to predict how net primary productivity and other ecosystem processes respond to abiotic drivers. These models are particularly useful in scaling from organismal to ecosystem levels but are still underdeveloped in their ability to capture community change, especially in grassland ecosystems. To represent compositional changes, these models must simulate competition, mortality, establishment, and reproduction of plant populations within communities. Yet, current ecosystem modeling approaches to forecast plant community change have derived from studies of forested systems and are either too coarse to capture fine-scale community dynamics (e.g. dynamic global vegetation models (DGVMs)) or too complex to be used at large spatial scales (e.g. forest gap models).