The impact of alternative trait‐scaling hypotheses for the maximum photosynthetic carboxylation rate (Vcmax) on global gross primary production. Issue 4 (23rd June 2017)

• Record Type: Journal Article
• Authors: Walker, Anthony P.; Quaife, Tristan; van Bodegom, Peter M.; De Kauwe, Martin G.; Keenan, Trevor F.; Joiner, Joanna; Lomas, Mark R.; MacBean, Natasha; Xu, Chongang; Yang, Xiaojuan; Woodward, F. Ian
• Is Part Of: New phytologist. Volume 215:Issue 4(2017)
• Page Start: 1370
• Page End: 1386
• Publication Date: 2017-06-23
• Subjects: assumption‐centred modelling -- co‐ordination hypothesis -- Dynamic Global Vegetation Model (DGVM) -- gross primary production (GPP) -- modelling photosynthesis -- plant functional traits -- terrestrial carbon cycle -- trait‐based modelling; Botany -- Periodicals; 580
• DOI: 10.1111/nph.14623
• Languages: English
• ISSNs: 0028-646X

Abstract

Summary: The maximum photosynthetic carboxylation rate ( V cmax ) is an influential plant trait that has multiple scaling hypotheses, which is a source of uncertainty in predictive understanding of global gross primary production (GPP). Four trait‐scaling hypotheses (plant functional type, nutrient limitation, environmental filtering, and plant plasticity) with nine specific implementations were used to predict global V cmax distributions and their impact on global GPP in the Sheffield Dynamic Global Vegetation Model (SDGVM). Global GPP varied from 108.1 to 128.2 PgC yr −1, 65% of the range of a recent model intercomparison of global GPP. The variation in GPP propagated through to a 27% coefficient of variation in net biome productivity (NBP). All hypotheses produced global GPP that was highly correlated ( r = 0.85–0.91) with three proxies of global GPP. Plant functional type‐based nutrient limitation, underpinned by a core SDGVM hypothesis that plant nitrogen (N) status is inversely related to increasing costs of N acquisition with increasing soil carbon, adequately reproduced global GPP distributions. Further improvement could be achieved with accurate representation of water sensitivity and agriculture in SDGVM. Mismatch between environmental filtering (the most data‐driven hypothesis) and GPP suggested that greater effort is needed understand V cmax variation in the field, particularly in northern latitudes.