Great uncertainties in modeling grazing impact on carbon sequestration: a multi-model inter-comparison in temperate Eurasian Steppe

Abstract

The impact of grazing activity on terrestrial carbon (C) sequestration has been noticed and studied worldwide. Recent efforts have been made to incorporate the disturbance into process-based land models. However, the performance of grazing models has not been well investigated at large scales. In this study, we performed a spatially explicit model uncertainty assessment in the world's largest pasture ecosystem, the temperate Eurasian Steppe. Five grazing models were explicitly incorporated into a single terrestrial biogeochemical model to simulate regional C consumption from grazing activity (C graze ). First, we summarized the underlying mechanisms and explicitly compared the general functions used to describe the processes in different models. Then, the models (five models with 12 simulations) were run in parallel using the same forcing data and livestock distribution map in 2006. Results indicated that the modeled regional C graze varied from 0.1-16.1 gC m −2 for the year. The corresponding ratios of C graze to aboveground net primary productivity ANPP and net primary productivity (NPP) ranged from 0.08%-24.6% and 0.028%-11.2%, respectively. Parameter sensitivity was further analyzed. Model outputs are highly sensitive to the intake rate (i.e. feeding rate of livestock per day), half maximum intake rate, and initial livestock weight. Our results indicate that great uncertainty exists in simulating C graze . We ascribed the major uncertainty to the different process description and poor parameterization. This study calls for more efforts to the comprehensive synthesis of usable dataset, the foundation of a standard observation system and the observe-based inter-comparison to evaluate models, which would facilitate more accurate assessment of C sequestration by pasture ecosystems and lead to better representation in earth system models.

Grazing alters the sensitivity of plant productivity to precipitation in northern temperate grasslands

“…However, much of current understanding of ecosystem sensitivity to precipitation variability is derived from responses of above‐ground net primary productivity (ANPP) to precipitation (Knapp et al, 2017b) and assessed independently of grazing effects (but see Irisarri et al, 2016), despite grazing being the predominant land‐use of grasslands (Asner et al, 2004). As such, understanding ecosystem responses to the interactive effects of co‐occurring factors is critical, and a reliable benchmark is needed for predicting ecosystem sensitivity robustly from ecosystem models (Dangal et al, 2017; Chen et al, 2018; Ma et al, 2020).…”

Explicit Representation of Grazing Activity in a Diagnostic Terrestrial Model: A Data‐Process Combined Scheme

“…The grazing process has recently been explicitly represented in global models (Chang et al, 2013; Luo et al, 2012; Pachzelt et al, 2013). However, large uncertainties still exist in modeling grazing activity (Chen et al, 2018; Fetzel et al, 2017; Zhou et al, 2017). A standard data set seems to be necessary to diagnose and benchmark the models.…”

“…Currently, several models have been incorporated into Global Vegetation Models and plan to be consider in future intermodel comparison projects (Erb et al, 2017; Pongratz et al, 2017). With the increasing model developments and applications, however, great uncertainties were found in both approaches at large scales (Chen et al, 2018; Fetzel et al, 2017). Therefore, a spatial‐explicit diagnostic tool or data set is urgently needed for model evaluation, benchmarking, and future improvement.…”

Identifying key processes and sectors in the interaction between climate and socio-economic systems: a review toward integrating Earth–human systems

“…In addition, the variety of livestock systems in operation globally complicates the discussion surrounding the effects of climate change (Lopez-i-Gelats 2014;Rivera-Ferre et al 2016). Inter-model uncertainty (e.g., what Chen et al (2018) demonstrated regarding carbon sequestration from grazing activity) also remains a problem.…”