Bayesian method predicts belowground biomass of natural grasslands

ABSTRACT

Belowground biomass accounts for most of the carbon fluxes between biosphere and atmosphere. However, the relative importance of geographical, climatic, vegetation, and soil factors to belowground biomass at the regional scale is not well understood. To improve our understanding and estimations of belowground biomass, we used multilevel regression modeling to estimate the primary productivity of natural grasslands and determine the effects of the above-mentioned factors on belowground biomass. Mean annual precipitation (MAP), longitude, soil bulk density (SB), and soil moisture content (SMC) explained 22.4% (highest density interval, HDI: 12.6–32.5%), 10.5% (HDI: 0.6–20.6%), 10.2% (HDI: 1.9–18.8%), and 13.1% (HDI: 1.5–25.2%) of the variation in regional belowground biomass, respectively. Our results clearly demonstrate that belowground biomass values of ecological communities exhibited the pattern meadow > steppe > desert steppe. MAP was the most important driver of productivity, and SMC was a goodpredictor of variations in productivity at the regional scale. Our results show that multifunctionality indices that appropriately account for the comprehensive responses of the multiple drivers of grassland ecosystems are important at the regional scale.

RÉSUMÉ

La biomasse racinaire compte pour la majorité des flux de carbone entre la biosphère et l’atmosphère. Toutefois, l’importance relative des facteurs liés à la géographie, au climat, à la végétation et au sol dans la production de la biomasse racinaire n’est pas bien comprise. Afin d’améliorer la compréhension et la prédiction de la biomasse racinaire, nous avons utilisé la modélisation par régression multi-niveaux pour estimer la productivité primaire de prairies naturelles et pour déterminer les effets des facteurs susmentionnés sur la biomasse racinaire. Les précipitations annuelles moyennes (MAP), la longitude, la masse volumique du sol (SB) et le contenu en humidité du sol (SMC) expliquaient respectivement 22,4% (intervalle à plus haute densité, HDI: 12,6-32,5%), 10,5% (HDI: 0,6-20,6%), 10,2% (HDI: 1,9-18,8%) et 13,1% (HDI: 1,5-25,2%) de la variation régionale de la biomasse racinaire. Nos résultats montrent clairement que les valeurs de biomasse racinaire des communautés écologiques suit le patron prairie > steppe > steppe désertique. Les MAP avaient l’effet le plus marqué sur la productivité et le SMC était un bon indicateur des variations de productivité à l’échelle régionale. Nos résultats montrent que les indices demultifonctionnalité tenant compte adéquatement des réponses complètes des nombreuses variables influençant les écosystèmes de prairies sont importants à l’échelle régionale.

KEYWORDS:

• Bayesian analysis
• regression
• belowground biomass
• richness

MOTS CLÉS:

• Analyse bayésienne
• régression
• biomasse racinaire
• richesse

Acknowledgements

The study was financially supported by the National Natural Science Foundation of China (41390463, 31260125, 41501094), the National Sci-Tech Support Program of China (2015BAC01B03), the National Sci-Tech Basic Program of China (2014FY210100), the Science and Technology Service Network Initiative of the Chinese Academy of Sciences (KFJ-EW-STS-005), and the Open Project Program of Breeding Base for State Key Laboratory of Land Degradation and Ecological Restoration of North-western China/Key Laboratory for Restoration and Reconstruction of Degraded Ecosystems in North-western China, Ministry of Education (2017KF007).

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [41390463];National Natural Science Foundation of China [41501094];National Natural Science Foundation of China [31260125]; Science and Technology Service Network Initiative of the Chinese Academy of Sciences [KFJ-EW-STS-005];National Sci-Tech Basic Program of China [2014FY210100];Open Project Program of Breeding Base for State Key Laboratory of Land Degradation and Ecological Restoration of North-western China/ Key Laboratory for Restoration and Reconstruction of Degraded Ecosystem in North-western China of Ministry of Education [2017KF007];National Sci-Tech Support Program of China [2015BAC01B03].