Search

Rhizoliths are the products of mineralization, petrification, or fossilization around and/or within plant roots. Among them, carbonate rhizoliths are the most common. Pristine carbonate rhizoliths with co-existing plant root relicts in the Tengeri Desert, NW China were studied, with a combination of intensive field observations and laboratory methods such as microscopy, scanning electronic microscopy, energy dispersive X-ray spectra, radiocarbon dating, and isotope mass spectrometer. The field observations revealed that the pristine rhizoliths are only present at the sites where Artemisia sphaerocephala Krasch are growing i.e. in swales among sand dunes. Soil moisture of the swales is the main controlling factor of rhizoliths formation. It is in turn affected by the soil physical properties, landscape position, and climate variability, consistent with the locations of sampled rhizoliths in the swales where Artemisia plants are mostly distributed. The 14C AMS dating indicated that the rhizoliths are much older (4000–5000 years) than their co-existing modern plant root relicts in agreement with field observations. Morphological, mineralogical and isotopic analyses revealed that carbon sources used for the rhizoliths formation were partially derived from decomposing plant roots but with significant contribution from dissolution of lithogenic carbonates. The calcium sources were suggested to be the in situ weathering of minerals (mostly lithogenic carbonates) and the pressure-dissolution of carbonates. Enough CO2 from the root decomposition have triggered carbonate accumulation around the root to form rhizoliths. Other minor chemical components of the root are S, N, P, which produce acidic water with the negative ions of SO42−, NO3–, PO43-, have also favored acidic soil environment and enhanced carbonate dissolution and mineral weathering. Redox environment around Artemisia roots were also observed to be a key factor for the pristine rhizolith formation. The pristine rhizoliths were preferentially formed in semi-closed redox condition with water nearly always available at intermediate depths. In addition, they were formed through carbonate epidiagenesis in shallow soils of the desert. Altogether, our results showed that the formation of the pristine rhizoliths was affected by the combination of several environmental factors. This led us to propose a conceptual model of rhizolith formation in desert soils.

Atmospheric carbon dioxide (CO2) absorption by desert soils has received increasing interests in recent years; however, the underlying physical and chemical mechanisms are not commonly acceptable. Here, we hypothesised that autotrophic carbon fixation of soil microbes contributes to this process. To test this postulate, we investigated the genomic and biochemical potential of autotrophic carbon fixation and traced atmospheric autotrophic carbon absorption using metagenomics and 13CO2 labelling approaches in the Mu Us Desert in northern China. More than 30000 genes involved in the six carbon fixation pathways (approximately 2% of the assembled metagenomes, in relative abundance) were found in the metagenome of the desert soil, and the relative abundance for genes encoding for the reductive citrate cycle was the highest among the six pathways. The main autotrophic microbes employing the six pathways belong to Actinobacteria, Proteobacteria, Chloroflexi, Acidobacteria, Gemmatimonadetes, Firmicutes, Thaumarchaeota, Nitrospirae, Planctomycetes, and Bacteroidetes, respectively. 13CO2 labelling revealed that the contents of microbially incorporated soil organic carbon (13C-SOC) and dissolved organic carbon were 0.572–1.45 mg kg−1 and 0.290–0.914 mg kg−1, respectively. Further, the 13C-SOC correlated with the relative abundance of genes of the total six pathways, reductive citrate cycle, 3-hydroxypropionate bi-cycle, and reductive acetyl-CoA pathway. Another in situ labelling experiment showed a significant increase in δ13C of SOC, and the incorporated carbon (13C) in SOC accounted for 3.85% of total atmospheric carbon absorption. These results showed that desert soil microbes containing genetic potential for autotrophic carbon fixation spread over a broad taxonomic range, and incorporated atmospheric carbon into organic components, which contributed to atmospheric carbon absorption. Although more research is required to accurately evaluate the portions of autotrophic carbon in the amount of atmospheric carbon absorption, the biotransformation of carbon from the atmosphere to soil via autotrophic carbon fixation represents a microbial pathway for persistent atmospheric CO2 absorption in desert soils, and further implicates an important carbon biochemical cycle for carbon accumulation in oligotrophic desert soils.

Modifications of rainfall patterns are expected to accompany global climate changes. It has been suggested that in dry regions changes in soil carbon emission induced by precipitation will affect soil carbon storage and atmospheric CO 2 concentration. However, our understanding of the responses of soil carbon emission [often as soil respiration ( R S )] to rainfall pulses is still limited regarding changes in soil respiration components [heterotrophic respiration ( R H ) and autotrophic respiration ( R A )] and under different precipitation patterns in arid and semiarid ecosystems. To evaluate the variations in soil carbon emission in response to rainfall pulses, we measured R S and its components in situ before/after precipitation in the Mu Us Desert, China. Rates of R S and its components were significantly enhanced by rainfall pulses, but gradually reverted thereafter. Moreover, the magnitudes of diel hysteresis for R S , R H , and R A with respect to soil temperature ( T S ) were modified by precipitation, and the effects of rainfall pulses on R S were influenced by antecedent soil water availability. In addition, the ratio of respiration components was changed by individual precipitation events, with an increase in the amount of each rainfall pulse causing a decrease in the proportion of R H to R S . Our results indicate that rainfall pulses in desert ecosystems have a major impact on soil carbon emission via changes in the magnitude and ratio of respiration components. We accordingly suggest that greater carbon emission and alterations in respiration components may occur with more extreme precipitation in desert ecosystems.

Anomalous CO2 absorption by soils in drylands has been reported worldwide. It has been reported that atmospheric carbon can first be converted into soil dissolved inorganic carbon, and then be carried and rapidly sequestered in groundwater through irrigation (artificial water input). Nevertheless, the deposition rate may be inapplicable in drylands without artificial water input, and the fate of the atmospheric carbon may also be different. Here, we assessed the amount of the local rainfall, evaluated the elapsed time of the natural water infiltration and the feldspar weathering-induced carbon consumption rate, and analysed the characteristics of soil inorganic carbon in the Mu Us Desert, northwest China, where we have previously studied abiotic atmospheric carbon absorption and the soil water input mainly derives from precipitation. The results showed that, even with extreme precipitation, the inputted water could not rapidly transport carbon to the first aquiclude, indicating that natural water input cannot deposit the majority of the atmospheric carbon absorbed by soil into groundwater. Further analysis indicated that feldspar weathering might be able to rapidly consume most of the atmospheric carbon. The consumed carbon would be converted into pedogenic carbonate and conserved in subsoil due to the transportation of inputted water. These results can provide a basis for determining the fate of atmospheric carbon absorbed by soils in drylands.

This study aimed at assessing whether patch type (i.e., under-shrub soil patch and inter-shrub soil patch) has an effect on soil microbes and how different shrub species altered the soil microbes through understanding soil microbial activity, biomass, and community structure. We characterized the soil microbes in under-shrub and inter-shrub soil patches in three shrublands (Artemisia ordosica, Salix psammophila, and Caragana microphylla), respectively, in the Mu Us Desert, China, using microbial activity indicators, chloroform fumigation-extraction analysis, and high-throughput 16S rRNA gene sequencing. Members of the phyla Proteobacteria, Actinobacteria, Acidobacteria, Planctomycetes, Bacteroidetes, Chloroflexi, Firmicutes, and Gemmatimonadetes were dominant. Inter-shrub soil patch differed from under-shrub soil patch in soil bacterial composition, microbial enzyme activity, and biomass, but not in diversity. Soil collected in A. ordosica shrubland exhibited the highest microbial enzyme activity, biomass, and diversity. Shrub species had significant effects on community structure, primarily the relative abundance of Proteobacteria, Actinobacteria, and Bacteroidetes. The results indicated that both shrub species and patch type had effects on soil microbial communities. In shrub-dominated desert ecosystems, spatial heterogeneity of soil nutrients and moisture might not be the main factors underlying variations in bacterial diversity. The different compositions of microbial communities in various shrublands provide a foundation for further research into the mechanisms of soil organic carbon accumulation.

It has been reported that soils in drylands can absorb CO2, although the patterns and mechanisms of such a process remain under debate. To address this, we investigated the relationships between soil CO2 flux and meteorological factors and soil properties in Northwest China to reveal the reasons for "anomalous" soil CO2 flux in a desert ecosystem. Soil CO2 flux increased significantly and exponentially with surficial turbulence at the diel scale under dry conditions (P

Revegetation with xerophilous shrubs is an effective approach to combat desertification in northwestern China; however, evaluation studies on fine-root properties of shrubs and soil organic carbon are limited. To gain a better understanding of revegetation practices, we investigated the vertical distribution of fine-root biomass, necromass, production, and effect on soil organic carbon (SOC) content in three shrub species (Salix psammophila, Hedysarum mongolicum, and Artemisia ordosica). In addition, we also estimated the fine root decomposition rate with litterbag techniques. The results showed that revegetation practices resulted in a significant increase in SOC content. Over a 10 year period of revegetation, the SOC content in S. psammophila, H. mongolicum, and A. ordosica plots increased by 0.87, 1.07, and 1.82 times, respectively, more than that in bare-land plot. Increase in total SOC content was mainly due to increase in light-fraction SOC, except for the A. ordosica plot. Variations in the short-term increase of SOC content after revegetations with the three shrubs on sand land might be explained by fine root decomposition rates, at least in part. A. ordosica may be a better species for SOC accumulation and sequestration in the study site. Additionally, fine-root biomass and production were not associated with more SOC content increase in shrub plots. The results suggest the mechanism of SOC accumulation and sequestration differed among shrub plots and highlight the effectiveness of different shrub species as revegetation materials in terms of SOC accumulation and sequestration.

Fine root (

It is generally considered that desert ecosystems release CO2 to the atmosphere, but recent studies in drylands have shown that the soil can absorb CO2 abiotically. However, the mechanisms and exact location of abiotic carbon absorption remain unclear. Here, we used soil sterilization, 13CO2 addition, and detection methods to trace 13C in the soil of the Mu Us Desert, northern China. After 13CO2 addition, a large amount of 13CO2 was absorbed by the sterilised soil, and 13C was found enriched both in the soil gaseous phase and dissolved inorganic carbon (DIC). Further analysis indicated that about 79.45% of the total 13C absorbed by the soil was trapped in DIC, while the amount of 13C in the soil gaseous phase accounted for only 0.22% of the total absorbed 13C. However, about 20.33% of the total absorbed 13C remained undetected. Our results suggest that carbonate dissolution might occur predominately, and the soil liquid phase might trap the majority of abiotically absorbed carbon. It is possible that the trapped carbon in the soil liquid phase leaches into the groundwater; however, further studies are required to support this hypothesis.