Changes in arsenic mobility and speciation across a 2000-year-old paddy soil chronosequence

Rice accumulates arsenic (As) when cultivated under flooded conditions in paddy soils threatening rice yield or its safety for human consumption, depending on As speciation. During long-term paddy use, repeated redox cycles systematically alter soil biogeochemistry and microbiology. In the present study, incubation experiments from a 2000-year-old paddy soil chronosequence revealed that As mobilization and speciation also change with paddy soil age. Younger paddies (≤100 years) showed the highest total As mobilization, with speciation dominated by carcinogenic inorganic oxyarsenic species and highly mobile inorganic thioarsenates. Inorganic thioarsenates formed by a high availability of reduced sulfur (S) due to low concentrations of reducible iron (Fe) and soil organic carbon (SOC). Long-term paddy use (>100 years) resulted in higher microbial activity and SOC, increasing the share of phytotoxic methylated As. Methylated oxyarsenic species are precursors for cytotoxic methylated thioarsenates. Methylated thioarsenates formed in soils of all ages being limited either by the availability of methylated As in young soils or that of reduced-S in older ones. The present study shows that via a linkage of As to the biogeochemistry of Fe, S, and C, paddy soil age can influence the kind and the extent of threat that As poses for rice cultivation.

Soil development following glacier retreat shapes metagenomic and metabolomic functioning associated with asynchronous C and N accumulation

Glacier retreat caused by global warming may result in the variation of soil organic carbon and nutrient cycling. Yet, the dynamic change of soil microbial functional profiles, especially C metabolism-related, with soil development following glacier retreat are still unclear. In the present study, we investigated the soil microbial communities, metagenomic functioning, and metabolomic profiles along the Hailuogou Glacier forefield representing a 120-year chronosequence. The alpha diversity indices of soil bacteria, protozoa and nifH genes showed an upward trend with increased soil ages, and the beta diversity of soil archaea, bacteria, fungi, protozoa, nifH and nirS genes were significantly correlated with soil ages, in which increasing soil C and P while decreased C/N and pH significantly contributed to the differences of soil microbial communities among the analyzed environmental variables. The metagenomic functional genes related to the metabolisms of Glycogen and Cellulosome, Iron Acquisition and Metabolism were significantly decreased with chronosequence, while the utilization of Xylose and Lactate, Potassium Metabolism, Sulfur Metabolism showing an upward trend with soil ages, in which soil C/N ratios and pH were the most influential factors. In addition, soil C and C/N ratios were also significantly correlated to metabolomic compositions, in which the complexity of the metabolite structure increased with soil ages. Our results indicate that glacier retreat may lead to the asynchronous C and N accumulation along the chronosequence, thereby affecting the metagenomic and metabolomic functioning of soil microbial communities related to C metabolisms during soil development following glacier retreat.

Soil microbial co-occurrence networks become less connected with soil development in a high Arctic glacier foreland succession

Classically, ecologists have considered that biota becomes more integrated and interdependent with ecosystem development in primary successional environments. However, recent work on soil microbial communities suggests that there may in fact be no change in network integration over successional time series. Here, we performed a test of this principle by identifying network-level topological features of the soil microbial co-occurrence networks in the primary successional foreland environment of the retreating high-Arctic glacier of Midtre Lovénbreen, Svalbard. Soil was sampled at sites along the foreland of inferred ages 10-90 years since deglaciation. DNA was extracted and amplicon sequenced for 16 s rRNA genes for bacteria and ITS1 region for fungi. Despite the chronologically-related soil pH decline and organic C/N accumulation, analysis on network-level topological features showed network integration did not change with inferred chronological ages, whereas network integration declined with decreasing pH and increasing total organic carbon (TOC) - both factors that can be viewed as an indicator of soil development. We also found that bacteria played a greater role in the network structure than fungi, with all keystone species in the microbial co-occurrence network being bacteria species. Both number and relative abundance of the keystone species were significantly higher when soil pH increased or TOC decreased. It appears that in the more extreme and less productive conditions of early primary succession, integration between members of soil biota into consortia may play a greater role in niche adaptation and survival. Our finding also emphasizes that ecosystem development is not simply a product of time but is influenced by locally heterogeneous factors.

Organic carbon accumulation in the glacier forelands with regard to variability of environmental conditions in different ecogenesis stages of High Arctic ecosystems

Recently deglaciated surfaces of glacier forelands are subjected to a variety of biotic and abiotic factors that lead to continuous soil formation. Until now, no attempt has been taken to analyse multiple factors that might affect soil development in the Arctic forelands. The main aim of this research was to determine the factors that influence soil development in the eight forelands of Svalbard. Moreover, the effects of both habitat type (glacier foreland and mature tundra) and geographical location on environmental variables treated as potential factors influencing soil formation were tested. In 2017, at each location a series of 1 m² plots was established; all 168 plots were investigated in terms of soil properties, spatial data, biological soil crusts (BSCs) properties, percent cover of BSCs and vascular plants. Stepwise multiple linear regression analysis using forward variable selection showed that soil development was significantly associated with six of fifteen analysed factors, i.e. BSC cover, carbon and nitrogen content in BSCs, soil pH, Topographic Wetness Index and foreland location. Two-way analysis of variance followed by Tukey's test revealed significant differences in studied environmental variables between habitat types and studied locations, showing that foreland soils still retain particular initial characters to differentiate them from tundra soil.