UC Berkeley

Soils store more carbon (C) than the atmosphere and biosphere combined, yet the fundamental mechanisms that regulate this vast pool of C remain elusive. Soil C and associated organic molecules are critical to soil fertility, water holding capacity, and the global C balance. Plants, particularly plant roots, are the primary source of C in soil. The fate of that C depends largely on the composition and activity of the soil microbial community, which transforms and cycles C. However, when C associates with soil minerals or is isolated in soil aggregates, C may be physically protected. Mineral-associated C can be thousands of years old, but mineral-associated C can also rapidly turnover. We examined how interactions between plants, soil microbial communities, and minerals control the persistence of C in soil.

Plants release C into soil as rhizodeposits during plant growth. After plants die, detrital litter is incorporated into soil organic matter (OM). The rhizosphere, or zone of root influence, is a zone of dynamic release of low molecular weight C compounds and a hotspot of microbial activity. The detritosphere, which we define as soil with dead plant litter, has more chemically complex C substrates and a microbial community distinct from the rhizosphere. A typical mineral in planted surface soil likely experiences times when a root grows past, and after the plant senesces, the surrounding soil shifts from a rhizosphere zone to a detritosphere zone. In ecosystems with Mediterranean-type climates with distinct wet growing seasons and dry seasons with little plant growth, shifts from plant growth to plant decay are largely seasonal. The impact of oscillation between a rhizosphere-dominated system and a detritosphere-dominated system on mineral-associated C chemistry and persistence is unknown. We designed two experiments to compare between a rhizosphere-dominated system and a detritosphere-dominated system, with controls allowing comparison to bulk soil with no added plant C.

In both the rhizosphere and detritosphere, mineral type controlled total C accumulation. We placed quartz, ferrihydrite-coated quartz, and kaolinite specimen minerals in soil microcosms with 13C-labeled Avena barbata rhizosphere soil, 13C-labeled A. barbata litter detritosphere soil, and unlabeled bulk soil. These specimen minerals represent a range of surface reactivity and surface area, and had no detectable C at the start of the experiments. We also used a density-gradient centrifugation method to separate native minerals from the soil that already had associated C from the field. The specimen minerals provided a snapshot of a single season of plant growth and decay, while the native minerals were a natural analog to minerals in the field. All specimen minerals accumulated C, but in the rhizosphere, the ferrihydrite accumulated the most C, while in the detritosphere, the kaolinite accumulated the most C by mineral mass. Mineral type mattered, but so did the C source. It was the interaction between the two that controlled the total quantity of C.

Carbon association with minerals was dynamic. Unlike the specimen minerals, native minerals did not significantly accumulate or lose total C in either the rhizosphere or detritosphere soil. However, we saw evidence for C exchange with these minerals, as well as with the specimen minerals, by comparing trends in C accumulation with 13C enrichment. We detected no significant difference in total C between minerals incubated in the rhizosphere compared with those incubated in bulk soil, and between minerals incubated in the detritosphere compared with those incubated in bulk soil. In the rhizosphere, kaolinite minerals did not accumulate C after 1 month incubation, but the percent root-derived 13C increased after 1 month. This suggests dynamic C exchange. Other evidence for dynamic C exchange in the rhizosphere were rapid changes in mineral-associated C chemistry and exchange of old C for 13C labeled root-derived C on the native minerals, despite no significant change in total C. Based on a mixing model that determined what fraction of C on the minerals was from the growing roots, we calculated that C turnover on native minerals could be 6.0 mg C per g mineral per year or greater. In the detritosphere, we also see evidence for dynamic C exchange, but to a much lesser degree than in the rhizosphere. Based on a mixing model that determined what fraction of C on the minerals was from added litter, we calculate that C turnover on native minerals could be 0.2 mg C per g mineral per year or greater. While there was no significant treatment effect on total C, treatment did impact mineral-associated C chemistry.

Both C source (rhizosphere, detritosphere, or bulk soil) and mineral type interacted to define the composition of mineral-associated C. This was particularly evident at the microscale, which we analyzed with a combined NanoSIMS-STXM/NEXAFS approach. We also found that mineral size, morphology, and ability to form microaggregates, in addition to mineral surface reactivity, influenced mineral-associated C chemistry. In the detritosphere, microaggregate formation may play a particularly important role in protecting litter that is only marginally degraded, including litter-derived C compounds such as phenolic C. We found that a portion of all mineral-associated C was lipids.

Lipid classes varied by both mineral type and C source. Lipidomic analysis suggested that many of these lipids were microbial, particularly fungal storage lipids and microbial membrane lipids. Evidence suggested that the lipid-rich fungal order Mucorales and bacterial order Streptomyces may play an important role in mineral-associated lipid formation. In the rhizosphere, fungal hyphae, likely arbuscular mycorrhizal fungi (AMF), acted as a conduit of root-derived C to the minerals, accounting for a significant fraction of total 13C associated with the minerals.

We found that it was the interaction of plant C source, microbial community, and mineral type that ultimately controlled the fate and potential persistence of mineral-associated C. While we observed similar trends in the rhizosphere and detritosphere, such as the importance of mineral type and evidence of dynamic C association, we also found important distinctions between the two systems. In the rhizosphere, we see evidence for rapid exchange of C with minerals in the presence of a growing root. Here, minerals with high surface reactivity, such as ferrihydrite, may accumulate more C because they lose less C. Other minerals with lower surface reactivity may lose more C due to microbial degradation and reaction with organic acids from the root. While adjacent to a growing root, the chemistry of mineral-associated C shifts to more oxidized C characteristic of microbially-processed compounds, reflecting the high microbial activity of the rhizosphere. In the detritosphere, C exchange with minerals appears slower. Aggregation was important in determining both total C and chemistry, particularly for kaolinite. If a new root grew by a mineral, we expect that any C not strongly bound chemically or physically through aggregation may be more likely to exchange with new root-derived C. Certain microbial orders may play an important role in impacting the fate of C. AMF, as conduits of root-derived C to mineral, and lipid-rich microbes such as the fungal order Mucorales and the bacterial order Streptomyces, show promise to enhance C association and persistence on minerals. The temporal and spatial oscillation between bulk, rhizosphere, and detritosphere soil in natural surface soil systems necessitates a holistic view of C contribution from different sources, and the resulting microbe-mineral interactions that control C fate and persistence.