Soil microbes play a crucial role in ecosystem function. Here, the effects of soil and water conservation measures on soil microbial community structures, biodiversity, and co-occurrence networks are investigated and compared. We sampled soils at three different depths (0–10 cm, 10–20 cm and 20–40 cm) in a citrus orchard that uses long-term soil and water conservation measures, which includes Bermuda grass strip intercropping (BS), Bermuda grass full coverage (BF), Radish–soybean crop rotation strip intercropping (RS) and clear tillage orchards (CT). Results demonstrated that BS and BF yields a significant increase in bacterial richness and diversity of fungal in soils, while BF contains more beneficial microbial taxa, especially those with degrading and nutrient cycling capabilities. Microbial community structures differed significantly among the applied measures. In addition, co-occurrence networks under BS, BF and RS were more complex and robust than that of CT, and the stability of the network in BF was the highest. Microbial interactive stability and potential interactions in bacterial networks were stronger than those of fungi. The distribution of dominant phyla showed that Chloroflexi and Ascomycota dominated the different soil and water conservation measures. Proteobacteria and Ascomycota are revealed to be keystone species in bacterial networks and fungal networks, respectively, while Proteobacteria was the keystone species in microbial networks. Though the relative abundance of Chloroflexi turned out to have increased among the four measures, the relative abundance for Proteobacteria, Acidobacteria and Actinobacteria all decreased along the soil profile, with Acidobacteria under BS to be an exception. Soils under BS and BF had higher total nitrogen, microbial biomass carbon and organic carbon than CT and RS. Organic carbon(C) and total nitrogen(N) in soil were the major drivers of these bacterial community patterns, while there was no significant correlation between them and fungi. Overall, BF increases soil nutrients and microbial diversity, and also promotes ecological stability and interrelations among microbial taxa that collectively improve soil quality in the citrus orchard studied. Therefore, we recommended BF to be an ideal application for citrus orchards of southeast China.

Degraded soil has a high carbon sink potential. However, the carbon sequestration capacity and efficiency of comprehensive control measures in soil erosion areas are still not fully understood, and this information is essential for evaluating the effects of adopted restoration measures. The objective of this study was to determine the restoration of soil organic carbon and active carbon components under the impact of soil erosion measures and reforestation following different restoration ages. A small watershed with four typical restored plots following the same control measures (combination measures with horizontal bamboo burl-groove + replanting trees, shrubs and grasses) but different restoration ages (4 years, 14 years, 24 years and 34 years) and two reference plots (bare land (carbon-depleted) and nearby undisturbed forest (carbon-enriched)) in subtropical China was studied. The results showed that the soil organic carbon contents at a 1 m soil depth and the dissolved organic carbon and microbial biomass carbon concentrations in the upper 60 cm of soils of the four restored lands were higher than those in the bare land. Furthermore, the restored lands of 4 years, 14 years, 24 years and 34 years had soil organic carbon stocks in the 1 m soil depth of 22.83 t hm⁻², 21.87 t hm⁻², 32.77 t hm⁻² and 39.65 t hm⁻², respectively, which were higher than the bare land value of 19.86 t hm⁻² but lower than the undisturbed forestland value of 75.90 t hm⁻². The restored forestlands of 34 years of ecological restoration also had a high potential of being a soil organic carbon sink. Compared to the bare land, the restored lands of 4 years, 14 years, 24 years and 34 years had soil organic carbon sequestration capacities of 2.97 t hm⁻², 2.01 t hm⁻², 12.91 t hm⁻² and 19.79 t hm⁻², respectively, and had soil organic carbon sequestration rates of 0.74 t hm⁻² a⁻¹, 0.14 t hm⁻² a⁻¹, 0.54 t hm⁻² a⁻¹ and 0.58 t hm⁻² a⁻¹, respectively. Our results indicated that the combined measures of horizontal bamboo burl-groove and revegetation could greatly increase carbon sequestration and accumulation. Suitable microtopography modification and continuous organic carbon sources from vegetation are two main factors influencing soil organic carbon recovery. Combination measures, which can provide suitable topography and a continuous soil organic carbon supply, could be considered in treating degraded soils caused by water erosion in red soil areas.

Atmospheric nitrogen (N) and acid deposition have become global environmental issues and are likely to alter soil respiration (R_s); the largest CO₂ source is from soil to the atmosphere. However, to date, much less attention has been focused on the interactive effects and underlying mechanisms of N and acid deposition on R_s, especially for ecosystems that are simultaneously subjected to elevated levels of deposition of both N and acid. Here, to examine the effects of N addition, acid addition, and their interactions with R_s, we conducted a two-way factorial N addition (control, CK; 60 kg N ha⁻¹ a⁻¹, LN; 120 kg N ha⁻¹ a⁻¹, HN) and acid addition (control, CK; pH 4.5, LA; pH 2.5, HA) field experiment in a subtropical plantation in China. Our results showed the following: (1) During the one-year observation period, the seasonal dynamics of R_s presented a single peak curve model, which was closely related to the surface soil temperature. (2) The simulated N deposition and acid deposition significantly decreased the R_s in the subtropical plantation. Compared to the CK plots, the LN and HN treatments reduced the annual mean values of R_s by 41% and 56%, and the annual mean values of R_s were inhibited by 26% and 31% in the LA and HA plots. The inhibition of N application on R_s was stronger than that of the simulated acid deposition. (3) Significant interactions between N addition and acid addition on R_s were detected, and R_s was significantly inhibited under four co-addition treatments. (4) The underlying mechanism and main reason for the responses of R_s to simulated N and acid deposition in this study might be the inhibition of soil microbial biomass and soil enzyme activity due to soil acidification under increased N and acid input.

Rainfall intensity (RI) and land cover type are two important factors that affect soil erosion and thus the transfer and loss of soil organic carbon (SOC). However, the in situ quantitative monitoring of SOC loss under natural rainfall and various land cover types restored on eroded lands has not been thoroughly examined. In order to further study the effects of rainfall changes and vegetation types on SOC loss in the red soil region of Southern China, the Jiangxi Eco-Science Park of Soil and Water Conservation in De’an County, Jiangxi Province, was taken as the research object. Considering natural rainfall and based on the long-term field in situ monitoring of rainfall and runoff and sediment data, we studied the effects of three land cover types (bare land, orchards, and grass cover) on surface runoff, sediment production, and SOC loss in relation to 1 hour of RI during natural rainfall in the red soil region of Southern China during rainy seasons of 2020 and 2021 (March to August). Compared with bare land plots, the orchard and grass cover plots had surface runoff reductions of 67% and 98%, respectively, and sediment reductions of 79% and 99% over the two rainy seasons, respectively. With an increasing RI over 1 hour, total SOC loss increased for each of the three land cover types. More SOC loss was associated with sediments, and the enrichment ratio of SOC in the sediments (ER_oc) decreased significantly. The ER_oc values decreased in the following order: bare land (1.23) > orchard (1.08) > grass cover (0.81). Bare land exhibited the highest proportion of SOC associated with sediment in the total SOC loss (P_s), at 68.69%, followed by the orchard plots, at 55.02%, and then the grass cover plots at 49.24%. With the transfer of land cover type from bare land to orchard and to grass cover (decreased soil loss intensity, SLI), more SOC was lost associated with runoff in the form of dissolved organic carbon (DOC); the values of ER_oc and organic carbon loss intensity (CLI) also decreased significantly. These findings are crucial to improving our understanding of the regulatory mechanisms of rainfall changes and land cover types on SOC loss during soil erosion.