Abstract:

Aims Glomalin related soil protein (GRSP) is a type of glycoprotein produced by arbuscular mycorrhizal fungi, playing a vital role in the regulation of soil physical structure and soil carbon stability. Previous studies have demonstrated the effects of soil physico-chemical properties on GRSP accumulation. To date, there is a lack of systematic research on how plant species composition of diversity and dominance affecting GRSP accumulation.

Methods A total of 72 plots in Harbin Experimental Forest Farm of Northeast Forestry University were surveyed in this study. The soil profile at 1 m depth was divided into five layers for soil sampling. Tree species diversity indices (Richness, Shannon-Wiener index, Simpson index, and evenness index) were calculated. The importance values (IVs) were calculated for dominant tree species according to the relative abundance, species frequency, and cross-area ratio to total stem cross-area. Soil physico-chemical properties were measured, including soil organic carbon (SOC), total nitrogen (TN), soil moisture content (MC), electrical conductivity (EC), and pH. All soil samples were assayed for easily extractable GRSP (EEG), total GRSP (TG), EEG/TG, EEG-C/SOC, and TG-C/SOC. The association between GRSP traits and soil physicochemical properties, IVs of dominant trees, and tree species diversity were ordinated using redundancy ordination (RDA) and variation partitioning analysis.

Results (1) TG and EEG correlated positively with SOC throughout the entire soil profile. The same was true with TN and MC at some soil depths, but the correlation was negative with EC and pH values. (2) In most soil layers, the greater EEG accumulation with higher EEG-C/SOC (proportion of C in EEG to SOC) and EEG/TG accompanied with the higher IV for Pinus tabulaeformisvar. mukdensis, but smaller IVs for Phellodendron amurense and Ulmus pumila. In contrast, the greater TG accumulation aligned with the higher IVs in Juglans mandshurica, P. amurense, and U. pumila but lower IVs in P. tabulaeformis var. mukdensis, Larix gmelinii and Pinus sylvestris var. mongolica. (3) The plant Simpson index, Shannon-Wiener index, and species richness correlated negatively (P < 0.05) with EEG-C/SOC, EEG-N/TN (proportion of N in EEG to TN), TG-C/SOC (proportion of C in TG to SOC), and TG-N/TN (proportion of N in TG to TN), and these indices had no relationship with EEG, TG, and EEG/TG. In addition, EEG/TG and EEG-N/TN correlated positively (P < 0.05) with plant evenness. These trends were similar across different soil layers in the 1 m profile. (4) RDA ordination and variation partitioning showed that biotic factors (plant diversity, IV of dominant species) explained 20.2% of GRSP variation, and soil factors explained 7.8%. Of the biotic factors, the IVs of the dominant plant species accounted for 16.4%, while species diversity only explained 0.4%. The evergreen coniferous species (eg, P. tabulaeformis var. mukdensisand P. sylvestris var. mongolica), rather than the broad-leaved species, usually accompanied the higher GRSP accumulation and contribution to SOC and TN (P < 0.01). The underlying mechanism contributing this difference possibly related to mycorrhizal symbiosis of trees, i.e., TG related negatively to ectomycorrhizal tree IVs, but related positively to arbuscular mycorrhizal tree IVs.

Conclusion Our findings highlighted that GRSP amounts and its contribution to soil carbon and nitrogen are mainly regulated by tree dominance rather than their diversity traits. This data underscored that GRSP-oriented soil management and evaluation can be promoted by adjusting the dominant species, a much simple indicator than species diversity indices.

Key words: glomalin-related soil protein, species diversity, importance value of dominant species, soil physico- chemical, redundancy ordination and variation partitioning