Soil has been considered as one of the most endangered resources on earth, which highly influences environmental health for human well-being. Soil contains a substantial part of terrestrial carbon (C) and plays an important role in terrestrial C cycling. Vegetation type has long been recognized among the important soil forming factors, and tree species has been proved to be of great importance for soil C stocks and even forms. Therefore, forest management such as selection of appropriate tree species for mitigating further climate change requires a better understanding of factors affecting soil biota that is an integral part of soil ecosystem and plays an essential role in ecosystem functioning. However, till now, quantitative estimates of tree species effects on soil C stocks, soil fauna communities, litter decomposition, and their interactions are still scarce, and the scientific basis for targeted use of tree species to maximize sequestration of soil C following forest management or afforestation under further climate change scenario is also limited.
In paper I, we quantitatively synthesized 850 observations from field studies that were conducted in a common garden or monoculture plantations to assess how tree species type (broadleaf vs. conifer), mycorrhizal association (arbuscular mycorrhizal (AM) vs. ectomycorrhizal (ECM)), and N-fixing ability (N-fixing vs. non-N-fixing), directly and indirectly, affect topsoil (with a median depth of 10 cm) C concentration and stock, and how such effects were influenced by environmental factors such as geographical location and climate.We found that (1) tree species type, mycorrhizal association, and N-fixing ability were all important factors affecting soil C, with lower forest floor C stocks under broadleaved (44%), AM (39%), or N-fixing (28%) trees respectively, but higher mineral soil C concentration (11%, 22%, and 156 %) and stock (9%, 10%, and 6%) under broadleaved, AM, and N-fixing trees respectively; (2) tree species type, mycorrhizal association, and N-fixing ability affected forest floor C stock and mineral soil C concentration and stock directly or indirectly through impacting soil properties such as microbial biomass C and nitrogen; (3) tree species effects on mineral soil C concentration and stock were mediated by latitude, MAT, MAP, and forest stand age.
In paper II, we evaluated the effects of tree species and their associated two types of mycorrhiza, namely arbuscular mycorrhiza (AM) and ectomycorrhiza (ECM), on soil fauna taxonomic and functional community composition using diversity, abundance, and biomass as proxies. We found that (1) soil fauna biomass as well as taxonomic and functional diversity and evenness were all significantly higher under AM tree species compared with ECM trees except for lime, which showed a pattern similar to that for AM tree species, but soil fauna abundance was higher under ECM trees; (2) the abundance and biomass were higher under AM tree species for macrofauna as a whole, but lower under AM tree species for mesofauna; (3) although tree species and mycorrhizal association significantly affected the abundance and biomass of different taxonomic or functional groups directly and/or indirectly through moderating resource availabilities of litter, forest floor, and soil, tree species effects varied among different taxonomic and functional groups; and (4) among the variables representing resource availability, litter N concentration and lignin:N ratio, forest floor Ca and Mg concentrations, and soil pH, Ca, and Mg were the major moderators controlling soil fauna abundance, biomass, and diversity.
In paper III, we conducted a field experiment using litterbags of different mesh sizes with foliar litter of two tree species associated with AM and three tree species associated with ECM fungi in six Danish common garden sites to assess how tree species and mycorrhizal association affect the effects of meso- and macrofauna on litter decomposition. We found that (1) litter species, mycorrhizal association, litterbag mesh size, and their interactions all significantly affected litter decomposition; (2) tree species do not only directly affected litter decomposition, but also indirectly through regulating microbial and soil fauna communities, with an overall pattern of higher mass loss and decomposition rate of litter from AM than ECM tree species regardless of mesh size; (3) access of meso- and macrofauna significantly enhanced litter decomposition, increasing the decomposition rate of AM and ECM tree species by 255% and 92% by mesofauna and 265% and 108% by all soil fauna, respectively, (4) except for lime litter (associated with ECM fungi) that showed a similar pattern with litter from AM tree species; and (5) litter with high Ca concentration stimulated soil fauna effects on litter decomposition, while litter with high C:N and lignin:N ratios was less or not at all stimulated by meso- and macrofauna access.
Altogether, our results revealed how tree species and their specific traits of mycorrhizal association, species type, and N-fixing ability influence soil C stocks at the global scale, and how tree species and mycorrhizal association shape soil fauna communities across or within different taxonomic and functional groups and affect the differential contributions of meso- and macrofauna to litter decomposition. Our findings would be useful to inform tree species selection in forest management or afforestation aiming to sequester more atmospheric C in soils, to predict the effects of forest management on the soil fauna community composition of abundance, biomass and diversity, and to better understand the importance of plant-fauna-soil interactions following forest management or afforestation in C sequestration and climate change mitigation.