Impact of plant root functional traits and associated mycorrhizas on the aggregate stability of a tropical Ferralsol

Elsevier, Geoderma, Volume 312, 15 February 2018
Demenois J., Carriconde F., Bonaventure P., Maeght J.-L., Stokes A., Rey F.
In many tropical regions, such as New Caledonia, soil erosion from anthropogenic activities and subsequent ecological restoration are major issues that require detailed soil and vegetation data for the production of management plans. To determine if some plant species are more useful for stabilizing soil aggregates and thus reducing erodibility, we examined three species endemic to New Caledonia, and measured how root traits and associated mycorrhizas and fungi influenced Ferralsol aggregate stability (MWD). The three species are hosts to different types of mycorrhizas and were: (i) the sedge Costularia arundinacea (Sol. Ex Vahl) Kük., an AMF (arbuscular mycorrhizal fungi) host, (ii) the shrub Tristaniopsis glauca Brongn. & Gris and (iii) the tree Arillastrum gummiferum (Pancher ex Brongn. & Gris) Baill., both of the latter are ectomycorrhizal fungi (ECM) hosts. Fungal abundance, aggregate stability, soil organic carbon (SOC), iron (Fe) and aluminium (Al) sesquioxides were measured in the soil beneath 20 individuals for each species, as well as in 20 control samples of bare soil. Root functional traits including root mass density (RMD), root length density (RLD) and percentage of fine roots were measured on all individuals. Results showed that plant species can significantly influence soil aggregate stability. MWD was greater in soil beneath Costularia characterized by high RMD, RLD, percentage of fine roots and fungal abundance, while MWD in the rhizosphere of Tristaniopsis and Arillastrum was similar to that of bare soil. Fe and Al were very high in all soil samples and are suspected of masking the influence of roots, fungi and SOC on MWD at the scale of isolated ECM-hosts. Therefore, MWD alone would not be a relevant predictor of restoration on such soil and further investigations should be carried out to identify a set of predictors useful for indicating the restoration of degraded soils on ultramafic substrates.