Relationships of decomposition to phylogenetic isolation belowground, PIB
PIB mainly related to the lower C loss at 8 months in analyses only accounting for PIA and PIB (Fig. 2). This relationship was partly direct, i.e. the abiotic trajectory T3, and was replaceable by tree phenological effects on microclimate, as predicted by hypotheses PIA7 and PIA8. Partly this relationship appeared to be mediated by decomposers (trajectories T1 and T2). First, PIB increased Acari abundance after 8 months, which in turn related to strongly decreased C loss (albeit the PIB*Acari interaction had the inverse effect). Second, PIB rendered the effect of fungal diversity on C loss more negative. These relationships are partly consistent with predictions of PIB1 and PIB3, i.e. a naiveté of decomposers in or transfer of toxins from the ambient phylogenetically distant litter (Table 1, consistent with decomposability of functionally uniform litter; Grossmann et al . 2020). At 14 months PIB did not relate to C loss (in analyses only accounting for PIA and PIB). Nevertheless, at 14 months PIB decreased Collembola abundance which in turn related positively to C loss. Possibly, this potential negative Collembola-mediated effect of PIB contributed to the negative effect of PIB on 13C after 14 months. Overall, there appears to be a temporal phylogenetic home-field advantage where C loss is stronger in a “home litter” produced by closely related neighbourhood (Aponte et al . 2012). As a consequence, oaks may have difficulties to penetrate into a phylogenetically distant neighbourhood.
Regarding mass loss and N loss, litter mostly decomposed equally well when exposed in a phylogenetically closely or distantly related belowground neighbourhood. Processes that decrease decomposition were likely compensated by such that increase decomposition, and we could indeed identify the corresponding opposite relationships when comparing different trajectories. Specifically, PIB directly increased mass loss and decreased N loss after 8 months, while decomposer-mediated effects were opposite: PIB decreased the effect of Collembola or Acari on mass loss, and PIB decreased microbial biomass itself decreasing N loss. Moreover, PIB might also have relatively little effect on mass loss, as in our system high PIB often corresponds to a gymnosperm-dominated neighbourhood with highly recalcitrant litter (Kaneko & Salamanca 1999; Cornwell et al . 2008; Berendse & Scheffer 2009). Being “trained” in decomposing such recalcitrant litter, decomposers in a high PIB environment might efficiently reduce the mass of any litter, with little preference for their “home” litter (Milcu & Manning 2011; Wallenstein et al . 2013). Finally, mass loss might be maintained in a high PIB environment by increased enzymatic activities of ectomycorrhizal fungi (Yguel et al . 2014; Martin Schädler & Daniel J. Ballhorn 2016). Overall, despite constancy in mass loss and N loss among low and high PIB neighbourhoods, the trajectories of decomposition have become different, and there is a “home-field functioning” (Vivanco & Austin 2008; Ayres et al . 2009). Shift in belowground trajectories of functioning while maintaining overall rate of functioning suggests redundancy among different trajectories of decomposition, and a flexibility to shift in neighbourhood. Nevertheless, different trajectories may depend on different litter traits and penetration into a phylogenetically distantly related neighbourhood might impose new selection pressures on litter traits (Guénon et al . 2017).