Introduction
There is a tendency of species to retain the ancestral niche of their lineage, i.e. niche conservatism, and different niches may hence be occupied by different lineages (Crisp & Cook 2012). Niche conservatism is the necessary condition for related species to coexist locally (Prinzing et al . 2016, 2017). However, if environmental conditions change rapidly and organisms are sessile and long lived then niche conservatism may produce local neighbourhoods that are distantly related, such as a tree immigrating into a zone that recently became abiotically suitable but is still occupied by distant relatives. This and other processes result in phylogenetic isolation between a focal organism and its neighbours (Cavender-Bares et al . 2009). Coexistence with distant relatives may change species interactions to the advantage or the disadvantage of the focal individual (Webb et al . 2002; Yguel et al . 2011; Verdú et al . 2012; Gerholdet al . 2015; Prinzing et al . 2017). However, it remains unknown what are the consequences of coexistence with distant relatives for the plant’s contribution to ecosystem functioning, such as the recycling of its own carbon (C) and nitrogen (N) through decomposition. For sessile, big, long lived organisms like trees that grow and germinate in their own litter such an ecosystem feedback would be essential for maintenance and reproduction - and thereby the capacity of species to track conserved niches under rapid environmental change (Ackerly 2003).
Plant litter decomposition is an important and complex ecosystem function depending on multiple abiotic (e.g. temperature and moisture conditions), biotic factors (i.e. plants and soil organisms) and their interactions (Coûteaux et al . 1995; García-Palacios et al . 2013; Santonja et al . 2015). This process is governed, firstly, by the quantity and quality (i.e.physico-chemical characteristics) of litter that was produced by plants, providing microhabitats and food resources for soil organisms. These soil organisms are then the main actors of the decomposition process controlled by i) the abundance and diversity of soil organisms (Fig. 1a, Trajectory T1, Cornwell & Weedon 2014) and ii) the efficiency by which these soil organisms decompose a given litter (Fig. 1a, Trajectory T2). Even though the process is largely driven by soil organisms, a part of the decomposition process is abiotically driven resulting from leaching losses of mineral ions and small organic molecules (Fig. 1a, Trajectory T3). These different trajectories will leave different statistical signatures (Table 1). Below, we will hypothesize how phylogenetic isolation of a focal tree might operate on each of these trajectories.
We hypothesize that phylogenetic isolation of a focal tree from its neighbors influences the decomposition of the tree’s litter due to aboveground mechanisms, i.e. mechanisms changing litter quality (Fig. 1b). Specifically, phylogenetic isolation aboveground (PIA) strongly reduces phytophagy of the focal tree (Yguel et al . 2011) and phytophagy is closely linked to the leaf litter quality (Schädleret al . 2003). For example, phytophagy can induce the production of secondary metabolites such as phenolics (Schultz & Baldwin 1982; Karban & Myers 1989; Coley & Barone 1996; Strauss & Agrawal 1999). These compounds negatively affect abundance and efficiency of decomposer organisms but also the abiotic decomposition of litter (Trajectories T1 and T2 in hypothesis PIA1, and PIA2 in Table 1). On the other hand, phytophagy can increase the leaf damage and in turn increase the accessibility of leaf litter for soil decomposers (PIA3, Cárdenas & Dangles 2012), as well as for moisture and desiccation (PIA4, Ritchieet al . 1998). Finally, phylogenetic isolation from neighbours may also impose a stress on the focal trees due to unfavorable competitive or microclimatic conditions (e.g. lower soil pH or moisture conditions). Such stress may trigger increased investment into, again, phenolics and/or carbon-rich compounds (Fernandez et al . 2016), impacting decomposers numbers and efficiency as well as abiotic decomposition of litter (PIA5 and PIA6).
We also hypothesize that phylogenetic isolation of a focal tree from its neighbors influences decomposition of the tree’s litter due to belowground mechanisms (Fig. 1b). On the one hand, a phylogenetically distant neighborhood produces a litter that is dominated by distantly related plant lineages and may select for soil biota particularly capable of decomposing this dominant litter (Ayres et al . 2009; Austin et al . 2014; Pan et al . 2015; Cheng & Yu 2020), but naïve for decomposing the litter of the focal tree. Decomposition of the focal tree’s litter might hence suffer from a phylogenetic away-field disadvantage (PIB1 in Table 1), a so far untested phenomenon equivalent to taxonomic away-field disadvantage that has been often demonstrated (Gholz et al . 2000; Negrete-Yankelevich et al . 2008; Vivanco & Austin 2008). On the other hand, mixing of litter from distantly related focal and neighbor trees may trigger complementarity effects: transfer of nutrients among litters might increase decomposer abundance, diversity and activity and hence decomposition (PIB2, Lummer et al . 2012; Handa et al . 2014; Porre et al . 2020). Transfer of toxins among litters, in contrast, might decrease decomposers and their efficiency and hence decomposition (PIB3, Hättenschwiler & Vitousek 2000; Gessner et al . 2010). Finally, some neighboring lineages may degrade the physical environment in which the focal tree decomposes, e.g . by decreasing soil pH (PIB4) and soil moisture conditions (PIB5) or by increasing thermal fluctuations due to delayed budburst (PIB7, Yguelet al . 2014) and thereby a shorter vegetation period (PIB8). Other neighboring lineages might improve the physical environment in which the focal tree decomposes, e.g. by decreasing soil soaking and hence increasing aeration (PIB6, Cornelissen et al . 2017; Dias et al . 2017). Overall, phylogenetic isolation of a focal tree affects the pool of decomposers, their resources and the physical background, but consequences for decomposition remain unknown.
As decomposition below a given tree will be the result of a mixture of both aboveground and belowground processes, we designed a litter bag experiment for trees growing among phylogenetically closely or distantly related neighbors, i.e. under low or high phylogenetic isolation (PI, Fig. 1c). To our knowledge, this is the first ‘phylogenetic litter-transplantation’ experiment. Such an approach exposes litter from both high-PI and low-PI trees under a given tree to identify effects of PI operating via aboveground processes. Moreover, the approach exposes litter from a given tree under both high-PI and low-PI trees to identify the effects of PI operating via belowground processes. We studied oaks (Quercus petraea ), a system for which effects of phylogenetic isolation on enemy pressure and budburst have been demonstrated already (e.g. Yguel et al . 2011; Yguelet al . 2014) and for which major shifts in ranges of suitable climate are predicted (Hansen et al . 2001; Iverson & Prasad 2001; Barton 2002).
To identify ecosystem consequences of coexistence with distantly related neighbours we tested a set of increasingly complex predictions from hypotheses in Table 1: (1) Phylogenetic isolation above (PIA) and/or belowground (PIB) per se changes the rate at which litter decomposes (in terms of mass loss, C-loss and N-loss). (2) PIA and/or PIB changes the trajectories by which litter decomposes: the abundances and diversities of different soil biota controlling decomposition, the efficiency by which a given group of soil biota at a given abundance decompose litter, the putatively abiotic effects that are not attributable to soil organisms considered (Fig. 1a). As soil biota we accounted for the dominant groups Acari, Collembola and microorganisms, and, for a subsample, for fungi. Moreover, trajectories invoking higher decomposer activities should lead to a strong relative accumulation of13C (Bowling et al . 2008) and we hence tested whether PIA and/or PIB changes the13C/12C signatures. (3) The effects of PIA or PIB can be replaced, and hence likely explained by, litter traits and/or environmental characteristics as outlined above and in Table 1.