Strong specificity and network modularity at a very fine phylogenetic scale in the lichen genus Peltigera

Abstract

Identifying the drivers and evolutionary consequences of species interactions is a major goal of community ecology. Network-based analyses can provide mathematical tools to detect non-random patterns of interactions, and potentially help predicting the consequences of such patterns on evolutionary dynamics of symbiotic systems. Here, we characterize the structure of a lichen network at a very fine phylogenetic scale, by identifying the photosynthetic partners (i.e., cyanobacteria of the genus Nostoc) of lichenized fungi belonging to a monophyletic section of a single genus (i.e., section Polydactylon of the genus Peltigera), worldwide. Even at such a fine phylogenetic scale, we found that interactions were highly modular and anti-nested, indicating strong preferences in interactions. When considering local Peltigera communities, i.e., datasets at small spatial scales with only a slightly broader phylogenetic range, interactions remained modular but were asymmetric, with generalist Nostoc partners interacting with specialized Peltigera species. This asymmetry was not detected with our global spatial scale dataset. We discuss these results in the light of lichen community assembly, and explore how such interaction patterns may influence coevolution in lichens and the evolutionary stability of the mutualism in general.

Species diversification and phylogenetically constrained symbiont switching generated high modularity in the lichen genus Peltigera

“…We thus calculated two widely used network metrics (i.e., modularity and nestedness), over a gradient of sampling effort ranging between ~ 50 and our full 1,026 thalli dataset. Confirming what we had found in prior analyses on the section Polydactylon of the genus Peltigera (Chagnon et al, ), we found that modularity and nestedness did vary with sampling effort, but with trends that increasingly diverged from random expectations as sampling effort increased (Figure ). In other words, further sampling would have only strengthened our conclusions about network structure.…”

“…We assembled a dataset of Peltigera-Nostoc interaction pairs summing up all information from the literature for which we could confidently assess the identity of the Peltigera species and Nostoc phylogroups or haplotypes (Chagnon et al, 2018;Lu et al, 2018;Magain, Miadlikowska, Goffinet, et al, 2017a;Magain, Miadlikowska, Mueller, et al, 2017b;Magain et al, 2018;Miadlikowska et al, 2014Miadlikowska et al, , 2018O'Brien, Miadlikowska, & Lutzoni, 2005;O'Brien, Miadlikowska, & Lutzoni, 2013;Pardo De la Hoz et al, 2018). The total dataset consisted of 1,026 thalli, or interaction pairs, representing 155 Peltigera species and 95 Nostoc phylogroups or haplotypes (Tables S1 and S2).…”

“…To infer a phylogeny for the entire genus Peltigera, we gathered published data from all currently validated species (Chagnon et al, 2018;Magain, Miadlikowska, Goffinet, et al, 2017a;Magain, Miadlikowska, Mueller, et al, 2017b;Magain et al, 2018;Miadlikowska et al, 2014Miadlikowska et al, , 2018Pardo De la Hoz et al, 2018) for seven loci (ITS, nrLSU, β-tubulin, RPB1, COR1b, COR3, COR16). We selected one representative per species with the highest number of sequenced loci, or two representatives per species when sets of available loci were mostly non-overlapping between two specimens of a species.…”

“…Because each locus was analysed separately in previous studies to detect potential topological conflicts, with appropriate corrections made when needed (Chagnon et al, 2018;Magain, Miadlikowska, Goffinet, et al, 2017a;Magain, Miadlikowska, Mueller, et al, 2017b;Magain et al, 2018;Miadlikowska et al, 2014Miadlikowska et al, , 2018Pardo De la Hoz et al, 2018), we concatenated these seven loci using in-house PERL scripts (Magain, 2018). Our concatenated dataset consisted of 205 specimens and 10,064 characters.…”

“…These lichenforming fungi (also referred to as mycobionts) are typically highly specialized on one or very few cyanobacterial phylogroup partners (also referred to as cyanobionts) that commonly associate with multiple Peltigera species (asymmetric specificity). The evolution of specialization has fuelled a large body of literature in ecology and evolution (Fisher, 1930;Levins, 1968;MacArthur, 1955;May, 1974) and one question that might be asked when looking at current patterns of specialization is whether selection towards such patterns was directional or stabilizing: in other words, is there a directional evolution towards specialization (Jaenike, 1990) or generalization (Waser, Chittka, Price, Williams, & Ollerton, 1996) or a combination of both, that is, specialization of Peltigera and generalization of Nostoc, which could be advantageous especially when symbionts are transmitted mostly horizontally (Chagnon et al, 2018;Lu et al, 2018;Lutzoni & Miadlikowska, 2009;Magain, Miadlikowska, Goffinet, Sérusiaux, & Lutzoni, 2017a)? Alternatively, has the specialization level of a guild been stable for a long time due to selection against higher, or lower, levels of specialization (i.e., stabilizing selection)?…”