Invasion is a community affair: Clandestine followers in the bacterial community associated to green algae, Caulerpa racemosa, track the invasion source

Abstract

Biological invasions rank amongst the most deleterious components of global change inducing alterations from genes to ecosystems. The genetic characteristics of introduced pools of individuals greatly influence the capacity of introduced species to establish and expand. The recently demonstrated heritability of microbial communities associated to individual genotypes of primary producers makes them a potentially essential element of the evolution and adaptability of their hosts. Here, we characterized the bacterial communities associated to native and non-native populations of the marine green macroalga Caulerparacemosa through pyrosequencing, and explored their potential role on the strikingly invasive trajectory of their host in the Mediterranean. The similarity of endophytic bacterial communities from the native Australian range and several Mediterranean locations confirmed the origin of invasion and revealed distinct communities associated to a second Mediterranean variety of C. racemosa long reported in the Mediterranean. Comparative analysis of these two groups demonstrated the stability of the composition of bacterial communities through the successive steps of introduction and invasion and suggested the vertical transmission of some major bacterial OTUs. Indirect inferences on the taxonomic identity and associated metabolism of bacterial lineages showed a striking consistency with sediment upheaval conditions associated to the expansion of their invasive host and to the decline of native species. These results demonstrate that bacterial communities can be an effective tracer of the origin of invasion and support their potential role in their eukaryotic host's adaptation to new environments. They put forward the critical need to consider the 'meta-organism' encompassing both the host and associated micro-organisms, to unravel the origins, causes and mechanisms underlying biological invasions.

Figure 1. PCA representing weighted Unifrac analysis of C. racemosa samples from Mediterranean Sea and Australia (invaded and native range), A-All samples including controls; B-Disinfected samples only.

Sample Codes: Al- Albany (Australia), CB- Cottesloe Beach (Australia), Lg- Liguria (Greece), AP- Agios Pavlo (Greece), Ml- Malta, M-Marseille, I- Illetas (Mallorca, Spain), EsC- Es Cargol (Mallorca, Spain), RI1- Rottnest Island 1 (Australia), RI2- Rottnest Island 2 (Australia), TN- Tunisia, VF- Villefranche (France).

Figure 2. Venn diagram representing bacterial communities shared within the three different treatments.

Numbers on each treatment represent the number of OTUs and percentages on overlapping areas represent the percentage of shared OTUs.

Figure 3. Maximum Likelihood tree from C. racemosa ITS calculated using the evolution Model TPM2+G with bootstraps calculated after 1000 replicates.

Alignments of each cluster were BLAST against Genebank nucleotide database and got the highest hits with sequences from Verlaque et al. 2003 study [AY334305-Cluster A and AY173118-Cluster B] C. prolifera samples from Mallorca were used as outgroup. Genbank accession numbers are represented within brackets.

Figure 4. Heatmap representing distribution of the main classes among samples.

Scale bar represents the percentage of sequences belonging to the OTUs represented in the heatmap.

Figure 5. Network of haplotypes from the most ubiquitous OTU (110) in the set of OTUs represented in Comamonadaceae family phylogenetic tree.

Network was drawn without keeping distance of links proportional to the number of mutations, in order to illustrate the clustering rather than the divergence