Self-financed Partner 8 - Drs. Thorsten Stoeck, Micah Dunthorn, and teams; University of Kaiserslautern, Germany.
P8 used and developed network and graph theory tools to analyze the BioMarKs ciliate data. Main aspects of the investigations were dispersal patterns of ciliates and detection of novel ciliate diversity. Generally, DNA and RNA-based analyses revealed similar patterns. Locations/geographic distance and different habitats were tested as possible factors shaping ciliate communities. Analyses showed that both factors have a structuring impact on the sampled communities. Significantly assortative communities were found throughout all network analyses.
Occupancy-abondance analyses gave a more detailed look on how ciliates were distributed along the BioMarKs sampling sites (Fig. 6). Only few cosmopolitan ciliates occurred at a similar evenness at all sampling sites. Whereas the vast majority of the investigated components showed some degree of endemism and occurred at three or less sampling sites. Ciliates follow the moderate endemicity hypothesis in their dispersal.
Figure 6: Three-dimensional plots displaying abundance, occurrence and evenness of sequences in connected components (CC). Each circle (n = 593) represents one CC. Color and position of a circle on the y-axis indicates its abundance, or the number of sequences in the CC. Occurrence refers to the number of BioMarKs sampling sites (maximum = 8) that are found in the CC. Evenness (estimated as Simpson index) describes how homogenously the different sampling sites are distributed in the CC (0 meaning uneven distribution, 1 meaning most homogenous distribution). CC with highest read abundance includes 10,255 sequences, occurring at three sampling sites with low evenness.
The structuring effect of habitats on ciliate communities was also tested with network tools. Results show that the proportion of components consisting exclusively of sediment sequences by far exceeded the proportion of subsurface or DCM exclusive components (Fig. 7). That is, benthic ciliate communities do not only share less ciliate diversity with communities observed in the water column, but also fall into many small exclusive components in the networks. Consequently, we suppose that adaption to a psammophilic lifestyle has lead to a higher ciliate diversity in benthic habitats than in pelagic habitats.
Figure 7: Composition of connected components (CC) in gene similarity networks. Each CC at each similarity threshold was analyzed to highlight the proportion of nodes, which derived from the three investigated habitats in this study (DCM, Sediment, Subsurface). The graphs show the proportion of all CC which exclusively consist of nodes of the same habitat.
Network analyses identified a large proportion of novel ciliate diversity. This proportion is represented by the dark blue nodes in Fig. 8, which are the communities that could not be assigned to any ciliate reference organism (cultured or environmental). Hence, after approximately two decades of microbial diversity studies, BioMarKs could still reveal a substantial amount of novel ciliate diversity.
Figure 8: Gene similarity network showing the giant connected component (GCC) of each dataset. GCC were constructed at the most inclusive sequence similarity threshold (≥85%) for cDNA (left) and DNA (right) network. Sequences in the network derived from cultured organisms (GenBank), environmental clones of former environmental studies (GenBank) and BioMarKs data. Nodes of the GCC represent Louvain communities (LC); the size of the node indicates the number of sequences in the respective LC, the color indicates the proportion of BioMarKs sequences in the respective LC. Only dark blue nodes represent LC consisting exclusively of BioMarKs sequences, highlighting the proportion of novel ciliate diversity only discovered in BioMarKs.
Based on the network analyses results, we identified two promising candidates for the detection of novel ciliate diversity. In order to recover the respective organisms we created specific primers and probes for amplification and in situ hybridization of the targeted ciliates. Phylogenetic analyses suggest both candidates belong to the class Spirotrichea and represent a potentially novel genus (inside the family Tontoniidae) and species (belonging to the genus Pelagostrobilidium), respectively. However, these results remain to be confirmed by morphological analyses of live sample material.
If the approach proves to be successful, more potential candidates might be identified in the network analyses and subsequently recovered in environmental samples.