Results
Sequences were obtained from the 30 samples (total of 180 samples in the three sites), 15 of them belonging to the Atlantic Ocean and 15 of them to the Pacific Ocean, these sequences were compared with the data available in NCBI from different geographical regions as described in table 1, for a total of 78 sequences to perform the analyses. The time-calibrated Bayesian phylogeny (Fig. 1B) placed the common ancestor of Ostreobium spp. in the Ordovician, around 500 Ma (95% HPD: 550– 450 Ma). The initial divergence within the genus were predicted to have occurred about 410 Ma (95% HPD: 420–350 Ma) during the Silurian. During the Mesozoic and Cenozoic occurred the most of the divergence events in Ostreobium s pp. Biogeographical results, shows many vicariant events that coincide whit the closure of the Panamanian Isthmus, nevertheless, this kind of result should be interpreted carefully, because the range of times that appear in error may show the event in different ages. Figure 2A shows a dominant event in the pacific that seems separate in two lineages towards Atlantic and pacific, but many of the groups at lower taxonomic level shows sequences founded in Pacific, Atlantic, Indo-Pacific and Red Sea.
The result of the analysis of species delimitation were similar, GMYC generated 21 groups with maximum 5 sequences and 25 individual categories, the mPTP produced 18 groups with maximum 3 sequences and 38 individual sequences, both analyses generated in common 15 groups (Fig. 2). Many of the individuals in the groups belonging to different geographic regions and different hosts. The results of the RASP analysis showed interesting patterns associated with the region of origin and the possible events of Ostreobium vicariance, out of a total of 146 nodes, 26 correspond to vicariousness (Fig. 2A) and 7 of them occurred in the proposed time interval. for the closure of the isthmus of Panama (15-3 Mya), 5 occur between individuals belonging to the Atlantic and the Pacific (nodes 32,94,101,133,139). Despite the difficulty of confirming it with certainty through our analyzes, the substantial oscillation in temperature and CO2 reduction known to have occurred in the late Miocene / early Pliocene, 3 Ma before the closure of the Central American seaway, could have initiated an expansion / bottleneck of the symbiont (LaJeunesse 2005). The rest of the vicariance events point to events that occurred during the Mesozoic, mainly in the Cretaceous and Triassic, at node 147 and 116 of figure 2 we observe the oldest events that are 293 Ma and 229 Ma ago respectively, although it is difficult to make inferences due to the temporal variation that the geographic regions have suffered, there is a slight tendency towards the separation of individuals from the Atlantic and the Pacific oceans.
Figure 3B shows the ancestry of the hosts of Ostreobium , so that it allows to give an idea about which hosts passed the symbiont and what was its main origin. It is observed how two convergence events occur on the individuals of the genus Hellioseris that acquiredOstreobium spp. from Agaricia , these individuals collected in the Atlántico locality are a perfect example of the low host specificity that Ostreobium presents, because although in this case the host passes between the same locality, this convergence phenomenon is repeated between Agaricia and other genera such asPocillopora or as hosts of the Red Sea. Ancestral reconstruction (Fig. 3B) result of the analysis in mesquite V 2.75, in the first node of the tree present uncertainly about the first possible host, however, the final groups of the tree defined the ancestor whit more confidence and showed pattern associates a host genus belonging to the same region, on the other hand, many of the transferences were acquired from genera from apart regions. Surprisingly, some branches evidence a pass from a genus and after a return at the same genus from the new host. The Atlantic individuals of the genus Agaricia in all cases passedOstreobium spp. to the genus Porites , from then onPorites managed to make this step between most of the genera, except Hellioseris and the substrates that did not correspond to the coral. In a study developed by Massé in 2018, an experiment is carried out at the level of coral development to know when it acquiresOstreobium spp., showing that are less than 5 days between the time of settlement of the larva of coral, until the moment in which the presence of the endolytic algae is confirmed. Given the speed of colonization and the lack of host choice in Ostreobium spp., it is predicted that the figure depicts ancestral host patterns in which Red Sea species acquire their symbiont from species belonging to the same genus from different geographic locations. In the group with the most individuals gathered by the GMYC analysis (Fig. 2A) we include individuals belonging to the Atlantic and Pacific locality of the genusPorites together with an individual of the genusPocillopora , this case is particularly interesting because there is a return to a previous host. Originally, it was observed that the ancestor is Agaricia (Atlantic) and that from this it makes a step towards Porites , however, a few nodes later it is observed how the passage from Porites towards Agaricia is made again confirming the generalist role that the endolithic algae has and the little importance that the genetic differentiation of the coral genera represents for their survival.