Neogene paleogeography provides context for understanding the origin and spatial distribution of cryptic diversity in a widespread Balkan freshwater amphipod
- Published
- Accepted
- Subject Areas
- Aquaculture, Fisheries and Fish Science, Biogeography, Evolutionary Studies, Molecular Biology, Zoology
- Keywords
- ancient lakes, gammarids, phylogeography, Bayesian analysis, Balkan Peninsula
- Copyright
- © 2017 Grabowski et al.
- Licence
- This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ Preprints) and either DOI or URL of the article must be cited.
- Cite this article
- 2017. Neogene paleogeography provides context for understanding the origin and spatial distribution of cryptic diversity in a widespread Balkan freshwater amphipod. PeerJ Preprints 5:e2741v1 https://doi.org/10.7287/peerj.preprints.2741v1
Abstract
Background. The Balkans are a major biodiversity and endemism hotspot, worldwide. Among the freshwater biota, amphipods are known for their high cryptic diversity. However, little is known about the temporal and paleogeographic aspects of their evolutionary history. We used paleogeography as a framework for understanding the onset of diversification in Gammarus roeselii: 1) we hypothesised that, given the high number of isolated waterbodies in the Balkans, the species is characterised by high level of cryptic diversity, even on a local scale; 2) the long geological history of the region might promote pre-Pleistocene divergence between lineages; 3) given that G. roeselii thrives both in lakes and rivers, its evolutionary history could be linked to the Balkan Neogene paleolake system; 4) we inspected whether the Pleistocene decline of hydrological networks could have any impact on the diversification of G. roeselii.
Material and Methods. DNA was extracted from 136 individuals collected from 26 sites all over Balkans. All individuals were amplified for ca. 650 bp long fragment of the mtDNA cytochrome oxidase subunit I (COI). After defining molecular operational taxonomic units (MOTU) based on COI, 50 individuals were amplified for ca. 900 bp long fragment of the nuclear 28S rDNA. Molecular diversity, divergence, differentiation and historical demography based on COI sequences were estimated for each MOTU. The relative frequency, geographic distribution and molecular divergence between COI haplotypes were presented as a median-joining network. COI was used also to reconstruct time-calibrated phylogeny with Bayesian inference. Probabilities of ancestors' occurrence in riverine or lacustrine habitats, as well their possible geographic locations, were estimated with the Bayesian method. A Neighbour Joining tree was constructed to illustrate the phylogenetic relationships between 28S rDNA haplotypes.
Results. We revealed that G. roeselii includes at least 13 cryptic species or molecular operational taxonomic units (MOTUs), mostly of Miocene origin. A substantial Pleistocene diversification within-MOTUs was observed in several cases. We evidenced secondary contacts between very divergent MOTUs and introgression of nDNA. The Miocene ancestors could live in either lacustrine or riverine habitats yet their presumed geographic localisations overlapped with those of the Neogene lakes. Several extant riverine populations had Pleistocene lacustrine ancestors.
Discussion. Neogene divergence of lineages resulting in substantial cryptic diversity may be a common phenomenon in extant freshwater benthic crustaceans occupying areas that were not glaciated during the Pleistocene. Evolution of G. roeselii could be associated with gradual deterioration of the paleolakes. The within-MOTU diversification might be driven by fragmentation of river systems during the Pleistocene. Extant ancient lakes could serve as local microrefugia during that time.
Author Comment
This is a submission to PeerJ for review.
Supplemental Information
Detailed information for all analysed individuals
Table S1. All analysed individuals with BOLD process ID, sample ID, Barcode Index numbers (BIN) and their COI haplotype membership, primers pairs used for amplification.
Divergence time for key nodes in the time-calibrated reconstruction of phylogeny
Table S2. Divergence time (Ma) for key nodes and rates of COI evolution in substitutions per site, per My estimated using Bayesian inference for different calibration schemes. Node affiliation given in Fig. S1. Calibration schemes based on secondary calibration points, standard rate and geological points (see material and methods). Lower and upper 95% highest posterior densities (95%HPD) are provided.
Results of the Bayesian node reconstruction
Table S3. Node ages (Ma) with lower and upper 95% highest posterior densities, geographical coordinates of the ancestral locations expressed as mean and standard deviation (SD), probabilities associated with occurrence in riverine (P(R)) or in lacustrine (P(L)) habitat (see Figure S1 for node identification).
K2p distances between MOTUs
Table S4. Mean Kimura two parameters (K2p) distances (below the diagonal) and standard error (SE) (above the diagonal) between MOTUs for 74 COI haplotypes. N and k = number of individual sampled and haplotypes per MOTU, respectively.
Fit of the diversification models to the Bayesian reconstruction of phylogeny
Table S5. Fit of the diversification models to Bayesian reconstruction of phylogeny, based on the Akaike information criterion (AIC).
Molecular genetic diversity and historical demography within MOTUs
Table S6. Molecular genetic diversity and historical demography based on mtDNA COI gene region (530 bp) within the 13 MOTUs (A-M). Analysis was done either at the scale of the entire MOTU (in bold) or a site or set of sites (in italic) within a MOTU.
Genetic differentiation between sites
Table S7. Genetic differentiation based on mtDNA COI gene region (530 bp) between sites (S1 vs S2) within MOTUs (A, C, E, G and K) present in more than one site within the morphospecies Gammarus roeselii in the Balkans.
Figure S1
Bayesian chronogram with node definition, calibration points and results of species delimitation methods.