Daphnia

Bellati, Adriana, Tiberti, Rocco, Cocca, Walter, Galimberti, Andrea, Casiraghi, Maurizio, Bogliani, Giuseppe & Galeotti, Paolo, 2014, A dark shell hiding great variability: a molecular insight into the evolution and conservation of melanic Daphnia populations in the Alps, Zoological Journal of the Linnean Society 171 (4), pp. 697-715 : 704-705

ALPINE DAPHNIA

Firstly, COI sequences of melanic individuals were compared using the Basic Local Alignment Search Tool (BLAST, http://blast.ncbi.nlm.nih.gov/Blast.cgi, Altschul et al., 1997) with all homologous sequences available in GenBank (http://www.ncbi.nlm.nih.gov/GenBank/) using the nucleotide blast (nblast) with default parameters. Sequences were also compared with the Barcoding of Life Database (BOLD, http://www.barcodinglife.com/index.php/databases, Ratnasingham & Hebert, 2007), using the engine BOLD-IDS for identification and checking the ‘All Barcode Records on BOLD’ option. Genetic distances amongst different COI haplotypes were estimated using a standard DNA barcoding method (i.e. Kimura’s two-parameter model correction, K2P, Kimura, 1980).

Phylogenetic relationships of the newly discovered pigmented populations were inferred both by maximum likelihood (ML) criterion and posterior probability statistic based on Bayesian inference (BI). We based our analyses only on the ND5 fragment (540 bp) because no COI sequences were available for the distinct geographical EuPC lineages and haplogroups. Homologous sequences were retrieved from GenBank, including 139 samples representative of all EuPC populations analysed so far ( Marková et al., 2007, 2013; Vergilino et al., 2009, 2011; Dufresne et al., 2011). Available sequences from the eight high-mountain populations sampled on the Alps (N = 14, see Table 1 and Fig. 3 for geographical distribution of samples) were used as a comparison to infer phylogenetic relationships of melanic individuals, together with 17 and 28 sequences belonging to the PYR and the HTM haplogroups, respectively ( Table 1). In order to validate morphological taxonomy, we also selected four sequences of D. middendorffiana (MIDD) and 15 sequences of Nearctic D. pulicaria (NAPC) , including haplotypes previously found in European pigmented populations (see Table 1 for accession numbers). Two sequences of European D. pulex Leydig, 1860 (EuPX) were considered as the outgroup, and the final alignment was performed using the CLUSTALW module of BIOEDIT 7.1 ( Hall, 1999).

The best-fit model Hasegawa-Kishino-Yano ( Hasegawa, Kishino & Yano, 1985) with gamma substitution parameter (+ G, α = 0.5490) was selected amongst 88 possible models of evolution according to the corrected Akaike information criterion (AICc; Hurvich & Tsai, 1989) implemented in jMODELTEST 0.1.1 ( Guindon & Gascuel, 2003; Posada, 2008). ML was performed using the BEST approach implement- ed in PhyML 3.0 ( Guindon et al., 2010), which combines the nearest-neighbour interchanges with the subtree pruning and regrafting algorithms to maximize tree likelihood. Statistical support for nodes was quantified by a nonparametric bootstrap test using 1000 replicates ( Felsenstein, 1985) and estimating Shimodaira-Hasegawa (hereafter SH)-like approximate likelihood ratio probabilities. BI was performed using MrBayes 3.2 ( Ronquist et al., 2012), which calculates posterior probabilities using a Markov chain Monte Carlo sampling approach ( Huelsenbeck et al., 2001; Altekar et al., 2004). Two independent runs implementing four chains were carried out starting from random trees and with a total length of 10 × 106 generations for each one. Trees were sampled every 100th generations and the earliest 20% of the data sampled were discarded as burn-in after checking tracer plots in TRACER 1.5 ( Rambaut & Drummond, 2007). Convergence of chains upon a stationary distribution was also checked by monitoring the standard deviation of split frequencies (= 0.0034) and the potential scale reduction factor (= 1.000). Average genetic sequence divergences (and relative standard errors, SE) between distinct lineages and haplogroups found in European mountain regions were calculated by uncorrected pairwise genetic distance estimations (p -distance) using MEGA 5 and setting 1000 bootstrap replicates ( Tamura et al., 2011).

Finally, intraspecific relationships amongst highmountain populations sampled in the Alps (including pigmented Daphnia specifically sampled for this study) were visualized by haplotype network analysis using TCS 1.21 ( Clement, Posada & Crandall, 2000) after collapsing individual sequences into haplotypes using the online web tool DnaCollapser 1.0 available on the FaBox site (http://users-birc.au.dk/biopv/php/fabox/). A parsimony criterion following the algorithm by Templeton, Crandall & Sign (1992) was used to estimate the number of mutational steps by which pairwise haplotypes differ, assessing also the minimum number of connections required to join together all the haplotypes in a single gene network.

Figures

Figure 3. Map of the Alps showing European Daphnia pulicaria populations (1–12) and haplotypes (A1–A10) considered in this study. Different colours and patterns identify haplotypes found in more than one population or together with other haplotypes. See Table 1 for details of populations.