Contrasting population genetic structure among freshwater-resident and anadromous lampreys: the role of demographic history, differential dispersal and anthropogenic barriers to movement

Abstract

The tendency of many species to abandon migration remains a poorly understood aspect of evolutionary biology that may play an important role in promoting species radiation by both allopatric and sympatric mechanisms. Anadromy inherently offers an opportunity for the colonization of freshwater environments, and the shift from an anadromous to a wholly freshwater life history has occurred in many families of fishes. Freshwater-resident forms have arisen repeatedly among lampreys (within the Petromyzontidae and Mordaciidae), and there has been much debate as to whether anadromous lampreys, and their derived freshwater-resident analogues, constitute distinct species or are divergent ecotypes of polymorphic species. Samples of 543 European river lamprey Lampetra fluviatilis (mostly from anadromous populations) and freshwater European brook lamprey Lampetra planeri from across 18 sites, primarily in the British Isles, were investigated for 13 polymorphic microsatellite DNA loci, and 108 samples from six of these sites were sequenced for 829 bp of mitochondrial DNA (mtDNA). We found contrasting patterns of population structure for mtDNA and microsatellite DNA markers, such that low diversity and little structure were seen for all populations for mtDNA (consistent with a recent founder expansion event), while fine-scale structuring was evident for nuclear markers. Strong differentiation for microsatellite DNA loci was seen among freshwater-resident L. planeri populations and between L. fluviatilis and L. planeri in most cases, but little structure was evident among anadromous L. fluviatilis populations. We conclude that postglacial colonization founded multiple freshwater-resident populations with strong habitat fidelity and limited dispersal tendencies that became highly differentiated, a pattern that was likely intensified by anthropogenic barriers.

Keywords: Lampetra; anadromy; barriers to migration; life history; microsatellite; speciation.

Fig 1

Map showing location of sampling sites 1–18 (see Table S1, Supporting information for detail). Inset is a detailed map of part of the Ouse subcatchment of the Humber catchment, showing sampling locations. Only sampled rivers are shown.

Fig 2

(a) Median-joining network showing 16 haplotypes found from 108 samples of Lampetra at six sampling locations. Note that Bann (Lf) is a freshwater-resident L. fluviatilis population. Lf = anadromous L. fluviatilis, Lp = L. planeri and Lf Res = freshwater-resident population of L. fluviatilis. Details of the sample locations are given in Table S1 (Supporting information). (b) Forty-six haplotypes from combined studies comprising of both L. fluviatilis and L. planeri. Circled groups show correspondence with clades identified in Mateus et al. (2011). Clades I–III consist of freshwater-resident L. planeri (but see Mateus et al. 2013b) with restricted distribution, and clade IV contains both freshwater-resident Lp and anadromous Lf with a wider distribution along with haplotypes identified in Espanhol et al. (2007) from France, Sweden and Germany (Lp and Lf H22) and France (Lp H28). Please see the open-access online paper for a colour version of this figure.

Fig 3

Matrix of pairwise F_ST values using 13 microsatellite loci, for all Lampetra populations sampled. Lf = anadromous L. fluviatilis, Lp = L. planeri and Lf Res = freshwater-resident population of L. fluviatilis. Table showing the actual values is included in Supporting information (Table S5). Numbers on axes are marked with a square to represent L. planeri and a circle to represent freshwater-resident L. fluviatilis.

Fig 4

structure bar plot generated from microsatellite data for three population clusters of lampreys. (a) Comparison between Lampetra fluviatilis and Lampetra planeri populations (K = 6); (b) L. planeri populations (K = 6); (c) Loch Lomond populations compared to a population of L. fluviatilis from the Humber catchment and freshwater-resident L. fluviatilis populations from the R. Bann in N. Ireland (K = 3). Please see the open-access online paper for a colour version of this figure.

Fig 5

Isolation by distance tests for correlation between genetic differentiation (based on microsatellites) showing (a) geographic distance between freshwater-resident Lampetra planeri populations (R² = 0.30, P < 0.05) and (b) geographic distance between anadromous Lampetra fluviatilis populations (R² = 0.40, P < 0.0001; i.e. excluding freshwater-resident Bann and Lomond Lf). Inclusion of freshwater-resident Lf populations in the analysis reduced the strength of the correlation (R² = 0.0841, P < 0.05)—not shown. (c) number of barriers between samples sites (R² = 0.8256, P < 0.0001) where LP–LP signifies comparison of numbers of barriers between L. planeri sampling sites, LP–LF is number of barriers between L. planeri and L. fluviatilis sampling sites, and LF–LF is the number of barriers between L. fluviatilis sampling sites. Only sites for which barrier information was available were included in the analysis (i.e. Lf and Lp for Wear, Dee, Derwent, Nidd, Ure, and Swale Lf only).