Iberian red deer: paraphyletic nature at mtDNA but nuclear markers support its genetic identity

Abstract

Red deer populations in the Iberian glacial refugium were the main source for postglacial recolonization and subspecific radiation in north-western Europe. However, the phylogenetic history of Iberian red deer (Cervus elaphus hispanicus) and its relationships with northern European populations remain uncertain. Here, we study DNA sequences at the mitochondrial control region along with STR markers for over 680 specimens from all the main red deer populations in Spain and other west European areas. Our results from mitochondrial and genomic DNA show contrasting patterns, likely related to the nature of these types of DNA markers and their specific processes of change over time. The results, taken together, bring support to two distinct, cryptic maternal lineages for Iberian red deer that predated the last glacial maximum and that have maintained geographically well differentiated until present. Haplotype relationships show that only one of them contributed to the northern postglacial recolonization. However, allele frequencies of nuclear markers evidenced one main differentiation between Iberian and northern European subspecies although also supported the structure of both matrilines within Iberia. Thus, our findings reveal a paraphyletic nature for Iberian red deer but also its genetic identity and differentiation with respect to northern subspecies. Finally, we suggest that maintaining the singularity of Iberian red deer requires preventing not only restocking practices with red deer specimens belonging to other European populations but also translocations between both Iberian lineages.

Keywords: Cervus elaphus hispanicus; Iberian glacial refugia; conservation genetics; paraphyletic taxa; phylogeny; phylogeography.

Figure 1

Phylogenetic reconstruction of the 45 red deer haplotypes found in this study (n = 682) from the 757 bp of the control region (D‐loop) analyzed. Branch lengths are Bayesian posterior probabilities and the scale bar represents the mean number of nucleotide changes per site. Main nodes labeled with numbers are used below for estimating divergence times (Table 4).

Figure 2

Genetic pairwise differentiation between red deer assemblages based on mtDNA haplotypes. See Table 3 for Fst values and significance.

Figure 3

Sampling sites and Iberian clusters: Map shows current distribution of red deer in Spain (dots indicate presence in UTM 10 × 10 km, after Palomo et al. 2007) and location of sampling localities indicated with numbers that correspond to those appearing in Table 1. Circles at each location depict the relative frequency of samples from both haplogroups. Shaded areas show possible natural distribution ranges for both haplogroups, while arrows indicate known restocking routes (information on restocking after Pérez et al. 1998; Carranza 2011, and J. Carranza unpublished data).

Figure 4

(A) Mismatch distribution of mtDNA haplotypes for each of the Spanish red deer phylogroups. The expected frequency is based on a population constant model and is represented by a continuous line. The observed frequency is represented by a dotted line. (B) Bayesian skyline plots showing the historical demographic trends for each of the Spanish red deer phylogroup. The y‐axis express population size, estimated in units of Neμ (Ne: effective population size, μ: mutation rate per haplotype per generation) in a log‐scale. Median estimates and confidence intervals are shown.

Figure 5

STRUCTURE software results. (A) All populations. Graph shows the log‐likelihood values, Ln Pr (X⁄K) (white points) and Ad hoc statistic ΔK (black points) showing the most probable number of genetic clusters (K). Membership coefficients of probability for each cluster are also showing. (B) Non‐Spanish populations. (C) Spanish populations. NOR: Norway; SCOT: Scotland; GER: Germany; SSP: Sierra de San Pedro; MT: Montes de Toledo; CAR: Cardeña; HOR: Hornachuelos. Different subspecies are coded with the following notation: atlan: C. e. atlanticus; scotic: C. e. scoticus; elaph: C. e. elaphus; hispan: C. e. hispanicus.

Figure 6

Unrooted neighbor‐joining tree for the obtained genetic clusters (Norway, Scotland, Germany, west Spain, central‐south Spain) based on microsatellite allele frequencies.

Figure 7

Three‐dimensional plot of the factorial correspondence analysis (FCA). Populations are coded with different colors: Sierra de San Pedro (gray), Montes de Toledo (pink), Cardeña (green), Hornachuelos (dark blue), Norway (yellow), Scotland (blue), and Germany (white). To facilitate interpretation, we indicate different areas in the plot with circles. Individuals from Scotland and Germany are mixed in the same space. One individual from Sierra de San Pedro appears within the central‐south Spain area and one from Cardeña in west Spain.

Figure 8

Relationship between geographical distance and genetic relatedness between individuals. Average estimates were taken for individuals separated in the following distance intervals: Same population; <100 km (individuals from Cardeña and Hornachuelos were compared); <200 km (individuals from Montes de Toledo‐Cardeña and Montes de Toledo‐Hornachuelos were compared); <300 km (individuals from Sierra de San Pedro‐Montes de Toledo, Sierra de San Pedro‐Cardeña and Sierra de San Pedro‐Hornachuelos were compared); >300 km (individuals from different countries were compared). Figure shows mean and standard deviation.