Introgression of mountain hare (Lepus timidus) mitochondrial DNA into wild brown hares (Lepus europaeus) in Denmark

Abstract

Background

In Europe the mountain hare (Lepus timidus) exists in Great Britain, Norway, Sweden, Finland, parts of the Alps and in Eastern Europe, but not in Denmark. Interspecific hybridization has been demonstrated between native Swedish mountain hares and introduced brown hares (Lepus europaeus). During the data collection in a study concerning Danish brown hares we identified 16 hares with a single very divergent haplotype.

Results

Phylogenetic analysis shows that the divergent Danish haplotype is most closely related to the Swedish mountain hare. The frequency of Lepus timidus mtDNA haplotype in the Eastern Danish hare populations is estimated to 6%.

Conclusion

In contrast to what is known, the Danish hare populations are not pure L. europaeus populations but include introgressed brown hares with Swedish L. timidus mtDNA. The most probable explanation of this is natural migration or translocation of introgressed brown hares from Sweden. The impurity of hare populations has implications for conservation and population genetics.

Background

The brown hare (Lepus europaeus Pallas, 1778) is widely distributed throughout Europe, up to 60°N, in Asia Minor and probably south to Israel. It is a popular game animal and its range has expanded to the east both by natural dispersion and by translocations to central and far-east Siberia [1]. Worldwide it has also been imported to many other countries, outside its natural distribution range, e.g. South and North America, Australia and New Zealand.

The brown hare is common all over Denmark and is the only Lepus species that exists in the country. Since 1960 the hare population has declined drastically in Europe and in Denmark. The primary cause of the decline is habitat changes in relation to agricultural intensification, whereas effects of climatic changes and predator abundance have increased by loss of year-round access to high-quality food and cover [2]. Additional factors like predators (especially foxes), birds of prey and traffic may have influenced the population locally [3]. In Denmark this is reflected by a large drop in annual number of hares shot from more than 400,000 before 1960 to 67,600 in 2004/2005 [4].

The general range of the mountain hare (Lepus timidus Linnaeus, 1758) covers much of the Palaearctic. In Europe it exists in Great Britain, Norway, Sweden, Finland, parts of the Alps and in Eastern Europe but it does not exist in Denmark [5]. During the late nineteenth century the brown hare was introduced in southern Sweden as a game animal [6]. In Scandinavia the mountain hare populations have been retreating northwards since the introduction of the brown hare, presumably as a result of a gradual competitive exclusion by the latter [7]. Mountain hares supposedly colonized Scandinavia through repeated immigration waves over two different post-glacial colonization routes; one from the south and one from the northeast [8].

Given the sympatric distribution of several hare species, e.g. in Sweden, Italy, and Spain, a number of studies focused on the possibility of interspecific hybridization. The first evidence of interspecific hybridization in hares concerns L. timidus and L. europaeus in Sweden [9–11] where unidirectional introgression of the native L. timidus mtDNA occurs in the introduced L. europaeus. However, introgression of mitochondrial DNA into other hare species has been demonstrated, e.g. the introgression of L. timidus mtDNA into L. granatensis, L. europaeus and L. castroviejoi in northern Iberia [12] and introgression of foreign mtDNA is likely to occur in several hare species from Asia [13]. These findings have lead to the conclusion that the genetic integrity of many European species largely depends on differences in behaviour and ecology which, at best, offer semi-permeable isolation [14]. In captivity, mountain hare females mate with brown hare males and produce viable offspring, whereas the reverse crosses do not happen spontaneously but can be performed successfully with insemination [15]. The F₁ hybrids are morphological intermediates between the two species [16].

During the data analysis in a study concerning Danish brown hares we identified 16 hares with a very divergent haplotype. Here we show that these 16 individuals from the Danish wild brown hare populations carry mountain hare mtDNA haplotypes.

Results

During alignment of 385 hare sequences, 16 sequences (6 males, 10 females) were clearly different from the rest and collapsed into a single haplotype. The 16 individuals were obtained from five of the eight populations sampled, and these five populations are all located on islands in the eastern part of Denmark (Figure 1). The frequency of sampled individuals with L. timidus mtDNA in those five populations averaged 6% (a crude estimate across all eight populations is 4.16%, Figure 1).

Figure 1

Location and frequency of the Lepus timidus haplotype in Denmark. Lepus timidus was found in five (large map) of the eight (small map) sampling locations.

A BLAST search against GenBank with the divergent haplotype mtDNA returned L. timidus as the best hit (99% similarity), but not a single L. europaeus sequence in the first 50 hits. The subsequent phylogenetic analysis (Figure 2) revealed that the divergent Danish haplotype is grouped with L. timidus samples from across Europe (100% posterior probability in all three independent runs) and most closely related to the Swedish L. timidus haplotypes (100% posterior probability in all three independent runs, green clade in figure 2). The other Danish haplotypes included in the phylogenetic analysis, the 19 L. europaeus haplotypes, are grouped with the L. europaeus clade as expected, and they are most closely related to German haplotypes (100% posterior probability from three independent runs, red clade in figure 2). Hence, the Danish haplotypes clearly originate from two different species.

Figure 2

Bayesian phylogeny of 370 bp from the mitochondrial D-loop of 20 Lepus timidus and 28 Lepus europaeus haplotypes from Europe, including the 19 Danish Lepus europaeus haplotypes, and the divergent Danish haplotype. The numbers show nodal support, given as posterior probabilities (for selected nodes, posterior probabilities from three independent runs are shown).

Discussion

The results show the presence of mountain hare mtDNA in Danish brown hare populations. From mtDNA alone we cannot decide whether these 16 individuals are true mountain hares or introgressed brown hares. This would require further analysis using nuclear gene sequences (e.g. transferrin) (there are no differential diagnostic microsatellite loci for the two species [17]). However, given a) the frequency of the L. timidus haplotype where it is present (6% in the five populations), b) that L. timidus is not reported to be a native breeding species in Denmark [5], and c) that the two species are able to hybridize [9], these individuals most likely represent introgressed brown hares.

The most likely origins of introgression are; 1) natural migration of mountain hares from Sweden followed by introgression, or 2) natural migration or translocation of introgressed brown hares from Sweden. However, since the documented distribution of mountain hares in Sweden does not include southern Sweden [9, 10], natural migration of mountain hares is less probable than introduction of introgressed brown hares (carrying mountain hare mtDNA) from Sweden.

The fact that The Baltic Sea occasionally freezes over in very severe winters facilitates natural migration of introgressed brown hares from southern Sweden across the sea barrier to the Eastern Danish islands (Figure 1). This is also supported by the results from the genetic analyses (Figure 2).

During the mid 1980'ies a network of hare breeding facilities was established in Denmark, and it was initiated with imported hares primarily from Italy, France, Hungary and Sweden [18]. No combined record is kept of the amount and origin of hares imported to Denmark. In 1993, when farming hares were prohibited in Denmark, there were 100 hare-farms in Denmark, with an annual export of up to 5,500 hares [19]. As approximately 15% of Swedish L. europaeus specimens from sympatric areas carried transmitted L. timidus mtDNA [10], it is possible that phenotypic brown hare specimens imported from Sweden for breeding purposes have carried the L. timidus mtDNA and later escaped to the natural hare population.

The detection of only one haplotype in the 16 individuals shows low mtDNA diversity in the total population of hybrids in Denmark. The geographical distribution of the 16 intogressed brown hares (Figure 1) and the lack of haplotype diversity support a common origin in southern Sweden.

The implications of introgressed brown hares in the Danish hare populations are currently unclear. It has been suggested that hares with an alien mtDNA have a lowered fitness as a result of a functional incompatibility between the cytoplasmatic mitochondrial genes and the cell nucleus [10]. In opposition to this, due to its observed high frequency in brown hares (93%) Melo-Ferreira et al. [20] suggested, that the ancient L. timidus mtDNA observed in the Iberian Peninsular might have some selective advantage depending on the nuclear background.

Thulin et al. [17] raised the question as to whether any 'pure' population of brown hares exists anywhere. Our findings give further support to this statement. This has important consequences for conservation and population genetics, e.g. problematic definition of management units, unclear fitness effects, genetically mixed populations and the inclusion of nuclear markers will be necessary for future hare studies.

Conclusion

Contrary to what is known, the Danish hare populations are not pure L. europaeus populations but include introgressed brown hares with Swedish L. timidus mtDNA. The most probable explanation is natural migration or translocation of introgressed brown hares from Sweden. The impurity of hare populations has implications for conservation and population genetics.

Methods

Sample collection

During 2003–2006 The National Environmental Research Institute of Denmark collected tissue samples from shot, wild hares at eight locations in Denmark (Figure 1). The hares were autopsied and several organs and muscle tissue samples were taken out for later analyses.

Sequencing

DNA was extracted from hare muscle tissue using a slight modification of the Chelex protocol [21]. The tissue was placed into 1.5 mL Eppendorf tubes containing 200 μl of 20% Chelex resin solution (BIO-Rad, Hercules, CA, USA). Tubes were vortexed briefly, boiled at 100°C for 20 min in a heating block, centrifuged 3 min at 13000 rpm and stored at minus 20°C.

A 494 bp fragment of the mtDNA D-loop (control region) was amplified via PCR, using the mammalian control region primers L15997 5'-CACCATTAGCACCCAAAGCT-3' located in the tRNA gene and H16498 5'-CCTGAAGTAGGAACCAGATG-3' [22]. PCR was performed in a total volume of 10 μl and contained 1 μl buffer (1.5 mM MgCl₂), 1.6 μl dNTP (1.25 mM of A, C, G, T, respectively), 0.5 μl of each primer (10 pmol/μl), 0.1 μl Taq polymerase (Amersham Pharmacia Biotech), topped up with distilled water to 9 μl and 1 μl template DNA (40–60 ng/μl) was added. Cycling conditions were 94°C for 3 min, and 35 cycles of 94°C for 30 s, 50°C for 40 s and 72°C for 60 s and a final extension step of 72°C for 7 min. Sequencing was conducted under BigDye™ terminator cycling conditions, the reacted products purified using ethanol precipitation and run using an Automatic Sequencer ABI 3730xl. Both strands were sequenced in all samples.

Haplotype editing and collapsing

Sequences were edited using Bioedit version 7.0.0 [23]. Identical haplotypes among the 385 sequences were found using DNAcollapser [24].

Phylogenetic analyses

Nine L. europaeus and 20 L. timidus sequences were downloaded from GenBank. Sequences with geographical information were selected to represent haplotypes from all over Europe and Russia. Together with the 19 Danish L. europaeus and the divergent haplotype, these 49 sequences (table 1) were aligned using Muscle [25] and cropped to the shortest sequence (370 bp). Phylogenies and nodal support were estimated using MrBayes, version 3.0b4 [26] under a Bayesian framework [27, 28] using a general time reversible substitution model (GTR + γ) and the coalescent branch length model. Bayesian analysis was initiated with random starting trees, run for 5 × 10⁶ generations, and the Markov chain was sampled every 1000 generations. Model parameters were estimated directly from the data and three independent replicates were conducted to avoid entrapment in local optima [29]. The initial 1,250 trees were discarded as "burn-in" and the remaining 3,750 trees were used to estimate nodal support as posterior probabilities.

Table 1 Sampling localities and accession numbers of D-loop sequences used in the phylogenetic analysis (figure 2).