Hybridisation and introgression of exotic Cervus (nippon and canadensis) with red deer (Cervus elaphus) in the British Isles
Smith, Stephanie Lindsay
Europe’s largest population of wild red deer (Cervus elaphus) resides in the British Isles and has been present since the end of the last ice age, c. 11,000BP. Since the mid-19th century, multiple introductions of Japanese sika (Cervus nippon) and wapiti (Cervus canadensis) have taken place across the British Isles. While wapiti introductions have generally gone extinct, sika have thrived and expanded and now often live in sympatry with red deer. Hybridisation between these species has been demonstrated in captivity and in the wild. This study sought to determine the extent of hybridisation and introgression between red and sika across large parts of the British Isles and elucidate some of its potential consequences. Chapter 2 addresses the extent of hybridisation and introgression across Scotland and NW England. A total of 2984 samples from the North Highlands, the central Highlands, the Hebrides, Kintyre and the English Lake District were genotyped at 22 microsatellite loci, which are highly diagnostic for red and sika and strongly diagnostic for red and wapiti and a mitochondrial marker that is diagnostic for red and sika, alongside 49 wapiti samples from Canada. Microsatellite data was analysed using the Bayesian clustering program Structure 2.3 to determine the extent of admixture between species. There was some evidence for very low-level introgression by wapiti into a small number of Scottish red deer (<0.2% of total). Only two areas (both in Kintyre, Argyll) showed extensive introgression with collapse of assortative mating between red and sika (50.4% and 61.8% of sampled individuals were hybrid in West Loch Awe and South Kintyre, respectively). However, rare and widely scattered individuals with low-level sika introgression or cytonuclear disequilibrium suggest hybridisation has occurred in several other places in mainland Scotland and Cumbria in the past without subsequent loss of assortative mating. Chapter 3 addresses the extent of hybridisation in Ireland. There are now an estimated 4,000 red deer in Ireland and their numbers are increasing. It has recently been determined that the red deer in Killarney, County Kerry are descended from an ancient (c. 5,000BP) introduction and therefore merit genetic conservation. Introduction of exotic species, including Japanese sika and North America wapiti, since the 19th century have primarily occurred via the now defunct Powerscourt Park, County Wicklow, which was the source of many translocations to the rest of Ireland as well as to the UK. 374 deer samples from across Ireland were analysed as in Chapter 2. Wapiti introgression was again very low, with trace amounts of introgression detected in a small proportion of samples (0.53%), whilst 41% of 197 deer sampled in Co. Wicklow and 47% of 15 deer sampled in Co. Cork were red-sika hybrids according to either their nuclear genome or mitochondrial haplotype. No pure red deer were detected in Co. Wicklow, suggesting that in this region the red deer has disappeared following hybridisation. Whilst no hybrids were detected among 37 red samples and 77 sika samples in Co. Kerry, the Co. Cork hybrids pose a threat to the Killarney populations due to their proximity. Chapter 4 investigates population genetic structure within red and sika populations across the British Isles and investigates whether low-level introgression by the other species influences the resolved population structure. Structure analysis was conducted separately using 2307 ‘pure’ red deer individuals and 752 ‘pure’ sika animals from the British Isles (defined as Q > 0.95 for red and Q < 0.05 for sika) and then on reduced sample sizes using more stringent purity criteria (Q ≥ 0.99 and Q ≤ 0.01). As might be predicted, the more stringent criteria removed individuals in areas known to contain advanced backcrosses. In red deer, there was some evidence for a loss of within-species population structure under the more stringent criteria, while for sika there was not. Datasets were also analysed using Discriminate Analysis of Principal Components; a multivariate method designed to infer and describe genetic population structure. In red deer, both analytical approaches confirmed the strong separation of the deer on Harris and Lewis from others, and there is support for clusters typified by the other Hebridean islands, Kintyre, central and North Scotland and the English sites. Among sika, both approaches supported the likelihood of three clusters which are presumably the result of bottleneck events as each introduction was made. Chapter 5 investigates the phenotypic consequences of hybridisation by three approaches. Firstly, carcass weight was regressed against genetically-determined hybrid scores (at two stringency levels, see Chapter 4) and heterozygosity (in terms of red and sika alleles). Among hybrids, carcass weight is linearly related to hybrid score (Q) and there is some evidence for a positive relationship with heterozygosity. This suggests that additive genetic variation explains variation in carcass weight to a greater extent than heterosis. Secondly, analysis of five case studies representing individual putative hybrids submitted by stalkers from areas without known hybridisation, two proved to be hybrids, while the other three were pure sika. Lastly, in regions known to contain hybrids, the accuracy of ranger-assigned phenotype averaged 78% and revealed that in Scotland accuracy tends to decline as an individual becomes more genetically intermediate; whilst in Co. Wicklow it is the identification of pure parental animals that is more challenging. In conclusion, the existence of rare and widely scattered advanced red-sika backcrosses with low-level nuclear introgression and/or mitochondrial introgression (e.g. in North of Scotland, Cumbria) highlight that some hybridisation events are followed by extensive backcrossing without the breakdown of assortative mating, while others are followed by the generation of a hybrid swarm (e.g. in South Kintyre, West Loch Awe, Co. Wicklow, Co. Cork). Phenotypic traits can become intermediate due to hybridisation and this may facilitate further gene flow and hybridisation. New molecular tools including next generation sequencing (NGS) will enable better understanding the hybridisation process and its phenotypic consequences in this and other systems.