In this study, we demonstrated that genetic monitoring can be used as a method to evaluate the effectiveness of a long-term culling operation to reduce an invasive species population. We used a combination of mtDNA to gain historical insights into the introduction and establishment of the grey squirrel in north Wales, and temporal microsatellite DNA data to assess if culling had altered the genetic diversity. Previous studies have shown that culling is an effective management strategy as it shows a reduction in population size (Signorile & Shuttleworth 2016). However, our study suggests a scenario of relative population stability despite intensive control efforts. This is likely a consequence of the high number of historical introductions of genetically diverse grey squirrels from a broad geographical area in North America, and continued gene flow from neighbouring populations within north Wales containing genetically differentiated populations.
mtDNA
A total of six mtDNA haplotypes were found within the north Wales population, three of which were new to this study. Twelve haplotypes were previously recorded throughout Britain, but these haplotypes were taken from a relatively small sample sizes; 14 from Stevenson et al (2013) and a further 73 from Stevenson-Holt & Sinclair (2015), and were representative of the populations in England and Scotland. However, given the high number of haplotypes found within north Wales alone, it is likely that the mtDNA diversity of the grey squirrel population is under recorded in Britain, and there are likely many other undocumented haplotypes present. Our mtDNA analysis showed that the haplotypes within north Wales are genetically diverse, and only two haplotypes, H9 and H15 positioned relatively close to each other on the network diagram (Fig. 3). Even the three haplotypes found on the island of Anglesey, H9, H12 and H14 are distinct from one another and likely originated from different introductions from North America. When comparing the British haplotypes to those recorded in North America (Moncrief et al 2012), the contemporary British population is highly diverse, with haplotypes from Britain found throughout the network which contained animals from its eastern and southern US distribution. Moncrief et al (2012) only sampled 69 animals from throughout this region, which again suggests that mtDNA within North America is also under recorded. However, despite the relatively low level of sampling, we found three haplotypes from North America in the contemporary British population, H8, H11 and H12. Haplotype H8 was previously recorded in Cumbria in northwest England, while H11 and H12 were recorded in Henbury in Northern England, a well-documented introduction site in Britain, from which other translocations took place (Stevenson-Holt & Sinclair 2015). The overlap of the H12 haplotype within all three studies can directly link grey squirrel movement patterns. This provides the information to trace the translocations from Maryland and Virginia in eastern North America to northern England and now to Wales. This possibly identifies a migration route of the grey squirrels from northern England to Clocaenog as this seems to be an entry point for moving populations into north Wales as it has the highest diversity of haplotypes in the areas sampled for this current study.
The high mtDNA diversity found within Britain today is testament to the large numbers of repeated introductions and translocations that took place over a 60-year period. This is further confounded by the species’ naturally high level of genetic diversity owing to its evolutionary history in North America where its genetic legacy appears not to have been imprinted by glacial history, limiting interpretation of its phylogeographic history (Moncrief et al 2012). Other Sciurus species have similar phylogeographic legacies including S. niger in North America and S. vulgaris in Eurasia (Grill et al 2009; Moncrief et al 2012; O’Meara et al 2018). All have a high number of mtDNA haplotypes that cannot generally be traced to commonly recognised glacial refugia. While this legacy may on the surface appear to be a contributing factor to the overall success of the grey squirrel in terms of its colonisation ability following repeated introductions, the same cannot be said for the red squirrel, who has also been reintroduced and translocated multiple times in parts of Britain and Ireland, and retains a similarly diverse genetic history (O’Meara et al 2018).
Microsatellite
In relation to the microsatellite data, the grey squirrel population in north Wales is overall genetically diverse and contains relatively high levels of allelic richness (averaging 6.7) and high levels of expected heterozygosity (average HE = 0.56), but rather low levels of observed heterozygosity (average HO = 0.35) were obtained. The lower levels of observed heterozygosity are likely indicative of diminishing genetic diversity, which may also be related to the high levels of inbreeding evidenced within the population which were high and significant across all loci and averaged 0.35. The levels of allelic richness recorded in north Wales are higher than previously reported for Northumberland in northeast England (AR = 4.3) and in East Anglia in the east of England (AR = 3.44). Levels of expected and observed heterozygosity were also higher for both populations in Northumberland (HE = 0.66; HO = 0.64) and East Anglia (HE = 0.71; HO = 0.77), and less of a difference between both expected and observed values of heterozygosity were observed (Signorile et al 2014). FIS values were also lower in both the Northumberland (0.04) and East Anglia (-0.15) populations (Signorile et al 2014) suggested that these populations are less inbred than populations in this current study. Given the high levels of inbreeding and the lower levels of observed heterozygosity in comparison to expected heterozygosity, it was anticipated that the population might currently or previously have experienced a genetic bottleneck. However, there was no evidence of a recent genetic bottleneck as evidenced by a normal ‘L’ shaped distribution of mode-shift test and non-significant heterozygote excess. It may be the case that the high levels of allelic richness are countering or protecting the population from entering a bottleneck. Bottlenecks are detected when levels of allelic richness decrease before a decrease in heterozygosity has occurred resulting in higher levels of heterozygosity than expected by the number of alleles (Luikart et al 1998). The high levels of inbreeding in absence of a genetic bottleneck is somewhat of a paradox, as invasive species are often expected to have overcome a genetic bottleneck during the invasion process (Frankham 2005), but in the case of the grey squirrel, it appears to have avoided a bottleneck, despite high levels of inbreeding, perhaps due to the high numbers of genetically diverse founders. Similar genetic legacies have been found in the invasive brown anole lizard (Anolis sagrei), which due to the effect of multiple introductions from different areas, resulted in higher genetic diversity in invasion sites than in its original native range (Kolbe et al 2004).
The impact of repeated culling over time appeared to illicit mixed results in relation to genetic diversity. For instance, observed and expected levels of heterozygosity generally decreased over time in the Gwynedd population between 2014 (HO = 0.410; HE = 0.612 and 2019 (HO = 0.353; HE 0.609), but actually increased in Clocaenog between 2011 (HO = 0.336; HE = 0.494) and 2015 (HO = 0.372; HE 0.551). However, variation also occurred within the sampling years, which is likely a consequence of geneflow and new animals entering the culled area, particularly in Clocaenog which is geographically closer to England and as was said earlier, is likely to have greater colonisation potential. This is further corroborated by an increase in the number of alleles and levels of allelic richness in the area, which actually increased during the culling period. Inbreeding levels were high and significant across all years sampled, and again no evidence of genetic bottlenecks were found. Indeed, the STRUCTURE and PCA results from this study have shown that the cull that took place in Gwynedd in 2014 contained a genetically distinct population, with further evidence of population differentiation occurring between later culls in Gwynedd and Cloceanog, suggesting a mechanism for how migrants can introduce new diversity into a population at arrival. It seems likely that the removal of animals creates a space whereby animals from nearby areas migrate to fill, and while Signorile et al (2016) found that grey squirrels exhibited relatively levels of migration, it seems that the removal or reduction of adjacent populations, encourages inward migration. In a conservation context, this mechanism is called a genetic rescue, whereby even a small number of new immigrants can contribute additional genetic variation to the remnant population reducing the risk of extinction (Whiteley et al 2014). In the case of the invasive grey squirrel, it may be a mechanism which supports and protects the genetic diversity of the species following intensive control efforts.
The animals that dispersed onto and were culled in Anglesey, although small in number, did not differ from the populations in Gwynedd or Clocaenog, and likely originated from both areas. The most recent animal culled in Anglesey in 2020 clustered with the animals culled from Gwynedd in 2014, suggesting that while this cluster appeared to have been removed by earlier culls, it is likely that it was not fully removed and may be recovering.
Overall implications and recommendations
We processed grey squirrel samples that had been culled in north Wales between 2011 and 2020. This culling effort was costly both financially and in terms of manpower to attempt to remove grey squirrels from the woodlands of Gwynedd and Clocaenog and all surrounding areas, and our results have indicated that the population is not likely to be exterminated in the near future without sustained regular culling efforts. The likelihood of new animals moving into a culled area is also high, as we demonstrated in our microsatellite data analysis, that once an area was cleared of a population, a new and genetically distinct population moved in to occupy that empty space. The high levels of mtDNA diversity, even within a small area of north Wales adds further genetic variability to the population. The implications of this means that control efforts have to take place at regular intervals to ensure that new squirrels do not invade a cleared area, requiring further financial and labour supports.
Despite the broad ecological similarity and complete niche overlap between the invasive grey squirrel and the native red squirrels, it is notable that both having been introduced, reintroduced, and translocated multiple times in Britain, the red squirrel continues to be much more susceptible to the negative effects of a decrease in population size than the grey squirrel. However, the grey squirrel appears to be able to recover from a reduction in population size as it increases migration and gene flow. This ability appears to be one of the biological factors that makes the grey squirrel such a successful invasive species (e.g. Shuttleworth et al. 2020).
One of the ways invasive species are thought to be so successful is because they often do not encounter stress to the same degree that they do in their native environment and may be released from predatory pressures. This can result in an increase in fitness which can counter other eco-evolutionary processes normally experienced by species in their native environments such as inbreeding (Colautti et al 2017). Intriguingly, recent studies in Britain and Ireland have shown that the presence of a predator, the pine marten (Martes martes) (Sheehy & Lawton 2014, Twining et al 2020, 2021) is associated with a reduction of grey squirrel populations and thus facilitates a natural recovery of the red squirrel. As a result, a number of pine marten reinforcement projects have taken place in Britain (Sheehy et al 2018; McNicol et al 2020a, 2020b). Captive bred pine martens have been released in the Gwynedd area of our study (Bamber et al 2020). A combination of increased predator presence and an improvement in habitat may provide the best circumstances to maintain the Welsh red squirrel population, while actively reducing the grey squirrel. The investment in the restoration of predatory species like the pine marten to the north Wales countryside may provide benefits not only to the conservation and management of native and invasive species but will provide further increases in biodiversity benefiting the wider ecosystem and society in general.