Investigating the Potential Origins of a Newly Established Goosander (Mergus Merganser) Breeding Population in Ireland

ABSTRACT

The presence of goosanders (Mergus merganser) was first recorded in Ireland in 1970, and they have been slowly increasing in numbers since. A small breeding population is now established in Wicklow Mountains National Park, County Wicklow, although the origin of the colonising individuals is currently unknown. This study amplified the mitochondrial COI gene from one goosander chick tissue sample and one scat sample from Ireland and four goosander scat samples from the River Deveron in Scotland. Phylogenetic analysis showed that the sequences generated from the Irish tissue and scat samples matched 100% to both the Scottish scat sequences and known Scandinavian sequences, indicating that the Irish goosander population likely spread from Britain and/or Scandinavia, perhaps due to increasingly favourable breeding. It is likely that avian colonising events will continue due to climate change. As such it is important that future conservation and biodiversity decisions are scientifically informed, in order to better understand changes in range and breeding habitats. Other than the tissue sample analysed, this study also validated non-invasive avian scat sampling as a valuable tool for species identification, without the need for invasive or destructive sampling.

INTRODUCTION

The Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services estimates that current global extinction rates exceed the average over the past 10 million years, and that this loss of species is likely to be further accelerated unless action is taken to tackle key drivers of biodiversity loss (Brondizio et al. 2019). Worldwide, approximately half of bird species are in decline (Lees et al. 2022), and across the European Union, breeding bird numbers have decreased by an estimated 18% since 1980 (Burns et al. 2021). In Ireland, 54 bird species, approximately a quarter of the Irish bird fauna, are assigned to the International Union for Conservation of Nature red list (Gillbert et al. 2021). Habitat loss, linked to long-term trends of increasing temperature in Ireland (Mateus and Potito 2022), is likely to play an important role in many of these threats (Colhoun and Cummins 2013). However, implementation of effective conservation and management policies is generally hindered by insufficient data for many species (Kindsvater et al. 2018).

Yet there are some bird species, for example the great spotted woodpecker (Dendrocopos major), that are expanding their historical ranges, slowly colonising Ireland (McComb et al. 2010; McDevitt et al. 2011). This is possibly in response to an expansion of favorable habitats and breeding conditions in Ireland (Parslow 1967; Harrison et al. 2003; Parkin and Knox 2009). Previous analysis of mtDNA from blood and feather samples for Irish populations of great spotted woodpeckers demonstrated a British origin for the colonising group (McDevitt et al. 2011). Investigating the origins of colonising populations can help inform conservation efforts of native species by assessing habitat needs and competition with potential colonising species (Heath et al. 2000), providing critical information for management and conservation (Hiley et al. 2013; Siegel et al. 2014).

Goosanders (Mergus merganser) are large-bodied sea ducks (Mergini, Anatidae) (Donne-Goussé et al. 2002) with a classic Holarctic distribution across Eurasia, the northern United States and Canada (Hoyo et al. 2013; BirdLife International 2025). They are principally piscivorous, although they will also take a range of aquatic prey, such as mollusks, crustaceans and insect larvae (Hoyo et al. 2013). They breed throughout the northern extent of their geographic range (Fig. 1), migrating south during the winter (Snow et al. 1998). The European portion of their range is estimated to support between 130,000 and 210,000 mature individuals (European Commission: Directorate-General for Environment 2022). Goo-sanders are currently expanding their geographic range in Europe (Keller 2009; Kajtoch and Bobrek 2014; Bordignon et al. 2018; Marchowski et al. 2022). Historical observations show goosanders first nesting in Scotland in 1871 (Baxter and Rintoul 1922). They had become a common nesting species in northern mainland Scotland within twenty years (Mills 1962), and gradually expanded their range throughout Scotland and northern England (Mills 1962; Parslow 1967). By 1970, there were recorded breeding populations in Wales (Meek and Little 1977), with estimates of up to 2,900 breeding pairs in Britain by the late 1990s (Gregory et al. 1997).

Fig. 1.

Breeding records of goosanders from Europe. Data taken from the European Breeding Bird Atlas 2 website (Keller et al.2020; EBCC 2022), showing European locations where goosanders have been recorded breeding; retrieved 18 June 2025.

Goosanders are uncommon in Ireland, although they are becoming more common as both wintering and breeding birds (BirdLife International 2025). Ireland historically represents the extreme western limit of the European range of goosanders (Meek and Little 1977), where they are typically summer migrants. There is some evidence that winter numbers in Ireland have been increasing since the 1980s (Balmer et al. 2013; BirdWatch Ireland 2025), and it is possible that some of the birds recorded in Ireland during the winter period represent a locally-breeding population (Lewis et al. 2019). The upper lake in Glendalough, County Wicklow remains the only known breeding site in Ireland (Wicklow Mountains National Park 2021), with two confirmed breeding events recorded between 2007–11 (Balmer et al. 2013) and eleven over-winterings in 2017/2018 (Fitzgerald et al. 2021).

The origin of the goosander breeding population in Glendalough Lake, County Wicklow has not been investigated to date. They have been hypothesised to belong to the breeding population found from Britain to western Russia (Lewis et al. 2019). However, other bird species migrate between Greenland and Iceland and Ireland and the British Isles (e.g., pink-footed goose; Anser brachyrhynchus) (Snow et al. 1998), and Holarctic birds have expanded ranges between Eurasia from North America via Greenland and/or Iceland (e.g., members of the herring gull species complex) (Liebers et al. 2004). As such, a potential relationship to North America populations cannot be excluded.

This study also examined the use of non-invasive sampling from scat to collect DNA. By relying on discarded DNA, non-invasive DNA techniques allow non-destructive sampling of genetic material and do not require capturing or directly interfering with focal species. This is desirable for rare or endangered taxa from small or transient populations. Fecal material contains DNA fragments from both the donor organism, principally from the intestinal epithelium, and potentially consumed plant and animal material (Tighe et al. 2018; Meyer et al. 2020; Tighe et al. 2020). Targeted primers can be used to extract and amplify DNA from either the animal itself or prey species it has consumed to address questions of phylogeny, population structure or ecological/trophic structure (Bellemain et al. 2005; Forgacs et al. 2019; Meyer et al. 2020; Bach et al. 2022; Treloar et al. 2023). Here, we evaluate hypotheses of the origins of the nascent Irish breeding population of goosander, via populations in northwestern Europe versus populations in northeastern North America, through a phylogenetic analysis of goosander mtDNA.

Materials and methods

Sample collection and storage

One goosander tissue sample was obtained from a chick found dead by the Irish National Parks and Wildlife Service, at the upper lake in Glendalough, in April, 2020. Further to this, five scat swab samples were collected, one from the resident goosander population in Wicklow National Park, Ireland in November 2020 and four from the Deveron River in Scotland, United Kingdom, in May 2021. The Scottish samples were collected during the March to May breeding season to avoid potentially sampling scat from migratory individuals (Table S1). Sterile swabs (Deltalab) were used to swab scat samples in the field. The swab tip was placed into sterile polypropylene tubes (Sarstedt 15ml), preloaded to the 3ml mark with silica beads (2.5–6.0mm – Fisher-Scientific) and kept at −20°C until DNA extraction.

Dna extraction and sequencing

Tissue and scat samples were subjected to a Chelex (Chelex 100 Bio-Rad Laboratories, Hercules, CA) DNA extraction protocol (Estoup et al. 1996). Each swab was placed in a sterile 1.5ml Eppendorf tube into which 12μl of proteinase K (20mg/ml) was added. Whilst on a stirrer plate, 400μl of Chelex (10% solution – Bio-Rad, analytical grade Chelex 100 resin) was then added. The samples were vortexed for 20 seconds and incubated at 56°C for 2 hours. Samples were then vortexed for 20 seconds and further incubated at 100°C for 15 minutes. Subsequently, samples were vortexed for 20 seconds and centrifuged for 1 minute at 20°C and 17,95g. A negative control of double distilled purified water was included in each extraction. The supernatant was removed to a sterile 1.5ml Eppendorf tube and stored at -20°C. A 749bp amplicon of the mitochondrial cytochrome oxidase subunit 1 (COI) gene was amplified using the primers BirdF1 (5’-TTC TCC AAC CAC AAA GAC ATT GGC AC-3’) and BirdR1 (5’-ACG TGG GAG ATA ATT CCA AAT CCT G-3’) (Hebert et al. 2004). A 30μl PCR master mix was prepared, in a UV-sterilised hood, consisting 3μl of Buffer (Kapa Biosystems), 1.2μl of 10mM dNTPs (Invitrogen), 1.2μl of each primer (10μM) (Integrated DNA Technologies), 0.12μl Taq polymerase (Kapa Biosystems), 1.2μl of Bovine Serum Albumin (BSA) (20mg/ml) (Thermo Scientific), 21.08μl of dd H₂O and 1μl of template DNA. PCR conditions were as follows: Initiation at 95°C for 2 minutes, followed by 40 cycles of 95°C for 30 s, 55°C for 30 s and 72°C for 1 min, with a final extension for 10 min at 72°C. Post PCR, samples were subject to electrophoresis on a 1% agarose gel stained with SYBR© Safe (Life Technologies) and amplicon size was assessed using a 100bp (Solis BioDyne) DNA ladder. The presence of an amplicon in the gel, at the correct molecular weight, was considered a successful amplification. Samples were subjected to commercial Sanger sequencing (Macrogen) in both directions. Following sequencing, forward and reverse strands were aligned and trimmed in Geneious version 10.2.3 using the Geneious alignment tool (Kearse et al. 2012). The resulting consensus sequences were then passed to BLASTn (https://blast.ncbi.nlm.nih.gov/Blast.cgi) for validation.

Phylogenetic analysis

We tested phylogenetic hypotheses of relatedness of the Irish breeding population of goosanders with other goosander populations around the world. We constructed a phylogenetic data matrix with COI sequences from the six Scottish and Irish samples (Table S1), with 17 COI goosander sequences downloaded from GENBANK. This constitutes the currently available COI sequence data for M. merganser. To this, we added data for four species in the Mergini (Table S2), with the common eider (Somateria mollissima) used as the outgroup (Donne-Goussé et al. 2002; Huang et al. 2016; Chen et al. 2023). Sequences were aligned with Clustal Omega (Sievers et al. 2011; Sievers and Higgins 2018). The resulting alignment was 720bp in length, and ModelFinder (Kalyaanamoorthy et al. 2017) chose the Tamura-Nei substitution model with empirical base frequencies and estimated invariant sites (TN+F+I) (Tamura and Nei 1993). Maximum-likelihood tree search with 1000 ultrafast bootstrap replicates (Minh et al. 2013; Hoang et al. 2018) was performed in IQTREE (ver. 1.6.12) (Nguyen et al. 2014).

Results

DNA was successfully extracted and amplified from the goosander tissue and all five scat samples. These samples showed a clear single amplicon at the correct molecular weight. Following Sanger sequencing, alignment and editing, unambiguous sequences were generated for the scat and tissue samples from Glendalough and the four scat samples from Scotland, with individual sequence lengths ranging from 569bp to 720bp. BLASTn validation verified that the five Irish and Scottish scat samples are a match to the tissue sample from Glendalough Lake, and that these match known goosander COI sequences with 100% matches to previously accessioned sequences on Genbank. The six COI sequences generated in this study were uploaded to Genbank (Accession numbers: PV815598 - PV815603).

The results of the phylogenetic analysis (Fig. 2) finds that the hooded merganser (Lophodytes cucullatus) nested within the genus Mergus, in agreement with previous work (Donne-Goussé et al. 2002; Huang et al. 2016; Chen et al. 2023). As such, some taxonomic revision within the Mergini is likely in order, with the genus Mergus being split to accommodate this paraphyly, or with hooded mergansers subsumed within Mergus, as in older taxonomies (Donne-Goussé et al. 2002).

The phylogenetic analysis resolves goosanders into three well-supported subclades. Irish and Scottish goosander COI sequences group with those from Scandinavia (Fig. 2, Fig. S1), and this European section of the goosander phylogeny is related to goosander samples from China and Korea. Internal resolution within this group is limited (Fig. 2a), with minimal branch length differences within this group (Fig. 2b), although the European goosanders do appear distinct from samples from east Asia. The lack of genetic information for populations between northwestern Europe and China limits the ability to resolve any hypotheses of finer scale structure. Distinct from the Eurasian group are two goosander clades: 1) North America and 2) Japan/Alaska (Fig. 2). This distinction parallels, although is not completely congruent with, the accepted division of M. merganser into Western Eurasian, East Asian and North American subspecies (Snow et al. 1998).

Discussion

Although limited by sample size, the sequences generated from the Irish and Scottish scat samples match both the tissue sample from an Irish goosander, as well as known, vouched goosander sequences on Genbank. Therefore, our results suggest that DNA from avian scat samples represents a viable nondestructive and non-invasive survey tool that can be employed to aid management and conservation decision making.

The phylogenetic analysis suggests a northwestern European origin of the Irish breeding population, based on the two individuals sequenced. Irish and Scottish goosander COI sequences cluster with those from Scandinavia, and this group of sequences, in turn, are most closely related to individuals from China and Korea. Importantly, this clade is phylo-genetically distinct from sequences from the clade uniting the COI sequences from North America and the Pacific Rim. While the internal branch lengths among COI sequences within each of the regional clusters are very short, and there is little support for internal topologies, the regions themselves show differentiation from one another. Our results suggest that the current Irish population is sourced from the UK/Scandinavia rather than a transatlantic dispersal. It remains to be seen whether COI sequences from Icelandic goosander populations cluster with those from Ireland/Scotland/Scandinavia, however previous research in which the mitochondrial control region and various nuclear markers were sequenced suggests that, whilst the Icelandic population is a member of the European sub-species of goosander, it is largely genetically distinct from other Northern European populations (Hefti-Gautschi et al. 2009).

Bird species in the northern hemisphere are changing their geographic ranges in the face of a changing global climate, with most of the change typically being a northward shift in breeding sites (Marchowski et al. 2022). However, in response to global warming, goosanders have paradoxically expanded southwards into parts of Europe such as Poland and the Balkans, taking advantage of the earlier thawing of ice cover (Dobrev et al. 2020; Marchowski et al. 2022). In the specific case of Ireland, range expansion for goosanders may reflect recent trends of reforestation (DAFM - Department of Agriculture Food and the Marine 2024), as is thought to be the case for great spotted woodpeckers, as both species prefer mature woodland habitat for breeding (McDevitt et al. 2011; Kočí and Krištín 2023). In addition, there is a general trend toward westward expansion in northern Europe of wet-lands birds in response to global warming (Soultan et al. 2022). These results suggest that the arrival of goosanders in Ireland marks a westward expansion of this species from Scandinavia and/or Britain and appears to be a continuation of this pattern. The present-day geographic range of goosanders now encircles the northern continents (BirdLife 2025). A simple linear concept of a ring species around the Arctic (as in Mayr 1942) is likely not an accurate representation the dispersal of goosanders (Liebers et al. 2004). Rather the present distribution around the northern hemisphere is likely the culmination of a set of overlapping dispersal events. Unfortunately, it is not possible to pinpoint the ultimate geographic origin of Mergus merganser with the lack of resolution among the three principal COI clades (Eurasian, North American, and Northern Pacific). Further sampling will be required, both in terms of genetic loci, such as SNPs or whole genome sequencing, and geographic spread, to increase the resolving

Fig. 2.

Phylogeny of goosander COI sequence data. A) Maximum likelihood topology with bootstrap values >50% for internal nodes indicated. B) Maximum likelihood topology with branch length data. Outgroup sequences given in black. Goosander sequences are highlighted by geographic group: Eurasia in red; Scandinavia/Britain in blue, with Ireland samples in green; Pacific Rim in purple; North America in orange. See text for details. While topologies within each of these groups are ambiguous due to short internal branch lengths, geographic regions form distinct cohorts. The Irish goosander samples clearly fall within the group from Scandinavia and Great Britain. And this group is clearly distinct from North American and the Pacific Rim.

power of the phylogeny to determine potential dispersal routes of this widely-distributed species (Liebers et al. 2004).

As global warming and climate change are predicted to continue, conservation managers will need to understand the mechanisms of change and how they may impact on species behavior and range (Traill et al. 2010). Effective bird conservation must not only consider the current vulnerability of a species/habitat, but also assess future challenges, to confront emerging climate related threats to habitat management (Siegel et al. 2014). Establishing the origin of expanding bird populations is one element in this understanding of expansion mechanisms.