The efforts of the acclimatisation movement of over a century ago are perhaps best known for the introduction and establishment in many countries of two bird species that developed into serious pests: the House Sparrow (Passer domesticus) and European Starling (Sturnus vulgaris). Another species was also widely introduced during this period, the European Goldfinch (Carduelis carduelis). This attractive member of the Fringillidae is native to Europe, western Asia, and extreme northern Africa. Multiple subspecies are divided into two groups: the western, black-crowned C. c. carduelis group of west and central Europe and the eastern, grey-crowned C. c. caniceps group of west and central Asia (Cramp and Perrins 1994; Clement et al. 2020).

European Goldfinches were introduced to the Azores around 1860 (Clarke 2006), Australia in 1863 (Acclimatisation Society of Victoria 1863; Ryan 1906), New Zealand in 1864 (Anonymous 1864), Bermuda in 1885 or 1893 (Prentiss 1896; Verrill 1902) and Uruguay in 1913 (Dias 2000). Populations in all these areas persist today, some of which have spread to nearby islands or countries (Dias 2000; Codesido and Drozd 2021).

The first introduction of European Goldfinches to North America was facilitated by Thomas S. Woodcock (Woodcock 1852). In late 1852, he brought 168 songbirds, amongst them 48 European Goldfinches, to New York City which were held in captivity over the winter (Woodcock 1853). Those that survived, including 16 European Goldfinches, were released on 20 April 1853 in Green-Wood Cemetery in Brooklyn (Woodcock 1853; Cleaveland 1866). None of the released birds was seen past late summer (Cleaveland 1866). The species first appeared in New York City’s Central Park in 1879 (Adney 1886) purportedly having crossed the Hudson River from Hoboken, New Jersey, where birds were said to have been released the previous year (Eaton 1914; Cruickshank 1942 Austin 1963). European Goldfinches were later reported as common in parts of New York City, with a maximum report of 50 in winter 1902 (Abbott 1902). Griscom (1923) wrote that they were gone from Central Park by 1907 and that very few were reported anywhere in the New York City area by the early 1920s. They did continue to be reported from Long Island in Nassau County, New York (Nichols 1936; Lincoln 1998). The highest numbers published for that area were 17 in 1938 (Eynon 1940) and ~ 24 in the mid-1940s (Lincoln 1998). The population there persisted until the 1950s when they disappeared due to habitat changes related to development (Elliott 1968). European Goldfinches were also reported as common in eastern Massachusetts by 1880 (Brewer 1879; Allen 1880), but apparently never achieved a self-sustaining population (Brewster 1906; Strohbach et al. 2014).

Elsewhere in North America, a temporarily successful introduction occurred in Oregon, where 40 or more pairs were introduced around Portland in 1889, 1892 and 1907 (Anthony 1891; Pfluger 1896); they thrived for some years, but did not persist (Jewett and Gabrielson 1929). Unsuccessful attempts to establish European Goldfinches in North America occurred in St. Louis, Missouri in 1870 (Widmann 1907); Cincinnati, Ohio in 1872–1874 (Langdon 1881); and Vancouver, British Columbia in 1908 or 1910 (Carl and Guiguet 1972).

Over the latter half of the 20^th century, scattered European Goldfinch sightings in North America were considered released or escaped birds. In the western Great Lakes region, Wisconsin had four published records prior to the 1990s (Jung 1936; Lound and Lound 1956; Frank 2004). In Illinois, up to seven European Goldfinches were reported in September 1953 in Chicago’s Lincoln Park, Cook County, including adults and dependent young (Binford 1993). These furnished the first Illinois records (contra Smith and Parmalee 1955).

Around 2001, sightings of European Goldfinches in the western Great Lakes region of the United States began to increase beyond occasional reports, a situation suspected at the time to originate from a cage bird dealer in Illinois (Dinsmore and Silcock 2004). Craves (2008) compiled records through mid-2006 from this region, revealing nest building activity in 2003 and successful fledging of young in 2005. In the decades since, European Goldfinch reports have continued to increase, but a full treatment of their status is lacking.

Here, we compile occurrence records of European Goldfinches from 2001–2021 to describe their: 1) distribution, 2) breeding status and phenology, 3) natural food sources and 4) apparent introduction pathway. We also briefly discuss their potential ecological impacts. Our intent is to summarise the early phase of establishment and to prompt formal study of this species in North America.

We compiled 7120 records of European Goldfinches across North America from 2001–2021 (Fig. 1). Observations were recorded in 41 U.S. States and nine Canadian Provinces (Suppl. material 1). The eastern, grey-crowned C. c. caniceps race made up < 1% of the entire dataset. The majority of records were from eBird (82%) and those directly received by JAC (9%). The bulk of the eBird records (88%) were accumulated from 2015–2021, while most of the directly reported records (71%) were from 2001–2006.

Figure 1.

Locations of records of European Goldfinches in North America, 2001–2021. A table of the number of records by U.S. State and Canadian Province with county totals is in Suppl. material 1.

Forty-seven percent of all records were from the western Great Lakes region and the majority of those (n = 2919) were from Illinois and Wisconsin (Fig. 2). Similar to the overall dataset, 90% of the records for this region were from eBird and those directly received by JAC. The bulk of the eBird records (90%) were accumulated from 2015–2021, while most of the directly reported records (74%) were from 2001–2006. Of the 2467 eBird records for the western Great Lakes region, 80% were “complete” checklists, in which the users indicated they recorded all species they were able to identify. Ninety-four percent of these complete checklists were from 2015–2021. About 40% of all records from the western Great Lakes region were from the last week in March through May. Only about 4% of records were from mid-August through October.

Figure 2.

Number of records of European Goldfinches in the western Great Lakes region, 2001–2021. Stacked bars show proportions from the core population in Illinois and Wisconsin, and all other States (Minnesota, Iowa, Missouri, Indiana, Kentucky, Michigan, Ohio, and Ontario west of -81°longitude).

The general pattern of records for the years 2007–2014 for both the overall dataset and the western Great Lakes region was a decreasing number of directly reported records and records in the grey literature and an increasing number of eBird records.

Initial records were loosely clustered in north-eastern Illinois and south-eastern Wisconsin with scattered outliers in neighbouring States. From 2004–2006, there were records across the entire region, after which outlying records declined, but continued in the vicinity of the original cluster (Fig. 3). Since about 2018, records were concentrated between the cities of Milwaukee, Wisconsin and Chicago, Illinois and mostly within 15 km of Lake Michigan.

Figure 3.

Records of European Goldfinches in the western Great Lakes region grouped in 3-year increments. Some points in these panels are outside our geographic boundaries of this region due to map projection distortions. An animated depiction of the accumulating records over time can be viewed at https://tinyurl.com/bdz8vxt3.

In the western Great Lakes region, every year in our time range had records with counts of at least two birds. Beginning in 2009, each year had records reporting counts of five birds. From 2013 on, each year had counts of at least 10 birds. Starting in 2016, each year had records of 20 or more birds. All counts of European Goldfinch ≥ 20 individuals were in Illinois or Wisconsin and nearly all were in winter (November through February). There are six records of ≥ 50 European Goldfinches with the highest single count being 95 on 1 January 2021 in Lake County, Illinois. There was a significant increase from 2001 to 2021 in the mean count of birds per record (0.12 ± 0.02 [SE]; R² = 0.64, t = 5.8, df = 19, P ≤ 0.001). The mean number of European Goldfinches per record in 2021 was 3.6.

We assigned breeding evidence codes for 2320 records in the western Great Lakes region from the months of March through August. Nineteen percent of these records were from March, 26% from April and 28% from May. The number of records decreased over the rest of the breeding period, with 12% in June and the remainder in July and August. The majority of these records were assigned codes in the Observed or Possible categories and, therefore, did not provide strong evidence of breeding (eBird 2021). The remainder (n = 274) were coded in the Probable or Confirmed categories. None of these birds was noted as C. c. caniceps either by the observer or in our review of media.

The majority (n = 268) of Probable or Confirmed records were located in northeast Illinois and southeast Wisconsin (Fig. 4). There was a single breeding record in the State of Ohio that falls outside the extent of Fig. 4: a pair of goldfinches frequented a feeder in Cuyahoga County in 2015, where an adult and immature were photographed on 22 August 2015. We plotted seven Probable records coded P (Pair in Suitable Habitat) from 2006 separately in Fig. 4; although pairs were initially present in these quadrangles, no breeding confirmations were ever reported in these areas and no records of multiple birds occurred in years after the initial reports. They are not included in the totals below.

Figure 4.

Locations with breeding evidence of European Goldfinches in the western Great Lakes region. From records categorised as Probable or Confirmed, the highest breeding evidence is plotted in USGS 7.5-minute quadrangles. Black represents Confirmed, grey represents Probable, crosshatched represents European Goldfinch pairs (Probable) recorded in 2006 in quads where no records of multiple birds occurred in later years. Quads may contain multiple breeding locations and some quads overlap county lines. A confirmed breeding record in Ohio is outside the map boundaries and not shown (see Results).

Illinois had 155 records with high-level breeding evidence. European Goldfinches were confirmed in 13 quads in four counties and coded as probable in five additional quads and two more counties. The majority of the records (n = 141) were from Lake County. Breeding evidence, a pair constructing a nest, was first reported in the State in 2003, there have been Probable and/or Confirmed records in the State every year since.

In Wisconsin, there were 113 records coded Probable or Confirmed. European Goldfinches were confirmed breeding in 16 quads in seven counties. They were coded as probable in four additional quads. Most records were from Racine County (n = 62) and neighbouring Kenosha County (n = 33). Breeding evidence was first reported in Wisconsin in 2004 and there have been Probable and/or Confirmed records in the State every year since 2009. A link to a map file showing the Confirmed records with county boundaries is in the Suppl. material 2.

Records showed that nest building was initiated as early as 1 March, with nest building activities extending to 27 July (Fig. 5). Occupied nests were reported from 13 May though 8 July, and fledged young were reported from 22 May through 30 August (Fig. 5).

Figure 5.

Phenology of breeding activities of European Goldfinches in the western Great Lakes region. Black vertical lines represent median dates. Boxes represent 1^st and 3^rd quartiles. Whiskers represents minimum and maximum dates, with solitary dots representing outliers. Breeding codes are as follows: FL (Recently Fledged Young), FY (Feeding Young), ON (Occupied Nest), N (Visiting Probable Nest Site), A (Agitated Behaviour), NB (Nest Building), CN (Carrying Nesting Material, C (Courtship Display or Copulation), P (Pair in Suitable Habitat), T (Territorial Defence), S7 (Singing Bird Present 7+ Days). Full definitions and application of these codes can be found at eBird (2021).

Our examination of media associated with European Goldfinch records resulted in 125 observations of European Goldfinches utilising natural food sources in eastern North America. We identified all plants to at least genus (Suppl. material 3, which provides scientific names) and all 16 plant taxa occur in the western Great Lakes region (USDA NRCS 2022). Ten non-native plant taxa, all of Eurasian origin, made up 80% of the observations. The remaining six plant taxa are native to North America, but represent species or genera that have been introduced from North America into at least parts of the native range of European Goldfinches (Royal Botanical Gardens Kew 2022). Most observations (89%) and most plant taxa (69%) were herbaceous plants and the rest were trees. All observations were of European Goldfinches eating seeds, except for a video of a bird eating the flower buds of a bigtooth aspen tree.

Import data from LEMIS indicate that nearly 159,000 European Goldfinches were imported into the U.S. for the purpose of commercial trade from 2000–2014. This number is conservative, as it does not include potential European Goldfinches which may have been amongst the > 16,000 birds listed under Carduelis sp. or the thousands of birds listed under even more generic terms. The country of origin of most birds, 60%, was given as Australia, all of which were coded as captive-bred. Another 34% were from Russia, of which 67% were coded as captive-bred and the rest having been taken from the wild (Fig. 6).

Figure 6.

Country of origin and number of European Goldfinches imported into the United States for the purpose of commercial trade, 2000–2014 (all available data). Dashed line represents birds imported from Australia by two California companies. Bars represent birds imported by a single Illinois company. Not shown are < 1700 birds imported from Canada, China, and New Zealand by four other entities.

Over 99% of the European Goldfinches were imported by three entities. The largest quantity, 39%, was imported by a company headquartered in McHenry County (Illinois Secretary of State 2022). This company was the sole importer of the European Goldfinches from Russia, receiving > 53,000, all in the years prior to 2006. From 2008–2013, the Illinois company received ~ 6100 wild-caught birds from Uzbekistan and 50 captive-bred birds from the same country in 2014. They also received a small number, < 3% of their total imports, of captive-bred birds from Spain ≥ 2005.

Another 30% of the total were imported by a California company which acted as the sales office of the Illinois company (California Secretary of State 2022; Moody’s Analytics 2022), all were from Australia. A second California company received 30% of total imports, which were also all from Australia, with the exception of 500 captive-bred birds from New Zealand.

The initial increase in reports of European Goldfinches in the western Great Lakes region began around 2002. This was followed by a large increase in the number of European Goldfinch and other non-native cage bird records in the Great Lakes region over the years 2004–2006 (Fig. 2, David 2005; Craves 2008). At that time, it was rumoured that an Illinois dealer (the same one we identified through import records) was either purposely releasing birds or had lax husbandry practices (Dinsmore and Silcock 2004; Fiske 2004). We believe a global outbreak of pathogenic avian influenza H5N1 may have also played role. The outbreak originated in Asia in late 2003 and wild migratory birds were implicated in its spread (Gauthier-Clerc et al. 2007). In 2005, there was a total embargo imposed on imports of birds from Russia (HHS/CDC 2006), the source of nearly all European Goldfinches imported by the Illinois company up to that point (Fig. 6). We speculate that a substantial number of European Goldfinches and other species may have been released by the Illinois company during this period due to concerns about captive birds having, getting, or transmitting avian influenza; the potential need for additional inspections or health screening; and/or possible restrictions on sales (Gilardi 2005; Senni 2005; Romagosa 2015).

Reports diminished between 2007 and 2014. This may have been due to the discontinuation of solicited records by JAC in 2006, or an actual decline in the number of birds in the region. Beginning in 2006, the number of European Goldfinches imported by the Illinois company was also greatly reduced because of the import embargoes, from a mean of around 8900 birds annually prior to 2006 to fewer than 900 a year thereafter (Fig. 6). If the Illinois company was the source of the European Goldfinches in the region, perhaps the reduced inventory resulted in fewer releases or escapes.

Beginning in 2015, the number of reported observations began to increase substantially in the western Great Lakes region, mostly in Illinois and Wisconsin. In part, this may be attributed to the rising usage of eBird, including the debut of their mobile app (Team eBird 2015). European Goldfinch records were also approved for public display on eBird maps in some parts of Wisconsin and Illinois around this time, thus increasing awareness. Additionally, participants in the Wisconsin Breeding Bird Atlas were specifically instructed to watch for the species and report them via the eBird atlas portal (Anich and Craves 2015).

The steady increase in reports may also reflect the start of a period of population growth after an initial lag phase (Crooks 2005; Aagaard and Lockwood 2014), similar to the pattern that occurred after the introduction of the House Finch (Haemorhous mexicanus) into eastern North America ~ 1940. After a ban on the sale of these protected migratory birds, native to western North America, at least one New York area dealer apparently released their stock (Elliott and Arbib 1953). The founding population was estimated at 80 birds (Veit and Lewis 1996). The eastern population of House Finches was estimated at 280 birds in 1951 (Elliott and Arbib 1953). Numbers increased 10-fold from 1962–1971, coincident with rapid range expansion (Bock and Lepthien 1976; Veit and Lewis 1996).

The distribution of European Goldfinch records from 2001–2006 (Fig. 3), appears to show widespread dispersal from north-eastern Illinois and south-eastern Wisconsin, a pattern that supports our theory of a substantial release of birds around 2005 originating in that area. The reduction in records and apparent geographic contraction that followed may represent the failure of scattered individuals to establish satellite populations away from the northern Illinois source. There is a possibility that the observed pattern might also be influenced by the increasing number of eBird records which are concentrated closer to population centres and popular birding sites and the discontinuation of solicited records, many of which came from more rural counties that have less eBird coverage.

With well over a decade of continuous breeding in Illinois and Wisconsin, it appears that European Goldfinches are establishing a self-sustaining population in this area. Our data indicate the nesting ecology of European Goldfinches in the western Great Lakes region is similar in many respects to those of previous North American populations as well as in the native range.

Cruickshank (1942) gave egg dates ranging from 26 April to 4 June for birds in the New York region. Our records of nest construction activities in March suggest a slightly earlier commencement of breeding, perhaps due to an advancement in egg laying over the ensuing decades. This has occurred in Britain, where the mean laying date of European Goldfinches moved up 20 days over the period 1968–2019 (Walker et al. 2020).

Cruickshank (1942) and Elliott (1968) reported several nests being built or with eggs in July. From this, Cruickshank concluded European Goldfinches were single-brooded, while Elliott felt these late nests indicated double-brooding. We had just two records of nests being constructed or occupied after late June in our dataset. Given that only 26% of our breeding season records were from June–August, we hesitate to draw any conclusions on how many broods are attempted by European Goldfinches annually in the western Great Lakes region.

Various methods have been used to study the diets of wild European Goldfinches or granivorous passerines, including prolonged field observations, gut flushing, and post-mortem stomach content analysis (Middleton 1970; Campbell 1972; Moeed 1975). The use of web-sourced photographs has recently been used to explore the diets of birds (Naude et al. 2019; Panter and Amar 2021). This method may suffer from bias, such as being photos more likely to be taken along roadsides, in open habitats, or in lower vegetation levels, but it can provide insight into a time- and cost-effective manner (Berryman and Kirwan 2021; Panter and Amar 2022). Our compilation of natural food sources derived from media associated with records from eastern North America (Suppl. material 3) suggests European Goldfinches may prefer the seeds of plant species not native to this region, but indigenous to their native range.

European Goldfinches are nearly entirely granivorous, specialising in the seeds of composites (Asteraceae) in their native range, strongly favouring thistles, burdocks, knapweeds, dandelions, and ragworts (Senecio spp.) (Newton 1967; Newton 1972 Cramp and Perrins 1994). Half of the plant taxa noted in our media review were composites, including all the taxa listed above, except for Senecio spp. All non-native taxa we identified except the grass Bromus inermis have been reported as food items by European Goldfinches in their native range (Newton 1967; Newton 1972; Cramp and Perrins 1994). A reliance on introduced Eurasian plants is a pattern noted in other regions where European Goldfinches have become established (Middleton 1970; Campbell 1972; Dias 2000). European Goldfinches that were previously established in North America in the New York region reportedly fed on burdocks, thistles, grasses, various garden composites, grasses, sweetgum, and larch (Larix spp.) (Elliott 1968), comparable to our findings.

There are many records of European Goldfinches at feeding stations. Many – especially those of multiple birds – are from feeders adjacent to appropriate habitats, such as parks or vacant land, including natural areas with known populations of European Goldfinches. Flocks of European Goldfinches, including juveniles, feeding on natural food sources much of the year suggest that while this species will utilise feeders, they readily adapt to and make use of widely available natural food sources and do not seem likely to be dependent on human provision of food.

We believe there is substantial evidence that the main source of the European Goldfinches in the western Great Lakes region was a dealer located in McHenry County, Illinois, ~ 80 km northwest of the city of Chicago (Illinois Secretary of State 2022). This company was the largest importer of European Goldfinches into the U.S. for the years 2000–2014, imported many other cage birds (Eskew et al. 2019) and was the only company with a direct connection to this region. In April 1998, there was a report of a European Goldfinch from rural Walworth County, Wisconsin at a home < 10 km from the McHenry County address of the Illinois importer (Frank 1998). Between 2001 and 2004, there were also records of Common Chaffinches (Fringilla coelebs), Eurasian Blue Tits (Cyanistes caeruleus) and Great Tits (Parus major) in McHenry and Walworth counties and in nearby Racine County, Wisconsin; these and other cage-bird species were generally found < 200 km from the importer’s location (David 2002; David 2004; Craves 2008). In her report of nesting Great Tits in McHenry County in 2003, Fiske (2004) suggested that the source of the birds was the McHenry County bird importer either accidentally or intentionally releasing exotic birds.

The majority of European Goldfinches imported into the U.S. from 2000–2005 were from Russia; all these birds were imported by the Illinois company. The country of origin listed in import records may not have been where wild-caught birds were actually collected, but rather the location from where birds harvested across a wider area are consolidated and exported (Eskew et al. 2020; Sinclair et al. 2021). It seems likely that many of the European Goldfinches imported by the Illinois company prior to 2006 were collected somewhere in eastern Europe and central Asia and perhaps from multiple locations throughout these regions.

Thus far, most non-native bird species have failed to develop permanent populations in their new environments (Zenni and Nuñez 2013; Aagaard and Lockwood 2016). Those species which have become successfully established have interacted with many other taxa, with varying impacts on the recipient ecosystems (Blackburn et al. 2009). For birds, the most important impacts include predation; competition for resources, including food and nest sites; hybridisation; interaction with other non-native species, usually the spread of invasive plants via frugivory (Baker et al. 2014; Evans et al. 2016); and disease transmission (Springborn et al. 2014; Martin-Albarracin et al. 2015). Despite this potential for disruption, relatively few introduced bird species, particularly songbirds, have had severe impacts, especially outside of island settings (Bauer and Woog 2011; Strubbe et al. 2011; Baker et al. 2014).

Our data offer limited insight into potential impacts of European Goldfinches on native ecosystems. European Goldfinches are not predatory and they appear to be compatible with native songbirds. Comments included with many records indicated European Goldfinches were only occasionally aggressive towards other birds and were more often described associating with other finches, especially American Goldfinches (Spinus tristis). European Goldfinches are not cavity nesters and are unlikely to compete for nesting sites.

The primary natural food sources we documented being used by European Goldfinches in eastern North America are common weedy species, such as burdocks, teasels, and various thistles which are considered invasive in North America (EDDMapS 2022). Given the large variety of widespread, abundant plants suitable for European Goldfinches and native granivorous birds, it seems doubtful that competition for food resources could be problematic. Studies of European Goldfinches and related granivores indicate they are seed predators with few viable seeds being passed through their digestive tracts (von Mariluan 1895; Roessler 1936; McCallum and Kelly 1990; Heleno et al. 2011). European Goldfinches are not frugivorous and, thus, highly unlikely to disperse fleshy-fruited non-native plants. Dispersal of troublesome plants is presumably minimal, but data on foraging ecology are needed to assess unforeseen biotic interactions.

Hybridisation is considered a potential threat introduced birds may impose on native species (Baker et al. 2014). While European Goldfinches are frequently crossed with other finches in captivity, these hybrids are apparently nearly always sterile and wild hybrids are rare (Hinde 1956a, b; McCarthy 2006), suggesting the potential for negative impacts is low. The asynchrony in the breeding cycles of American and European Goldfinches may reduce the opportunity for hybridisation between these species; in eastern North America, American Goldfinches do not begin nesting until late June with a peak in the second half of July (Nickell 1951; Middleton 1978).

Our data shed little light on the potential role of European Goldfinches as reservoirs of disease or pathogens. There were no comments in the over 7000 records we compiled nor any images in the media we reviewed suggesting a sick bird, which might indicate infection with West Nile virus or Mycoplasma gallisepticum, the bacteria that cause conjunctivitis in House Finches and related Fringillids (Farmer et al. 2005). Pathogen screening and surveillance studies are needed to determine any disease risks posed by European Goldfinches.

In other countries where they have become established, European Goldfinches are not considered a major threat. They are not listed as an invasive species in Bermuda (Bermuda Department Environment and Natural Resources 2022) or the Azores (Marchante et al. 2020) and they are not included on the New Zealand Pest Register (Biosecurity New Zealand/Tiakitanga Pūtaiao Aotearoa 2022). In Australia, they are considered a “low interest” exotic species (DCCEEW 2007). Uruguay lists them as an invasive species of medium priority undergoing risk assessment (Aber et al. 2012) with no data on impacts (Aber et al. 2014).