We describe a highly isolated population of hairstreak butterfly from Waterton Lakes National Park, Alberta, Canada, as a new species, Satyrium curiosolus sp. nov., previously recognized as Satyrium semiluna (Half-moon Hairstreak). We propose “Curiously Isolated Hairstreak” as the common name due to its disjunct and unusual distribution. Previous whole-genome analyses revealed S. curiosolus has extremely low genomic diversity and is highly divergent from the nearest S. semiluna populations in British Columbia and Montana, more than 400 km distant. Further analysis suggested prolonged inbreeding and isolation for up to ~40,000 years BP. Ecological niche modeling indicated that S. curiosolus occupies environmental conditions that are distinct from S. semiluna, suggesting niche divergence driven by long-term geographical and ecological separation. While host plant and ant associations have not been definitively resolved, they likely differ between S. curiosolus and S. semiluna. As part of this description, we provide whole-genome consensus sequences for each individual of the type series and identify 21,985 single nucleotide polymorphisms (SNPs) that are divergently fixed between S. curiosolus and S. semiluna, including 117 unlinked SNPs distributed across the genome as putative diagnostic markers. Previously listed as Endangered in Canada as the Waterton population of S. semiluna, S. curiosolus should retain this conservation status due to its extreme isolation, small population size, and flatlined genomic diversity. We propose species recognition as a testable hypothesis under the General Lineage Concept and recommend further research to explore the taxonomy, ecological relationships, and conservation of the greater species complex, including S. curiosolus, S. semiluna, and S. fuliginosa.

Butterfly, Curiously Isolated Hairstreak, endangered species, genomics, Half-moon Hairstreak, niche divergence, Sagebrush Sooty Hairstreak

The northernmost populations of a North American butterfly, the Half-moon Hairstreak (Satyrium semiluna Klots; sometimes called “Sagebrush Sooty Hairstreak”), have received recent study by MacDonald et al. (2025), but their taxonomic status remains in question. Although S. semiluna is “apparently secure” across its range in the USA (COSEWIC 2006, 2022; ECCC 2016; NatureServe 2024), the species’ northern range limit extends into Canada, where it is listed as Endangered under the “Species at Risk Act”. All but one Canadian population occur in south-central British Columbia, with an estimated aggregate abundance of 5,000–15,000 individuals. Based on continuity of both suitable habitat and the species’ occurrence records, British Columbia populations are presumably well connected to others south of the USA–Canada border and likely represent an example of political boundaries dictating protection rather than range-wide conservation concern. British Columbia populations have been recommended for downlisting to Threatened (COSEWIC 2022). The single other Canadian population persists on a ~300 ha alluvial fan (Blakiston Fan) in Waterton Lakes National Park, Alberta, where it is isolated from all other S. semiluna populations by a distributional gap of more than 400 km. This population was recently recommended for uplisting to Critically Endangered based on its uniqueness, small size, and considerable isolation (COSEWIC 2022).

The Alberta population is small, with genomically based estimates of contemporary effective population size (N_e) around 500 individuals and surveys suggesting that between 1,000 and 10,000 adults fly annually (COSWEIC 2022; MacDonald et al. 2025). Aside from enigmatic island insect populations, such as the Lord Howe Island stick insect (Dryococelus australis (Montrouzier)) (Priddel et al. 2003) and some Hawaiian drosophilids (O’Grady and DeSalle 2018), few if any other insects have been documented with such a small population size and high degree of long-term isolation. The Alberta population’s environmental and ecological associations are also unique, adding to its scientific interest. Blakiston Fan receives an average summer precipitation of ~200 mm, while the locations of all other S. semiluna populations in the central-northern portion of the species' range receive between 32 and 154 mm (mean = 71 mm) (MacDonald et al. 2025). This difference in precipitation manifests in different habitat characteristics—Populations throughout British Columbia and the USA inhabit steppe-like habitats dominated by big sagebrush (Artemisia tridentata Nutt.). In contrast, occupied habitat at Blakiston Fan is best described as prairie/grassland dominated by sedges, grasses, and herbaceous plant species.

Another possible axis of niche divergence is larval host-plant association. Populations throughout British Columbia and the USA Pacific Northwest feed on silky lupine (Lupinus sericeus Pursh) and possibly Pacific lupine (Lupinus lepidus Lindl.) (James and Nunnallee 2011), while the Alberta population feeds only on silvery lupine (Lupinus argenteus Pursh), even though L. sericeus is common at the site. Host associations of most other populations east of the continental divide in the USA are unknown and require investigation. Myrmecophily presents a third possible axis of niche divergence (MacDonald et al. 2025). Larvae of the Alberta population exhibit a mutualistic relationship with Lasius ponderosae Schär, Talavera, Rana, Espadaler, Cover, Shattuck & Vila. In British Columbia, L. ponderosae is absent in S. semiluna habitat, and larvae associate with Formica and Camponotus species. Similar associations with Formica and Camponotus have been observed in California (Runquist 2012).

Given the Alberta population’s small size and considerable isolation, inbreeding depression and loss of adaptive potential were identified by Parks Canada and the Half-moon Hairstreak Conservation Committee as likely threats to its long-term persistence. In these situations, genetic rescue is often assumed to be an effective conservation strategy (Storfer 1999; Weeks et al. 2011; Frankham et al. 2017; Ralls et al. 2020; Clarke et al. 2024). To assess whether genetic rescue is indeed appropriate for the Alberta population, MacDonald et al. (2025) generated the first chromosome-level genome assembly for the species and whole-genome resequencing data for the Alberta population, British Columbia populations, and the nearest USA population in Montana. Based on genetic divergence, environmental and ecological differences and a very long inferred history of isolation with no evidence of contemporary or recent gene flow, we, together with Parks Canada and the Half-moon Hairstreak Conservation Committee, determined that the Alberta population satisfies requirements of a distinct species that has long been on an independent evolutionary trajectory. Species-level recognition highlights the unique ecology and evolution of this butterfly, demonstrates a clear need for continued consideration under the “Species at Risk Act” and International Union for the Conservation of Nature, and provides an important case study on the utility of genomics in taxonomy. Genomics has an increasingly important role in taxonomic descriptions (Fennessy et al. 2016; Nater et al. 2017; Zhou et al. 2018; Stanton et al. 2019). However, chromosome-level genome assemblies for new species, along with whole-genome consensus sequences for the type series, remain rare (see Brandão‐Dias et al. 2022).

All types (Fig. 2) are deposited in the University of Alberta E.H. Strickland Entomological Museum (UASM).

Here we summarize the taxonomically relevant methods of MacDonald et al. (2025). Eight individuals were collected from Blakiston Fan, Alberta, four from Richter Pass, British Columbia, three from Anarchist Mountain, British Columbia, and four near Red Lodge, Montana (Parks Canada Agency Research and Collection Permit WL-2021-39,020, Nature Conservancy Canada Research Permit NCC_BC_2021_SS001, and Nature Trust of British Columbia Permit #3461). Four Alberta individuals were used to generate a chromosome-level reference genome assembly using PacBio HiFi long-read sequencing (Pacific BioSciences, Menlo Park, California, USA) and Omni-C proximity ligation (Dovetail Genomics, Scotts Valley, California, USA).

Whole-genome resequencing of individuals from Blakiston Fan (n = 4) and the geographically nearest populations from Richter Pass (n = 4), Anarchist Mountain (n = 3), and near Red Lodge (n = 4) was performed on an Illumina NovaSeq S1 platform, with a target coverage of ~20×. Reads were aligned to our reference genome assembly and used to identify millions of single nucleotide polymorphisms (SNPs). Population structure and degree of admixture was assessed using PCA and the program “structure” (Pritchard et al. 2000) and genetic divergence among inferred genomic clusters was estimated using F_ST (Weir and Cockerham 1984). Genetic diversity was estimated using individual-based heterozygosity and nucleotide diversity (π), while the proportion of an individual’s genome within runs of homozygosity over 0.1 Mb (F_ROH) served as an index of inbreeding. Historical effective population size (N_e) was inferred from each individual's genome sequence using the pairwise sequentially Markovian coalescent (PSMC) (Li and Durbin 2011).

Nuclear whole-genome consensus sequences (fastq format) were generated for each individual using individual-level BAM files (produced in genotype calling) and the mpileup command (-C 50, -Q 30, and -q 30) in samtools (Danecek et al. 2021). This was piped into the vcf2fq command from vcfutils.pl using our genome assembly as the reference. Filtering included sites with inferred consensus quality < 20 and a read depth less than 8× or greater than two times each individual sample’s mean coverage, calculated from BAM files using samtools “depth”.

A series of MaxEnt models (Phillips et al. 2006) were generated to assess environmental and ecological associations of S. semiluna across the central and northern extent of the species’ range. To assess niche divergence of the Alberta population relative to others within this modelling extent, a MaxEnt model was trained excluding Alberta occurrences and used to predict habitat suitability for all S. semiluna occurrences, including Alberta. If the predicted suitability of Blakiston Fan was substantially lower than the locations of all other occurrences, niche divergence was inferred (Campbell et al. 2022).

Our reference genome assembly was highly contiguous, spanning 1.25 Gb across 86 scaffolds, with an N50 of 56.2 Mb. Whole-genome resequencing of 15 individuals produced > 1.4 billion high-quality reads, yielding a dataset of 41,083,914 variants, with 23,889,641 SNPs retained after filtering. PCA and “structure” cleanly split all individuals into three populations with no evidence of admixture (Fig. 1a, b). F_ST values indicated substantial genetic divergence between Alberta and British Columbia (0.424), Alberta and Montana (0.292), and British Columbia and Montana (0.322). The two British Columbia sites showed no divergence, suggesting a high degree of gene flow (F_ST = −0.004). Mean heterozygosity was lowest in Alberta (0.083), compared to British Columbia (0.216) and Montana (0.154), and nucleotide diversity (π) was also lower in Alberta (0.003) than in British Columbia (0.008) and Montana (0.005) (Fig. 1b). This suggests much larger population sizes and broad-scale population connectivity in British Columbia and Montana compared to Alberta. Runs of homozygosity were 5–70 times more abundant in Alberta individuals, with individual F_ROH values averaging 0.192 in Alberta, 0.006 in British Columbia, and 0.033 in Montana (Fig. 1b). PSMC indicated that the Alberta population has been very small, isolated, and stable from 40 to 5 kya, with effective population size (N_e) estimated between 1,000 and 5,000 individuals (Fig. 1c). In contrast, British Columbia and Montana populations experienced large expansions toward the end of this time period, suggesting broad-scale connectivity. MaxEnt models predicted habitat suitability with high accuracy (AUC_ROC = 0.94) and identified mean summer precipitation as the most important environmental variable predicting S. semiluna occurrences (Fig. 1d). When Alberta occurrences were excluded from model training, predictive accuracy increased (AUC_ROC = 0.97). Using this model to predict habitat suitability at Blakiston Fan resulted in an estimate of 0.003, much lower than the locations of all other S. semiluna occurrences. This was interpreted as evidence of niche divergence.

Figure 1. 

Summary of genomic and niche analyses from MacDonald et al. (2025) A principal component analysis (PCA) using a dataset of 108,283 physically unlinked single nucleotide polymorphisms (SNPs) separated sequenced individuals into three discrete clusters. Weir and Cockerham’s (1984) F_ST values are shown between the three clusters B clustering analyses using the program structure (Pritchard et al. 2000) of all individuals found an optimal K value of 2, splitting Alberta and Montana from British Columbia; hierarchical runs excluding British Columbia identified an optimal K value of 2 with virtually no admixture between Alberta and Montana. Here, we combine hierarchical runs into a single admixture plot. Average heterozygosity and an estimate of inbreeding (F_ROH) for each individual is shown to the left of the admixture plot. Analyses of runs of homozygosity (ROH) and the proportion of each individual’s genome contained in ROH > 0.1 MB (F_ROH) suggested that historical inbreeding has been much more prevalent in the Alberta population (mean F_ROH = 0.192) than in British Columbia (F_ROH = 0.006) or Montana (F_ROH = 0.033) populations, suggesting a long history of isolation C the pairwise sequentially Markovian coalescent (PSMC) from 2.5 mya until 5 kya of the three identified clusters, with each individual’s genome serving as an independent sample. Years before present is shown on the x-axis and estimated effective population size (N_e) on the y-axis. The Alberta population flatlined between 1,000 and 5,000 individuals from 40 to 5 kya, indicating complete isolation. British Columbia and Montana both experienced substantial increases in N_e, suggesting broad-scale connectivity. The approximate duration of the Wisconsin glaciation is shown in blue (Clayton and Moran 1982; Bischoff and Cummins 2001) D) Predicted S. semiluna habitat suitability, predicted using 17 environmental variables, landcover data, and various terrain indices. “Research-grade” iNaturalist occurrences and the collection locations of sequenced individuals, excluding the Alberta population, were used to parameterize the model. Blakiston Fan had suitability value of 0.003, while other S. semiluna populations inhabited areas of much higher suitability. Environmental conditions at Blakiston Fan are therefore atypical for the species, indicating niche divergence.

Figure 2. 

A–D dorsal and ventral wing surfaces of the Satyrium curiosolus type series. Bodies of these specimens were used in genomic DNA extractions. Sequence data were used to generate a whole-genome consensus sequence for each specimen. Metadata for each specimen are given under “Type material” E composite photographs of pinned specimens (female left, male right), showing the dorsal wing surface on the left forewing and hindwing and the ventral wing surface on the right forewing and hindwing F S. curiosolus larvae being attended to by Lasius ponderosae ants G a freshly eclosed S. curiosolus on silvery lupine (Lupinus argenteus); H) Photograph of Blakiston Fan, Alberta, Canada I Calibrite ColorChecker Classic, photographed with the same setup and settings used to photograph the type series. The scale bar (bottom left) is 1 cm, against which A–E are scaled.

Genetic divergence, environmental and ecological divergence, and a very long history of isolation with no evidence of contemporary or recent gene flow are sufficient to recognize the Alberta population as a distinct taxonomic entity. We propose its recognition as a new species.

Satyrium curiosolus warrants recognition as a distinct taxonomic entity. We evaluated whether to describe it as a subspecies or a species based on two main criteria. Braby et al. (2012) defined subspecies by the combination of partial isolation of allopatric lineages, phenotypic distinctiveness, and at least one fixed, diagnosable character. This definition is rooted in the General Lineage Concept (GLC), which considers species as independently evolving lineages supported by multiple lines of evidence including criteria often associated with various species concepts (De Queiroz 1998, 2007). To ensure an objective comparison, we assess explicit criteria from these concepts (Table 1), taking them at face value as described in their original publications. While one interpretation of the GLC is that it lacks any specific criteria, like intrinsic reproductive isolating mechanisms or fixed morphological characters (De Queiroz 1998), practical application necessitates that we identify and score multiple criteria in making a species/subspecies determination. Thus, we focus on criteria used by alternative species concepts relevant to the “grey zone” of speciation (De Queiroz 2007).

Allopatry and isolation are critical properties for many subspecies and species concepts. Here, all evidence indicates that divergence between S. curiosolus and the geographically nearest S. semiluna populations is non-clinal, with no evidence of contemporary or recent gene flow. Satyrium curiosolus is completely isolated today, and coalescent-based analyses suggest this isolation may extend up to 40,000 years BP. Given this considerable isolation, traditional considerations of potential or actual reproductive isolation (Mayr 1942) are difficult to apply, and secondary contact with S. semiluna is improbable given both species’ relatively low vagility and the magnitude of range shift required (MacDonald et al. 2024). From a geographical standpoint, there is no possibility of hybridization and gene flow. This is a hypothesis that could be falsified by the discovery of genetic and environmental/ecological intermediates between S. curiosolus and S. semiluna. We consider this unlikely, given the prominent butterfly survey effort in the region. Nonetheless, we and others will continue to search for undiscovered populations. Satyrium curiosolus exhibits ecological distinctiveness, including unique environmental, host plant, and ant associations. Morphological distinctiveness was not extensively investigated but has been suggested by other species experts—Kondla (2004) noted that males from Blakiston Fan (S. curiosolus) have a smaller wingspan than other S. semiluna populations investigated (<25 mm vs > 30 mm) and less conspicuous ventral spotting. Beyond size and wing pattern, our most concrete diagnosable characters include 21,985 SNPs that are divergently fixed for alternate nucleotides between S. curiosolus and the nearest S. semiluna populations. For simplicity, we identified a subset of 117 SNPs evenly spaced across the genome. Future resequencing of additional S. curiosolus and S. semiluna individuals may reveal some of these SNPs to be “near fixed” even though they are fixed in our sample set—this would not invalidate them as diagnostic characters, but simply suggest that they have not evolved to complete fixation. As with any set of diagnostic characters, our evaluation of genomic diagnostic characters may change as new information becomes available.

Complete absence of gene flow, and a long history of isolation combined with genomic, environmental, and ecological differentiation, satisfies many of the criteria associated with alternate species concepts and unified under the GLC, notably those originating from the phylogenetic species concept (including variants introduced by Rosen 1979; Baum and Shaw 1995; Nelson and Platnick 1981; Cracraft 1983), the ecological species concept (Van Valen 1976), and the evolutionary species concept (Simpson 1951; Wiley 1978). The alternative—retaining S. curiosolus as a subspecies—would imply ongoing or potential gene flow, which is demonstrably absent, making species recognition the most taxonomically defensible classification (Braby et al. 2012). Still, we extensively debated the appropriate classification, initially considering subspecies as a more prudent and conservative option, anticipating that further evidence and analyses may substantiate species-level recognition. However, assuming that subspecies rank is the more prudent or conservative classification was potentially problematic. This assumption treats subspecies as an intermediate stage rather than objectively evaluating evolutionary independence based on multiple lines of evidence. Subspecies status often reflects structured intraspecific variation with some degree of contemporary or recent gene flow, not clear lineage separation (Mayr 1963; Braby et al. 2012). Under our view of the GLC, species recognition is a testable hypothesis, not a permanent designation, and should be based on the strength of evidence rather than a default bias toward subspecies as a sort of evolutionary null. Treating subspecies as a holding category creates an asymmetrical burden of proof, requiring disproportionately strong evidence for species recognition while subspecies designations persist under weaker criteria. This bias could obscure evolutionary significance and delay the recognition of independent lineages by misrepresenting biodiversity.

We acknowledge John Acorn for preparation and photography of pinned specimens, Lacey Hébert and Llewellyn Haines for fieldwork support, Steve Kohler for providing Montana specimens, Natasha Lloyd for logistical and permit support, Janet Sperling for assistance with lab work, including DNA barcoding of specimens to confirm species-level identifications, Eric Runquist for consultation, members of the UCLA Shaffer lab for analysis comments and critiques, the Waterton Lakes National Park Ecological Restoration Team for leading habitat restoration programs and assisting population monitoring, and the Half-moon Hairstreak Conservation Committee for discussions on impacts of this research. We also wish to acknowledge two anonymous reviewers whose comments greatly improved the quality of the manuscript.