New Genetic Toolkit Tracks Golden Eagles From a Single Shed Feather

A single feather drifting from an eagle’s nest may soon tell scientists nearly everything they need to know about the bird that dropped it. Researchers have unveiled a new genetic tool that allows them to identify individual golden eagles (Aquila chrysaetos), determine their sex, assess family relationships, and even flag samples that actually came from a different raptor species, all from DNA extracted from feathers collected without ever touching the birds. The study, published in the journal Ecology and Evolution, describes a panel of 96 genetic markers tailored specifically to the golden eagles of Fennoscandia, the boreal expanse spanning Norway, Sweden, and Finland, and demonstrates that it performs reliably even on the low-quality DNA that plagues non-invasive sampling.

The golden eagle occupies a precarious position in the Fennoscandian landscape. As an apex predator, it is long-lived, slow to reproduce, and needs enormous territories, characteristics that make it acutely vulnerable to human pressures. The species typically lays only one or two eggs, and young birds may not settle in a territory of their own until roughly five years of age. Survival during the early years is low, climbing steeply only as birds approach adulthood. This life history means that even modest increases in adult mortality can push a population into decline. Today the main breeding strongholds lie in the north, with smaller populations in southernmost Sweden and Denmark and a notable concentration of pairs on the Swedish island of Gotland.

The region’s eagles have already survived one brush with disaster. Persecution, habitat destruction, and the accumulation of environmental toxins drove the population down sharply in the early twentieth century, with local extinctions in several areas. Legal protections introduced from the 1970s onward allowed numbers to recover steadily, and the population is now considered stable, although breeding attempts and nesting success still swing widely from year to year depending on prey availability and weather. Recovery, however, does not erase history. Populations that pass through a bottleneck can lose genetic diversity, which in turn makes them more sensitive to future environmental shocks. Whether the Fennoscandian eagles carry such scars has been debated, with some microsatellite-based studies finding no bottleneck signature and others, drawing on museum specimens, documenting a loss of haplotypes and comparatively low variation relative to central Asia and the Caucasus.

Part of the uncertainty stems from the limitations of the genetic markers traditionally used. Microsatellites, the workhorses of wildlife genetics for decades, are individually informative but demand long, intact stretches of DNA, exactly what degraded non-invasive samples often lack. Shorter-read SNP markers, by contrast, tolerate fragmented DNA far better, and results from SNP panels are easier to compare across laboratories, a considerable advantage when a population ranges across three countries with separate monitoring programs. The trade-off is that each SNP carries less information than a microsatellite, so more markers are needed to achieve comparable power. The new panel meets that requirement by packing 96 markers onto a nanofluidic genotyping platform that can process large sample sets quickly and cheaply.

To build the panel, the team sequenced DNA from 96 golden eagles sampled across Sweden, Norway, Denmark, and Finland, working with the NGO Golden Eagle Sweden, the Swedish Veterinary Agency, and the Norwegian Institute for Nature Research. Most samples came from chicks ringed at nests or adults captured for other research projects, supplemented by tissue from birds found dead. RAD-sequencing initially surfaced nearly 30,000 candidate SNPs, which the researchers whittled down through filtering for biallelic markers, allele frequencies, Hardy-Weinberg equilibrium, and coverage. Validation on 81 independent samples, including 40 pulled-feather samples from Finnish nestlings and 41 Norwegian samples representing blood, bone, claw, eggshell, footpad, feather, and muscle tissue, left 95 final SNPs plus one marker from a previous study that determines sex.

The validation results were striking. Blood, footpad, and tissue samples achieved 100 percent amplification success, and the crucial non-invasive sources performed almost as well: shed, molted feathers reached 98 percent amplification, and plucked feathers 96.2 percent, a difference so small it was statistically indistinguishable. Because molted feathers actually edged out plucked ones on average, the authors argue that field workers can rely entirely on feathers gathered from the environment rather than plucking them from live birds, sparing both eagles and handlers the stress and injury risk of direct contact. Less useful were eggshells, bones, and claws, which each came from a single sample and showed amplification success ranging from roughly 62 to 89 percent.

When it comes to telling individual eagles apart, the panel is exceptionally precise. The probability that two unrelated individuals would share the same genotype across all 95 markers is about 5.11 times 10 to the minus 30, and even full siblings can be discriminated with a probability of identity of roughly 4.06 times 10 to the minus 20. Exclusion probabilities for parentage analysis exceeded 0.99997 whether or not one parent was already known. Genotyping accuracy was similarly strong: across ten feather samples run twice, the average error rate was just 1.979 percent, and seven of the ten samples showed zero mismatches and 100 percent amplification. The one problematic sample, with an error rate above 10 percent, also had the lowest amplification success, reinforcing a well-established principle of non-invasive genetics that genotyping quality tracks sample quality and that strict amplification thresholds are essential.

Sex determination proved equally dependable. The CHD1ZW marker correctly assigned the sex of 40 of 41 previously sexed samples, with the single failure attributable to missing data rather than a mismatch, and it worked consistently in replicate runs. Assessing family relationships revealed both the panel’s strengths and its limits. Simulations of 10,000 pairs per relationship class showed that the panel correctly classifies close relatives, meaning parent-offspring pairs or full siblings, about 90 percent of the time, and unrelated pairs about 88.6 percent of the time. Half-siblings, however, were correctly identified in only 77.1 percent of simulated cases, blurring into both the unrelated and first-order categories. The overall classification accuracy of 86.5 percent means the panel can credibly distinguish unrelated birds from close relatives, which is what matters most for estimating territory turnover and dispersal, but researchers should treat half-sibling assignments with caution, echoing earlier findings that roughly 100 SNPs separate close kin from strangers but not degrees in between.

The panel also delivered a first broad genetic picture of the Fennoscandian population using its own markers. Observed and expected heterozygosity were closely matched across most regions, inbreeding coefficients hovered near zero, and genetic differentiation among populations was low, at FST around 0.04, consistent with earlier evidence of gene flow both within Fennoscandia and between Fennoscandia and the Alps. Even Gotland, despite its island geography, showed no elevated inbreeding or depressed diversity, though the authors caution that only seven Gotland samples were available and that the number of markers may be too small to detect subtle inbreeding. A Finnish outlier with a heterozygote deficit almost certainly reflects a sample size of just two birds rather than a genuine biological signal, since a larger Finnish microsatellite study found no such deficit. Notably, the panel doubled as a species check: white-tailed eagle feathers, which can be mistaken for golden eagle feathers in the field, produced genotypes with very low heterozygosity and abundant missing data, fingerprints that let researchers filter out misidentified samples before they contaminate an analysis.

The broader significance extends well beyond eagles. Norway and Sweden already monitor wolverines, brown bears, and wolves through non-invasive genetic sampling, and a comparable system for golden eagles promises data on reproductive success, territory turnover, and gene flow that nest visits alone cannot provide. Because SNP results travel cleanly between laboratories, the panel opens the door to coordinated monitoring across Norway, Sweden, and Finland, and to building shared databases for a population that ignores national borders. In Japan, where microsatellite studies of the local golden eagle emphasized the value of continuous genetic monitoring for catching early signs of bottlenecks and declining connectivity, the same logic applies. For a species still shadowed by forestry, infrastructure expansion, and climate change, a tool that converts a molted feather on the forest floor into a named, sexed, and relatedness-mapped individual could become the difference between spotting a population’s decline early and reading about it in the statistics afterward.

Subject of Research: Development and validation of a 96-SNP genetic panel for non-invasive monitoring of the Fennoscandian golden eagle population

Article Title: A Genetic Tool for Non‐Invasive Monitoring of Fennoscandian Golden Eagles (Aquila chrysaetos)

Article References: Lindberg, B., Kleven, O., Kvist, L., Norman, A., Köningsson, H., Jacobsen, K.-O., Nilsson, P.-O., Singh, N. J., & Spong, G. (2026). A Genetic Tool for Non‐Invasive Monitoring of Fennoscandian Golden Eagles ( Aquila chrysaetos ). Ecology and Evolution, 16(9), Article e74386. https://doi.org/10.1002/ece3.74386

Image Credits: AI Generated

DOI: 10.1002/ece3.74386

Keywords: golden eagle, Aquila chrysaetos, SNP panel, non-invasive genetic sampling, Fennoscandia, genetic monitoring, wildlife forensics, feather DNA, relatedness, conservation genetics, raptor ecology, nanofluidic genotyping