Birds are one of the most likely dispersal vectors for plants in Antarctica. We studied the nesting behavior of south polar skua (Catharacta maccormicki) and brown skua (Catharacta antarctica lonnbergi) to assess their potential role in ornithochory in the Argentine Islands - Kyiv Peninsula region. Nest samples were collected during 2009-2020 years in the Argentine Islands - Kyiv Peninsula region including all islands and coasts of the Graham Land from the Lemaire Channel to the islands of Berthelot Islands from north to south and extending from west to east from the Roca Islands, Cruls Islands, Rasmussen Point to the coast. We found that skuas utilize different nest building materials, including bryophytes, vascular plants (hairgrass Deschampsia antarctica), and lichens. In south polar skua nests, mosses and lichens dominate in the nest material; in brown skuas Deschampsia antarctica and mosses dominate. Both bird species likely collect nest components from nearby vegetation formations (<1 m distant). We conclude that C. maccormicki and C. antarctica lonnbergi are not selective in their choice of plant species, simply using the materials that dominate near the nest. Therefore, both species carry these materials from nearby sites, and only occasionally bring them from distant places.  In conclusion, for both species we did not find any evidence to support their involvement in long-distance ornithochory (stomatochory) in the region.

Our fieldwork was done in the maritime Antarctic, in the Argentine Islands-Kyiv Peninsula region. The study area included all islands and coastlines of Graham Coast from Lemaire Channel to Berthelot Islands on the north-south direction with Roca Islands, Cruls Island, Rasmussen Point on the west-east direction side. Field surveys and sampling were done during seven seasons between 2009 and 2020. The name of each scientific season includes two years (for example 2009/2010), because the main scientific work is carried out from December to April.

In every season the majority of samples from south polar skua nests were collected on Galindez Island where the Ukrainian Antarctic station “Academic Vernadsky” is located. We estimate that the island holds at most 30 south skua nests per season. Depending on the ice situation in different years the sampling was also carried out on other islands in the Argentine Islands archipelago and nearby sites.

The number of south polar skua and brown skua nests examined varied between years. The selection of nests collected depended on the ice conditions during the season. On Galindez Island, samples of all south polar skua nests were collected in each study year, while on the other sites collecting depended on island availability in a given year. The brown skua is quite rare in the region, as it is the southern edge of its nesting area; therefore all known and available nests were examined every year. The nests of the birds were observed late in the nesting season, during a short period to reduce birds disturbance. GPS coordinates and pictures of each nest were taken.

During four seasons 2013/2014, 2015/2016, 2018/2019, 2019/2020, plant formations in which each nest was located were registered. All nests were analyzed for the percentage composition by different nesting materials. In the field, we quantified the nest composition (percentage of the total fraction of bryophytes, hairgrass D. antarctica, lichens, feathers, limpet shells, bones and algae). Additionally, in the seasons 2013/2014, 2015/2016, 2018/2019, 2019/2020 samples of material of approximately 10 × 10 cm were collected from all or from the majority of nests. In these samples, moss composition was determined in detail. Moss and lichen species were identified based on standard taxonomic guides (Øvstedal and Smith 2001, Ochyra et al. 2008).

 Statistical analysys

For statistical analysis, the average values of the defined classes of materials of the south polar skua nests for each season were compared. We calculated the test value as c² = (observed value – expected value)²/ (expected value). Results of these pairwise comparisons are expressed in relative units (Table 2). The value of the criterion statistics according to the criterion of median (Mood median test) was determined for each pair of distribution of nest contents by years (for the limit a = 0.05 of the χ² distribution, when comparing the two distributions, the limit value is 3.84) (Pollard 1982, Corder and Foreman 2014). 

To assess how much the contents of different components of nests located on different moss banks persist or differ over time, we formed a series of pairwise spatial differences in the modules for data sets by percentages of mosses, lichens, and feathers. The other components of the nest were not included in the analysis, as their values were mostly zero.

A series of pairwise spatial differences for the nests between the 2013/2014 and 2015/2016 seasons were compared by the regression technique by CurveExpert 1.3 software. Test value was calculated by formula F_{1, n-2}=t²_n-2=(n-2)R2/(1-R2), where n=52, for the upper limit a=0.05 of the F-distribution point value is 4.08 for n = 52.