Article | Open | Published:

Vagrant birds as a dispersal vector in transoceanic range expansion of vascular plants

Scientific Reportsvolume 9, Article number: 4655 (2019) | Download Citation

Abstract

Birds are thought to be important vectors underlying the disjunct distribution patterns of some terrestrial biota. Here, we investigate the role of birds in the colonisation by Ochetophila trinervis (Rhamnaceae), a vascular plant from the southern Andes, of sub-Antarctic Marion Island. The location of O. trinervis on the island far from human activities, in combination with a reconstruction of island visitors’ travel history, precludes an anthropogenic introduction. Notably, three bird species occurring in the southern Andes inland have been observed as vagrants on Marion Island, with the barn swallow Hirundo rustica as the most common one. This vagrant displays long-distance migratory behaviour, eats seeds when insects are in short supply, and has started breeding in South America since the 1980s. Since naturalised O. trinervis has never been found outside the southern Andes and its diaspores are incapable of surviving in seawater or dispersing by wind, a natural avian dispersal event from the Andes to Marion Island, a distance of >7500 km, remains the only probable explanation. Although one self-incompatible shrub seems doomed to remain solitary, its mere establishment on a Southern Ocean island demonstrates the potential of vagrancy as a driver of extreme long-distance dispersal of terrestrial biota.

Introduction

Successful long-distance dispersal events are extremely rare, difficult to observe directly, and thus typically only reconstructed by phylogeographic means1,2. Notable exceptions are observations of pumice rafting following natural disasters such as tsunamis or volcanic eruptions3,4. While the role of abiotic vectors such as oceanic currents and wind in long-distance dispersal events is increasingly acknowledged5, biotic dispersal remains difficult to reconstruct6. For example, pelagic birds have repeatedly been suggested as a likely long-distance dispersal vector of propagules across oceans7,8, although the behaviour of marine birds does not match with the dispersal syndromes of inland terrestrial species9,10. Some exceptions do exist however, such as when seabirds come into contact with shoreline vegetation11. The mechanisms that underlie long-distance dispersal events, and especially the role of birds therein, is therefore still a matter of ongoing discussion7,12,13.

A suitable system to study the potential mechanisms underlying long-distance dispersal events are the Southern Ocean islands. These sub-Antarctic islands are scattered throughout the Southern Ocean and are amongst the most remote places on Earth. Even though the terrestrial biota of these islands are relatively species-poor, they display a pattern of inter-island similarity that indicates long-distance dispersal in an eastward direction14,15. This directional dispersal is likely the consequence of strong westerly winds and the West-wind Drift7,16,17,18. Indeed, wind has been suggested as the dispersal vector of recent arrivals of arthropod species to Marion Island19. Since sub-Antarctic climate change is resulting in increasing temperatures, the milder climate is additionally resulting in more suitable conditions for propagules of new species to establish as part of natural range expansion20,21. However, because the majority of new arrivals to the Southern Ocean islands is the result of anthropogenic dispersal22, observations on natural long-distance dispersal events are masked by those of anthropogenic origin, making it difficult to determine whether a new introduction is human-mediated, or the result of a natural dispersal event.

In spite of the remoteness of the Southern Ocean islands from major landmasses, historic and current anthropogenic activities have resulted in a steady rate of colonisation events of non-native species to the islands by means of an anthropogenic vector23,24. While not all colonisation events result in successful establishment, often due to the harsh climatic conditions in this region, a large number of species have shown to do extremely well on the islands, often with detrimental impact on the indigenous biota22,25,26. Therefore, in recent years an increasing amount of effort has been invested in preventing anthropogenic dispersal of propagules and early eradication of populations of newly introduced species. For example, on South African National Antarctic Programme expeditions no fresh food is allowed, while all cargo, footwear and hiking gear is cleaned and checked for presence of soil and propagules prior to landing. Also, a number of range-restricted alien species are being actively controlled with herbicides27. Despite these prevention efforts, some propagules will inevitably slip through and colonise successfully22,23.

When a new species previously unknown to a Southern Ocean island is found close to anthropogenic structures such as huts, hiking trails or scientific stations, this typically indicates that the dispersal vector was anthropogenic20,25. However, a non-natural dispersal event cannot be assumed per se. For example, since propagules of terrestrial biota are unlikely to survive long-term inundation in saline water, wind is the most likely long-distance dispersal vector for some recently discovered populations of wind-dispersed species such as lichens, bryophytes and ballooning arthropods19,28. For non-wind-dispersed biota, pelagic and migratory birds are assumed to be the main vectors in natural long-distance dispersal events9,12,13. Recent discoveries of plant species that are known to other islands of the South Indian Ocean Province have thus been assumed to have been transported by birds travelling between the Southern Ocean islands29,30.

The recent discovery of an unidentified shrub at a remote location on Marion Island, a Southern Ocean island, however, did not fit any of these explanations. The enigmatic appearance of the species indicated that it had originated from outside the South Indian Ocean Province, while the remote location of this small population is contrary to an anthropogenic dispersal event. Since detailed data are available on human activities on this island and on resident and migratory birds, this discovery provides a unique model system to explore an alternative dispersal route.

In this study, we investigate the role of long-distance dispersal events in the colonisation process of Southern Ocean islands. We do this by reconstructing the dispersal route of a recently discovered unidentified shrub on a Southern Ocean island. We first identify the species using a combination of plant anatomy and phylogenetic techniques to determine the region of origin, and assess its arrival time on Marion Island. We then reconstruct historic visitors’ movements to the island to assess whether the dispersal event was of anthropogenic or natural origin. Since birds constituted a potential dispersal vector of the plant species, we compare bird occurrence data from the plant species’ region of origin with those of its new habitat. We then discuss the role of birds, and in particular that of vagrants, as vectors of long-distance dispersal events.

Results

Plant morphological analysis

The canopy cover of the unknown shrub had increased from 1.11 m² to 2.09 m² to 2.46 m² between 2004, 2015 and 2017, respectively (Fig. 1a). A linear model fitted to this growth rate indicated — with a 90% confidence interval of 1.6 y (linear model: canopy cover = 0.099 × year – 196.8; F = 59.95, P = 0.082, r²adj = 0.967) — that the shrub had sprouted in late 1992.

Figure 1
Figure 1

Overview of the plant morphology of Ochetophila trinervis on Marion Island: (a) some 25 years after establishment, the habitus remains a dwarf shrub; (b) detail of a twig showing the 3-nerved abaxial venation pattern; and (c) partial transection of a leaf, with arrows indicating the location of epidermis stomata on both leaf sides.

The unknown shrub displayed a set of traits typical of the tribe Colletieae (Rhamnaceae): decussate leaves subtending two serial buds each, and an upper bud giving origin to a short spine (Fig. 1b). The connate stipules of opposite leaves, the more or less spatulate leaf shape, the entire leaf margin and the presence of three main veins suggested a close proximity to Ochetophila trinervis (Gillies ex Hook. & Arn.) Poepp. ex Endl.31, a self-incompatible perennial shrub with wind- and insect-pollinated hermaphroditic flowers32. The leaf sections and leaf clearings indicated that the unknown shrub had an anatomical structure similar to O. trinervis33. The unknown shrub had stomata in the upper and lower leaf epidermis (Fig. 1c). Upper epidermis stomata were also found for the 3-month old O. trinervis seedlings, but not for the older specimens.

Phylogenetic analysis

The final alignment of 24 sequences had a length of 829 bp, of which 638 bp (77%) were identical. The Bayesian and ML analysis of the tribe Colletieae, based on trnL-F sequence data, resulted in congruent phylogenies. As such, only the consensus tree of the Bayesian analysis is presented (Fig. 2). The tribe itself received support of 1.00/100% (posterior probability/bootstrap value). The Trevoa/Retanilla and Adolphia/Discaria clades, as well as the genera Colletia and Ochetophila have support of 0.9/80% and above. The unknown shrub is a sister taxon of O. trinervis with a support of 0.98/77% (labelled as “sp. Marion Island”). In the final alignment, the sequences of O. trinervis and the unknown shrub are identical.

Figure 2
Figure 2

Majority-rule consensus tree (Bayesian analysis) of Colletieae, including sp. Marion Island, based on trnL-F sequence data, indicating that the Marion Island species is the same species as the Ochetophila trinervis specimen from South America. Numbers above nodes represent posterior probabilities, numbers below are bootstrap values of the ML analysis. Posterior probabilities below 0.80 and bootstrap values below 60% are not indicated.

Biogeographic analysis

We collated 47 georeferenced records for O. trinervis. All of these observations were located in the lowlands to montane areas of the southern Andes (Fig. 3). No records were found outside this range. Particularly, there were no herbarium records of this species in the online database of the South African National Biodiversity Institute (http://newposa.sanbi.org; date accessed: September 14, 2018). The nearest naturally occurring population was near Lake Pueyrredón (Argentina, 47°26′10″S,71°55′23″W) observed by DM. The geodesic distance between the O. trinervis observation on Marion Island and this nearest natural population was ~7530 km. The Euclidean distance from O. trinervis to the research station on Marion Island was 9.6 km, and to the nearest field hut 1.6 km (Fig. 3, inset).

Figure 3
Figure 3

Overview of the known distribution range of Ochetophila trinervis in South America (▲) and the location of Marion Island (•) in the sub-Antarctic Indian Ocean (Projection: Lambert azimuthal equal-area for the South Pole). The arrow depicts the Antarctic Circumpolar Current (ACC), whereas the dashed lines represent the major oceanic fronts: Subtropical Front (STF), Subantarctic Front (SAF), Polar Front (PF), and Southern Antarctic Circumpolar Current Front (SACCF). The inset depicts the geographical dimensions of Marion Island (Projection: Transverse Mercator), the location of the permanently manned base (▪), the current and former field huts (▫), and of the only known O. trinervis shrub (•).

Dispersal analysis

We contacted 35 of 123 persons that were listed as visitors to Marion Island in the period 1986–1991. Of the 18 respondents, all indicated that they had not visited South America prior to their visit to Marion Island. Pre-Marion Island travel history was mostly limited to Namibia and Angola as part of military service.

Chown and Froneman34 report 86 bird species for the Prince Edward Islands, which includes all breeding and vagrant species (Supplementary Table S1). Some 56 of these are also listed on the bird checklists of Chile or Argentina. The vast majority of these species, however, have a strictly marine or coastal habitat preference, which does not overlap with the distribution range of O. trinervis. Only three species have an inland terrestrial habitat preference and a distribution range that overlaps with O. trinervis, namely: cattle egret Bubulcus ibis, pectoral sandpiper Calidris melanotos, and barn swallow Hirundo rustica. Each of these three bird species has a cosmopolitan distribution range, shows migratory behaviour, and has been recorded as a vagrant on Marion Island. All other vagrants on Marion Island with a terrestrial habitat preference possess a breeding range that is strictly limited to the African continent.

Discussion

We successfully identified a vascular plant new to a South Indian Ocean Province island using a combination of plant morphological comparison and DNA analysis, and estimated its time of arrival to the island. Typically, the discovery of a species new to a sub-Antarctic island is reason for concern, since the vast majority of such discoveries reveal a further spread of non-indigenous species across this biogeographical zone20,22,35. Below we discuss the identification process, whether its arrival on the island was the result of an anthropogenic or a natural dispersal event, and how these findings contribute to our understanding of long-distance dispersal events.

Our molecular analysis closely resembled previous phylogenetic classifications of the Colletieae with the genus Ochetophila as a distinct and well-supported clade31,36,37. Since the DNA sequences and leaf characteristics of the unknown shrub on Marion Island were almost identical to the reference material (Figs 1 and 2), we identified the specimen with high confidence as Ochetophila trinervis (Rhamnaceae: Colletieae). This record is remarkable because the natural distribution range of O. trinervis is strictly limited to the southern Andes32,38. Since only a single individual of O. trinervis was found on Marion Island, no genetic diversity analysis could be conducted at the population level to estimate time since arrival. Also, none of the systematic or ad hoc biological surveys conducted between 1965 and 2004 revealed a species anything like this one25. We are therefore confident that O. trinervis is a recent arrival to Marion Island, and that it has gone undetected for some 12 years due to its inconspicuous growth form and remote location on the island. A recent arrival alone, however, does not determine whether an introduction was anthropogenic or natural. While an anthropogenic dispersal event appears likely given that most new arrivals to Southern Ocean islands are exotics20,39, and many are transported in propagule form by visitors to the island23, the result of natural dispersal or of human-assisted dispersal is ambiguous. For example, O. trinervis has never been recorded as a non-indigenous species outside its natural distribution range, whereas most new records of non-native species to the Southern Ocean islands are species that are cosmopolitan or invasive elsewhere18. We are therefore cautious to exclude a natural dispersal event a priori.

On Southern Ocean islands and Antarctica, non-indigenous species are generally first detected around research stations, field huts or in disturbed habitats20,40. Such spatial pattern fits the expectation that newly introduced species first establish in close proximity to their point of introduction41, often facilitated by some form of disturbance25,42. The O. trinervis shrub, however, was found in an undisturbed habitat far from such potential point of introduction (Fig. 3), on an island which, due to its young volcanic substrate, is difficult to traverse. Moreover, non-indigenous species frequently have large distribution ranges, are associated with anthropogenic activities, and have been introduced to multiple areas outside their natural distribution range22,43. In its natural distribution range, however, O. trinervis — a common shrub or treelet, frequently forming woods of hundreds of individuals — is associated with riverine rather than anthropogenic habitats44. Finally, our visitors’ survey indicated that no respondent had visited the southern Andes prior to their Marion Island voyage. Although contemporary visitors to Southern Ocean islands generally have a prior travel-history to other cold-climate areas18, during the 1980s travelling between South Africa and South America was costly and cumbersome due to geopolitical circumstances45,46. We therefore consider the likelihood that a visitor accidentally (or even deliberately) transported an O. trinervis propagule from the southern Andes to Marion Island as highly improbable.

Natural extreme long-distance dispersal events are usually by means of wind dispersal, oceanic currents, or avian dispersal12,13,47. The diaspore morphology of O. trinervis, a dehiscent explosive fruit capsule containing 1–3 smooth wingless seeds of approx. 2 mm in diameter, as typically observed for species with a ballistic dispersal syndrome48, precludes attachment to floating debris, animal fur or human clothing32,44. This genus and the family Rhamnaceae have also not been discovered in comprehensive assays of seeds associated with the clothing and equipment of visitors to the sub-Antarctic islands or Antarctic continent23,49. Moreover, viable seeds of closely related taxa are likely to sink, while saline water swiftly reduces seed viability, precluding hydrochory as a possible dispersal syndrome50. Finally, the location of the shrub on the east-side of the island, corresponds with the area where most avian vagrants are observed51. Therefore, transoceanic dispersal via anemochory or hydrochory is highly unlikely; leaving avian dispersal as only parsimonious explanation.

Marine and pelagic birds are often suggested as a likely vector in historic and recent transoceanic dispersal events due to the extreme long distances that these birds can cover7,8,10. However, these birds do not have the behavioural traits to acquire an O. trinervis diaspore from the southern Andes. Of the three bird species with an inland terrestrial habitat preference and observed both on Marion Island and in Chile/Argentina (Supplementary Table S1), the pectoral sandpiper feeds on arthropods and other invertebrates, while the cattle egret predominantly feeds on insects52. Although the barn swallow Hirundo rustica is also primarily insectivorous, this species is known to feed on fruits and seeds when insects are in short supply53. Indeed, barn swallows are the most common vagrant bird species on Marion Island51, and have even been observed as far south as King George Island, Antarctica54. Moreover, even though the barn swallow predominantly breeds in the Northern Hemisphere, this species has switched its migratory behaviour by establishing sedentary breeding populations in Argentina as recently as the early 1980s55. Finally, although it is unclear whether the barn swallows that have been observed on Marion Island are of the subspecies Hirundo rustica subsp. erythrogaster — as typically observed in South America — the strong circumpolar winds in easterly direction suggest that vagrancy from South America, facilitated by a storm event, is possible16,35. Indeed, weather events and growing populations have been associated with transoceanic vagrancy56,57. This unique combination of having recently switched its migratory behaviour and being one of the most common vagrants on Marion Island leaves the barn swallow as the most likely avian dispersal vector between the southern Andes and Marion Island.

For a newly discovered species to be considered as an indigenous species, both the origin of the propagule and its dispersal vector need to be determined20. Even though we had no physical evidence of a barn swallow depositing an O. trinervis propagule on Marion Island, all other indicators pointed to an avian-mediated natural dispersal exclusively. Indeed, migratory birds have been observed to transport ingested vascular plant seeds across oceans13, suggesting that vagrants can play a similar role as a vector of transoceanic colonisation. Interestingly, as a member of the Colletieae, this species can form a symbiosis with nitrogen-fixing Frankia bacteria, which can have significant local impact on nutrient-poor soils44. Since O. trinervis does not readily reproduce vegetatively and, under the current climatic conditions, is unlikely to produce any flowers32, the future impact of this single shrub is unlikely to be significant. Until and unless this small population goes extinct, we propose the status of O. trinervis as an indigenous species to Marion Island and the Southern Indian Ocean Province.

There is no clear evidence yet linking climatic change with increased natural colonisation rates of Southern Ocean islands58. However, climate change is expected to speed up island colonisation across long-distances due to changing migratory behaviour12,21. Natural range expansion of a vascular plant from the southern Andes to a Southern Ocean island fits the narrative that vagrancy is an important driver of transoceanic dispersal56. In addition, climate barriers are lowering, resulting in the Southern Ocean islands and the Antarctic as increasingly suitable for newcomers59. Since propagules of alien species are common and have large ecological niches, vagrant birds are expected to disproportionally disperse alien rather than indigenous species and with increased frequency60.

Methods

Study area

Marion Island — one of the two Prince Edward Islands (South Africa) — is located ~1750 km south of the African continent in the Southern Ocean (Fig. 3). It is a rugged 293 km2-sized oceanic island belonging to the Kerguelen Biogeographic Province34. The nearest neighbouring islands to the Prince Edward Islands are Bouvet Island at ~2540 km to the west, and the Crozet Islands at ~1080 km to the east. The Antarctic Circumpolar Current ensures that Marion Island’s climate is cool and wet with an average annual temperature of 6.4 °C and just over 2000 mm of precipitation per annum34.

The vegetation of Marion Island is relatively species poor due to a combination of climatic constraints and its remoteness from potential species pools. Still, 23 vascular plant species have been recorded as indigenous to the island with an additional 18 species as introduced by humans25,27. The island is characterised by low-growing vegetation such as mire and fernbrake habitats, and fellfield vegetation with cushion-forming Azorella selago, bryophytes, and lichens. For a detailed description of the vegetation types on Marion Island see Chown and Froneman34.

Plant morphological analysis

On March 18, 2004, an unknown shrub was discovered by Greg Hofmeyr in fernbrake vegetation on the southeast area of the island (Fig. 3 inset; 37°50′02″E, 46°57′40″S). The shrub’s canopy dimensions were measured and a herbarium specimen was collected. However, the shrub remained unidentified and it has since then been listed as a species with an unclear introduction status34.

To obtain a species identification, we collected a small piece of a branch with leaves in May 2015. Part of the sample was preserved in 98% ethanol for morphological analysis. The remainder of the sample was preserved in silica gel for genetic analysis. We also measured the dimensions of the canopy in April of 2015 and 2017 to assess growth expansion since its discovery (Fig. 1a).

A tentative identification, based on colour photos (Fig. 1b), suggested that the unknown shrub belonged to a taxon within the Colletieae (Rhamnaceae). Using standard plant anatomical methods, full-grown, healthy and morphologically representative leaves were either embedded in paraffin and transversally sectioned to study the anatomical structure (Fig. 1c), or cleared to study the distribution of stomata and the nerviation pattern. Results were compared with anatomical data on Colletieae33. To estimate the plant-anatomical development stage of the unknown shrub, freshly collected leaf samples from specimens of Ochetophila trinervis (syn.: Chacaya trinervis, Discaria trinervis) of different ages (3-month and 1-year old saplings, and an adult tree) were collected from the species’ native range (San Carlos de Bariloche, Argentina; legit E. Chaia and S. Alzogaray, October 2016), processed as above, and included in the comparison.

Phylogenetic analysis

To confirm the species identification, we compared the genetic material of the unknown shrub with sequence data for the trnL intron and trnL-F spacer region of Colletieae from Aagesen, et al.37, Burge, et al.61, Kellermann and Udovicic62, and Richardson, et al.36. All species in the tribe were included with the exception of Colletia spartioides Bertero ex Colla and C. spinosissima Gmel. Outgroups of related tribes and genera were selected according to the most recent family-wide phylogeny63. DNA of the unknown shrub was extracted from silica-dried leaf material using the ISOLATE II Plant DNA Kit (Bioline, London). For Colletia hystrix, Retanilla ephedra and Retanilla stricta, the same DNA extract was used as described in Aagesen, et al.37. All samples are listed in Supplementary Table S2.

Two PCR amplifications of cpDNA were performed using primers trnL c and d, and trnL e and f64. PCR amplifications were performed in 35 μl reactions using the MyTaq™ HS DNA polymerase kit of Bioline reactions containing: 1.5 μl of each 10 µM primer, 7 μl of 5 × reaction buffer, 0.25 μl of polymerase (1.25 U). The cycling profile was as follows: 2 min at 95 °C, then 37 cycles of 20 s at 95 °C, 20 s at 58 °C, 20 s at 72 °C, with a final extension of 2 min at 72 °C. The quality and approximate size of products were checked by electrophoresis on a 1.5% agarose gel. DNA sequences were generated in both forward and reverse direction from direct sequencing of PCR products through the Beijing Genome Institute (China).

To analyse the molecular data we used Geneious 8.1.9 (created by Biomatters, http://www.geneious.com). After initial automatic alignment, the alignment was refined manually. Sequences for both DNA regions, trnL intron and trnL-F spacer, were aligned separately. Jagged ends of the alignments were trimmed and both regions were combined for analysis. Polynucleotide regions could not be meaningfully aligned and were excluded from the dataset. Three indel regions were treated as informative and removed after coding. Maximum likelihood (ML) of phylogeny estimation was calculated with the plugin PhyML 2.2.365, using the GTR + I + G model of nucleotide substitution and calculating bootstrap support with 100 replicates, 6 substitution rate categories, and an NNI topology search. The Bayesian analysis was run with the plugin MrBayes 3.2.266 using four Markov Chain Monte Carlo heuristic searches of 1.1 mio generations performed in four independent runs. Subsampling frequency was set to 200, burn-in length to 100,000. The substitution model used was GTR + I + G. Parameter trace files were examined to ensure convergence of separate runs. Posterior probabilities are indicated to illustrate support of nodes.

Biogeographic analysis

To determine the distance between the nearest naturally occurring population of O. trinervis and the location of the unknown shrub on Marion Island, we collated all georeferenced observations of O. trinervis and its synonyms into a single geodatabase. For this we obtained records from Tortosa38 and GBIF (https://www.gbif.org/), supplemented with personal observations by DM. We specifically searched for the occurrence of specimen outside the expected distribution range. We then measured the geodesic distance between the nearest O. trinervis record and the unknown shrub.

Dispersal analysis

To determine whether O. trinervis could have been introduced by means of an anthropogenic vector, we obtained a list of names of visitors to Marion Island from the Antarctic Legacy of South Africa (http://blogs.sun.ac.za/antarcticlegacy/) for the time period 1985–1991. The Antarctic Legacy of South Africa keeps a record of all individuals that have visited the island every year. Since no current contact details for visitors during this time period were available, we contacted as many persons as possible on this list via social media (https://www.facebook.com/groups/marionisland/). We asked when they visited Marion Island and which territories outside southern Africa were visited before their visit to the island. Only if the latter question included South America did we ask which areas in South America had been visited, which areas of Marion Island had been visited, and if gear such as shoes, boots or bags had been used during both visits.

To identify which bird species might be a potential vector for the dispersal of O. trinervis to Marion Island, we compared the checklist of bird species from the Prince Edward Islands34, with those of the countries in which O. trinervis occurs naturally: Chile and Argentina (http://www.museum.lsu.edu/~Remsen/SACCCountryLists.htm). To harmonise species names, we adopted the nomenclature of the IOC World Bird List version 7.1 (http://www.worldbirdnames.org/). We identified bird species as potential dispersers based on the following criteria: a distribution range overlapping with O. trinervis, a non-marine habitat preference, and a long-distance migration behaviour. Bird distribution ranges and habitat preferences were obtained from GBIF, while migration and feeding behaviour was assessed from literature.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    le Roux, J. J. et al. Relatedness defies biogeography: the tale of two island endemics (Acacia heterophylla and A. koa). New Phytol. 204, 230–242, https://doi.org/10.1111/nph.12900 (2014).

  2. 2.

    Cook, L. G. & Crisp, M. D. Directional asymmetry of long-distance dispersal and colonization could mislead reconstructions of biogeography. J. Biogeogr. 32, 741–754, https://doi.org/10.1111/j.1365-2699.2005.01261.x (2005).

  3. 3.

    Bryan, S. E. et al. Rapid, long-distance dispersal by pumice rafting. PLoS One 7, e40583, https://doi.org/10.1371/journal.pone.0040583 (2012).

  4. 4.

    Carlton, J. T. et al. Tsunami-driven rafting: Transoceanic species dispersal and implications for marine biogeography. Science 357, 1402–1406, https://doi.org/10.1126/science.aao1498 (2017).

  5. 5.

    Cowie, R. H. & Holland, B. S. Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. J. Biogeogr. 33, 193–198, https://doi.org/10.1111/j.1365-2699.2005.01383.x (2006).

  6. 6.

    Viana, D. S., Santamaría, L., Michot, T. C. & Figuerola, J. Allometric scaling of long-distance seed dispersal by migratory birds. Am. Nat. 181, 649–662, https://doi.org/10.1086/670025 (2013).

  7. 7.

    Gillespie, R. G. et al. Long-distance dispersal: a framework for hypothesis testing. Trends Ecol. Evol. 27, 47–56, https://doi.org/10.1016/j.tree.2011.08.009 (2012).

  8. 8.

    Valdebenito, H. A., Stuessy, T. F. & Crawford, D. J. A new biogeographic connection between islands in the Atlantic and Pacific Oceans. Nature 347, 549–550, https://doi.org/10.1038/347549a0 (1990).

  9. 9.

    Nogales, M., Heleno, R., Traveset, A. & Vargas, P. Evidence for overlooked mechanisms of long-distance seed dispersal to and between oceanic islands. New Phytol. 194, 313–317, https://doi.org/10.1111/j.1469-8137.2011.04051.x (2012).

  10. 10.

    Cruden, R. W. Birds as agents of long-distance dispersal for disjunct plant groups of temperate western hemisphere. Evolution 20, 517–532, https://doi.org/10.2307/2406587 (1966).

  11. 11.

    Sebastian, P., Schaefer, H., Lira, R., Telford, I. R. H. & Renner, S. S. Radiation following long-distance dispersal: the contributions of time, opportunity and diaspore morphology in Sicyos (Cucurbitaceae). J. Biogeogr. 39, 1427–1438, https://doi.org/10.1111/j.1365-2699.2012.02695.x (2012).

  12. 12.

    Viana, D. S., Santamaría, L. & Figuerola, J. Migratory birds as global dispersal vectors. Trends Ecol. Evol. 31, 763–775, https://doi.org/10.1016/j.tree.2016.07.005 (2016).

  13. 13.

    Viana, D. S., Gangoso, L., Bouten, W. & Figuerola, J. Overseas seed dispersal by migratory birds. Proc. R. Soc. B-Biol. Sci. 283, 20152406, https://doi.org/10.1098/rspb.2015.2406 (2016).

  14. 14.

    Chown, S. L., Gremmen, N. J. M. & Gaston, K. J. Ecological biogeography of Southern Ocean islands: species-area relationships, human impacts, and conservation. Am. Nat. 152, 562–575, https://doi.org/10.1086/286190 (1998).

  15. 15.

    Greve, M., Gremmen, N. J. M., Gaston, K. J. & Chown, S. L. Nestedness of Southern Ocean island biotas: ecological perspectives on a biogeographical conundrum. J. Biogeogr. 32, 155–168, https://doi.org/10.1111/j.1365-2699.2004.01169.x (2005).

  16. 16.

    Sanmartín, I., Wanntorp, L. & Winkworth, R. C. West Wind Drift revisited: testing for directional dispersal in the Southern Hemisphere using event-based tree fitting. J. Biogeogr. 34, 398–416, https://doi.org/10.1111/j.1365-2699.2006.01655.x (2007).

  17. 17.

    Muñoz, J., Felicísimo, A. M., Cabezas, F., Burgaz, A. R. & Martínez, I. Wind as a long-distance dispersal vehicle in the Southern Hemisphere. Science 304, 1144–1147, https://doi.org/10.1126/science.1095210 (2004).

  18. 18.

    Moon, K. L., Chown, S. L. & Fraser, C. I. Reconsidering connectivity in the sub-Antarctic. Biol. Rev. 92, 2164–2181, https://doi.org/10.1111/brv.12327 (2017).

  19. 19.

    Lee, J. E., Terauds, A. & Chown, S. L. Natural dispersal to sub-Antarctic Marion Island of two arthropod species. Polar Biol. 37, 781–787, https://doi.org/10.1007/s00300-014-1479-9 (2014).

  20. 20.

    Hughes, K. A. & Convey, P. Determining the native/non-native status of newly discovered terrestrial and freshwater species in Antarctica – Current knowledge, methodology and management action. J. Environ. Manage. 93, 52–66, https://doi.org/10.1016/j.jenvman.2011.08.017 (2012).

  21. 21.

    Walther, G.-R. et al. Ecological responses to recent climate change. Nature 416, 389–395, https://doi.org/10.1038/416389a (2002).

  22. 22.

    Chown, S. L. et al. Continent-wide risk assessment for the establishment of nonindigenous species in Antarctica. Proc. Natl. Acad. Sci. USA 109, 4938–4943, https://doi.org/10.1073/pnas.1119787109 (2012).

  23. 23.

    Lee, J. E. & Chown, S. L. Breaching the dispersal barrier to invasion: quantification and management. Ecol. Appl. 19, 1944–1959, https://doi.org/10.1890/08-2157.1 (2009).

  24. 24.

    Shaw, J. D., Spear, D., Greve, M. & Chown, S. L. Taxonomic homogenization and differentiation across Southern Ocean Islands differ among insects and vascular plants. J. Biogeogr. 37, 217–228, https://doi.org/10.1111/j.1365-2699.2009.02204.x (2010).

  25. 25.

    le Roux, P. C. et al. Human activities, propagule pressure and the alien plants in the sub-Antarctic: evidence in support of management. Biol. Conserv. 161, 18–27, https://doi.org/10.1016/j.biocon.2013.02.005 (2013).

  26. 26.

    Chown, S. L. et al. Climate change and elevational diversity capacity: do weedy species take up the slack? Biol. Lett. 9, 20120806, https://doi.org/10.1098/rsbl.2012.0806 (2013).

  27. 27.

    Greve, M., Mathakutha, R., Steyn, C. & Chown, S. L. Terrestrial invasions on sub-Antarctic Marion and Prince Edward Islands. Bothalia 47, a2143, https://doi.org/10.4102/abc.v47i2.2143 (2017).

  28. 28.

    Marshall, W. A. Biological particles over Antarctica. Nature 383, 680, https://doi.org/10.1038/383680a0 (1996).

  29. 29.

    Turner, P. A. M., Scott, J. J. & Rozefelds, A. C. Probable long distance dispersal of Leptinella plumosa Hook.f. to Heard Island: habitat, status and discussion of its arrival. Polar Biol. 29, 160–168, https://doi.org/10.1007/s00300-005-0035-z (2006).

  30. 30.

    Ryan, P. G., Smith, V. R. & Gremmen, N. J. M. The distribution and spread of alien vascular plants on Prince Edward Island. Afr. J. Mar. Sci. 25, 555–562, https://doi.org/10.2989/18142320309504045 (2003).

  31. 31.

    Kellermann, J., Medan, D., Aagesen, L. & Hilger, H. H. Rehabilitation of the South American genus Ochetophila Poepp. ex Endl. (Rhamnaceae: Colletieae). N. Z. J. Bot. 43, 865–869, https://doi.org/10.1080/0028825X.2005.9512996 (2005).

  32. 32.

    Medan, D. & Devoto, M. Reproductive ecology of a perennial outcrosser with a naturally dissected distribution. Plant Syst. Evol. 254, 173–184, https://doi.org/10.1007/s00606-005-0335-3 (2005).

  33. 33.

    Medan, D. Anatomía y arquitectura foliares de Discaria (Rhamnaceae). Kurtziana 18, 133–151 (1986).

  34. 34.

    Chown, S. L. & Froneman, P. W. The Prince Edwards Islands. Land-sea interactions in a changing ecosystem. 1st edn, i-470 (Sun Press, 2008).

  35. 35.

    Chown, S. L. et al. The changing form of Antarctic biodiversity. Nature 522, 427–434, https://doi.org/10.1038/nature14505 (2015).

  36. 36.

    Richardson, J. E., Fay, M. F., Cronk, Q. C. B., Bowman, D. & Chase, M. W. A phylogenetic analysis of Rhamnaceae using rbcL and trnL-F plastid DNA sequences. Am. J. Bot. 87, 1309–1324, https://doi.org/10.2307/2656724 (2000).

  37. 37.

    Aagesen, L., Medan, D., Kellermann, J. & Hilger, H. H. Phylogeny of the tribe Colletieae (Rhamnaceae) — a sensitivity analysis of the plastid region trnL-trnF combined with morphology. Plant Syst. Evol. 250, 197–214, https://doi.org/10.1007/s00606-004-0204-5 (2005).

  38. 38.

    Tortosa, R. D. Una especie polimorfa de Discaria: D. chacaye (G. Don) comb. nov. (Rhamnaceae) y sus híbridos presuntivos. Parodiana, 79–98 (1983).

  39. 39.

    Webb, D. A. What are the criteria for presuming native status? Watsonia 15, 231–236 (1985).

  40. 40.

    Lewis Smith, R. I. & Richardson, M. Fuegian plants in Antarctica: natural or anthropogenically assisted immigrants? Biol. Invasions 13, 1–5, https://doi.org/10.1007/s10530-010-9784-x (2011).

  41. 41.

    Lockwood, J. L., Cassey, P. & Blackburn, T. M. The more you introduce the more you get: the role of colonization pressure and propagule pressure in invasion ecology. Divers. Distrib. 15, 904–910, https://doi.org/10.1111/j.1472-4642.2009.00594.x (2009).

  42. 42.

    Haussmann, N. S., Rudolph, E. M., Kalwij, J. M. & McIntyre, T. Fur seal populations facilitate establishment of exotic vascular plants. Biol. Conserv. 162, 33–40, https://doi.org/10.1016/j.biocon.2013.03.024 (2013).

  43. 43.

    van Kleunen, M. et al. Global exchange and accumulation of non-native plants. Nature 525, 100–103, https://doi.org/10.1038/nature14910 (2015).

  44. 44.

    Reyes, M. F., Gobbi, M. E. & Chaia, E. E. Reproductive ecology of Ochetophila trinervis in Northwest Patagonia. Funct. Plant Biol. 38, 720–727, https://doi.org/10.1071/fp11010 (2011).

  45. 45.

    Pirie, G. H. Aviation, apartheid and sanctions: air transport to and from South Africa, 1945–1989. GeoJournal 22, 231–240, https://doi.org/10.1007/BF00192821 (1990).

  46. 46.

    Pirie, G. H. Southern African air transport after apartheid. J. Mod. Afr. Stud. 30, 341–348, https://doi.org/10.1017/s0022278x00010752 (1992).

  47. 47.

    Nathan, R. Long-distance dispersal of plants. Science 313, 786–788, https://doi.org/10.1126/science.1124975 (2006).

  48. 48.

    Medan, D. Fruit morphogenesis and seed dispersal in the Colletieae (Rhamnaceae) I. The genus Discaria. Bot. Jahrb. Syst. Pflanzengesch. Pflanzengeogr. 105, 205–262 (1985).

  49. 49.

    Huiskes, A. H. L. et al. Aliens in Antarctica: Assessing transfer of plant propagules by human visitors to reduce invasion risk. Biol. Conserv. 171, 278–284, https://doi.org/10.1016/j.biocon.2014.01.038 (2014).

  50. 50.

    Keogh, J. A. & Bannister, P. Transoceanic dispersal in the amphiantarctic genus Discaria: an evaluation. N. Z. J. Bot. 31, 427–430, https://doi.org/10.1080/0028825X.1993.10419521 (1993).

  51. 51.

    Oosthuizen, W. C., Dyer, B. M. & de Bruyn, P. J. Vagrant birds ashore at the Prince Edward Islands, southern Indian Ocean, from 1987 to 2009. Afr. J. Mar. Sci. 31, 445–450, https://doi.org/10.2989/AJMS.2009.31.3.18.1006 (2009).

  52. 52.

    Hockey, P., Dean, R. & Ryan, P. Roberts birds of southern Africa. 7th edn, 1–1296 (John Voelcker Bird Book Fund, 2005).

  53. 53.

    Underhill, L. G. & Hofmeyr, J. H. Barn Swallows Hirundo rustica disperse seeds of Rooikrans Acacia cyclops, an invasive alien plant in the Fynbos Biome. Ibis 149, 468–471, https://doi.org/10.1111/j.1474-919X.2007.00598.x (2007).

  54. 54.

    Korczak-Abshire, M., Lees, A. & Jojczyk, A. First documented record of barn swallow (Hirundo rustica) in the Antarctic. Pol. Polar. Res. 32, 355–360, https://doi.org/10.2478/v10183-011-0021-9 (2011).

  55. 55.

    Garcia-Perez, B., Hobson, K. A., Powell, R. L., Still, C. J. & Huber, G. H. Switching hemispheres: a new migration strategy for the disjunct Argentinean breeding population of Barn Swallow (Hirundo rustica). PLoS One 8, e55654, https://doi.org/10.1371/journal.pone.0055654 (2013).

  56. 56.

    la Sorte, F. A. & Fink, D. Projected changes in prevailing winds for transatlantic migratory birds under global warming. J. Anim. Ecol. 86, 273–284, https://doi.org/10.1111/1365-2656.12624 (2017).

  57. 57.

    Veit, R. R. Vagrants as the expanding fringe of a growing population. Auk 117, 242–245, https://doi.org/10.1642/0004-8038(2000)117[0242:VATEFO]2.0.CO;2 (2000).

  58. 58.

    Frenot, Y. et al. Biological invasions in the Antarctic: extent, impacts and implications. Biol. Rev. 80, 45–72, https://doi.org/10.1017/S1464793104006542 (2005).

  59. 59.

    Duffy, G. A. et al. Barriers to globally invasive species are weakening across the Antarctic. Divers. Distrib. 23, 982–996, https://doi.org/10.1111/ddi.12593 (2017).

  60. 60.

    Reynolds, C., Cumming, G. S., Vilà, M. & Green, A. J. Birds as key vectors for the dispersal of some alien species: Further thoughts. Divers. Distrib. 23, 577–580, https://doi.org/10.1111/ddi.12549 (2017).

  61. 61.

    Burge, D. O. et al. Diversification of Ceanothus (Rhamnaceae) in the California Floristic Province. Int. J. Plant Sci. 172, 1137–1164, https://doi.org/10.1086/662028 (2011).

  62. 62.

    Kellermann, J. & Udovicic, F. Large indels obscure phylogeny in analysis of cpDNA (trnL-F) sequence data: Pomaderreae (Rhamnaceae) revisited. Telopea 12, 1–22 (2008).

  63. 63.

    Hauenschild, F., Matuszak, S., Muellner-Riehl, A. N. & Favre, A. Phylogenetic relationships within the cosmopolitan buckthorn family (Rhamnaceae) support the resurrection of Sarcomphalus and the description of Pseudoziziphus gen. nov. Taxon 65, 47–64, https://doi.org/10.12705/651.4 (2016).

  64. 64.

    Taberlet, P., Gielly, L., Pautou, G. & Bouvet, J. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Mol. Biol. 17, 1105–1109, https://doi.org/10.1007/bf00037152 (1991).

  65. 65.

    Guindon, S. et al. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst. Biol. 59, 307–321, https://doi.org/10.1093/sysbio/syq010 (2010).

  66. 66.

    Huelsenbeck, J. P. & Ronquist, F. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755, https://doi.org/10.1093/bioinformatics/17.8.754 (2001).

Download references

Acknowledgements

This work was supported by South African National Research Foundation Grant 110734. The plant specimen was collected under permit number 6/2015 issued to M.G. Eugenia Chaia kindly collected, at the authors’ request, material of Ochetophila trinervis for anatomical purposes. Greg Hofmeyr provided data from 2004. Norberto H. Montaldo helped with the compilation of avian information. We acknowledge support by Deutsche Forschungsgemeinschaft and Open Access Publishing Fund of Karlsruhe Institute of Technology.

Author information

Author notes

    • Jesse M. Kalwij

    Present address: Institute of Geography and Geoecology, Karlsruhe Institute of Technology, Reinhard-Baumeister-Platz 1, 76131, Karlsruhe, Germany

Affiliations

  1. Centre for Ecological Genomics and Wildlife Conservation, Department of Zoology, University of Johannesburg, Auckland Park, 2006, South Africa

    • Jesse M. Kalwij
  2. Cátedra de Botánica General, Facultad de Agronomía, Universidad de Buenos Aires, Buenos Aires, Argentina

    • Diego Medan
  3. Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina

    • Diego Medan
  4. State Herbarium of South Australia, Department for Environment and Water, GPO Box 1047, Adelaide, South Australia, 5001, Australia

    • Jürgen Kellermann
  5. The University of Adelaide, School of Biological Sciences, Adelaide, South Australia, 5005, Australia

    • Jürgen Kellermann
  6. Department of Plant and Soil Sciences, University of Pretoria, Private Bag X20, Hatfield, 0028, South Africa

    • Michelle Greve
  7. School of Biological Sciences, Monash University, Clayton, Victoria, 3800, Australia

    • Steven L. Chown

Authors

  1. Search for Jesse M. Kalwij in:

  2. Search for Diego Medan in:

  3. Search for Jürgen Kellermann in:

  4. Search for Michelle Greve in:

  5. Search for Steven L. Chown in:

Contributions

J.M.K., M.G. and S.L.C. conceived the ideas; J.M.K. and M.G. collected the data; D.M. and J.K. conducted the morphological and genetic analysis; J.M.K. led the writing. All authors contributed to the data interpretation and writing.

Competing Interests

The authors declare no competing interests.

Corresponding author

Correspondence to Jesse M. Kalwij.

Supplementary information

About this article

Publication history

Received

Accepted

Published

DOI

https://doi.org/10.1038/s41598-019-41081-9

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.