Introduction
The traditional delimitation of the genus Medicago L. underwent significant revision several decades ago, when species of Trigonella L. that shared the characteristic tripping mechanism of the flowers with Medicago – the so-called “medicagoid Trigonella” – were transferred to Medicago (Small 1987; Small et al. 1987; Small and Jomphe 1989; Bena 2001). This reclassification substantially increased the number of recognized species within the genus. In its current circumscription, the genus comprises 91 species and hybrids, predominantly distributed across temperate and subtropical regions of Eurasia and the Mediterranean (POWO 2025).
Medicago includes species of considerable economic importance, most notably the forage crop alfalfa (M. sativa L.), and has therefore been the focus of extensive research across a wide range of disciplines, including molecular phylogenetics (Bena 2001; Chen et al. 2021). A comprehensive global monograph of the genus was published relatively recently (Small 2011), yet new taxa continue to be described. The most recent example is M. tetraprostrata J.S.Erikss. & B.E.Pfeil, a cryptic species believed to result from hybridization between M. prostrata Jacq. and M. sativa (Eriksson et al. 2017).
Over the past few decades, an unexpectedly large number of new Medicago species have been described, the majority of which originate from the Middle East – widely recognized as the primary centre of diversity for the annual taxa within the genus (Small 2011). These include M. syriaca E.Small, native to Syria and southwestern Iran (Small 1990a); M. rigiduloides E.Small, ranging from the eastern Mediterranean to Central Asia (Small 1990b); M. heldreichii E.Small and M. huberi E.Small from southwestern Turkey (Small et al. 1987); M. laxispira Heyn from Iraq; M. makranica Heyn from Pakistan (Heyn 1984); and M. hypogaea E.Small, which occurs from the eastern Mediterranean to Iraq (Small and Brookes 1984). While this concentration of recent discoveries is consistent with the region’s status as a biodiversity hotspot for the genus, it also underscores how incompletely Medicago has been studied in its core range. Notably, molecular data remain unavailable for many of these species, limiting our ability to rigorously test morphology-based species delimitations and to assess their phylogenetic placement and biogeographic origins with confidence.
The presence of spiny fruits in many Medicago species has facilitated their unintentional spread as weeds across much of the world, primarily through epizoochorous dispersal. In particular, the fruits readily attach to animal skins – most notably the fleece of sheep and other wool-bearing animals – allowing species to be inadvertently transported over long distances to regions outside the genus’ native range, such as Australia, South Africa, and South America. As a result, species like M. polymorpha L. have become naturalized in many parts of the world. Medicago species are among the most characteristic so-called “wool aliens”, a term used for alien plants introduced via contaminated wool. In Australia, for instance, they are considered some of the most common and troublesome contaminants in raw wool, posing significant challenges for processing (Australian Wool Testing Authority 1986). In addition to the morphological adaptations that enhance dispersal, Medicago seeds are water-impermeable (“hard seeds”), a trait that enables them to persist in the soil seed bank for extended periods – often decades, or even over a century – without losing viability (Small 2011).
In eastern Belgium, the valley of the Vesdre River, particularly around the town of Verviers, was historically a major centre of the wool processing industry. For over a century, raw wool was imported from distant regions, primarily from the Southern Hemisphere (notably Australia, South Africa, and South America), but also from the Middle East and North Africa (Visé 1942, 1958). During processing, the wool was washed in local facilities, and the wastewater – containing dirt, as well as fruits and seeds – was discharged into the river. Downstream of the factories, under favourable conditions (especially during hot, dry summers), many alien plant species germinated. In this way, exotic species were introduced continuously into the local flora for more than a century. This phenomenon of wool-mediated plant introduction has received considerable attention in other parts of Europe as well (e.g. Hayward and Druce 1919; Probst 1949; Lousley 1961; Ryves 1974). However, from the late 1960s onwards, the direct discharge of untreated wastewater into the river ceased for environmental reasons, and the wool alien flora gradually disappeared.
In July 2021, the Vesdre Valley was severely affected by catastrophic flooding, which caused widespread destruction throughout the region. As a consequence, sediments along the riverbanks were displaced, exposing buried seeds that had remained dormant for decades. In the summer of 2022, a remarkably rich and diverse adventive flora emerged on the newly disturbed alluvial soils. This flora was composed predominantly of species from the regions that historically supplied raw wool – Australia, South Africa, and South America – but also included taxa from North Africa, the Middle East, and the Indian subcontinent (Verloove et al. 2023, in press). In addition to widespread cosmopolitan weeds, several rare and endemic species were encountered, some with remarkably restricted native ranges. Notably, several taxa had not even been formally described at the time of their introduction, including, for example, Sida sivarajanii Tambde, Sardesai & A.K.Pandey, a species recently described from India (Tambde et al. 2020).
Among the species that emerged following the 2021 floods was a taxon of Medicago previously unknown to science. Extensive comparison with all morphologically similar and phylogenetically related Medicago species, based on the most recent global monograph (Small 2011) and publicly available molecular data, revealed no match. The Vesdre specimen is thus recognized as a species new to science. Its discovery highlights the value of combining molecular and morphological evidence to document previously unrecorded diversity.
Discussion
Medicago peregrina is distinct both at the molecular and morphological level, representing a lineage separate from all previously described Medicago species.
Distinctiveness of <italic><tp:taxon-name><tp:taxon-name-part taxon-name-part-type="genus" reg="Medicago">Medicago</tp:taxon-name-part> <tp:taxon-name-part taxon-name-part-type="species" reg="peregrina">peregrina</tp:taxon-name-part></tp:taxon-name></italic>
<italic>Molecular phylogenetic evidence</italic>
Molecular phylogenetic analyses confirm that the Vesdre specimen represents a distinct Medicago lineage, with closest affinities to M. heyniana (GA3ox1 and matK) and a broader relationship to M. polymorpha, M. sativa, and allied taxa (see Results). These findings corroborate the utility of ITS, GA3ox1, and matK for species-level delimitation in Medicago (Steele et al. 2010; Chen et al. 2021).
<italic>Morphological distinctness</italic>
Medicago peregrina differs from M. heyniana in several diagnostic traits.
<italic>Seed morphology</italic>
M. heyniana possesses rugose, transversely ridged seeds – a key character that led to the establishment of sect. Heynianae. In contrast, M. peregrina has smooth seeds that are significantly smaller in size (ca 1.8 vs 3 mm) (Fig. 4C).
<italic>Fruit morphology</italic>
Fruits of M. peregrina are smaller (6 mm vs 8–12 mm), blacken at maturity (vs remaining yellowish grey), are hairy (vs glabrous), and typically consist of only two loosely coiled turns (vs 2.5–4.5) (Figs 2E, 2G, 3, 4A, 4B).
<italic>Stipules</italic>
Entire to shallowly toothed near the base in M. peregrina, vs deeply laciniate in M. heyniana (Fig. 2D).
<italic>Floral characters</italic>
The new species bears consistently single, smaller flowers (4.5 vs 5–9 mm), whereas M. heyniana often produces one or two (Fig. 2A).
<italic>Indumentum</italic>
Medicago peregrina is conspicuously covered with papillose hairs that, at low magnification, appear multicellular – a feature that is relatively rare within the genus (for terminology, see Hewson 1988) (Fig. 4D–F). Though we were unable to examine this trait in M. heyniana directly, Greuter (1970) described its indumentum as consisting of simple, subappressed hairs. Given the striking appearance and potential diagnostic value of papillose hairs, it is unlikely that their presence would have gone unnoticed. The occurrence of these papillose hairs – superficially resembling multicellular hairs and otherwise reported only for a few Medicago species such as M. arabica (L.) Huds. and M. ciliaris (L.) All. (Mehregan et al. 2002) – further supports the distinctness of M. peregrina. This observation also raises the question whether similar reinterpretations might apply to other Medicago taxa previously described as bearing multicellular hairs, and whether such homoplastic structures reflect convergence or point to deeper, unresolved relationships within the genus. Detailed SEM-based studies of M. arabica, M. ciliaris, and related taxa will be essential to clarify whether their trichomes are truly multicellular or, likewise, papillose in nature.
<italic>Study limitations</italic>
Standard approaches for describing a new species typically require multiple samples from across the native range (e.g. Möller et al. 2024). In this case, only a single individual was available. Nevertheless, the combination of molecular and morphological evidence strongly supports recognition of Medicago peregrina as a distinct species, though its full genetic and morphological variation, as well as geographic range, remain unknown.
Historical context and the phenomenon of wool alien plants
Describing a species as new to science without knowledge of its area of origin is uncommon practice. One illustrative example is Artemisia verlotiorum Lamotte, an invasive species first described from France, which was later found to be native to China (Verloove et al. 2021). Similarly, researchers studying wool alien flora have often described species based on plant material imported with raw wool, without any certainty regarding their native habitat. In most cases, these species were later identified in their areas of origin, often after attention had been drawn to them following their formal description as new species.
Numerous examples exist of such occurrences. For instance, Thellungia advena Stapf, now recognized as Sporobolus advenus (Stapf) P.M.Peterson, was first described from Solothurn, Switzerland (Stapf 1920) and subsequently identified as originating from eastern Australia. Several species of Lepidium were similarly described from plant material imported with wool. Lepidium calycinum Godron (an illegitimate name, now replaced by L. auriculatum Regel & Körn.) (Godron 1853) and L. pseudodidymum Thell. (Thellung 1914) were initially discovered in France and Scotland, respectively, before being documented in their native regions. Likewise, Lepidium peregrinum was first described from Scotland (Thellung 1913; Hayward and Druce 1919), only to be found later in Australia. Similarly, L. fallax Ryves, described from England, is almost certainly native to South America, although it has yet to be formally recorded there (Ryves 1977). Additionally, the Australian species Millotia depauperata Stapf was originally described from plant material found in Scotland (Stapf 1910; Short 1995).
Other taxa, once described from wool alien plant material, have since been synonymized. Notable examples include Nassella flaccidula var. glomerata Hack. ex Druce (now N. pubiflora (Trin. & Rupr.) É.Desv.) and Schkuhria advena Thell. (now S. pinnata (Lam.) Kuntze ex Thell.). More broadly, in instances where plant species are described without clear knowledge of their area of origin, they are often given epithets such as advena, aliena, or similar derivatives (e.g. Wipff and Veldkamp 1999).
Rare species as wool aliens
It is evident that common and widespread weeds are more likely to be transported to other regions as contaminants in raw wool. Nevertheless, rare species or those with a very limited native range have also regularly been found in wool alien floras. Several recent examples were discovered along the banks of the Vesdre River, including Abutilon nobile Domin, a species endemic to central Queensland (the Hughenden-Emerald area) in Australia, and Sida sivarajanii, a recently described species with a restricted distribution in parts of India (Tambde et al. 2020; Verloove et al. in press). In addition to the newly described Medicago peregrina, an unidentified Lupinus species – likely originating from the Andes and not matching any currently known taxa (A. Planchuelo, Argentina, pers. comm.) – was also found. Furthermore, two South African species of Hermannia L., known but not yet formally described (D. Gwynne-Evans, South Africa, pers. comm.), were also recorded along the Vesdre River. These likely represent rare taxa with restricted native ranges.
A similar case from the historical literature is that of Lepidium peregrinum, a species now nearly extinct in its native Australia (Scarlett 1999) but originally described from Scotland as a wool alien. Although the presence of such geographically restricted species in wool-importing regions may seem paradoxical, it is a phenomenon that has long puzzled researchers (e.g. Hayward and Druce 1919: xxvi), but it can be easily explained. In the past, sheep grazed over vast, unfenced areas, granting them access to locations often inaccessible to humans. This extensive mobility likely facilitated the unintentional collection and transport of seeds or fruits from species that had not yet been formally described.
Origin of <italic><tp:taxon-name><tp:taxon-name-part taxon-name-part-type="genus" reg="Medicago">Medicago</tp:taxon-name-part> <tp:taxon-name-part taxon-name-part-type="species" reg="peregrina">peregrina</tp:taxon-name-part></tp:taxon-name></italic> and the wool alien hypothesis
The discovery of Medicago peregrina fits the well-known pattern of wool alien species: plants described from regions far from their native habitat, arriving as contaminants in raw wool. While we cannot be absolutely certain that this species was introduced via wool, the historical dominance of wool-processing in the Vesdre Valley (Visé 1942, 1958), the scarcity of alternative industrial vectors, and the simultaneous occurrence of nearly 200 other unusual adventive species – most of Australian origin – strongly support this explanation (Verloove et al. in press). Taken together, these observations make a wool-related introduction pathway by far the most plausible scenario.