The first record of Salsola paulsenii (Chenopodiaceae) in Ukraine, with taxonomic and nomenclatural comments on related taxa

Salsola paulsenii (Chenopodiaceae) is reported for Kyiv: several plants were found (September 2017) in a sandy ruderal area of the northwestern part of Obolon, along railway tracks leading to the grain elevator and brewery facilities of the Obolon Joint-Stock Co. Some co-occurring plants may represent hybrids of S. paulsenii and S. tragus. Data on morphology and distribution of S. paulsenii and related species are provided. If established, this species may become an invasive weed in Ukraine. Recent changes in taxonomy and nomenclature of Salsola sensu lato are discussed. Nomenclatural information is provided for species of the Ukrainian flora formerly reported in Salsola sensu lato and now placed in Salsola sensu stricto (incl. Kali), Soda, Caroxylon, Climacoptera, and Pyankovia.

Most species of Salsola sensu stricto occur in the continental arid regions of Asia, with some taxa extending or native to western coastal Europe, the Mediterranean region, coastal Asia, and Australia.Despite the revision by Rilke (1999), the taxonomy of certain entities of the S. tragus aggregate remains problematic.In particular, there are probably some additional taxa (including cryptic species) in continental Asia, yet undescribed or already described but not recognized in recent taxonomic treatments.In addition to S. australis R. Br. recently taxonomically restored by Hrusa and Gaskin (2008), several other Australian species such as S. macrophylla R. Br. and infraspecific taxa of "S.kali" described by Bentham (1870), which were included in a polymorphic S. kali by Wilson (1984) and in S. tragus by Rilke (1999), will be probably also restored following S. pontica as a synonym of S. tragus in that "coastal" sense.Michalková and Letz (2014) characterized that coastal taxon as having "distinctly succulent leaves, bracts and bracteoles; bracts long spine-tipped".However, anyone can see that the lectotype of S. tragus (LINN 315.3; digital image available from http://linnean-online.org/3160/), in contrast to another original specimen (LINN 315.5, not the lectotype!), is a plant with rather narrow and almost non-succulent leaves and not so prominently spiny bracts and bracteoles, quite well corresponding to the current concept of S. tragus.The issue of lectotypification of S. tragus was exhaustively discussed by Tzvelev (1993), Rilke (1999) and other authors, so there is no need to change again its current well-established application.
Several species of Salsola were unintentionally introduced by human activities (mainly agriculture and trade) far beyond their native ranges and became invasive weeds in many parts of the world.In particular, the iconic tumbleweeds of the North American Southwest are in fact quite recent invaders.The most widespread weedy species S. tragus (also widely known in the past as S. pestifer A. Nelson, S. ruthenica Iljin, nom.illeg., S. kali subsp.ruthenica Soó, S. iberica (Sennen & Pau) Botsch.ex Czerep., and under the misapplied names S. australis auct.non R. Br. and S. kali auct.p.p., non L.) started its colonization of North America only closer to the end of the 19 th century (Beatley, 1973;Crompton, Bassett, 1985;Mosyakin, 1996Mosyakin, , 2003;;Beckie, Francis, 2009 and references therein).However, as it has been demonstrated by recent thorough studies using molecular, morphological and ecological approaches (Ryan, Ayres, 2000;Gaskin et al., 2006;Ryan et al., 2007;Hrusa, Gaskin, 2008;Smith et al., 2013;Welles, Ellstrand, 2016a, b), in southwestern North America (and in California in particular) several species introduced from Eurasia (S. tragus, S. paulsenii Litv.) and Australia (S. australis) are involved in active microevolutionary processes resulting in the emergence of highly invasive weedy races, such as the recently identified allohexaploid S. ryanii Hrusa & Gaskin (= Kali ryanii (Hrusa & Gaskin) Brullo & Hrusa) that originated from hybridization of diploid S. australis and tetraploid S. tragus.Now this hybridogenous species is rapidly spreading in California (Welles, Ellstrand, 2016a).Moreover, it has been recently demonstrated that "this neo-invasive is the result of three independent admixture events and that gene flow among the individual lineages is just beginning, leading to withinspecies admixture" (Ellstrand, Rieseberg, 2016: 834; more thorough studies (see preliminary data in Borger et al., 2008;Chinnock, 2010).Synonymization of the Australian taxon S. macrophylla with the Pontic (and eastern Mediterranean?) S. pontica (Pall.)Degen (= Kali ponticum (Pall.)Sukhor., S. tragus subsp.pontica (Pall.)Rilke) cursorily done by Rilke (1999) and later accepted by Galasso and Bartolucci (2014), who coined the combination Kali macrophyllum (R. Br.) Galasso & Bartolucci, seems to be at least questionable.Their identity has not been proven yet beyond doubt, and until such a proof is available, I prefer to use the name S. pontica for our plants.This taxon, which was described from the Black Sea shores of Crimea, Ukraine, as S. kali var.pontica Pall., belongs to a problematic group of Pontic-Mediterranean coastal taxa currently known as S. pontica (either sensu lato or sensu stricto) and Kali dodecanesicum C. Brullo, Brullo, Giusso & Ilardi (Brullo et al., 2015b).No nomenclatural combination in Salsola is currently available for the latter taxon; however, it is quite possible that it is in fact conspecific with S. controversa Tod.ex Lojac.described from Sicily (Lojacono Pojero, 1904: 271-272), which in that case will be the correct name for this Mediterranean species (if it is indeed specifically distinct from S. pontica), or for the whole Pontic-Mediterranean coastal group (if treated as one species, incl.S. pontica).Of course, the identity of two earlier species-rank names (S. macrophylla described from Australia and S. brachypteris Moq.described from Java, Indonesia) mentioned by Rilke (1999) in synonymy of her S. tragus subsp.pontica should be critically assessed as well.It should be noted that in IPNI (2017-onward) and in some other sources S. controversa was reported with the authorship "Tod.ex Nyman -Consp.Fl.Eur.3: 631.1881"; however, that name was mentioned by Nyman in synonymy only and without any description or diagnosis; thus, it was not validly published in 1881.In contrast, Lojacono Pojero (1904) accepted that species in his Flora Sicula and provided its rather detailed description.
There are two alien species of Salsola currently known in Ukraine: S. collina Pall.(Kali collinum (Pall.)Akhani & Roalson, as "collina") and S. paulsenii, the Welles, Ellstrand, 2016b).Also, the first cases of fieldevolved Glyphosate-resistant races of weedy S. tragus were recently reported from Montana and Washington states of the USA (Kumar et al., 2017), which indicates the ongoing microevolutionary processes resulting in herbicide resistance in some naturalized populations of the species in North America.
Salsola sensu lato in Ukraine: previous data Iljin (1952) reported for Ukraine seven species of Salsola (plus one currently recognized species, S. pontica (Pall.)Degen, was accepted as a variety of S. tragus, and S. collina Pall.was mentioned in a footnote), of which only three species still remain in that genus.The names of taxa accepted in main relevant floristic publications after Iljin (1952) are summarized in Table 1.It shows that currently the Ukrainian taxa listed by Iljin (1952) and  Iljin (1952: 386-387).Baranova and Khilova (1990) provided on their distribution map a dot in eastern Ukraine (corresponding to Chuguyiv?), but no precise locality or herbarium specimens were reported.Reliable and herbarium-confirmed collections of S. collina were made later in several habitats along railways and near grain elevators in Kyiv (Mosyakin, 1988(Mosyakin, , 1991)).Since then S. collina is known in Ukraine from several other scattered localities (e.g., Slavuta, Khmelnytsky Region, railway tracks at a grain elevator; collected by Mosyakin in 2001 and reported in Gubar, 2004: 71) and in the Southeast (Ostapko et al., 2010).It still remains in Ukraine a rather uncommon "railroad' plant forming occasionally, however, quite large but local colonies, predominantly in gravel and crushedstone track ballast, or sometimes in nearby sandy habitats.This species is occasionally found in Europe beyond its native range (Aellen 1960(Aellen -1961;;Tzvelev, 1996;Rilke 1999;Vidéki, 2005 etc.) and is spreading in North America (Mosyakin, 1996(Mosyakin, , 2003;;Blondeau et al., 2007;Oldham et al., 2011, and references therein).

The first record of Salsola paulsenii in Ukraine
During a routine survey of ruderal habitats within the city of Kyiv for search and monitoring of alien species, on 18 September 2017 I found S. paulsenii in the northwestern part of the city, Obolon (Minsky administrative district), along railway tracks near the grain elevator and brewery facilities of the Obolon Joint-Stock Co. (Latitude: N 50° 30' 34.1879"; Longitude: E 30° 28' 47.0765").There were four large and welldeveloped plants growing between railway tracks (Figs 1, 2) and several smaller and rather underdeveloped plants (some probably affected by herbicides) in a narrow sandy strip along the tracks, ca. 10 meters from the larger plants.Accompanying species were Salsola tragus L. (several morphotypes), Sisymbrium S. tragus sensu lato were morphologically transitional toward individuals of S. paulsenii growing nearby.These atypical plants grew in stands of S. tragus sensu stricto and, being habitually similar to typical forms of that species, had elongated tips of perianth segments (tepals) forming a column (Fig. 3).However, these perianth tips were shorter and less rigid than in typical S. paulsenii (Table 2).Morphologically these plants fit descriptions of variable presumable hybrids between S. tragus and S. paulsenii (reported as S. gobicola Iljin, fide Rilke 1999), or probably the morphotype reported from southwestern North America under the informal name "S. paulsenii lax form" (see Hrusa, Gaskin, 2008;Ayres et al., 2009 and references therein).The identity and origin of the North American "lax form" remain obscure.Since that form contained alleles from S. tragus, S. paulsenii, and S. australis, plus two unique alleles, it may be a complex hybrid involving three or more species (Ayres et al., 2009).Morphological loeselii L., Grindelia squarrosa (Pursh) Dunal., Setaria viridis (L.) P. Beauv., Eragrostis minor Host, Polygonum aviculare L. aggr., Chenopodium betaceum Andrz.(= C. strictum auct.non Roth; for nomenclatural and taxonomic updates see Mosyakin, 2017), C. album L. aggr.; less commonly also Artemisia vulgaris L., Atriplex tatarica L., Bassia laniflora (S.G.Gmel.)A.J. Scott (= Kochia laniflora (S.G.Gmel.)Borbás), Centaurea stoebe L. aggr., Digitaria sanguinalis (L.) Scop., Erigeron canadensis L. (= Conyza canadens (L.) Cronquist), Medicago sativa L., and some other plants rather common in ruderal habitats in Kyiv.
Salsola paulsenii differs from the related widespread species S. tragus (rather common in Ukraine) in several quite reliable morphological characters, which are discussed below.A closely related species, S. praecox (Litv.)Iljin (= S. kali var.praecox Litv.; S. paulsenii subsp.praecox (Litv.)Rilke; Kali praecox (Litv.)Sukhor.), which is restricted mainly to Middle/Central Asia, differs from S. paulsenii in being usually less robust and smaller, more pubescent, normally having shorter and narrower leaves, smaller anthers, and different phenology (flowering/fruiting in April-August, as compared to June-November for S. paulsenii).
The main characters distinguishing S. tragus, S. paulsenii, and S. pontica, as well as S. australis and S. ryanii, which may be expected in Ukraine in the future as occasional aliens, are summarized in Table 2.Among these characters, the most obvious ones are the long protruding column formed by the tips of perianth segments above the wings, and bracts gradually tapering into and ending with stiff spines up to 3 (-3.5)mm long (Fig. 2).In S. tragus the perianth tips are lax and much shorter, and bracts are more abruptly narrowed into a weaker and shorter spine or mucro (Fig. 4).The mature winged perianth in S. paulsenii is usually wider than that in S. tragus; although sometimes fruits with very diverse wings can be found even on one plant: broadly winged, with wings eroded or dentate at margins, to almost wingless.All these variations were observed in the specimens collected in Kyiv in the Obolon locality.The mentioned wing characters partly depend on developmental stage, position of flowers/ fruits on the branches, and probably some other factors.Moreover, perfectly winged fruits in S. paulsenii are usually developed throughout the plant (almost from the base) and at the end of the flowering/fruiting season (Table 2).Lower and middle branches of S. paulsenii are characters reported for S. gobicola in the protologue (Iljin, 1955) seem to be perfectly intermediate between those of S. paulsenii and S. tragus.Additional studies are needed to confirm or refute the possible hybrid nature and origins of those morphologically rather variable Eurasian and North American plants.
Mature fruits of S. paulsenii are enclosed in five persistent perianth segments, three of which usually bear rather wide horizontal wing-like appendages.The species is thus well-adapted to dispersal by wind (anemochory).However, in our case it is evident that diaspores of S. paulsenii were introduced to the locality with grain transported by railroad to the elevator and brewery.Possible immediate geographical sources of introduction are discussed below.Considering the potential invasive nature of S. paulsenii and the availability of nearby sandy habitats suitable for further spread of the species, I decided to eradicate the discovered plants by manual weeding; representative specimens were herbarized and deposited at KW.   3. Plant morphologically intermediate between Salsola paulsenii and S. tragus: close-up of a branch with fruits and bracts/bracteoles.Note the long tips of perianth segments (but less rigid than in S. paulsenii, and reflexed at apices), the bracts and bracteoles more abruptly contracted into shorter spines, and irregular (eroded or dentate) wing margins regarding the actual occurrence of S. paulsenii in North America and commented that American records of this species need confirmation; however, she also stated that the identification of a specimen from California (Twisselmann 17472 at B) can be confirmed with certainty ["…die amerikanischen Angaben dieser Art eine Überprüfung benötigen!Hier kann ein Beleg aus Kalifornien Twisselmann 17472 (B) zweifelsfrei bestätigt werden."].Since then, in addition to identifications of North American specimens made or confirmed by Botschantzev and Mosyakin, the rather widespread occurrence of S. paulsenii in southwestern North America has been further confirmed by integral studies with application of molecular and micromorphological methods (Gaskin et al., 2006;Hrusa, Gaskin, 2008;Ayres et al., 2009).Moreover, S. paulsenii participates in microevolutionary processes involving some other alien species of Salsola, and occasionally hybridizes with S. tragus, forming variable hybrids, some known as usually rather long, horizontally spreading to slightly arcuate, with secondary branches forming almost a 90° angle with the primary branches.Finally, it should be taken into account that S. paulsenii is a rather variable species, identification of morphologically intermediate (presumable hybrids), deviant or underdeveloped plants may be problematic, and thus consulting reliably identified herbarium specimens for comparison is advised.
The main portion of the native range of S. paulsenii covers mostly sandy, sandy-steppe, or sandy-saline habitats of the desert and semidesert zones all around the Caspian Sea (including eastern Azerbaijan and adjacent regions of Russia), Middle and Central Asia (Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, Uzbekistan, eastern and northern Iran, Afghanistan, western Pakistan, western China, southwestern Mongolia) (Iljin, 1936;Grubov, 1966;Pratov, 1972;Rilke, 1999: 159, map in Fig. 75;Freitag, Rilke, 1997;Freitag, 2001;Zhu et al., 2003;Sukhorukov, Akopian, 2013;Sukhorukov, 2014 etc.).In the European part of Russia S. paulsenii was reported as a native species from Astrakhan, Orenburg and Volgograd administrative regions and the Kalmyk Republic (Tzvelev, 1996;Sukhorukov, 2014), and as a rare alien collected in Moscow along railroads (Ignatov et al., 1990;Mayorov et al., 2012).However, Sukhorukov (2014: 335) commented that the species is common in Astrakhan Region but only old herbarium specimens are known from Orenburg and Volgograd regions, and the records from Moscow Region in fact belong to S. tragus.
The species is also introduced and naturalized in the southwestern states of the USA (Beatley, 1973;Evans, Young, 1980;Mosyakin, 1996, 2003, andreferences therein), where it is known under the vernacular name "Barbwire Russian-thistle", because of its rigid and very prickly branches.Rilke (1999: 158) expressed her doubts Considering that, it is quite possible that diaspores of S. paulsenii were brought to Kyiv with grain from southern Ukraine, which is also most probably true for Grindelia squarrosa, an alien of North American origin that is now naturalized and widespread in the southern and some other regions of Ukraine, and which was cooccurring with S. paulsenii in the Obolon locality.If it is the case, there might be yet unnoticed localities of S. paulsenii (introduced earlier from the native range or from North America?) in other regions of Ukraine, especially areas of commercial grain production in the south.Judging from its ecological and biogeographic peculiarities and the history of its spread in North America, S. paulsenii has a potential to become a noxious environmental (or even agricultural?)weed in Ukraine, especially in the southern regions.The potential spread of that species, however, may be limited by its ecological preferences (mostly open sands and also sandy soils).Despite that, it may occupy in Ukraine the sandy habitats ecologically identical or similar to those already successfully invaded by another prickly alien, Cenchrus longispinus (Hack.)Fernald, which is now locally common and invasive in southern regions of Ukraine, and also in Kyiv (see Mosyakin, 1995;Schanzer et al., 2017).Salsola paulsenii may also hybridize with local weedy races of S. tragus, forming potentially invasive stabilized hybrids.Because of that, Ukrainian botanists and agronomists should keep their eyes open for this potentially dangerous new weed.

Possible sources of introduction of Salsola paulsenii to Ukraine, and future prospects
In view of the current knowledge of distribution patterns of S. paulsenii and considering that its diaspores were evidently brought with grain used for beer brewing at Obolon Co., there are three possible geographical sources for the present introduction of the species to Kyiv: (1) directly from the primary native range in Middle/Central Asia or southeasternmost European Russia; (2) from the secondary range in North America, and (3) from some unidentified areas in southern Ukraine where this species may be spreading unnoticed.We cannot exclude any of those possibilities.However, at present Ukraine is not an importer but exporter of grain (wheat, barley etc.).For example, official data indicate that in 2016-2017 Ukraine exported barley to various countries (the most important being Saudi Arabia, China, Jordan, countries of the European Union, Libya etc.).The total Ukrainian export of grain and grain products at present is also overwhelmingly higher than import: for example, 38 million tons of export against just 171 thousand tons of import (only 0.4% of export!) for the period of July 2016 -April 2017 [Agravery -Agrarian Information Agency: http://agravery.com/uk/posts/show/import-zerna-v-ukrainu-za-10-misacivsklav-lise-171-tis-tonn (text in Ukrainian)].Imported grain is represented mainly by rice and corn (maize).These changes in Ukraine's export/import of grain considerably diminished the occurrence of occasional "grain immigrants" (ephemerophytes, non-established alien plants) in the alien flora of Kyiv, as compared to the 1980s and 1990s (Mosyakin, 1991, and unpublished data).Importation of agricultural products from Russia to Ukraine recently shrunk dramatically; the same is true for import of agricultural products from Kazakhstan to Ukraine, where the share of grain is also now nearly negligible [AgGeek: http://aggeek.net/ru/markets/id/ukraina-i-kazahstan-shans-podelit-shkuru-russkogomedvedja-154/ (text in Russian)].
Direct introduction from North America to Kyiv appears to be less possible because importation of grain and grain products from the US and Canada is now very limited.However, agricultural companies of Ukraine, which operate mainly in the southern and central regions of the country, continue importation of highquality seed (mainly sunflower, corn, soybeans etc.), Мосякин С.Л.Первая находка Salsola paulsenii (Chenopodiaceae) в Украине и номенклатурнотаксономические комментарии о родственных таксонах.Укр. бот. журн., 2017, 74(5)

Fig. 1 .
Fig. 1.Salsola paulsenii growing at railway tracks in Kyiv, Obolon.All photographs made in the field in the same locality

Fig.
Fig. 3. Plant morphologically intermediate between Salsola paulsenii and S. tragus: close-up of a branch with fruits and bracts/bracteoles.Note the long tips of perianth segments (but less rigid than in S. paulsenii, and reflexed at apices), the bracts and bracteoles more abruptly contracted into shorter spines, and irregular (eroded or dentate) wing margins

Fig. 4 .
Fig. 4. Salsola tragus: close-up of a branch with fruits and bracts/bracteoles: forms with white (a) and rose-colored (b) wings.Note the short and weak tips of perianth segments, and the bracts and bracteoles more abruptly narrowed into spines

Table 1 . Salsola sensu lato in Ukraine: names used in major floristic treatments, and currently accepted names
Salsola collinawas probably found for the first time in Ukraine by M.V. Klokov and M.I.Kotov near railway tracks in Chuguyiv, Kharkiv Region (mentioned inIljin 1952:  387, in adnot., with an exclamation mark indicating that Iljin has seen the specimen).