Atlas Florae Europaeae notes, 35. Further critical notes on Cytisussect.Tubocytisus (Fabaceae) in Europe

Abstract A few species names in Cytisussect.Tubocytisus are re-assessed and taxonomically evaluated. Diagnostic characters are discussed and the species status of C.absinthioides Janka, C.eriocarpus Boiss., C.frivaldszkyanus Degen, C.jankae Velen. and C.smyrnaeus Boiss. is confirmed. The holotype of Cytisustriflorus Lam. was found to belong to C.hirsutus L. rather than to the C.ratisbonensis group as currently treated. Cytisuslasiosemius Boiss. is not the correct name for C.frivaldszkyanus Degen, but another synonym of C.hirsutus. Cytisuslitwinowii V.I.Krecz., which was known solely from the holotype, is a synonym of C.austriacus L. s.str. Chamaecytisuspseudojankae Pifkó & Barina, reported from a small area shared between Albania, Greece and North Macedonia, is treated as a subalpine variant of C.austriacus. Cytisustmoleus Boiss. is removed from the synonymy of C.eriocarpus and added to the synonymy of C.pygmaeus Willd. Cytisusfalcatussubsp.albanicus Degen & Dörfl. and C.pubescens Gilib. are synonymised with C.hirsutus. Cytisusmicrophyllus Boiss. is moved from C.austriacus s.l. to the synonymy of C.frivaldszkyanus, and C.pindicola (Degen) Halácsy to the synonymy of C.jankae. Chamaecytisuscalcareus (Velen.) Kuzmanov is accepted as Cytisuscalcareus (Velen.) Sennikov & Val.N.Tikhom., comb. nov., and its distribution is circumscribed. Cytisushirsutusvar.ciliatus (Wahlenb.) Hazsl. and C.polytrichusvar.subglabratus Val.N.Tikhom. & Sennikov, var. nov. are recognised as glabrous variants of the corresponding species. Lectotypes of C.ciliatus, C.hirsutissimus K.Koch, C.jankae, C.lasiosemius, C.pubescens, C.rhodopeus J.Wagner ex Bornm. and C.thirkeanus K.Koch are designated. Cytisuspolytrichus is reported from the Western Caucasus in place of C.wulffii auct.

Cytisus sect.Tubocytisus DC. (= Chamaecytisus Link) is the largest part of Cytisus s.l.Its species number varies greatly according to the accepted concept, ranging from about 30 (Cristofolini and Troía 2006) to 43 (Govaerts et al. 2021).The species in this group may be very similar to each other, being different in minor characters of dimensions and pubescence (Cristofolini 1991;Cristofolini and Troía 2017).This fact poses a natural difficulty in the taxonomic delimitation of this group and is responsible for wide discrepancies and contradictions in taxonomic assessments between individual researchers (e.g.Gibbs (1970); Tzvelev (1987); Cristofolini (1991); Pifkó (2009)).
Published treatments of Cytisus sect.Tubocytisus varied in detail, but remained consistent in one major feature, i.e. a high level of taxonomic splitting, resulting in narrowly delimited taxa with faint, but constant differences in pubescence, dimensions, leaf shape and habit (Sennikov and Tikhomirov 2024a).Certain deviations observed between particular treatments may be better explained by some material being inaccessible to individual researchers, thus accounting for lumping of single species or misinterpretation of particular species names.
In the present contribution, we provide notes on some species of Cytisus, mostly in Central and Eastern Europe and the Balkans, which require taxonomic or nomenclatural corrections.This study is based on our examination of the original material and protologues of relevant species names, which allowed us to match otherwise discrepant taxonomic decisions made by various researchers (e.g.Gibbs (1970); Cristofolini (1991); Pifkó (2005Pifkó ( , 2009)); Pifkó and Barina (2016)).
This revision contributes taxonomic and nomenclatural corrections to the mapping programme for "Atlas Florae Europaeae".

Material and methods
This study is based on herbarium specimens, examined by traditional morphological method.The diagnostic characters used in this study are the same as in Cristofolini (1991) and Sennikov and Tikhomirov (2024a).
The synonymy is based on our examination of original material available through online resources (JSTOR, JACQ) and protologues.Type designations follow the latest rules of botanical nomenclature (Turland et al. 2018).New typifications are illustrated by scanned images of herbarium specimens.
Species descriptions are omitted.Instead, diagnostic characters are discussed and comparison tables are provided for species groups.
Country-level species distributions are compiled from reliable literature and accessible herbarium specimens (B, BR, H, JE, K, L, LE, LY, MA, MW, PRC, RB, U, W, WU), which were examined largely online as scanned images via JSTOR (https://www.jstor.org)and JACQ Virtual Herbaria (https://www.jacq.org).We also used human observations documented by photographs, which were available online via iNaturalist (https://www.inaturalist.org/).The distributions in the Balkans may be incomplete due to insufficient level of local studies and limited availability of herbarium material.Some species with critically revised circumscriptions are mapped.The list of specimens or observations examined and used in mapping is made available through Internet Archive (Tikhomirov and Sennikov 2023).

Cytisus hirsutus group
Taxonomy.The diagnostic character of this species group is long patent (horizontally spreading) stiff hairs on calyces and pedicels.This group requires a thorough revision on the account of its high morphological variability.In our notes, we concentrate on selected species whose type material is known to us.Type.Italy.Sassari: Olbia ("Prope Olbyam in Galloprovincia"), Herb.Burser XXII: 5 (lectotype UPS, designated by Cristofolini and Jarvis (1991: 498)).
Taxonomy.This species has dimorphic inflorescences (Cristofolini 1991) and leaves densely hairy above.Cristofolini (1991) included various glabrescent forms into this species, which we prefer to exclude because such forms are not parts of the infraspecific variability in the material that we have examined.
Distribution.Europe: mountain areas from western France to the Eastern Carpathians longitudinally, from southern Poland to southern Italy latitudinally (Cristofolini 1991;Cristofolini and Troía 2017).
Notes on nomenclature.In the protologue of Cytisus supinus, Linnaeus (1753) cited three synonyms borrowed from Clusius (1601), of which one synonym ("Cytisus VII.species altera Clus.hist. 1. p. 96") was cited twice.This erratic way of citation evokes the idea of corrupted references.We checked these double-cited references against the relevant synonyms in Bauhin (1671), which were linked with Clusius (1601) by Linnaeus (1753) and in the earlier treatments of Clusius (1583).The first instance of this reference, cited by Linnaeus (1753), belongs to Cytisi VII.species altera (Clusius 1601: 97), which is not accompanied by any illustration.The second citation actually refers to Cytisus VII (Clusius 1601: 96) with an illustration, which was designated by Cristofolini and Jarvis (1991: 498) as a lectotype of C. supinus.Although Cristofolini and Jarvis (1991) cited Cytisus VII.species altera as the lectotype, they unambiguously referred to the same illustration as Linnaeus, thus making the same technical citation error.We provide a correct citation here.
The protologue of Cytisus triflorus was based on the only cited specimen collected by Martin Vahl in Naples in 1785 (collection date from Lanzoni (1930)).This specimen was designated as a lectotype by Skalická (1986), but is most likely the holotype.
The species name Cytisus triflorus was misfortunately resurrected from oblivion by Skalická (1986) and accepted by Cristofolini (1991) for a segregate of C. ratisbonensis s.l., which is superficially similar to and often confused with C. hirsutus.Skalická (1986) examined the type specimen of this species name on the basis of a photograph which apparently did not show its features of pubescence.We requested a high-quality scanned image of the type from P-Lam; its examination revealed that the calyces, pedicels and petioles of this plant are covered by long upright setose hairs, which do not cover the plant tissues.These hairs are clearly distinct from the subappressed pubescence of dense thin hairs in the C. ratisbonensis group, which completely covers the plant parts, and correspond to the characters of C. hirsutus.Since the usage of this plant name after Skalická (1986) is relatively new and unstable (e.g. in Eastern Europe, the name C. lindemannii is still used for this species: Czerepanov (1995), Fedoronchuk (2019)) and the taxonomy of the C. ratisbonensis group has been in flux, the disappearance of this species name will not be of principal inconvenience for the users of plant nomenclature.
Cytisus pubescens Gilib.was originally introduced in Gilibert (1782), which is included in the list of suppressed works, thus disavowing valid publication of all new names of species and infraspecific taxa published in this book.This species name was validly published in a revised version of the same book (Gilibert 1793) which was reprinted from its original, also suppressed edition (Gilibert 1785).Since the reprint was not explicitly suppressed, its species plant names are considered validly published and may compete for priority (e.g.Ardenghi (2015)).
There are no extant herbarium specimens associated with the protologue of C. pubescens (Shiyan et al. 2013).The only element of its original material in existence is an illustration cited in the protologue, Cytisus VII (Clusius 1601: 96).This illustration is drawn from plants occurring in Spain ("praesertim Baetica"; this Roman Province largely corresponds to Andalucia) and represents C. hirsu tus (Cristofolini and Jarvis 1991).Although Gilibert (1793) clearly described a plant of the C. ratisbonensis group under his C. pubescens, the illustration cited in the protologue mandates the reduction of this species name to a synonym of C. hirsutus, which is formally effected here by lectotypification.
Cytisus falcatus was described as a relative of C. hitsutus (Waldstein & Kitaibel, 1812).Its pods are hairy and leaflets are sparsely hairy above, thus indicating the synonymy with C. hirsutus rather than C. ciliatus as treated by Micevski (2001) and Pifkó (2005).Cristofolini (1991) erroneously added C. falcatus to the synonymy of C. triflorus (which was a member of the C. ratisbonensis group in his sense).
The main collection of K.Koch was acquired to B in 1913 (Ulbrich 1917) and subsequently destroyed with few exceptions (Lack 1978).The specimens of Cytisus described by Koch survived at LE only (Edmondson and Lack 1977), and this material is designated as a lectotype of C. hirsutissimus here.Thirke labelled his collections with very generic designations.but Koch (1846) recorded that Thirke's collecting activities took place around Trabzon and, to a lesser extent, Samsun in 1843.
We traced two specimens from the original collection of C. hirsutissimus at LE.As the protologue states that calyces of this species are covered by horizontally spreading hairs (Koch 1846), thus corresponding to the diagnostic characters of C. hirsutus, we designate a specimen (LE 00013762) whose characters are in complete agreement with the protologue.Some authors (Kreczetowicz 1940;Grossheim 1952;Portenier and Solodko 2002) treated C. hirsutissimus as endemic to the Caucasus, which reportedly differed from the East European C. lindemannii (= C. elongatus) in longer pedicels and a patent (vs.subappressed) pubescence of the whole plant.These minor and variable characters cannot be considered species-specific, and C. hirsutissi mus of these authors was correctly identified with C. triflorus (Cristofolini 1991).Gibbs (1970) placed C. hirsutissimus in the synonymy of C. hirsutus on account of its lateral inflorescences (his treatment maintained the difference between C. hirsutus and C. supinus, thus artificially dividing a single species with dimorphic inflorescences, whereas C. triflorus is a species with monomorphic lateral inflorescences).Our designated lectotype confirms the latter synonymisation.
Cytisus lasiosemius Boiss.was described from Asiatic Turkey ("inter Samsun et Tekekoi [Tekkeköy]", now Bayraktepe National Park, Samsun Province).In the protologue, Boissier (Tchihatcheff 1860) compared the new species with C. su pinus (= C. hirsutus), and distinguished it from the latter by acute leaflets and hairy standard.These characters are variable within C. hirsutus, and Gibbs (1970) rightly placed C. lasiosemius to the synonymy of his C. supinus.On the contrary, Cristofolini (1991) accepted C. lasiosemius as a priority name for C. frivaldsz kyanus Degen, which also has rather patent hairs.This treatment cannot be accepted because the pubescence of C. lasiosemius is composed of long, sparsely situated horizontal hairs on its stems, petioles and pedicels, typical of C. hirsutus, whereas the pubescence of C. frivaldszkyanus is very densely covering the stems, petioles and pedicels and consists of both long and short curved hairs, like in the C. ratisbonensis group (Sennikov and Tikhomirov 2024a).We confirm the opinion of Gibbs (1970) and add C. lasiosemius to the synonymy of C. hirsutus.
The original material of C. lasiosemius consists of a few specimens collected by P.A. Tchihatcheff in Turkey during 1858 (Tchihatcheff 1860).These specimens are accompanied by tiny field tickets with different field numbers, thus indicating that they are different gatherings.Niketić (2021) designated a complete herbarium sheet at P with three gatherings as a lectotype, which is inadmissible.We restrict this choice to a single gathering numbered 629.(Micevski 2001).
Notes on taxonomy and distribution.This taxon was described from the vicinities of Liptovský Hrádok in present-day Slovakia (Wahlenberg 1814) and occurs in the mountains surrounding the Pannonian Plain and in the Balkans (Holub and Bertová 1988;Pifkó 2009 and our data).Cytisus ciliatus is closely related to C. hirsutus, but differs from the latter by the upper side of its leaf laminae and by pod surfaces being glabrous or nearly so (vs.regularly hairy).So far, we have no evidence that the distribution of hairy and glabrous plants of C. hirsutus is separate; this distinction denotes the same casual loss of pubescence as observed in some other species of Cytisus (C.ruthenicus var.Some authors (Bernard 1977) interpreted the name Cytisus glaber as corresponding to C. hirsutus, which cannot be true because of its leaves glabrous above.Judging from the glabrous leaves of the plant and its occurrence in "Austria", C. glaber is an earlier (albeit illegitimate and therefore unusable) synonym of C. ciliatus Wahlenb.(C.hirsutus s.l.).Tzvelev (1987) formally accepted Chamaecytisus glaber (with C. elongatus mis-added to its synonymy) and applied it to west Ukrainian and cultivated plants of Central European origin with erect stems, leaves glabrous above, lateral inflorescences and patent pubescence, which agrees with our interpretation.
Cytisus serotinus is a plant with the leaves glabrous above, which belongs to the C. hirsutus group.It was originally recognised due to its presumed late flowering season, but merely coincides with C. ciliatus.
Notes on nomenclature.Wahlenberg (1814) distinguished Cytisus ciliatus from C. hirsutus, which was the original name for his material, by the pubescence of its leaves and pods.In the collections of UPS, where the Herbarium of Wahlenberg is housed, two specimens of the original material were found, both corresponding to the original description and the provenance cited in the protologue.One specimen bears precise collection data, but the draft name of the taxon (C.hirsutus [...] glabris) written by Wahlenberg, whereas the second specimen bears the final plant name (C. ciliatus), but generalised collection data ("e montibus Carpaticis") written by C.P. Thunberg.As both specimens correspond to the taxon as circumscribed by Wahlenberg and are undoubtedly linked with the protologue, we prefer the specimen with exact provenance from the author's collection as a lectotype.
Despite all searches, we were not able to trace any herbarium material linked with the protologue of C. glaber (Linnaeus filius 1782), in which a species with the leaves glabrous above and slightly hairy below was described from "Austria".The only original element, an illustration of "Cytisus glaber, siliqua angusta" in Bauhin and Cherler (1650: 373) was rejected by Cristofolini (1991) as conflicting with the original description (calyces depicted as campanulate, whereas the protologue stated the calyx being "oblongus subventricosus"), although this presumed conflict may be explained by the crude nature of this drawing.So far, this species name remains untypified and interpreted on the basis of the protologue (Tzvelev 1987).
A later synonym belonging to the same taxon is C. serotinus Kit.ex DC. (Candolle 1825), described from historical "Hungary" without a further specification.Pifkó (2005) designated a lectotype at BP; since no specimens were cited by Candolle as syntypes, his only specimen used for the original description is the holotype, and the lectotype at BP has no standing.The only original specimen in Candolle's herbarium at G is lacking a precise provenance, which can be derived from comparisons with the main collections of P. Kitaibel kept at BP (Jávorka 1957) and from the diaries of Kitaibel (Gombocz 1945;Lőkös 2001).
Three specimens identified as C. serotinus are preserved in the herbarium of Kitaibel at BP (Pifkó 2005), collected near Mukachevo in present-day Ukraine and at Gödöllő in present-day Hungary.Kitaibel (Lőkös 2001) also mentioned that he collected this species near Szatmár (now Satu Mare in Romania, near the border with Hungary and Ukraine).The specimen at G-DC is dated as received in 1815 and seemingly was collected during that year on the way from Mukachevo to Satu Mare (Lőkös 2001).Notes on taxonomy and distribution.Cytisus polytrichus sharply differs from C. hirsutus in its creeping stems, small leaves and constantly axillar flowers (Cristofolini 1991).

Cytisus polytrichus
Plants of this species have been known from the Western Caucasus under a wrong name, C. wulffii auct.(Kreczetowicz 1940;Grossheim 1952).The latter species is endemic to the Crimea and differs from C. polytrichus in appressed (vs.strictly patent) hairs on its leaves and calyces (Sennikov and Tikhomirov 2024a).
Notes on nomenclature.Krytzka et al. (1999) designated the only suitable specimen at LE as lectotype, following the unpublished annotation by N.N.

Table 1
Taxonomy.The diagnostic characters of this species group are erect stems, dense capitate inflorescences and long thin silky hairs on calyces and pedicels.The knowledge on this group is highly incomplete, especially regarding the variability of Cytisus austriacus L. s.l.Type.Historical Hungary ("Ungaria").Herb.Burser XXII: 3, left-hand specimen (lectotype UPS, designated by Cristofolini in Turland and Jarvis (1997: 468)).Distribution.Europe: mountainous regions from Austria to western Ukraine and from southern Poland to Greece and European Turkey, with the presence in southern East European uplands; Asia: Turkey, Russian Caucasus (Gibbs 1970;Tzvelev 1987;Cristofolini 1991).

Cytisus austriacus
Notes on taxonomy.This species is highly variable in respect of the pubescence on its leaves and calyces and is currently recognised in a broad sense, with some infraspecific taxa (Cristofolini 1991).Our current treatment is focused on the typical plants, corresponding to C. austriacus s. str.
A short-leaved variant of the species was separated as C. austriacus subsp.microphyllus "(Boiss.)Boiss." by Cristofolini (1991), probably because of Bal dacci 315 (BM 000750880) which was the basis for the treatment of C. austri acus var.microphyllus in Baldacci (1899).This collection from Mt. Smolikas in north-western Greece consists of subalpine plants of C. austriacus s. str.which  have regrown after damage and developed smaller leaves, otherwise being in agreement with the type.Notes on nomenclature.Cytisus litwinowii V.I.Krecz.was described as a local endemic of the Central Russian Upland, confined to calcareous substrates (Kreczetowicz 1940).This plant was originally distinguished because of its lesser developed pubescence and golden-yellow flowers, which are smaller than in C. blockii V.I.Krecz.(= C. kerneri Błocki).Another reason to distinguish this plant as a separate taxon was its confinement to the area of presumably relic pine forests and shrublands of the steppe area of Central European Russia, which reportedly harboured endemic taxa of Tertiary age (Kozo-Polansky 1931).However, this area of endemism has been confuted by other researchers, who considered its age being early postglacial and its relics being taxonomically indistinct (Grosset 1964).Among the presumed endemics of this territory, Daphne julia K.-Pol.turned out to be a synonym of D. cneorum L. (Grosset 1964) and Tanacetum alaunicum K.-Pol.was synonymised with Chrysanthemum zawadskii Herbich (Tzvelev 1994), whereas Cotoneaster alaunicus Golitsin appeared to be a synonym of C. integerrimus Medik.(Sennikov 2011).
Further authors (Heywood and Frodin 1968;Tzvelev 1987) accepted C. lit winowii and distinguished it from C. austriacus, which also occurs in Central European Russia, by its leaflets glabrous or very poorly (sparsely) pubescent above (vs.densely appressed-hairy in C. austriacus).Following these authorities, C. litwinowii was accepted in major compilations (Yakovlev et al. 1996;Govaerts et al. 2021).
We examined the holotype of C. litwinowii at LE and realised that the leaflets of this plant, which had grown in the shade, are regularly pubescent above, but the hairs are poorly recognisable due to overpressing.As pubescence of leaflets was the main diagnostic characters for C. litwinowii and no other material of the taxon is known, but the holotype, we reduce it to a synonym of C. austriacus.The placement of C. litwinowii in the synonymy of C. blockianus Pawł.(Cristofolini 1991), which was accepted by some databases (Roskov et al. 2006), cannot stand because the latter species does not occur east of the Carpathians (Tzvelev 1987).Besides, the bright flower colour of C. litwinowii agrees particularly with the characters of C. austriacus, rather than the pale flower colour of C. blockianus (Tzvelev 1987).Pifkó and Barina (2016) described C. pseudojankae Pifkó & Barina as a strongly branching plant with undeveloped axillar shoots, small, narrowly lanceolate leaflets and laxly appressed pubescence, which they compared with the C. austriacus aggr., but placed in the C. eriocarpus aggr.Such plants were considered endemic to a restricted area near Lake Prespa at the borders of Albania, North Macedonia and Greece (Pifkó and Barina 2016;Bergmeier et al. 2020).According to the description and drawing of C. pseudojankae in Pifkó and Barina (2016), this taxon is very similar to C. austriacus in its strong and upright stems (vs.weak and ascending stems in C. eriocarpus s.l.), habit and narrowly lanceolate leaf shape.
The original material of C. pseudojankae (Pifkó and Barina 2016) consists of plants superficially looking like having lateral flowers; however, these plants are typical members of the C. austriacus group with capitate inflorescences, and the seemingly lateral flowers observed in C. pseudojankae are a result of its abundant branching, with the uppermost branches, much abbreviated, going to flower and thereby forming a pseudolateral inflorescence.Their leaves are similar to those of the plants treated as C. austriacus subsp.microphyllus by Cristofolini (1991).(Diklić 1972;Kuzmanov 1976;Micevski 2001;Assyov and Petrova 2012;Barina et al. 2018;Niketić 2021).Fig. 7.

Cytisus jankae
Notes on taxonomy.Cristofolini (1991) placed C. jankae next to C. austriacus, thus indicating their affinity.Both species share capitate inflorescences, lanceolate leaves and rather appressed pubescence on all green parts, but C. jankae differs from C. austriacus s.str.by its constantly small size and prostrate habit.Its recent subordination to C. heuffelii (Niketić 2021), which differs in its calyx being 7-8 mm long (vs.10-13 mm long in C. jankae), is hardly justified.
According to their original material, C. pindicola belongs to the synonymy of C. jankae as typified here.The synonymisation of C. pindicola with C. frivaldsz kyanus proposed by Barina et al. (2018) is not supported by their diagnostic characters (Table 1).
Notes on nomenclature.The original material of Cytisus jankae Velen., mounted as a single specimen (PRC 451243), is highly heterogeneous and consists of six fragments of small plants with stems ascending from woody caudices, with capitate inflorescences and narrow leaves, which are referable to three species.In spite of its apparent heterogeneity, this entire specimen has been recently designated as a lectotype of the species name (Niketić 2021).
Two linear-leaved fragments (top centre, bottom left) on this specimen belong to C. absinthioides Janka, which is another species of the Balkans.This species is sometimes (Cristofolini 1991;Govaerts et al. 2021) merged with C. eriocarpus Boiss.(syn.C. smyrnaeus Boiss.), which is characterised by its leaflets being broadly obovate to elliptic rather than narrowly lanceolate and is totally different in its habit and long spreading pubescence.Cytisus absin thioides is characterised by typically upright, strongly branched stems, regular presence of abbreviated sterile shoots in the leaf axils, small flowers (with calyces 7-8 mm long), rather short subpatent pubescence on the stems and dense appressed pubescence of silvery appearance on the leaflets.
Two plants on the left and right sides are characterised by decumbent to ascending stems, narrowly lanceolate or oblanceolate leaflets and subpatent  pubescence on stems and calyces, with less developed sterile shoots in leaf axils.These plants correspond to C. pygmaeus Willd., occurring in the Balkans and Turkey.
The plant mounted above the label is similar to C. pygmaeus, but differs from the latter in a densely appressed pubescence, the feature corresponding to the original description of C. jankae which reads "foliolis linearibus vel lineari-spathulatis ... calycis adpresse sericei ..." (Velenovský 1890).The small fragment alongside the label probably belongs to the same species.As this plant is in good agreement with the protologue, we designate it as a lectotype of C. jankae.
Other low-growing and small-leaved variants presumably belonging to the same group are C. pseudopygmeus Davidov and C. georgievii Davidov, described from the Pontic part of Bulgaria (Davidoff 1902) and synonymised with C. jankae by Kuzmanov (1976).We refrain from any assessment of these species names because we were not able to examine any original material.
Cytisus pindicola (Degen) Halácsy agrees with the type of C. jankae, but slightly differs from the latter in slightly shorter hairs on stems (0.7-1 mm long vs. 1-2 mm long in C. jankae) and leaves (0.5-0.8 mm long vs. 0.8-1.5 mm long in C. jankae) and in shorter calyces (8-10 mm long vs. 10-13 mm long in C. jankae).Cytisus pindicola was previously placed in a subspecies of C. austri acus (Cristofolini 1991, as C. austriacus subsp.microphyllus), but differs from the latter in shorter leaves and a different habit.
The original material of Cytisus austriacus var.pindicola Degen (Baldacci 1899) consists of four gatherings which were distributed under a single number, as Baldacci 110.K.I.Christensen intended to designate a lectotype at W, but the only specimen in that collection is a mixture of four indistinguishable gatherings (Reich et al. 2021).Lectotypification is advisable with Degen's material at BP. Distribution.Europe: Balkan Peninsula (Bulgaria, Greece, North Macedonia, Serbia) (Kuzmanov 1976;Assyov and Petrova 2012).The occurrences outside Bulgaria are confirmed or reported here (Fig. 8).Pifkó and Barina (2016) removed the report of Chamaecytisus calcareus from Albania in favour of their C. pseudojankae, which we synonymise with C. austriacus.
Notes on taxonomy.This miniature plant belongs to the C. austriacus group because of its terminal inflorescences, which are rather dense and surrounded by floral leaves.It differs from C. austriacus by its short habit, much smaller and shorter, subelliptic (vs.lanceolate) leaves, and from C. jankae by the same shape of leaves (although of similar size) and by subpatent (vs.appressed) pubescence of calyces.This species was omitted by Cristofolini (1991) and is currently recognised only in Bulgaria (Kuzmanov 1976;Assyov and Petrova 2012).
Notes on taxonomy.Cytisus absinthioides strikingly differs from any other species of the C. austriacus group by its habit, resembling some plants of Arte misia due to its tall branched stems with regularly developed sterile branches in leaf axils and dense appressed sericeous pubescence on its leaves and calyces.Its calyces and pods are distinctly small (Janka 1872).

Cytisus frivaldszkyanus
Notes on taxonomy.This species with subpatent to patent pubescence was accepted by Cristofolini (1991), but under a wrong name, C. lasiosemius, probably because of the unavailability of the type collection of the latter species name.
Notes on nomenclature.Degen (1893) described Cytisus frivaldszkyanus from a few localities in present-day Bulgaria, citing four syntype gatherings.The examined material is fairly homogeneous, and the application of the species name is unambiguous.So far, we refrain from lectotypification because the main collection of Degen at BP has not been examined by us.
Cytisus rhodopeus was first mentioned in the synonymy of C. eriocarpus by Degen (1893) and validly published by Bornmüller (1925) without any descriptive matter, but with a reference to the description of C. absinthioides in Velenovský (1891).Five syntypes from Bulgaria were cited in the original description (Velenovský 1891), which deviated much from the description of the true C. absinthioides provided by Janka (1872) by a longer calyx (13-15 mm long vs. 7-8 mm long in C. absinthioides) with patent (vs.appressed) hairs.Through the kindness of P. Mráz, we traced a specimen in the collection of J. Velenovský at PRC, which exactly corresponds to the protologue by its diagnostic characters and taxonomic references on its label (to C. absinthioides Janka and "C.eriocarpus Boiss.var.", as Velenovský (1891) also noted a relationship with the latter species).This specimen fully reflects the taxonomic concept of Velenovský (1891) and is designated as a lectotype of C. rhodopeus here.Cristofolini (1991) accepted C. austriacus subsp.microphyllus "(Boiss.)Boiss." as the correct name for a small-leaved segregate of C. austriacus, citing C. pindicola (Degen) Halácsy in its synonymy.The type collection of C. micro phyllus Boiss. is quite dissimilar from C. pindicola and belongs to C. frivaldsz kyanus because of its strong suberect stems, partly obovate (vs.lanceolate) leaflets and pods with nearly patent (vs.appressed) hairs.

Cytisus pygmaeus group
Table 2 Taxonomy.The diagnostic characters of this species group are mostly prostrate habit and pseudolateral inflorescences.This group is very poorly known and may be an artificial assemblage of superficially similar species.Their distributions need to be verified due to common confusions and misidentifications.Distribution.European and Asiatic Turkey, Bulgaria, Greece (Kuzmanov 1976;Cristofolini 1991;Assyov and Petrova 2012), Romania (Fig. 13).Other European records, from North Macedonia and Serbia (Diklić 1972;Micevski 2001), seem to belong mostly to C. jankae or C. calcareus.A record of C. jankae from Romania (Grinţescu 1957) is treated as belonging to C. pygmaeus here.
Notes on taxonomy.The leaves of this species may vary slightly from oblong-lanceolate to oblanceolate.Plants with the leaves looking more lanceolate were described as C. pygmaeus and C. chrysotrichus, whereas plants with rather oblanceolate leaves were named C. tmoleus and C. thirkeanus.This difference, albeit very subtle, led Cristofolini (1991) to classify C. pygmaeus as a subspecies of C. austriacus, whereas he placed the plants described as C. tmoleus to C. eriocar pus.Having examined some material from Asiatic Turkey, we observed both types of leaves in the same plants; this makes the distinction practically impossible.
The pubescence on calyces of C. pygmaeus is variable, ranging from semi-patent to subappressed.The type collection of C. pygmaeus has clearly semi-patent hairs.
Niketić (2021) provisionally accepted the occurrence of C. pygmaeus in Serbia, although the relevant materials have not been examined.Micevski (2001) listed it among doubtful records in North Macedonia.The collections identified as C. pygmaeus which we examined from the Balkans belong to C. jankae, and we assume that the distribution of C. pygmaeus in Europe may be much more limited than it is currently believed.
Notes on nomenclature.Willdenow (1802) described the species without mentioning floral characters.His indication of "Galatia" in the protologue corresponds to the fruiting specimen of D. Sestini in Willdenow's personal collection.A duplicate of this collection was separated to HAL, which allowed Pifkó and Barina (2016) to designate a lectotype at B.
Notes on taxonomy.This species is very similar to C. frivaldszkyanus due to its abundant patent pubescence.However, it differs from the latter in its broadly elliptic to obovate, nearly rotund leaflets, which are apically subrotund (vs.elliptic-lanceolate to obovate, broadly acute in C. frivaldszkyanus).Cytisus eriocarpus is similar to C. hirsutus, from which it differs in its pubescence (abundant short hairs mixed with long patent hairs vs.only long patent hairs in C. hirsutus) and smaller subrotund leaflets, as already noted in the protologue (Boissier 1843).
Notes on nomenclature.Gibbs (1970) inadvertently designated a specimen at K as the lectotype of C. eriocarpus.

Table 1 .
Diagnostic characters in the Cytisus austriacus group.