Description from Flora of China
Subshrubs to perennial or annual herbs. Stems often weak and clambering, often notably prickly or "sticky" (i.e., retrorsely aculeolate, "velcro-like"). Raphides present. Leaves opposite, mostly with leaflike stipules in whorls of 4, 6, or more, usually sessile or occasionally petiolate, without domatia, abaxial epidermis sometimes punctate- to striate-glandular, mostly with 1 main nerve, occasionally triplinerved or palmately veined; stipules interpetiolar and usually leaflike, sometimes reduced. Inflorescences mostly terminal and axillary (sometimes only axillary), thyrsoid to paniculiform or subcapitate, cymes several to many flowered or infrequently reduced to 1 flower, pedunculate to sessile, bracteate or bracts reduced especially on higher order axes [or bracts sometimes leaflike and involucral], bracteoles at pedicels lacking. Flowers mostly bisexual and monomorphic, hermaphroditic, sometimes unisexual, andromonoecious, occasionally polygamo-dioecious or dioecious, pedicellate to sessile, usually quite small. Calyx with limb nearly always reduced to absent; hypanthium portion fused with ovary. Corolla white, yellow, yellow-green, green, more rarely pink, red, dark red, or purple, rotate to occasionally campanulate or broadly funnelform; tube sometimes so reduced as to give appearance of free petals, glabrous inside; lobes (3 or)4(or occasionally 5), valvate in bud. Stamens (3 or)4(or occasionally 5), inserted on corolla tube near base, exserted; filaments developed to ± reduced; anthers dorsifixed. Inferior ovary 2-celled, ± didymous, ovoid, ellipsoid, or globose, smooth, papillose, tuberculate, or with hooked or rarely straight trichomes, 1 erect and axile ovule in each cell; stigmas 2-lobed, exserted. Fruit on pedicels sometimes elongating during development, green, gray, or infrequently white (to red, orange, or black), mostly dry to leathery schizocarps, infrequently spongy, rarely ± fleshy and berrylike, ellipsoid to subglobose; schizocarps separating into 2 indehiscent mericarps, each with 1 seed, subglobose, ellipsoid-oblong, or reniform, smooth and glabrous to tuberculate and/or covered with trichomes often hooked and clinging; seeds small, grooved ventrally (i.e., adaxially); testa membranous; endosperm corneous; embryo curved; cotyledons leaflike; radicle terete, inferior.
Galium is by far the largest and most widespread genus within the tribe Rubieae (subfamily Rubioideae). According to the most recent contributions (Natali et al., Opera Bot. Belg. 7: 193-203. 1996; Ehrendorfer et al., Fl. Iranica 176: 1-287. 2005; Bremer & Eriksson, Int. J. Pl. Sci. 170: 766-793. 2009; Soza & Olmstead, Taxon 59: 755-771. 2010), this tribe is closest to Theligoneae, Putorieae, and Paederieae, and includes the following genera treated (or mentioned) in the present flora: Asperula, Cruciata Miller, Galium, Kelloggia, Leptunis, Microphysa, Phuopsis, Rubia, and Sherardia Linnaeus.
So far, the genera Cruciata and Sherardia have not been found in China yet but may be expected there because of their partly weedy character and widely adventive occurrence. They are included in the key below for future reference but not among the full generic presentations. Sherardia arvensis Linnaeus is widely distributed in warm temperate and high-elevation tropical regions and can be separated from Asperula, Phuopsis, Leptunis, or Galium by its terminal capitate inflorescences enclosed by leaflike bracts, its clearly developed calyx with 6 acute lobes, and its pink or violet corollas with well-developed funnelform tubes and 4 lobes.
Among the few Cruciata species, the W Eurasiatic C. pedemontana (Bellardi) Ehrendorfer appears occasionally as an adventive in warm temperate regions. It is common, e.g., in SE North America, and could be found in China too. Cruciata can be separated from Galium by its flowering stems with vegetative apices and the inflorescences consisting only of lateral axillary cymes on middle and lower stem nodes. These cymes are equal to or shorter than the subtending leaves when fully developed. In contrast, the inflorescences are mostly terminal and axillary and longer than the leaves in Galium.
The characters relevant for the taxonomy of Galium and other Rubieae deserve some comments. Life and growth forms are important, particularly with respect to the differentiation into half-shrubs, herbaceous perennials, and annuals. Stem and leaf posture, consistency, shape, and indumentum (e.g., pubescent or retrorsely aculeolate with recurved microhairs) are often quite diverse and may vary within species or even populations. The true leaves are always opposite and 2, but interpetiolar stipules may vary from inconspicuous and divided or simple to enlarged and leaflike, forming whorls of 4 or up to 6 and more. During seedling and shoot development all these taxa pass through the 2- and 4-whorl stage, but some taxa remain at this stage, while others continue to develop more numerous whorl elements toward the middle of their stems. This is a most informative differential character within Rubieae. Other relevant features relate to leaf shape, venation, texture, and particularly indumentum. Here, the presence of longer or shorter microhairs (use a lens) on surfaces and particularly margins as well as their forward or backward direction is of taxonomic importance.
Other morphological characters decisive for Rubieae taxonomy concern the inflorescences (e.g., the position and structure of the cymes). Flower shape is essential for the traditional separation of the genera Asperula (with salverform, funnelform, or cup-shaped corollas) and Galium (with ± rotate corollas). It is now clear that there are transitions between these character states and that even closely related taxa may differ in this respect. So far, it has been possible to provisionally maintain Asperula and Galium by the transfer of obviously misplaced taxa and by using the presence or absence of bracts and bracteoles as a differential character for the two genera (see Ehrendorfer et al., loc. cit. 2005).
The indumentum of ovaries and fruit as well as fruit consistency also vary strongly within Rubieae. Informative are, for example, ± fleshy berries (as in Relbunium Bentham & J. D. Hooker, Rubia, and certain Galium taxa) vs. dry schizocarps or the presence vs. absence of hairiness and whether the trichomes are hooked (i.e., the fruit disperse as "stick-tights" on animals) vs. straight. However, the distinction between all these structures is arbitrary, and there are even transitions between trichomes and tuberculate protuberances of various shapes as well as between hairy and glabrous. All this is well illustrated by Yang and Li (Bull. Natl. Mus. Nat. Sci., Taichung 11: f. 1. 1998). Furthermore, ovary and fruit indumentum and surface structures may change during development and sometimes vary genetically within species or even within populations, as in several Galium species. In general, authors in other regions have documented infraspecific variation from glabrous to densely hairy or tuberculate fruit but traditionally have only separated plants with hooked trichomes into different species. However, intrepid Chinese authors have easily combined these latter morphotypes, e.g., in G. dahuricum sensu W. C. Chen (in FRPS 71(2): 255. 1999), whereas Fl. Japan (3a: 238-239. 1993) distinguished G. manshuricum on the basis of this character. Only careful studies and field observations can clarify such cases, as in G. spurium, where the infraspecific variation of fruit, either smooth, tuberculate, or covered with hooked hairs, has been proven.
Further differential characters for the taxonomy of Rubieae come from the fields of palynology (e.g., number of colpi), karyology (e.g., deviations from the normal chromosome base number x = 11 in Asperula sect. Cynanchicae (Candolle) Boissier with x = 10 or in Galium sect. Aparinoides (Jordan) Grenier with x = 12; common occurrence of polyploidy), and reproductive biology. Most of the perennial Rubieae taxa have conspicuous hermaphroditic or andromonoecious flowers and inflorescences and are insect-pollinated and self-incompatible outbreeders (e.g., Phuopsis or G. boreale and G. verum). Nevertheless, for several annuals with small and inconspicuous flower aggregates selfing and autogamy have been documented (e.g., G. aparine, G. spurium, and Sherardia arvensis). Furthermore, polygamodioecy and dioecy occur in some groups (e.g., G. elegans). Up to now, only few and insufficient data from all these fields are available for Asian Rubieae species and have not been mentioned in FRPS. Nevertheless, such data are significant and will have to be addressed in more detailed future systematic Rubieae studies from this region.
The α-taxonomy of Rubieae in E Asia is still in a problematic state. A general survey of the collections at the herbaria KUN, MO, PE, W, and WU has revealed the existence of many very polymorphic, complex, and insufficiently understood species groups. Therefore, the present treatment has to be regarded as provisional.
A particularly critical case concerns several Galium species described by H. Léveillé from 1904-1917 (see Lauener & Ferguson, Notes Roy. Bot. Gard. Edinburgh 32: 103-115. 1973). These descriptions are most fragmentary and the relevant types are not yet studied sufficiently (but see Mill, Edinburgh J. Bot. 53: 193-213. 1996). Relevant taxa in alphabetical order are G. blinii (see under that name), G. bodinieri (see under G. blinii and G. rebae), G. cavaleriei (see under G. asperifolium), G. comarii (see under G. dahuricum), G. esquirolii (see under G. asperifolium), G. hongnoense (see under G. spurium), G. mairei (see under G. elegans), G. martini (see under G. bungei), G. quinatum (see under G. blinii), G. remotiflorum (see under G. bungei), and G. venosum (see under G. bungei).
The treatment of Galium for the Flora of Taiwan by Yang and Li (Bull. Natl. Mus. Nat. Sci., Taichung 11: 101-117. 1998; Fl. Taiwan, ed. 2, 4: 254-259. 1998) is not satisfactory in several aspects: keys and descriptions are rather idealized and lack carefully observed ranges of morphological variation for the taxa; species are circumscribed more narrowly and based on different characters than used by other authors in the region (e.g., presence vs. absence of leaf indumentum is considered variable within species by most other authors); the treatment is not well reconciled with continental Galium taxonomy (e.g., there are no references to the Russian floras, and names synonymized by others are used without explanation); and at least two names based on types from Taiwan are missing.
With respect to a more "natural" and general taxonomic classification of the Rubieae and Galium, a number of recent morphological, karyological, palynological, and particularly DNA-analytical studies (e.g., Natali et al., loc. cit.; Robbrecht & Manen, Syst. & Geogr. Pl. 76: 85-146. 2006; Bremer & Eriksson, loc. cit.; Soza & Olmstead, loc. cit.) are available. They show that Theligonum should be placed into a separate tribe (Theligoneae), that the tribe Rubieae is monophyletic, and that Kelloggia (as subtribe Kelloggiinae, still with normal Rubiaceae stipules, calyx teeth, and 3-colpate pollen, but already with hooked trichomes on the dry mericarps) occupies a basal position in Rubieae. The Central American genus Didymaea J. D. Hooker (still with normal stipules but with the calyx already lacking, 5-coplate pollen, and seeds separating from the fleshy pericarp) represents a link to the genus Rubia in the true Rubiinae. Their stipules are nearly always leaflike, the pollen is polycolpate, and the seeds never separate from the pericarp. Rubia, a well-circumscribed and certainly monophyletic genus, is always perennial, has 5-lobed corollas, and berrylike fruit.
The remaining Rubiinae are also monophyletic as a whole, but their traditional genera Asperula, Bataprine Nieuwland, Callipeltis Steven, Crucianella Linnaeus, Cruciata, Galium, Leptunis, Mericarpaea Boissier, Microphysa, Phuopsis, Relbunium, Sherardia, Valantia Linnaeus, and Warburgina Eig are all essentially interdigitated. They are difficult to separate and can hardly be brought into concordance with available phylogenetic data. These advanced Rubiinae tend to develop more and more apomorphic character profiles, i.e., change from perennial to annual, increase in numbers of leaflike stipules from 4 to numerous, loss of bracts and prophylls in the inflorescences, reduction from 5-lobed to (3 or)4-lobed corollas, specialization of mericarps, etc. As shown by the most comprehensive phylogram available so far (Natali et al., loc. cit.: f. 2; Soza & Olmstead, loc. cit.: f. 1, 2) and new findings (unpubl.), these more apomorphic Rubiinae form a polytomy or a grade with seven parallel clades. The most basal clade (1) consists of the monotypic Galium sect. Cymogalia Pobedimova only. The following Sherardia clade (2) includes Crucianella, Phuopsis, Sherardia, and several sections of Asperula together with Leptunis. Separate clades are formed by G. sect. Depauperata Pobedimova (3), A. sect. Glabella Grisebach, including G. sect. Aparinoides (4), and A. sect. Asperula (5). The Cruciata clade (6) consists not only of the genera Cruciata and Valantia but also of all sections of Galium (including the traditional genera Bataprine, Microphysa, and Relbunium) that form whorls of 2 leaves and normally not more than 2 even-sized leaflike stipules. Finally, the G. sect. Galium clade (7) comprises this and various other sections of Galium, which regularly develop whorls of leaves and leaflike stipules with 5 to more elements.
From the above data and the fact that a number of major groups of Rubiinae have not been DNA-analyzed yet, it is obvious that it is still difficult and partly impossible to harmonize DNA-supported clades with the traditional genera and sections. Thus, extensive changes are expected for generic and sectional circumscriptions within Rubiinae in the future. Therefore, we refrain from taxonomic changes for the present flora, list taxa in alphabetical order, and only supplement phylogenetic comments. Thus, the present treatment in principle follows FRPS (71(2): 216-286. 1999), mainly based on Pobedimova et al. (Fl. URSS 23: 287-381. 1958), but also considers Ehrendorfer et al. (loc. cit. 2005). In order to make comparison with available phylogenetic data and present infrageneric classification easier, relevant information is inserted as a "Taxonomic Conspectus" before the individual species descriptions. It was not until this volume was ready for the press that the need for the nomen novum, Galium glabriusculum, was discovered; therefore, this species alone is outside of the alphabetical order.
Here the key to species of Galium is extensively revised from that of FRPS. It includes all of the Chinese Galium species with full ranges of differential character variation. Furthermore, it keys out all other Rubieae genera which are easily confused with Galium and are documented or can be expected in China. Details on the genera Asperula, Leptunis, Microphysa, Phuopsis and Rubia can be found where they are listed in alphabetical order, references to Cruciata and Sherardia appear in the comments above.
Several species are keyed out more than once in the present key because they are circumscribed by combinations of characters rather than by unique features. Furthermore, many Galium species are markedly variable because of genetic differentiation (e.g., G. bungei, G. elegans) but also because of phenetic plasticity due to different environmental conditions. References to the number of leaves and leaflike stipules in whorls as well as leaf measurements refer to middle stem regions. Measurements of organs with hairy surfaces (e.g., leaves, fruit, and mericarps) here apply to the solid surface of the structure and do not include the trichomes. The terms "leaf whorl," "ovary," and "uncinate trichome" follow common usage in Galium. Infraspecific taxa are adopted from FRPS in order to facilitate future and more detailed work on this group and comparison with other floras. They are not included in the following main key but are subordinated under the relevant species in alphabetical order and keyed out there.
Taxonomic conspectus of the Rubieae (excluding Kelloggia and Rubia)
In FRPS (71(2): 216-286. 1999) the taxa of Galium were placed in the following sections (designated here by capital letters): G. sect. Depauperata (A), G. sect. Aparine (B), G. sect. Pseudaparine Lange (C), G. sect. Cymogalia (D), G. sect. Trachygalium (E), G. sect. Leptogalium Lange (F), G. sect. Platygalium (G), G. sect. Galium (H), G. sect. Leiogalium (I), G. sect. Trichocarpa (Pobedimova) Pobedimova (J), G. sect. Asperuloides Pobedimova (K), and G. sect. Brachyantha (Boissier) Pobedimova (L). Some species of uncertain position were classified as dubious (M). Species accepted in the present treatment but lacking in FRPS are designated as (Z). For comparison, a current (e.g., Jelenevsky et al., Novosti Sist. Vyssh. Rast. 35: 174-187. 2003; Ehrendorfer et al., Fl. Iranica 176: 1-287. 2005), DNA-analytically supported (e.g., Natali et al., Opera Bot. Belg. 7: 193-203. 1996; Soza & Olmstead, Taxon 59: 755-771. 2010; Ehrendorfer, unpubl.) but still provisional taxonomic conspectus is presented below. It lists all species here accepted under their clades and sections. For the clades 1-6 one should compare the comments above, for the sections compare Ehrendorfer et al. (loc. cit.). The placement of Galium species into sections (or lack of placement) by FRPS is indicated by showing the relevant letters used above in parentheses after the species names.
Galium sect. Cymogalia Pobedimova s.s.
39. Galium paradoxum (D)
Galium sect. Depauperata Pobedimova
16. Galium exile (A; incl. G. songaricum sensu FRPS)
Galium sect. Aparinoides (Jordan) Grenier
27. Galium karakulense (E)
25. Galium innocuum (as G. trifidum: E; incl. "G. palustre")
Asperula sect. Asperula
Galium sect. Platygalium W. D. J. Koch s.l.
9. Galium bungei (E; incl. G. martini: M)
45. Galium salwinense (E)
12. Galium crassifolium (A)
31. Galium linearifolium (E)
21. Galium hirtiflorum (Z)
20. Galium glandulosum (A)
18. Galium forrestii (D)
44. Galium rupifragum (Z)
35. Galium morii (D)
54. Galium tarokoense (A)
34. Galium minutissimum (M)
36. Galium nankotaizanum (M; incl. G. maborasense)
47. Galium serpylloides (A)
29. Galium kinuta (G)
24. Galium hupehense (M)
30. Galium kunmingense (Z)
17. Galium formosense (as G. kwanzanense: M)
15. Galium elegans (D)
63. Galium yunnanense (M)
26. Galium kamtschaticum (D)
10. Galium chekiangense (as G. nakaii: G)
40. Galium platygalium (K)
32. Galium maximoviczii (K)
7. Galium boreale s.l. (G)
60. Galium turkestanicum (M)
Galium sect. Hylaea (Grisebach) Ehrendorfer s.l.
37. Galium odoratum (J)
4. Galium asperuloides (A)
22. Galium hoffmeisteri (as subsp. of G. asperuloides: A)
14. Galium echinocarpum (A)
53. Galium takasagomontanum (M)
59. Galium triflorum (A)
58. Galium trifloriforme (Z)
Galium sect. Trachygalium K. Schumann s.l.
48. Galium sichuanense (Z)
13. Galium dahuricum (as "G. davuricum": F; incl. G. comarii, G. niewerthii, G. pseudoasprellum)
56. Galium tokyoense (as var. of "G. davuricum": F)
41. Galium prattii (M)
52. Galium taiwanense (M)
3. Galium asperifolium (I)
6. Galium blinii (as syn. of G. asperifolium var.: I; incl. G. quinatum: M)
51. Galium sungpanense (A)
42. Galium pusillosetosum (A)
1. Galium acutum (M)
43. Galium rebae (Z)
33. Galium megacyttarion (Z)
5. Galium baldensiforme (A)
49. Galium glabriusculum (A)
28. Galium karataviense (as G. rivale s.l.: K)
61. Galium uliginosum (F)
Galium sect. Leiogalium Ledebour
38. Galium paniculatum (M; incl. G. xinjiangense: J)
Galium sect. Orientigalium Ehrendorfer
8. Galium bullatum (I)
Galium sect. Galium
23. Galium humifusum (L)
62. Galium verum (H)
11. Galium consanguineum (as G. majmechense: H)
46. Galium saurense (M)
Galium sect. Aparine W. D. J. Koch s.s.
50. Galium spurium (as G. aparine var. tenerum: B)
2. Galium aparine (B)
Galium sect. Kolgyda Dumortier s.s.
57. Galium tricornutum (as G. tricorne: B)
Galium sect. Microgalium Grisebach
19. Galium ghilanicum (Z)
55. Galium tenuissimum (C)
More than 600 species: worldwide, mostly in meridional to temperate but also in alpine and arctic regions or in subtropical and tropical zones at higher elevations; 63 species (23 endemic, four of unconfirmed occurrence) in China.
(Authors: Chen Tao (陈涛); Friedrich Ehrendorfer)