A reassessment of Pinus Subgen. Pinus in China

A recent study of Pinus subgen. Pinus for the Flora of China project recognized 10 native species with four new combinations, all at subspecies rank, and determined one new homonym, and six new synonyms. A neotype is designated for P. kesiya. Doubtful or little-known species are listed.

in the Flora of China. Additionally, P. taeda of N America and P. thunbergiana of Japan, which are naturalized in China, are also included. Descriptions, distribution and ecology of species, subspecies or varieties will be available in the English and updated version of the Flora. For practical reasons, new taxa published after June 1994 are not included for discussion.
Previously this species was recorded only in the monsoon belt of the Himalayas from Bhutan to Pakistan (Critchfield & Little, 1966;Farjon, 1984). It was found distributed in Gyirong in south Xizang (Tibet) and was classified as a rare species in China (Fu & Jin, 1992 Pinus sylvestris (Scots pine) is the most widely distributed pine species; it grows throughout northern Eurasia, from Scotland and Spain in the west, to NE China (Critchfield & Little, 1966). The morphology of the Scots pine is very variable; more than 150 variants have been described. There have been different opinions on the taxonomic status of its populations in NE China. Until recently, Cheng & Fu (1978) treated these as var. mongolica while Kitagawa (1979) accepted only var. sylvestris. Some botanists (Chang & Li, 1982;An & Zhang, 1986) recognized both var. mongolica and var. sylvestris in China. By matching herbarium material throughout the distribution range, it seems that the variation of this species in NE China also follows a clinal pattern, as it does in Europe and the Far East. To make the Flora compatible with other modern floras, such as Flora Europaea (Tutin, Burges, Chater et al., 1993) and the Russian checklist (Czerepanov, 1995), var. mongolica is not recognized here. Further study is needed to clarify the affinity of the Chinese populations of this species. Morphological study (Wang, 1988) and molecular analysis (Szmidt & Wang, 1993) indicate that var. sylvestriformis is closer to P. densiflora than to P. sylvestris. Pinus thunbergiana is naturally distributed in the coastal areas of Japan and southern Korea. It was introduced into China in the early 1900s and is now widely cultivated in Liaoning, Shandong and Zhejiang provinces as a reafforestation tree. It is also planted in Dalian, Hanzhou, Lushan, Nanjing, Wuhan, Shanghai and other major coastal cities as an ornamental tree.

Pinus thunbergiana
In the literature, Pinus thunbergii Parl. was commonly used, but it is a later homonym of P. thunbergii Lamb.

Pinus massoniana
Lamb., Descr. Gen. Pinus 1: 17 (1803). Type: [S Africa, Cape of Good Hope] a specimen 'brought by Mrt Francis Masson from the Cape of Good j Hope, where it was raised from seeds which had been sent from China' (whereabouts '• unknown). Lectotype: Lambert, Descr. Pinus 1: 1.12 (1803) (selected by Farjon, 1993). Pinus crassicorticea was based on two specimens from northern Guangxi, the western boundary of P. massoniana, with slightly thicker needles (0.7-0.8mm diam. in the former vs. 0.6-0.8mm in the latter) and thicker bark (up to 10cm). It was described that in P. crassicorticea the first-year shoots bear with two or three nodes. However, this can sometimes be found in the southern part of the range of P. massoniana in Guangdong, especially in dry areas (Cheng & Fu, 1978). It was also stated that there was an internal resin canal in the needles of P. crassicorticea. This is a somewhat variable character. However, a re-examination of the type showed that the resin canals are all marginal.  Pinus luchuensis, P. taiwanensis and P. hwangshanensis were described from the Ryukyu Islands, Taiwan and mainland China respectively. There has been no consensus on their inter-relationships for a considerable time. Wu (1956) accepted P. hwangshanensis as a variety of P. luchuensis, and P. taiwanensis as a synonym of the latter. However, Cheng et al. (1975) and Cheng & Fu (1978) treated P. hwangshanensis as a synonym of P. taiwanensis, and populations of P. hwangshanensis in Guangxi with both marginal and medial resin canals were recognized as a new variety of P. taiwanensis, var. damingshanensis. Among non-Chinese authors, Critchfield & Little (1966) accepted all three species, while Farjon (1984) did not mention P. hwangshanensis. Silba (1984) at first recognized only one species, P. luchuensis, but later (Silba, 1986) treated the other two taxa as varieties of the latter (even though his combination, P. luchuensis var. taiwanensis was not validated). It is true that the three taxa are similar to one another because of their seed cones, medial resin canals and relatively thicker needles. There are differences, though not very conspicuous, which make them distinct (Table 1). It is therefore better to treat them as one species with three subspecies.
There is an ambiguity in citing the authorship of Pinus hwangshanensis. Tsoong's paper is available in most western institutions, so that the compiler of Index Kewensis attributed the name to Tsoong. In fact, Hsia is the author of this name and she did publish a separate paper with exactly the same description but a different type citation. Both Tsoong and Hsia's papers appeared in 1936, but Hsia's was in the first issue of the Chinese Journal of Botany and presumed earlier than Tsoong's. Farjon (1993) reached the same conclusion but he ignored there was a specimen cited.  P. mukdensis was said to differ from P. tabuliformis in having dark grey bark and greyish brown or dark grey twigs. However, such characters are also found in various populations of P. tabuliformis. 7b. var. henryi (Mast.) C.T. Kuan, Fl. Sichuanica 2: 113 (1983 1485, 1486, 1487, 1488, 1489, 1490, 1494, 1495, 1497, 1498 (all  The status of Pinus henryi has been uncertain as it is an endemic taxon distributed in a remote area (Daba Shan) in central China. Wu (1956) suggested that its relationship was with P. massoniana. However, the needles are shorter and thicker in P. henryi than in P. massoniana. Furthermore, the apophyses of the seed scales are more prominent, and the bark is deeply furrowed into irregular squares in P. henryi. These characters suggest that this taxon is more closely related to P. tabuliformis than to P. massoniana. According to Kuan (1983), P. henryi and P. tabuliformis have naturally overlapping ranges in the Daba Shan area in Sichuan province. It is therefore more appropriate to treat it as a variety of the former.
P. massoniana var. wulingensis was described from Wuling Shan of the Daba Shan area. It differs from P. massoniana by having shorter needles (5-7cm long) and smaller seed cones .5 x 2-3cm), and from P. taiwanensis by its leaves with marginal resin canals. However, the original authors failed to compare it with P. henryi, with which it is conspecific. Mast, in J. Linn. Soc, Bot. 37: 416 (1906) Morphological study and molecular analysis suggested this is a hybrid of P. yunnanensis and P. tabuliformis (Wu, 1956;Wang & Szmidt, 1990). However, P. densata is still best treated as a species (for nomenclature see Note 1 of Art. H.3.4. of the ICBN).  The delimitation of P. yunnanensis (Yunnan pine) is sometimes questionable because of its close relationships with P. kesiya and P. insularis (Wu, 1956;Silba, 1984). However, P. yunnanensis may be distinguished in having thicker leaves, more or less drooping branches, and uninodal and shiny reddish brown first-year shoots. The chemical constituents are also different (Mirov, 1967;Farjon, 1984).

Pinus yunnanensis
Yunnan pine is basically restricted to the Yunnan plateau at elevations of 600-3000m. Two varieties were included in Flora Reipublicae Popularis Sinicae, var. tenuifolia in south-eastern Yunnan, western Guangxi and Guizhou, with thinner, longer and accordingly pendulous needles, and var. pygmaea, a shrub 0.4-2m tall with several trunks, distributed usually at higher elevations. The nomenclature and taxonomy of the Khasia pine have a confused history. Firstly, P. kesiya Royle ex Gordon was sometimes regarded as a nomen seminudum (Wu, 1956). However, Gordon's protologue presented a sufficient diagnosis: 'the cones resemble those of P. insignis, but they are not near so large, much flatter, and the scale not so prominent'. Therefore, most authorities (Critchfield & Little, 1966;Styles & Burley, 1972;Laubenfels, 1988;Farjon, 1993) accept its validity.

Pinus kesiya
Secondly, the Chinese populations, commonly called Simao pine in China, were named P. kesiya var. langbianensis (Cheng & Fu, 1978), typified by a specimen from central Vietnam. It was stated that the Simao pine and var. langbianensis differ from var. kesiya by having thinner bark, fissuring into irregular scaly plates. Field observations indicate that the bark of Simao pine may be as thick as 3cm. Further, examination of the type material of var. langbianensis and material from the Khasia mountains shows var. langbianensis should be part of the distribution of the Khasia pine.
Finally, with regard to its relationship with P. insularis Endl., the latter was usually merged with P. kesiya (e.g. Styles & Burley, 1972;Laubenfels, 1988), or for those who treated P. kesiya Gordon as a nomen nudum, P. insularis was used (Merrill, 1941;Wu, 1956;Silba, 1984). All herbarium specimens examined show that morphologically they are very similar in bearing longer and thinner needles in groups of three, fasciculate, multi-nodal first-year shoots and seed scales with prominent apophyses. It is therefore concluded that they are best considered conspecific. However, chemical differences may separate them (Farjon, 1984). In view of their geographical distribution, it seems more reasonable to accept two subspecies, subsp. insularis 1 from the Philippines and subsp. kesiya from China (southern Yunnan, south-eastern Tibet), Bangladesh, Bhutan, Burma, Cambodia, north-eastern India (Khasia), Laos, Nepal and Vietnam.
It is most likely that no herbarium specimen was preserved when Gordon published P. kesiya. It was 'raised from seeds presented to the Society by Dr. Royle, F. H. S.'. Therefore, a neotype is designated here, which is the same as that designated as lectotype for P. kasya Royle ex Parl. and P. khasya Royal ex J.D. Hooker, and the latter are made into two nomenclatural synonyms, to further the current usage of P. kesiya. According to Laubenfels (1988), the similar pines of mainland SE Asia and the Philippines differ from those of Sumatra noticeably by having a 'grass stage' for the seedling (after it emerges, the seedling grows for a season in height, then it grows without increasing in height in the second and third years). The needles of the mainland pines are 19-25cm long, and the seeds are nearly twice as heavy as those of the Sumatran ones. P. merkusiana Cooling & Gaussen was proposed to accommodate them, which, however, is a name published after 1958, without a designated type and thus not validly published (ICBN Art. 37.1). An earlier name, P. latteri, is already available. 'Grass stage', as a developmental feature, may not be a good character to separate species; this may partly explain why Laubenfels still treated P. latteri and P. merkusiana as synonyms of P. merkusii. However, in view of other morphological differences, such as longer needles and larger cones and seeds, as well as geographical distribution, P. latteri is here treated as a subspecies of P. merkusii.