Picea brachytyla (Franch.) Pritz.

Picea brachytyla was described in 1899 from Farges 806, gathered in Kangding in west-central Sichuan, China. It has traditionally been understood to have a discontinuous but wide range in western China, from northwest Hubei through southern Shaanxi, southern Gansu, Sichuan, to northwest Yunnan and extreme southeast Tibet. Variation across this large area has resulted in a number of named forms, but most recent treaments have tended to recognise only a single variety, var. complanata (Mast.) W.C.Cheng ex Rehder, restricted to the western part of the range (e.g. Dallimore & Jackson 1966; Bean 1976; Krüssmann 1985; Farjon 1990; Fu, Li & Mill 1999; Debreczy & Rácz 2011; Farjon 2017).

Phylogenetic studies appear to tell a different story. Recent phylogenies of Picea all agree that the genus includes three major clades, and that Picea brachytyla belongs to a clade containing mostly so-called ‘flat-leaved’ species from Asia (but not including those from Japan); that is to say its nearest relatives include members of the P. likiangensis species group, P. farreri, P. spinulosa, etc. (Ran, Wei & Wang 2006; Lockwood et al. 2013; Shao et al. 2019). What these and other studies have failed to agree on, however, are the details of the relationships between these species. On the other hand, together with more focussed studies such as Ru et al. (2016), Shen, Ran & Wang (2019) and Lyu et al. (2020) they provide evidence that spruce populations in northwest Yunnan and southeast Tibet, hitherto referred to P. brachtyla var. complanata, do not form a monophyletic clade with P. brachtyla.

This is a more intricate problem than it might first appear. The ‘var. complanata’ sampled in these studies are those populations in northwest Yunnan and adjacent regions, but var. complanata was not described from this region; it is simply the name that has been applied to material distributed here. Var. complanata was described – originally at species rank – from Wilson’s collections around Daxiangling, Ya’an Prefecture, central Sichuan (W 3030, 3031) (Sargent 1916). Picea brachtyla was itself first described from this same area of mountains west of the Chengdu Plain, and recent investigations have demonstrated that these populations represent only a single entity (Ru et al. 2016; Lyu et al. 2020). Var. complanata must, therefore, become a synonym of P. brachytyla, rather than a separate species to accommodate western populations. In his essay The Taxads and Conifers of Yunnan, Wilson (1926) seems to suggest that the name P. complanata should be applied to material distributed in western Yunnan, but unfortunately Masters chose Sichuan material for the type when he described P. complanata in 1906, and this concept must stand. Such are the rules of botanical nomenclature.

None of the other names that have been described, and which have since fallen into the synonymy of Picea brachytyla s.l., can be applied to the northwest Yunnan populations either, as they were all founded on collections made in west-central Sichuan: Picea ascendens Patschke (W 3024 from Wenchuan County), P. pachyclada Patschke (W 1896 from west of Xinjin) and P. sargentiana Rehder & E.H.Wilson (W 4048 from Wenchuan County or west of the same) (Sargent 1916). P. sikiangensis, a name of uncertain status and application, was used by Jean Hoch in reference to material from northwest Yunnan with glandular hairs (K. Rushforth pers. comm. 2020); it is treated here as a synonym of P. likiangensis subsp. balfouriana but its status requires further investigation. Neither of the late John Silba’s varieties of P. spinulosa (viz. var. pseudobrachytyla, var. yatungensis) are referable here (Silba 1990). It is regrettable that neither Ru et al. (2016) nor Lyu et al. (2020) sampled populations of P. brachytyla from further north in Gansu, Shaanxi or Hubei, the former study due to prohibitive costs, the latter because they could not locate any populations, which is interesting in itself.

This leaves the identity of the spruce populations in northwest Yunnan, that resemble Picea brachytyla in some respects but which have unequivocally been shown not to belong here, an open question. Lyu et al. (2020), Ru et al. (2016) and Lockwood et al. (2013) each found that these populations nest within the P. likiangensis species group, but could not resolve their exact placement. Lyu et al. (2020) point out that some studies (Shao et al. 2019; Shen, Ran & Wang 2019) have inferred a close relationship with P. farreri and postulate that northwest Yunnan populations ‘might be transferred to P. farreri if they form a monophyletic group’ (Lyu et al. 2020) but further research is needed. During research for this account no living trees were seen that were traceable to this area and that also resembled P. brachytyla. Dickson 13, introduced to the Royal Botanic Garden Edinburgh in 2013 as ‘Picea brachytyla var. complanata’ and widely distributed among sites participating in the International Conifer Conservation Programme under Edinburgh’s accession 20132162, is P. likiangensis (pers. obs.).

Going forward, contemporary workers should apply the name Picea brachytyla only to populations distributed in central and northern Sichuan eastward through southern Gansu, southern Shaanxi and into southern Hubei. Material from this area, or at least from Sichuan, is well represented in cultivation. Wilson collected extensively here, and many of the names listed in synonymy were founded on his collections. His seed collections, gathered on various expeditions, came under W 1282, 1530, 2076, 2078, 2079, 2081, 4048, and 4050 (Sargent 1916).

In the UK and Ireland Picea brachytyla would prove well suited to the climate across these islands, but particularly to milder, wetter, western regions where fine trees are now well distributed: 33.5 m tall × 93 cm dbh at Westonbirt, Gloucestershire in 2024, 32 m × 90 cm at Stourhead, Wiltshire in 2021, 29.5 m × 82 cm at Strathallan, Perthshire in 2017, 28 m × 90 cm at Stanage, Powys in 2023, and 27.5 m × 1.06 m at Tregrehan, Cornwall in 2024 (Tree Register 2024). P. brachytyla is curiously rare in North America; no Wilson material survives at the Arnold (Arnold Arboretum 2024) but it is represented there by material raised from Chinese Academy of Forestry seed sent in 1980, the best of which is now 26.8 cm dbh (Arnold Arboretum 2025). The Oregon nursery Bucholz & Bucholz began offering P. brachytyla from 1980 (Jacobson 1996) perhaps from the same sending, which was widely distributed, including via commercial channels. SICH 956 and 1126 were both gathered in Liangshan Yi, Sichuan, and introduced (as var. complanata) to multiple collections.

For W.J. Bean this was among the handsomest of spruces ‘with leaves of an unusually cheerful shade of green, contrasting with the vividly glaucous under­surface’ (Bean 1976). When branches are being buffeted by wind this creates a beautiful two-tone effect, which is also obvious on a still day looking up into the canopy of mature trees from the base. It seems to be a strong grower everywhere, and being late into growth is rarely troubled by late frosts (Bean 1976).