Picea likiangensis
Sponsor
Kindly sponsored by
Sir Henry Angest
Credits
Tom Christian (2025)
Recommended citation
Christian, T. (2025), 'Picea likiangensis' from the website Trees and Shrubs Online (treesandshrubsonline.
Common Names
- Lijiang Spruce
- Likiang Spruce
- 丽江云杉 (li jiang yun shan)
- 原变种 (yuan bian zhong)
Synonyms
- Picea likiangensis var. bhutanica Silba
- Picea likiangensis subsp. bhutanica Silba (Silba)
- Picea likiangensis var. forrestii Silba
- Picea likiangensis subsp. forrestii Silba (Silba)
- Picea yunnanensis Lacass.
Infraspecifics
Other taxa in genus
- Picea abies
- Picea × albertiana
- Picea alcoquiana
- Picea asperata
- Picea asperata aggregate
- Picea brachytyla
- Picea breweriana
- Picea chihuahuana
- Picea crassifolia
- Picea engelmannii
- Picea farreri
- Picea × fennica
- Picea glauca
- Picea glehnii
- Picea jezoensis
- Picea koraiensis
- Picea koyamae
- Picea likiangensis aggregate
- Picea linzhiensis
- Picea × lutzii
- Picea mariana
- Picea × mariorika
- Picea martinezii
- Picea maximowiczii
- Picea mexicana
- Picea meyeri
- Picea morrisonicola
- Picea neoveitchii
- Picea obovata
- Picea omorika
- Picea orientalis
- Picea polita
- Picea pungens
- Picea purpurea
- Picea rubens
- Picea × saaghii
- Picea schrenkiana
- Picea sitchensis
- Picea smithiana
- Picea spinulosa
- Picea wilsonii
Tree to 50 m tall, to 2–2.6 m dbh. Bark rough and scaly from a young age, yellowish-grey in young trees, fissured and grey in older trees, freshly exposed bark orange-brown. Crown pyramidal or broad conical. First order branches long, slender, spreading, relatively sparsely set in older trees; second order branches spreading or ascending at first. Branchlets slender (stout in leading shoots and vigorous young trees), rather stiff, buff to pale yellowish-grey, darkening and becoming grey with age, prominently ridged and narrowly grooved, sparsely pubescent; pulvini 1–1.5 mm. Vegetative buds ovoid-conical or conical, 4–6 × 3–4 mm, slightly resinous; bud scales small, triangular, appressed, reddish-brown, persisting for several years. Leaves spreading radially, relatively sparse on shoot, directed slightly forward, weakly parted beneath shoots, 8–17 × 1–1.5 mm, linear, straight or slightly curved, rigid, transversely rhombic in cross-section, apex acute, pungent (except on shaded shoots); amphistomatic, with fewer lines of stomata above than below; leaf colour bluish-green above, more glaucous beneath. Pollen cones axillary, 2–2.5 cm long, rose-red at first, yellowish at maturity. Seed cones ovoid-oblong, sessile, base oblique, apex obtuse, 7–15 × 4–5.5 cm cm wide with opened scales; magenta at first, ripening through purple to pale purplish-brown. Seed scales thin and flexible, obovate or broadly obtrullate, 1.5–2.6 × 1–1.7 cm at mid-cone; lower surface smooth or striated, glabrous; upper margin rounded to obtuse, entire, sometimes weakly denticulate. Bracts ligulate, 2 mm long, entirely included. Seeds ovoid-conical, 2–4 mm long, dark brown; seed wings ovate-oblong, 7–14 mm long, light brown. (Farjon 2017; Rushforth 2008; Fu, Li & Mill 1999).
Distribution China NW Yunnan (Lijiang Range) and adjacent SW Sichuan, perhaps just into SE Xizang (Tibet)
Habitat On the Lijiang Shan Picea likiangensis co-dominates mixed forest on deep fertile soils at 2500–3800 m asl with Abies forrestii and Quercus guyavifolia. Other associates here include Taxus wallichiana, Tsuga forrestii, Quercus monimotricha, Griffitharia vestita and several species of Sorbus, Rhododendron, and Berberis.
USDA Hardiness Zone 8
RHS Hardiness Rating H6
Conservation status Least concern (LC)
Taxonomic note Until recently Picea likiangensis was considered to include up to five varieties (the type var. likiangensis, and var. hirtella W.C.Cheng; var. montigena (Mast.) W.C.Cheng; var. rubescens Rehder & E.H.Wilson; var. linzhiensis W.C.Cheng & L.K.Fu) distributed across the geographically complex mountainous regions of SW China. Most of these taxa were described in the early 20th century when study of this complex region’s phytogeography was still in its infancy. Rushforth reviewed this species complex in 2008 (Rushforth 2008) and found var. linzhiensis sufficiently distinct to raise it to species rank (P. linzhiensis) and this view has been broadly accepted. Among the remaining varieties Rushforth found no more than two entities: the type, restricted to a relatively small area of NW Yunnan and SW Sichuan; and another, more widely distributed in the same area but at higher altitudes, and differing quite markedly in morphology, which he circumscribed as P. likiangensis subsp. balfouriana (Rehder & E.H.Wilson) Rushforth (Rushforth 2008 and q.v.). Rushforth’s revision has greatly simplified the P. likiangensis complex and we follow his taxonomy here, but it is important to note that several other major works on conifers do not, e.g. Debreczy & Rácz (2011) and Farjon (2017) (from whence Plants of the World Online (2024) derives its own treatment). See also the article ‘Picea likiangensis aggregate’. Throghout the text below, we refer to Lijiang Spruce (P. likiangensis subsp. likiangensis) and Balfour Spruce (subsp. balfouriana) to avoid ambiguity.
The French missionary, explorer and botanist Père Jean-Marie Delavay discovered Picea likiangensis (subsp. likiangensis) on the Yulongxue Shan, part of the Lijiang Range in Yunnan, China, in 1884. It was described to science in 1899, based on Delavay’s specimens (Delavay 1031). George Forrest later introduced it to cultivation from the same locality in 1910 (Bean 1976). Bean’s statement that Wilson had introduced it from western Sichuan in 1904 (W 1834) is an error; W 1834 was collected from high altitude (‘3300–4600 m’ – Sargent 1916) and seems more likely to have represented subsp. balfouriana. Original trees under F 6746 survive at the Royal Botanic Garden Edinburgh, while others from various early 20th century introductions are still relatively frequent in collections. Both in gardens and literature trees referred to simply as ‘P. likiangensis’ usually belong here (Grimshaw & Bayton 2009).
Some mature examples are not unattractive trees, although there are several reasons to damn Lijiang Spruce with faint praise. Handsome specimens tend to be the exception to the rule that this species ages rapidly and badly. The rather open crowns of mature trees suggest they are prone to losing branches, and that even in a genus intolerant of shade Lijiang Spruce stands out as extremely sensitive, while the spreading branches and somewhat glaucous foliage can recall the aesthetic of Sitka (P. sitchensis) with, at least in Britain, all its vilified connotations.
In youth Lijiang Spruce can be a very vigorous tree, but it is prone to grow gracelessly, especially grown in the open when, more often than not, it quickly develops a displeasing, rounded, dumpy outline from which long extension shoots protrude in an alarming way. Such is the fate of one unfortunate plant at Edinburgh, a re-propagation from Forrest’s original F 6746, planted by King Charles III when still Prince of Wales on a prominent, open, sunny bank near the alpine houses. Its royal connections have spared this tree the removal that, on aesthetic grounds, it so urgently deserves (when in years to come it begins to shade the alpine houses, sense may finally prevail). Close planting in groups may help draw young plants up, but this is likely to come at the cost of reduced cone production, and the young seed cones, brilliant magenta at first before ripening through various purple-blue hues, are the chief ornamental feature.
None of the many reintroductions of Lijiang Spruce have produced more aesthetically pleasing trees. Collections include those made from the Lijiang Range under ACE 2385 and the various LPE collections introduced via Edinburgh in 2003, but only LPE 54 was ever widely distributed (Royal Botanic Garden Edinburgh 2024). The SICH collections seen in the preparation of this account (SICH 645, 704, 1126) appear to fit subsp. balfouriana, as would be expected based on geography, but because of historic uncertainty (see taxonomic note) many institutions still label SICH material simply P. likiangensis.
Aesthetics aside, the various introductions of Lijiang Spruce, from Forrest’s through to the present day, have demonstrated this is a reasonably tough tree. It is still planted for its perceived novelty but it is not so hardy as Balfour Spruce and tends to enjoy richer living on better soils in cool areas with high rainfall and humidity. It seems to be unsuited to a continental climate; multiple accessions at the Arnold Arboretum have failed to reach maturity (Arnold Arboretum 2024).
subsp. balfouriana (Rehder & E.H.Wilson) Rushforth
Common Names
Balfour Spruce
Synonyms
Picea balfouriana Rehder & E.H.Wilson
Picea likiangensis var. balfouriana (Rehder & E.H. Wilson) Hillier ex Slavin
Picea likiangensis var. rubescens Rehder & E.H.Wilson
Picea hirtella Rehder & E.H.Wilson (non Humboldt, Bonpland & Kunth)
Picea likiangensis var. hirtella (Rehder & E.H.Wilson) W.C. Cheng
Picea sikangensis W.C.Cheng
Tree 15–40(–50) m tall. Crown narrow, moderately dense. Bark breaking into small, square flakes. Young branchlets densely pubescent and orange- or reddish-brown. Leaves grey-green or bluish-green, arrangement dense, appressed above the shoot, 0.15–0.2 cm, thick. Cones small (4.5–10 × 3–4 cm), usually purplish-brown at maturity but sometimes yellowish-brown. Seed scales thin and leathery. (Farjon 2017; Grimshaw & Bayton 2009; Rushforth 2008; Fu, Li & Mill 1999).
Distribution China S Qinghai; W Sichuan; E Xizang (Tibet); northwestern-most Yunnan
RHS Hardiness Rating: H7
USDA Hardiness Zone: 5b-6
Taxonomic note The treatment of the Picea likiangensis complex adopted here follows Rushforth (2008) in lumping the varieties hirtella, montigena and rubescens (as applied by Fu, Li & Mill (1999) and Farjon (1990; 2017)) into a single entity. The rank at which the resulting taxon is treated has significant bearing on the name to be applied. At specific level P. montigena Mast. has priority, though there are doubts over the validity of this name as the type specimen may be of hybrid origin (Bean 1976). At the rank of subspecies balfouriana (Rehder & E.H.Wilson) Rushforth is the correct epithet, while at the rank of variety rubescens Rehder & E.H.Wilson has priority. Rushforth (2008) argues convincingly against the use of variety, given the geographical and ecological differences separating this taxon from P. likiangensis sensu stricto; it could logically be treated as a species but the nomenclatural uncertainties, and the very close relationship with P. likiangensis, make subspecies the more straightforward approach.
Balfour Spruce (Picea likiangensis subsp. balfouriana) was perhaps first introduced by Ernest Wilson in 1904 under W for Veitch 1834 and W 1836 (as ‘P. likiangensis’) (Sargent 1916) but no surviving trees are known. He collected it again in 1908 under W 2082 and W 2084 (both originally introduced as P. hirtella) and again in 1910 under W 4080 (this collection, from west of Kangting, would give rise to the name P. balfouriana Rehder & Wilson). W 4084 was also introduced in 1910, as Picea likiangensis var. montigena, and trees in several collections still bear this name or some variation of it (including an original tree and several propagules from it, all growing at the Arnold Arboretum, USA – Arnold Arboretum 2024) but all known plants from W 4084 are in fact P. asperata (pers. obs.; pers. comm. K. Rushforth).
Many older trees in collections are probably traceable to W 4080; examples survive at Edinburgh and Dawyck in the UK, and there are several Harry Smith introductions in Scandinavian collections and grafts from several of these have been distributed, including two at the Sir Harold Hillier Gardens in Hampshire, UK (Grimshaw & Bayton 2009). In the landscape arboretum at Gothenburg, a forest plot of Balfour Spruce raised from Harry Smith collections was planted in 1953–54; by 2005 the best of 48 extant trees was 22 m × 37 cm dbh (Aldén 2006).
Compared with their Lijiang Spruce contemporaries these Balfour Spruces have, on the whole, maintained fuller crowns which, together with their shorter leaves, more strongly forward and more densely set on the shoot, and often a darker glaucous-green, creates quite a different aesthetic. On closer inspection the densely hairy shoots, combined with these other vegetative characters, help to distinguish Balfour and Lijiang Spruces quite easily (Rushforth 2008).
It would be misleading to suggest that even with such improvements, Balfour Spruce is a thing of beauty, but a well-furnished tree in good health is certainly handsome and young trees are much more photogenic than similarly aged Lijiang Spruce. Furthermore Balfour Spruce is the hardier subspecies, coming from higher elevations and generally a more continental climate zone, and is more likely to thrive and persist in colder climates and on poorer soils. As in Lijiang Spruce, perhaps the chief attraction are the young seed cones, which are briefly a brilliant magenta before they begin to darken and then transition through purple to dark blue, finally maturing brown.
Balfour Spruce is most likely to be confused with Picea purpurea. Distinguishing them on vegetative characters alone is difficult; in general terms the shoots of P. purpurea are much more densely hairy than those of Balfour Spruce, and the leaves are generally shorter, greener, and more densely set on the shoots, often obscuring it entirely. The seed cones of Balfour Spruce, when available to consult, are larger than those of P. purpurea (Rushforth 2008). For further discussion on this point and on the species complex to which Balfour and Lijiang Spruce belong, see the article ‘Picea likiangensis aggregate’.
