Gymnosperm Database
Link to jump to start of content Home Topics Bookstore Links Site Map Contact Us



Tsuga mertensiana growing by a lake in Washington [C.J. Earle].


Tsuga dumosa growing on Erlang Shan near Luding, Sichuan [Daniel Winkler].


Valid HTML 4.01 Transitional


(Endlicher) Carrière 1855

Common names

Hemlock (see Remarks), pruche [Canadian French] (Taylor 1993), ツガ属 (tsuga) [Japanese].

Taxonomic notes

The ten species in this treatment are:

Syn: Pinus Linnaeus sect. Tsuga Endlicher 1847 (Taylor 1993).Tsuga is often described as including two subgenera, subgen. Hesperopeuce (Lemmon) Ueno (syn. genus Hesperopeuce Lemmon; includes only T. mertensiana), and subgen. Tsuga (includes all other species). The subgenera are distinguished by both cone and foliage characters, but as noted below, are not supported by recent molecular studies, which generally divide the genus into Asian and North American clades. One species formerly described as an aberrant member of this genus, T. longibracteata W.C. Cheng, is now treated in a separate genus Nothotsuga, in some respects intermediate between Tsuga and Keteleeria.

A detailed molecular study by Havill et al. (2008) examined chloroplast and nuclear ribosomal nucleic acid sequences across a large dataset including many individuals from every extant species of Tsuga. That analysis validated the outgroup status of Nothotsuga and found strong support for a clade containing the North American species T. canadensis, T. heterophylla, and T. mertensiana, and an Asian clade containing all the other species, with Tsuga carolinana nested among them. T. caroliniana was also placed in the Asian clade in the analysis by Leslie et al. (2012); this result is morphologically and biogeographically implausible, and suggests that further molecular work in the genus would be informative.

The analysis by Havill et al. (2008) also supported the hypothesis that T. formosana is a good species (many authors have regarded it as synonymous with, or a variety of T. chinensis); their analysis placed it as sister to T. dumosa, a distinctive species of western China, and instead found that T. chinensis shares a clade with the southern Japanese hemlock T. sieboldii. The same analysis supported a close relationship between T. forrestii and T. chinensis, and did not support a common hypothesis that T. forrestii may have arisen by hybridization between T. chinensis and T. dumosa. Rather, T. dumosa was placed differently in both chloroplast DNA and and ribosomal ITS trees, suggesting a hybrid origin for this taxon; Havill et al. (2008) suggest an (unknown) paternal parent of an Asian species and a (now extinct) maternal parent of a European species. This result is compatible with the relatively western distribution of T. dumosa and also helps to explain the placement of T. caroliniana in the Asian clade.

Molecular methods have also been used to determine the probable ages of divergence from a common ancestor for all species of Tsuga. Those data are compatible with a Neogene divergence (i.e., less than about 23 million years ago) for all species, though late Paleogene (somewhat older) is more likely for the divergence between the North American (sans caroliniana) and Asian clades (Leslie et al. 2012). The same analysis (Leslie et al. 2012) confirmed the highly distinct lineage of Nothotsuga, finding an early Paleogene divergence; other studies (e.g. Gernandt et al. 2008) have pushed that divergence back as early as the late Cretaceous (ca. 80 million years ago). These chronologies have substantial uncertainty due to the skimpy fossil record for Tsuga, which yields few calibration points for molecular clock analyses.


Trees: Evergreen with a conic to irregularly ovoid (in some Asian species) crown; leading shoot usually drooping.
Bark: Gray to brown, scaly, often deeply furrowed.
Branches: Horizontal, often arranged in flattened 'sprays' and arched downward.
Shoots: Short (spur) shoots moderately developed; young twigs and distal portions of stem flexuous and pendent, roughened by peglike projections persisting after leaves fall.
Leaves: Leaves borne singly, persisting several years, ± 2-ranked or radiating in all directions, flattened to somewhat angular; abruptly narrowed to a petiolelike base, set on peglike projections, these angled, projected forward, sheath absent; apex acute, rounded or notched; stomata in two bands below; upper surface free of stomata except in subgen. Hesperopeuce; resin canals 1. Buds mostly rounded at apex, not resinous. Cotyledons 4-6.
Pollen cones: < 8 mm long, solitary, globose, brown, borne on year-old twigs.
Seed cones: Also borne on year-old twigs, maturing in 5-7 months, shedding seeds and falling soon thereafter or persisting for several years; pendent, ovoid or oblong (oblong-cylindric in subgen. Hesperopeuce), sessile or nearly so.
Cone scales: Persistent, shape various, thin, leathery, glabrous (pubescent in subgen. Hesperopeuce), lacking apophysis and umbo; bracts small, included.
Seeds: Ca. 3-5 × 2-3 mm, with numerous small resin vesicles; winged, with wing thin, 5-10 mm. x=12 (Taylor 1993, M.P. Frankis e-mail 1999.02.07).

Distribution and Ecology

Temperate North America and eastern Asia. Species of Tsuga are native to relatively moist climates where water stresses are minimal. Most are conspicuous, if not dominant, members of the communities in which they occur (Taylor 1993).

Big tree

Probably T. heterophylla.


Probably T. mertensiana.



"Hemlock wood is moderately strong and pliable and lacks resin ducts. With the decline of associated species considered superior in commercial value, hemlocks have become important in the timber industry, especially for pulp. Hemlocks are also widely used for horticultural purposes; numerous cultivars have been developed" (Taylor 1993).



The Latin name is from the Japanese name for T. sieboldii. T. canadensis and T. heterophylla were discovered early (before 1800) and until 1855 were simply treated as species of Pinus, but in 1842 T. sieboldii, the southern Japanese hemlock, was described as Abies tsugae. In 1847 Endlicher transferred it into Pinus sect. Tsugae, and then in 1855 Carriere gave us the modern name for the genus.

My friend Michael Frankis (M.P. Frankis, e-mail 1999.02.07) has advised me that the common English name "hemlock" refers to the percieved similarity in the smell of the crushed foliage of T. canadensis to that of the poisonous umbelliferous herb, water hemlock Conium maculatum. The two plants are of course totally unrelated, and Tsuga is not poisonous. I have tested this supposition in my back yard, where I have growing both Tsuga heterophylla and Conium maculatum. My plants do indeed resemble each other in odor, apart from a distinct resinous element in the scent of the Tsuga. I don't know for certain that this resemblance, observed in two plants native to western Washington, remains true in eastern North America or in Britain, but it seems plausible. The Conium leaf, by the way, resembles that of certain Cupressaceae in form (especially Thujopsis), a fact noted by Charles F. Partington in The British Cyclopedia of 1836, who says that the "hemlock spruce" is "so called from its branches in tenuity and position resembling the foliage of the common hemlock." André (1964) also asserted a resemblance between the foliage of Conium and that of the conifer, and that this was the reason the genus name name Conium was applied to poison hemlock. So there are several convergences between conifers and this flowering plant.

We now turn to Roman times, when the plant we now call Picea abies was known simply as Picea, because it was such a common tree in the land of Picea, which we now call Prussia. The common name "spruce" is also derived from the word "Prussia", as is the French name of hemlock, "pruche." But when did they start to give that name to the hemlock?

The first hemlock known to western science was T. canadensis, which would have been encountered by European immigrants to what is now maritime Canada and New England, places where this tree grows near the sea. It is thus likely that doctors or apothecaries, who would have been interested in the medicinal uses of the foliage of both plants, here first noted the resemblance in odor between Tsuga and Conium, and gave the plant its common name. Soon, though, it came to the attention of botanists. Linnaeus in 1763 acknowledged the existence of the species with the name Pinus canadensis, but it is likely that it was known to the botanical world much earlier (by the mid-1600s its presence in the New World would have been noted) and began to be known as "hemlock spruce", a fine example of an English binomial: the adjective (today we might say "epithet") "hemlock" modifying the generic noun "spruce." This gives a clue to the origin of both its French and English names. Speaking only of conifers (not angiosperms), the French domains in the New World were dominated by conifers of the genus Picea, with abundant P. rubens, P. glauca and P. mariana across the landscape. The English domains, on the other hand, were dominated by various species of Pinus. So the common name "hemlock spruce" suggests a northern origin for the term "hemlock", perhaps French, migrating thence to the English. At any rate, the name "hemlock spruce" was in wide use by the time Menzies used it to describe the new western hemlock (Tsuga heterophylla) he found in 1792, although the genus Picea would not be formally described for another 32 years, and the segregation of Tsuga would not occur for an additional 31 years.


André, Jacques. 1964. La résine et la poix dans l'antiquité, technique et terminologie. L'antiquité classique 33(1):86-97.

Gernandt, David S., Susana Magallón, Gretel Geada López, Omar Zerón Flores, Ann Willyard, and Aaron Liston. 2008. Use of simultaneous analyses to guide fossil-based calibrations of Pinaceae phylogeny. International Journal of Plant Sciences 69(8):1086–1099.

Havill, Nathan P., Christopher S. Campbell, Thomas F. Vining, Ben LePage, Randall J. Bayer, and Michael J. Donoghue. 2008. Phylogeny and biogeography of Tsuga (Pinaceae) inferred from nuclear ribosomal ITS and chloroplast DNA sequence data. Systematic Botany 33(3):478–489.

See also

Farjon (1990).

Last Modified 2016-02-15