Betula cordifolia Regel

Betula cordifolia has until quite recently been regarded as a mere variety or northern montane form of Betula papyrifera – even Furlow (1997) in Flora of North America questioned whether it deserves specific status. However, its distinctness, both genetic and morphological, has been confirmed by Brittain & Grant (1965, 1967) and Grant & Thompson (1975). As a diploid it is unlikely to interbreed with hexaploid B. papyrifera, though it probably provides that species with one of its constituent genomes, one of the reasons why the two can be difficult to differentiate. To Canadian and American naturalists, however, the distinctness of B. cordifolia has never been in question, as reflected in its vernacular name: Speckled Birch refers to the twigs, which – unlike those of Betula papyrifera – are characteristically speckled with comparatively large lenticels. The somewhat fastigiate, erect, rather stiff growth and large rounded buds are also very characteristic. The cordate leaves, particularly on the spur shoots, are sometimes reminiscent of Mulberry, or even Davidia.

Less obvious characters that separate the species from B. papyrifera are the pendent young female catkins (initially erect in B. papyrifera), of which the central scale lobes are oblong and parallel-sided (more or less triangular in B. papyrifera) and the lateral lobes forward-pointing (very variable in attitude in B. papyrifera). Less reliable, but also commonly cited as a character is the greater number of leaf veins (9–12, against <9 in B. papyrifera), though this varies greatly with vigour and environment; the cordate leaf-base is also found in some individuals of B. papyrifera.

Confusion with Betula papyrifera means that the species’ distribution is not fully known, though it is abundant in Newfoundland and Labrador and Atlantic Canada, occurring separately from B. papyrifera, generally at higher altitudes or further north. Its presence in the far north of Labrador and Quebec is inferred. In Maritime Canada it grows as scattered trees on the hills, forming conspicuous orange-coloured patches in autumn amongst the ubiquitous dark-green Picea mariana; in the Laurentides it even forms patches of climax forest. Extending westwards into Ontario and even into Minnesota, it can be seen growing in the hills along the northern side of Lake Superior. In Vermont it is scattered sporadically through the dominant scrub of rather low B. alleghaniensis, but becomes scarce in the Appalachians south of New England. Unexpectedly, however, there are relict populations of large trees far to the south, on Black Mountain and neighbouring Mount Mitchell – further south than any existing records of Betula papyrifera; these trees are imposing, with freely peeling bark, but the population appears to be dwindling, since replacement seedlings appear to be absent.

Originating from a northerly region with short summers, Betula cordifolia grows well in the British Isles, though it can be killed back by late spring frosts and is reported to be sensitive to drought. Growth is less rapid than in B. papyrifera, but the trees are sturdier in appearance and of neat, upright habit with an ability to withstand exposure. The peeling reddish-cream to dark red-brown bark (generally darker than in B. papyrifera) and the autumn colour are both attractive. In cultivation, individuals of B. cordifolia come into fruit at a younger age than those of B. papyrifera of similar provenance. Collections made by Kenneth Ashburner grow well in Devon and at Ness Botanic Garden, as do others from Tony Schilling gathered on Mount Washington, New Hampshire in 1979 (Schilling 2379), and at RBG Edinburgh from a collection made by T.E. Clarke.

The species was described in 1861 by Eduard von Regel (1815–1892), director of the Imperial Botanical Gardens in St Petersburg, and an authority on Betula. Curiously, Regel gave the type locality as Novaya Zemlya in northern Russia, where the specimen (in De Candolle’s herbarium) was supposed to have been collected by de la Tylaie in 1826 – obviously not credible. However, further investigation reveals that the type (Geneva 698032) actually gives the location as ‘Terre-neuve’ (i.e. Newfoundland), evidently mistranslated into Russian by Regel.

In northern locations where it grows below B. glandulosa scrub, the two species interbreed to produce the wide-spreading or prostrate hybrid B. × minor (Tuck.) Fernald. Reports of tetraploids in B. cordifolia are likely to be due to the great difficulty in distinguishing between B. cordifolia and hybrid B. cordifolia × B. papyrifera.