There are thought to be at least five species of Engelhardia, distributed in southern and southeastern Asia, but the genus is rather poorly understood, due to inadequate herbarium collections, and this number may change with further investigation. They are medium to large, deciduous to evergreen trees. The terminal buds are oblong and naked. The leaves are paripinnate (rarely imparipinnate) with 2–14 leaflets, the leaflet margins entire or serrate. Engelhardia species are monoecious, rarely dioecious; the inflorescences are terminal or axillary, on old or new growth; the male and female flowers held on separate spikes or combined in panicles; the male spikes are pendulous, solitary or clustered, the female spikes are erect or recurved. The male flowers are subtended by a 3-lobed bract, with 1–4 sepals (rarely none), no petals, and 3–15 stamens. The female flowers, subtended by an enlarged 3-lobed bract, have 4 sepals, fused to the ovary, and no petals. The fruiting spikes are pendulous, to 40 cm or more in some species; the fruit is a three-winged nutlet (Lu et al. 1999).
Engelhardia is a genus almost entirely unknown to Western gardeners, and probably is indeed only for ‘the speculative plant addict’ (Hudson 2004). Some have the potential to be highly ornamental. In their native habitats most are large trees, and some may have medicinal properties (Mabberley 1997).
Gathering information on these trees is hampered by alternative spellings of the generic name. The French botanist Jean-Baptiste Louis Claude Théodore Leschenault de La Tour (never was the standard abbreviation ‘Lesch.’ more welcome) first used the spelling Engelhardia (published without explanation of the name’s origin, in 1826, the year of his death), but used Engelhardtia in work published posthumously, in 1829, mentioning the name N. Engelhardt (IPNI 2019). Without taking a definite view on this subtle issue, we follow the majority of contemporary sources (for example Lu et al. 1999; Kozlowski et al. 2018) in using Leschenault’s first spelling, though Engelhardtia was used by Grimshaw & Bayton (2009).
The familiar temperate zone genera of Juglandaceae (Juglans, Carya, Pterocarya etc.) belong to Subfamily Juglandoideae. Engelhardia belongs to its sister group, Subfamily Engelhardioideae. These are in general trees of warmer climates: there are two centres of diversity, in Central America and South East Asia. In fruit, Engelhardia species recall Pterocarya and Cyclocarya in their pendulous catkins of winged nutlets, but they can easily be distinguished from these by the three nutlet wings (two in Pterocarya, a disc-wing in Cyclocarya) and by the solid rather than chambered pith (Kozlowski et al. 2018).
Here we mention three taxa, which may have the greatest potential for cultivation, at least on the southeastern fringes of our North American area. Mild winters and long, warm, moist summers are indicated. Based on his experiences in Cornwall, Tom Hudson notes that ‘basically all species would prefer rather more warmth, and a shorter winter, but I have a hunch that if a plant was going to reach a critical size it would at least survive’ (pers. comm. 2019).
Evergreen, to 30 m. Leaflets dark green, leathery with entire margins, glabrous with yellowish pelatate scales beneath. Inflorescences terminal on new growth. Variable, with a very wide distribution, from eastern Pakistan to Indonesia by way of southeast China and Vietnam. This species is sometimes placed in its own, monotypic genus as Alfaropsis roxburghiana (Lindl.) Iljinsk.: that view has some support from molecular data, and is increasingly adopted (for example Kozlowski et al. 2018).
As mentioned in New Trees (Grimshaw & Bayton 2009), young plants of two provenances were tested at Tregrehan, Cornwall, one from China in 2002 (under the synonym E. fenzelii Merrill), and one from Taiwan in 2004. They proved very slow-growing, with new growth late to emerge, although the new leaves are an attractive bronze colour. The Taiwanese plant reached 120 cm tall but is now dead; of two Chinese plants, one was killed by deer, the other survives but still only about 30cm tall (T. Hudson, pers. comm. 2006, 2019). Zhang et al. (1998) pointed to this species (again as E. fenzelii), as a candidate for cultivation in the American Southeast.
Deciduous, to 12 m. Leaflets with irregularly serrate or entire margins; underside tomentose, and with conspicuous glandular scales; rachis and petiole tomentose. Inflorescences lateral on old growth. India to Cambodia, reaching 1000 m altitude in southwest Yunnan: this variety includes the most northerly populations of a widespread and variable species. We have no evidence that it has been tried in our area.
Deciduous, to 20 m. Leaflets with entire margins; underside hairy or glabrous, without conspicuous glandular scales. Inflorescences lateral on old growth. Variable: several varieties have been described. Very widely distributed from the Himalayas to Malaysia and the Phillipines, by way of southern China.
According to Tom Hudson (pers. comm. 2006, 2019), E. spicata is spectacular when seen in full flower in Yunnan; his early experiments with cultivation in Cornwall have been hampered by root problems. Zhang et al. (1998) highlighted E. spicata var. aceriflora (Reinwardt) Koorders & Valeton, which reaches 1700 m altitude in Yunnan, as another plant with potential for the American Southeast.