A genus of six living species (one interspecific hybrid is known), native to riparian habitats in Asia.
Deciduous trees, typically medium sized to large, and fast-growing, often multistemmed and sometimes suckering freely. Bark furrowed, branchlets with chambered pith. Young shoots are robust and vigorous, often extending beyond the expanded leaves, taking on a distinctively ‘naked’ appearance, the bare shoot being terminated by a bud, which will expand as the shoot continues to grow. Terminal bud scales may be present but early-falling, or entirely absent; although this difference is useful in identification, in practice it is not easy to determine on fast-growing shoots. Leaves pinnate, often large. Leaflets elliptic to oblong, shortly acuminate at the apex, with serrate margins, variably pubescent (pubescence most visible on young growth). The number of leaflets can vary quite markedly across a single tree; they are largest on vigorously-growing young specimens. Plants monoecious, with flowers in pendulous catkins. Female catkins terminal on new growth, male catkins lateral on old growth or at the base of new growth. Female catkins elongate in fruit, with 20–80 winged nutlets, yellow-green at first, becoming brown (Grimshaw & Bayton 2009; Kozlowski et al. 2018).
The long, pendulous infructescences of wingnuts place them among the finest fruiting trees for late summer. On the other hand, the propensity of several species to make quite horrifying thickets through suckering cannot be ignored. Nevertheless, a well chosen species in an appropriate site, given a firm arboricultural hand, can make a handsome year-round tree.
The wingnuts’ closest relatives are the walnuts (Juglans), the hickories and pecans (Carya), and a single species each of Cyclocarya and Platycarya (Manos et al. 2007; Mostajeran et al. 2017). Pterocarya is easy to distinguish from both Juglans and Carya by its smaller, winged nuts on hanging catkins, and from Carya by its chambered pith, easily seen by splitting a twig lengthwise. The nut of Cyclocarya has a wing that uniquely forms a conspicuous disc right around it. Platycarya differs again as it has much smaller, narrow-winged nuts, a solid pith and a strikingly-erect male catkins and infructescences. The recent work of Kozlowski et al. (2018) is very helpful in giving a broad overview of the family, as well as copiously illustrated descriptions of the species and their wild status.
These genera diverged early in the Tertiary period, with unambiguous Pterocarya fossils going back at least to the Eocene epoch (34–56 million years ago). Fossils dating from the later Tertiary have been found widely across North America, Europe and Asia (Manchester 1987), suggesting a wide distribution in the warm temperate forests of the northern continents. The onset of the Pleistocene with its cycles of glaciation (Milne & Abbot 2002) led to Pterocarya becoming extinct in North America and Europe, and its range is Asia reduced. The range of the one surviving West Asian species contracted as late as the Middle Ages (Mostajeran et al. 2017), coinciding with a period of ‘climate chaos’ (Bradley et al. 2003). The ranges of several species are at present heavily pressed by human activity (Kozlowski et al. 2018) – this is in every sense a genus with a relict distribution.
Wingnuts are primarily trees of moist places found near rivers and streams. Such riparian habitats pose the twin challenges of waterlogging and periodic disturbance. Various responses to waterlogging have been documented; for example, periods of inundation stimulated lenticel production in P. stenoptera seedlings (Yang & Li 2016), presumably facilitating air movement into the aerenchyma tissue, which allows oxygen to reach the interior of roots even when air has been displaced from the soil by water. They are fast-growing, shade intolerant trees with light, rather weak wood, which may be very susceptible to fungal decay when wet: when considering alternatives to Ash (Fraxinus excelsior) as a timber tree, Wilson et al. (2017) described the wood of P. fraxinifolia as being more like that of alder or poplar, also genera rich in riparian trees. They have a strong tendency to produce suckers, epicormic shoots and water shoots. All this adds up to a picture of trees adapted to respond to disturbance by rebuilding damaged or destroyed canopies quickly and, in terms of resource use, cheaply.
Pterocarya flowers are pollinated by wind. The nut wings may assist wind dispersal, but longer distance seed dispersal of these riverside trees seems to be by water. In a population genetic analysis of P. fraxinifolia, Yousefzadeh et al. (2018) concluded that despite river flow being the main agent of seed dispersal, genes were well-mixed within river basins, up-river pollen dispersal in the wind presumably balancing down-river seed movements.
Like walnuts, Pterocarya leaves contain the alkaloid juglone; the levels may not be high enough to inhibit growth of plants under the canopy, as seen in some walnuts (Leuty 2012) but they are high enough for at least two species to be used for poison-fishing (Apel 2001; Hadjmohammadi & Kamel 2006). They also share diseases with Juglans. For example, they have been shown to be susceptible to thousand cankers disease, a beetle-borne fungal disease emerging as a threat to walnut production in the US (Seybold 2017). The two genera are sufficiently close that P. stenoptera can be used as a rootstock for commercial walnuts, to the end of increasing Phytophthora resistance (Browne et al. 2011).
Of the six living species, P. fraxinifolia stands alone as a tree of West Asia, restricted to a scatter of forests rich in Tertiary relict species. Of the remaining East Asian species, P. stenoptera is very widespread in China while P. rhoifolia, probably endemic to Japan, experiences more severe winter climates and regular snow than the others. P. hupehensis and P. macroptera have more restricted ranges in central and southern China, while P. tonkinensis is perhaps best considered a southern form of P. stenoptera, intergrading with it and extending its range through southernmost China into Vietnam and Laos – it is the most tropical of the wingnuts. Relationships within the genus require further research. The phylogenetic work of Mostajeran et al. (2017), while not a complete study, casts doubt on the validity of Manning’s (1978) classification. One further species, P. paliurus Batalin is treated here as Cyclocarya paliurus (Batalin) Iljinskaya, following Flora of China (Lu et al. 1999) and other contemporary work.
One interspecific hybrid is known, P. × rehderiana, a chance hybrid of cultivated P. fraxinifolia and P. stenoptera in cultivation. There is every possibility that other hybrids may arise from seed set in arboreta, although the high levels of seed set seen even in isolated Pterocarya individuals may lower the chances (N. Macer, pers. comm. 2019). One intergeneric hybrid has been made, between Juglans regia and an unidentified Pterocarya (McGranahan et al. 1986). This distant cross was only made possible by embryo rescue techniques, the aim being rootstocks for the commercial walnut industry. Two hybrids were made, one initially vigorous, but whether on its own roots, used as a rootstock or grafted onto a walnut it would not survive more than one or two seasons – ‘the hybrid seemed to have a fatal flaw’ (G. McGranahan, pers. comm. 2019). Other workers have tried and failed.
All six Pterocarya species are grown in specialist collections within our area. Specimens are sometimes found misidentified, and Dirr (2009) casts an aura of generalised doubt around the identification of wingnuts in North America. Two species, P. fraxinifolia and P. stenoptera, as well as P. × rehderiana, are seen well beyond arboreta. P. macroptera v. insignis seems poised to become a significant garden tree, while P. rhoifolia seems worthy of more widespread urban planting. While many are proving adaptable, all are at their best on fertile, moist soils in sheltered sites. Young trees are susceptible to damage in cold winters, and even established specimens are susceptible to late frosts (Bean 1976). Cold limits their cultivated range in northern Europe and central/northern North America.
Propagation is best from seed, but for selected clones vegetative methods are necessary. Firm softwood cuttings of some species can be rooted in early summer with the assistance of rooting hormone, but others (such as P. macroptera) will callus but not root under this regime (M. Fillan, pers. comm. 2008). Foster (2019) reports that epicormic shoots of P. macroptera v. insignis pulled from the base of the trunk when semi-ripe will root in mist or under plastic, but should be left in the cuttings tray, frost-free and dry, until growth restarts the following spring. Digging of suckers, layering or hardwood cuttings in winter are further options.
|1a||Rachis winged at least in part.||2|
|1b||Rachis not winged.||3|
|2a||Full length of rachis winged, wings flat. Terminal leaflet usually absent. Nut wings narrow to linear.||stenoptera|
|2b||Rachis partly or irregularly winged, with erect ridges or reduced wings. Terminal leaflet usually present. Nut wings rounded to ovate.||× rehderiana|
|3a||Terminal bud naked, branchlets without bud scale scars.||4|
|3b||Terminal bud with 2-4 caducous bud scales, branchlets with narrow bands of bud scale scars.||6|
|4a||Terminal leaflet absent. Nut wings narrow to linear.||tonkinensis|
|4b||Terminal leaflet present. Nut wings orbicular-ovate to elliptic-rhomboid.||5|
|5a||Leaf under 30 cm. Leaflets (3-)5-11(-15).||hupehensis|
|5b||Leaf 20-60 cm. Leaflets 9-21(-27).||fraxinifolia|
|6a||Leaflets 11-21. Rachis and petiole no more than pubescent at first. Nut wings to 2cm long.||rhoifolia|
|6b||Leaflets 7-13. Rachis and petiole tomentose at first. Nut wings to 3 cm long.||macroptera|