Flora of North America species comparison

Clonal aspen groves develop rapidly following fires and other disturbances and may quickly decay in their absence as infections are transmitted through the connecting root system. Populus tremuloides is the most widely distributed tree in North America, found throughout cold and cool-temperate regions from coast to coast and from within the Arctic Circle to the north rim of the Valley of Mexico. It ranges from sea level in the north and east to the north slopes of high mountains in the southernmost part of its range. The southerly locations, like that on Mt. Livermore in the Davis Mountains, Texas, the most southerly stand in the flora area, may be Pleistocene relicts. Groves are often occupied by single clones and show no sexual reproduction but persist and spread by root suckers. Clone formation commonly results also in striking differences in appearance and phenology of adjacent groves or blocks of trees (B. V. Barnes 1969). Some individuals display a particularly rich, yellow autumn coloration that makes them a standout among North American trees, particularly in the West, where this richness was the basis for segregation of P. tremuloides var. aurea. There do not appear to be consistent regional differences within the species that would justify recognition of subspecies or varieties (Barnes 1975). Instead, there is as much variation from clone to clone within a region as there is among regions.

Populus tremuloides hybridizes with both the native P. grandidentata (P. ×smithii B. Boivin) and the Eurasian P. alba (P. ×heimburgeri B. Boivin) in southeastern Canada and the northeastern United States (B. V. Barnes 1961; T. A. Spies and Barnes 1982). Populus ×smithii occurs as far west as the Niobrara River valley, Nebraska, ca. 350 km west of the nearest present populations of P. grandidentata. Preformed leaves are more ovate than those of P. tremuloides and have larger teeth. Populus ×heimburgeri has transiently tomentose twigs, buds, and abaxial leaf surfaces. Contrary to some published reports (E. Lepage 1961; T. C. Brayshaw 1965b; B. Boivin 1966b; F. G. Bernard 1968), P. tremuloides does not hybridize naturally with P. angustifolia, P. balsamifera, or P. deltoides. The correct identification of such specimens is discussed under each of the purported parents.

The closely related Eurasian aspen, Populus tremula Linnaeus, is sometimes cultivated in North America, particularly as a columnar staminate clone (‘Erecta’). Its leaves are very similar in shape to those of P. tremuloides and usually have slightly larger teeth. Buds are often minutely hairy. Artificial hybrids between P. tremula and P. tremuloides, P. ×wettsteinii Hämet-Ahti, are sometimes grown for plantation forestry, particularly in the Great Lakes region.

(Discussion copyrighted by Flora of North America; reprinted with permission.)

Species ca. 30 (8 in the flora).

Populus has six well-marked sections, of which four occur in the flora area: swamp poplars [sect. Leucoides Spach (P. heterophylla)]; balsam poplars [sect. Tacamahaca Spach (P. angustifolia, P. balsamifera, and P. trichocarpa)]; cottonwoods [sect. Aigeiros Duby (P. deltoides and P. fremontii)]; and aspens [sect. Populus (P. grandidentata and P. tremuloides)]. Species within a section usually have separate distributions and hybridize freely where they come in contact. Species of different sections often have overlapping ranges and do not hybridize, except that members of sect. Aigeiros hybridize with all species of both sect. Leucoides and sect. Tacamahaca with which they are sympatric (J. E. Eckenwalder 1984). All known natural hybrids in the flora area are discussed under their parent species. Although some were originally described as species, they are not self-perpetuating. Because they can persist for decades by clonal growth, they can often be found in the absence of one or both parents.

All poplars are capable of clonal expansion, producing new trees from sprouts of root systems (soboliferous habit). Different species vary greatly in their propensity for cloning. The cottonwoods, Populus deltoides and P. fremontii, rarely produce root-borne shoots under natural conditions, with most clonal suckers arising from buried branches (S. B. Rood et al. 1994); the aspens, P. grandidentata and P. tremuloides, regularly form characteristic, dome-shaped clonal groves, some hectares in extent, with thousands of individual stems (B. V. Barnes 1966). Other species fall between these extremes. As a consequence of clonal growth, whole patches of trees may be solely staminate or pistillate and have uniform leaf morphology and phenology, sometimes making it difficult to find individuals of both sexes (Barnes 1969).

Collecting representative materials and identifying poplars is also complicated by seasonal heterophylly (variation in leaves along a shoot over the course of a growing season; W. B. Critchfield 1960; J. E. Eckenwalder 1980c, 1996). Two principal categories of leaves may differ in some characteristics, including size, shape, and marginal teeth. Preformed leaves (also known as early leaves) overwinter in buds as usually 3–10 tiny rolled up leaves before expanding with spring flush following flowering. Neoformed leaves (also known as late leaves) are initiated and expand during the growing season as shoots continue to lengthen after spring flush. Some neoformed leaves of rapidly growing suckers and young trees are the largest produced by each species, often more than twice as long as the largest leaves of mature trees. Preformed leaves are clustered at shoot bases, separated by internodes less than 1 cm. On long shoots (and all leaves on clonal sucker shoots in their first year), neoformed leaves are produced with internodes of 2 cm or more throughout the growing season until formation of winter buds, which contain next year’s preformed leaves. By the time winter buds form, most preformed leaves may already have fallen, leaving only neoformed leaves. As a result of this seasonality and unisexuality, individual staminate and pistillate trees should be marked and collected on three separate occasions: 1) at flowering; 2) when preformed leaves are mature (and when capsules are just opening on pistillate trees); and 3) with formation of mature winter buds, but before leaves have turned color and developed their abscission layer.

Poplar trees vary greatly in proportion of preformed and neoformed leaves during their life cycle and in how distinct these two leaf types are. Ironically, Populus heterophylla is one of the least heterophyllous species; its name alluding to variation among preformed leaves. In contrast, species with large teeth on preformed leaves, such as P. fremontii and P. grandidentata, are often strongly heterophyllous, with smaller and more numerous teeth on neoformed leaves (N. L. Britton 1886; J. E. Eckenwalder 1996). The most dramatic example of heterophylly in poplars occurs in the introduced Eurasian P. alba, which has angular-ovate preformed leaves and maplelike, palmately 5-lobed neoformed leaves that are unique in the genus. In all species, teeth of neoformed leaves have larger glands at the tips than those of preformed leaves. These glands are connected to leaf venation and, in combination with basilaminar glands on the adaxial side of a leaf blade at its junction with the petiole, secrete resins onto young, expanding leaves (J. D. Curtis and N. R. Lersten 1978).

Poplars are often grown for ornament, shelterbelts, timber, pulp, and specialty wood products. Breeding of poplars, especially using interspecific crosses, has become an important source of stock for plantation forestry. Selected clones of some crosses, such as Populus ×generosa (discussed under P. trichocarpa) and P. ×canadensis (discussed under P. deltoides), are among the fastest growing and most productive temperate trees. All native species (except P. heterophylla) and some natural hybrids are cultivated to some extent within their native ranges. In addition to native poplars, some Eurasian species and hybrids are commonly cultivated and may persist and spread clonally to form prominent landscape components. Because most of these taxa are represented in North America almost entirely by single clones and are solely pistillate or staminate, they are not truly naturalized here, despite their occasional prominence in the landscape. Eurasian taxa represented exclusively by staminate individuals produce no seed, and any seeds dispersed by the taxa that are solely pistillate of necessity have arisen from pollination by related (usually native) species. The most frequently encountered cultivated Eurasian species are included in the keys and are briefly described and discussed, along with any known natural hybrids, in the discussion for the most closely related native species.

Publication of the complete draft sequence of the Populus trichocarpa genome (G. A. Tuskan et al. 2006) has helped consolidate the position of Populus among the elite ranks of “model” organisms used for genetic, evolutionary, developmental, physiological, plant pathological, and herbivory related studies, among others. The published genome sequence may also lead to improved understanding of relationships within the genus but has not yet influenced molecular phylogenetic studies, which have suffered from limited sampling, although some interesting insights on wider past hybridization, than any seen today, have emerged (R. L. Smith and K. J. Sytsma 1990).

When A. Cronquist (1968, 1981) formulated the classification used in much of this flora, treating Salicaceae in a monotypic Salicales associated with Violales, it was already strongly suspected, based on floral development, leaf architecture, and shared secondary metabolites, that Populus and Salix were derived from within the predominantly tropical family Flacourtiaceae, as traditionally circumscribed (A. D. J. Meeuse 1975; W. S. Judd et al. 1994; J. E. Eckenwalder 1996). There are strong similarities between Populus and the monotypic eastern Asian genera Idesia, Itoa, and Poliothyrsis, and also with the more widespread Xylosma, found in all tropical regions except Africa, all are members of tribe Flacourtieae in traditional circumscriptions (D. E. Lemke 1988; D. J. Mabberly 1997). Molecular studies have amply confirmed these relationships and greatly extended them in showing that these and other flacourt genera are more closely related to Populus and Salix than they are to many other genera within traditional Flacourtiaceae (M. W. Chase et al. 2002). In fact, Flacourtiaceae are so heterogeneous that the family has been split into two, about half of the tribes being aligned with Populus and Salix in an enlarged Salicaceae (the earlier name, even though Flacourtia itself is included in the family). Chase et al. restricted tribe Saliceae to Populus and Salix, but their tribe Flacourtieae is paraphyletic and some of its genera (including Itoa and Poliothyrsis) will have to be transferred to Saliceae to establish consistent monophyly of groups within the family. Populus and Salix are sister taxa within this tribe but, of the two genera, some poplar species have retained more plesiomorphic features than have any of the willows, including more stamens and carpels, a less modified perianth, and broader leaves with more complex venation. Most of these species, belonging to two generally more southerly sections not found in the flora (sect. Abaso, with P. mexicana, and sect. Turanga, with P. euphratica, P. ilicifolia, and P. pruinosa), have strongly heteroblastic developmental leaf change, in which broad adult leaves expected through familiarity with the common temperate representatives of the genus, including all species in the flora except P. angustifolia, are preceded by willowlike, narrow, seedling and juvenile leaves (Eckenwalder 1980c, 1996b).

The earliest known fossil poplars, found in latest Paleocene and Eocene sediments within the flora area, are assigned to sect. Abaso and run the full gamut of leaf widths, some seemingly heteroblastic and others with narrow, fully adult leaves (J. E. Eckenwalder 1980c; S. R. Manchester et al. 2006). The common ancestor of Populus and Salix was probably somewhat more similar to Populus than to Salix, and this is partially corroborated by the recently described Eocene genus Pseudosalix (L. D. Boucher et al. 2003). Based on its morphology, this plant is the closest known relative of Populus and Salix and co-occurs with both genera in the Green River Formation of Utah and Colorado, so it is too late to represent their common ancestor. Nonetheless, its features are reasonably intermediate between traditional Salicaceae and their newly accepted relatives among extant traditional Flacourtiaceae. Populus strongly supports the concept of an enlarged Salicaceae and equally so an enlarged tribe Saliceae that would include some genera retained by M. W. Chase et al. (2002) in tribe Flacourtieae against their own evidence, which, admittedly, did not contain all relevant genera. The botanical community already recognizes a different set of family relationships than presented in the Flora of North America North of Mexico.

Keys. Three different keys are provided here to permit more effective identification throughout the annual phenological cycle in Populus: a key to flowering specimens that may also be used (but not optimally) for leafless, winter dormant specimens; a key to fruiting specimens; and a key to specimens with mature leaves.

Flowering in early spring is short-lived in poplars, with abscission of staminate catkins and pistillate floral bracts, and shrivelling of stigmas all occurring before emergence of the leaves, much as in precocious Salix species (in which bracts are not caducous). During anthesis, some winter characters involving twig colors and textures, leaf scars, and features of buds generally remain intact, so such features are used as supplementary characters in the key to flowering specimens. Characters of flower buds can also be helpful in winter identification but they are absent or obscured with emergence of the inflorescences at anthesis and are omitted from this key. The key includes characters of both staminate and pistillate individuals, which generally do not occur together on a single specimen. The characters of those rare individuals with mixed catkins or hermaphroditic flowers may be anomalous in other ways as well, and such individuals may not be readily identifiable.

At the time of fruit maturation, in late spring to early summer, staminate individuals are not distinguishable from non-fruiting pistillate individuals. Furthermore, while preformed leaves may then be expanded enough to reveal their mature characteristics, there may well not yet be any neoformed leaves present. Because the time of fruit maturation is not optimal for identifying non-fruiting individuals of Populus by their vegetative characters, the key to leafy specimens is designed primarily for late summer use, after the seeds and fruiting catkins have been shed. Because trees of the commonly cultivated, introduced P. nigra, P. simonii, and P. ×canadensis are rarely anything but staminate in North America (except in arboreta, botanical gardens, or poplar plantations), these taxa are excluded from the fruiting key.

Leafy specimens are best identified in late summer or early autumn when winter buds are forming and neoformed leaves are present and preformed leaves have not yet been shed. The key to leafy specimens is most effective at this time but should prove useful whenever mature, or nearly mature, leaves are present.

Three commonly encountered introduced Eurasian species (Populus alba, P. nigra, and P. simonii) are included in the keys but are not given full treatments in the text, where they are discussed following the most closely related native species. Other, less common, introduced species that are only found clearly under cultivation are omitted from the keys as are the relatively numerous different hybrid combinations (except for three introduced hybrids that have much the same status as the prominent introduced species: P. ×canadensis ‘Eugenei,’ P. ×canescens, and P. ×jackii ‘Balm of Gilead’). There are more different natural hybrid poplar combinations in the flora than there are poplar species. Although these hybrids can be found across the continent, they are usually much less common than their parents and typically grow with both, which gives a clue to their identity. The natural hybrids are not included in the keys because their numbers, additional variability, and general intermediacy between the parent species would greatly reduce ease of use of the keys and confidence in identification for the sake of relatively few specimens. Instead, hybrids are discussed following the descriptions of their parent species. A specimen that does not quite key out in the appropriate seasonal key here may well be a hybrid and the discussion of the species it comes closest to should be consulted for distinguishing characteristics. Overlapping parental distributions are also helpful, although, on rare occasions, hybrids may be found far beyond the natural range of one (or both) of their parent species.

(Discussion copyrighted by Flora of North America; reprinted with permission.)