Sex and the Single Salix

by Thomas Landis, David Dreesen & R. Kasten Dumroese

This article, reprinted with the authors’ permission, was originally printed in Native Plants Journal, Volume 4, Number 2 (Fall 2003), and can be viewed with more graphics at


Most restoration projects strive to create a sustainable plant community but exclusive use of vegetatively propagated material may be preventing this goal. The dioecious willows and cottonwoods of the Salicaceae are widely used in riparian restoration projects. Hardwood cuttings have traditionally been used to propagate these species in nurseries, and live stakes, branched cuttings, and poles are also used in bioengineering structures for bank stabilization. Woody cuttings are collected either from the project site or from stooling beds in nurseries during the winter dormant period. Unfortunately, little attention has been given to the sex of the donor plants. The potential problem is that a proper mixture of male and female plants may not be present in the hardwood cuttings or rooted cuttings destined for the restoration site—in the worst case they may be entirely 1 sex or the other. Fortunately, it is relatively easy to distinguish male and female plants. Collecting cuttings from many different plants and from a known ratio of males and females will ensure that the resultant plants will be able to produce viable seeds and achieve the ultimate goal of a sustainable plant community. The Salicaceae consists of hundreds of species of woody trees, shrubs, and subshrubs but contains only 2 genera: the willows (Salix spp. L.) (Zasada and others 2003) and the poplars, cottonwoods, and aspens (Populusspp. L.) (Wycoff and Zasada 2003). This plant family is noteworthy for several reasons: 1) individual plants are dioecious (each plant is either male or female)—rarely are plants both (hermaphroditic) (Rowland and others 2002; Zasada and others 2003); 2) members of the Salicaceae are keystone species in many riparian plant communities; and 3) they are commonly propagated by cuttings rather than seeds. Between 300 to 500 species of Salix and 36 species of Populus exist, with centers of species diversity being in the north temperate and subarctic regions of the world.Demand for willow and cottonwood plant materials has been increasing greatly in the last decade because of an increased interest in riparian restoration. A wide variety of plant materials of Salix and Populus are being utilized, and most of these use hardwood cuttings as propagules. In the western US, we use a large range of sizes of hardwood cuttings for stabilizing and revegetating stream banks (Table 1). Note that live stakes are stuck directly in stream banks, and the branched cuttings used in bioengineering structures like brush mattresses, fascines, and vertical bundles are expected to root, stabilize soil, and eventually revegetate the site (Dreesen and others 2002; Hoag and Landis 2002). This is vegetative propagation, and like all forms of vegetative propagation, has serious consequences.

Table 1

Type of Cuttings  Length (metric [English])  Use of cuttings
Microcuttings 5 to 10 cm (2 to 4 in) Nursery propagation
Live stakes 30 to 46 cm (12 to 16 in) Direct sticking on site
Branched cuttings (whips) 0.6 to 1.8 m (2 to 6 ft) Bioengineering structures on site
Poles 3.6 to 4.9 m (12 to 16 ft) Outplanted on site


The ultimate objective of any restoration project is to produce a plant community with the greatest possible genetic diversity, and one that is also self-sustaining. Our concern is that all of the above uses of willow and poplar involve cuttings and therefore vegetative propagation (see sidebar). The Salicaceae are ideal for cuttings because, with the exception of quaking aspen (Populus tremuloides Michx.) and Scouler willow (Salix scouleranaBarratt Ex Hook.) in the northern Rocky Mountains (Edson and others 1995), they root very easily. Sexual propagation is preferred, however, in restoration projects because seeds contain a mixture of genetic characteristics so that offspring will consist of both male and female plants. In contrast, vegetative propagation produces exact clones of the mother plant. This is of particular concern with dioecious plants, such as Salixand Populus, because all the progeny produced by vegetative propagation will have the same sex as their parent (Figure 2). If care is not taken to collect cuttings from a broad area, genetic diversity may also suffer.

Willow propagation techniques and their challenges

Most restorationists and nursery workers collect dormant cuttings of willow and cottonwood without any consideration to the sex of the parent plant. In nature, these species often reproduce naturally from root sprouts or buried branches and, as a result, adjacent plants on the project site may be from the same clone. Branches often break off parent plants during floods, become buried farther downstream, and root into new plants. If few genetically different plants are growing in an area to start with, perhaps the result of grazing, flood control, agricultural practices, or harvesting (Karrenberg and others 2002), all the willows or cottonwood plants in a riparian community can be from only a few parents or even a single parent. This was found to be the case with both lanceleaf cottonwood (Populus x acuminata Rydb. Pro sp.) and yellow willow (Salix lutea Nutt.) on the Hopi Reservation in northeastern Arizona. Genetic testing revealed that all the plants at the project site were clones of the same individual. This illustrates that natural asexual reproduction tends to prevail in areas where secondary succession occurs. Along major rivers, however, where periodic flooding occurs and wipes out the existing vegetation, primary succession prevails and Salix and Populus populations are usually of seedling origin.The unisex problem becomes even worse when willow and cottonwood cuttings are brought back to the nursery and used to start stooling beds. Because cuttings will be collected from these beds for many years, ignoring the sex of donor plants will seriously bias the sexual composition of the cuttings. Walk through the stooling beds in your local nursery next spring when the willows are flowering and you might be surprised.In summary, if care is not taken to ensure that a balanced mix of male and female plants are used for collecting cuttings, then the resultant plant materials will contain a disproportionate ratio of males and females. When large numbers of these same plants are used in riparian restoration projects, the longterm sustainability of the plant community will be jeopardized.


Because sexual and genetic diversity are critical in ecological restoration, the sex of donor plants should be identified in the field and sexual propagation should be used whenever possible. Unfortunately, a biochemical or genetic test to distinguish the sex of a donor plant is unavailable (Zasada and others 2003). Therefore, only field solutions are possible.

Identify the Sex of Parent Plants on Project Sites

By marking known clones of each sex in advance, cuttings or seeds of both sexes can be collected. It is simplest to determine the sex of willow and cottonwood plants when they are flowering. Depending on species, willow catkins may appear before (precocious), during (coetanous), or after (serotinous) new leaves appear in spring. This timing mechanism, along with habitat and elevation, discourages hybridization between species—natural hybrids have only been observed in small areas of overlap between closely related species that are usually separated by elevational distributions (Argus 1964; Argus 1973; Dorn 1974; Brunsfeld and Johnson 1985). Identifying anthers in male catkins and pistils in females is easy. Verifying the sex of the parent plants is also easy when they are producing characteristic capsules and cotton.During the winter dormant season, it is possible to identify the sex of dormant willows and cottonwoods by examining the size and location of floral buds, which are typically found in the upper branches just below the terminal vegetative bud. We have had good luck sexing cottonwood trees in the field using a pole pruner to collect branch tips from upper portions of the crown, and examining the floral buds with a razor blade and 10X hand lens. Male buds are typically larger than female buds and the floral structure can also be checked by slicing buds with a razor blade. Willows have much smaller buds, however, and so it is better to examine the cut buds under a dissecting scope. Once an individual plant is sexed, cuttings are collected, kept cool and moist in white plastic bags, and labeled. In keeping with tradition, we use pink flagging labels for female cuttings and blue for males.

Live Stakes

Collecting Propagules

Labeling plants of known sex also makes it easy to come back and collect seeds from females. It is also important to identify the species, especially in the willows. Cottonwood flowers in spring (Wycoff and Zasada 2003) but willows can be divided into species producing seeds in the spring or summer and those that disperse seeds in fall (Zasada and others 2003). Seeds can be processed and sown in either bareroot beds or containers (see Dawes 2003; Day and others 2003; Dreesen 2003). We have found it better to sow seeds of spring and summer flowering willows immediately but seeds from fall flowering willows have a cold dormancy requirement that may require cold-moist stratification (Zasada and others 2003).If you are unable to make frequent visits to the restoration site when seeds are maturing, another option is to bring stem cuttings of known male and female plants back to the nursery and force them to produce seeds. These cuttings should be collected during the winter dormant season when rooting success is highest and floral buds are completely formed. The trick is to be able to distinguish male and female plants at this time, and to collect mature cuttings with floral buds.


The dioecious nature of the Salicaceae requires that special measures must be taken to ensure that genetic and sexual diversity is maintained during propagation. The critical thing is to identify the sex of willows and cottonwoods when collecting cuttings in the field. Then, a sexually and genetically diverse mixture of cuttings can be obtained for bioengineering structures, direct sticking, or for establishing stooling beds in the nursery. Seed propagation should be encouraged whenever possible. Seeds can be collected from the project site, or it is relatively easy to force seed production from rooted cuttings in the nursery. Using these procedures, one can collect cuttings, force seeds, and grow genetically and sexually diverse willow and cottonwood seedlings in as little as 3 to 4 years.


We thank Tara Luna for her expertise in identifying willows in the field and comments on earlier drafts.


Argus GW. 1964. Preliminary reports on the flora of Wisconsin. No. 51, Salicaceae. The genus Salix. Wisconsin Academy of Science 53:217–272.Argus GW. 1973. The genus Salix in Alaska and the Yukon. National Museum Canada Publications Botany. No. 2.Brunsfeld SJ, Johnson FD. 1985. Field guide to the willow of east-central Idaho. Moscow (ID): University of Idaho Forest, Wildlife and Range Experiment Station. Bulletin No. 39. 95 p.Dawes D. 2003. Using a shop vacuum to clean Salicaceae seeds. Native Plants Journal 4:140.Day RA, Walter RP, Kozar JJ, Bricker SJ, Bowers JG. 2003. Propagation protocol for bareroot bigtooth and quaking aspen from seeds. Native Plants Journal 4:125–128.Dorn RD. 1974. A systematic study of some North American Salix [dissertation]. Laramie (WY): University of Wyoming.Dreesen DR. 2003. Propagation protocol for container willows in the southwestern US using seeds. Native Plants Journal 4:118–124.Dreesen D, Harrington J, Subirge T, Stewart P, Fenchel G. 2002. Riparian restoration in the Southwest: species selection, propagation, planting methods, and case studies. In: Dumroese RK, Riley LE, Landis TD, technical coordinators./p>National proceedings: forest and conservation nursery associations—1999, 2000, and 2001. Fort Collins (CO): USDA Forest Service, Rocky Mountain Research Station. Proceedings RMRS-P-24. p 253–272.Edson JL, Leege-Brusven AD, Wenny DL. 1995. Improved vegetative propagation of Scouler willow. Tree Planters’ Notes 46(2):58–63.Hoag JC, Landis TD. 2002. Plant materials for riparian restoration. In: DumroeseRK, Riley LE, Landis TD, technical coordinators. National proceedings: forest and conservation nursery associations—1999, 2000, and 2001. Fort Collins (CO): USDA Forest Service, Rocky Mountain Research Station. Proceedings RMRS-P-24. p 33–43.[ITIS] Integrated Taxonomic Information System. 2002. URL: (accessed 18 Jun 2003).Karrenberg S, Edwards PJ, Kollmann J. 2002. The life history of Salicaceae living in the active zone of floodplains. Freshwater Biology 47:733–748.Rowland DL, Garner ER, Jespersen M. 2002. A rare occurrence of seed formation on male branches of the dioecious tree, Populus deltoides. American Midland Naturalist 147(1):185–187.Wycoff GW, Zasada JC. 2003. Populus L.: poplar, cottonwood, aspen. In: Woody plant seed manual (on-line database). URL: (accessed 15 Apr 2003).Zasada JC, Douglas DA, Buechler W. 2003. Salix L.: willow. In: Woody plant seed manual (on-line database). URL: (accessed 15 Apr 2003).


Thomas D Landis
National Nursery Specialist
USDA Forest Service
Cooperative Programs
2606 Old Stage Road
Central Point, OR 97502-1300

David R Dreesen
USDA Natural Resources Conservation Service
Los Lunas Plant Materials Center
1036 Miller Street, SW
Los Lunas, NM 87031

R. Kasten Dumroese
Research Plant Physiologist
USDA Forest Service
Southern Research Station
1221 S Main Street
Moscow, ID 83843