Propagating native Salicaceae for riparian restoration on the hopi reservation

52

NATIVEPLANTS | SPRING 2006 he Hopi Reservation is located T in northeast Arizona (Figure 1A) where the tribe has been working to eradicate exotic salt-cedar (Tamarix ramosissima Leneb. [Tamari- caceae]) and Russian-olive (Elaeagnus angustifolia L. [Elaeagnaceae]) from streams and wetlands. Comprising about 2% of the reservation, these ripar- in Arizona ian and wetland communities are eco- logically and culturally valuable for livestock grazing, wildlife habitat, tradi- tional gathering, and ceremonial use (Lomadafkie 2003). Although the initial eradications were successful, the salt- cedar are resprouting. At the first Inter- tribal Nursery Council meeting in 2001, Thomas D Landis, David R Dreesen, Jeremy R Pinto, the tribe asked the USDA Forest Service and R Kasten Dumroese for help in propagating willows (Salix L. [Salicaceae]) and cottonwoods (Populus L. [Salicaceae]) to plant in these areas. During initial visits to project areas on ABSTRACT the reservation, we identified the principal riparian trees and shrubs: Fremont cot- The USDA Forest Service, USDA Natural Resources Conservation Service (NRCS), and tonwood, Goodding’s willow, coyote wil- the Hopi Tribe Office of Range Management have been working together on low, and arroyo willow (Table 1). Tribal native plant restoration projects in northeastern Arizona. The aggressive exotic members also took us to remote sites plants, Russian-olive (Elaeagnus angustifolia L. [Elaeagnaceae]) and salt-cedar where we found small stands of lanceleaf (Tamarix ramosissima Ledeb. [Tamaicaceae]), have invaded many wetland and ripar- cottonwood and quaking aspen (Table 1). ian areas on the Hopi Reservation, excluding willows (Salix L.), cottonwoods (Populus It is important to note that many of the L.), and other native plants. The tribe has been mechanically removing the invasives and has asked for help in propagating native species to plant in these project areas. wetland and riparian areas on the Hopi Although much information is available on how to collect willows and cottonwoods Reservation are geographically isolated and propagate them, some unique challenges exist on Hopi lands. Some species are and not always contiguous (Figure 1A). In common, while others are very rare and in some cases only a few individual plants addition, the aggressive invasion of salt- exist. The scattered locations of streams, wetlands, and seeps must be considered cedar and Russian-olive (Figure 1B) has during plant material collections to ensure that both genetic and sexual diversity are severely reduced and separated the popu- adequately represented. Another challenge is the determination of target plant stock lations of native willows and cotton- types that are appropriate on the diverse hydrologic conditions on the various proj- woods. From our field observations, we ect sites. Collected plant materials were taken to the NRCS Plant Materials Center in suspected that several of the existing plant Los Lunas, New Mexico, for both seed and vegetative propagation. stands comprised only one sex, and some- Landis TD, Dreesen DR, Pinto JR, Dumroese RK. 2006. Propagating native Salicaceae for riparian restoration times only a single individual (Pinto and on the Hopi Reservation in Arizona. Native Plants Journal 7(1):52–60. Landis 2003). One extended stand of arroyo willow along Bluebird Canyon K E Y W O R D S (Figure 1C) appeared to contain only restoration, culturally significant plants, invasive species, Intertribal Nursery Council, female plants, while a small grove of Salix, Populus lanceleaf cottonwood at Deer Springs was observed to be all males (Table 1). N O M E N C L A T U R E Removal of the salt-cedar and Russian- USDA NRCS (2005) olive has been completed on some sites, for instance, Polacca Wash (Figure 1D). Opposite. Old male catkins, like these on Salix exigua from the Hopi Reservation in Arizona, can be used to determine the sex of dioecious species during winter dormancy, and ensure that branch This area has also been fenced to keep collections include both sexes. Photo by Thomas D Landis cattle out and so is ready for outplanting, 53

NATIVEPLANTS | SPRING 2006 but the salt-cedar is already resprouting. Some test outplantings have proven suc- cessful so our challenge now is to pro- duce enough plant materials, of the proper stocktype, and have them ready when the sites have been prepared.

THE PROPAGATION STRATEGY

Once the native plants had been identi- fied, the next step was to determine where to propagate them. Because the Hopi do

A not have their own nursery, we contacted the Los Lunas Plant Materials Center (PMC) in New Mexico, which is operated by the USDA Natural Resources Conser- vation Service (NRCS) (Figure 1A). They have extensive experience growing native plants in greenhouses and outdoor fields for riparian restoration projects (Dreesen and others 2002). Most Salicaceae are vegetatively propa- gated with woody cuttings, and non- rooted cuttings are also widely used as live stakes or pole cuttings in riparian restora- tion projects. Because all members of the

B Salicaceae are dioecious, we had concerns about using vegetative propagation (Lan- dis and others 2003). Using only cuttings could compromise our objective of restor- ing the riparian and wetland areas on the Hopi Reservation with plants of the great- est possible genetic diversity. We wanted to pro-duce plant communities that were self-sustaining and so decided to produce all our plant material from seeds. C D Photos by Thomas D Landis Seed Production Options Seeds of the willow family are char- Figure 1. Riparian and wetland restoration sites on the Hopi Reservation are widely separated (A) acteristically very small and fragile and and overrun with the exotic trees salt-cedar and Russian-olive (B). Several sites, such as Bluebird therefore store poorly (Wycoff and Canyon, are isolated in sandstone canyons and all the arroyo willows were observed to be the Zasada 2003; Zasada and others 2003). same sex (C). The exotics have been removed at some sites, such as Polacca Wash, and the area fenced to exclude cattle, but the salt-cedar is resprouting (D). To our knowledge, seeds of willows, cot- tonwoods, and aspen had never been collected on the Hopi Reservation, so there was no chance of finding a supply of local seeds. We were also ignorant of the exact flowering periods for our tar- get species at these high elevations, which ranged from 1433 to 2073 m (4700 to 54 6800 ft). With limited time and funding,

NATIVEPLANTS | SPRING 2006 PROPAGATING NATIVE SALICACEAE ON THE HOPI RESERVATION TABLE 1

List of important Salicaceae (cottonwoods, willows, and aspen) found on the Hopi Reservation, Arizona.

Scientific name Common name Form Abundance Flowering period Sex of cuttings collected Males Females

Populus fremontii S. Wats. Fremont cottonwood Large tree Common Late May to June X X

Populus x acuminata Lanceleaf cottonwood Large tree Very rare Late April to May X Rydb. (pro sp.) [angustifolia x deltoides]

Populus tremuloides (Michx.) Quaking aspen Small tree Very rare Late May to June

Salix gooddingii (Ball) Goodding’s willow Small tree Uncommon Late May to June X

Salix exigua (Nutt.) Coyote willow Shrub Common May to July X X

Salix lasiolepsis (Benth.) Arroyo willow Shrub Rare March to April X Photos by Thomas D Landis visiting all the projects areas to collect probably because the developing flower seeds would have been impossible. capsules on these sexually mature cuttings created a drain on carbohydrate reserves. Collect Mature Cuttings Coyote willow was the most preco- and Produce Seeds at Nursery cious and produced flowers and some So, our strategy was to identify male seeds the first season (Figure 2B), but and female plants on the Hopi project most cuttings only grew leaves; we had to areas during the winter dormant period, wait until the second year to get apprecia- A collect mature cuttings with floral buds, ble numbers of seeds. The female arroyo and root them at the Los Lunas PMC willow produced some viable seeds the (Landis and others 2003). second year, which is interesting because We visited the project areas in late win- we collected no male cuttings. It is possi- ter and early spring of 2003 and 2004 and ble that the seeds resulted from apomixis learned to sex cottonwoods and aspen by or hybridizing with other willows at Los dissecting and examining sexual buds. Lunas PMC that flower at the same time. Because the willow buds were smaller, we With only male clones of the Goodding’s made tentative sex identifications by look- willow, no seeds were produced. The ing for dried-up catkins. Branch ends con- newly stuck Fremont cottonwood cut- taining floral buds (Figure 2A) were tings produced male and female flowers collected from trees with a pole pruner. but no seeds formed even after attempts to The woody cuttings were taken to the Los hand pollinate. In the second season after Lunas PMC and rooted in containers in a rooting, the Fremont cottonwood did not greenhouse with moderate success. Coy- produce any flower buds or capsules. In ote willow and lanceleaf cottonwood had addition, most of the mature Fremont good rooting success (80% to 90%), cottonwood cuttings never developed B whereas the rooting of Goodding’s willow good apical dominance and so were diffi- and arroyo willow were moderately suc- cult to manage. Figure 2. Mature Fremont cottonwood cut- cessful (75%). The sexually mature cut- Our initial plan was to mix the rooted tings (A) rooted reasonably well and some flowered but no seeds were produced after 2 tings of Fremont cottonwood had much cuttings from different locations, allow seasons. Coyote willow was the only species poorer rooting, about 65% after 3 mo, and them to flower and cross-pollinate, and to produce a reasonable number of catkins by the end of the first growing season, less produce locally adapted but genetically and some seeds the first year (B). Our initial plan to develop willow seed production areas than a third of the original cuttings were diverse seeds. A detailed propagation plan at Los Lunas PMC is still possible, but we had producing vigorous plants. The poor per- is outlined in Pinto and Landis (2003). to reconsider with Fremont cottonwood. formance of Fremont cottonwood is This may be possible with the willows, but 55

THOMAS D LANDIS AND OTHERS NATIVEPLANTS | SPRING 2006 A BC Photos by Thomas D Landis

Figure 3. Collecting Fremont cottonwood seed capsules in the field proved to be relatively easy (A), but willows had to be carefully identified beforehand (B). The cottonwood and Goodding’s willow seeds germinated well and produced shippable seedlings in only 4 mo (C).

the poor results with the Fremont cotton- in Ray Leach Cone-tainers™ (164 cm3 plants of arroyo willow and female wood forced us to reconsider. In addition, [10 in3]). Even though the Fremont cot- plants of lanceleaf cottonwood and concern about cross-pollination from tonwood and Goodding’s willow seeds establish them in seed production plant- other willows at Los Lunas PMC further were collected in mid-June, we were still ings at Los Lunas PMC. Then, we can reduced our confidence in developing able to produce large seedlings by the end foster cross-pollination and produce long-term seed production areas. of September—a growing season of only seeds of greater diversity for our restora- 4 mo (Figure 3C). In fact, we decided to tion efforts. Based on our experiences Field Seed Collection top prune these seedlings to maintain a with Fremont cottonwood, however, we In spring 2004, we returned to the Hopi favorable root-to-shoot balance. are uncertain how long it will take to Reservation and collected seeds from produce seeds of lanceleaf cottonwood. some of the project areas. In particular, we Dealing with Clonal Stands wanted to increase our genetic diversity by As previously mentioned, we sus- collecting from remote locations in the pected that the arroyo willow and lance- THE CHALLENGE OF western part of the reservation. We found leaf cottonwood stands were all of the QUAKING ASPEN abundant seeds of Fremont cottonwood same sex (Table 1) and perhaps even a (Figure 3A) at Pasture Canyon which, single clone. To test this hypothesis, we Conventional Techniques because of the great distance from Keams collected leaf samples of arroyo willow In the willow family, aspen is unique in Canyon and other collection areas (Figure in Bluebird Canyon and lanceleaf cot- that stem cuttings root poorly, so, nurs- 1A), should give us a good mix of genetics. tonwood at Deer Springs. Sample collec- eries have had better luck forcing sprouts We also collected seeds of Goodding’s wil- tion was done according to the from root sections and getting them to low at Blue Canyon; however, because procedure recommended by the USDA root (Dreesen and others 2002). We col- exotic willows such as Salix babylonica Forest Service National Forest Genetics lected root sections from quaking aspen at (auct. non L.) had invaded at some sites, all Laboratory (NFGEL) (Hipkins 2003). 2 locations on the Hopi Reservation but willows had to carefully identified before When the samples were processed at they did not produce sprouts. This may be seed collection could begin (Figure 3B). NFGEL, the results confirmed our due to the timing of the collections or the Direct seed propagation turned out to hypothesis that the extreme isolation of lack of vigor in the parent trees. be easier than we had imagined. The seed some of the project sites on the Hopi Therefore, we were excited when we capsules were cleaned using the proce- Reservation has resulted in clones that noticed aspen catkins on some of the dure of the Los Lunas PMC (Dreesen are genetically and sexually identical. trees in Aspen Canyon (Figure 4A). 56 2003) and seeds were sown immediately Our ultimate objective is to locate male When they were taken to Los Lunas

NATIVEPLANTS | SPRING 2006 PROPAGATING NATIVE SALICACEAE ON THE HOPI RESERVATION A Photo by Thomas D Landis

B Photo by Cavid R Dreesen

C

D

Figure 4. The small amount of aspen seeds that we collected on the Hopi Reservation was nonviable (A), but we were able to find some on the surrounding Navajo Reservation. Using the few seedlings that came from that collection, we tried a new “stacked propaga- tion” technique to “bulk-up” the number of plants (B–D). 57

THOMAS D LANDIS AND OTHERS NATIVEPLANTS | SPRING 2006 A

C B Photos by Thomas D Landis

Figure 5. Hydrologic zones have been delineated for riparian areas on the Hopi Reservation (A). Large stock types such as pole cuttings will be best in the flood-prone zones (B). Smaller, and less expensive, container plants would be best in wetland areas with a high water table and low erosion potential (C).

PMC and cleaned, however, the catkins aspen seedlings and the fact that severed collect more aspen seeds from the Hopi yielded no viable seeds. This may have roots will form new shoots. We created a sites. We will also plant some of the been due to poor pollination or mois- stack of Styroblock® containers with a Navajo aspen plants at these sites to ture stress from the extended drought in 4-l (1-gal) aspen seedling inserted in the encourage eventual cross-pollination. the region. On a subsequent trip, some top block. The lower blocks were filled viable seeds were collected from health- with a growing medium of composted ier aspen stands on the surrounding pine bark, pumice, and peat moss, and a DEFINING TARGET Navajo Reservation. This time, the thin layer of media was also sandwiched PLANT MATERIALS catkins did yield some viable seeds and in between the blocks (Figure 4B). By FOR RIPARIAN RESTORATION around a dozen seedlings were grown in next spring, the roots of the aspen 262 cm3 (16 in3) DeePots™ containers seedling should have grown down Currently, we are working with the Hopi and subsequently transplanted into 4-l through the cavities in the lower blocks to determine what is the ultimate stock (1-gal) containers for further growth. and totally occupied all the cells. At this type for the various outplanting sites. point, we will run a sharp knife blade The 2 primary considerations when Stacked Propagation between the blocks and sever the roots considering target stock types for ripar- In discussions with Larry LaFleur of (Figure 4C). The pruned root systems ian restoration are hydrologic zones and Smoky Lake Nursery, we learned about should form new shoots, which will the effect of erosion during flood events. a new vegetative propagation method develop into shippable aspen plants for quaking aspen that we are calling (Figure 4D). Hydrologic Zones “stacked propagation” (LaFleur 2004). In spite of its novelty, this is still veg- For restoration purposes, the use of This technique takes advantage of the etative propagation and so, to ensure hydrologic zones helps to account for 58 rapid and extensive root growth of wide genetic variation we will still try to the presence of subsurface water at the

NATIVEPLANTS | SPRING 2006 PROPAGATING NATIVE SALICACEAE ON THE HOPI RESERVATION TABLE 2

Tentative target plant materials for riparian and wetland sites on the Hopi Reservation, Arizona.

Species Stocktype Container volume Suitable outplanting sites Outplanting tool See figure 5a

Fremont TreePots™ 3875 cm3 (1 gal) Far overbank and transitional Tile spade cottonwood zones, abandoned meanders, PVC “tall pots” 6145 cm3 (375 in3) oxbow bends, and wide Tile spade overbank areas

Poles n/a Large-bit auger

Lanceleaf TreePots™ 3875 cm3 (1 gal) Far overbank and transitional Tile spade cottonwood zones, abandoned meanders, PVC “tall pots” 6145 cm3 (375 in3) oxbow bends, and wide overbank areas. Upland zones Poles n/a for shelterbelts Large-bit auger

Quaking aspen TreePots™ 3875 cm3 (1 gal) Far overbank and transitional Tile spade zones PVC “tall pots” 6145 cm3 (375 in3) Tile spade

Goodding’s TreePots™ 3875 cm3 (1 gal) Far overbank and transitional Tile spade willow zones abandoned meanders, PVC “tall pots” 6145 cm3 (375 in3) oxbow bends, and wide Tile spade overbank areas

Poles n/a Large-bit auger

Coyote willow Whips n/a Bank and overbank zones Small-bit auger

Ray Leach Cone-tainers™ 164 cm3 (10 in3) Bank zone Tile spade

Arroyo willow Whips n/a Bank and overbank zones Small-bit auger

Ray Leach Cone-tainers™ 164 cm3 (10 in3) Bank zone Tile spade various times of the year (Figure 5A). willow was rarer, but coyote willow was a table of target plant materials for the On Hopi lands, all plants of the willow ubiquitous and the most dominant Hopi project sites (Table 2). For cotton- family are found within reach of plant in the flood zone. woods, aspen, and the tree-type Good- groundwater. Large trees, such as Fre- ding’s willow, the best stock type would be mont cottonwood, lanceleaf cotton- Erosive versus Non-erosive Sites the 4-l (1-gal) TreePot™ or a custom- wood, and Goodding’s willow are Each project site must be evaluated made “tall pot” of 10 cm (4 in) diameter located in the far overbank and transi- for the maximum effect of water erosion PVC pipe with a 0.6 to 0.9 m (2 to 3 ft) tional zones (Table 2) where their deep during high water events. This is a major depth. Pole cuttings would also be an root systems can access water as it drops factor when considering stock types option, and we have started pole stooling in the dry season. Quaking aspen were because even high-quality nursery blocks at Los Lunas PMC for Fremont cot- found in the far overbank and transi- plants can be lost in a single flood event. tonwood and plan another for Goodding’s tional zones where their root systems For high water erosion sites, long pole willow. For large stream courses with the have access to the water table. The cuttings (Figure 5B) or deep containers highest erosion potential, pole cuttings smaller shrub willows occur in the bank like PVC “tall pots” are the best option. would be outplanted in the overbank and and overbank zones because of their transitional zones (Figure 5A). Because extensive fibrous root systems and their Tentative Stock Types lanceleaf cottonwood and quaking aspen flexible stems that will move with the Combining hydrologic information are relatively rare, trial outplantings of force of the high water flows. Arroyo with species characteristics, we developed large container sizes and poles will be 59

THOMAS D LANDIS AND OTHERS NATIVEPLANTS | SPRING 2006 Pinto JR, Landis TD. 2003. Propagating plant made in the overbank and transitional Propagation from Seeds materials for the Hopi Reservation, Arizona. zones. Lanceleaf cottonwood may also be Seeds of all species in the willow family In: Riley LE, Dumroese RK, Landis TD, tech- useful for shelterbelts around crops or can be cleaned easily but should be sown nical coordinators. National proceedings: structures (Table 2). immediately. With all species, seed propa- forest and conservation nursery associa- Of course, the cost increases exponen- gation is the easiest and best way to main- tions—2003. Ogden (UT): USDA Forest tially with larger stock types, and so, we tain genetic and sexual diversity. Sowing Service, Rocky Mountain Research Station. may try some trial outplantings with in miniplugs with subsequent transplant- Proceedings RMRS-P-38. p 80–84. smaller container sizes. Because of their ing to larger containers is space efficient, [USDA NRCS] USDA Natural Resources Conser- aggressive growth habits, arroyo willow but direct sowing in larger containers like vation Service. 2005. The PLANTS database, and coyote willow in the Ray Leach con- Ray Leach Cone-tainers™ eliminates the version 3.5. URL: http://plants.usda.gov tainers may prove useful. This may be need for transplanting and may result in a (accessed 15 Jul 2005). Baton Rouge (LA): especially effective in the wetland areas shorter crop cycle. National Plant Data Center. Wycoff GW, Zasada JC. 2003. Populus L.: poplar, with low erosive potential, such as Polacca cottonwood, aspen. In: Woody plant seed Wash (Figure 5C). At this site, water levels Stacked Propagation manual. URL: http://www.wpsm.net have actually increased since the salt-cedar This new technique is a good way to (accessed 15 Oct 2004). has been removed, and they do not drop bulk-up limited plant material and should Zasada JC, Douglas DA, Buechler W. 2003. as much during the dry season. work for all species in the willow family. It Salix L.: willow. In: Woody plant seed still suffers, however, the major drawback manual. URL: http://www.wpsm.net of all vegetative propagation of Salicaceae (accessed 15 Oct 2004). SUMMARY: —limited sexual and genetic diversity. LESSONS LEARNED SO FAR A U T H O R I N F O R M A T I O N Collecting Cuttings REFERENCES for Seed Production Thomas D Landis Our initial strategy of collecting Dreesen D, Harrington J, Subirge T, Stewart P, National Nursery Specialist (Retired) Fenchel G. 2002. Riparian restoration in mature cuttings for seed production back USDA Forest Service the Southwest: species selection, propa- at the nursery is more effective for willow JH Stone Nursery gation, planting methods, and case stud- than for cottonwood or aspen. Many wil- 2606 Old Stage Road ies. In: Dumroese RK, Riley LE, Landis TD, low species are very precocious and will Central Point, OR 97502-1300 technical coordinators. National proceed- often produce seeds the first year or defi- [email protected] ings: forest and conservation nursery nitely by the second. Cross-pollination associations—1999, 2000, and 2001. Ft David R Dreesen between species is a concern, however, Collins (CO): USDA Forest Service, Rocky Horticulturist especially because willows are insect- Mountain Research Station. Proceedings USDA Natural Resources pollinated. Mature cottonwood cuttings RMRS-P-24. p 253–272. Conservation Service do not root as well, and it is uncertain how Dreesen DR. 2003. Propagation protocol for Los Lunas Plant Materials Center long it will take to produce catkins and container willows in the southwestern US 1036 Miller Street SW seeds. Still, if seed collection is difficult, using seeds. Native Plants Journal Los Lunas, NM 87031 4:117–123. this becomes a viable option and is an [email protected] excellent way to maintain or even increase Hipkins V. 2003. General collection guide for vegetative materials. Placerville (CA): genetic diversity of the resultant seedlings. Jeremy R Pinto USDA Forest Service, National Forest Tribal Nursery Coordinator Genetics Laboratory. Seed Collection in Field [email protected] LaFleur L. 2004. Personal communication. This is the simplest and most cost- Smoky Lake (AB): Smoky Lake Nursery. R Kasten Dumroese effective method if care is taken to collect Manager. Research Plant Physiologist from several trees over the range of project Landis TD, Dreesen DR, Dumroese RK. 2003. [email protected] sites.With willows, careful identification is Sex and the single Salix: considerations necessary to identify and exclude exotic for riparian restoration. Native Plants Jour- USDA Forest Service species. Seed collection is more problem- nal 4:109–116. Southern Research Station atic if access to the project sites is difficult, Lomadafkie S. 2003. Personal communica- 1221 South Main Street tion. Kykotsmovi (AZ): Hopi Tribe. Wet- and especially if the timing of seed pro- Moscow, ID 83843 duction is unknown. lands Coordinator. 60

NATIVEPLANTS | SPRING 2006 PROPAGATING NATIVE SALICACEAE ON THE HOPI RESERVATION