Riparian Revegetation Using Native Seed: Feasibility Studies on the Lower Colorado River Grabau, Matthew R. 1 , Milczarek, Michael A. 2, Garnett, G. 3 , Bunting, Daniel P. 4 , Karpiscak, Martin 4 , Lewis, Elizabeth 4 , and David Quanrud 4 1 GeoSystems Analysis, Inc., Tucson, Arizona, USA, [email protected] 2 GeoSystems Analysis, Inc., Tucson, Arizona, USA 3 US Department of the Interior, Bureau of Reclamation, Boulder City, Nevada, USA 4 The University of Arizona Office of Arid Lands Studies, Tucson, Arizona, USA Ongoing Efforts: 1. Long-term monitoring of small-scale field study plots to determine irrigation regime effects and long-term vegetation dynamics. 2. Additional small-scale study plots seeded with Goodding’s willow alone to eliminate competition with cottonwood. More intensive weed control is also being implemented. Proposed Future Work: 1. Pilot-scale revegetation project on the lower Colorado River to evaluate scaling effects and logistics for large-scale seed collection, storage, and application. Conclusions: 1. Viability of Fremont cottonwood, Goodding’s willow, and coyote willow seed can be extended to greater than two years by freezing. 2. Soil conditions (bulk density, texture and fertility) and seeding rates can significantly affect plant establishment, growth, and species diversity. In general, sandy or compacted soils reduce plant vigor; higher seeding rates result in greater plant establishment, although the rate of establishment decreases at higher seeding rates. 3. Direct seeding resulted in dense cottonwood and willow establishment. Best results were observed with hydroseeding of un-cleaned seed onto furrows. 4. Tree establishment rates of approximately 10% were observed in field plots for Fremont cottonwood seeded at 25 PLS/ft 2 . 5. Tree establishment rates of approximately 1% were observed in field plots for Goodding’s willow at 150 PLS/ft 2 . 6. Saltcedar (Tamarix ramosissima), an introduced species, established at higher density than seeded native species. However, after one growing season saltcedar was primarily in the cottonwood understory and has experienced higher mortality than Fremont cottonwood. 7. Revegetation by direct seeding appears feasible using standard seed collection, preservation, application, and irrigation techniques. Additionally, direct seeding can increase genetic diversity and reduce costs compared to vegetative propagation techniques. 8. Intensive grass and broadleaf weed management is necessary in retired agricultural fields prior to, and after revegetation by direct seeding. Results Restoring native riparian plant communities is a major objective of management agencies in the West, with significant plans to revegetate areas currently farmed or dominated by invasive species with cottonwood and willow (e.g. USBR 2004). Vegetative propagation and subsequent planting of potted plants or rooted cuttings is currently the standard method of revegetation (Raulston 2003). If direct seeding can be achieved, restoration costs might be dramatically reduced while increasing tree density and maximizing genetic diversity within restoration areas. Passive revegetation from seed occurs in natural and managed riparian ecosystems where moist, bare soil is available during seed dispersal as a result of favorable hydrologic conditions. Examples include reintroduction of periodic flooding (e.g. Nagler et al. 2005) or managed drawdown of ponds during periods of seed dispersal (e.g. Roelle et al. 2001). Direct seeding of cottonwood and willow has not yet been implemented in large-scale restoration due to perceived limits of seed viability (Young and Clements 2003) and an unproven record of success (Raulston 2003). The Bureau of Reclamation is conducting feasibility studies to assess revegetation of riparian trees along the lower Colorado River using native seed. Feasibility studies conducted to date consist of a three year, phased germination, greenhouse, and field study program focused on Fremont cottonwood (Populus fremontii), Goodding’s willow (Salix gooddingii), and coyote willow (Salix exigua). Introduction Methods Discussion Study Locations and Implementation References We thank the Bureau of Reclamation LCR office for their continued support and funding of these efforts. In particular thank you to Barbara Raulston for initiating the project. Thanks to Dr. Stephen P. McLaughlin (The University of Arizona Office of Arid Lands Studies) for his guidance on seed preservation techniques. Thanks to Dr. Leslie Gunatilaka for accommodating research efforts at the University of Arizona Office of Arid Lands Studies. Thanks to Bill Seese, Mike Oldham and all Cibola NWR staff for their support in implemented field studies. Thanks to Havasu NWR, Bill Williams NWR, and the Ahakhav Tribal Preserve for permission and guidance for native seed collection. Thanks to Riverbottom Farms for assistance in field preparation and irrigation. Thanks to all of the University of Arizona student employees for their assistance in all phases of the study. Without all of these individuals, this project would not have been a success. Nagler, P. L., O. Hinojosa-Huerta, E. P. Glenn, J. Garcia-Hernandez, R. Romo, C. Curtis, A. R. Huete, and S. G. Nelson. 2005. Regeneration of native trees in the presence of invasive saltcedar in the Colorado River Delta, Mexico. Conservation Biology 19:1842- 1852. Raulston, B. E. 2003. Habitat restoration on the lower Colorado River, demonstration projects: 1995-2002. US Department of the Interior, Lower Colorado Regional Office, Boulder City, Nevada. 36 pp. Roelle, J. E., D. N. Gladwin, and B. S. Cade. 2001. Establishment, growth, and early survival of woody riparian species at a Colorado gravel pit. Western North America Naturalist 61:182-194. USBR. 2004. Lower Colorado River Multi-Species Conservation Plan, Volume II: Habitat Conservation Plan. Final. December 17. (J&S 00450.00.) Sacramento, CA. Young, J.A., and C.D. Clements. 2003. Germination of Seeds of Fremont Cottonwood. Journal of Range Management 56:660-664. 0% 20% 40% 60% 80% 100% Apr-06 Jul-06 Oct-06 Feb-07 May-07 Aug-07 Dec-07 Mar-08 Jun-08 Germination Trial Date Viability Cleaned, Ambient T, w/ O2 Uncleaned, Frozen, w/ O2 Uncleaned, Frozen, No O2 Seed Collection Storage Treatment: ~0% viability for all room temperature storage after 19 w eeks. Phase 1: Germination Study 0% 20% 40% 60% 80% 100% Apr-06 Jul-06 Oct-06 Feb-07 May-07 Aug-07 Dec-07 Mar-08 Jun-08 Germination Trial Date Viability Cleaned, Ambient T, w/ O2 Uncleaned, Frozen, w/ O2 Uncleaned, Frozen, No O2 Seed Collection Storage Treatment: ~0% viability for all room temperature storage after 17 w eeks. Figure 2: Coyote willow seed viability over time. 0% 20% 40% 60% 80% 100% Apr-06 Jul-06 Oct-06 Feb-07 May-07 Aug-07 Dec-07 Mar-08 Jun-08 Germination Trial Date Viability Cleaned, Ambient T, w/ O2 Uncleaned, Frozen, w/ O2 Uncleaned, Frozen, No O2 Seed Collection Storage Treatment: ~0% viability for all room temperature storage after 17 w eeks. Figure 3: Goodding’s willow seed viability over time. Figure 1: Fremont cottonwood seed viability over time. Phase 1 and Phase 2 Seed Collection: Lower Colorado River Seed Processing, Germination, and Greenhouse Studies: The University of Arizona Southwest Center for Natural Products Research and Commercialization, Tucson, Arizona Germination Studies 7-Gallon Pot Studies Phase 1: Germination study (April 2006-August 2008). Objective: Determine potential for long-term seed storage. ● Collection of native seed from various locations on the lower Colorado River (LCR) ● Experimental preservation methods: room-temperature storage, freezing, and oxygen removal. ● Periodic viability analysis over time. ● Germination tests in incubators and on soil beds at different temperatures. Phase 2: Greenhouse study (May 2006-September 2006). Objectives: 1) Determine establishment rates from direct seeding of Fremont cottonwood, Goodding’s willow, and coyote willow. 2) Determine if seed cleaning increases establishment rates. 3) Determine the effects of soil texture, compaction, and organic fertilizer on establishment and growth. 4) Determine the effect of seeding rates on growth and species composition. ● Loose soil, compacted soil, or sandy soil. ● Cleaned or un-cleaned seed. ● Biosol (organic fertilizer) at 0 kg/ha or 1680 kg/ha. ● Total seeding rates of 129 PLS/m 2 , 646 PLS/m 2 , or 3229 PLS/m 2 , 1/3 each of Fremont cottonwood, Goodding’s willow, and coyote willow. Phase 3: Small-scale field study (May 2007-September 2009). Objectives: 1) Determine establishment rates for direct seeding of riparian species. 2) Determine the effectiveness of large-scale seeding and irrigation methods. ● Seeding rate of 1076 PLS/m 2 as 40% Goodding’s willow, 40% coyote willow, and 20% Fremont cottonwood. ● Hydroseeding and broadcast seeding onto furrowed or level plots. ● Sprinkler irrigation for three weeks or surface irrigation only. ● In-situ soil moisture content, temperature, and EC monitoring network: Soil temperature and EC at 15 cm below ground surface (bgs) in each plot. Soil moisture at 15, 45, and 91 cm bgs. Groundwater elevation monitoring through a network of ten piezometers. ● Initial and long-term analysis of vegetation success: Establishment, growth, and survival. Split-block factorial design: Phase 2: Greenhouse Study Without Organic Fertilizer Loose Soil Compacted Soil Sandy Soil Loose Soil Compacted Soil Sandy Soil With Organic Fertilizer (Biosol TM ) Results POFR Stems/m 2 SAGO Stems/m 2 SAEX Stems/m 2 POFR g/m 2 SAGO g/m 2 SAEX g/m 2 Soil Type Loose 238 A 110 A 93 B 435 A 122 A 13 AB Compacted 265 A 117 A 175 A 258 B 86 A 20 A Sand 244 A 95 A 114 B 217 B 9 B 7 B Seed Type Cleaned 328 A 137 A 207 A 368 A 74 A 20 A Un-Cleaned 170 B 78 B 48 B 239 B 71 A 6 B Biosol Rate 0 kg/ha 284 A 140 A 161 A 268 B 67 A 13 A 1680 kg/ha 214 B 74 B 93 B 339 A 78 A 14 A Seeding Rate 129 PLS/m 2 32 C 16 C 51 B 116 C 41 B 13 A 646 PLS/m 2 119 B 82 B 105 B 229 B 78 AB 15 A 3229 PLS/m 2 596 A 223 A 226 A 566 A 98 A 12 A Least-Squared Means Table 2: Linear ANOVA model results after one growing season for Fremont cottonwood (POFR), Goodding’s willow (SAGO), and coyote willow (SAEX). Major Findings: • Soil type affected tree biomass (loose>compacted>sand). • Compacted soil decreased root penetration. • Application of cleaned seed resulted in twice the establishment of un-cleaned seed. • Organic fertilizer increased growth rates in sand, but did not increase growth rates in loose or compacted soil. • Organic fertilizer decreased tree establishment rates. • Higher seeding rates resulted in higher stem density, but favored POFR dominance (out-competition of willow species). • Higher seeding rates resulted in reduced POFR growth rates. • Higher seeding rates reduced establishment and growth of volunteer species. Phase 3: Small-Scale Field Study 0 2 4 6 8 10 12 14 16 Border Furrow Border Furrow Border Furrow Broadcast Hydroseed Hydroseed Cleaned Seed Un-Cleaned Seed Treatment Dry Biomass, grams/m 2 POFR TARA Grass and Sedge Average Dry Biomass: 106 g/m 2 Figure 4: Fremont cottonwood (POFR) and saltcedar (TARA) dry biomass after four months of growth. July 2007-Two Months of Growth in Furrow Plot: Cottonwood Establishment Along High-Water Mark June 2009-POFR Density and Height After Two Growing Seasons Table 3: Linear ANOVA model results after two growing seasons for Fremont cottonwood (POFR), Goodding’s willow (SAGO), and coyote willow (SAEX). Results POFR Stems/ m 2 SAGO Stems/m 2 SAEX Stems/m 2 TARA Stems/m 2 POFR Crown Cover TARA Crown Cover POFR Canopy Cover TARA Canopy Cover POFR Average Height, cm TARA Average Height, cm Sprinklers No Sprinklers 13.9 A 0.288 A 0.113 A 22.4 A 0.74 A 0.11 A 0.76 A 0.47 A 212 A 76 A Sprinklers 11.8 A 0.082 A 0.072 A 6.39 B 0. 41 B 0.13 A 0.43 B 0.25 B 150 B 75 B Seed Treatment Un-cleaned Hydroseed 17.9 A 0.437 A 0.166 A 14.9 A 0.62 A 0.13 A 0.65 A 0.38 A 176 A 80 A Cleaned Hydroseed 10.4 A 0.004 B 0.105 A 17.0 A 0.57 A 0.12 A 0.59 A 0.38 A 199 A 77 A Cleaned Broadcast 10.2 A 0.115 B 0.007 A 11.3 A 0.54 A 0.10 A 0.56 A 0.32 A 167 A 69 A Surface Irrigation Method Border 10.3 A 0.152 B 0.035 A 16.2 A 0.46 B 0.13 A 0.48 B 0.34 A 154 B 74 A Furrow 15.3 A 0.218 A 0.150 A 12.6 A 0.69 A 0.10 A 0.72 A 0.38 A 207 A 76 A Least-Squared Means Major Findings: • POFR establishment of 7% of PLS rates. • SAGO and SAEX establishment of less than 1% of PLS rates. • Sprinklers did not increase tree establishment and decreased growth rates. • Hydroseeding eliminates the need for seed cleaning. • Furrow irrigation encouraged higher growth rates and more even tree distribution. • Higher growth and survival rates for Fremont cottonwood than Goodding’s or coyote willow. • Volunteer saltcedar establishment was higher than seeded tree establishment. • Saltcedar growth rates were lower than cottonwood. Acknowledgements Phase 3 Small-Scale Field Studies: Cibola National Wildlife Refuge, Cibola, Arizona Site Characterization Small-Scale Study Implementation Soil Texture Soil Salinity Vegetation Monitoring Site Preparation and Seeding