summer 2007 Devising Strategies for Effective Mangement and Removal of Arundo donax in Portugal Research Overview Arundo donax in Portugal The invasion of exotic plants is a global epidemic that has the potential to permanently alter the physical, economic, and cultural landscapes of our world. Invasive plants are particularly harmful in Mediterranean riparian systems, like Portugal and California, where the native plant systems tend to support higher rates of biodiversity than any other local ecosystems (Salinas, 2002). In particular, the Giant Reed (Arundo donax) is a riparian species that has been aggressively encroaching on Portuguese channels and floodplains for a number of centuries. This aggressive invasive has caused a host of problems from displacing native vegetation, to increasing fire risk, and reducing flood conveyance. In the summer of 2007, I spent three weeks in Portugal observing and researching Arundo growth and management in three different watersheds: the Rio Real, the Sado, and the Lage. Specifically, I conducted interviews, mapped and visited outbreak sites, and reviewed existing policies in order to determine the feasibility of beginning a formal management effort of Arundo in Portugal. Since Arundo is not widelyrecognized in Portugal as an environmental hazard, my research in California (which focused on the hazard potential and the difficulties implicit in Arundo management) as well as my personal experience with the overgrowth of Arundo in my home watershed informed the direction of my research. Based on what I already knew, and what I learned in Portugal, I outlined the hazards caused by Arundo infestation in Portugal and defined legal and technical strategies for increasing and encouraging management of Arundo in Portuguese watersheds. The Giant Reed (Arundo donax) is classified as a perennial hydrophyte, which ranges nine to thirty feet tall, and grows in cane-like clumps that frequently transform into large infestation colonies (Dudley, 2000). The plant is a member of the Poacea (grass) family and closely resembles bamboo with its stout hollow culms that range.5-1.5 inches (1-4 cm) thick and its shallow rooting system (Coffman 2007). Due to its intolerance for freezing, the plant is typically found in elevations below 1,000 ft (304 meters) and along the margins of low-gradient riparian streams (less then 2 percent) (Dudley, 2000). Reproducing asexually through the Arielle finds Arundo donax in the Lage River, north of Oeiras. at University of California, Berkeley 1
at University of California, Berkeley displacement from managed habitats (Dudley, 2000). Arundo donax has long been acknowledged as a problem and a nemesis within Mediterranean climates. The natural flood regimes of Mediterranean streams coupled with anthropogenic disturbances such as the clearing of floodplain and riparian zones have provided the plant with the means and the territory to expand far beyond its intended habitat. The ability of the plant to transpire massive amounts of water (1.2-7.5 m/year, three times more than most native riparian species) combined with the dryness of its leaf and stem material makes it both a fire hazard and a drain on the limited water resources of Mediterranean climates (Iverson 1994, Coffman 2007). Furthermore, heavy infestations of Arundo can obstruct river flow, distribution of rhizome and stem particles, Arundo uses its close proximity to water sources to transport and spread its reproductive matter downstream. A hardy plant, Arundo is capable of surviving in a multitude of soil types (however, it grows particularly well in moist, well-drained, fresh/semi-saline soils) and its propugales can regenerate after enduring 40.5 percent moisture loss and deep burial under several feet of alluvium (Dudley, 2000 and Else, 1996). Arundo also has the distinction of being known as one of the fastest growing terrestrial plants in the world, with its capacity to attain growth rates of 2.3 ft (70 cm) per week and 4 in (10 cm) a day (Dudley, 2000). Also notable, is the fact that Arundo s habitat range expands beyond the margins of riparian vegetation and that its invasive takeover of riparian corridors is almost certainly [a result] of [Arundo] escapes and summer 2007 2
leading to an increase in flood and erosion damage to river neighboring properties. In addition to this, several authors have noted that Arundo provides very little habitat for native wildlife species (particularly in California, where it has been most widely studied), has the tendency to out-compete native plant species, and grows so densely that it physically restricts wildlife and human passage (Coffman 2007, Bell 1997). Even though Arundo donax has long been acknowledged as a distinctly Mediterranean problem, Portugal has no formal management or prevention techniques in place to deal with the impacts and spread of Arundo (T. Ferreira personal communication). My three-week research efforts showed me that Arundo has an extensive and far-reaching presence in Portugal. However, despite the obvious high-presence of Arundo in Portugal, there is very little interest from landowners and institutions to address (or even acknowledge) that Arundo could be negatively impacting in their environment. This is despite the fact that recent resarch has shown that invasive species bear a strong connection to the low richness and cover of endemic species observed in Portugal s watersheds (Aguiar et al 2006). Many Portuguese land and water agencies report to have high amounts of aquatic invasive species (like Arundo) and various problems due to the high presence of these species (i.e. like block waterways, etc). Furthermore, there are numerous reports from California-based scientists and institutional agencies that have identified Arundo as a severe fire and water hazard in drought-prone watersheds -- like Mediterranean watersheds. Recently, with the passage of the EU-wide Water Framework Directive (a directive aimed at improving the ecological status of European rivers) regional experts have begun advocating for an invasive vs. endemic rating system to be included within the parameters of the WFD s good ecological status test (Aguiar et al., 2006). Whether or not this rating system ends up being included in the WFD, the suggestion that invasive species are an applicable measure of watershed health holds enormous implications. Particularly since invasive species (like Arundo) vastly outnumber Portugal s native riparian plant communities. As a result, if this rating system were ever included with the parameters of the WFD, immediate steps would have to be taken to establish formal procedures in invasive management and removal. This research effort was an attempt to frame some of the steps that would be necessary for Portugal to take if the WFD included Arundo as a watershed metric. Methods In order to identify political and technical strategies for successfully managing Arundo in Portugal, I undertook the following: Research the extent of infestation of Arundo in Portugal. Since the extent of Arundo is largely unmapped in Portugal, three very different river sites were visited to gain a holistic and specific idea of how bad the infestation was. The first river visited was the Rio Real, located in the Ribeiros do Oeste watershed. In the Rio Real, agriculture was the dominant land use type with vineyards and pear harvesting being the major crop types. The second and third rivers visited were the Sado and Lage rivers. The former located in the Tajo-Sado basin, where I visited a site located immediately below a dam infrastructure located within a protected national park. The latter site, the Lage, was located in a largely urbanized, touristy region, where the stream was heavily channelized in order to protect buildings from unexpected floods and high run-off events. In the Rio Real, with the assistance of students from UC Berkeley and Universida de Técnica de Lisboa (UTL), I conducted a remote sensing exercise using 2004 realcolor aerial photograph (.5 meter/per pixel) of the Rio Real watershed. The photographs were loaded into ArcGIS 9.2, and then visually checked for all visible Arundo stands. The visible stands were then measured at University of California, Berkeley summer 2007 3
and delineated for the Rio Real and two of its largest tributary systems. Site visits were then conducted to confirm the rough accuracy of the mapping. At the Sado and the Lage, a number of Arundo outbreaks were photographed and recorded in order to assess the behavior of Arundo in those different land-use environments. Gauge scientific, landowner, and institutional perceptions of Arundo in Portugal. Interviews were held with technical experts from within the Portuguese invasive research community. I also met with farmers in the Rio Real watershed to discuss their reasons for allowing Arundo growth in their channels and discuss possible management solutions. I also, with the assistance of Cecilia Simões (ITL), translated a recent document regarding institutional perceptions of the growth of aquatic invasives (like Arundo) in Portugal. Find legislative and economic resources suitable for initiating and sustaining ongoing Arundo removal efforts. With the help of students from UC Berkeley and Universida de Técnica de Lisboa (ITL), I conducted a literature review looking at recent and current Portuguese legislation which could instigate Arundo management efforts. Identify removal procedures that could effectively remove Arundo while preventing further deterioration of water quality. Based on research findings from California, I identified effective removal procedures that could work within the legislative guidelines of Portugal law. Results An effective strategy for beginning a large-scale removal effort of Arundo in Portugal will depend upon: 1) timely documentation of the extent of Arundo in every Portuguese watershed, 2) implementation of existing (direct and implied) Arundo political measures and 3) effective application of Arundo removal techniques. Documentation Mapping invasive s and their spatial distribution is a necessary first step of any successful invasive management strategy (Dudley 2007 pers. comm., DiPietro 2002). However, this task can be laborious and exceed the technical and financial means of the Portuguese landowners and institutions charged with the Arundo removal task. Based on my conversations with scientists and landowners, I came to the conclusion that mapping Arundo had to be lowcost (both financially and technically), timely, and as accurate as possible. Therefore, I began experimenting with high-resolution, low-cost photographs to identify Arundo. At the Rio Real site, I identified 68.3 hectares (approx. 168.77 acres) using 2004 real-color aerial photographs (.5 meter/per pixel) and manual identification techniques using ArcMAP 9.2 software. Using this method, I identified Arundo (and another non-native grass species: Phragmites australis) in 57 kilometers (approx. 35.4 miles) of stream in less than 3 days. Visual site visits confirmed that for the most part this was an effective technique for rapid, large-scale Arundo documentation. The possibility of using free high-resolution sources from websites like Google Earth (a similar, but even lower cost procedure then the method used in Rio Real) was discussed and presented to a number of scientists and landowners. Overall, based on the Arundo mapping effort in Rio Real and the site visits made at the Lage and Sado basins, I concluded that the extent of Arundo infestation in Portugal was extreme and much greater than the overall extent in California. Consequently the hazard potential that I could observe (in terms of wide scale erosion and fire events) was much greater then I initially predicted and expected. I believe immediate action is required in order to quantify the extent of damage Arundo is having in Portugal, and hopefully salvage the habitat of the few native riparian species left in Portugal. at University of California, Berkeley summer 2007 4
Legend Rio Real Main branch Rio Real Watershed Potential Vegetation Vegetation Class Arisaro-Querceto broteroi Asparago aphylli-querceto broteroi S. Geoséries ribeirinhas 5.000 VEGETATION OF THE RIO REAL WATERSHED Meters at University of California, Berkeley summer 2007 5
IDENTIFIED PRESENCE OF ARUNDO DONAX IN THE RIO REAL WATERSHED at University of California, Berkeley summer 2007 6
Political Measures The EU s Water Framework Directive (WFD) could possibly provide the necessary motivational framework for instigating and enforcing Arundo removal in Portugal. Even though the WFD doesn t directly mention invasive species as a specific interest for water body health, the importance of Arundo removal is implied in Annex 5 Table 1.2 where high status rivers are noted as having riparian zones [which] correspond totally or nearly totally to undisturbed conditions. Since Arundo is a non-native invasive species which thrives and creates disturbance in the riparian corridor, the presence of Arundo in Portugal comprises the morphological quality and ability for Portuguese rivers to pass the WFD s good ecological status test. The WFD is not the only decree which can be applied towards the effort to remove Arundo in Portugal. In 1994 Executive Law #46 (amended to Executive Law #234 in 1998) obligated the Ministério do Ambiente, do Ordenamento do Território e do Desenvolvimento Regional (MAOTDR) to ensure that streams were clean and unblocked. Given Arundo s propensity to encroach into stream channels and constrain river flow, Arundo removal falls under the direct jurisdiction of MAOTDR and its affiliated agencies. Furthermore, Executive Law #565 passed in 1999, regulates the introduction of non-native species and directly identifies Arundo as an invasive species whose growth should be prohibited and curtailed by MAOTDR. These two legislative decrees form the jurisdiction of Arundo removal on a national and regional level. Furthermore, it should be noted that both these decrees were widely cited during Portugal s widespread effort to remove the floating invasive Jacinto aquático (water hyathica) from shipping and irrigation channels. The widespread reduction of Jacinto since the late 90 s attests to the possibility of applying the same legislation to Arundo removal in Portugal. Technical Measures In order to successfully restore Portugal s native riparian corridors, a great deal of Arundo would have to be removed from Portugal s floodplains. Based on the plant s dispersal method (i.e. downstream distribution of reproductive material during flood events), removal should be conducted after the flood season in late spring and early summer. Removal efforts should begin upstream and continue downstream. Mechanical removal efforts, such as uprooting the entire plant or cutting the above ground biomass and covering it with tarps, were selected as being the least environmentally damaging methods of removing Arundo. Bank stabilization methods should be utilized to decrease erosion and protect Portugal s levee structure. Competitive growth and resource studies conducted by Gretchen Coffman at UCLA confirm that several native riparian species have the capacity to prevent Arundo regrowth and be used as bank stabilization plants. Coffman identified a member of the Salix spp. (willow) species as being a species capable of withstanding the invasive growth patterns of Arundo. Bank stabilization projects that include willow stakes and plantings have a higher chance of restoring Portugal s native riparian structure. Other plants which showed a slight to extremely aggressive competitive interactions with Arundo include: Fraxinus angustifolia, Alnus glutinosa, Tamarix africana, and Sambucus nigra. Given Arundo s weak competitive abilities with other plants and its rapid recruitment on barren flood/fire damaged floodplains, active vegetation management and rapid disaster replantings could reduce the amount of Arundo re-establishment in Portugese floodplains. Future research topics include a thorough assessment into the affordability and efficiency of various Arundo removal techniques (ex: mechanical vs. chemical vs. biocontrol). Financial resources for Arundo removal (ex: funding from carbon sequestration, wildlife habitat at University of California, Berkeley summer 2007 7
funds, and recreational grants) must also be identified. Furthermore, and perhaps most importantly, further outreach efforts should be taken to notify the Portugese population (scientist, landowner, and institution alike) about the threats Arundo poses to their waterways. Acknowledgments This research was supported by a grant from the University of California Pinto-Fialon Endowment, administered by the (PSP). Special appreciation goes out to Matt Kondolf (UCB), Clara Landeiro (ITL), Eike Flebbe (ITL), Cecilia Simões (ITL), Melissa Parker (UCB), Ilídio Moreira (ISA), and Teresa Ferreira (ISA) for their time and guidance in this collaborative work. Selected Bibliography Aguiar, F. C. et al. 2006. Alien and Endemic Flore at reference and non-reference sites in Mediterraneantype streams in Portugal. Aquatic Conservation: Marine and Freshwater Ecosystems, Wiley InterScience. 2006. Bell, G.P. 1997. Ecology and management of Arundo Donax, and approaches to riparian Habitat restoration in southern California. Brock, J.H. et al, editors. Plant Invasions: Studies from North America and Europe. Blackhuys, The Netherlands. 103-113. Boose, A.B., Holt, J.S. 1999. Environmental effects on asexual reproduction in Arundo Donax. Weed Research. 39. 117-127. Coffman, G. 2007. Factors Influencing Invasion of Giant Reed (Arundo Donax) in Riparian Ecosystems of Mediterranean-type Climate Regions. Thesis. University of California, Los Angeles. Dudley, T. L. 2000. Arundo Donax. Bossard, C.C. et al. editors. Invasive Plants of California s Wildlands. University of California Press, Berkeley. 2000.53-58. Dudley, T. L. personal communication. February 25- April 18, 2007. DiPietro, D. 2002. Mapping the Invasive Plant Arundo donax and Associated Riparian Vegetation Using Hyperspectral Remote Sensing. Thesis. University of California, Davis. Else, J. 1996. Post-Flood Establishment of Native Woody Species and an Exotic, ArundoDonax, in a Southern California Riparian System. Thesis. San Diego State University. Ferrira, T. personal communication. April 10, 2007. Iverson, M. E. 1993. Effects of Arundo Donax on water resources. Jackson, N.E. et aleditors. Arundo Donax Workshop Proceedings, Ontario, CA. November 1993. 19-25. Salinas, J. M. et al. 2002. Riparian Plant Restoration in Summer-Dry Riverbeds of Southeastern Spain. Restoration Ecology 10 (4). December 2002. 695-702. Tickner, D.P. et al. 2001. Riparian plant invasions: hydrogeomorphological control and Ecological impacts. Prograss in Physical Geography 25 (1). 2001. 22-52. at University of California, Berkeley summer 2007 8
at University of California, Berkeley summer 2007 9