Restoring functional riparian ecosystems: concepts and applications

Riparian ecosystems provide vital interfaces between uplands and rivers (Gregory et al., 1991; Naiman et al., 2005), sustain regional biodiversity (Naiman et al., 1993; NRC 2002; Sabo et al., 2005) and support key ecosystem services for society (Bagstad et al., 2012). Overland and subsurface freshwater flow wind through watersheds and floodplains and ultimately into streams and rivers. The hydrologic connections between land and water maintain riparian structure and function and sustain the ecosystem processes and services upon which many organisms, including ourselves, are dependent. Once these critical hydrologic connections are broken, riparian ecosystems, and their associated ecosystem services, decline (Nilsson and Berggren, 2000). Despite policy and funding support, many attempts to restore functioning riparian ecosystems fail (Williams et al., 1997). Countless ecologists and practitioners have pointed to the difficulty of restoring riparian ecosystems when the underlying causes of degradation have not been addressed and ameliorated, when practices have focused narrowly on the re-creation of specific floodplain elements (e.g. stream meanders) and when hydrologic connectivity has not been re-established (Clewell and Rieger, 1997; Molles et al., 1998; Hobbs and Harris, 2001; Ruiz-Jaen and Aide, 2005; Bernhardt et al., 2007; Jähnig et al., 2011). In sum, the projects failed because they were not process based (Beechie et al., 2010). Flood pulsing is an engine that drives riparian structure and function (Junk et al., 1989; Naiman and Décamps, 1997; Poff et al., 1997; Middleton, 1999). Flooding sculpts the floodplain, as sediment and vegetation are scoured through erosional forces. As floodwaters move across the floodplain, and encounter vegetation, energy is expended, and flood velocity declines. Flooding flushes the floodplain of accumulated salt, saturates soils, recharges alluvial groundwater and provides water to vegetation (Gregory et al., 1991; NRC, 2002; Naiman et al., 2005). Sediment is deposited when flood waters recede, which decreases stream turbidity and builds floodplains. Floodwaters can be rich in nutrients, and riparian ecosystems act as a buffer to absorb or dissipate the excess nutrients through the combined biogeochemical cycling activities of plants and soil microbes (Peterjohn and Correll, 1984; Gregory et al., 1991; Naiman et al., 2005). Fluvial geomorphic processes create temporal and spatial heterogeneity within the riparian landscape thereby sustaining a rich assortment of plant and animal life. Indeed, riparian ecosystems can be 'hotspots' of biodiversity (Naiman et al., 1993; Naiman and Décamps, 1997). The variation in water availability and in flood frequency and magnitude along a spatial gradient from river to upland allows for a diverse assemblage of plant functional types. Flood pulsing connects plant communities when seeds and other propagules entrained in floodwaters are dispersed across riparian landscapes (Boudell and Stromberg, 2008; Nilsson et al., 2010). These diverse and productive plant communities provide habitat to a variety of animals from permanent to temporary residents (NRC, 2002; Naiman et al., 2005). Riparian ecosystems are degraded by vertical, lateral and longitudinal hydraulic disconnection (Ligon et al., 1995; Ward and Stanford, 1995; Scott et al., 1999; Nilsson and Svedmark, 2002; Stromberg et al., 2006), with an accompanying loss of ecosystem services. Groundwater extraction that vertically disconnects the floodplain from its river stresses organisms (e.g. phreatophytic trees) that rely upon a readily available water resource. Levees disrupt the lateral hydrologic connections between rivers and riparian zones by forming an embankment that channelizes rivers and prevents overbanking (Middleton, 1999; Gergel et al., 2002; NRC, 2002). Impoundment by dams, like channelization, changes the flow and sediment regimes of the river downstream and impacts riparian ecosystems upstream by permanent inundation or by rising water tables (Ligon et al., 1995; Nilsson and Berggren, 2000; NRC, 2002; Johnson et al., 2012; Volke et al., in press). Sometimes, flow alteration is extreme, ceasing flow altogether. The changes from the river's natural flow regime can radically impact fluvial geomorphology, resulting in decreased diversity of fluvial landforms, increased erosion and channel incision, and decreased sediment and nutrient transport (Florsheim et al., 2008; Wohl et al., in press). Additionally, the timing, magnitude and duration of managed flow regimes are often out-of-sync with the life cycles of riparian species, causing population declines (Mahoney and Rood, 1998). Degradation of riparian ecosystems also occurs when land use practices remove riparian vegetation or alter watershed hydrology (NRC, 2002). Plant communities and soil structure are altered when vegetation is overgrazed and the ground trampled by livestock (Elmore and Kauffman, 1994). Banks degrade and collapse when cattle congregate along the stream, causing the channel to widen and leading to a suite of instream changes that affect aquatic organisms. Riparian ecosystems are narrowed when land is cleared for crops leaving only a remnant riparian buffer. Associated wetlands are drained to provide productive land for cultivation. Excess fertilizer in storm run-off flows can overwhelm the nutrient processing capacity of narrow or degraded riparian buffers, allowing the nutrient-enriched waters to enter the stream (Lovell and Sullivan, 2006). Riparian ecosystems in urban environments are influenced by many anthropogenic stressors such as vegetation removal, stream channelization, non-point and point source pollution and storm drainage that can exceed the natural peak flow capacity of the stream, leading to significant changes in channel morphology (Wenger et al., 2009). The combined effects lead to a suite of symptoms referred to as 'urban stream syndrome' (Meyer et al., 2005). Collectively, hydrologic disconnection and landscape modification create novel conditions for organisms (Poff et al., 1997). This may result in conditions that are better suited to upland species (Greene and Knox, 2014) or to those from other biogeographic regions (Stromberg et al., 2007) than to the original riparian species. Owing to their recognized importance, riparian ecosystems are often the focus of restoration efforts (NRC, 1992; Bernhardt et al., 2005). The Clean Water Act, National Environmental Protection Act, the Farm Bill and Endangered Species Act in the United States (NRC, 2002); the Water Framework Directive in Europe ([EC] European Commission, 2000) and internationally, The Ramsar Convention (Mattews, 1993) are frequently the policy basis that supports restoration efforts. In the United States alone, an average of more than one billion dollars per year was spent on riparian restoration projects between 1990 and 2003 (Bernhardt et al., 2005). It can be a lucrative business with whole industries focused on site evaluation, restoration and mitigation, and project monitoring. Despite the policy foundation, and availability of funding and expertise for riparian restoration, the successful restoration of degraded riparian ecosystems is faced with many challenges. There are numerous stakeholders invested in riparian ecosystems, with sometimes conflicting interests (Egan et al., 2011). Tension occurs between municipal and agricultural stakeholders and managers of floodplain ecosystems, although unexpected allegiances can develop as farmers and conservationists, both fight against long-distance water 'grabs' by cities. Residents rely on the services provided by rivers, and so local rivers are often diverted for irrigation, impounded to form reservoirs or leveed as part of a risk-management strategy to control flooding and provide a safer environment for floodplain residents. Additionally, many degraded riparian ecosystems exist in fragmented metropolitan environments where competition between stakeholders for land is intense (Bernhardt and Palmer, 2007). Thus, hydrologic connectivity, the foundation of a functioning riparian ecosystem, is rarely completely re-established, and riparian ecosystems if restored, typically exist as disjointed and functionally disconnected narrow strips within an altered landscape matrix. Compounding these issues, the science and practice of restoration ecology is relatively new (Jordan et al., 1987; Young et al., 2005), with early efforts at river restoration often failing to take an integrative process-based or ecological approach. River restoration emerged primarily from the practice of hydrological engineering for flood control (Palmer and Bernhardt, 2006) and to support fisheries through environmental flow allocation before 'spilling over' into the floodplain to encompass experimental flooding. Many practitioners focused on the restoration of specific floodplain structural elements such as stream meanders during 'hydrogeomorphological' restoration (Bernhardt and Palmer, 2011) and the planting of a selection of riparian and upland species during landscape restoration (Bernhardt et al., 2005). Often, practitioners were unable to remedy the underlying causes of riparian degradation, which derived from poor watershed planning and involved multiple municipalities and stakeholders (Roni et al., 2008). As a result of historical land use patterns, continuing anthropogenic impacts and early approaches to stream restoration, restored riparian ecosystems are frequently a pale facsimile of riparian ecosystems associated with free-flowing, wild rivers. Increasingly, restoration practitioners and scientists are calling for better integration and implementation of ecological theory with restoration policy and practice (Cairns, 1991; Higgs, 1997; Middleton, 1999; Bernhardt et al., 2007; Beechie et al., 2010). Years of restoration practice, research and monitoring have pointed to the need for process-based holistic approaches that focus on remedying the underlying causes of degradation, restoring structure and function and returning key hydrologic and geomorphic processes (Middleton, 1999; Bernhardt et al., 2007; Beechie et al., 2010; Bernhardt and Palmer, 2011; Wohl et al., in press). Many articles and more than a few books have been published on the need for the re-establishment of hydrologic connectivity (Molles et al., 1998; Middleton, 2002; Hughes and Rood, 2003; Rood et al., 2003); realistic restoration goals and adaptive, long-term monitoring and management of restoration projects (Kondolf, 1995; Hobbs and Harris, 2001; Ruiz-Jaen and Aide, 2005; Bernhardt et al., 2007; Jähnig et al., 2011); and collaborations between policymakers, managers, practitioners and scientists (Clewell and Rieger, 1997; Aronson et al., 2007; Bernhardt et al., 2007). This special issue of Ecohydrology arose from the 'Restoring functional riparian ecosystems: Concepts and applications' symposium at the 2013 5th World Conference on Ecological Restoration, in Madison, Wisconsin, United States. Participants discussed the ecological issues and challenges related to process-based approaches to restoring riparian ecosystems. We continue this conversation as we focus on case studies for a variety of restoration issues that practitioners, managers, ecologists and society face today. Several papers in this issue point to gaps in our knowledge about freshwater ecosystems and the need for further study and long-term monitoring. Springer et al. (2015) describe the ecological characteristics and anthropogenic impacts on freshwater springs in Alberta, Canada. They find that the springs are valuable in conserving plant diversity yet are understudied and at risk due to groundwater extraction and other impacts. Dufour et al. (2015) report on the role that abandoned channels play in sustaining regional biodiversity and call for management practices that allow for migration and avulsion of river channels. Nilsson et al. (2015) review Finnish and Swedish boreal streams that were restored after years of use for timber floating and conclude that two decades is not enough time for riparian organisms to recover and that long-term monitoring and an overall landscape perspective are needed. The importance of hydrogeomorphic processes and constraints for restoration is also discussed. Rood et al. (2015) find that forested riparian communities provide greater erosional resistance to flooding than grasslands; hence, conserving riparian forests would stabilize stream banks. Riparian tree species such as Populus sp. have life cycle events that are tied to regional climatic flood regimes (Mahoney and Rood, 1998), and Tiedemann and Rood (2015) report on the loss of Populus colonization sites on the flow-regulated Boise River, United States, caused by attenuated flows that failed to create the necessary surfaces for regeneration. Based on findings from their manipulative field experiment, they suggest that, along with managed flow releases, clearing and re-grading of the floodplain may be necessary for successful Populus recruitment. Dixon et al. (2015) report the outcome of a 500-year flood on passive restoration of Populus and associated species along the regulated Missouri River, United States. Despite the magnitude of the flood, its effectiveness for Populus recruitment was limited by the geomorphic legacies of 60 years of flow regulation, the extreme duration of flooding and its divergence from historical flow patterns. In a related paper, Johnson et al. (2015) point out that successful passive restoration through managed flow releases requires more than changes in flood frequency and magnitude on the Missouri River. Flow releases are devoid of sediment, leading to channel incision, loss of hydrologic connectivity and alteration of floodplain structure and function. Passive process-based restoration may not suffice to bring degraded riparian ecosystems to their former glory unless efforts focus more intensively on geomorphological structure and processes. It is the new, the novel or hybrid environment, that confounds our efforts to restore degraded riparian ecosystems to historical conditions, and ecologists increasingly are discounting the past as a restoration template. Conducting restoration in novel environments leads us to re-assess our definitions of restoration success, provides opportunities to learn from surprises and stimulates the development of innovative restoration approaches that take advantage of infrastructure and resources available in the novel system. These challenges and opportunities are both powerfully illustrated in urban environments where anthropogenic change is pervasive. Beauchamp et al. (2015) describe the influence of initial plantings and the restoration process on community assembly and soil properties in urban riparian sites. They report significant differences in tree community composition between paired restored and unrestored sites in the Baltimore metropolitan area; however, they also found that tree seedling composition between the sites became more similar over time. Soils of the restored sites had lower levels of Pb and Zn, but similar levels of Ca in comparison with unrestored urban riparian soils. Their results underscore the complexity of urban riparian restoration as species from surrounding sites disperse into restored sites over time, and the legacy of soil toxins from urban land use remains. While the reintroduction of riparian species into radically changed environments may be fraught with difficulties, innovative restoration approaches may develop and unintended positive consequences may be revealed when we focus our energies on how to restore hydrogeomorphic processes within these novel environments. Middleton et al. (2015) describe how increased discharge from a river diversion, originally intended to prevent the influx of oil released during the Deep Water Horizon spill in 2010, ultimately reduced salinity and increased productivity in Taxodium distichum (bald cypress) swamps. Another novel approach to functional restoration is the use of storm water and effluent discharge to support urban riparian forests and marshlands as described by Bateman et al. (2015). In this paper, the authors find that, in combination with actively restored sites, these 'accidentally' restored sites contributed to overall regional diversity in an urban landscape matrix and warrant protective measures to secure the novel water sources. It is a daunting challenge to implement process-based restoration in a world that is undergoing urbanization and climate change and in which multiple stakeholders prioritize different endpoints of riparian restoration. Successful restoration planning and implementation should solicit input from multiple and diverse stakeholders, and we must develop approaches that encourage their participation (Wohl et al., 2005). Broadbent et al. (2015) present a blueprint for the use of ecological endpoints as a tool to assess how stakeholders value the ecosystem services provided by riparian ecosystems in the arid southwestern United States. This approach is of particular importance when we include climate change impacts because the public is divided on whether or not they view climate change as a concern (Riffkin, 2014). Convincing a skeptical public to have a flexible attitude towards restoration outcomes while practitioners devise adaptive restoration approaches to mitigate and prepare for the effects of climate change will also be challenging. Ecologists have long discussed the moving target that is ecosystem restoration. The poignancy of this task is underscored when we include regional projections for climate change in our restoration plans. Both Perry et al. (2015) and Boudell et al. (2013) call for flexibility in restoration planning and implementation, and the need for valuing alternative recovery pathways, as we restore degraded riparian ecosystems in an uncertain world. Perry et al. (2015) also discuss the variety of tools that are available to assist planners with adjusting restoration approaches as needed. Boudell et al. (2013) describe the challenge of finding reference sites when targets are ambiguous and where the biotic community, owing to elements such as seed banks, has flexible, alternative development and recovery pathways. We extend thanks to the contributors to the conference symposium and to this special issue. We hope that the papers in this issue will stimulate further research on process-based riparian restoration approaches and discussion on how to implement these approaches in novel, changing environments.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.