Final Report Summary - ESFFORES (EVALUATING SUCCESS OF FLOODPLAIN FOREST RESTORATION)
River-floodplain forests provide a wide array of ecosystem services essential for human well-being, such as conservation of biodiversity, flood and nutrient/pollutant buffering, and primary productivity. Unfortunately, most floodplain forests have been replaced with urban, industrial, agricultural and forestry development and are today limited to poorly functional, fragmented remnant patches. In the early 1980s, ecological restoration came up as a solution to deal with this environmental issue. The challenge was to make human use of floodplains compatible with the existence and self-sustainable and healthy riparian forests. Many restoration projects were implemented worldwide since then, specifically designed to deal with a variety of environmental problems affecting floodplain forests, including their direct loss by floodplain occupation with economic activities, but also invasion by exotic species, alteration of the natural stream flow regime cause by regulation, overexploitation of water resources, etc. The great variety of restoration approaches and a historic superficial and underfunded monitoring are responsible for a lack of systematic objective and standard criteria to evaluate success or failure of restoration actions, which ultimately jeopardizes their effectiveness. The ESFFORES project (Evaluating Success of Floodplain Forest Restoration) aimed at exploiting the great potential of the variety of restoration projects that remain unevaluated to assess their effectiveness and discuss about the existence of “universal” criteria for declaring success. By concentrating the efforts in evaluation of results instead of in goals definition and techniques for implementation, as restoration ecology had traditionally done, we were able to multiply the replication of traditional evaluations (usually local and site-specific), incorporating larger spatial scales and longer times elapsed since the end of restoration works. This set the basis for 1) finding indicators responsive to restoration and 2) discerning the factors leading to success or failure. The inclusion of very different restoration strategies from different world regions at large spatial and long temporal scales allowed us to discern “universal” patterns for both questions 1) and 2).
We first examined the actual extent to which recommendations for project monitoring after riparian vegetation restoration were being followed across the globe and where we might improve as a discipline, by systematically reading a total of 169 papers published in the international, peer-reviewed literature (González et al., 2015). We found five big groups of restoration strategies (in decreasing order of importance): hydro-geomorphology, active plant introduction, control of exotic species, abandonment of economic activities and control of herbivory/grazing. The indicators belong to three main groups: biodiversity (e.g. richness, species associations), structure (e.g. plant species abundance, physiognomy) and, much less frequent, processes (e.g. demography, nutrient cycling). As expected, we confirmed that published evaluations had been mainly local and covering less than 5-6 years. Surprisingly, we found that trajectories of change were more frequently assessed than static comparisons with control and reference conditions. Broadening the assessment approach (e.g. metrics and criteria used, factors of success explored, number of projects assessed, larger spatial and temporal scales and reporting failure) was formally recommended to increase their value for adaptive management and for academic use to test ecological principles.
In a second step, we evaluated the massive efforts to control invasive Eurasian Tamarix trees (tamarisks) in southwestern U.S. rivers carried out for the last two decades (González et al., In Preparation a, b), as Tamarix control is the most common strategy of floodplain forest restoration in this ecoregion. This tree has been associated with very negative effects on the ecosystem. In a collaborative study of 14 institutions, we monitored the vegetation response to restoration along hydrologic, soil and climatic gradients across six southwestern U.S. states. Tamarix control switched the dominance of exotics over natives, but more due to the removal of the former than to the enhancement of the latter. Tamarix abundance, a proxy of invasion severity, was greatly determined by climatic gradients. Native cover was not consistently promoted neither by any of the different removal techniques (mechanical, burning, biocontrol) nor by active revegetation. In addition, the four methods, especially those generating the highest disturbance, promoted secondary invasions of noxious plants, especially forbs. The abundance of riparian species was much lower in restored sites than in reference conditions, suggesting that ecological restoration in southwestern U.S. rivers has focused too much on weed control, overlooking local hydrology (the main driver of riparian plant communities). These results have helped to inform future restoration of tamarisk-infested rivers, as they have been shared with local stakeholders (e.g. the Tamarisk coalition). They also represent a very valuable experience to deal with present and future plant invasions in European rivers.
In a third step, we assessed the vegetation response to different strategies to restore floodplains of highly regulated rivers upon abandonment of commercial activities. This is the most common situation of floodplain forest restoration in the southwestern European ecoregion. In the Ebro river (NE Spain; González et al., 2016a) and the Garonne river (SW France; González et al., 2016b), we assessed restoration approaches applied in the last 20 years including channel widening, floodplain excavation, and abandonment of agriculture, gravel extraction and hybrid poplar plantations, with and without plantation of native species. We found that channel widening led to plant communities closely resembling those found in natural gravel bars, including new recruits of keystone tree species. Floodplain excavation resulted in pioneer, poor, non-strictly riparian communities. Immature, degraded, alternative stable states predominated in most of the abandoned agricultural and cleared poplar plantations, regardless of time elapsed since restoration and whether poles of native species were planted. Differences between rivers existed: a slight convergence of the recovering plant communities towards reference riparian communities was observed in the Garonne but not in the Ebro. This could be mainly due to a drier, hotter climate and the presence of xerophytic (steppic-halophytic) vegetation in upland areas of the latter river. However, in the two rivers, forest-like communities relatively similar to mature, natural riparian references were attainable spontaneously when hybrid poplars were allowed to resprout after clear-cutting, or abandoned without cutting.
Altogether, the results of the American and European ecoregions showed that:
1) Similar plant diversity metrics can be used to evaluate restoration success of very different restoration approaches, but some metrics are more responsive to restoration than others. The best metrics are those related to disturbance (i.e. punctual change of environmental conditions), namely key species (poplars, willows, tamarisks and other riparian species) and ecological groups (opportunistic, ruderal, light-demanding, etc.) that indirectly represent conditions frequently found in riparian zones.
2) Restoration outcomes in floodplain forest restoration resulting from different strategies can be partially predicted by readily measurable from environmental (especially climatic and hydro-geomorphic) and management (restoration approaches) quantitative, readily measurable parameters; with the former group operating at larger spatial scales than the latter. Management alone cannot restore pre-disturbance conditions, but can greatly influence vegetation successional trajectories. An unexpected result was that restoration success was not necessarily more likely with active revegetation.
3) In general, and for relatively short time frames < 20 years used in this study, environmental and management factors are more determining in structuring plant communities than time since restoration.
4) The future of success evaluations in floodplain forest restoration relies upon i) how to relate the changes in ecosystem structure induced by restoration, relatively easy to monitor, with changes in the associated functions, which are the ultimate but hardly ever tested goal of ecological restoration (e.g. conservation of biodiversity, flood and nutrient/pollutant buffering, primary productivity); and ii) how to provide environmental managers and restoration practitioners with criteria that, is not only responsive to restoration (“structural” indicators) but irrefutably represents the functionality of ecosystems improved by restoration.
Literature cited:
González et al. 2015. Restoration of riparian vegetation: a global review of implementation and evaluation approaches in the international, peer-reviewed literature. J Environ Manage 158:85-94.
González et al. In Preparation a. Vegetation response to removal and biocontrol of invasive Tamarix in southwestern U.S. rivers: A collaborative study including 416 sites.
González et al. In Preparation b. Secondary invasions of noxious weeds promoted by restoration-related disturbance in riparian systems.
González et al. 2016a. Strategies to restore floodplain vegetation after human activities have been abandoned. Restoration Ecology In Press.
González et al. 2016b. Poplar plantations along regulated rivers may resemble natural riparian forests after abandonment: a comparison of passive restoration approaches. Restoration Ecology In Press.
We first examined the actual extent to which recommendations for project monitoring after riparian vegetation restoration were being followed across the globe and where we might improve as a discipline, by systematically reading a total of 169 papers published in the international, peer-reviewed literature (González et al., 2015). We found five big groups of restoration strategies (in decreasing order of importance): hydro-geomorphology, active plant introduction, control of exotic species, abandonment of economic activities and control of herbivory/grazing. The indicators belong to three main groups: biodiversity (e.g. richness, species associations), structure (e.g. plant species abundance, physiognomy) and, much less frequent, processes (e.g. demography, nutrient cycling). As expected, we confirmed that published evaluations had been mainly local and covering less than 5-6 years. Surprisingly, we found that trajectories of change were more frequently assessed than static comparisons with control and reference conditions. Broadening the assessment approach (e.g. metrics and criteria used, factors of success explored, number of projects assessed, larger spatial and temporal scales and reporting failure) was formally recommended to increase their value for adaptive management and for academic use to test ecological principles.
In a second step, we evaluated the massive efforts to control invasive Eurasian Tamarix trees (tamarisks) in southwestern U.S. rivers carried out for the last two decades (González et al., In Preparation a, b), as Tamarix control is the most common strategy of floodplain forest restoration in this ecoregion. This tree has been associated with very negative effects on the ecosystem. In a collaborative study of 14 institutions, we monitored the vegetation response to restoration along hydrologic, soil and climatic gradients across six southwestern U.S. states. Tamarix control switched the dominance of exotics over natives, but more due to the removal of the former than to the enhancement of the latter. Tamarix abundance, a proxy of invasion severity, was greatly determined by climatic gradients. Native cover was not consistently promoted neither by any of the different removal techniques (mechanical, burning, biocontrol) nor by active revegetation. In addition, the four methods, especially those generating the highest disturbance, promoted secondary invasions of noxious plants, especially forbs. The abundance of riparian species was much lower in restored sites than in reference conditions, suggesting that ecological restoration in southwestern U.S. rivers has focused too much on weed control, overlooking local hydrology (the main driver of riparian plant communities). These results have helped to inform future restoration of tamarisk-infested rivers, as they have been shared with local stakeholders (e.g. the Tamarisk coalition). They also represent a very valuable experience to deal with present and future plant invasions in European rivers.
In a third step, we assessed the vegetation response to different strategies to restore floodplains of highly regulated rivers upon abandonment of commercial activities. This is the most common situation of floodplain forest restoration in the southwestern European ecoregion. In the Ebro river (NE Spain; González et al., 2016a) and the Garonne river (SW France; González et al., 2016b), we assessed restoration approaches applied in the last 20 years including channel widening, floodplain excavation, and abandonment of agriculture, gravel extraction and hybrid poplar plantations, with and without plantation of native species. We found that channel widening led to plant communities closely resembling those found in natural gravel bars, including new recruits of keystone tree species. Floodplain excavation resulted in pioneer, poor, non-strictly riparian communities. Immature, degraded, alternative stable states predominated in most of the abandoned agricultural and cleared poplar plantations, regardless of time elapsed since restoration and whether poles of native species were planted. Differences between rivers existed: a slight convergence of the recovering plant communities towards reference riparian communities was observed in the Garonne but not in the Ebro. This could be mainly due to a drier, hotter climate and the presence of xerophytic (steppic-halophytic) vegetation in upland areas of the latter river. However, in the two rivers, forest-like communities relatively similar to mature, natural riparian references were attainable spontaneously when hybrid poplars were allowed to resprout after clear-cutting, or abandoned without cutting.
Altogether, the results of the American and European ecoregions showed that:
1) Similar plant diversity metrics can be used to evaluate restoration success of very different restoration approaches, but some metrics are more responsive to restoration than others. The best metrics are those related to disturbance (i.e. punctual change of environmental conditions), namely key species (poplars, willows, tamarisks and other riparian species) and ecological groups (opportunistic, ruderal, light-demanding, etc.) that indirectly represent conditions frequently found in riparian zones.
2) Restoration outcomes in floodplain forest restoration resulting from different strategies can be partially predicted by readily measurable from environmental (especially climatic and hydro-geomorphic) and management (restoration approaches) quantitative, readily measurable parameters; with the former group operating at larger spatial scales than the latter. Management alone cannot restore pre-disturbance conditions, but can greatly influence vegetation successional trajectories. An unexpected result was that restoration success was not necessarily more likely with active revegetation.
3) In general, and for relatively short time frames < 20 years used in this study, environmental and management factors are more determining in structuring plant communities than time since restoration.
4) The future of success evaluations in floodplain forest restoration relies upon i) how to relate the changes in ecosystem structure induced by restoration, relatively easy to monitor, with changes in the associated functions, which are the ultimate but hardly ever tested goal of ecological restoration (e.g. conservation of biodiversity, flood and nutrient/pollutant buffering, primary productivity); and ii) how to provide environmental managers and restoration practitioners with criteria that, is not only responsive to restoration (“structural” indicators) but irrefutably represents the functionality of ecosystems improved by restoration.
Literature cited:
González et al. 2015. Restoration of riparian vegetation: a global review of implementation and evaluation approaches in the international, peer-reviewed literature. J Environ Manage 158:85-94.
González et al. In Preparation a. Vegetation response to removal and biocontrol of invasive Tamarix in southwestern U.S. rivers: A collaborative study including 416 sites.
González et al. In Preparation b. Secondary invasions of noxious weeds promoted by restoration-related disturbance in riparian systems.
González et al. 2016a. Strategies to restore floodplain vegetation after human activities have been abandoned. Restoration Ecology In Press.
González et al. 2016b. Poplar plantations along regulated rivers may resemble natural riparian forests after abandonment: a comparison of passive restoration approaches. Restoration Ecology In Press.