Article | Open

Restoring tides to reduce methane emissions in impounded wetlands: A new and potent Blue Carbon climate change intervention

  • Scientific Reports 7, Article number: 11914 (2017)
  • doi:10.1038/s41598-017-12138-4
  • Download Citation
Received:
Accepted:
Published online:

Abstract

Coastal wetlands are sites of rapid carbon (C) sequestration and contain large soil C stocks. Thus, there is increasing interest in those ecosystems as sites for anthropogenic greenhouse gas emission offset projects (sometimes referred to as “Blue Carbon”), through preservation of existing C stocks or creation of new wetlands to increase future sequestration. Here we show that in the globally-widespread occurrence of diked, impounded, drained and tidally-restricted salt marshes, substantial methane (CH4) and CO2 emission reductions can be achieved through restoration of disconnected saline tidal flows. Modeled climatic forcing indicates that tidal restoration to reduce emissions has a much greater impact per unit area than wetland creation or conservation to enhance sequestration. Given that GHG emissions in tidally-restricted, degraded wetlands are caused by human activity, they are anthropogenic emissions, and reducing them will have an effect on climate that is equivalent to reduced emission of an equal quantity of fossil fuel GHG. Thus, as a landuse-based climate change intervention, reducing CH4 emissions is an entirely distinct concept from biological C sequestration projects to enhance C storage in forest or wetland biomass or soil, and will not suffer from the non-permanence risk that stored C will be returned to the atmosphere.

Introduction

Methane emissions from wetlands, predominantly from freshwater systems such as peatlands and tidal fresh and low salinity wetlands, comprise about 1/3 of global CH4 emissions from all sources1. Soil microbial respiration and low rates of methane oxidation in anaerobic, water-saturated soils result in substantial CH4 emissions in those freshwater settings. Depending on the timescale of analysis, climate warming due to CH4 emissions can partially or entirely offset climatic cooling due to C sequestration in freshwater wetland soil2. In contrast, in saline wetlands, including salt marshes, saline mangroves, and seagrass beds, CH4 emissions are typically minor because abundant sulfate ion in seawater limits microbial CH4 production and emission3. Thus, with high rates of net C storage and minor CH4 emission, saline wetlands generally have a strong cooling effect on climate3,4,5,6.

Commonly, however, tidal exchange of saline water between the coastal ocean and emergent, tidal wetlands, including salt marshes and mangroves (herein referred to as “tidal wetlands”), has been blocked or restricted by human activity, and those alterations can dramatically freshen and degrade the ecosystem7,8. Both drainage and impoundment of tidal wetlands have been practiced for a wide range of purposes during the past several centuries of human civilization7,8,9. Blockage or restriction of tidal flows, through installation of dikes or tide gates, is a common method to protect coastal infrastructure; to drain tidal wetlands for farming, mosquito control, and development; or to raise or manage water tables and reduce salinity for aquaculture, mosquito control, rice production, and wildfowl management. Inadvertent tidal restrictions also occur due to road, railroad and other infrastructure development, with affected wetlands landward of transportation corridors often becoming freshened and flooded due to retention of freshwater drainage from the watershed8. As a result of those many causes for complete or partial tidal restriction in tidal wetlands, inhabited and developed coastal landscapes typically contain a patchwork of unaltered tidal wetlands interspersed with drained, impounded, and partially restricted wetlands, ranging in size from <0.1 km2 to multiple km2 (Fig. 1). In many cases, if the alteration occurred decades or centuries ago, the parcels may no longer be recognized as former salt marsh by local residents or by the National Wetlands Inventory10, but rather as fresh or brackish ponds or wetlands where impounded, or, where drained, former wetland may appear as woodland, shrubland, developed land, or agricultural land.

Figure 1
Figure 1

Satellite photographs of tidally-restricted wetlands on the U.S. Atlantic coast, demonstrating a range of histories, causes and sizes: (a) The Herring River basin within the National Park Service Cape Cod National Seashore, Massachusetts. The solid black outline indicates 4 km2 of former tidal salt marsh. The estuary and marsh were diked in the first decade of the 20th century, in an unsuccessful attempt to control mosquitos. Portions of the basin are now7 drained marsh, and are characterized by acid-sulfate soil, loss of elevation, and colonization by native and non-native grass, shrub and trees. Other portions are impounded, freshened and colonized by native and non-native fresh and brackish wetland grasses. The wetland is proposed for tidal restoration, and pre-restoration carbon cycle and ecological data collections are underway. (b) An impoundment on ~0.05 km2 of former salt marsh on the south shore of Cape Cod, Massachusetts, caused by a 19th century railroad that was subsequently converted to a bicycle path. Similar incidental impoundments are common and widespread along coastal transportation corridors. (c) The Prime Hook National Wildlife Refuge in the state of Delaware. The Refuge is a 41 km2 complex of managed wetland and open water. In response to repeated storm damage, plans were in development to upgrade the artificial berm that impounded 16 km2 of freshwater impoundment (https://www.fws.gov/refuge/prime_hook/). Prior to completion of the upgrades, Hurricane Sandy breached the artificial berm in 2012 and restored saline tidal flows to the impoundment. Subsequently, deliberate salt marsh restorations were undertaken. The inadvertent breach likely resulted in significant GHG emission reductions. Maps created by georeferencing each image from Google Earth (image source credits: (a) Google, Landsat/Copernicus, copyright DigitalGlobe; (b) Google, Landsat/Copernicus; (c) Google, USGS), base map and scale bars were added using ArcGIS for desktop version 10.3.1 (http://www.esri.com), and images arranged using Adobe Photoshop CC 2017 vers. 18.0.1 (http://www.adobe.com/products/photoshop.htm).

Restriction or blockage of tidal flows dramatically alters wetlands as habitat for flora and fauna8, and further causes major changes in microbial and chemical processes in wetland soil and water7,11. Here we emphasize that those changes can also alter the direction and magnitude of greenhouse gas (GHG) exchange between the wetland and the atmosphere. The fate of soil C in wetlands, and emissions of CH4 and CO2, largely depend on water salinity and water table elevation relative to the soil surface, with temperature and plant productivity also having important influence. Under flooded or impounded conditions, soil becomes anaerobic, effectively to the soil surface. Under those conditions, if blockage of tidal flows has caused water salinity to decrease from >18 psu (~50% seawater) to <18, then a substantial increase in CH4 emission is expected to occur9. Salinity is a critical driver because seawater contains abundant sulfate ion, and where sulfate is abundant, sulfate reduction tends to outcompete methanogenesis as a metabolic pathway for microbial decomposition of organic matter. Additionally, sulfate can serve as the terminal electron acceptor in the oxidation of methane to carbon dioxide12, further reducing CH4 emissions under saline conditions. Though measurements are uncommon, CH4 emissions in inundated, low salinity portions of tidally-restricted wetlands are elevated13,14, and similar to the average of published rates for tidal wetlands at salinity <18 psu3. Drainage of both tidal and non-tidal wetlands, on the other hand, drains water from soil pore-spaces and exposes soil organic matter to air and associated oxygen, thereby promoting aerobic microbial respiration of C stocks that have accumulated over centuries to millennia, with CO2 as the primary product. Drainage-associated soil CO2 emissions are intense per unit area, and drained tidal wetlands15 and terrestrial peatlands16 are significant drivers of climate change. Drainage and restoration of terrestrial (non-tidal, freshwater) peatlands are not the focus of the present study, except as an informative comparison to GHG consequences of tidal restriction and restoration of tidal wetlands, since rewetting (re-flooding through restoration of natural hydrology) of terrestrial peatlands, to protect soil C stocks from ongoing respiration and loss as CO2 emissions, is considered to have great potential as a landuse-based climate intervention17.

To examine the intensity of GHG emissions due to tidal restriction in salt marshes, as well as the potential for emissions reductions through tidal restoration, we modeled resulting changes over time in the cumulative mass of atmospheric CO2 and CH4 emitted or sequestered, and their radiative forcing (RF, a measure of the change in energy in the global atmosphere). The concepts around changes in soil C stocks and in GHG emissions with impoundment or drainage of salt marsh, and with restoration of tidal flows, are diagramed in Fig. 2. Here, “tidal restoration” describes removal or opening of dikes, tide gates, or under-sized culverts to renew tidal water exchange between the salt marsh and coastal ocean, and thus restore natural, or nearer to natural, water level and salinity. We first compared the intensity, per unit area, of cooling (net negative RF) in natural Blue Carbon ecosystems (seagrass beds, salt marshes and mangroves) to intensity of warming due to enhanced CH4 and CO2 emissions in impounded and drained salt marsh, respectively (Fig. 3). As further context, we also modeled: climatic cooling (negative RF) from net C storage in U.S. forests; warming (positive RF) from tailpipe emissions of a typical U.S. automobile; and warming (positive RF) from enhanced CO2 emissions in drained, terrestrial peatlands. We utilized RF calculations in this study to enable comparison of the climatic impact of CO2 and CH4 emissions or emissions reductions that are sustained over a period of decades to centuries. Use of standard global warming potentials (GWP) to compare the impact of different GHG is not appropriate in the context of sustained emissions that are characteristic of ecosystems and their management4. Rather, standard GWP calculations are intended as a policy tool based on analysis of the atmospheric perturbation lifetime of a discrete pulse of GHG emission. Standard GWP calculations underestimate the climatic impact of sustained CH4 emissions or emissions reductions relative to sustained changes in CO2 emissions.

Figure 2
Figure 2

Conceptual model of carbon cycle processes and greenhouse gas flux changes in response to hydrological management in tidal wetlands. (a) In unaltered, or successfully restored, salt marsh, sulfate ion inhibits methane emission, and high rates of net CO2 uptake, soil C storage, and soil elevation gain occur. (b) Salt marsh drainage increases exposure of soil carbon stocks to oxygen, and results in a rapid rate of aerobic respiration to CO2, loss of carbon stock, and loss of elevation. (c) Impoundment commonly leads to freshening and increased water level, and those conditions are likely to cause an increase in methane emission. Effects of impoundment on soil carbon stocks and rate of soil carbon storage are not well-known, and herein rates are assumed to be similar to natural salt marsh. Diagrams created using Adobe Illustrator CC and arranged using Microsoft PowerPoint vers. 15.32. Diagram symbols were provided courtesy of the Integration and Application Network, University of Maryland Center for Environmental Science (ian.umces.edu/imagelibrary/).

Figure 3
Figure 3

Twenty-year cumulative radiative forcing (picowatt-years per m2 of earth surface) due to net effect of carbon sequestration and methane emission in typical coastal wetlands and in hydrologically-altered wetlands. Radiative forcing is calculated per hectare (104 m2) of wetland, with continuous emissions over a 20-year period, based on emission factors provided in Table 1. As described in the text, scenarios were considered based on two emission factors for carbon dioxide in drained marsh and three emission factors for methane in freshened or impounded wetlands. Those EF are shown here as separate bars. For context, emissions in wetlands are compared to terrestrial forest C sequestration and to emissions from an automobile. RF from an automobile was calculated based on twenty years of tailpipe emissions from one U.S. automobile50 at 4,690 kg CO2 y−1.

In the present study, emission factors (EF), or typical CH4 and CO2 emission rates for each ecosystem type or condition, were derived from the literature and from IPCC18 tier 1 EF (Table 1). To represent the range and variability of CH4 emissions in low salinity tidal wetlands, and therefore expected range of emissions and EF in freshened, impounded wetlands, we modeled the time course of RF using three different CH4 EF based on the median and average of published emission rates3 at salinity <18, as well as the IPCC EF for low salinity tidal wetlands. The wide range in literature estimates for CH4 emissions in low salinity tidal wetlands is likely due to variations across sites in other drivers of emissions, including water level, plant productivity and temperature. Therefore, it is important to examine a range of EF for these low salinity wetlands, to represent the diversity of emissions. For CO2 EF in drained and rewetted terrestrial and tidal wetlands, we used IPCC Tier 1 EF from drained and rewetted inland organic soils and tidal wetlands. Note that there is uncertainty regarding the length of time over which the high rate of CO2 emission will occur following drainage. The duration will depend in part on the depth of drainage and whether further, deeper drainage occurs following initial drainage19, thereby exposing new soil C to oxidation. We therefore used two CO2 EF. EF1 assumed that drainage-associated respiration had gone to completion during decades to centuries of drainage, and current and future drainage-associated CO2 emission is zero. EF2 assumed that C stocks were not depleted, and CO2 emission continued over the time period of analysis.

Table 1: Methane and carbon dioxide emission rates utilized as inputs to model radiative forcing in a series of wetland settings and restoration scenarios.

Results

Radiative Forcing per Unit Area

Results indicate that climatic warming, or positive RF, due to tidal restriction (both impoundment and drainage) is large on a per unit area basis relative to the magnitude of cooling from C sequestration in forests and in unaltered coastal wetlands (Fig. 3). For instance, over a period of 20 years following initial alteration, CH4 emissions from an impounded and freshened salt marsh can result in a cumulative RF that is a factor of ~3 to 20 greater than the magnitude of climatic cooling due to net C stock increase in continental U.S. forests and to soil C sequestration in unaltered salt marsh, and a factor of 1.5 to 10 greater than sequestration in natural mangrove. Over those 20 years, CH4 emissions from one hectare of impounded wetland are equivalent to RF from 20 years of continuous tailpipe emissions of 1.7 to 6.3 automobiles (Fig. 3).

To quantify the potential for emissions reductions through wetland management, we calculated the time-course, over decades to centuries, of cooling (negative RF) predicted to occur due to CH4 and CO2 emissions reductions resulting from tidal restoration in tidally-restricted salt marsh, based on the cumulative difference between the contrasting management action and no action scenarios (Fig. 4). We compared climatic cooling due to tidal restoration to cooling from other wetland carbon management options, including creation of new salt marsh or sea grass beds as biological C sequestration projects, and to rewetting of terrestrial peatland to cease the high rate of CO2 emissions from drained peatland soils. In general, the tidal restorations in salt marshes, to restore natural salinity and water levels, will be dramatically more effective at cooling climate than other wetland-based climate change interventions (Fig. 4). The time course of cumulative RF resulting from reduction in the rate of CH4 production in previously flooded and freshened soils (Fig. 4a,d), and reduction in CO2 production in previously drained soils (Fig. 4b,e) indicate sizeable, rapid and sustained climatic cooling. Despite the high rate of C sequestration in salt marshes and mangroves, building and planting of new marsh, mangrove or seagrass bed, provides relatively modest climatic cooling (Fig. 4c,f). Avoided CO2 emissions from rewetting of drained, terrestrial peatlands (Fig. 4c,f) can be substantial, but are partially offset by resumption of CH4 emissions in soil rewetted with fresh water20 (Table 1).

Figure 4
Figure 4

Time-course of cumulative radiative forcing modeled as the difference between emissions under scenarios of no action and of restoration (Table 1). Negative values indicate cumulative climatic cooling (picowatt-years per m2 of earth surface) per square meter of wetland restored, over a 30-year period (a,b,c) and a 200-year period (d,e,f). Methane emission reductions due to tidal restoration in impounded (flooded and freshened) former salt marsh (panels a and d) result in large reduction in RF, particularly on a timescale of decades to a century. Carbon dioxide emission reductions due to rewetting of drained salt marsh (b,e) and terrestrial peatland (c,f) also have relatively large climate benefit, but depend on assumptions regarding the period of time that the wetland would have been maintained in a continuously drained condition, and the period of time that the finite carbon stock in the drained soil would continue to respire (uncertainty regarding timeframe of benefit indicated by dashed lines; see text).

Geography of Tidally-Restricted Wetlands and Scale of Emissions

There is substantial potential for GHG reductions based on the intensity of emissions, and on the areal extent of tidally-restricted wetlands in developed landscapes. Tidally-restricted wetlands are widespread throughout the inhabited world (e.g.7,8,13,14,15,18,19,20,21,22,23,24,25,26,27,28,29,30,31). Infrastructure and activities that alter wetland and coastal hydrology have been closely associated with intensity of human development and economic activity in terrestrial and marine portions of the coastal landscape31,32, and therefore are expected to increase in the future, particularly in portions of the world that are undergoing rapid development23,31. Quantitative geographical data on tidally-restricted wetlands are limited, and indeed the global area of tidal wetlands as a whole is not well-constrained15. Thus, it is not possible to estimate global scale GHG emissions due to restrictions. As an indication of scale, we used mapped and areal data along portions of the U.S. Atlantic coast on managed impoundments and on incidental impoundments associated with transportation infrastructure. The analysis resulted in an estimate that those tidal restrictions have caused a reduction in salinity, and enhancement of CH4 emissions, in approximately 27% of tidal wetlands on the U.S. Atlantic coast (Table 2). The impoundment and freshening result in enhanced CH4 emissions in the range of 28,000 to 145,000 tonnes (t) CH4 y−1. It is worth emphasizing once again that these are sustained emissions, and thus RF due to those rates of continuous emission over a 20-year period are equivalent to 20 years of continuous emissions from 0.6 to 3.1 million automobiles (see Methods). Note that this is an estimate of CH4 emissions alone, and does not consider changes in C sequestration rate in soil, nor loss of existing soil C stocks and enhanced soil CO2 emissions, due to changes in water level and salinity. Further, the area and emissions calculations consider only the effect of freshening and impoundment of tidal wetlands, and do not include area or emissions of tidal wetlands that have been drained, unintentionally nor intentionally, for purposes such as agriculture or land development.

Table 2: Geography of tidally restricted wetlands on the Atlantic coast of the United States.

Discussion

Key features of avoided wetland methane emissions

The results of this study indicate that, despite a high rate of C sequestration in coastal wetlands, in many cases tidal restoration in salt marshes will have dramatically greater potential per unit area as a climate change intervention than the other examined ecosystem management actions. Coupled with the common, widespread occurrence of tidal restrictions, there is significant potential for GHG emissions reductions through tidal restoration in salt marshes. Though we did not address tidal restriction and restoration in mangroves, many of the processes and rates presented for salt marshes would be expected to be similar in mangroves. We further note that avoiding tidal restrictions with future coastal development would have similar and perhaps greater benefits in terms of avoided GHG emissions than the restoration scenarios examined in this study. Here we discuss several features of tidal restoration, and particularly avoided methane emissions, that highlight further advantages relative to enhanced CO2 sequestration in other landuse-based climate change interventions, due to key aspects that result in rapid, substantial, and sustained reduction in RF:

First, CH4 has a radiative efficiency, or instantaneous heat absorption capacity, that is a factor of ~73 greater than that of CO2, and a factor of ~94 greater with consideration of indirect effects of CH4 on atmospheric chemistry33. Second, reductions in both CH4 and CO2 emissions are soil microbial responses, and thus, though more data are needed, onset following restoration of natural water level and salinity is likely to be rapid14,21 relative to development of restored C storage capacity following full ecosystem recovery or development in biological sequestration projects34. In the present study, ecosystem recovery was represented as a lag of 5 years prior to onset of the full rate of C sequestration in salt marsh soil (Table 1, note k). Thus, climatic cooling due to avoided CH4 emissions is relatively large on short timescales. In contrast, in projects that promote biological C sequestration in wetland soil, or similarly in forest biomass, accrual of climate change mitigation benefits is a process of gradual accumulation on a timescale of centuries (see “New salt marsh” curve in Fig. 4c and f), due to both the gradual rate of ecosystem development following restoration, and to the modest radiative efficiency of CO2, coupled with a long atmospheric perturbation lifetime. As a result, at 15 years post restoration, for instance, negative RF (cooling) due to CH4 reduction (Fig. 4a) was a factor of 12 to 59 greater than that due to CO2 sequestration in new salt marsh (Fig. 4c).

Third, in the absence of restoration, enhanced CH4 emissions in an impounded and freshened wetland can be expected to continue indefinitely, due to a continuous supply of new organic matter substrate for methanogenesis from ongoing primary production35. Thus, although CH4 is considered to be a short-lived climate pollutant, with an atmospheric perturbation lifetime of only 12.4 years, restoration can avoid a long-term, sustained emission, and the cumulative mass of avoided CH4 and magnitude of climatic cooling will continue to increase indefinitely through time (Fig. 4d). In contrast, although avoidance of CO2 emissions by tidal restoration in drained salt marshes can have an RF in the first few decades that is as great as avoided CH4 in freshened marshes, the duration and ultimate cumulative magnitude of enhanced CO2 emissions in a drained wetland is limited by the mass of the C stock contained in soil above the level of drainage. Therefore, where wetlands have been drained for decades to centuries, the benefit of restoration may be relatively minor (Fig. 4b emission factor 1). Where pre-restoration CO2 respiration is rapid, re-wetting will be of substantial benefit (Fig. 4b emission factor 2), but the anticipated duration of that benefit will be uncertain (Fig. 4e). Based on IPCC Tier 1 marsh soil C stocks and respiration rates in drained marshes, the stock in the top meter of soil would be consumed within approximately 3 decades following drainage18.

A fourth critical feature of wetland tidal restoration as a climate change intervention is that the CH4 emissions thus avoided will have what can be referred to as “inherent permanence”: following tidal restoration and onset of reduced methane emissions, the emissions thus avoided cannot be cancelled, even if emissions were to resume in the future due to a change in ecosystem status. For instance, if a tidal restriction were re-established 30 years after restoration, emissions would resume, but the biosphere and atmosphere would maintain for extended time a deficit of heat and of GHG equivalent to 30 years of reduced emissions. Further, since GHG emissions from tidally-restricted wetlands are caused by human activity, they are anthropogenic emissions, and reducing anthropogenic wetland emissions will have an effect on climate that is equivalent to reduced emission of an equal quantity of fossil fuel GHG. Therefore, avoiding CH4 emissions as a landuse-based climate change intervention is a distinct concept from biological C sequestration projects to enhance C storage in forest or wetland biomass or soil. Biological C sequestration accumulates climate benefit at a relatively slow rate, and the C sink must persist for a century or more to have significant impact on climate (see Fig. 4f, new salt marsh example). During that time, there is a risk that the stored C will be rapidly returned to the atmosphere, through processes such as fire or ecosystem degradation and organic matter decomposition. As a result, the monetary and climate change mitigation value of biological C sequestration projects must be discounted in C markets and in trading programs to account for the “non-permanence risk” that stored C will be returned to the atmosphere36,37. To be clear, this discussion is not intended to suggest that temporary reductions in CH4 emissions will achieve effective management of climate change, but rather to highlight a distinction between avoided emissions and C sequestration.

Implications for GHG emission inventories and reduction programs

Tidal restoration to avoid CH4 emissions is a relatively new concept for Blue Carbon management, and warrants consideration within GHG markets and emission offset programs, national and local-level efforts to reduce emissions, and national GHG inventories based on IPCC guidance. Within international climate change treaty negotiations there has been increasing emphasis on co-benefits accruing from GHG mitigation actions that involve improved landscape management, including forests, soils and potentially coastal and marine ecosystems. In 2013, the International Panel on Climate Change released the 2013 Supplement to the 2006 IPCC Guidelines for National GHG Inventories: Wetlands18, to guide accounting of emissions and management of wetland soils. The Supplement includes guidance on CO2 emissions for drained wetlands, but does not consider GHG emissions from impoundments in general, nor freshened or impounded wetlands due to tidal restriction. The U.S. has recently completed the first national inventory worldwide to include coastal wetlands in a National GHG Inventory, and the UN Framework Convention on Climate Change has asked countries to provide feedback on experiences with application of the Supplement, to inform future expansion of the accounting guidance where science is available. The analysis presented here indicates that tidally restricted wetlands meet the primary criteria for inventoried ecosystems in that they are managed landscapes, with emissions and sinks of substantial magnitude and rate of change. If other countries ultimately follow suit, then inclusion of those emissions in the U.S. Inventory will promote widespread recognition and management of the issue, and justify development of CH4 EF for tidal restrictions and impoundments in IPCC guidance for GHG inventories. Note that since ecological restoration through tidal reconnection has been a widespread practice in some parts of the world, restorations that were conducted since the baseline year, typically 1990, may represent considerable, but as yet unrecognized, emissions reductions.

Wetland C and GHG management can also play a role as an additional method for reducing regional, state and local GHG emissions, including through regulatory or voluntary carbon markets or cap & trade programs. Verified Carbon Standard (VCS) has developed a Methodology for crediting tidal wetland carbon offset projects, designed for both voluntary and regulatory C markets worldwide, and within that methodology there is accommodation for avoided CH4 projects as a creditable activity37. The implications of the results presented in the present study may support continued development of concepts related to crediting of climate change mitigation through avoided CH4 emissions in wetlands, to reflect the distinctions between avoided CH4 and C sequestration projects:

First, at present, in the VCS methodology, there must be reasonable expectation that a wetlands-based, emissions offset project will have a lifetime of at least 100 years. The mandated lifetime is in response to the concept that in projects that promote C sequestration in soil or biomass, the sequestered C must be kept out of the atmosphere for an extended time to have significant climate change mitigation value. However, in the context of avoided CH4 emissions, the 100-year time frame is arbitrary, since such offset projects do not rely on storage of C in a soil or biomass reservoir, and they accrue considerable, long-term and quantifiable mitigation benefit on timeframes of decades. Given that the persistence of any project for 100 years is difficult to assure, a shortened minimum lifetime for avoided CH4 projects could increase utilization of tidal wetlands as offset projects.

Second, the application of a 100-year CH4 global warming potential of 21 to calculate carbon dioxide equivalents (CO2e) of avoided CH4 emissions underestimates the climate change mitigation value of the avoided CH4 to a substantial degree. As an example, if we were to apply that GWP value in a CO2e approach to scenarios 1 and 3 in Table 1 (tidal restoration in salt marsh to avoid CH4 emissions vs. creation and planting of new salt marsh to enhance C sequestration), the range of CH4 emission factors would result in a calculated GHG reduction benefit for avoided CH4 (scenario 1) that is similar to the C sequestration benefit in scenario 3 (scenario 1 = 0.7 to 3.5 times the value of scenario 3; not shown). The ratio would remain constant for the 100-year lifetime of the projects. That result is in contrast to the more nuanced calculation of change in radiative forcing that we noted previously in the manuscript, in which reduction in RF (cooling) due to CH4 reduction (Fig. 4a) was a factor of 12 to 59 greater than RF reduction due to CO2 sequestration in new salt marsh (Fig. 4c), at 15 years post restoration. Even at 100 years post restoration, the benefit of avoided CH4 emissions (scenario 1) remains a factor of 3 to 11 greater (Fig. 4d and f). As noted previously, a standard GWP approach is inappropriate for sustained emissions or sinks that are typical of ecosystems4. A 100-year GWP that is calculated based on simulation of a single pulse of CH4, with an atmospheric lifetime of 12.4 years, integrates 12 years of heat retention across 100 years of impact, and ignores the fact that with sustained emissions a pool of CH4 would be maintained in the atmosphere indefinitely, rather than for just 12 years.

There is ongoing debate regarding the value, in terms of addressing climate change, of reducing CH4 emissions vs CO2 emissions, and the debate is relevant to selection of a GWP value for CH4, and crediting timeframe, in C markets and offsets. Though there is not a simple answer, it is clear that aggressive reduction in emissions of CH4 and other short-lived climate pollutants (SLCP) can reduce the rate of increase in global temperature during the current century, and can reduce peak temperature if coupled with rapid elimination of CO2 emissions38. Further, it has been shown that, although CH4 is a short-lived climate pollutant, during the time that it resides in the atmosphere, a portion of the heat that it traps within the biosphere will be retained in oceans and contribute to sea level rise for several centuries39. However, if actions to reduce emissions of SLCP result in delayed reductions in CO2 emissions, then higher peak temperatures will result38. Therefore, society is likely to benefit most from simultaneous, but separate, efforts to reduce both CO2 and SLCP emissions, to manage both short-term and long-term climate change. Utilization of offsets and emissions trading programs therefore requires relative evaluation of pollutants with distinct interactions with climate. Solutions might be to limit maximum utilization of offsets in C markets, as is done in California’s Cap and Trade Program40, or to develop separate reduction and trading programs for CO2 and SLCP. For instance, the state of Massachusetts has specific emissions reduction goals for natural gas transmission41, New Zealand has separate targets for methane reductions in their Intended Nationally Determined Contribution42, and many national and subnational governments have programs to reduce methane emissions from agricultural sources43. As described in the present study, degraded tidal wetlands can be significant sources of both CO2 and CH4, while intact wetlands are generally a strong CO2 sink. Thus, wetland management can contribute to management of both short and long-lived climate pollutants.

Finally, beyond C markets and national GHG inventories, awareness of anthropogenic emissions in tidally-restricted wetlands, and the potential for mitigation through ecosystem restoration, may be of significant value to land management agencies with wetlands, tidal impoundments, and coastal infrastructure under their purview, such as the U.S. Fish and Wildlife Service, the National Park Service, local and state government entities, and private land owners. In wetlands that are intentionally managed, in many cases opening tidal restrictions would be prohibited by ongoing important land uses, or where restoration would put developed land at risk of flooding. But, increasingly, decisions are being forced, by accelerating sea level rise and increasing storm intensity, regarding whether to spend resources to preserve and upgrade aging dikes and tide gates that maintain wetlands in drained or impounded condition (for example see Fig. 1c, Prime Hook NWR). Management decisions and objectives are based in part on anticipated changes in the value of ecosystem services to society, including coastal protection, quality and type of habitat, recreational utility, and others. Reduction in GHG emissions has significant value to society, and can be highlighted as a benefit, regardless of whether the benefit is monetized. Even where opening tidal restrictions is contraindicated, there still may be potential for management of water levels and salinity at targeted times of the year, specifically for the purpose of reducing GHG emissions and promoting C sequestration, while maintaining utility of the landscape for those other uses or protecting infrastructure from flooding.

In other circumstances, the decisions may be less complicated in terms of tradeoffs of services, such as where tidal restriction is due to inadvertent, transportation-related obstruction of flow, and thus there is no current beneficial use of the impaired wetland (for example, see Fig. 1b). In some cases, ecosystem restoration through opening of tidal restrictions can be consistent with coastal resilience planning, since exposure to tidal flow will tend to promote resumption of natural accretion of wetland elevation in response to sea level rise, resulting in enhanced protection of the infrastructure landward of the wetland.

There is significant potential for C and GHG management in coastal wetlands, but continued advances are needed on the scientific basis for policy frameworks and for quantification of emission factors in response to management actions. There has been particularly little study of changes in soil and biomass C stocks, and of GHG fluxes, in response to tidal restriction and tidal restoration in salt marshes and mangroves. Further, in all areas of the world the state of knowledge is poor regarding the geographic distribution of tidal wetlands, of salinity within those wetlands, of occurrence of tidally-restricted salt marshes and mangroves, and of the potential for tidal restoration. The findings presented here therefore suggest a research agenda to study GHG emissions and the fate of carbon stocks in natural, restricted and restored wetlands, across geographic settings, as well as mapping occurrence and characteristics of tidal wetlands and tidally-restricted wetlands.

Methods

Radiative Forcing—To calculate radiative forcing, we modeled the time-course of change in the atmospheric inv entory of each GHG per square meter of wetland, based on emission rate and on rate of destruction or removal from the atmosphere. Methane was modeled as a single reservoir with atmospheric perturbation lifetime of 12.4 years, and first order removal due to oxidation to CO2. RF from management of methane emissions includes forcing from the resulting changes in both the pool of atmospheric CH4 and CO2 produced through CH4 oxidation. We calculated the mass of CH4 in the atmosphere, per square meter of wetland at time t, according to Neubauer and Megonigal4 eq. 1:

M C H 4 C , ( t ) = F C H 4 C dt+ [ M C H 4 C , ( t 1 ) × e ( d t / Τ C H 4 ) ] ,
(1)

where M C H 4 C is the mass of atmospheric CH4 (g C m−2), F C H 4 C is the emission factor (g C m−2 y−1), dt is the time step (0.2 y), and TCH4 is the atmospheric perturbation lifetime of CH4.

Mass of atmospheric CO2 was modeled based on a synthesis of models, utilizing four non-interacting reservoirs with distribution of emissions among the reservoirs and atmospheric perturbation lifetimes in the reservoirs as in Joos et al.44, Table 5 and IRF model within that publication. The mass of CO2 in the atmosphere was calculated as:

M C O 2 C , ( t ) = i = 1 4 f i ( F C O 2 C d t + [ M C O 2 C _ i , ( t 1 ) × e ( d t / Τ C O 2 _ i ) ] ,
(2)

where f i is the fraction of CO2 emissions distributed to the ith reservoir, and other terms are as defined for the CH4 model.

In both models, the instantaneous RF was calculated at each time step as the product of the mass of each gas in the atmosphere and its radiative efficiency33 (1.75 × 10−15 W m−2 (kg CO2)−1 and 1.28 × 10−13 W m−2 (kg CH4)−1 with CH4 radiative efficiency adjusted by a factor of 1.65 to account for indirect effects45. Cumulative RF is the sum of instantaneous RF over a given time period. N2O emissions were not considered due to insufficient data.

Geography—Large scale, geographic data on tidal restrictions are rare, but a large number of examples and case studies of restriction and restoration occur within the literature. As an indication of scale, approximately 50% of pre-development tidal saltmarsh area in the U.S. has been destroyed by human activities, with impoundment and drainage as prominent causes23, and a quarter of remaining coastal wetlands in the state of Massachusetts, for example, is landward of nearly 600 transportation-related tidal restrictions46,47,48,49. To estimate the scale of tidally restricted wetlands area and the proportion of existing wetlands that are restricted, we compiled published information on the occurrence along the U.S. Atlantic coast of managed wetland impoundments and incidental, full or partial impoundments caused by transportation infrastructure (Table 2). In the northeast U.S., wetland and impoundment areas landward of primarily transportation-related restrictions comprise 29% of total tidal wetland. In the southeast U.S. 10% of total tidal wetland area is diked and impounded for waterfowl and mosquito management. The tidal wetlands in the regions and states represented in the analysis comprise 55% of the total tidal wetland on the U.S. Atlantic coast, and thus are expected to be approximations of the entire coast and may be similar to many developed coastal regions elsewhere. Based on the assumption that similar levels of transportation-related and diked restrictions occur throughout the U.S. east coast, we estimate that 39% of total tidal wetland area on the U.S. Atlantic coast, ~3,800 km2, is above tidal restrictions or blockages. Montague et al.23, reported that 87% of the southeast U.S. mosquito and waterfowl impoundments included in Table 2 had salinity less than 16 psu. Thus, they are expected to be significant methane sources. It is not known what proportions of restricted wetlands related to transportation infrastructure are in freshened, drained, or relatively unaltered condition. In Massachusetts 70% of tidally restricted wetland area is described as freshwater marsh or fresh shrub, with 52% of the vegetated area colonized by invasive Phragmites australis, while the remainder remains brackish to saline marsh48,49. In general, where purposeful actions have not been taken to drain tidally-restricted wetlands, such as is the case with incidental restrictions caused by transportation infrastructure, flooding and freshening are the most likely result. Thus, we can approximate that 70% of restricted wetland area compiled here, or 2,650 km2, is in a flooded/freshened condition. To calculate minimum and maximum estimates of CH4 emission from those wetlands, we multiplied the area by the minimum and maximum annual emissions rates of CH4 from impounded wetlands (Table 1, Scenario 1, EF1 and EF3 pre-restoration minus post-restoration). For context, the annual emission rates from 2,650 km2 of freshened wetlands were compared to average annual tailpipe emissions from a U.S. automobile41, at 4,690 kg CO2 y−1, with the comparison calculated as cumulative RF over a 20-y period of continuous emissions at those rates. The resulting anthropogenic wetland emission estimate was 28,000 to 145,000 tonnes CH4 y−1, with RF equivalent to 20 years of continuous emissions from 0.6 to 3.1 million automobiles (not shown).

Data availability

The data that support the analyses presented are included in tables and in cited references. Results of radiative forcing calculations are available upon request from the lead author.

Additional Information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    Bridgham, S. D., Cadillo-Quiroz, H., Keller, J. K. & Zhuang, Q. Methane emissions from wetlands: biogeochemical, microbial, and modeling perspectives from local to global scales. Glob. Change Biol. 19, 1325–1346 (2013).

  2. 2.

    Neubauer, S. C. On the challenges of modeling the net radiative forcing of wetlands: reconsidering Mitsch et al. 2013. Landsc. Ecol. 29, 571–577 (2014).

  3. 3.

    Poffenbarger, H. J., Needelman, B. A. & Megonigal, J. P. Salinity Influence on Methane Emissions from Tidal Marshes. Wetlands 31, 831–842 (2011).

  4. 4.

    Neubauer, S. C. & Megonigal, J. P. Moving Beyond Global Warming Potentials to Quantify the Climatic Role of Ecosystems. Ecosystems 18, 1000–1013 (2015).

  5. 5.

    Chmura, G. L., Anisfeld, S. C., Cahoon, D. R. & Lynch, J. C. Global carbon sequestration in tidal, saline wetland soils. Glob. Biogeochem. Cycles 17, 1111 (2003).

  6. 6.

    Howard, J. et al. Clarifying the role of coastal and marine systems in climate mitigation. Front. Ecol. Environ. 15, 42–50 (2017).

  7. 7.

    Portnoy, J. W. Salt marsh diking and restoration: Biogeochemical implications of altered wetland hydrology. Environ. Manage. 24, 111–120 (1999).

  8. 8.

    Roman, C. T. & Burdick, D. M. (eds). Tidal Marsh Restoration: A Synthesis of Science and Management. (Island Press, 2012).

  9. 9.

    Montague, C., Zale, A. & Percival, H. Ecological Effects of Coastal Marsh Impoundments - a Review. Environ. Manage. 11, 743–756 (1987).

  10. 10.

    National Wetlands Inventory. Available at: https://www.fws.gov/wetlands/ (Accessed: 12th May 2017).

  11. 11.

    Portnoy, J. W. & Giblin, A. E. Effects of historic tidal restrictions on salt marsh sediment chemistry. Biogeochemistry 36, 275–303 (1997).

  12. 12.

    Megonigal, J. P. & Schlesinger, W. H. Methane-limited methanotrophy in tidal freshwater swamps. Glob. Biogeochem. Cycles 16, 1088 (2002).

  13. 13.

    Gatland, J. R., Santos, I. R., Maher, D. T., Duncan, T. M. & Erler, D. V. Carbon dioxide and methane emissions from an artificially drained coastal wetland during a flood: Implications for wetland global warming potential. J. Geophys. Res.-Biogeosciences 119, 1698–1716 (2014).

  14. 14.

    Helton, A. M., Bernhardt, E. S. & Fedders, A. Biogeochemical regime shifts in coastal landscapes: the contrasting effects of saltwater incursion and agricultural pollution on greenhouse gas emissions from a freshwater wetland. Biogeochemistry 120, 133–147 (2014).

  15. 15.

    Pendleton, L. et al. Estimating Global ‘Blue Carbon’ Emissions from Conversion and Degradation of Vegetated Coastal Ecosystems. Plos One 7, e43542 (2012).

  16. 16.

    Hooijer, A. et al. Current and future CO2 emissions from drained peatlands in Southeast Asia. Biogeosciences 7, 1505–1514 (2010).

  17. 17.

    Couwenberg, J., Dommain, R. & Joosten, H. Greenhouse gas fluxes from tropical peatlands in south-east Asia. Glob. Change Biol. 16, 1715–1732 (2010).

  18. 18.

    Hiraishi, T. et al. Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands. Intergovernmental Panel on Climate Change, Switzerland. Available at: Wetlands_Supplement_Entire_Report.pdf. (Accessed: 12th May 2017).

  19. 19.

    Deverel, S. J., Ingrum, T. & Leighton, D. Present-day oxidative subsidence of organic soils and mitigation in the Sacramento-San Joaquin Delta, California, USA. Hydrogeol. J. 24, 569–586 (2016).

  20. 20.

    Vanselow-Algan, M. et al. High methane emissions dominated annual greenhouse gas balances 30 years after bog rewetting. Biogeosciences 12, 4361–4371 (2015).

  21. 21.

    Konisky, R. A., Burdick, D. M., Dionne, M. & Neckles, H. A. A regional assessment of salt marsh restoration and monitoring in the Gulf of Maine. Restor. Ecol. 14, 516–525 (2006).

  22. 22.

    Zhong, Q. et al. Carbon Dioxide Fluxes and Their Environmental Control in a Reclaimed Coastal Wetland in the Yangtze Estuary. Estuaries Coasts 39, 344–362 (2016).

  23. 23.

    Kennish, M. J. Environmental threats and environmental future of estuaries. Environ. Conserv. 29, 78–107 (2002).

  24. 24.

    Crain, C. M., Gedan, K. B. & Dionne, M. Tidal restrictions and mosquito ditching in New England marshes in Human Impacts on Salt Marshes: A Global Perspective (eds Silliman, B. R., Grosholz, E. D. & Bertness, M. D.) 149–69 (University of California Press: Berkely, CA, USA 2009).

  25. 25.

    Ritter, A. F. et al. Ecological signatures of anthropogenically altered tidal exchange in estuarine ecosystems. Estuaries Coasts 31, 554–571 (2008).

  26. 26.

    Howe, A. J., Rodriguez, J. F. & Saco, P. M. Surface evolution and carbon sequestration in disturbed and undisturbed wetland soils of the Hunter estuary, southeast Australia. Estuar. Coast. Shelf Sci. 84, 75–83 (2009).

  27. 27.

    Ma, Z. et al. Rethinking China’s new great wall. Science 346, 912–914 (2014).

  28. 28.

    Lee, J. & An, S. Effect of dikes on the distribution and characteristics of Phragmites australis in temperate intertidal wetlands located in the South Sea of Korea. Ocean Sci. J. 50, 49–59 (2015).

  29. 29.

    Moreno-Valcarcel, R., Oliva-Paterna, F. J., Bevilacqua, S., Terlizzi, A. & Fernandez-Delgado, C. Long-term effects of tidal restriction on fish assemblages in east Atlantic coastal marshlands. Mar. Ecol. Prog. Ser. 543, 209–222 (2016).

  30. 30.

    Zedler, J. B. (ed.) Handbook for Restoring Tidal Wetlands. (CRC Press, 2000).

  31. 31.

    Prigent, C. et al. Changes in land surface water dynamics since the 1990s and relation to population pressure. Geophys. Res. Lett. 39, L08403 (2012).

  32. 32.

    Gittman, R. K. et al. Engineering away our natural defenses: an analysis of shoreline hardening in the US. Front. Ecol. Environ. 13, 301–307 (2015).

  33. 33.

    Myhre, G. et al. Anthropogenic and Natural Radiative Forcing. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Stocker, T. F. et al., Eds pp. 659–740)(Cambridge Univ. Press, Cambridge, 2013).

  34. 34.

    Cornell, J. A., Craft, C. B. & Megonigal, J. P. Ecosystem gas exchange across a created salt marsh chronosequence. Wetlands 27, 240–250 (2007).

  35. 35.

    Whiting, G. & Chanton, J. Primary Production Control of Methane Emission from Wetlands. Nature 364, 794–795 (1993).

  36. 36.

    Kim, M.-K., McCarl, B. A. & Murray, B. C. Permanence discounting for land-based carbon sequestration. Ecol. Econ. 64, 763–769 (2008).

  37. 37.

    Methodology for Tidal Wetland and Seagrass Restoration, v1.0 | Verified Carbon Standard. Available at: http://database.v-c-s.org/methodologies/methodology-tidal-wetland-and-seagrass-restoration-v10 (Accessed: 12th May 2017).

  38. 38.

    Shoemaker, J. K., Schrag, D. P., Molina, M. J. & Ramanathan, V. What Role for Short-Lived Climate Pollutants in Mitigation Policy? Science 342, 1323–1324 (2013).

  39. 39.

    Zickfeld, K., Solomon, S. & Gilford, D. M. Centuries of thermal sea-level rise due to anthropogenic emissions of short-lived greenhouse gases. Proc. Natl. Acad. Sci. USA 114, 657–662 (2017).

  40. 40.

    Cap-and-Trade Program | California Air Resources Board. Available at: https://www.arb.ca.gov/cc/capandtrade/capandtrade.htm (Accessed: 12th May 2017).

  41. 41.

    Massachusetts Department of Environmental Protection, Available at: http://www.mass.gov/eea/docs/dep/air/laws/ghgregs16-notice.pdf (Accessed: 12th May 2017).

  42. 42.

    United Nations Framework Convention on Climate Change, Available at: http://www4.unfccc.int/Submissions/INDC/Published%20Documents/New%20Zealand/1/New%20Zealand%20INDC%202015.pdf (Accessed: 12th May 2017).

  43. 43.

    Mandatory GHG Emissions Reduction Programs | Coalition on Agricultural Greenhouse Gases (C-AGG). Available at: http://www.c-agg.org/pages/mandatory-ghg-emissions-reduction-programs/ (Accessed: 12th May 2017).

  44. 44.

    Joos, F. et al. Carbon dioxide and climate impulse response functions for the computation of greenhouse gas metrics: a multi-model analysis. Atmospheric Chem. Phys. 13, 2793–2825 (2013).

  45. 45.

    Myhre, G. et al. Anthropogenic and Natural Radiative Forcing. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Stocker, T. F. et al., Eds Available at: www.climatechange2013.org (Accessed: 12th May 2017).

  46. 46.

    Durey, H., Durey, H., Smith, T. & Carullo, M. Restoration of tidal flow to salt marshes: The Massachusetts experience in Tidal Marsh Restoration: A Synthesis of Science and Management (eds Roman, C. T. & Burdick, D. M.) 165–172 (Island Press/Center for Resource Economics, 2012).

  47. 47.

    Wetlands Restoration and Banking Program, Boston, Massachusetts, Atlas of Tidally Restricted Marshes, North Shore Massachusetts. Available at: http://www.mass.gov/envir/massbays/pdf/moris/north_shore_tidally_restricted_marshes_atlas.pdf (Accessed: 12th May 2017).

  48. 48.

    Cape Cod Commission, Cape Cod Atlas of Tidally Restricted Salt Marshes. Available at: http://www.capecodcommission.org/resources/coastalresources/TidalAtlas.pdf (Accessed: 12th May 2017).

  49. 49.

    Costa, J., Rockwell, J. & Wilkes, S. Atlas of Tidally Restricted Salt Marshes in the Buzzards Bay Watershed, Massachusetts. Buzzards Bay Project National Estuary Program (2002). Available at: http://buzzardsbay.org/smatlasmain.htm (Accessed: 12th May 2017).

  50. 50.

    US EPA, Greenhouse Gas Emissions from a Typical Passenger Vehicle. Available at: https://www.epa.gov/greenvehicles/greenhouse-gas-emissions-typical-passenger-vehicle-0 (Accessed: 12th May 2017).

  51. 51.

    US EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2014. Available at: https://www3.epa.gov/climatechange/Downloads/ghgemissions/US-GHG-Inventory-2016-Main-Text.pdf. (Accessed: 12th May 2017)

  52. 52.

    Chaffee, C., Ferguson, W. & Cole-Ekberg, M. Salt marsh restoration in Rhode Island in Tidal Marsh Restoration: A Synthesis of Science and Management (eds Roman, C. T. & Burdick, D. M.) 157–164 (Island Press/Center for Resource Economics, 2012).

Download references

Acknowledgements

Research supported by the USGS Coastal & Marine Geology Program, USGS Land Carbon Program, and NOAA Science Collaborative grant #NA09NOS4190153. We thank the Bringing Wetlands to Market team for helpful discussions, S. Brosnahan, J. O’Keefe Suttles and K. Befus for aid with figures. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Author information

Affiliations

  1. U.S. Geological Survey, Woods Hole Coastal & Marine Science Center, 384 Woods Hole Road, Woods Hole, MA, 02543, USA

    • Kevin D. Kroeger
  2. Silvestrum Climate Associates, LLC, 150 Seminary Drive, 1E, Mill Valley, CA, 94941, USA

    • Stephen Crooks
  3. University of Rhode Island, Department of Biological Sciences, 120 Flagg Road, Kingston, RI, 02881, USA

    • Serena Moseman-Valtierra
  4. Marine Biological Laboratory, Ecosystems Center, 7 MBL Street, Woods Hole, MA, 02543, USA

    • Jianwu Tang

Authors

  1. Search for Kevin D. Kroeger in:

  2. Search for Stephen Crooks in:

  3. Search for Serena Moseman-Valtierra in:

  4. Search for Jianwu Tang in:

Contributions

K.D.K. conducted the analyses and led writing of the manuscript. S.C., S.M.V. and J.T. helped to conceive and draft the manuscript.

Competing Interests

The authors declare that they have no competing interests.

Corresponding author

Correspondence to Kevin D. Kroeger.

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Creative Commons BY

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.