National Academies Press: OpenBook

An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report (2022)

Chapter: 3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches

« Previous: 2 Environmental Trends and Indicators
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

Chapter 3
Assessing Cumulative Effects of Restoration: Current and Emerging Approaches

INTRODUCTION

It is challenging to assess the progress of ecological restoration against the backdrop of ongoing environmental change and periodic acute events. Over 50 years ago, the National Environmental Policy Act of 1969 (NEPA) offered one approach—considering the impacts of a project or multiple projects compared to a world without these efforts (the “no action alternative” or “future without project”). This concept of a future without action can also be applied to projects that have benefits—such as restoration. This method for environmental planning develops a future vision of “new” baseline conditions.

Based on understanding of environmental trends (see Chapter 2), the future Gulf of Mexico (GoM) Coast will be substantially different than it is today, with or without restoration actions. Restoration actions taken now will continue to affect the GoM Coast, and environmental changes and interacting stressors have the potential to confound future assessments of the effectiveness of these restoration efforts (Hobbs and Norton, 1996; Manning et al., 2006; Moreno-Mateos et al., 2020; Palmer et al., 2016). Using projected “future without project” environmental condition as the appropriate baseline for comparison with a future that includes large-scale restoration efforts instead of today’s conditions, which are changing rapidly, can be quite valuable. Modeling related to this type of approach has started to occur in the GoM region (e.g., Meselhe et al., 2022), but it is not widespread.

Large-scale restoration in coastal watersheds, such as parts of the Greater Everglades,1 the Louisiana Coast,2 Tampa Bay,3 Galveston Bay,4 Mobile Bay,5 and Mississippi Sound6 are very involved undertakings. As regional restoration actions, such as those associated with Deepwater Horizon (DWH) settlements, grow in number and complexity, scientists and resource managers charged with restoring large ecosystems, including those outside the Gulf (e.g., Allan et al., 2013; Ortiz et al., 2018), are finding that the cumulative impacts of multiple environmental stressors need to be considered when assessing restoration success.

Restoration itself can also produce cumulative effects that interact with stressor impacts (Diefenderfer et al., 2021). Further, the effects of restoration may be beneficial, or undesirable and unplanned (Seddon et al., 2021). Failing to consider cumulative impacts of stressors may result in ecological surprises, which are the unanticipated behaviors of ecosystems. The relative occurrence of ecological surprises can increase as an ecosystem’s capacity to absorb impacts diminishes (Filbee-Dexter et al., 2017; Paine et al., 1998).

___________________

1 See https://www.nps.gov/ever/learn/nature/cerp.htm.

2 See https://coastal.la.gov/our-plan/2017-coastal-master-plan/.

3 See https://tbep.org/.

4 See https://gbep.texas.gov/.

5 See https://www.mobilebaynep.com/assets/pdf/FINAL-CCMP-11.25.2019.pdf.

6 See https://www.mdeq.ms.gov/wp-content/uploads/2017/09/2016-Addendum-FINAL-10.31.2016.pdf.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

As watershed- and estuary-scale restoration efforts have matured, the cumulative benefits of multiple restoration efforts of diverse types have helped to counteract the negative impacts of multiple stressors (Beck et al., 2019; CERP, 2014; Côté et al., 2016; Diefenderfer et al., 2016). Restoration and management strategies have traditionally focused on singular objectives, such as improving water quality or preventing erosion on a local or site-specific scale (Daoust et al., 2014), but it is now understood that achieving multiple restoration objectives—such as improving ecosystem structure and function, and diversifying and maximizing ecosystem services—requires a systematic and multidisciplinary approach (Diefenderfer et al., 2003; Gann et al., 2019; Hodgson et al., 2019; Moreno-Mateos and Comin, 2010; Neeson et al., 2015; Thom et al., 2011).

Since the passage of NEPA, the term “cumulative effects” has generally been defined as the collective impact of past, present, and future human activities on the environment (Spaling and Smit, 1993). The definition typically has a negative connotation because of a history of research documenting interacting human-related stressors and greatly declining ecosystem function (Darling and Côté, 2008; Halpern et al., 2015; Lotze et al., 2006; Luoma et al., 2001). However, the concept of the cumulative effects of restoration (see Box 3.1) has a positive connotation, in that the collective effects of multiple activities may contribute to a net positive change in ecosystem form or function. The approaches to evaluating the cumulative effects of large-scale restoration that have been tried include spatial analysis of big data, specialized indices, and lines of evidence (Beck et al., 2019; Diefenderfer et al., 2016; Raposa et al., 2018). Large-scale restoration efforts, consisting of formal or coordinated projects, have developed methods according to their goals and objectives and have been guided by the attributes of specific places envisioned for restoration (Achete et al., 2017; Allan et al., 2013; Konisky et al., 2006; NASEM, 2021). Overall, there is a lack of consensus about a standardized approach to evaluating cumulative effects of restoration, although there have been many calls for developing such an approach (Fischer et al., 2021; Jones et al., 2018; Love et al., 2017).

On the Gulf Coast, for example, after the DWH oil spill, extensive stakeholder engagement in the GoM identified appropriate aims for the use of settlement funds (Mabus, 2010; Walker et al., 2012). Much of the focus of Gulf Coast environmental restoration to date is on taking an ecosystem approach to the recovery of habitat conditions and associated species (DWH NRDA Trustees, 2016; RESTORE Council, 2016; NFWF, 2020). In other regions with similar aims, ecosystem restoration activities have also included intentionally facilitating the synergistic interactions of species (Eger et al., 2020; Halpern et al., 2007) or habitats (Sobocinski and Latour, 2015). Such holistic approaches present opportunities to achieve cumulative effects larger than the sum of the parts (i.e., synergistic effects). Many cumulative effects of environmental management are unplanned and/or uncontrolled (Filbee-Dexter et al., 2017), and in the worst case have negative outcomes relative to managers’ aims for improved functions benefiting species and ecosystems. In some cases, negative outcomes may be addressed with well-coordinated adaptive management involving program managers, restoration practitioners, and research scientists (Littles et al., 2022; Wilber and Bass, 1998).

Approaches to assessing “cause and effect” between multiple restoration and associated management actions and their outcomes at the ecosystem scale and larger are the subject of the remainder of this chapter.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

Recognizing that assessment resources are limited, for instances in which relatively few or disconnected projects are being implemented in a given area (thus increasing the likelihood of the restoration signal being lost in the noise of background stressors), or in cases of restoration especially geared toward learning and experimentation (e.g., pilot projects), the evaluation of cumulative effects may be less important. For all other restoration, the approaches are described in this chapter, which includes:

  • a detailed overview of antagonistic and synergistic effects of restoration actions;
  • a description of modes and pathways of several types of cumulative effects;
  • the use of hypotheses summarized by ecosystem conceptual models;
  • an introduction to multiple lines of evidence and causal criteria frameworks and in turn, a description of the various tools needed to develop multiple lines of evidence;
  • reflections on restoration planning and endpoints, including constraints; and
  • a case-study discussion of cumulative effects assessment in the annually occurring GoM hypoxic zone.

ANTAGONISM AND SYNERGISM IN RESTORATION EFFORTS

Diverse pressures in estuarine and coastal waters, both natural and anthropogenic, can generate multiple stresses on an ecosystem’s structure and function (O’Gorman et al., 2012). The effect of those multiple stressors can be additive (equal to the sum of their individual effects), synergistic (greater than the sum of their individual effects), or antagonistic (less than the sum of their individual effects) (Breitburg and Riedel, 2005). Antagonistic, additive, and synergistic stressors can be judged to be either beneficial or detrimental relative to program goals and objectives (Côté et al., 2016; Piggott et al., 2015). The cumulative effects of restoration efforts may also be additive, synergistic, or antagonistic and similarly judged beneficial or detrimental depending on program goals and objectives.

The idea of synergistic and antagonistic effects of multiple stressors in ecological systems and ecosystem management is well established. For example, Crain et al. (2008) synthesized 171 studies that manipulated two or more stressors in marine and coastal systems and concluded that the more stressors there were, the greater the need to account for complex interactions in both ecological studies and conservation actions. Teichert et al. (2016) evaluated the impact of nine stressor categories on fish ecology in 90 estuaries and concluded that targeting mitigation of synergistic stressors needed to be a restoration priority. In their meta-analysis of multiple stressors on seagrasses, Stockbridge et al. (2020) emphasized that understanding and accurately predicting the complex nature of stressor interaction is important in conservation, concluding that the focus needs to be on mitigating those stressors for which the greatest benefit is derived.

Antagonistic interactions may be less common than synergies, and they appear to be less understood or reported, but they are equally relevant to habitat restoration as are more additive effects (Côté et al., 2016). Understanding complex interactions at both species and community levels has been shown to enhance the effectiveness of restoration efforts in numerous habitats, from salt marshes to seagrass meadows to mangrove forests and coral reefs (Eger et al., 2020; Renzi et al., 2019; Silliman et al., 2015). Figure 3.1 shows how such complex interactions could take place among common restoration communities of the Gulf Coast, using nutrient input reduction as an example. The figure also shows how external sources affecting the four restoration communities (many of which are discussed in Chapter 2) are often less controllable than stressors local to the communities.

As discussed in the remainder of this section, exploring how to make use of ecological synergies and avoid antagonistic interactions as part of restoration efforts could improve benefits and efficacy of their implementation. This effort may also avoid costly ecological surprises following restoration investments. When assessing the impacts of synergistic or antagonistic effects of restoration activities, it is useful to separately consider two different types of restoration that occur on vastly dissimilar spatial scales—habitat and watershed.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Image
FIGURE 3.1. Chronic and acute inputs, restoration communities, and synergistic and antagonistic interactions. A box and arrow diagram showing likely and potential interactions, both synergistic (+) and antagonistic (–) between commonly used restoration communities (shown as five navy blue boxes) and two categories of inputs to these communities (Chronic and Acute). Some common antagonistic and synergistic interactions among restoration communities are also indicated. The diagram indicates important effects occurring in restoration communities from both external and often less controllable sources, as well as from interactions among restoration types. WWTP refers to wastewater treatment plants; H20 refers to water loads; SLR refers to sea level rise; SAV refers to submerged aquatic vegetation; HABs refers to harmful algal blooms.

Coastal Habitat–Scale Synergism and Antagonism

The restoration of productive coastal nursery habitats—such as oyster reefs, salt marshes, and seagrass meadows—is common. Such restoration, however, remains experimental relative to local acute and chronic inputs and still carries a risk of failure. Figure 3.1 shows some common antagonistic and synergistic interactions among restoration communities, including important effects occurring in restoration communities from both external and often less controllable sources, as well as from interactions among restoration types. Here the scale is many orders of magnitude less than that of entire watersheds, likely on the order of square meters to several hectares.

There is accumulating evidence that locating nursery habitats in close proximity to one another may positively enhance secondary productivity relative to similar habitats restored at greater distances. This has been documented most clearly for fish, shrimp, and crabs, which, because of their commercial importance, have been frequently studied. One early example is the enhanced biomass of such species in seagrass meadows adjacent to North Carolina salt marshes (Irlandi and Crawford, 1997), but a number of other examples in different locations also exist (e.g., Berkström et al., 2012; Olson et al., 2019; Sobocinski and Latour, 2015). To understand how this can occur, Gilby et al. (2018) and others have shown that many fish move daily, or with tidal cycles, between coastal nursery habitats, such as marshes, mangroves, seagrasses, and/or reefs (cf. Boström et al., 2011; Potter, et al., 2015; Olds et al., 2017). The degree to which adjacent habitats can enhance productivity is affected by the length of time in which the habitats become dry during ebb tides (Grabowski et al., 2005; Peterson et al., 2003). It is also known that better-connected ecosystems often support more fish than those that are isolated (Gilby et al., 2021; Nagelkerken et al., 2015; Olds et al., 2017) (Figure 3.2).

However, restoring habitats in close proximity could result in functional redundancy such that the combined nursery habitat benefits sum to less than those expected of the two separate habitats (Geraldi et al., 2009; McDonald et al., 2016). To resolve this issue, focused research on critical uncertainties would be needed. Functional redundancy is an example of antagonism because each hectare restored would produce less than the last. At present, there is more evidence for synergistic than antagonistic effects of adjacent

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

nursery habitats. However, the extent to which synergism or functional redundancy occurs in adjacent nursery habitats, whether natural or restored, is a topic requiring more study and one with broad implications for developing the spatial arrangement of restored coastal habitats (Barnett and Belote, 2021).

The potential for synergistic or antagonistic interactions has implications for designing the spatial configuration of large habitat restoration projects or suites of projects. Even without considering biological effects, hydrology itself is characterized by nonlinear processes (Allan, 2004). For restoration, this is expressed by such examples as the synergistic effects of dike breaching on floodplain hydrology and consequences of the spatial position of dike breaches along the tidal–fluvial gradient, shown by Diefenderfer et al. (2012). As the spatial arrangement of restoration sites is increasingly being considered under planning processes for environmental management (Gilby et al., 2018; Lester et al., 2020; Lin and Kleiss, 2007), rigorous testing of the effects of different spatial combinations of nursery habitat restoration and those in the Gulf’s different tidal regimes is now possible.

Estuary- and Watershed-Scale Synergism and Antagonism

Watershed alterations at very large scales can be expected to produce a huge number of both direct and indirect effects. Alterations of freshwater flow are one type of watershed alteration that can alter salinity regimes, nutrient delivery, water clarities, and sediment deposition rates and produce many changes in the receiving waters and their ecologies (Carle et al., 2020; Dorado et al., 2015). This type of watershed alteration has been proposed in Louisiana as a means of delivering sediments to rapidly eroding coastal wetlands, aiming to delay the loss of productive wetlands to sea level rise and land subsidence. Many additional changes occur as the diverted freshwater alters salinities, temperature, and other water quality parameters at both the origin and the destination of rerouted waters, with potentially cascading changes for the functioning of resident flora and fauna (see Figure 3.3 and Box 3.2).

Image
FIGURE 3.2. Example of synergism in coastal restoration. Fish move among complex habitats such as seagrass meadows and salt marshes in coastal seascapes (dark grey arrows). By restoring these habitats in close proximity to one another we might improve the habitat values, productivity and the carrying capacity of coastal seascapes for fish and fisheries (light grey arrows and circles). Symbols courtesy of the Integration and Application Network, ian.umces.edu/symbols. SOURCE: after Gilby et al. 2018, Figure 4.
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

Also at the watershed scale, reducing nutrient inputs via upgrades to wastewater treatment facilities and better control of non-point-source inputs from agriculture and urban landscapes can produce large changes in downstream environments by reducing algal productivity and increasing water clarity, which can facilitate the increase of submerged vegetation and its rich floral and faunal associates (shown in Figure 3.3). As described in Chapter 4, documented examples of large-scale restoration with these types of water quality and habitat goals include Tampa and Sarasota bays in Florida, and Galveston Bay in Texas, where reducing nutrient inputs led to increased light availability and seagrass extent. In these Gulf examples, multiple nutrient reduction projects implemented over several decades are associated with decreased algae concentrations (as measured by chlorophyll a concentration), increased water clarity, and resulting increases in seagrass acreage. Furthermore, net-beneficial interactions between nutrient reduction projects and habitat restoration projects were documented in Tampa Bay (Beck et al., 2019).

Although improved conditions associated with implementation of multiple restoration efforts have been observed for some examples at the estuary/watershed scale as described above, no documented examples of antagonistic and/or synergistic effects that may have contributed toward observed improvements in estuary-wide conditions in the GoM were identified. Restoration and management programs that were initiated years ago were not typically designed to detect synergistic or antagonistic effects, resulting in an information and data gap. At the project scale, monitoring and reporting key parameters using comparable methods would allow combining monitoring data for larger-scale analysis (for parameters, see, e.g., NASEM, 2017; DWH NRDA Trustees, 2019; RESTORE Council, 2021a).

ASSESSING THE CUMULATIVE EFFECTS OF RESTORATION

Why are cumulative effects of large-scale restoration efforts so difficult to measure and quantify? Key factors include how the magnitude of changes compares to the sensitivity of the detection method and how biological and environmental conditions can dampen outcomes (CEQ, 1997). In their recent paper on advancing understanding of the cumulative effects of large-scale restoration, Diefenderfer et al. (2021) provide a framework for assessing landscape/estuary-scale restoration progress despite inherent challenges to detecting them; in doing so, they provide an approach to also improve ecosystem outcomes beyond what might have been possible to achieve with independent, site-scale projects. To develop their cumulative effects framework, the authors modified the traditional stressor-based framework of characterizing cumulative effects discussed previously (Chapter 1) (CEQ, 1997) to identify cumulative restoration benefits. This committee has modified their work with examples from the GoM, which are detailed in the next section.

Modes of Cumulative Effects: Systemic, Spatial, and Temporal Effects

Table 3.1 presents the systemic, spatial, and temporal “modes” of cumulative effects, which categorize the general ways cumulative effects are expressed in ecosystems; the table is based on Diefenderfer et al. (2021) and modified with examples from the GoM. This table also includes the various “pathways” associated with each main mode:

  • Systemic cumulative effects: The systemic approach to realizing cumulative effect benefits in large-scale restoration includes three means of accruing ecological benefits identified in the ecological literature: compounding or cascading, triggers and thresholds, and indirect effects.
  • Spatial cumulative effects: Spatial approaches recognize changing spatial patterns of populations, ecosystems, and landscapes, and the cross-boundary and space crowding effects often present in complex systems and influencing habitat restoration for threatened and endangered species.
  • Temporal cumulative effects: Temporality is built into accruing cumulative restoration benefits following restorative actions intended to catalyze natural processes to advance restoration, recognizing time lags and/or the opposite—time crowding—will occur.
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Image
Figure 3.3. A logic-flow diagram summarizing five altered effect pathways associated with enhanced freshwater inflows to Gulf estuaries. See Box 3.2 for a detailed explanation of each pathway. SOURCE: Committee.
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

Table 3.1 can be used for a variety of purposes, such as considering all pathways by which long-term trends, acute events, and stressors may affect a restoration project or projects—and in turn also considering how multiple restoration efforts will affect each other via one of these pathways.

The Influence of Hot Spots, Hot Moments, and Ecosystem Control Points on Cumulative Effects

Given the importance of spatial patterns and temporal dynamics in affecting the cumulative effects of restoration (as shown in Table 3.1), the concept of hot spots and hot moments (HSHMs) is also relevant for cumulative effects assessment. This concept was originally proposed by McClain et al. (2003),who described spikes in rates and reactions in elemental cycling in biogeochemical processes, such as denitrification (Groffman et al., 2009), over space (hot spots) and time (hot moments). The applications of HSHMs in the field of ecosystem restoration to date have been limited even though HSHMs provide an opportunity to increase the benefits gained through restoration projects and programs, given that the potential returns of all locations are not equal in light of events such as natural disasters (e.g., hurricanes).

The HSHM concept was used by Kannenberg et al. (2020) who found that hot moments in ecosystems’ gross primary production (GPP) constituted a significant percentage (up to 12 percent of the annual budget) of carbon (C) assimilation. HSHM dynamics occur along coastal interfaces, including the Gulf Coast, and have the potential to accelerate—or, conversely, limit—process-based cumulative effects of restoration (e.g., biogeochemical reaction rates, carbon storage, decomposition) (Ward et al., 2020). The “blue carbon” or global carbon sink functions of mangroves, salt marshes, and seagrasses are thereby modified by HSHMs (Bertram et al., 2021). Bernhardt et al. (2017), acknowledging that spatial and temporal aspects of spots and moments almost always co-occur, developed the concept of ecosystem control points. The use of HSHMs and ecosystem control points for evaluation of post-DWH restoration of the GoM coast offers an opportunity for synergistic knowledge development. Their applicability is supported by observations to date of stressor effects in Chesapeake Bay and hypoxia (discussed below). Table 3.2 provides a short description of HSHMs and related GoM examples.

Because all three previously identified modes of cumulative effects (in Table 3.1) may occur simultaneously, this committee introduces here a new mode of cumulative effects:

  • Spatiotemporal-topological cumulative effects: The spatiotemporal-topological approach accepts that systemic, spatial, and temporal cumulative effects modes can and perhaps often occur simultaneously, as illustrated by the three modes through which they are expressed: (1) hot spots, places with anomalous ecosystem functions; (2) hot moments, times of anomalous ecosystem functions, recurring or not; and (3) ecosystem control points, which are combined “spot moments.” The topological reference refers to the spatial patterns of these HSHMs and ecosystem control points, and their adjacency or feedback effects across ecosystems (Bernhardt et al., 2017). Acute events (see Chapter 2), while not synonymous with hot moments, may be causally linked to their expression.

In general, identifying HSHMs of ecosystem functions and services during the course of a restoration project could provide important clues about the environmental or climate drivers or stressors associated with these temporary events (Kannenberg et al., 2020).

The four types of ecosystem control points or integrated spot moments that are particularly important to ecosystem dynamics, based on the biogeochemical topography of a landscape, are termed permanent, activated, export, and transport (Bernhardt et al., 2017); see Table 3.2 for short descriptions and GoM examples. A variety of models, from conceptual to machine learning models, can be implemented to understand the drivers of the control points.

Measuring or monitoring these control points is likely to generate the data necessary to improve model predictions for event timing and magnitude, allowing restoration planners and managers to take advantage of beneficial outcomes and avoid deleterious effects. The contribution of these HSHMs and ecosystem

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

TABLE 3.1. Main Characteristics of the Systemic, Spatial, and Temporal Modes of Cumulative Eeffects with Gulf Coast Examples

  Cumulative Impacts of Stressors (NEPA Perspective, Negative Impact) Cumulative Effects of Ecosystem Restoration and Management (Positive or Beneficial Effect)
Modes of Cumulative Effects Cumulative Pathway Cumulative Impacts of Environmental Stressors (Brief Description) Gulf Coast Examples of Cumulative Impacts of Stressors or Degradation Cumulative Effects of Large-Scale Ecosystem Restoration and Management (positive or beneficial) Gulf Coast Examples of Cumulative Effects of Large-Scale Ecosystem Restoration and Management (positive or beneficial)
Systemic Compounding Effects arising from multiple sources or pathways Reef-fish food webs post-DWH (Chagaris et al., 2020) Mobile Bay habitat alterations fisheries—blue crab (Jordan et al., 2008) In ecosystems altered by restoration, multiple internal or external drivers and stressors produce linear or nonlinear, antagonistic or synergistic effects and feedback Tampa Bay and Galveston Bay improvements in water quality (Beck et al., 2020; Greening et al., 2014; HARC, 2020), hydrogeomorphic effects on Everglades wading bird habitat function (Beerens et al., 2015; Pearlstine et al., 2020)
Triggers and Thresholds Fundamental changes in system behavior or structure Marsh edge erosion post-DWH Silliman et al., 2016) Sea level rise and coastal storm thresholds for fish, wildlife, and plant species (Powell et al., 2017) Thresholds are points in restoration response functions at which small changes in drivers or stressors or sudden changes in state variables yield abrupt shifts between alternate ecosystems states; triggers are environmental drivers or stressors that produce nonlinear system-state responses Tampa Bay and Florida Peninsula, threshold for light for seagrass (Choice et al., 2014; Dixon, 2000)
Indirect Secondary effects Clean-up activities post–oil spill negatively impact habitat (NRC, 2013) Restoring physical processes has biological effects, often including linkages between primary and secondary production Oyster reef restoration aids seagrass establishment (Sharma et al., 2016)
Spatial Landscape pattern Change in landscape pattern Changes in landscape patterns of Pensacola estuarine drainage area (Yang and Liu, 2005) Reduced fragmentation, increased patch size, and restored connectivity and configuration influence ecosystem process and population dynamics Connecting coastal nearshore and offshore habitats in GoM (Peterson et al., 2020)
Cross boundary Effects occur away from the source Nitrogen export from corn belt to GoM fosters hypoxia (McLellan et al., 2015) Restoration influences system states or processes outside of restored sites, including interactions between restoration sites Export from upper Mississippi River Basin (Robertson and Saad, 2021)
Space crowding High spatial density of effects Nonpoint source phosphorus management in Florida coastal waters (Yang and Toor, 2018) Multiple restoration projects are implemented within the same geographic domain, with overlapping areas of influence and interaction Mobile Bay or landbridge in Louisiana (Gregory Steyer, presentation to the committee, November 10, 2020)
Temporal Time lags Delayed effects DWH impacts on coastal and nearshore fisheries—long-term implications (Murawski et al., 2021) Important interactions and biota appear long after restoration alters drivers, stressors, or components as the system adapts of influence and interaction Proposed sewage treatment plant and septic system upgrades (Alabama Gulf Coast Recovery Council, 2019)
Time crowding Frequent and repetitive effects Pulsed inflow of diverted water timed according to river stage (Day et al., 2009; Gledhill et al., 2020) The frequency or duration of restoration actions affects the ecosystem, or restoration alters the timing of stressors Hydrological variability, including pulse events (Liu et al., 2021a; Montagna et al., 2018)

NOTE: The cumulative effects of stressors are shown together with the corresponding potential or actual cumulative effects of large-scale ecosystem restoration. Table adapted from Diefenderfer et al. (2021), which in turn used definitions from CEQ (1997). Gulf Coast examples of multistressor impacts and cumulative restoration effects were identified and developed by this committee.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

TABLE 3.2. Conceptual Application of the HSHM and Ecosystem Control Points Paradigms to Disturbance Processes, Stressors, and Restoration on the GoM Coast

Cumulative Pathway Short Description Gulf Coast Natural Disturbance Process Examples Gulf Coast Anthropogenic Stressor Gulf Coast Restoration Examples
         
Hot Spots Spatial, “patchy” Kannenberg et al. (2020) Marsh dieback, windfalls, marsh fires Fluid withdrawal for hydrocarbons that causes subsidence hotpots and inundation hotspots, marsh fires Marsh creation, landbridge and barrier island restoration
Hot Moments Temporal, “flashy” Kannenberg et al. (2020) Riverine floods, hurricane-related wind, precipitation, and storm surges, marsh fires Bonnet Carré release into Mississippi Sound, COVID-19 lockdown (anthropause) effect on water quality, marsh fires Peak freshwater flow through manmade diversions and siphons delivered to marshes
         
Permanent Ecosystem Control Points Sustained high biogeochemical rates relative to landscape Describes all tidal wetland channels Freshwater withdrawal from rivers limiting freshwater contributions to the coast Shoreline stabilization
Activated Ecosystem Control Points High transformation rates when conditions are optimized The receiving area of freshwater floods Denitrification accelerates as a result of warming temperatures Oyster reef restoration, species-specific habitat creation (e.g., ponding)
Export Ecosystem Control Points High accumulation capacity, with threshold for high export Natural crevasses of the river Failure of retention systems causes release of nutrients or harmful compounds (e.g., Lake Okeechobee) Man-made river diversions at the source
Transport Ecosystem Control Points High transport capacity contributes disproportionately River-borne sediments contribute to marsh Marine-derived sediment and nutrient deposition Excess nutrients can stress marsh health, climate change–driven saltwater intrusion Nutrient regulation through sewage treatment

control points may not be trivial during the course of a restoration trajectory, and thus, they may need to be included while analyzing the effectiveness of restoration projects across a landscape or seascape (Kannenberg et al., 2020). Further, developing a better understanding of the combination of drivers and stressors responsible for triggering HSHMs and ecosystem control points for a particular ecosystem could help to incorporate them in a predictive modeling framework for assessing the cumulative effect of restoration projects. HSHMs and control points can be used to help address risk in projects during cost-benefit or other aspects of planning.

In summary, control points and HSHMs allow restoration planners to tailor plans with the aim of achieving beneficial cumulative effects while avoiding harmful ones. If they are ignored at the planning stage of restoration projects or programs, outcomes may differ by orders of magnitude from predictions (Petersen et al., 2008). Furthermore, understanding HSHMs is a foundation for the effective development of hypotheses and the design of monitoring systems to measure cumulative effects. HSHM and control points, while localized, can also have landscape-scale effects and measurable signals in their vicinity can overwhelm background-level annual averages over much larger areas. Awareness of HSHMs and considering them in planning efforts can also assist with avoiding ecological surprises.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

THE ROLE OF CONCEPTUAL MODELS IN DEVELOPING HYPOTHESES

The ecological restoration literature includes many experiments conducted at small scales for the restoration of particular plant communities or the recovery of habitat for a species of interest (e.g., Martin et al., 2021; McDonald et al., 2016). However, relatively few coastal restoration projects are designed to test hypotheses or monitored long enough to do so (Waltham et al., 2021). Many restoration projects are initiated on the basis of a perceived shared understanding of what works, which may be based on unrelated locales or conditions. Systematic experimental restoration with replication is lacking (Howe and Martínez-Garza, 2014). In large-scale public restoration, modeling alternative restoration actions often substitutes for on-the-ground experimentation (Buenau et al., 2014; Diefenderfer et al., 2012).

The aim of evaluating progress resulting from large-scale ecosystem restoration is to determine whether, relative to chronic background trends and acute events (discussed in Chapter 2), the cumulative restoration effects are on track to meet goals or not. If they are not, the inevitable question is, “Why not?” To help inform adaptive management of the trajectory, it can be valuable to consult the conceptual model that guided restoration design.

Conceptual Models

Conceptual models are graphical representations of interrelationships between drivers, pressures, stressors, restoration actions, and ecosystem response, based on one or a series of hypotheses (Gentile et al., 2001; Suter, 1999). They are often used to represent understanding of the current and future states of the ecosystem, which are crucial for determining restoration project priorities and assessing future projections. Preparing a conceptual model can enhance understanding of the current state of the ecosystem and raise questions about underlying assumptions. The initial understanding of the ecosystem at the time of the restoration design will subsequently be tested by restoration actions (Brudvig and Catano, 2021).

The use of conceptual models is in longstanding practice in environmental management and was previously recommended for effective post-DWH GoM monitoring (NASEM, 2017). The Monitoring and Adaptive Management Procedures and Guidelines Manual (DWH NRDA Trustees, 2019) provides guidance on the use of conceptual models at the project level. The applicability of those procedures and guidelines to evaluation of cumulative effects of large-scale restoration bears emphasis here. For example, the model in Figure 3.1 (above) represents current expectations for potential synergistic and antagonistic effects as a result of interactions among habitats being restored—knowledge that may be tested by future studies and assessment of the effectiveness of future actions. Aronson et al. (2017) posit that expanding restoration to the landscape scale needs conceptual tools that consider not just ecological processes, but also policy and management activities that could either support or hinder the realization of restoration goals. The experience of coastal and estuarine systems with restoration programs in progress indicates the value of beginning simple modeling efforts—such as conceptual models—early on, before moving to more complex versions (Brudvig and Catano, 2021). More complexity is warranted when understanding based on related research programs, data availability from comprehensive monitoring, and management questions have been developed and evolved.

Conceptual models can help bridge the disconnect between a vision for restoration of a target species or geography and the reality of implementation that happens one project at a time across long periods of time (DiGennaro et al., 2012; Fischenich, 2008). Conceptual models are most helpful when developed in the early stages of a restoration project or program by systematically listing all the active variables, drivers, and stressors within the broader ecosystem. As new monitoring data are collected at different stages of the project or program, refining the conceptual model may become necessary (Olander et al., 2018).

Failure to achieve coastal restoration goals may occur for many reasons, including lack of understanding of initial conditions, design flaws, implementation challenges, and/or unexpected environmental changes. For example, a recent synthesis for tidal marsh restoration found that tidal marsh restoration is still fundamentally an experimental activity that is unique to each site, without universal standards (Waltham

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

et al., 2021). Key environmental stressors of geographic variation among salt marshes include seascape configuration, the length of time when a habitat is wet or dry, riverine input, salinity, sediment supply and geomorphology, climatic region, and vegetation composition (Ziegler et al., 2021).

The Disaster-Pressure State-Ecosystem Services-Response-Health conceptual model developed by Sandifer et al. (2017) assesses ecosystem services and human health outcomes in the GoM after disasters that produce large-scale ecosystem injuries. The model connects disaster events (e.g., hurricanes, oil spills) to ecosystem pressure, states, services, and, ultimately, an array of responses. Conceptual models are also recommended for inclusion in all Natural Resource Damage Assessment (NRDA) Monitoring and Adaptive Management Plans.7 However, O’Farrell et al. (2017) concluded that these conceptual models could be used effectively across ecosystems and projects and shared among restoration managers when prepared in a uniform framework. One such example of an explicit uniform framework is the Delta Regional Ecosystem Restoration Implementation Plan (DRERIP), which links a conceptual model to an action evaluation plan and a decision support tool (DiGennaro et al., 2012).

O’Farrell et al. (2017) also reviewed the current status of ecosystem-based fishery management modeling efforts for the GoM and identified 45 models in six classes, ranging from simple conceptual and qualitative models to models considering bottom-up and top-down interactions, whether biogeochemical based or coupled hydrodynamic and ecological model platforms. Other classes were extensions of single-species models, dynamic multispecies models, and aggregated whole ecosystem models. Another example is the Gulf of Mexico Integrated Ecosystem Assessment program of the National Oceanic and Atmospheric Administration (NOAA), which uses a simple conceptual framework to guide indicator development.8 At the estuary/watershed scale, Tampa Bay scientists and resource managers have adopted a nutrient management strategy based on a conceptual model linking required light levels for sustaining healthy seagrass meadows with chlorophyll a concentrations and nitrogen loading levels (Greening et al., 2014; discussed further in Chapter 4). At regional and larger levels, the Chesapeake Bay Program and its partners have used conceptual models throughout the development and implementation of their action plan to help guide research, project implementation, and assessment of cumulative effects, as discussed in the following section (Linker et al., 2002, 2013; Shenk et al., 2012, 2013).

A potential tool for aiding the development of conceptual models is to develop a solid understanding of the restoration area by considering an ecohydrogeomorphic classification of the GoM. As described in Brinson (1993), classification simplifies the concept of a wetland or aquatic system, recognizing that while each potential area may be unique, such areas can also be placed into categories that share functional properties. The result of reducing the apparent complexity allows for improved communication among researchers and managers. Ecohydrogeomorphic classes also clarify the relationship between ecosystem structure and function. This allows for the comparison of systems with a set of common characteristics, and also acknowledges that some cross-boundary cooperation will be needed to generate useful assessments. There are several classifications in the GoM region that may be considered. For example, Pendleton et al. (2010) divided the U.S. Gulf Coast into eight regions, based on geomorphology, geology, and ecology. The definition of ecoregions led by James Omernik (Omernik and Griffith, 2014) may be useful in classifying the impacts of watershed contributions in the Gulf Region, while various older studies which presented classifications of coastal systems may still contribute much to our understanding (Lugo and Snedaker, 1974; Odum and Copeland, 1972).

How Modeling Advanced Restoration of the Chesapeake Bay and Its Watershed

The Chesapeake Bay Program focuses on nutrient load reductions, a compounding cumulative effect (see Table 3.1, above), to eliminate a spectrum of eutrophication impacts (cross-boundary cumulative

___________________

7 See https://www.gulfspillrestoration.noaa.gov/sites/default/files/2018_01_TC_MAM_Procedures_Guidelines_Manual_12-2017_508_c.pdf.

8 See https://www.integratedecosystemassessment.noaa.gov/regions/gulf-of-mexico.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

effects). During its more than 35 years of operation, there have been substantial increases in population in the 64,000 mi2 watershed, as well as many land-use changes, such as forest losses and increases in impermeable surfaces (Clune et al., 2021). Total nitrogen and total phosphorous loads have decreased by about 27 percent and 23 percent, respectively, during the past 25-year period (Testa et al., 2018). Reductions in atmospheric deposition of total nitrogen (Eshleman and Sabo, 2016) and substantial upgrades in wastewater treatment plant operations that removed both nitrogen and phosphorous have been responsible for much of these declines (Testa et al., 2018). As a result of load reductions, water quality conditions (e.g., nutrient and algal chlorophyll a concentrations) have improved in many, but not all, areas of the estuary, hypoxic/anoxic conditions have decreased in size and persistence in the bay and its large tributaries, and communities of submerged aquatic vegetation (SAV) have started to recover (Testa et al., 2018). Restoration experience in the Chesapeake Bay indicates that estuarine water quality “memory” is short (seasons to years rather than decades); nutrient load reductions exhibit the strongest effects close to load-reduction locations; and season and location are also important, with high-salinity areas responding to small load reductions during the nutrient-limited summer season.

The Chesapeake Bay Program uses a variety of models, from qualitative conceptual models to complex spatially explicit simulation models. Initial understanding of nutrient dynamics and estimates of needed load reductions for the Chesapeake Bay were based on a sequence of relatively simple mass-balance computations (Boynton et al., 1995; Nixon, 1987; Smullen et al., 1982), each having limited spatial resolution and accuracy; over time, modeling efforts improved as understanding improved and better datasets became available (e.g., Lee et al., 2013). Initial coupled hydrodynamic–water quality models had limited spatial resolution and were flawed with regard to estuarine biogeochemistry (D’Elia et al., 2003; Hydroqual, 1981), but they served to stimulate testing and guide improvements (e.g., Brady et al., 2013; Cerco and Cole, 1993). The current Chesapeake Bay coupled hydrodynamic–water quality model now has considerable spatial detail and is used as a primary management tool in nutrient reduction efforts (Hood et al., 2021).

Although in a different region of the country, the management of the Chesapeake Bay Program, including its scientific and political processes, has lessons and perspectives for considering Gulf of Mexico restoration efforts, in part because of many situational similarities. Both programs also cover multiple states sharing a common resource and includes many actors working together across political boundaries, including with multiple state resource agencies and federal counterparts. Issues addressed by the Chesapeake Bay Program in common with the GoM are many and include water and sediment quality, habitat loss (especially oysters and seagrass), coastal development and fisheries disruptions, and of course climate change. Further, while the Chesapeake Bay modeling effort is more mature than efforts in the Gulf of Mexico, significant contributions toward prioritizing restoration efforts are under way. For example, the generation of the Louisiana State Master Plan, which is updated once every 5 years, is dependent on a complex modeling system that utilizes landscape modeling through an Integrated Compartment Model (ICM), surge and wave models, and risk models (White et al., 2017).

The Chesapeake Bay example suggests that value can be added by modeling early with conceptual and simple mass-balance models representing initial hypotheses about the ecosystem, then using those models as a framework for synthesizing data toward improved understanding, more comprehensive models and assessments, effective adaptive management decisions, and, ultimately, measurable estuarine restoration. This 35-year restoration effort adopted the idea of starting with initial analysis and modeling, adding complexity as scientific understanding and management needs allowed and demanded (Linker et al., 2013).

AN APPROACH FOR CONSIDERING THE CONSEQUENCES OF LARGE-SCALE RESTORATION

Evaluating the effects of a restoration effort often involves one body of water or watershed and is therefore unreplicable (Waltham et al., 2021). This means that the usual design with which most ecologists and environmental scientists are familiar, an experiment with randomly allocated treatments and replica-

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

tion, is not possible (Howe and Martínez-Garza, 2014). However, this does not mean that rigorous analysis of system-wide restoration projects cannot be done, only that strict assignment of causes and effects cannot be made via standard methods of statistical analysis alone.

To compensate for this inability to use traditional experimental designs, the lack of reference conditions, the lack of replication, the difficulties in establishing causality and often the shortage of appropriate data, Diefenderfer et al. (2011, 2016) proposed an evidence-based evaluation methodology that utilizes multiple lines of evidence and causal criteria. These methodologies have previously been used in biological risk assessment (Suter et al., 2002, 2010). Downes et al. (2002) recommended that causal criteria be used in ecosystem restoration, and they have since been used effectively in freshwater ecology (Norris et al., 2012; Webb et al., 2015). Although the methods have been in existence since their introduction for health research in the 1960s (Hill, 1965; U.S. Department of Health, Education, and Welfare, 1964), multidisciplinary applications like these are still expanding (Ludwig et al., 2010; Wickwire and Menzie, 2010).

Multiple Lines of Evidence

Multiple lines of evidence are intended to confer built-in redundancy, where each line of evidence functions as an “umbrella” under which relevant analyses are collected for evaluation of key hypotheses. The following discussion is built on the lines of evidence proposed by Diefenderfer et al. (2016, p. 7) for salmon habitat restoration and reconnection in the Columbia River estuary, which were intended to be “universally applicable to large-scale ecosystem restoration programs,” are applied here for the more general GoM purpose of recovering aquatic species through coastal habitat restoration (see Table 3.3). The organizing principle consists of seven general lines of evidence, designated a–g, which are intended to encompass the typical kinds of indicators monitored to evaluate large-scale restoration effectiveness, and related analytical and modeling methods. As applied to large-scale ecosystem restoration, the lines of evidence were intended to be used in synthesis of the composite data and analyses to help distinguish association from causation (Diefenderfer et al., 2016). A suite of monitored indicators and analyses is developed for each line of evidence to evaluate hypotheses represented in the conceptual model.

The Chesapeake Bay example, discussed above, can be used to illustrate the concepts of multiple lines of evidence, causality, and thresholds. From the beginning of the Chesapeake Bay Program in the 1980s, extensive monitoring data analysis and various types of modeling, such as mass-balance models and large-scale numerical models, were undertaken. These tools were used to develop multiple lines of evidence, which were used to determine that excess nutrients and eutrophication were responsible for extensive loss of seagrass beds and associated habitat. Initially, it was thought that excess phosphorus was responsible for the eutrophication, but further investigation showed that nitrate was a critical component as well, showing the importance of understanding causality. Finally, it was demonstrated that the Chesapeake Bay had exhibited the characteristics of exceeding a threshold when it flipped from a benthic-oriented system to a water column–dominated system.

In the Chesapeake Bay Program example, several specific lines of evidence came up, including:

  • an emphasis on long-term monitoring in tidal and nontidal waters and from atmospheric sources (line a) and of the landscape (line g);
  • research aimed at improving understanding of issues related to water quality and habitat restoration (line e); and
  • a coupled and evolving suite of models (line b) designed for a number of purposes.

Together, these exemplars support the needed synthesis of diverse datasets to test and forecast effects of nutrient load reductions in a complex environment (Linker et al., 2002, 2013).

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

TABLE 3.3. Description of Seven Lines of Evidence for the Recovery of Aquatic Species through Habitat Restoration

  Data Summary Analysis Synthesis Evaluation
Line of Evidence Monitored Indicators Analyses Causal Criteria Cumulative Effects Category
  1. Research on Critical Ecological Uncertainties
Various Summarize advances in understanding cause–effect associations; iterative improvement of the conceptual model Plausibility, temporality, specificity, coherence, exposure pathway, predictive performance Indirect, time lags, compounding
  1. Evidence-Based Review of the Literature
Species presence, residence time, survival, prey availability, diet, stomach fullness, growth Systematic global literature search, filtering, review, and scoring based on formal criteria Strength and consistency, plausibility, specificity, analogy, coherence, predictive performance Not applicable to cumulative effects
  1. Physics-Based and Ecosystem Models
Water-surface elevation, particulate organic matter export Hydrodynamic modeling of inundation patterns and particulate organic matter export Strength and consistency, plausibility, gradient, temporality, coherence, exposure pathway Space crowding, indirect, time lags, cross-boundary, nonlinear, compounding
  1. Meta-Analysis of Restoration Action Effectiveness
Water-surface elevation, water temperature, sediment accretion, vegetation similarity, species presence Qualitative assessment of action-effectiveness studies in the restoration program; analysis of data from historically reconnected sites Strength and consistency, gradient, specificity of association, coherence, predictive performance Landscape, time lags
  1. Analysis of Data and Modeling of Target Species
Presence, diet, stomach fullness Comparative analysis of stomach contents; detections of migratory species Plausibility, gradient, coherence, exposure pathway Cross-boundary, indirect, compounding
  1. Modeling of Cumulative Net Ecosystem Improvement
Prey, biomass production, prey and biomass export, area of habitat restored Additive modeling of change in function, restored area, and probability of success Plausibility, coherence, exposure pathway Landscape, compounding
  1. Change Analysis on the Landscape Setting
Forest cover, impervious surface Remote-sensing data analysis of forest cover and urbanization change trajectories in watersheds Plausibility, coherence Landscape

Causal criteria are described in the following section.

See Tables 3.1 and 3.2.

NOTE: Causal criteria used in the synthesis are defined and described in more detail in the text below.

SOURCES: Adapted from Diefenderfer et al. (2016). Included are associated analyses of monitored indicators, the causal criteria used for synthesis (Hill, 1965; Dorward-King et al., 2001), and the cumulative effects categories used for evaluation (CEQ, 1997). Causal criteria employed in the synthesis are defined and described in more detail in the following section.

Causal Criteria

The problem of how to establish causality in coastal ecosystems is larger than statistical methods alone can resolve. Downes et al. (2002) suggest that restoration programs first define causal criteria and decide how they will be examined and measured, and then review available literature for effects of human activity and extract the information needed to evaluate response variables using each of the causal criteria. Causal criteria are aspects of the associations between two variables (Hill, 1965). Factors to be considered when deciding whether an observed statistical association is causal include:

  • temporality (the effect follows the cause),
  • strength of association (the magnitude of the effect),
  • consistency of association (documented by multiple observers under various circumstances),
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
  • dose–response relationships or biological gradient (gradient in the cause and response level),
  • consistency of evidence through replication of findings and other knowledge,
  • specificity of the association (limited to particular sites and/or effects),
  • biologic plausibility (understanding of the mechanism),
  • complete exposure pathway (the cause can reach the receptor),
  • coherence of evidence (lack of conflict between cause-and-effect interpretation and known facts),
  • experimentation (manipulation of the cause),
  • analogy (comparison to similar systems),
  • predictive performance (prediction of restoration outcomes), and
  • consideration of alternate explanations.

Not every criterion needs to be satisfied. In fact, the only necessary criterion is temporality. A positive statistical association between an exposure and an outcome does not necessarily mean that the exposure is the cause of the outcome. Causality is more than a “link”; it is a demonstration that an exposure(s) (restoration method) is responsible for a specific outcome(s). For every exposure–outcome relationship, there will always be gradations of evidence and certainty, and observed links or associations can be due to many factors. Causal inference is not purely objective, and it always includes a subjective judgment of the degree to which the evidence satisfies each criterion, leading to the ultimate conclusion of the likelihood that a particular causal relationship exists.

Although the cumulative effects of GoM restoration are not amenable to classical statistical analysis because the coast itself or individual estuaries, bays, and watersheds are the experimental units and cannot be replicated, causal criteria may be used to help distinguish among potential causes of an observed change in the ecological system. The aim of using causal criteria would be to separate the effects of acute environmental events or chronic environmental trends from the collective actions by those implementing GoM restoration.

TOOLS FOR GATHERING MULTIPLE LINES OF EVIDENCE

This section describes the types of tools that can be applied to collect data for, analyze, and model the multiple lines of evidence described in Table 3.3 (above). As with the evaluation of the cumulative impacts of stressors on ecosystems across geographic scales (Hodgson and Halpern, 2018), multiple analytical methods are needed to encompass the complexities of the cumulative effects of ecosystem restoration (Diefenderfer et al., 2016). Many of these tools are in use by GoM restoration project and program managers. Applications of each tool have the potential to produce results that contribute to evaluation of one or more lines of evidence. This section will not cover study design methods standard to ecosystem restoration—in particular the value of reference sites versus control sites, and before versus after data, which are described in Effective Monitoring to Evaluate Ecological Restoration in the Gulf of Mexico (NASEM, 2017).

Research on Critical Ecological Uncertainties

Table 3.3 summarizes the use of research to understand critical ecological uncertainties as a line of evidence. As discussed in the Chesapeake Bay example, uncertainties are often acknowledged at the outset, during development of ecosystem conceptual models. They can also arise during project and program implementation, especially when things go wrong or there are unexpected outcomes (Ebberts et al., 2018). Uncertainties can be traced back to data collection and modeling methods, problems with geographically scaling knowledge, changes in trends monitored through time series, and lack of quantification of drivers and stressors (Brudvig and Catano, 2021). Any scientific method suitable to address research hypotheses falls into the category of critical ecological uncertainty research. One frequently used technique is bench- or mesocosm-scale experiments, which contribute to restoration designs by isolating various factors in the ecosystem, revealing and quantifying relationships (Fry et al., 2017; Peralta et al., 2003).

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

The issue of how to assess the efficacy of a system-wide manipulation carried out without replication was considered by Carpenter (1993) in a review of his and colleagues’ manipulations of lakes. Carried out over many years, nutrient inputs and predator population were manipulated to alter and understand whole-lake structure and function. They intended to discover whether lakes changed nonrandomly after the manipulations, and whether the manipulations were responsible for any changes that occurred. Because of the lack of replication, standard statistical analysis could not be done and random assignment of treatments and alternative analyses were employed.

To test the first hypothesis that nonrandom changes occurred postmanipulations, Carpenter et al. (1989) used time-series analyses to test whether observed changes were outside the limits historically recorded in the manipulated lakes. If nonrandom changes were detected, the second hypothesis was tested by comparisons with similar types of lakes and model outputs, and small-scale mesocosm experimentation was employed to collect data that allowed ecological interpretations to be made. Ultimately, knowledge gained from this work was able to be used to restore damaged lakes to their original conditions.

Evidence-Based Review of the Literature

Richard H. Norris built a meta-analysis framework, EcoEvidence, to assess environmental cause and effect around causal criteria with information synthesized from multiple publications (Norris et al., 2012; Webb et al., 2015). Several quantitative methods are incorporated in the method to more objectively assess published environmental research (as shown in Table 3.3). Scores are based on the quality of the study design and the number of all available published evidence concerning a given hypothesis. The method has been applied to analyses of river flow alteration and effects on the biota from frequency, magnitude, and duration of flows (Greet et al., 2011; Webb et al., 2012). However, as in other evidence-based endeavors, such as medicine, the need to integrate the results of such weighted evidence is “relentlessly situated and contextual” (Wieringa et al., 2017). Much the same argument could be made in regard to the spatiotemporally variable conditions of the GoM.

Physics-Based and Ecosystem Models

There are many physics-based and numerical models that can be useful in the development of a modeling framework during restoration planning and implementation (Table 3.3). This line of evidence can be expanded to include ecosystem models due to increasing development and use of model linkages across physical and biological regimes, such as understanding the potential for impacts from long-term environmental trends, identifying possible synergistic and antagonistic stressors, quantifying cumulative effects, and facilitating cross-site comparative analysis. Linkages among natural resources are complex, even on small scales. Addressing these complex relationships is difficult, and even the most sophisticated modeling tools may need data that are not readily available in many areas.

Physics-based models are mathematical representations of materials and energy flow through the ecosystem and are used to capture active physical processes within a large-scale domain (e.g., Jaiswal et al., 2020). Multiple parameters or variables are often involved, with interrelationships bounded by either empirical or mathematical submodels. Physical models—such as hydrodynamic models (Hodges, 2014), sediment and nutrient transport models (Flynn, 2001; Merritt et al., 2003), and water quality models (Moriasi et al., 2015)—have been used to examine alternative coastal restoration project designs, the impact of sea level rise on coastal ecosystems, water column characteristics, soil erosion potential and soil–water dynamics, movement of riverborne sediments, carbon flux dynamics in wetlands, and other hydrological and geomorphological processes (e.g., Brown and Pevey, 2019; Burchard et al., 2006; Hiatt et al., 2018; Leach et al., 2021; Passeri et al., 2015; Rogers et al., 2012; Wassmann et al., 2006). NOAA’s Ecosystems & Fisheries-Oceanography Coordinated Investigations (EcoFOCI) program is one example of a physical and biophysical model-

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

ing effort that uses mathematical modeling to synthesize physical and biological data across an ecosystem (Dougherty et al., 2010).

Physics-based models can also be a useful tool for modeling active physical processes within a restoration site at various stages of restoration activities, although they need a large amount of field or simulated data. Ecosystem models include complex interactions among ecosystem components (e.g., species, habitats), ecosystem processes (e.g., drivers, pressures, stressors), and ecosystem services (e.g., carbon sequestration, biomass, population) (see summary Table 1 in O’Farrell et al., 2017; Geary et al., 2020). The interrelationships between various components and processes are either mathematically or empirically driven (Edwards, 2001). A large amount of field and simulation data are needed to parametrize ecosystem models, which ultimately provide important clues regarding temporal trends in overall ecosystem conditions or services (Denman, 2003; Matear, 1995; Peters and Okin, 2017). Integrated modeling can be used to bring together the potential elements of conceptual, physical, and population model results (EPA, 2008; Fulton, 2010; Laniak et al., 2013).

These complex models are typically developed to help define and assess multiple factors at larger spatial scales, including synergistic and antagonistic effects and background trends (Johnston et al., 2000, 2017). As an example, community models that deal with spatiotemporal dynamics of biotic assemblies in relation to underlying ecosystem or environmental processes are a combination of ecosystem and population models. The ICM, mentioned previously, is an example of a dynamic ecosystem model that simulates changes in wetland hydrology, species cover, and elevation along the Louisiana coast (de Mutsert et al., 2021). One of the challenges with ecosystem models is that scaling across the landscape can be problematic because of the amounts and variety of data needs related to model parametrization. Without the availability of large amounts of field data, ecosystem modelers may have to spatially extrapolate the relationships established at a site, which can be prone to high uncertainty (Geary et al., 2020).

Meta-analysis of Restoration Action Effectiveness

In environmental sciences, many studies do not report the statistics needed for formal, quantitative meta-analysis (Table 3.3) of restoration efforts (Greet et al., 2011; Norris et al., 2012). In conducting a global meta-analysis of wetland restoration, researchers found only 70 studies with sufficient information in the scientific literature from 1970 to 2010, despite including estuarine, lacustrine, palustrine, and riverine wetlands (Meli et al., 2014). The design of restoration experiments and hence the use of formal meta-analysis is further limited by the impossibility of finding true replicates in nature (Howe and Martínez-Garza, 2014).

Publication in scientific literature usually lags behind the timing needed to support ongoing decision making in a given restoration program, which can prevent formal meta-analysis of published data from contributing to decisions (Diefenderfer et al., 2011). However, meta-analysis of interim reports produced by restoration projects has been done in other systems and was seen as a valuable source of information for adaptive management and course correction (Diefenderfer et al., 2016). Hence, the term “meta-analysis” is used in this report to include the assessment of interim reports and data, as well as for the traditional use of the published scientific literature. The results of many GoM restoration projects initiated after DWH for restoring coastal nursery habitats, such as oyster reefs, salt marshes, and seagrass meadows, are still under way or in the process of being analyzed, and in lieu of formal meta-analysis, a qualitative meta-analysis of interim reports could be informative.

As an example, reports developed during the initial phase of the multi-State/EPA Chesapeake Bay Program (1978–1983), some of which were later transformed into peer-reviewed publications, played an important and timely role in designing portions of the current program. Reports by Heinle et al. (1980), for example, described some early water quality trends and made clear the need for a comprehensive and long-term monitoring program; Stevenson and Piper (1979) summarized knowledge concerning the causes and consequences of SAV declines in the bay and, again, indicated the need for long-term habitat monitoring and research. Thus, gray literature—particularly rigorously reviewed federal reports—can be a very

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

important resource for assessing restoration effectivness in the GoM. The Strategic Conservation Assessment (SCA) tool9 funded by the RESTORE Council has developed a conservation planning inventory tool that could potentially serve as a repository for these types of assessments and reports.

Analysis of Data and Modeling of Target Species

Often, habitats for threatened or endangered species are a focus of coastal restoration, and data on species are collected accordingly, particularly to connect species recovery with habitat restoration. Depending on the target species, a wide variety of such data may be warranted, with associated analyses and modeling (see Table 3.3, above). Population models, for example, are mechanistic models used to predict or simulate population dynamics of species within an ecosystem due to changes in habitat characteristics from a set of drivers and stressors. These models can also be used to analyze species vulnerability, movement, and individual traits. Since population dynamics are intricately linked to habitat characteristics, restoration projects can utilize population models to analyze improvement in ecosystem services in terms of species dynamics with a restored habitat. For example, a marsh restoration site can use a bird population model to examine the changes in species dynamics and individual traits before and after the restoration projects. Population models have been widely used in the GoM to project changes in plant or animal populations in response to types and severity of background trends, particularly in developing Gulf-wide or regional fisheries and endangered species management plans. Population models have also been used at smaller spatial scales to examine potential functional effects (such as changes in the level of primary and secondary productivity) associated with an oyster reef or seagrass bed restoration project.

___________________

9 See https://www.quest.fwrc.msstate.edu/sca-project.php.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

The accuracy and precision of population models depend on the quantity and quality of monitoring data collected on the species and understanding of critical factors driving population changes. A report from the National Academies of Sciences, Engineering, and Medicine (2017) linked the monitoring requirements for mobile species having large spatial ranges with monitoring requirements for cumulative effects beyond the project scale. It discussed inadvertent impacts of restoration on wide-ranging species and how some types of restoration activities might produce harmful effects, and detailed monitoring methods to help restoration planners and wildlife managers minimize impacts and benefit these species.

Modeling Cumulative Net Ecosystem Improvement

As shown in Table 3.3, one line of evidence involves estimating whether and to what degree an ecosystem may or may not have improved because of an intervention, such as restoration. This section provides a brief overview of several models commonly used by agencies for habitat change. The calculation of cumulative net ecosystem improvement (CNEI) is one recently introduced example. This model was developed for ecosystems in the Pacific Northwest (Diefenderfer et al., 2016) and is based on the earlier Net Ecosystem Improvement Index (Thom et al., 2005, 2011). Because this additive function considers multiple restoration efforts in a particular geographic area, it can be a useful tool for assessing the cumulative effects of multiple restoration projects relative to a specific target function. CNEI takes into account the project area, the number of restoration projects within that area, changes in ecological function, and the probability of long-term restoration success. In this model, examples of changes in ecological function could be an outcome such as seagrass biomass or secondary production such as the density of juvenile invertebrates in a seagrass meadow.

A strength of this model is that it can be used in any ecosystem (Diefenderfer et al., 2016). The key decisions are which metrics are most appropriate to use for assessing change in ecological function and the probability of success, based on known conditions and past responses to restoration actions. A suitable reference site and the ability to estimate success are also needed. There are techniques employing CNEI ideas that have been developed to objectively evaluate wetland functions in the northern GoM: the Wetland Value Assessment and the hydrogeomorphic approach, which are described below. For the most part, these techniques are used to evaluate the potential effects of restoration efforts, but they could potentially be used as one “line of evidence” in an overall cumulative effects investigation.

The Wetland Value Assessment, an example of metrics already assessed for other restoration tools that could be used in the GoM, was developed under the Coastal Wetlands Planning, Protection and Restoration Act (CWPPRA) program to determine the benefits of proposed wetland restoration projects. In this approach, habitat quality and quantity are measured for baseline conditions. Then, “future without project” and “future with project” conditions are predicted, based on modeled data and/or the best professional judgment of the team. This approach uses variables considered important to the suitability of a particular habitat type for supporting a diversity of fish and wildlife species. Each model consists of (1) a list of variables considered important for characterizing fish and wildlife habitat in a particular wetland type, (2) a suitability index graph for each variable, and (3) a mathematical formula that combines the suitability indices for each variable into a single value for wetland habitat quality. Modules exist for cypress-tupelo swamp, fresh/intermediate marsh, brackish marsh, saline marsh barrier islands, barrier headlands, and coastal chenier/ridge. This methodology has been used heavily in Louisiana (USFWS, 2006).

The hydrogeomorphic approach, rather than focusing only on fish and wildlife functions of wetlands, seeks to evaluate the physical, chemical, and biological characteristics of wetlands. Model development begins with the classification of the wetland system into regional wetland subclasses based on hydrogeomorphic factors (Brinson, 1993), followed by the creation of a functional profile that describes the characteristics of the regional subclass, its functions, and the ecosystem and landscape attributes that influence each function. Reference wetlands are selected from a defined geographic area, assessment models are developed, and the models are calibrated using the reference wetlands (Smith et al., 1995). Similar to the Wetland Value Assessment, the models produce a numerical value that is multiplied by area, though functions remain

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

independent, not summed. Although the hydrogeomorphic approach is not widely used because of its detail and data requirements, such models have been developed for several GoM wetland regional subclasses, including the Northwest GoM tidal fringe wetlands (Shafer et al., 2002) and tidal fringe wetlands along the Mississippi and Alabama Gulf coasts (Shafer et al., 2007), as well as numerous interim models for the wetlands in the vicinity of Galveston, Texas.

Both of these wetland evaluation techniques are designed for individual projects, not cumulative impact assessment, where the focus shifts from functions performed at the wetland scale to the larger watershed scale. However, the functional indices in the hydrogeomorphic models may be used in conjunction with other methods designed specifically to assess cumulative impacts, and its concepts have been used in other parts of the country, such as the assessment of several of the watersheds in southern California (Smith, 2003).

The Biological Condition Gradient (BCG) approach was developed by the U.S. Environmental Protection Agency (EPA) to define and communicate existing conditions of aquatic biological resources in order to meet requirements of the U.S. Clean Water Act (Davies and Jackson, 2006; EPA, 2016), and has potential for applications in estuarine, reef, and watershed restoration (Box 3.3). Originally applied to benthic stream organisms in the U.S. Northeast, the model estimates biological conditions at a site along a continuum, from natural or undisturbed to severely altered by anthropogenic stress. Each level is defined by an empirically derived description that can be interpreted by experts and practitioners, regardless of location or habitat type (Cicchetti et al., 2017; Yee et al., 2020). Because of the consistent and structured steps defined in the model, this approach has been applied to different regions and ecosystems, including streams (Davies and Jackson, 2006), estuaries (Cicchetti et al., 2017), and coral reefs (Bradley et al., 2020).

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

Change Analysis on the Landscape Setting

Data-driven models, often referred to as machine learning (ML) or deep learning (DL) models, can be quite effective in teasing apart complex relationships among ecosystem drivers or stressors and response, and in identifying patterns among different datasets. However, their performance depends on the types and amounts of training data available from a site (Clark and Gelfand, 2006; Crisci et al., 2012; Goldstein and Coco, 2015). Once these data-driven models are trained on a dataset comprising representative variables with a wide range and standard deviation, they can be used for long-term predictive modeling and monitoring. They can also be scaled up to satellite data for large-scale mapping (O’Connell et al., 2021; Sejnowski, 2020). These types of data-centric, predictive models have become a valuable tool for the ecology and environmental science community in recent years (Hampton et al., 2013; O’Connell et al., 2021; Rammer and Seidl, 2019). New refinement procedures are constantly being developed to improve their predictive ability and reduce overfitting problems, which is a common problem in these types of models when the model overperforms in the training phase and underperforms in the prediction phase (Willcock et al., 2018). One of them is ensemble modeling, for which several machine learning models are run, and their output is combined using a rule-based algorithm to produce the most accurate prediction (Sagi and Rokach, 2018).

Machine learning models are predictive models based on computational algorithms that are trained with a large number of input variables (e.g., physical, chemical, biological, ecosystem data) to disentangle complex and nonlinear relationships (see reviews by Hampton et al., 2013; Thessen, 2016). DL models often refer to a family of learning algorithms that use multiple hidden layers to build relationships for prediction, instead of using one established method to develop relationships among biophysical, biochemical, ecological, and environmental datasets (Thessen, 2016). Most ML/DL models are supervised learning algorithms that can use a variety of datasets from heterogeneous sources by compiling a large training dataset even without a set of hypotheses (Hampton et al., 2013; Rammer and Seidl, 2019). Similarly to traditional or existing physics-based and ecosystem models (discussed above), these models need a large amount of training data; however, they allow a degree of heterogeneity in the way the data were collected and processed (Crisci et al., 2012; Vinuesa et al., 2020). That characteristic brings a greater amount of flexibility to these new modeling tools compared with the rest of the modeling tools discussed above, which utilize controlled datasets and extensive parametrization. These new-generation models are gradually being used and implemented across a variety of ecosystem assessment studies (Ryo et al., 2020).

Data-driven modeling tools to assess cumulative effects of large-scale restoration projects need large amounts of data, whether field, simulated, climate, and other biophysical data; satellite-derived spatial or point data; or a combination of all data (Crisci et al., 2012; Goldstein and Coco, 2015; Vinuesa et al., 2020). Today’s world of big data, pervasive sensing, massive computing capacity, ML based on artificial intelligence (AI),, and DL create an opportunity to transform ecological modeling and large-scale ecosystem synthesis studies (Humphries et al., 2018).

In recent years, there has also been a substantial expansion in federally funded ecological site networks, which provide large heterogeneous datasets for such data-driven models to be implemented. Examples of such networks include AmeriFlux,10 FLUXNET,11 Long Term Ecological Research Network,12 National Ecological Observatory Network,13 Long-Term Agroecosystem Research Network,14 Ecosystem Phenology Camera Network,15 and Global Lake Ecological Observatory Network16. These data generation networks can support data-centric modeling at a larger—that is, watershed or estuary—scale; however, their direct application to small restoration sites at a project scale could result in higher prediction uncertainty and

___________________

10 See https://ameriflux.lbl.gov/.

11 See https://fluxnet.org/.

12 See https://lternet.edu/.

13 See https://www.neonscience.org/.

14 See https://ltar.ars.usda.gov/.

15 See https://phenocam.sr.unh.edu/webcam/.

16 See https://gleon.org/.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

entail new site-specific calibration and validation. In addition, while many of these networks may not be currently available for Gulf sites, these examples provide program managers with options for future data collection for their sites based on their program needs. For example, the PhenoCam network17 provides system architecture, camera specifications, mounting instructions, and source code in an open-source manner for easy implementation at any site.

In addition, satellite- and drone-based remote sensing, and a wide array of environmental citizen science projects are generating tremendous amounts of spatial and point data at different scales (Corbane et al., 2015; Ridge et al., 2020). Some of these data types are available in an open-source manner, including climate, biophysical, and satellite-derived datasets from federal agencies, such as the MERRA-2 database18 from the Global Modeling and Assimilation Office of the National Aeronautics and Space Administration (NASA) or NASA’s Earthdata database19. However, other data needs to be collected at an appropriate spatiotemporal scale. In the GoM, big datasets for ocean and coastal waters are available from the Gulf of Mexico Coastal Ocean Observing System,20 which collects and archives thousands of data points from a variety of sensors each year. On a smaller scale, although it is still substantial relative to datasets for large-scale restoration across the United States, Louisiana has developed the Coastwide Reference Monitoring System21, with 390 sites that collect data on the ecological condition of coastal wetlands.

These data-driven models can be valuable to analyze cumulative effects across multiple restoration projects at a variety of scales, provided sufficient training data are available at the initial stages. Although these models have not yet been widely adapted by researchers and restoration managers for large-scale cumulative impact assessment in the GoM, they have been increasingly popular for many types of ecosystem monitoring and predictions in the past decade, including wetland biomass, primary production, species dynamics, habitat suitability, and driver-response characterization studies (Huang et al., 2021; Michaels et al., 2019; O’Connell et al., 2021; Ridge et al., 2020; Shiu et al., 2020).

There are some examples of ML- and DL-based ecosystem models that exist for the GoM for different types of ecosystems, physical parameters, and biota. One such example is the supervised machine learning by emergent self-organization map analysis model proposed by Engle and Brunner (2019) to analyze the geochemistry of water samples collected from oil and gas wells in the northern GoM. Shiu et al. (2020) proposed a deep neural network for automated detection of marine mammal species, which enhanced the accuracy of the detection by orders of magnitude when compared with other detection algorithms. Trifonova et al. (2019) tested a data-driven model—dynamic Bayesian network models—with different levels of structural complexity and a varying number of hidden variables to predict ecosystem dynamics in GoM. Through this model, they discovered meaningful interactions among ecosystem components and their environment and examined how climate perturbations affect these relationships. One caveat is that these AI-based ML/DL models tend to indicate the correlation between ecosystem variables, not causation. One rapidly growing subfield in AI is explainable AI (xAI), which is used to decipher the complex output of ML models at various scales (Kakogeorgiou and Karantzalos, 2021; Phillips et al., 2021; Ryo et al., 2021). For example, Ryo et al. (2021) provide a summary of xAI tools available for ecologists for application in species distribution modeling at different scales. The use of models based on xAI that use field datasets or remote sensing datasets is a rapidly growing field in ecosystem modeling and environmental monitoring.

REFLECTIONS ON RESTORATION PLANNING AND ENDPOINTS

Ecosystem restoration is planned and carried out by various types of governmental, nongovernmental, and academic institutions that have differing constraints and the focus has historically been on individual project sites, ecosystems, or species (Clewell and Aronson, 2013; Roman and Burdick, 2012). The approach-

___________________

17 See https://phenocam.sr.unh.edu/webcam.

18 See https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2.

19 See https://www.earthdata.nasa.gov/.

20 See https://gcoos.org.

21 See https://lacoast.gov/crms/Home.aspx.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

es to planning and evaluating large-scale restoration discussed above are not standard in the current restoration planning paradigm (Diefenderfer et al., 2021). Planning by the DWH funding entities required the invention of systems for both prioritizing restoration projects and evaluating their outcomes, with many projects proposed at ecoregional (e.g., Chenier Plain) or statewide coastal-zone scales—or focused on migratory or highly mobile species that also necessitate a broad view. Such post-DWH projects are implemented at a much larger scale than most restoration conducted anywhere previously and have the potential to be more cost-effective with good planning and management (Neeson et al., 2015). Landscape-scale restoration planning is the exception, not the rule, and in the continental United States is generally carried out by long-term partnerships led by federal and state agencies at scales, such as the Greater Florida Everglades, Missouri River system, San Francisco Bay/Sacramento Delta, or Puget Sound (e.g., Fischenich et al., 2018). In contrast, DWH funds are available for a finite period with projects implemented in a decentralized manner across multiple administrative processes.

Applied Restoration Strategies for Landscape Stressor Constraints

As discussed in Chapter 2 and seen in the acute and chronic inputs shown in Figure 3.1 (above), stressors affecting an ecosystem occur on a variety of spatial and temporal scales. These stressors (defined in Box 2.1 in Chapter 2) affect the physical and biological conditions that impact ecosystem structure and function (Groffman et al., 2004; Twilley et al., 2019). For example, changes in estuarine inflow can interfere with the sediment-trapping function of salt marsh plants, which in turn affects the physical condition, ecosystem structure, and function of the marsh. The landscape processes affecting ecosystem structure and function can be estimated and then employed to approximate the relative degrees of stress and thereby prioritize areas for particularly beneficial types of restoration actions (Diefenderfer et al., 2009; Roni et al., 2018).

The degree of stress in an ecosystem, as well as the larger landscape in which it resides, can be helpful for prioritizing of actions (NRC, 1992). Possible actions include, after Thom et al. (2005):

  • restoration, which is the return of an ecosystem to a close approximation of its previously existing condition (NRC, 1992);
  • enhancement, which is any improvement of a structural or functional attribute (NRC, 1992);
  • preservation (or protection), which is the exclusion of activities that may negatively affect the system;
  • conservation, which is the maintenance of biodiversity and natural ecosystem processes; and
  • creation, which is the development of an ecosystem that is historically not present in a given geographic area.

Depending on the degree of stress both at a site and in the landscape processes supporting ecosystem functions at the site, specific management actions can be recommended to maximize the probability of success (Diefenderfer et al., 2009; Thom et al., 2005) (see Table 3.4). For example, sites with low degrees of local (site-scale) and surrounding landscape stress are the best candidates for conservation and preservation. Restoration is suitable for sites with medium to high degrees of stress, provided that landscape-scale stressors are minimal or can be treated to minimize their impact on the recovering site. Sites with high degrees of local and landscape-scale stress are unlikely to benefit from full restoration and instead the prudent action would be to focus on enhancement or creation.

Impairments at the landscape scale, which cannot be ameliorated, constrain restoration potential (NRC, 1992), as well as the types of cumulative effects (see Table 3.1, above) that may be seen after restoration efforts are completed. Outcomes of the more active land- and water-management strategies (restore, enhance, and create) may interact with the conservation and preservation of resources in the vicinity.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

TABLE 3.4. Ranking the Effectiveness of Restoration Approaches Based on Site-Scale and Landscape-Scale Stress

Landscape-Scale Stressors
Site-Scale Stressors Low Medium High
Low Conserve
Preserve
Conserve
Enhance
Restore
Preserve
Enhance
Conserve
Medium Restore
Enhance
Conserve
Preserve
Enhance
Restore
Conserve
Enhance
Create
High Restore
Enhance
Enhance
Restore
Create
Enhance

SOURCES: Adapted from Diefenderfer et al., 2009; based on Thom et al., 2005.

Constraints of Conventional Planning on Understanding Ecological Endpoints

The expectations for potential long-term and large-scale effects of restoration need to be tempered by real assessments. For example, since the passage of NEPA, environmental planning requires documentation, including the no-action alternative also known as the “future without project” or “without condition scenario” (see discussion above). Specifically, this projected future baseline condition, accounting for the types of trends described in Chapter 2, allows for an assessment of the benefits specifically attributable to the project that are expected to accrue over the project life span and for completion of an incremental cost analysis comparing the cost of a project to the projected benefits (USACE, 2000). In the Gulf, however, due to the complexity of stressors—such as relative sea level rise, freshwater inflows, hurricane frequency, and hurricane intensity (see Chapter 2)—this standard project-planning, traditional workflow may not achieve conventionally useful outcomes.

In standard project planning, the future without project is often a relatively perfunctory part of the planning process, and this component is often not well funded; nevertheless, Yoe (2012) points out that “the preparation of the without condition scenario [future without project] [is] the single-most critical analytical task in the planning process” (p. 1). The U.S. Army Corps of Engineers has developed guidance documents detailing some of the procedures and processes that are helpful to include in such an assessment (USACE, 2014, 2018). In a changing environment like the Gulf, a complete future without project assessment, in contrast with traditional approaches, may need to entail a large part of project planning to be useful. The future conditions report in the Louisiana Coastal Master Plan, for example, considers multiple sea level rise predictions, a range of subsidence rates, and a range of hurricane impacts (CPRA, 2017).

Once a valid range of future conditions has been determined, an additional challenge to conventional restoration planning is presented by the rapidly changing ecological conditions in the GoM relative to restoration goals and objectives. Such predictions depend upon the collection and archiving of quality long-term datasets (see Chapter 2). The development and maintenance of this planning process has needed and will continue to need significant capital investment, but it is essential to determining if a proposed project will perform as designed into the future.

Traditionally, a selected restoration project yields projected benefits, or net ecosystem improvement (Thom et al., 2005) or net ecological gain (National Infrastructure Commission, 2021), often measured in

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

habitat units or other “environmental currency” per dollar spent (USACE, 2000). In contrast, in the rapidly changing Gulf, it is possible that there may not be a net increase of benefits over the project life span. This poses a challenge to the traditional planning paradigm. Currently, there are no universally accepted guidelines for performing incremental cost analysis to compare projects that (1) maintain current conditions but yield no new benefits or (2) projects in which the action merely slows the rate of degradation. The outcome of many of the coastal land-building projects in Louisiana were designed with this second category in mind—that is, to slow, rather than stop, the rate of land loss. Acknowledgment of other social effects, including community resilience, and economic benefits of restoration projects, including avoided costs, could continue to be explored and addressed. Further work to determine restoration endpoints is a constraint for all restoration activities, particularly those in the dynamic GoM.

Importance of Detecting Mismatched Scales in Evaluating Restoration

The total area of wetlands in the GoM is about 18,261 km2 (Turner and Rabalais, 2019). The RESTORE Council FY2020 Annual Report to Congress, which summarized the accomplishments for 2018–2020 from funding under the 2015 Initial Funding Funded Priorites List and State Expenditure Plans awarded to date (RESTORE Council, 2021a), reported 8.5 km2 of wetland restoration. This is about 0.05 percent of the current wetland acreage in the Gulf. Still, as discussed relative to synergistic effects, coastal wetlands have a disproportionately high effect on fisheries production, carbon sequestration, and other beneficial functions (Bauer et al., 2013; Windham-Myers et al., 2018). Even combining land acquisition, areas that are under contract to apply best management practices, nonwetland habitat restoration, oyster habitat restored, and areas where invasive species were removed to the wetland restoration number (RESTORE Council, 2021a), the total was 234 km2. The DWH NRDA Trustees, in summaries of data pulled from NOAA’s Data Integration Visualization Exploration and Reporting (DIVER) portal, have created, restored, or enhanced 2,350 acres of marsh, beach, and dune habitats. Project summary information from the National Fish and Wildlife Foundation’s Gulf Environmental Benefit Fund indicates that approximately 170,000 acres of wetlands and other coastal habitats have been protected, restored and/or enhanced. Of course, this is only the preliminary funding and granting effort, and much more will be completed in the future, but it does illustrate the difficulties of scale and points to the necessity of explicitly determining an appropriate scale for cumulative effects assessment and the need to scale restoration efforts to a size appropriate for addressing the ecosystem problem. Further, a lack of evidence of cumulative effects at the regional scales does not necessarily point to a failure of individual restoration effects, but rather may be due to an insufficient relative scale of change.

CASE STUDY OF CUMULATIVE EFFECTS IN THE ANNUALLY RECURRING HYPOXIC ZONE IN THE GULF OF MEXICO

The Mississippi Basin provides an example of the cross-boundary effects, time lags, antagonistic and synergistic effects, and time crowding discussed in this chapter. Agricultural activities in the midwestern sections of the United States exert strong cross-boundary effects on eutrophication more than 1,000 miles away in the GoM. The concept of time lags is evident in the timing of nutrients that leave the midwestern fields during snow melt, spring rains, and spring flood events and cause hypoxia in the Gulf that generally peaks a couple of months later during the summer. As an example of an antagonistic effect, freshwater inflows from the Mississippi River can have a positive effect on adjacent swamps and bottomland hardwood forests, but too much freshwater entering the estuaries can be deadly to oysters. The repeated openings of the Bonnet Carré Spillway (six times in 10 years) in the 2010s, subjecting oysters in Mississippi Sound to multiple years of injury from excess freshwater, illustrates the idea of time crowding. And, perhaps most importantly, the Mississippi River Basin is instructive of the difficulties of scale that are involved when there are many restoration efforts that, even cumulatively, are small relative to the overall area of concern.

The Mississippi-Atchafalaya River Basin is the world’s fourth largest; it drains portions of 32 states and covers 41 percent of the contiguous United States. The region is home to 57 percent of U.S. farmland,

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

including about 180 million acres of corn and soybean fields, which contributes the majority of anthropogenic nutrients (nitrogen and phosphorus) to the Gulf of Mexico. In turn, 60 percent of the nitrogen nutrient input to the GoM comes from agricultural input (farm fertilizer, agricultural inputs from legume crops, and confined animal manure).22 An annually occurring hypoxic zone forms during the summer months as a result of these excess nutrients (Robertson and Saad, 2021), leading to algae growth and subsequent decomposition. As the decomposition process occurs, oxygen is depleted in the water, resulting in stress or death to any life that cannot move out of the zone.

Since 1985, a yearly snapshot of the size of the zone is available from a research cruise over a 2-week period each summer (see Figure 3.4). This single annual measurement of the hypoxic zone is an indicator of cumulative effects of human and natural systems on water quality and ecosystem health across large scales and long time periods. The annual hypoxic zone in the GoM illustrates a wide array of cumulative effects, which are illustrated in Table 3.5.

In this example of the traditional assessment of the cumulative effects of stressors causing environmental degradation, multiple synergistic and antagonistic effects of stressors can mask the ability to detect cumulative effects. While the quantity of nitrogen delivered to the Gulf is an important predictor of the size of the hypoxic zone, discharge rates in May of each year, as well as the presence of hurricanes and ocean currents, can generate antagonistic and/or synergistic cumulative effects (cross-boundary and compounding; see Table 3.5, which can have enormous consequences for the observed zone size in any given year. Excess nutrients are anthropogenic in source and their delivery downstream depends on precipitation upstream. A drought year can cause the following year’s delivery of nutrients to be unusually small, even if the total nutrient usage was large. Likewise, years of flooding and extensive rainfall can result in large nutrient deliveries the following spring, despite little or no increase in the application of nitrogen and phosphorus upstream (in Table 3.1, “Time Lag”). In addition to precipitation upstream, hurricanes and ocean currents can significantly impact the magnitude of the measured hypoxic zone. Given that precipitation is more variable and hurricanes are intensifying with climate change, it seems likely that these effects will continue and become more pronounced. Because of the presence of these significant confounding effects, rather than

Image
FIGURE 3.4. Size of GoM hypoxic zone. The year-to-year variability is due to a variety of synergistic and antagonist effects. HTF refers to the hypoxia task force. SOURCE: https://www.noaa.gov/news-release/larger-than-average-gulf-of-mexico-dead-zone-measured.

___________________

22 See https://www.usgs.gov/media/images/sources-nitrogen-delivered-gulf-mexico-0.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

focusing on a single-year measurement of the hypoxic zone, most analysts consider the 5-year running average in order to smooth out some of these effects.

The second key point of this example is that multiple modes and pathways of cumulative effects of restoration activities will often occur simultaneously and can reduce or augment the effects of individual restoration projects. In the context of the hypoxic zone, restoration activities are those that change land use, farming practices, and/or water delivery across the Mississippi River Basin with the goal of reducing nutrient inflows to the GoM; examples are provided in Table 3.5.

The third key point of this example is that, in the absence of a sufficient scale of restoration activities, cumulative effects of restoration activities may not be measurable and/or will be easily obscured by background variability. In 1997, the national Mississippi River/Gulf of Mexico Watershed Nutrient Task Force was created to coordinate activities, set targets for reductions in the size of the zone, and support a variety of related actions. In 2001, an Action Plan was released and over the next two decades, a number of reassessments and updates to it have been published.23

The official goal for the 5-year average size of the zone is depicted in Figure 3.4 and lies at the line labeled “hypoxia task force (HTF) goal.” In this figure, the goal is not currently being met and has not been met in any single year. Further, every 5-year average far exceeds it. Is the failure to meet this goal evidence that individual restoration efforts were not effective? To answer this question, it is first necessary to determine whether the set of restoration projects and activities that have been implemented since 1985 are of sufficient scale and/or duration to have had a measurable effect at the large-watershed scale. Second, it may be that adequate change in farming actions has occurred, but there may be significant lag times between changes in farming practices throughout the watershed and a clear, measurable signal that these changes are reducing the average size of the hypoxic zone, particularly considering the confounding effect of weather variability that contributes to high variability of the zone’s size.

TABLE 3.5. Role of Cumulative Effects in the GoM Hypoxic Zone

Cumulative Effect (from Table 3.1) Examples Relevant to Restoration Efforts to Reduce the Hypoxic Zone
Compounding Multiple wetland restoration projects upstream can generate additive nutrient reductions.
Triggers and Thresholds Once the temperature is warm enough, harmful algae blooms (HABs) grow and bloom.
Indirect, Secondary Effects Land use and agricultural practice changes that generate nutrient reductions have co-benefits in the form of carbon sequestration in soils and provision of habitat for pollinators, birds, and other wildlife—but some may have negative secondary effects too.
Landscape Pattern Restoration of a wetland in ideal topographical locations can act to catch water flowing from many agricultural fields, allowing nutrient recycling from a broad area.
Cross Boundary One watershed flows into the other. Nitrogen, phosphorous, and sediment flow downstream, but at different rates and paths.
Space Crowding Locations with multiple conservation practices in place, such as vegetative drainage ditches that feed into a retention pond, provide multiple opportunities for nutrients to settle or be cycled.
Time Lags Phosphorus moves more slowly than nitrogen in some systems, so while there are time lags associated with both nutrients, they should differ from each other.
Time Crowding One large rainfall event followed soon by another will have nonlinear erosion effects.

___________________

23 See https://www.epa.gov/ms-htf/history-hypoxia-task-force; https://www.epa.gov/ms-htf/looking-forward-strategy-federal-members-hypoxia-task-force.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×

Insight on this question comes from work done by the state of Iowa to document the extent of conservation actions that have been taken to address nutrient loss from the state to the Gulf.24 The Iowa Nutrient Reduction Strategy (INRS, 2017) estimated that over 90 percent of Iowa’s 22 million acres of row crop would need to be treated with conservation actions to reach the state’s goal of a 41 percent reduction in nitrogen and 29 percent reduction in phosphorus loads leaving the state. Currently, 3 percent of Iowa’s farmlands are treated. There is no evidence to suggest producers in other states have done more than Iowa farmers to address nitrogen losses from their fields.25 With this context in mind, it is not surprising that there is no measurable signal in the annual size of the hypoxic zone of these minor changes implemented to date.

A LONG-TERM VIEW

Traditional large-scale environmental management frameworks, derived from legal and regulatory requirements, were focused on remediating human activities with the capacity to harm whole regions. However, activities at this scale, such as restoration, may have positive effects on an ecosystem (Daoust et al., 2014; Diefenderfer et al., 2011, 2016, 2021). In light of the United Nations Decade on Ecosystem Restoration, which began in 2020, this type of widespread investment in restoration per se may become more commonplace (Aronson et al., 2020; Waltham et al., 2020). The assessment of cumulative restoration effects at regional or other intermediate scales depends on the ability to match the scale of improvements with the scale of detection methods.

The effort to maximize successful large-scale restoration efforts is a daunting task not only because of the normal complexities of these coastal ecosystems, but also because of trends in important chronic and acute influences on these ecosystems. As shown in the case studies, these complexities are present in the long-term and ongoing Chesapeake Bay Program and in the development of hypoxia in the GoM. Likewise, these complexities are present and have posed formidable challenges for the collective restoration investments across the Gulf Coast since the DWH. Approaches to meet the challenges detailed in this chapter include a framework of the modes of cumulative effects in ecosystems, causal criteria used in judgment, and the evaluation of analyses representing multiple lines of evidence. Currently, the comprehensive restoration framework on the Gulf Coast subsequent to DWH represents a unique and crucial opportunity to experiment with, develop, and demonstrate measurable restoration outcomes at a regional scale.

___________________

24 See https://store.extension.iastate.edu/Product/15915.

25 See https://www.epa.gov/ms-htf/report-nonpoint-source-progress-hypoxia-task-force-states.

Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 59
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 60
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 61
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 62
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 63
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 64
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 65
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 66
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 67
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 68
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 69
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 70
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 71
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 72
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 73
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 74
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 75
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 76
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 77
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 78
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 79
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 80
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 81
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 82
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 83
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 84
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 85
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 86
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 87
Suggested Citation:"3 Assessing Cumulative Effects of Restoration: Current and Emerging Approaches." National Academies of Sciences, Engineering, and Medicine. 2022. An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report. Washington, DC: The National Academies Press. doi: 10.17226/26335.
×
Page 88
Next: 4 Applications of Synthesis and Cumulative Effects Assessment in the Gulf of Mexico »
An Approach for Assessing U.S. Gulf Coast Ecosystem Restoration: A Gulf Research Program Environmental Monitoring Report Get This Book
×
Buy Paperback | $42.00 Buy Ebook | $33.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Valued for its ecological richness and economic value, the U.S. Gulf of Mexico is under substantial pressure from human activities. The Deepwater Horizon platform explosion and oil spill significantly damaged Gulf ecosystems and led to the largest ecological restoration investment in history. The unprecedented number and diversity of restoration activities provide valuable information for future restoration efforts, but assessment efforts are hampered by many factors, including the need to evaluate the interaction of multiple stressors and consider long-term environmental trends such as sea level rise, increasing hurricane intensity, and rising water temperatures.

This report offers a comprehensive approach to assess restoration activities beyond the project scale in the face of a changing environment. A main component of this approach is using different types of scientific evidence to develop "multiple lines of evidence" to evaluate restoration efforts at regional scales and beyond, especially for projects that may be mutually reinforcing (synergistic) or in conflict (antagonistic). Because Gulf of Mexico ecosystems cross political boundaries, increased coordination and collaboration is needed, especially to develop standardized data collection, analysis, synthesis, and reporting. With these improvements, program-level adaptive management approaches can be used more effectively to assess restoration strategies against the backdrop of long-term environmental trends.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!