A New Ecosystem-Based Cumulative Effects Assessment Framework to Enhance Strategic Environmental Assessment of Marine Spatial Plans

The scoping stage is crucial in determining the assessment’s success (Bragagnolo and Geneletti 2012; Canter and Ross 2014; Morrison-Saunders et al. 2014; Polido and Ramos 2015). It defines the geographical, temporal, and assessment scope for evaluating the plan alternatives and identifies key issues by consulting with relevant environmental authorities. This stage involves understanding how the plan will interact with other existing and future plans and projects in the area, the implications of their potential effects, including cumulative and transboundary effects, and how climate change may exacerbate these conditions. However, setting appropriate spatial and temporal boundaries and relevant methodologies to capture all potential effects has proven to be challenging (Therivel and Ross 2007; Bragagnolo and Geneletti 2012; Bidstrup et al. 2016). Also, the absence of a ‘systems perspective’ often leads to oversimplified assessments that fail to consider how changes to one species can affect an entire ecosystem (e.g., through food web effects), resulting in inadequate assessment components that do not fully address the (potential) risk of ecosystem effects over time (Gunn and Noble 2011; Declerck et al. 2022). EBA principles, which highlight ecosystem interconnectedness, offer the potential to make the SEA scoping stage, including scoping consultations, more holistic. These include the consideration of the full ecological integrity, biodiversity, functioning, interconnectedness, and resilience of the marine ecosystem (Table 1, EBA principle 2). The prioritization of ecological integrity in SEA can enhance its role in balancing environmental objectives with development, recognizing and safeguarding well-functioning and highly connected ecosystems (both habitat and species) which are central to supporting multiple sectors and ecosystem services. Similarly, EBA principles that advocate for the inclusion of all current and future marine uses and activities (including climate change uncertainties), considering their potential direct, indirect and interactive effects on resources and ecosystem services at appropriate scales, and their interactions, are key (EBA principles 6 & 11). Comprehensive consideration of all marine uses and activities, the pressures they generate, ecosystem components, and the wider ecosystem upon which they depend, is needed to meaningfully inform the baseline stage. Setting up appropriate spatio-temporal scales during the scoping stage is crucial to ensure appropriate data are identified and/or collected for relevant maps to be produced. Finally, this SEA stage is also crucial for identifying knowledge and data gaps and addressing any uncertainties related to climate change conditions which can contribute to the spatio-temporal variability of pressures and ecosystem components.

The baseline stage aims to provide a comprehensive picture of the current state of the environment and future trends (without plan implementation) focusing on the key issues identified during scoping. It requires examining ecosystem components across various topics such as biodiversity, human health, fauna, flora, water, air, and climate (CEC 2001). The resultant baseline forms the basis for selecting SEOs and assessing (cumulative) effects. Therefore, the baseline should offer detailed relevant information and data about high-sensitivity habitats and species (i.e., ecosystem components), including existing pressures from current plans, projects, or activities, and expected interactions with pressures resulting from the proposed plan (DHLGH 2022). Although the use of best-available knowledge and the precautionary principle are mandated by the SEA Directive, EBA principles can help reinforce this (Table 1, EBA principle 9). This principle emphasizes the need to transparently capture what is known/unknown, what needs to be known, and how to proceed in the face of uncertainty.

The baseline stage should be strongly informed by the scoping stage EBA principles on ecological integrity, comprehensive consideration of marine uses and activities, and climate change. This is crucial to ensure that SEAs move beyond the static representation of baseline conditions, commonly used to describe biodiversity in environmental assessment practice and conservation management (Partidário 2012; Gillon et al. 2016; Kovac et al. 2018), and move towards baselines that capture the fluid and dynamic nature of coastal and marine environments. This includes capturing the full range of ecosystem components, from static benthic habitats to highly mobile marine wildlife, and recognizing that baseline/current conditions encompass a wide range of pressures and impacts arising from both marine and land-based human activities. It also requires the consideration of potential interactions with additional future pressures within the portfolio of existing pressures affecting the ecosystem.

The MSFD lists key activities, pressures, and impacts on ecosystems that should be considered for management (CEC 2008), and valuable datasets have been produced to support these efforts (e.g., EEA 2020, OSPAR 2023a, b). Additionally, the EU Marine Observation and Data Network (EMODnet) portal (www.emodnet.eu) centralizes marine data from numerous organizations, focusing on human activities. These data improve marine baseline information and provide clearer insights into the spatio-temporal characteristics of pressures, which are often underrepresented in SEAs (González 2008; González et al. 2011). Also, the OSPAR Assessment Portal (https://www.ospar.org) provides thematic assessments that can provide valuable information on indicators and thresholds for different ecosystem components and pressures. Enhanced data sharing and utilization, including greater involvement of marine scientific expertise and local knowledge in SEA processes, are also crucial for building more comprehensive and accurate marine baselines in SEA practice. This also helps identify gaps in data and inform subsequent stages about the associated uncertainties and guiding future decisions towards a precautionary principle.

The wider scope of SEA often results in less detailed baselines, or a lack of specificity with regards to thresholds/tipping points needed for effective and robust assessments (Therivel and Ross 2007). To address this, some SEAs use SEOs to structure assessments, providing a framework for evaluating how a plan’s objectives align with or deviate from these objectives. While SEOs are not a formal requirement under the SEA Directive, they align with the Directive’s requirement to formulate environmental objectives and are widely adopted in Ireland and the UK (González 2008; Ravn Boess 2024). SEOs are developed for each environmental topic (e.g., biodiversity or human health) and are often derived from objectives, targets, and indicators established in local policies, and/or global strategies, such as Sustainable Development Goals (UN 2015). Nevertheless, SEOs often lack the specificity needed to effectively measure how a plan contributes to or detracts from environmental objectives over the long term (González 2008). The EBA principle (Table 1, EBA principle 1) that advocates for SMART—Specific, Measurable, Achievable, Realistic, and Time-bound—objectives linked to targets and indicators, can help address this by facilitating the denotation, spatial-specificity and trackability of SEOs.

Since SEAs set the rules for development consent (Therivel and Ross 2007), they must offer specific and realistic guidance on how future development should proceed to achieve these goals. Adopting SMART objectives can ensure that SEOs are more precise, targeted and efficient, particularly when they are closely aligned with the current state of the baseline. In this context, as noted by González (2008), spatial data from baseline maps can play a crucial role in shaping SEOs by incorporating spatial dimensions, and GIS can further assist by visualizing and prioritizing areas that need specific attention. For example, if an SEO focuses on achieving environmental objectives under the MSFD or WFD, the SEA should establish which particular objectives apply in which location/context. This ensures the marine baseline accurately reflects the various geographical areas within the assessment’s boundaries, including relevant pressures and ecosystem components thresholds or tipping points within the plan’s timescale. This approach would enable SEOs to be tailored not only to broader environmental goals, but also to the specific scope of the plan/program/policy, thus enhancing their relevance and making them more operational.

The impact assessment stage is arguably the most critical part of a SEA process, as it evaluates the potential effects of plans/programs/policies alternatives, followed by a more detailed examination of the preferred alternative (informed by scoping, baseline, and SEOs) to understand and guide the long-term use and management of the implementation area. Traditionally, SEA impact assessments have relied on SEO-based matrix-based methods (González and Geneletti 2021; DHLGH 2022). While matrices provide a clear overview at a strategic level, they fall short in capturing the complex spatio-temporal distribution of effects, essential for a robust CEA in SEA (González 2008; González 2012). Current SEAs have been limited in their ability to identify specific locations where effects might occur or their varying impacts on interconnected ecosystem components with particular sensitivities. These limitations highlight the need for improved environmental assessment methods that offer a more comprehensive, spatially-explicit, and quantifiable understanding of potential direct and indirect cumulative effects (González 2008; González et al. 2011).

The adoption of EBA principles, specifically principles 7 and 10 from Table 1, can help in addressing some of these methodological shortcomings. These principles focus on understanding ecosystem resistance, resilience, along with cumulative human impacts and environmental drivers, to predict ecosystem health. They also advocate for the use of risk assessment and management frameworks to identify and quantify social, economic, and ecological risks and trade-offs. Incorporating these principles can highlight the need for CEA methods that enhance the understanding of the specific sensitivity of different ecosystem components to pressures, including their unique behaviors and variations across spatial and temporal scales. The joint consideration of pressures, ecosystem components, and their interrelations is often insufficiently addressed in SEAs for MSP (Pinkau and Schiele 2021; Declerck et al. 2022), yet is essential for holistic assessments (Dubé et al. 2013). Risk-based approaches have been considered valuable for addressing these complexities and uncertainties and are also recommended by the MSFD (Brignon et al. 2022; Declerck et al. 2023). Therefore, it is crucial to explore how such methods could improve CEA within SEA practice. The following section further examines the specific characteristics of advanced CEA approaches towards better-informed planning decisions.

Most of the reviewed papers define their assessment scope using Halpern’s et al. (2008) mapping approach of activities/pressures and ecosystem components by identifying key components within a case study area through expert consultation. However, a key limitation of this method is its static nature, which overlooks dynamic aspects of cumulative effects in the marine environment, such as pressure dispersal across the water column and transboundary impacts—critical factors when setting boundaries for impact assessments. To address this, a limited number of studies have incorporated additional factors that account for the 4D dimensions of coastal and marine environments. For example, some studies use hydrodynamic modeling for pressures (e.g., Menegon et al. 2018a, 2018b), and others evaluate pressure dispersal as part of risk assessments (e.g., Piet et al. 2023). Such methods are particularly valuable in cases where there is no obvious spatial overlap between human activities and ecosystem components, but the pressures may still affect the ecosystem components. Time, however, remains a less explored aspect of cumulative effects (Willsteed et al. 2017). For instance, Battista et al. (2017) apply a temporally static model, potentially limiting its use in the context of evolving environmental changes, such as intensified climate change. While spatio-temporal variability of ecosystem components and pressures is often discussed and assessed, for instance through temporal overlap (i.e., when ecosystem components and pressures are likely to co-occur), the need for more robust models that accurately reflect ecosystem and climate change dynamics is increasingly recognized (Declerck et al. 2022, 2023).

Understanding human and environment interactions begins with a thorough assessment of activities/pressures and receptors/ecosystem components, commonly in consultation with marine experts (e.g., Battista et al. 2017; Piet et al. 2023). From a CEA risk assessment perspective, establishing linkages between these components is crucial (Judd et al. 2015; Muñoz et al. 2018; Stelzenmüller et al. 2018). Some of the reviewed methods support expert consultation with predefined lists of activities and ecosystem components from the MSFD and use visual tools to establish linkages or impact chains. For instance, they use network analyses such as the Drivers-Pressures-State-Impact-Response (DPSIR) framework (e.g., Robinson et al. 2014; Furlan et al. 2020), or Sankey Diagrams (e.g., Menegon et al. 2018a) to build linkages and graphically represent pathways and potential interactions. Given current knowledge limitations and the scarcity of pressure-impact and receptor-vulnerability spatial data available, interactions are challenging to identify. The majority of reviewed methods address only additive effects, as synergistic effects across multiple pressures remain difficult to evaluate, and antagonistic effects are often under-considered. However, some approaches, such as those by Battista et al. (2017), have begun to tackle these complexities by incorporating expert judgment to quantify synergistic and antagonistic effects through the use of an “additional threat modification factor” in pressure-receptor evaluations. Furlan et al. (2019) offer insights into spatial modeling of pressures by deriving GIS-based maps of spatial pressure intensity distribution, building on Furlan et al. (2018). They use a combination of ‘Choquet integral’⁴ techniques and advanced multicriteria decision analysis (MCDA) to spatially evaluate pressure interactions and estimate potential antagonistic and synergistic effects from multi-hazard interactions. Other methods use the number of links per pressure-receptor to indicate the likelihood of synergistic or antagonistic effects, but given the limited understanding of these interactions, the authors recommend interpreting outputs with caution (e.g., Piet et al. 2023).

Cumulative effects can result from a combination of direct and indirect impacts, with both linear and non-linear consequences on ecosystem components and ecosystems. Risk-based approaches assess the likelihood of these effects by evaluating pressure intensity and the specific sensitivity of an ecosystem component to such pressures (Stelzenmüller et al. 2010). Most reviewed methods (largely based on the Halpern et al. (2008) approach) consider factors such as resistance, recovery time, scale, frequency, and intensity of pressures in their assessments. More recent risk-based methods build on this foundation, using similar principles (though terminology may vary across methods, such as recovery time vs. resilience) while refining techniques to deepen insights into pressure-ecosystem component interactions and inform management. For instance, impact chains created through tools like DPSIR or Sankey Diagrams are evaluated with marine expertise, using categorical risk-based scoring systems from other studies to weigh the contribution of each link and prioritize management objectives on the greatest threats (e.g., Robinson et al. 2014). Other methods also incorporate additional parameters such as magnitude, hazard, and receptor behavior (e.g., ability of receptors to evade pressures), to provide more realistic sensitivity scores that reflect the unique vulnerabilities of ecosystem components to pressures (e.g., Furlan et al. 2019, 2020; Piet et al. 2023). Battista et al. (2017) further includes attributes like habitat connectivity and food web structure in sensitivity assessments to address direct and indirect impacts, as well as cascading effects. However, the focus of the majority of the reviewed papers remains on direct, linear, and negative impacts. Overall, assessing more complex indirect and non-linear impacts, such as regime changes (due to exceeding tolerance levels, thresholds, and/or tipping points), or food-web changes with cascading effects across various pressure-ecosystem interactions, remains a significant challenge.

Climate change is frequently treated as an additional pressure, using spatial layers such as sea surface temperature, ocean acidification, or UV radiation (Halpern et al. 2008). More recent approaches have investigated how climate change may interact with and/or amplify other pressures, including modeling multiple climate change scenarios over longer timeframes (Battista et al. 2017; Furlan et al. 2019). However, future scenarios of climate change and human development are inherently uncertain and hard to predict. Limitations in scientific data and knowledge underpinning cumulative effects interaction pathways and magnitude increase this uncertainty. Probabilistic models like Bayesian Networks are potentially valuable tools for adaptive marine management, as they can effectively incorporate related uncertainties (e.g., Furlan et al. 2020). Dynamic Bayesian Networks in particular can be especially useful for addressing uncertainties across different planning scenarios as they can incorporate temporal dimensions in the analysis (e.g., Furlan et al. 2020; Declerck et al. 2022).

Stelzenmüller et al. (2018) noted that GIS has been widely used to support spatially-explicit CEA. It remains essential for visualizing impacts (Depellegrin 2016), and combining it with tools like MCDA, Bayesian Networks, and expert input can significantly enhance CEA outputs (e.g., Furlan et al. 2019, 2020; Declerck et al. 2022). For instance, Furlan et al. (2020) demonstrated the effectiveness of integrating GIS’s visualization capabilities with the modeling and prediction functions of Bayesian Networks. Such GIS-based Bayesian Network framework supports evaluating the probability and uncertainty of cumulative effects under various ‘what-if’ scenarios, including different climate change conditions. Moreover, Furlan et al. (2019) and Declerck et al. (2022) have emphasized that combining GIS with Dynamic Bayesian Networks can be a powerful approach for evaluating cascading effects and spatio-temporal dynamics in marine environments. This combination can allow for the analysis of how relationships and impacts evolve over time, which is crucial for managing risks that vary across space and time. Declerck et al. (2022, 2023) also advocate for increased monitoring efforts to generate the spatial data required to enable their routine use for CEA.

The reviewed CEA methods have been predominantly developed for marine management and MSP, with limited focus on SEA and EIA, suggesting a disconnect between marine research and environmental assessment practice. While some methods claim to support SEA and EIA by centralizing relevant marine data (e.g., Menegon et al. 2018b), they primarily focus on specific MSP stages and do not elaborate on their application in SEA or EIA stages. Of the reviewed papers, the only ones specifically focused on enhancing CEA within SEA were Declerck et al. (2022, 2023). They stress the need for better alignment between CEA efforts in SEA and MSFD, advocating for an approach that integrates the ecosystem-based approach defined by OSPAR and the MSFD into SEA and EIA for a more holistic assessment framework at the European level.