A spatial decision framework for coastline restoration prioritization: integrating ecosystem dynamics, stability, and ecological capacity

Coastal protection and restoration are essential for achieving sustainable development in coastal regions facing intensifying climate and development pressures. However, under the implementation of coastline occupation-compensation balance policies, the lack of systematic approaches for identifying restoration priority areas remains a major challenge. We developed a multi-dimensional spatial decision framework to identify Priority Areas of Coastline Restoration (PACR) for the mainland coastline of Shandong Province, China, based on Landsat imagery (1984-2025, 30-m resolution). The framework integrates three key dimensions: Coastline Fragmentation Degree (CFD), Stability of Coastline (SC), and Coastal Ecological Carrying Capacity (CECC), enabling a comprehensive assessment of restoration priorities under heterogeneous coastal conditions. Our results revealed that: (1) CFD exhibited significant scale dependency, following a power-law scaling relationship with observational scale (R²>0.98), indicating that coastline fragmentation assessment is highly sensitive to spatial scale. (2) SC displayed pronounced spatial heterogeneity, with high- and extremely-high-stability areas accounting for 33.74% and 25.92% of the total coastline, respectively. (3) CECC showed a U-shaped temporal trajectory from 2010 to 2025, declining before 2016 and increasing thereafter, suggesting a shift in the balance between coastal development and ecological management. (4) The PACR framework identified southern Qingdao, northern Weifang, and eastern Rizhao as high-priority restoration areas, whereas the Yellow River Delta and other ecologically intact coastal zones showed relatively low restoration priorities. This study provides a spatially explicit decision-support framework for optimizing coastline restoration prioritization under occupation-compensation balance policies, offering new insights into ecosystem-based coastal management and climate-resilient shoreline governance.

Coastal zones represent one of the most dynamic interfaces between terrestrial and marine systems, providing critical ecosystem services—including carbon sequestration, storm surge mitigation, water purification, and habitat provision, while supporting intensive human activities and economic development (Barbier et al., 2011; Costanza et al., 2014; Ramesh et al., 2015). As prominent geomorphic boundaries between land and sea, coastlines integrate signals from both natural processes and anthropogenic disturbances. Their evolution reflects variations in sediment transport, hydrodynamic conditions, and geomorphological processes, while also recording the cumulative impacts of port construction, reclamation, and coastal urbanization (Zhu et al., 2018; Li et al., 2024a; Zhuang et al., 2024). However, rapid coastal development has substantially transformed natural coastline structures worldwide, resulting in habitat degradation, landscape fragmentation, and biodiversity loss (Dang et al., 2022; Yang et al., 2023), while also increasing exposure to climate-related hazards such as sea-level rise and intensifying storm surges (Neumann et al., 2015; Basnayake et al., 2026). Therefore, maintaining and restoring coastal ecosystems has become a critical challenge for achieving sustainable and climate-resilient coastal management.

China provides a representative example of intensive coastal transformation. Driven by rapid economic development and increasing demands for coastal space, extensive reclamation and infrastructure construction have profoundly reshaped coastal landscapes. From 1990 to 2020, China’s artificial coastline length increased from 10,022 km to 17,661 km, whereas natural coastline decreased from 20,019 km to 15,317 km, reducing the national natural coastline retention rate from 66.7% to 42.3% (Zhu et al., 2021; Li et al., 2023). Such large-scale conversion from natural to engineered coastlines has not only reduced coastal habitats but also disrupted landscape connectivity, altered hydrodynamic processes, and weakened ecosystem resilience (Cunha et al., 2021; Mishra et al., 2024). Consequently, restoring degraded coastlines has become essential for maintaining ecological integrity while accommodating sustainable coastal development (Liu et al., 2018).

To address coastline degradation, China has established a coastline occupation–compensation mechanism that requires ecological restoration to offset natural coastline losses caused by development activities (Zhang et al., 2024; Zhang and Que, 2026). However, unlike terrestrial land resources, coastlines represent complex ecological systems whose values cannot be adequately quantified by spatial length alone. Their ecological functions depend on habitat quality, landscape connectivity, geomorphological stability, and ecosystem service capacity (Li et al., 2024a). Therefore, length-based compensation does not necessarily achieve ecological equivalence between occupied and restored coastlines. In practice, coastline occupation-compensation remains largely constrained by administrative decision-making rather than ecological optimization (Zhang et al., 2025; Yu et al., 2026). Restoration sites are often selected according to local project requirements, with insufficient consideration of regional ecological patterns and long-term restoration performance. Consequently, high-value natural coastlines may be lost while compensation sites provide limited ecological benefits or fail to maintain restoration outcomes. This situation indicates that current occupation–compensation practices may achieve a numerical balance rather than an ecological balance, where restored coastline length matches the occupied length but ecological functions remain unequal (Jones et al., 2010; Zhou et al., 2023). The fundamental challenge has therefore shifted from determining how much coastline should be compensated to identifying where restoration can achieve ecological equivalence, physical persistence, and long-term sustainability.

Recent advances in coastal assessment and restoration science have provided important foundations for addressing this challenge. Yin et al. (2022) demonstrated that ecological carrying capacity assessment can be incorporated into territorial spatial planning, highlighting the importance of ecological constraints in spatial optimization. At the global scale, Basnayake et al. (2026) evaluated coastal vulnerability across 362,424 locations and identified geomorphology, tidal range, and coastal slope as dominant controlling factors, emphasizing the role of physical stability in restoration prioritization. Ren et al. (2026) analyzed 69 managed realignment projects worldwide and revealed that restoration success depends on the coordinated alignment of sediment supply, elevation conditions, and tidal creek connectivity. Liu et al. (2021) further demonstrated that sediment availability is a key determinant of nature-based restoration effectiveness. Furthermore, at the regional scale, Gan and Wang (2025) proposed a coastal ecological restoration zoning framework based on landscape ecological risk and ecosystem service value for Shanghai, while Shen et al. (2024) integrated landscape ecological risk with ecosystem service value to identify restoration zones in Fangchenggang. Despite these advances, existing studies have primarily focused on individual aspects of restoration—including ecological capacity, coastal vulnerability, or engineering feasibility—without integrating them into a unified spatial prioritization framework. A comprehensive framework that integrates coastline spatial structure, physical stability, and ecological capacity to support restoration site prioritization remains lacking.

To address this limitation, coastline restoration planning requires a function-oriented spatial decision framework that considers three fundamental requirements of successful restoration: reducing historical degradation, ensuring physical persistence, and maintaining long-term ecological sustainability. First, coastline fragmentation degree (CFD) characterizes historical degradation and spatial disconnection by quantifying natural coastline continuity and artificial coastline intensification (Li et al., 2018; Zhang et al., 2026). It identifies regions where restoration can effectively improve landscape connectivity and ecological integrity. Second, coastline stability (SC) evaluates restoration feasibility and persistence by incorporating coastline utilization intensity and multi-temporal coastline dynamics, reflecting whether restoration efforts can withstand future environmental and climate pressures (Zhao et al., 2023). Third, coastal ecological carrying capacity (CECC) defines the ecological boundary for sustainable restoration by integrating ecosystem service capacity, environmental tolerance, and anthropogenic pressures (Li et al., 2024b). Together, these dimensions establish a hierarchical decision pathway from degradation diagnosis (CFD), restoration feasibility assessment (SC), to long-term ecological sustainability evaluation (CECC), transforming coastline restoration from a length-based compensation approach toward an ecosystem-function-oriented optimization strategy.

Accordingly, this study develops a CFD–SC–CECC-based spatial decision framework to identify priority coastline restoration sites and support ecological-function-oriented coastline occupation–compensation. Specifically, this study aims to: (1) characterize the spatial and temporal evolution of coastline fragmentation and determine appropriate assessment scales using remote sensing and GIS techniques; (2) evaluate coastline stability based on coastline utilization intensity and multi-temporal coastline changes to identify physically feasible restoration areas; (3) construct a coastal ecological carrying capacity assessment model integrating multi-source environmental and socioeconomic data; and (4) integrate CFD, SC, and CECC to prioritize restoration locations based on ecological degradation, physical persistence, and sustainability considerations. This study provides a transferable spatial optimization approach for transforming coastline occupation-compensation from quantitative replacement toward ecological function equivalence, offering scientific support for adaptive, climate-resilient coastal governance in rapidly changing coastal regions.

Data source

Coastline extraction data source

We selected the Landsat 5 TM and Landsat 8 OLI/OLI-2 remote sensing images from 1984, 1990, 1995, 2000, 2005, 2010, 2015, 2020 and 2025 as the basic data (Supplementary Appendix 3.1.1 A). All images have a spatial resolution of 30 m with cloud cover below 10%. Following radiometric calibration, atmospheric correction, and geometric correction, the preprocessed images were used for coastline extraction and subsequent CFD and SC analyses. The data are from the United States Geological Survey (http://www.usgs.gov) and the Geospatial Data Cloud (https://www.gscloud.cn). http://www.gscloud.cn/.

In this study, the Mean High Tide Line (MHTL) was adopted as the indicator coastline. Meanwhile, based on the MHTL interpretation markers established by Hou Xiyong et al (Hou et al., 2016), as shown in Supplementary Appendix 3.1.1 B, the mainland coastline of Shandong Province from 1984 to 2025 was visually interpreted. Based on the existing classification system of coastlines and the distribution characteristics and structure of the mainland coastline in Shandong Province, this study divided the coastline into natural coastline, port and wharf coastline, reclamation coastline, aquaculture embankment, salt field embankment, transportation coastline and urban coastline. Based on this method, the theoretical maximum allowable interpretation error for 30-m resolution imagery is 28.28 m. A validation using 100 randomly selected sample points yielded a mean error of 10.14 m with a standard deviation of 21.44 m, meeting the accuracy requirements for coastline analysis at the provincial scale.

Data related to CECC

We established a hierarchical CECC assessment framework for Shandong’s coastal zone, structured across three criterion layers (Pressure-State-Response), four factor layers (environmental stress, demographic pressure, coastal resources, and economic development-social support), and twenty quantified indicators, with full methodological specifications detailed in our previous study (Li et al., 2024b).

Construction of coastline restoration priority assessment framework

PACR requires an integrated evaluation framework that simultaneously accounts for multiple characteristics of coastal systems. Here, we developed the Coastline Restoration Priority Index (CRPI) by integrating three complementary dimensions: CFD, SC and CECC. These dimensions were selected to characterize the spatial heterogeneity of coastline degradation, physical persistence, and ecological support capacity, respectively.

The spatial distribution of PACR was determined from the numerical characteristics of CRPI. The overall framework consisted of four major steps: (1) quantifying coastline fragmentation using CFD; (2) evaluating coastline stability based on coastal morphological changes and human utilization intensity; (3) constructing a multi-dimensional CECC evaluation system using environmental, resource, demographic, and economic indicators; and (4) integrating the three components to calculate CRPI and identify priority restoration areas.

Before integration, all indicators were normalized to eliminate dimensional differences. A weighted evaluation method was subsequently applied to combine CFD, SC, and CECC into the CRPI. Detailed calculation procedures for each component are described in Sections 3.2.1–3.2.4, and the overall workflow is illustrated in Figure 2.

Quantification of CFD

CFD was used to characterize the fragmentation characteristics of coastline spatial patterns. In this study, CFD was quantified based on the Index of Coastline Type Diversity (ICTD) and Patch Density (PD).

ICTD was calculated to represent the diversity characteristics of different coastline types within each assessment unit (Park and Guldmann, 2020; Hu et al., 2023), while PD was used to quantify the spatial distribution density of coastline patches (Cai et al., 2022). The two indicators were integrated according to the established methodology to derive the CFD value for each spatial unit, as expressed in (Equations 1–3):

Where CFD represents coastline fragmentation, with a value range of [0,1]. A i and L i represent the number of i-th coastlines and the length of i-th coastlines respectively. When there is only a single type of coastline in the study area, the value of CFD is 0. The larger the value of CFD, the higher the degree of fragmentation of the coastline in the study area.

Analysis of SC

The Stability of Coastline index (SC) reflects the capacity of a coastline to maintain its structural characteristics under the combined influence of natural processes and human activities. Since coastline stability is jointly influenced by external disturbance intensity and the resulting morphological changes, two indicators, the Index of Coastline Utilization Degree (ICUD) and Area Changes in the Coastal zone (ACC), were selected for SC quantification. Specifically, ICUD represents the intensity of human-induced disturbance, while ACC characterizes the spatial response of coastline evolution driven by erosion, accretion, and land reclamation.

1. ICUD calculation

Based on field investigations and expert knowledge, according to the differences in the degree of influence of human activities, we established the ICUD model (Mu et al., 2024), assigned different human activity intensity indexes to each type of coastline respectively (Table 1) (activity indices ranging from 1 to 4), and gave a comprehensive calculation formula for the degree of development and utilization of the coastline, as shown in (Equation 4).

Where ICUD denotes the degree of development and utilization of the coastline, L i and C i respectively represent the length of the i-th type coastline and the corresponding human force intensity index.

2. The changes in ACC caused by variations in the position of the coastline

The area enclosed by the coastline and land in 1984 and 2025 was compared to obtain the size of the area eroded or silted up along the coast. Since the analysis unit of this study was a 5 km coastline section (determined by the content of Section 4.1), several 5 km coastline sections of the 2025 coastline were spaced at intervals to draw profile lines perpendicular to the 1984 coastline. The area enclosed by two adjacent profile lines and the coastlines of the two years was the ACC along the 5 km coastline caused by the positional changes between 1984 and 2025. The ACC calculation for each coastline section is as shown in (Equation 5).

Where ACC is the area changes in the coastal of each coastline section. S 1 and S 2 respectively represent the expansion area and erosion area along the coast of a certain coastline section, and they are positive and negative respectively.

3. SC quantification

SC is quantified through algorithmic weighting of integrated measured by ICUD and ACC, yielding the composite stability index as formalized in (Equation 6).

Where SC denotes stability of coastline, The larger the value, the more stable the coastline is. ICUD and CCA are all classified into five grades. both scaled 1–5. Higher values denote lower development intensity or reduced area change. W 1 and W 2 respectively correspond to the weights of the above two, reflecting their complementary roles in characterizing coastline stability.

CECC quantification

CECC reflects the capacity of coastal ecosystems to support human activities while maintaining ecological functions. In this study, a multi-dimensional indicator system involving environmental pressure, coastal resources, population, and economic-social factors was constructed to evaluate the comprehensive state of coastal carrying capacity. Following the framework established in our previous work (Li et al., 2024b), the twenty indicators were first standardized according to their positive and negative attributes. Subsequently, the weights of each indicator were determined using the entropy method and analytic hierarchy process (AHP), and the state-space model was applied to calculate the CECC. Detailed calculation procedures can be found in our previous study (Li et al., 2024b). Owing to data availability constraints, CECC was evaluated at the prefecture-level city (municipal) scale. To integrate CECC with the segment-based CFD and SC, each 5-km coastline segment was assigned the CECC value of the city in which it is located.

The prefecture-level resolution of CECC reflects the administrative scale at which its underlying indicators are reported. Indicators such as population, GDP, environmental protection investment, and per capita water availability are compiled from statistical yearbooks and municipal bulletins that are only available at the administrative scale; disaggregating them to 5-km coastline segments is therefore neither feasible nor meaningful. This administrative-scale assessment is consistent with established practice in coastal carrying-capacity studies (Liu et al., 2020; Yin et al., 2022; Li et al., 2024b). Within the three-tier decision pathway, CECC functions as a city-level background constraint that modulates restoration priority across cities, whereas CFD and SC provide segment-level spatial discrimination; CECC is accordingly integrated as a continuous soft constraint (with a relatively low weight) rather than as a hard threshold, which avoids the arbitrariness of imposing an external cut-off value.

Multi-criteria spatial decision framework for PACR

Based on the quantified CFD, SC, and CECC, a multi-criteria spatial decision framework was developed to identify PACR. Firstly, the numerical values of CFD, SC, and CECC were classified into five levels using the natural breaks (Jenks) method, and corresponding scores were assigned (Table 2).

Subsequently, the three indicators were integrated through a weighted calculation to derive the CRPI, which characterizes the relative priority of coastline restoration. The detailed calculation formula is presented in (Equation 7). Finally, PACR were identified according to the spatial distribution characteristics of CRPI. The weights of CFD, SC, and CECC were determined using the entropy method, which calculates indicator weights based on the information contribution of each variable. (Equations 8–10) describe the specific calculation procedures.

It should be noted that the scoring direction is unified across the three indicators before aggregation: after classification, a higher grade score consistently denotes a lower restoration priority. Specifically, high CFD values (severe fragmentation) are assigned low scores, whereas high SC and CECC values (stable coastline and high carrying capacity) are assigned high scores (Table 2). Consequently, the additive weighted summation in (Equation 7) involves no polarity conflict, and a higher CRPI value unambiguously denotes lower restoration priority.

The relatively high weight of CFD (0.6) reflects not only its greater spatial variability but also its central role in diagnosing the fragmentation and artificialization of natural coastlines, which is the primary target of the occupation–compensation balance policy (Zhang et al., 2024); the lower weight of CECC (0.2) is consistent with its function as a city-level background constraint rather than a segment-level discriminator. The classification thresholds in Table 2 are based on the 2025 values of the three indicators (the most recent year), because the prioritization assessment reflects the current state of the coastline; by contrast, the multi-year mean CECC values reported in Section 4.3 serve two distinct purposes—tracing the temporal U-shaped trajectory and providing the multi-year samples required for the entropy-based weighting of the 20 CECC indicators.

where W CFD, W SC, and W CECC are the indicator weights determined by the entropy method (0.6, 0.2, and 0.2, respectively). Higher CRPI values denote lower restoration priority. P ij represents the proportion of the j-th indicator, X ij is the standardized result of the indicator, E j represents the information entropy of the j-th indicator, and W j reflects the weight of each indicator in the entropy method.

Scale effects and spatiotemporal characteristics of CFD

Scale effects in CFD

Our results demonstrate that the scale dependence of CFD is primarily driven by the pronounced influence of observation scale on the quantification of coastline fragmentation. Based on the length and structural composition of the mainland coastline of Shandong Province, seven observation scales (1, 5, 10, 15, 20, 25 and 30 km) were selected to quantify coastline fragmentation. Across all years from 1984 to 2020, CFD exhibited a robust power-law scaling relationship with observation scale (R²>0.980), where S denotes the observation scale, b is the scaling exponent characterizing the scale dependence of coastline fragmentation, and a is a constant. Increasing the observation scale consistently elevated CFD by incorporating more coastline types within individual segments, whereas the growth rate substantially slowed beyond 5 km (Figure 3). Notably, the 5 km scale captured 66% of homogeneous coastline segments, including 52% of segments within 1–5 km, enabling effective characterization of the dominant fragmentation patterns.

Accordingly, 5 km was identified as the optimal spatial scale for assessing mainland coastline fragmentation in Shandong Province and this scale was also selected as the basic spatial unit for subsequent coastline assessment.

Spatiotemporal characteristics of CFD along the mainland coastline of Shandong Province

Over the past four decades (1984-2025), the mainland coastline of Shandong Province exhibited a distinct phased evolution of CFD, characterized by an initial increase followed by a gradual decline, with values ranging from 0.2256 to 0.3195 (Figure 4). During 1984-2000, CFD increased rapidly from 0.2256 to the maximum value of 0.3195, indicating a substantial decline in coastline continuity. This increase was mainly associated with the extensive expansion of aquaculture embankments, reclamation activities, and port infrastructure in Binzhou, Weifang, and Qingdao, which transformed continuous natural coastlines into fragmented artificial coastal units. Since 2010, CFD has gradually decreased, reaching 0.2753 in 2025. This decline reflected improvements in coastal spatial management, including the optimization of coastline utilization, regulation of intensive aquaculture, and implementation of coastline protection policies and the occupation-restoration balance mechanism. These measures reduced uncontrolled artificial coastline expansion and contributed to the recovery of coastline integrity.

Spatially, CFD exhibited a persistent pattern of higher fragmentation in the eastern and southern coastal regions and lower fragmentation in the northwestern coast (Figure 5). The most significant increases in fragmentation were observed in southeastern Qingdao, eastern and southern Weihai, and southeastern Rizhao, where urbanization, port development, and intensive aquaculture strongly reshaped coastline morphology. Regionally, the Binzhou northern coast-eastern Dongying (Yellow River Estuary) maintained relatively low CFD values due to the dominance of natural muddy coastlines, with low-CFD areas accounting for 86.18% in 2025, although localized aquaculture and reclamation activities increased high-CFD areas by approximately 5% compared with 1984. The Yellow River Estuary-Northwestern Yantai (Longkou) region experienced the most pronounced fragmentation, with low-CFD areas decreasing from 62.30% to 27.63% and high-CFD areas increasing from 19.70% to 63.16%, driven by the conversion of reclaimed land into aquaculture ponds, salt-field embankments, and coastal development. Along the Longkou-southern Yantai coast, urban expansion, port construction, and bay aquaculture reduced low-CFD areas by 10.81% and increased high-CFD areas by 8.22%. Meanwhile, high-CFD zones along the southern Yantai-Rizhao coast were mainly concentrated around Jiaozhou Bay and Rizhao Port, where rapid artificial coastline expansion associated with urbanization disrupted natural coastline continuity, increasing high-CFD areas by 6.06%. Overall, the spatiotemporal evolution of CFD reflects the transition of Shandong’s coastal development from intensive exploitation toward spatial optimization and ecological conservation.

Analysis of SC in Shandong Province

The numerical characteristics and spatial distribution of the ICUD in Shandong Province in 2025, as well as the ACC and SC of each coastline section from 1984 to 2025 are shown in Figures 6, 7.

In 2025, the overall ICUD of the Shandong mainland coastline exhibited a high level of human utilization intensity, with high-value (14.23-17.48) and extremely high-value (17.48-20.00) areas accounting for 27.68% and 32.45% of the total coastline, respectively (Figure 6a). At the municipal scale, most coastal cities, except Dongying and Weihai, displayed a similar distribution pattern to the provincial average, characterized by the predominance of high and extremely high ICUD classes. Spatially, extremely high ICUD values were mainly concentrated along the northern coast of Weifang, the northwestern coast of Yantai, and the southern coast of Qingdao. This pattern was closely associated with intensive coastal utilization activities, particularly the expansion of aquaculture ponds, reclamation structures, and port infrastructure. In contrast, the eastern part of Dongying maintained relatively low ICUD values due to the extensive distribution of deltaic wetlands with high ecological connectivity and a relatively intact natural coastline (Figure 7a). Weihai exhibited a more balanced distribution among different ICUD classes, reflecting the spatial differentiation of coastal functional management. Specifically, the southwestern and northeastern coastal zones have prioritized ecological conservation and natural coastline maintenance, whereas the eastern and southeastern sectors have experienced more intensive development activities, resulting in a spatial pattern characterized by the coexistence of conservation and utilization.

At the ACC level, the Shandong mainland coastline was dominated by extremely low ACC values (0.00-5.44) in 2025, accounting for 83.61% of the total coastline (Figure 6b), indicating that most coastal segments experienced limited seaward expansion during the study period. Except for Binzhou and Dongying, all other coastal cities exhibited distribution patterns consistent with the provincial-scale characteristics. Binzhou showed pronounced spatial heterogeneity in ACC, with coastal segments primarily classified into low (5.44-15.91) and moderate (30.09-54.01) levels. The high-value and moderate-value areas accounted for 30.53% and 24.42%, respectively, and were mainly concentrated along the northeastern coast of Binzhou (Figure 7b). The elevated ACC values in this region were closely associated with the seaward expansion of artificial coastal structures, particularly aquaculture embankments. In Dongying, moderate ACC values were mainly distributed along the eastern coast, reflecting localized coastline advancement within the deltaic environment. Furthermore, the most pronounced seaward expansion occurred along the northern coast of Binzhou, eastern coast of Dongying, and northern coast of Weifang, with a cumulative expansion area of approximately 2050 km².

At the SC level, the Shandong mainland coastline exhibited SC values ranging from 1 to 5 in 2025. Moderate (2.5-3.0), high (3.0-4.0), and extremely high (4.0-5.0) SC classes dominated the coastline, accounting for 30.25%, 33.74%, and 25.92%, respectively (Figure 6c), indicating an overall high degree of coastline stability. Spatially, high SC values were mainly concentrated in the Yellow River Delta (east Dongying), southern Yantai, and northern and southern Weihai coasts (Figure 7c), whereas low-value areas were primarily distributed in northeastern Binzhou, northeastern Dongying, and northern Weifang. Spatial autocorrelation analysis further revealed significant clustering patterns of SC hotspots, cold spots, and corresponding secondary hotspot/coldspot areas (Figure 7d), suggesting that SC exhibits strong spatial dependence rather than a random distribution. These patterns likely reflect the combined influence of regional geomorphological settings, natural environmental conditions, and long-term coastline evolution processes.

Spatio-temporal characteristics of CECC and subsystem carrying capacity

Based on the CECC framework, the spatiotemporal evolution of coastal ecological carrying capacity (CECC) and its four subsystems in Shandong Province from 2010 to 2025 is illustrated in Figure 8, with annual variation rates summarized in Supplementary Appendix 4.3. The results reveal pronounced stage-dependent dynamics and spatial heterogeneity in the ecological carrying capacity of the Shandong coastal zone.

Using the mean CECC values of the seven coastal cities to represent the provincial-scale condition exhibited a distinct U-shaped trajectory during 2010-2025, with values ranging from 0.481 to 0.679. The year 2016 represented a critical transition point in the long-term evolution of CECC. During 2010-2013, CECC experienced the most pronounced decline, with an annual change rate of −15.34%. After 2016, CECC gradually recovered, exhibiting annual growth rates of 10.75% and 18.10% during 2016–2019 and 2020-2025, respectively. These patterns indicate a transition in the ecological carrying state of the Shandong coastal zone from a phase of increasing pressure accumulation toward a phase of recovery and improvement.

At the subsystem level, different dimensions of carrying capacity exhibited distinct evolutionary patterns. From 2010 to 2025, the cumulative change rates of environmental carrying capacity (ECC), population carrying capacity (PCC), coastal resource carrying capacity (CRCC), and economic-social carrying capacity (ESCC) were −30.35%, −44.42%, 81.99%, and 24.10%, respectively. Among these subsystems, the continuous enhancement of ESCC provided an important support for CECC recovery, whereas the declines in ECC, PCC, and CRCC reflected persistent pressures on coastal environmental and resource systems. Specifically, except for PCC, all other subsystems declined during 2010-2013, while ECC, CRCC, and ESCC gradually improved between 2016 and 2020, suggesting that ecological restoration and improved resource-use efficiency contributed positively to the enhancement of CECC.

Further analysis of long-term variations showed that the changes in overall ECC, PCC, CRCC, and ESCC in Shandong Province during 2010–2025 were −0.101, −0.175, 0.189, and 0.081, respectively. These results demonstrate divergent responses among different subsystems, with resource carrying capacity and socioeconomic support capacity showing increasing trends, while environmental and population-related pressures remained evident.

At the municipal scale, most coastal cities exhibited evolutionary patterns consistent with the provincial trend, characterized by a U-shaped trajectory and minimum CECC values around 2016, indicating that this transition point was regionally consistent. However, substantial spatial disparities existed among cities. The average CECC values during the study period followed the order of Weihai > Qingdao > Yantai > Dongying > Rizhao > Binzhou > Weifang. At the subsystem level, cities showed differentiated evolutionary characteristics. ECC declines were most pronounced in Yantai and Rizhao, with changes of −0.211 and −0.175, respectively. Except for Yantai, PCC decreased in all coastal cities, with changes ranging from −0.164 to −0.214. Meanwhile, Weihai exhibited the most substantial improvements in CRCC and ESCC, with increases of 0.382 and 0.107, respectively.

Overall, the ecological carrying capacity of the Shandong coastal zone experienced a transition from an initial decline associated with accumulated pressures during rapid coastal development to a recovery phase characterized by ecological restoration, optimized resource utilization, and enhanced socioeconomic coordination. Nevertheless, considerable spatial heterogeneity remains due to differences in resource endowments, industrial structures, and development intensity among coastal cities.

PACR identification in Shandong Province based on multi-criteria spatial decision-making model

The overall spatial characteristics of PACR in Shandong Province

Based on the spatial distribution characteristics of CRPI, the PACR of the Shandong mainland coastline was classified into five levels using the natural breaks method: extremely high (1.4-2.0), high (2.0-2.7), moderate (2.7-3.5), low (3.5-4.2), and extremely low (4.2-5.0). The lengths and proportions of different PACR classes along the Shandong coastline in 2025 are summarized in Table 3.

The results showed that PACR was dominated by the extremely low class, accounting for 28.12% of the total coastline, whereas extremely high PACR segments represented only 7.52%. This indicates that areas with severe coastal risk conditions were spatially concentrated rather than widely distributed. Based on the PACR assessment of 546 coastal segments along the Shandong mainland coastline (Figure 9), substantial spatial differentiation was observed among different PACR categories. Spatially, extremely low PACR areas were mainly distributed in the eastern Dongying (Yellow River Delta), northwestern Yantai, and southern Weihai. The Yellow River Delta, protected as a national nature reserve, has maintained extensive natural coastline and ecological continuity, with relatively limited anthropogenic disturbance, resulting in lower restoration priority. In contrast, the Jinghai Bay area in southern Weihai has undergone long-term planning and intensive management of aquaculture-related coastline utilization, forming a relatively stable coastal utilization pattern and reducing ecological pressure on coastline systems. In comparison, extremely high PACR areas were primarily concentrated in Jiaozhou Bay, southeastern Qingdao, and eastern Rizhao. As a strategically important bay in northern China, Jiaozhou Bay supports intensive port activities, industrial development, and urban expansion across both terrestrial and marine spaces. These intensive anthropogenic pressures have resulted in high coastline artificialization and fragmentation, contributing to elevated restoration priority.

Spatial characteristics of PACR at the municipal scale

At the municipal scale, the spatial distribution of PACR is shown in Figure 10. Significant spatial disparities were observed among coastal cities in Shandong Province. Binzhou, Dongying, Yantai, and Weihai were predominantly characterized by moderate or lower PACR levels, whereas other coastal cities generally exhibited moderate or higher levels. High and extremely high PACR zones were mainly distributed along northern Weifang, southern Qingdao, and eastern Rizhao, indicating regions with higher restoration priority.

Extremely low PACR areas accounted for more than 32% in Binzhou, Dongying, and Weihai. These regions are characterized by favorable natural resource conditions, relatively high coastal ecological carrying capacity, and limited coastline fragmentation, providing advantageous conditions for ecological restoration and natural recovery. In contrast, Qingdao and Rizhao were dominated by high and extremely high PACR classes. As major economic centers along the Shandong coast, these cities have experienced intensive port development, industrial expansion, and frequent land-sea economic activities. Long-term anthropogenic pressures, including coastline artificialization, land reclamation, and increased pollutant inputs, have imposed substantial stress on coastal ecosystems, highlighting these areas as important targets for coastline restoration.

Regarding spatial extremes, the maximum CRPI values in all seven coastal cities exceeded 4.2. Areas with CRPI values of 5.0 were mainly concentrated in northern and southwestern Weihai, where continuous bedrock coastlines or sandy coastlines coexist with high CECC values, indicating strong ecological stability. Conversely, the minimum CRPI values (1.4) occurred in the western and northeastern Jiaozhou Bay and eastern Laoshan areas. Due to intensive terrestrial and marine development, these regions have experienced substantial degradation of natural ecosystems and were identified as priority areas for coastline restoration. At the city-average scale, Dongying exhibited the highest CRPI value, reaching 4.1. This pattern is primarily attributed to the presence of the Yellow River Delta National Nature Reserve, where relatively weak anthropogenic disturbance and effective ecological protection have maintained high ecosystem integrity in the coastal zone.

Spatial hotspot analysis of PACR in Shandong Province

The Getis-Ord Gi* statistic was applied to characterize the local spatial clustering patterns of PACR along the Shandong mainland coastline, identifying statistically significant clusters of high and low CRPI values and providing spatial insights for coastal restoration planning (Figure 11). This approach enables the differentiation of significant hotspots and coldspots, thereby supporting the identification of spatially continuous restoration priority zones.

The results showed that high-CRPI hotspots (99% confidence level; corresponding to low restoration-priority segments) were primarily concentrated in the Yellow River Delta and the northeastern and southeastern coastal areas of Weihai, whereas secondary hotspots (90-95% confidence levels) were mainly distributed around these core hotspot regions, forming continuous spatial clusters. This pattern indicates that low-priority (high-CRPI) segments were not randomly distributed but tended to occur in groups under similar environmental settings and coastline utilization conditions. In contrast, low-CRPI coldspots (99% confidence level; corresponding to high restoration-priority segments) were mainly located along the southern and southeastern coasts of Qingdao, with secondary coldspots (90-95% confidence levels) surrounding these core areas. These regions generally corresponded to areas with intensive anthropogenic activities, high coastline artificialization, and elevated ecological pressures, resulting in concentrated high-priority (low-CRPI) clusters.

Overall, both high-CRPI hotspots and low-CRPI coldspots exhibited significant spatial aggregation, demonstrating strong spatial dependence along the Shandong mainland coastline. The Getis-Ord Gi* analysis further suggests that coastal restoration priorities are spatially interconnected rather than confined to individual coastline segments. Therefore, restoration planning should move beyond isolated site-based interventions and adopt spatially coordinated strategies targeting continuous hotspot-coldspot clusters to enhance regional ecological restoration effectiveness.

Building on these cold-spot clusters, continuous ‘restoration priority belts’ can be delineated by buffering the core cold spots (99% confidence level) and merging adjacent clusters; these belts can then be reconciled with administrative boundaries and marine functional zoning to assign restoration responsibilities and align interventions with the ‘one-map’ territorial spatial plan.

Mechanisms shaping spatial patterns of PACR

The identification of PACR requires explicit consideration of scale effects, particularly because coastal fragmentation and coastline configuration are inherently scale-dependent. Our results demonstrate that the quantification of coastline fragmentation varies substantially with observational scale, which is primarily constrained by three interacting factors: (1) the natural continuity of coastlines shaped by regional topography and geomorphological settings; (2) the spatial organization of production and utilization spaces driven by human activities; and (3) administrative boundary constraints associated with coastal management units (Chang et al., 2018; Toledo et al., 2025). Therefore, restoration prioritization based on a single fixed spatial scale may overlook critical spatial heterogeneity. A multi-scale assessment framework integrating natural geographical units and socio-administrative boundaries is essential for aligning ecological restoration objectives with practical governance requirements.

Existing studies have progressively advanced our understanding of coastal ecological constraints, physical vulnerability, and restoration feasibility. Ecological carrying capacity assessments have underscored the importance of ecological thresholds in coastal spatial planning (Yin et al., 2022), while vulnerability assessments have highlighted the role of geomorphological characteristics and environmental exposure in determining coastal risks (Basnayake et al., 2026). Furthermore, restoration-oriented studies have demonstrated that sediment dynamics, elevation conditions, and tidal connectivity strongly influence restoration effectiveness (Ren et al., 2026; Liu et al., 2021). However, these approaches generally focus on individual dimensions of coastal systems and provide limited support for spatially explicit restoration prioritization. By integrating coastline spatial configuration, coastal change characteristics, and ecological carrying capacity, the PACR framework developed in this study extends previous assessments from evaluating coastal conditions toward identifying where restoration actions should be prioritized, thereby bridging a critical gap between assessment and decision-making.

Although our multi-criteria spatial decision model improves the comprehensiveness of PACR identification, it also reveals the inherent complexity of interpreting restoration priorities across heterogeneous coastal settings. For example, Yantai Port and Qingdao Port have undergone extensive reclamation and coastline modification historically; however, their current land-use patterns have become relatively stable through functional transformation. Therefore, restoration priorities in these areas should not be determined solely by historical disturbance intensity, but should also consider current ecological conditions and future management objectives. Conversely, naturally stable bedrock coasts along the Jiaodong Peninsula, despite their high inherent geomorphological stability, have experienced increasing fragmentation due to scattered human interventions, underscoring the need for targeted ecological management even in geomorphologically resilient settings. Furthermore, PACR delineation should be integrated with territorial spatial planning and marine functional zoning to ensure compatibility between restoration interventions and long-term regional development strategies.

Coastal evolution is jointly regulated by natural processes and anthropogenic pressures (Dang et al., 2022; Hu and Wang, 2022; Zhu et al., 2023). Analysis of the CECC indicator system revealed that socioeconomic factors—including per capita GDP, environmental protection investment, and passenger transport volume, exhibited stronger explanatory power for CECC variations than natural factors. Meanwhile, natural conditions such as per capita water availability, vegetation net primary productivity, and nearshore water quality provided essential ecological support. These findings suggest that human activities have become dominant regulators of coastal ecological carrying capacity in highly developed coastal regions. Moreover, different coastline types exhibited distinct responses to anthropogenic pressures (Figure 12). Muddy coasts generally maintained lower fragmentation levels due to their continuous geomorphological characteristics and relatively limited structural disturbance, whereas bedrock coasts showed the highest fragmentation intensity because of localized development activities. Sandy coasts exhibited intermediate responses. This differentiation underscores the importance of shifting from uniform, disturbance-oriented management toward adaptive, scale-dependent, and ecosystem-based coastal governance tailored to specific coastline types.

Coastline zoning restoration strategies based on PACR

Nature-based Solutions (NbS) emphasize the central role of ecosystem functions and natural processes, which is highly consistent with the ecological principles underlying the PACR identification framework (Seddon et al., 2020; Unguendoli et al., 2023). Therefore, integrating NbS with PACR-based coastal zoning provides an effective pathway for achieving sustainable coastal development and enhancing coastal climate resilience. According to the spatial distribution characteristics of PACR, differentiated restoration strategies should be implemented according to regional ecological conditions and anthropogenic pressures.

High-PACR areas—exemplified by the southern and southeastern coastal zones of Qingdao—are characterized by high coastline fragmentation and relatively low CECC (Figure 8). These areas warrant prioritized, proactive restoration. Key interventions include restoring salt-tolerant vegetation communities in intertidal zones to enhance biodiversity and establish natural wave attenuation systems (Lennon and Sealey, 2022), improving nearshore habitat conditions through ecological enhancement measures, and reconstructing hardened coastlines using ecological seawall approaches to facilitate land-sea material exchange and habitat recovery (Suedel et al., 2022). Meanwhile, spatial zoning of intensive development areas and ecologically valuable coastal zones is essential for balancing human utilization and ecosystem conservation through spatial segregation, functional complementarity, and adaptive management (Leone and Zoppi, 2019).

In contrast, low-PACR areas within protected ecosystems—such as eastern Dongying— exhibit high SC and CECC but low CFD, indicating relatively intact coastal ecosystem structures and limited anthropogenic disturbance. Restoration strategies in these regions should emphasize strict conservation and natural recovery rather than intensive engineering intervention. These areas should be incorporated into ecological conservation redlines, while ecological corridors and regional connectivity should be maintained to safeguard long-term ecosystem stability (Wang et al., 2025). In medium-PACR areas, complex interactions between coastal erosion, habitat fragmentation, and human disturbance require adaptive restoration approaches. Priority should be given to restoring key ecological nodes and enhancing natural resilience with minimal intervention. Measures such as ecological seawalls in erosion-prone sections, removal of unnecessary barriers, and restoration of salt-tolerant vegetation corridors can improve coastline connectivity and ecological functions (MacArthur et al., 2019). Furthermore, PACR-based indicators can support refined coastline management by establishing occupation-control and ecological compensation mechanisms.

It is worth emphasizing that the low PACR assigned to the Yellow River Delta reflects the integrity of its shoreline morphology (low CFD, high SC, and high CECC) and therefore a low priority for morphological restoration. This does not conflict with the delta’s status as a national priority for wetland conservation and functional restoration, which operates on a distinct restoration dimension—wetland ecosystem functions rather than shoreline morphology—and is thus outside the scope of the PACR framework.

The implementation of coastal restoration nevertheless remains constrained by socioeconomic trade-offs and governance challenges. Local governments must balance ecological objectives with economic costs while increasing stakeholder participation (McAfee et al., 2021; Fan et al., 2023). For example, the removal of artificial structures or relocation of aquaculture facilities may encounter resistance due to conflicts with local livelihoods and development demands (Hynes et al., 2018). In addition, certain artificial structures, including ports and wharves, have become integrated into existing coastal systems, making selective retention a more practical option than complete removal (Dafforn et al., 2015; Portz et al., 2024). Given the pronounced spatial aggregation of PACR, restoration planning should explicitly account for spatial dependence and extend interventions beyond individual coastline segments to encompass adjacent coastal zones as integrated management units. Encouraging public participation, strengthening coastal ecological awareness, and establishing long-term institutional incentives can transform communities into continuous contributors to restoration actions (Czajkowski et al., 2015; Liu et al., 2025), ultimately enhancing the overall resilience and functional integrity of coastal ecosystems.

The PACR framework can be directly translated into the operational mechanisms of the coastline occupation–compensation balance policy. High- and extremely-high-priority segments (e.g., Jiaozhou Bay, southern Qingdao, and eastern Rizhao) can be designated as candidate ‘offset supply zones’, where restoration is most likely to yield ecological-recovery shorelines that satisfy the like-for-like and additionality principles of ecological compensation (Yu et al., 2017). Extremely-low-priority segments within ecologically intact areas (e.g., the Yellow River Delta) should be designated as ‘no-occupation zones’ and safeguarded under natural-coastline retention targets and marine ecological redlines. Provincial regulations already provide differentiated compensation ratios—for example, Guangdong Province requires 1:1.5 restoration for occupied natural shoreline when the retention rate falls below the control target, and Hebei Province applies 1:1.1–1:1.2 ratios under a ‘like-for-like or better’ rule—and the PACR framework can inform the spatial allocation of these compensation sites, shifting site selection from administrative discretion toward ecological-function-oriented optimisation.

Limitations and future perspectives

This study has several limitations that should be acknowledged and addressed in future research. First, although this study developed a PACR spatial decision framework by integrating historical coastline changes, current ecological conditions, and restoration-related considerations, uncertainties remain due to data availability and indicator representation. Several indicators were derived from statistical yearbooks and regional datasets, which may introduce uncertainties at finer spatial resolutions. Additionally, the coastline extraction accuracy (mean error: 10.14 m; standard deviation: 21.44 m for 30-m imagery) may propagate uncertainty into the CFD and SC calculations, particularly for short coastline segments. Furthermore, although the selected indicators represent major factors influencing PACR, additional variables associated with coastal processes and climate-driven processes—including sea-level rise projections, storm surge frequency, and long-term coastline retreat rates—could further improve the comprehensiveness of future assessments.

Second, the PACR framework was primarily designed for provincial-scale coastal assessment. When applied to smaller units such as county-level coastal zones, the indicator system and weighting scheme should be recalibrated according to local geomorphological characteristics, data availability, and socioeconomic conditions. Therefore, future applications should incorporate region-specific indicators and recalibrate model parameters to enhance the applicability of PACR across different coastal environments.

Despite these limitations, this study provides a spatially explicit framework for supporting coastal restoration prioritization and sustainable coastline management. By integrating coastline morphological characteristics, coastal stability, and the CECC assessment system encompassing socioeconomic conditions and ecosystem functions, the proposed framework enables a comprehensive evaluation of natural processes, ecological capacity, and anthropogenic pressures. This integrated approach provides scientific support for implementing coastline occupation-compensation balance policies and improving adaptive coastal governance (Axon et al., 2024). Future research should prioritize three directions: (1) incorporating climate-related variables—including sea-level rise projections, extreme storm events, and long-term hydrodynamic changes—to enhance predictive capability (Wang et al., 2024); (2) validating the framework through case studies in other coastal provinces with contrasting conditions; and (3) developing dynamic, real-time PACR assessment tools that integrate continuous remote sensing monitoring with periodic updates of socioeconomic indicators.

This study developed a spatial decision framework for identifying coastal restoration priorities along the mainland coastline of Shandong Province by integrating GIS-based spatial analysis and multi-criteria evaluation, which provides a scientific foundation for adaptive coastal management and sustainable coastline governance. The PACR framework advances coastline occupation–compensation practice by shifting the focus from length-based quantitative replacement toward ecological-function-oriented spatial optimization. The results revealed that coastline fragmentation exhibited significant scale dependence, with a power-law relationship between fragmentation degree and observation scale during 1984-2025. Spatially, coastline fragmentation was higher in the eastern and southern coastal regions but lower in the northwestern coast, reflecting the combined effects of coastal geomorphology, nearshore hydrodynamic conditions, and anthropogenic disturbances. Coastline stability showed evident spatial heterogeneity, with high-stability areas mainly distributed in eastern Dongying, southern Yantai, and northern Weihai, whereas low-stability zones were concentrated in northeastern Binzhou, northeastern Dongying, and northern Weifang. In addition, the ecological carrying capacity of the Shandong coastal zone exhibited an overall U-shaped temporal trajectory, with 2016 representing a key transition point. The restoration priority assessment identified significant spatial differentiation and autocorrelation, with higher-priority restoration areas mainly located in northern Weifang, southern Qingdao, and eastern Rizhao, while the Yellow River Delta, northwestern Yantai, and southern Weihai showed relatively lower restoration priority due to ecological constraints and geomorphological characteristics. Overall, this study demonstrates that coastal restoration planning should consider the coupled effects of coastline morphology, ecological processes, human activities, and spatial scale dependence, and provides a scientific basis for adaptive, climate-resilient coastal management and restoration prioritization in rapidly changing coastal regions.