Introduction

Marine benthic habitats are physically distinct areas of seabed that are inhabited by particular organisms1. They serve as vital links between productive shallow environments and the abyssal seabed, transport organisms and lithogenic materials, and thus play important roles in key ecosystem processes as well as providing valuable ecosystem services2,3,4. Therefore, it is crucial to have good scientific knowledge of marine benthic habitats.

However, accurately mapping these habitats presents substantial challenges due to the vastness and complexity of marine environments, compounded by technological limitations and high resource demands. Currently, only 5–10% of the seabed can be mapped at resolutions comparable to those on land5. In response to these challenges, the Marine Strategy Framework Directive (MSFD), issued by the International Council for the Exploration of the Sea (ICES), identifies marine habitat maps as essential for marine management and research. This directive is driven by the urgent need for effective management actions, such as benthic habitat protection, marine spatial planning, and the assessment of human impacts on the ocean6,7,8. The significance of marine habitat mapping is increasing recognized globally, with notable attention in regions such as Europe7,9,10, Asia6,11, and Australia12.

Benthic habitats are typically characterized by essential abiotic variables and distinct biotic communities. Two primary approaches are commonly employed to map these habitats: the top-down and bottom-up approaches. The top-down approach utilizes side-scan sonar, multibeam sonar, and other acoustic tools to categorize habitat types across large spatial scales. While this method efficiently and accurately maps habitat extents, it primarily focuses on variations in abiotic features13,14. In contrast, the bottom-up approach relies on in situ sampling data to synthesize habitat types via cluster analysis15. Although this method captures finer biotic variations at smaller spatial scales, it may fall short in terms of accuracy and spatial coverage16. Recently, a combined approach has gained traction among researchers, leveraging the broad coverage and stable, accurate boundaries of the top-down approach along with the distinct biotic features of the bottom-up approach10,17. This integrative approach offers the best estimate of benthic habitats based on all available knowledge at a certain time18,19.

In China, substantial ecological data have been gathered through various marine monitoring efforts, particularly since the 1950s20. However, much of the existing research focuses on specific biological communities or localized habitats, often lacking a comprehensive classification framework that encompasses the entire ecosystem. Additionally, the absence of detailed coastal and marine habitat maps limits the practical applicability of findings for management and conservation purposes. To address these challenges, this study presents a systematic classification scheme for marine benthic habitats, based on an integrative analysis of both biotic and abiotic features. By utilizing extensive data on topography, geology, hydrology, chemistry, biology, and anthropogenic activities, we have produced the first detailed marine benthic habitat map of Bohai Bay, China. This study addressed the following questions: (1) Which biotic and abiotic features best represent benthic habitats, and (2) How can benthic habitat maps support management decisions for ecological conservation and benthic resource utilization? The results aim to provide practical cases and insights for the promotion and application of benthic habitat mapping in China.

Study area

Bohai Bay is a semi-enclosed bay located in the western part of the Bohai Sea in China, with an average water depth of approximately 12.5 m21. For the purposes of this study, the Bohai Bay is defined by the coordinates 117°30ʹ to 119°E and 37°58ʹ to 39°12ʹN, following the ecological zoning scheme established by Huang22. This area encompasses approximately 11,000 square kilometers (Fig. 1).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Study area location and spatial distribution of samples.

The bay is primarily characterized by a gulf accumulation plain geomorphic type, with the seabed terrain that slopes from the periphery towards the center. Throughout the year, hypersaline water from the northern shore enters into the bay, where it transitions into hyposaline water before exiting along the southern shore. This hydrological process results in a counterclockwise current within the bay23.

Additionally, numerous rivers along the coast transport significant quantities of sediment into the bay. These sediments accumulate, forming extensive silty mudflats that provide habitats for a wide variety of benthic organisms and migratory seabirds24,25. Notably, Bohai Bay is also home to the largest living oyster reef community in China, which is attributed to the retreat of the sea approximately 2500 years ago26.

Methods

Habitat classification scheme

Developing a benthic habitat classification scheme is a fundamental step in benthic habitat mapping, providing a structured framework for organizing and interpreting ecological data. Globally, several schemes have been proposed, including the IUCN Global Ecosystem Typology27, the Coastal and Marine Ecological Classification Standard (CMECS) in the USA28, and the European Nature Information Service (EUNIS)29. These schemes are widely recognized for their hierarchical structure, organizing habitat units into nested levels30. Higher levels typically rely on broad-scale remote sensing or sonar-derived data to reflect abiotic features, while finer levels incorporate localized in-situ data to capture biotic features.

In this study, a two-level benthic habitat classification framework was developed for Bohai Bay, China, integrating both abiotic and biotic features. This framework, illustrated in Fig. 2, enables a comprehensive characterization of nearshore benthic habitats.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Conceptual framework of marine benthic habitat mapping in the Bohai Bay, China.

Abiotic features

The abiotic classification is structured around three key features: tidal influence, light availability, and substrate type.

First, based on tidal and light regimes, three primary zones were delineated. Littoral zone (intertidal) refers to areas alternated between exposure to air and seawater, supporting amphibious organisms that distinguish them from those found in other zones. Infralittoral zone (subtidal photic zone) refers to areas with sufficient light for photosynthesis (≥ 1% surface irradiance)31, supporting photophilic organisms such as seagrass meadows32. Circalittoral zone (subtidal aphotic nearshore zone) refers to areas lying below the photic limit yet remaining influenced by wave energy, supporting nutrient cycling and other ecological processes.

Second, each of the aforementioned zones is further classified based on substrate types. Substrates are grouped into three broad categories: geological, biogenic, and anthropogenic. These categories are further refined based on factors such as grain size, material composition, and structural origin28.

Biotic features

The biotic classification is based on benthic communities and can be categorized into four main categories. Reef biota refer to communities associated with reef-building fauna, sufficiently abundant to construct identifiable biogenic substrates. Benthic faunal communities refer to faunal communities associated with the seabed but not abundant enough to contribute to substrate formation, as in reef biota. Submerged plants refer to rooted vascular plants and sessile macroalgae that typically remain submerged but may be intermittently exposed during low tides. Semi-terrestrial plants refer to plant communities capable of thriving in the coastal intertidal zone while maintaining adaptations to terrestrial conditions.

Abiotic feature mapping

The spatial delineation of abiotic zones (littoral, infralittoral, and circalittoral) was performed using sequential boundaries: the mean high tide line, the mean low tide line, the photic zone boundary, and the 30 m isobath (representing the nearshore extent)32. Mean high tide line was derived from the coastline data of 2020 Third National Land Survey, which combined topographic surveys and field verification. Mean low tide line was approximated by extracting the 0 m isobath from the global bathymetry grid product (https://oceancloud.nmdis.org.cn/) using the multi-beam approach in the study area. Although the 0 m isobath may not perfectly match the mean low tide line, it can be very close in areas with gently sloping shallow terrains, particularly within tidal flats. The photic zone boundary, defined as the depth where 1% of surface light reaches the seabed, was derived from MODIS L3 product (https://oceancolor.gsfc.nasa.gov/l3/) following the approach by Son & Wang33. The 30 m isobath was extracted from the same global bathymetry grid product to delineate the nearshore extent. Given that the maximum depth in Bohai Bay is less than 30 m, the circalittoral zone coincides with the outer boundary of the study area. In this study, all spatial data layers were processed in ArcGIS 10.8.1. We first generated a polygon vector layer of the study area using the mean high tide line and the outer boundary of the Bohai Bay Ecoregion proposed by Huang22. Then this layer was clipped using the mean low tide line and the photic zone boundary line to obtain the extents of the littoral, infralittoral, and circalittoral zones.

Substrate types were categorized into three main groups: geological, biogenic, and anthropogenic. Geological substrates were mapped using global ocean sediment product (https://oceancloud.nmdis.org.cn). This dataset is based on in-situ marine sediment data collected during national marine special investigations from 1997 to 2020. It was produced through steps including data integration, quality control, gridding interpolation, vectorization processing, and manual review corrections. The data are in polygon vector format, with a resolution of 1:100,000. Biogenic and anthropogenic substrate data are also in polygon vector format, deriving from 2019 side-scan sonar survey and the National Sea Area Dynamic Supervision and Management System respectively. To be noted, anthropogenic substrate layer characterized by human modifications that alter ecological attributes, were excluded from biotic analysis. We corrected the boundaries of geological substrates based on the coastline and overlaid the biogenic and anthropogenic substrate polygons to create an integrated substrate type map.

Finally, the abiotic feature map was generated combining the layers of abiotic zones and substrate types. Each polygon was assigned a comprehensive type based on these attributes.

Biotic feature mapping

The biotic features of Bohai Bay were categorized into reef biota (oyster reefs), benthic faunal communities, and semi-terrestrial plants (tidal salt marshes). The distribution data of oyster reefs were derived in polygon vector format using 2019 side-scan sonar survey data. Tidal salt marsh data were also in polygon vector format using the data from 2023 National Coastal Ecosystem Survey, which integrated remote sensing identification with field verification. Benthic faunal communities were classified into four groups: endangered, key protected, economically important, and dominant communities. Endangered, key protected, and economically important communities were identified following the IUCN Red List of Threatened Species (https://www.iucnredlist.org/), China’s National Key Protected Wild Animals List (https://www.forestry.gov.cn/c/www/gkzfwj/272561.jhtml), and National Key Protected Economic Aquatic Plant and Animal Resources List (http://www.moa.gov.cn/govpublic/YYJ/202401/t20240123_6446366.htm). Dominant communities were analyzed using PRIMER version 6, employing steps including fourth-root transformation, similarity matrix creation, group-average linking cluster analysis, Similarity Profiles (SIMPROF), and Similarity Percentages (SIMPER)16,34,35. Recent benthic data (during the summers of 2021 to 2023, 215 stations in total) collected from the National Ecological Early Warning and Monitoring Program were used to reflect current benthic community status. All benthic samples were collected, processed, preserved, counted in accordance with the Specifications for Oceanographic Survey Part 6: Marine Biological Survey36. Marcrobenthic samples were collected using a 0.05m2 grab sampler, while intertidal organisms were sampled using a 25 × 25 cm intertidal quadrat. Samples were then washed through a 0.5 mm mesh sieve and fixed in a 5% formalin solution for subsequent identification and counting. A total of 236 benthic species were identified. No endangered or key protected communities were identified in Bohai Bay. Economically important communities include Ruditapes philippinarum, Acetes chinensis, and Scapharca communities. Dominant communities include Potamocorbula laevis, Moerella rutile, Gattyana pohaiensis, Nopherus ambigua, Mactra chinensis, Amphioplus japonicus, and Lineidae communities.

The habitats of benthic faunal communities were mapped using the niche theory-based method proposed by Tappin37, which posits that each species depends on a specific set of environmental conditions for its long-term survival38. The BIOENV method in PRIMER version 6 was employed to identify the relevant abiotic variables. We selected five variables for analysis, including bottom water temperature, bottom dissolved oxygen, current velocity, geomorphic type, and substrate type10,16,35. Other available variables of water depth, transparency, salinity, and nutrients were excluded due to the relatively low correlation coefficients. Data of these variables were obtained from the National Marine Ecological Early warning and Monitoring (during the summers from 2006 to 2023, 519 stations in total) to represent the mean environmental state. Environmental thresholds were defined as the mean ± standard deviation of samples associated with each community, capturing typical environmental variability. Habitat suitability was modeled as the percentage of abiotic variables meeting the niche requirements of each community37,38. All abiotic variables were first interpolated to generate 0.5 × 0.5 km gridded data. Based on the environmental thresholds for each benthic community, the reclassification tool was used to assign a value of 1 to grid cells falling within the threshold range, and 0 to those outside it. Then, using raster calculation tools, the number of abiotic variables meeting the niche requirements was calculated, and the results were finally normalized to a scale of 0–100% to derive habitat suitability. Potential habitats were identified where suitability exceeded 80%39,40, and observed samples confirmed the presence of the respective communities.

In this study, we integrated the distribution maps of various benthic communities into a single, unified biotic feature map. Although one community cannot occur exclusively in a specific habitat, it always invariably comprises multiple associated species. Here, to highlight the dominant features of each area and align the mapping process with policymaker needs, we established a spatial overlay priority considering the factors of occupancy duration, ecological protection, and resource utilization. First, based on the occupancy time of a habitat, oyster reefs and tidal salt marshes were assigned the highest priority. They both create a habitat for numerous organism communities, and also contribute to form the biogeochemistry environment to different extent. Second, for communities of key protected species, threatened species, economically important species, and dominant species, we overlaid the habitats in an order that prioritizing habitats of protected and threatened species, followed by those of economically important species, to align with the conservation priority policy in China. Finally, we mapped the habitats of dominant species in the remaining areas. In practice, multiple layers were first transformed into polygon vector format and merged into a single layer using a union tool. Then, each resulting polygon was manually assigned a biotic feature based on the established prioritization order.

Comprehensive habitat mapping

The abiotic and biotic maps were merged using the union tool, retaining the attribute information from both layers. Each resulting polygon was manually reviewed to ensure consistency between the abiotic and biotic features. Only biotic features consistent with the corresponding abiotic features were retained, any mismatched biotic information was excluded. A comprehensive benthic habitat type was then assigned to each polygon, resulting in the final integrated habitat map.

Results

Abiotic feature map

In this study, we utilized three key abiotic features including tidal influence, light availability, and substrate type to develop the abiotic feature map of Bohai Bay. Seven distinct type was identified, and anthropogenic substrate was also illustrated in Fig. 3.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Abiotic feature map of Bohai Bay, China.

Marine benthic habitat map

A total of 20 marine benthic habitat categories were identified within Bohai Bay, China, as shown in Fig. 4 and detailed in Table 1.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Marine benthic habitat map in the Bohai Bay, China.

Table 1 List of marine benthic habitats in the Bohai Bay, China.
Full size table

The littoral zone is primarily composed of two substrate types: mud and sand. Littoral mud habitats (94% of total littoral zone), enriched with nutrient-rich sediments, support a diverse array of benthic biota. Mollusks, such as Potamocorbula laevis, Moerella rutila, and Ruditapes philippinarum, dominate the fauna. Additionally, sporadic tidal salt marsh communities, featuring genera including Spartina sp., Tamarix sp., and Suaeda sp. are present. Littoral sand habitats, primarily located near the Caofeidian area, are also characterized by anthropogenic substrates. However, due to the lack of biological sampling in these sandy areas, specific biotic communities could not be determined.

The infralittoral zone is predominantly covered by mud (85% of total infralittoral zone), with smaller areas of sandy and biogenic substrates. The most extensive habitat in this zone is infralittoral mud, dominated by Lineidae and Nopherus ambigua Communities, which represent 14% of the total area and are mainly distributed in the western part of the bay. In the northern region of the infralittoral zone, three distinct communities are present: infralittoral mud dominated by Gattyana pohaiensis Communities, infralittoral mud dominated by economically important Scapharca communities, and Crassostrea gigas Communities on infralittoral biogenic substrate. The southern infralittoral zone contains extensive sandy substrates, which serve as habitats for economically important Acetes chinensis communities. Mollusk communities, such as those of Moerella and Mactra, are also widely distributed in these sandy areas.

The circalittoral zone is predominantly composed of muddy substrate (96% of total circalittoral zone). Sandy substrates, occupying approximately 4% of the circalittoral zone, are found in the eastern region of Bohai Bay, where Amphioplus japonicus Communities dominate. This distribution aligns with the preference of Amphioplus Communities for sandy habitats.

Discussion

Ecological features and spatial distribution of Bohai Bay

In this study, a two-level benthic habitat classification framework was developed for Bohai Bay, China, to spatially depict its main ecological features. Broadly, the bay can be categorized into three distinct abiotic zones: littoral, infralittoral, and circalittoral zone.

The littoral zone, occupying 21% of the total area, is primarily shaped by sediment deposition from numerous rivers flowing into bay. This zone is most prominent along the southern coastline, where it reaches a maximum width of 25 km. Mud dominates this zone by constituting 94% of the substrate, supporting ideal habitats for benthic biota such as mollusks. In addition, most anthropogenic substrates are located within this zone, especially near major ports such as Caofeidian, Tianjin, and Huanghua Port. Apart from that, large portions of the southern littoral zone have been converted into aquaculture farms, leading to fragmentation of natural mudflats. This transformation can have profound impacts, potentially altering the distribution and functioning of these ecosystems41,42.

The infralittoral zone, dominating 60% of the total area, lies within the photic area with sufficient light penetration to support photosynthetic activity and the growth of primary producers. This zone encompasses all the substrate types, including muddy, sandy, biogenic, and anthropogenic. Muddy substrates remain the dominant type, and sandy substrates are more concentrated in the southern and northeastern parts of the bay, where higher hydrodynamic energy influences sediment deposition and benthic community structure43. Biogenic substrates are located in the northwest, with anthropogenic substrates of artificial reefs around, which can play an important role in supporting the habitat restoration26.

Previous studies have highlighted worms, mollusks, and crustaceans as the dominant benthic taxa in Bohai Bay36,41,44,45,46, which is similar to this study. Specific studies by Zhou et al. reported that the dominant species in the nearshore waters of northwestern Bohai Bay in 2008 were Abrina lunella, Nucula paulula, and Theora fragilis47. Cai et al.42 and Li et al.48 identified Musculus senhousei as the dominant species from 2011 to 2014. Liu et al. found that Moerella iridescens and Sternaspis scutate were dominant in 201945. This study confirms the dominance of mollusks in Bohai Bay, particularly in the littoral and adjacent infralittoral zones. However, the species composition observed differs from previous findings. While tolerant species such as Potamocorbula laevis and Moerella iridescens remain prevalent in the littoral zone, the presence of environmentally sensitive species such as Ruditapes philippinarum in the northwestern region suggests ecological improvement. The successful establishment of larger mollusks alongside tolerant species indicates a positive shift in ecological conditions.

The circalittoral zone, characterized by limited light reaching the seabed, is predominantly located in the eastern offshore areas. However, due to the absence of biological samples, its biotic communities remain undetermined.

Uncertainty analysis

The results of this study, including the abiotic feature map and comprehensive benthic habitat map of Bohai Bay, are subject to uncertainties arising from data accuracy, methodological assumptions, and modeling limitations. Acknowledging these uncertainties is essential for improving future habitat mapping efforts.

Data accuracy and resolution represent a primary source of uncertainty. For example, the delineation of the circalittoral zone relied on MODIS L3 satellite data with a spatial resolution of approximately 4 km. While this resolution is suitable for regional scale mapping, it may fail to capture fine-scale variations in light penetration, especially in nearshore and shallow areas. Consequently, the boundaries of the circalittoral zone may involve some degree of uncertainty. Substrate classification depended on the integration of historical sediment survey data. Although the station density (Fig. 1) provided the highest possible coverage from field surveys across such a large area, spatial and temporal inconsistencies in the data remain a challenge. Similar issues also occurred in the benthic biotic and abiotic datasets.

The use of threshold-based habitat suitability modeling introduces additional uncertainty. Abiotic thresholds, such as bottom temperature, dissolved oxygen, and current velocity, were defined as the mean ± standard deviation of observed data. While this approach captures typical environmental conditions, it may overlook ecological responses to extreme or dynamic environmental variations. Furthermore, the model assumes uniform sensitivity of all identified species to abiotic variables, which may not accurately reflect real world conditions. Figure 5 illustrates the modeled probabilistic distributions of biotic communities in the infralittoral zone, showing that the extents of suitable abiotic conditions are significantly broader than the actual distribution of community samples. To address this discrepancy, we recalibrated the modeling results by accounting for the spatial distribution of observed samples. Only habitat patches with actual sampling stations were retained, and the proportion of sampling stations corresponding to each habitat type (as shown in Fig. 4) was indicated in Fig. 5. This approach is strictly based on field observations to avoid over-extrapolation, while it may underestimate potential suitable habitats.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Modelled probabilistic distribution of dominate benthic fauna in infralittoral mud in the Bohai Bay. (The probability shown in each figure refers to the percentage of abiotic variables matching the niche for specific community).

Implications of marine benthic habitat map

The marine benthic habitat map presented in this study can serve as indicators of the spatial distribution of key abiotic and biotic features. Additionally, it also contributes to improved recognition, monitoring, protection, and management, thereby supporting evidence-based decision-making in marine governance.

This study identified the distributions of oyster reefs (Crassostrea gigas Communities on infralittoral biogenic substrate) and tidal salt marshes (Spartina Communities in Littoral mud, Tamarix Communities in Littoral mud, and Suaeda Communities in Littoral mud). Oyster reefs are characterized by their complex three-dimensional biological structure, which offers crucial sanctuaries for fish, crustaceans, and mollusks, serving as feeding and habitation sites49. Tidal salt marshes, on the other hand, provide essential habitat and foraging sites for various marine animals and seabirds, while also offering ecosystem services. In Bohai Bay, both oyster reefs and tidal salt marshes have historically declined due to land reclamation, harbor construction, and overfishing26. However, recent years have seen gradual recovery of these habitats, driven by the establishment of marine protected areas and ongoing ecological restoration efforts. The benthic habitat map provides details on the distributions of these habitats, enabling a deeper understanding of their changes and aiding in the formulation of conservation strategies.

Benthic habitat mapping is also a crucial tool for ecosystem-based management50,51. It identifies areas that support economically important species, such as Ruditapes philippinarum, Scapharca sp., and Acetes chinensis, which are vital for ecological protection, especially given the gradual decline of biological resources in the bay. Additionally, dominant habitats, such as the littoral mud dominated by Moerella rutila Communities and the infralittoral sand dominated by M. jedoensis and Mactra chinensis Communities represent hotspots for biological resource utilization and are prime candidates for fishery activities. In contrast, habitats like infralittoral mud dominated by Lineidae and Nopherus ambigua Communities, although having limited biological resource value, are more tolerant of anthropogenic disturbance and may be considered as a priority area for marine economic development.

Moreover, the benthic habitat map aids in the planning of monitoring activities. For example, the western region of Bohai Bay exhibits limited data coverage (see Fig. 1), suggesting the need for more focused monitoring and survey efforts in this area. Enhancing the accuracy and reliability of future assessments in this region will be critical for effective ecosystem management and conservation efforts.

Conclusion

This study presents a two-level benthic habitat classification scheme and produced the first comprehensive benthic habitat map of Bohai Bay based on the best available data. A total of 20 distinct benthic habitat types were identified, providing critical insights that can inform ecosystem-based management by identifying priority conservation areas, potential zones for resource utilization, and areas requiring further monitoring and survey. Despite the contributions of this study, some data gaps remain. These gaps highlight the need for enhanced marine monitoring efforts, particularly in underrepresented regions. Meanwhile, other resources like photos, videos, and other forms of ground-truthing efforts are also needed. Additionally, the adoption of machine learning-based habitat models, such as MaxEnt or random forests, would better capture the complex, non-linear relationships between species distributions and environmental variables, contributing to a more reliability habitat maps.