Evaluating Ecosystem Characteristics and Ecological Carrying Capacity for Marine Fauna Stock Enhancement within a Marine Ranching System

1. Introduction

Inshore water areas provide a wide range of essential ecological services to humanity, including the provision of food and raw materials, environmental purification, climate regulation, and cultural enrichment [1,2]. Although they cover only 18% of the Earth's surface, these areas contribute to 25% of the planet's primary productivity and account for 90% of global fishery yields [3]. In particular, inshore waters are of vital importance to China, playing a critical role in ensuring food security and maintaining ecological balance [4,5]. However, in recent years, human activities such as industrialization, urbanization, agricultural expansion, and climate change have severely degraded China's inshore habitats and fishery resources, resulting in reduced fishery yields, diminished biodiversity, and the simplification of food webs [6].

In response to the challenges posed by the degradation of inshore ecosystems, China has proposed the establishment of a marine ecological civilization and implemented a range of measures aimed at protecting and restoring marine ecosystems. These initiatives include the restoration of nearshore environments, the development of marine ranching projects, the enforcement of seasonal fishing moratoriums, and the regulation of fishing vessels—efforts designed to enhance the stability, diversity, and sustainability of marine ecosystems [7,8,9]. Among these, the development of marine ranching projects is viewed as a key strategy for transforming and upgrading China’s marine fisheries while safeguarding the marine environment [10].

Unlike many other countries, China places greater emphasis on the development of artificial reefs, hatchery releases, and the establishment of online monitoring systems. The goal is to create a comprehensive industrial development framework for marine ranching that encompasses site selection, planning and layout, habitat restoration, resource conservation, and safety assurance [11]. Effective assessment of the structural and functional characteristics, as well as the carrying capacity of marine ranching ecosystems, is a fundamental prerequisite for the successful construction and management of marine ranching in China [6]. This is not only crucial for holistic planning but also for ensuring the scientific construction and management of these ecosystems [12,13,14]. Specifically, these assessments provide essential data to guide decisions on the construction area, type, and scale of marine ranching projects. For example, they help in selecting appropriate hatchery release species, estimating optimal stocking densities, and determining the appropriate scale for artificial reef construction [15,16,17,18,19]. Moreover, these assessments are critical for ensuring the long-term sustainability of marine ranching through effective management, which includes determining strategies for sustainable harvest based on carrying capacity [20,21,22].

Due to limitations in carrying capacity assessment technology, most marine ranching projects in China have not conducted comprehensive evaluations of carrying capacity [23]. Existing assessments have largely focused on biomass community structure, water quality, and the enhancement effects of individual species [18,24,25,26,27,28], with little attention to the overall performance of marine ranching systems [8,14,29]. Even system-level evaluations generally compare marine ranching with small adjacent areas, lacking a robust evaluation system and clear grading standards [28,30,31,32,33,34], which undermines the credibility of the results. As a result, most marine ranching projects rely heavily on empirical knowledge, with insufficient scientific foundations for selecting hatchery release species, estimating stocking densities, and designing habitat construction strategies. This leads to significant ecological and economic risks [6,14,35,36,37,38,39]. Therefore, there is an urgent need for the development of systematic evaluation and simulation methods for marine ranching ecosystems to enhance the scientific rigor and sustainability of marine ranching projects in China.

Ecological Network Analysis (ENA) is a systems-based methodology that quantifies the structure and function of marine ecosystems by analyzing the material and energy flow relationships among all components within the ecosystem food web [40,41,42,43,44,45,46,47,48]. This approach has proven to be a valuable tool for assessing ecosystem health and for linking ecological and socio-economic systems. It is widely applied in marine ecosystem management and ecological restoration, providing critical insights for decision-making and policy development [49,50,51].

The Ecopath with Ecosim (EwE) model is widely used to calculate ENA indices for marine ecosystems. It consists of three main components: Ecopath, Ecosim, and Ecospace. Ecopath is primarily employed to analyze the material and energy flows between various ecosystem components, while Ecosim simulates the dynamics of ecosystem food webs under different pressures. The EwE model has been extensively applied in recent studies to assess and simulate marine ranching ecosystems [52,53].

In this study, a marine ranching ecosystem and its control counterpart in the Beibu Gulf of China were selected as research subjects. Ecopath models were constructed for both ecosystems to evaluate their energy flow, trophic structure, and ecological carrying capacity. An index system based on ENA and a fuzzy comprehensive evaluation model was developed to assess the status of these ecosystems. SuiTable Atock enhancement groups for the marine ranching ecosystem were identified, and ecosystem dynamics were simulated under different stock enhancement strategies. This research aims to provide essential technical and theoretical support for the comprehensive evaluation and ecosystem-based stock enhancement of marine ranching. Furthermore, it seeks to contribute to the sustainable and high-quality development of marine ranching in the Beibu Gulf and across China.

2. Materials and Methods

2.1. Introduction of the Marine Ranching

The marine ranching, named “Jinggong”, is located to the southwest of the Guantou Ridge in Beibu Gulf, with coordinates ranging from 21°25'31.26″N to 21°25'58.96″N and from 108°54'15.39″E to 108°56'38.37″E. The specific location of the marine ranching is illustrated in Figure 1. Covering an area of 4.80 km², the marine ranching has an average water depth of approximately 9 meters. It was primarily established through artificial reef construction and hatchery releases. Moreover, an online environmental monitoring system has been implemented to track water quality. Since December 2020, approximately 20,000 cubic meters of artificial reefs have been deployed, categorized into four types, as shown in Figure 2, with materials predominantly composed of reinforced concrete. The control ecosystem is located about 2 kilometers south of the marine ranching.

2.2. Construction of Ecopath Models

2.2.1. Introduction of the Ecopath Model

Please refer to the Appendix A: Introduction of Ecopath model.

2.2.2. Survey of the Marine Environment

Surveys of the biotic and abiotic environment in the Jinggong marine ranching and control ecosystem were conducted in spring and autumn of 2023. Phytoplankton biomass was assessed by measuring chlorophyll a according to standard procedures [54]. Zooplankton samples were obtained through vertical tows using plankton nets with a mesh size of 169 µm. Benthic and swimming organisms in mud substrate areas were surveyed following the Specifications for Marine Surveys GB/T12763.6-2007. Swimming organisms and macrobenthos in reef areas were surveyed using SCUBA diving techniques for hand collection and video recording [32,55], meanwhile, sea urchins and gastropods were collected from reefs using 0.5 × 0.5 m quadrats, while attached organisms such as mussels, barnacles, and oysters were collected by scraping with a small knife. Hard-to-observe organisms, such as Charybdis japonica, were sampled using ground cage nets. Trawl surveys were also conducted around the reef area, with trawl data serving as an important supplement to ground cage net data. Taxa, biomass, and abundance of zooplankton, benthic, and swimming animals were measured for all collected samples.

2.2.3. Functional Group Division

Adhering to the functional group division principle of the Ecopath model [56], the marine ranching ecosystem was categorized into 23 functional groups, including pelagic fishes, large and medium demersal fishes, sparids, leiognathidae, small demersal fishes, corpaenidae, gobiidae, mantis shrimps, large crabs, other crabs, Metapenaeopsis Barbata, other shrimps, cephalopods, sea urchins, gastropods, barnacles, oysters, mussels, other bivalves, other benthos, zooplankton, phytoplankton, and detritus. The phytoplankton group served as the primary producer group, while all other groups functioned as consumers except for detritus. In contrast, the control ecosystem comprised 19 functional groups, exhibiting similar compositions to the marine ranching, but excluding sparids, barnacles, oysters, and mussels. The specific species of each functional group are listed in Appendix E: tables (Table A1 and Table A2).

2.2.4. Data Sources of Functional Groups

The Ecopath models were designed to model a period of 1 year. Biomass for each functional group was expressed as wet weight in t/km². Input data for biomass (B), production to biomass ratios (P/B), consumption to biomass ratios (Q/B), and diet composition of each group were estimated using data obtained from field surveys and literature sources. The methods employed to acquire these data are detailed in the Appendix E: tables (Table A3). The diet composition of each consumer group was displayed in Table A4 and Table A5. The unassimilated ratio of consumption (Ui) for mussels, barnacles, oysters, and other bivalves was set at 0.40, while the group of other benthic organisms was set at 0.35, and all other consumer groups were set at 0.20 [56,57,58]. Fishery data were provided by the Qinzhou Agriculture and Rural Bureau of Guangxi Zhuang Autonomous Region.

2.2.5. Model Balancing and Uncertainty

We used the estimated ecotrophic efficiency (EE) value of each functional group (which was <1) in the first attempt to balance the model. If the estimated EE value exceeded 1, the consumed biomass was greater than that produced. When this occurred, the input data of the diet composition for each consumer group was modified with small changes, with each change not exceeding 0.05. Furthermore, we ensured that most of the P/Q values (the gross food conversion efficiency, the ratio between production and consumption) were in the range of 0.1−0.3. We also ensured that the respiration to assimilation (R/A) and production to respiration (P/R) ratios in the model were <1; the respiration to biomass (R/B) ratio was higher in active species than in sedentary groups [56,59,60]. The pre-balanced diagnosis was also used to identify issues in the model structure and in data quality before balancing the models [61].

2.3. Construction of Indices System

2.3.1. Description of ENA Indices

This study established an index system to evaluate the ecosystem status of marine ecosystems based on ENA indices. The ENA indices were categorized into three groups. Firstly, indices representing the function of the ecosystem include Detritivory/Herbivory (D/H), the average transfer efficiency among different trophic levels (TE), and relative ascendancy (A/C). Secondly, indices representing the characteristics of ecosystem food web structure include connectance index (CI), system omnivory index (SOI), Finn's cycling index (FCI), and average path length (APL) [62]. Lastly, indices representing the maturity of the ecosystem include total primary production/total respiration (TPP/TR), total primary production/total biomass (TPP/TB), and total biomass/total system throughput (TB/TST) (Table 1). The meanings of each index are described in Appendix C: Introduction of ecological network analysis indicators.

2.3.2. Classification of Ecosystem Status Levels

Due to the lack of historical data on ENA indices in the study areas, directly evaluating the current status of the ecosystems was challenging. An alternative approach was to establish standards through inter-ecosystem comparison. Therefore, published literature on ecosystem assessment based on ENA for the world's coastal marine ecosystems (139 ecosystems in total, including estuaries, bays, islands, straits, oyster reefs, artificial reefs, etc.) was collected as much as possible and presented in Appendix E: tables (Table A6). The ecosystems were divided into 5 levels using the quintiles method. For positive type indicators, where a higher value indicates better system performance, the first, second, third, fourth, and fifth quintiles denote the critical values for the “poor”, “relatively poor”, “medium”, “relatively good”, and “good” grades, respectively. For negative type indicators, where a higher value indicates worse system performance, the first, second, third, fourth, and fifth quintiles denote the critical values for the “good”, “relatively good”, “medium”, “relatively poor”, and “poor” grades, respectively (Table 2).

Among the indices, D/H, TE, TB/TST, CI, SOI, FCI, and APL indicate a better ecosystem status as their values increase. Regarding A/C, Ulanowicz et al. (2009) proposed that the optimal trade-off value is 0.4596 [63], values below this threshold are positively correlated with improved system performance. Since none of the cases examined in this study exceeded this threshold, A/C was considered a positive indicator. For TPP/TR and TPP/TB, values greater than 1 suggest lower ecosystem maturity. The weighting of the indicators follows the methodology outlined by Zeng et al. (2021) [64]. As some of the ecosim scenarios in our study did not include fishing activities, we excluded the indicator of mean TL of the catch, as considered by Zeng et al. (2021) [64], and proportionally redistributed its weight among the remaining ecosystem function indices.

2.3.3. Evaluation of the Ecosystem Status of the Ecosystems

Fuzzy comprehensive evaluation (FCE) was used to calculate the composite score of all indicators. The final evaluation result was determined based on the principle of maximum membership degree. The specific evaluation steps are referenced from the methods of Tobor-Kapłon et al. (2007), Dong et al. (2021), and Wu and Hu (2020) [65,66,67].

2.4. Evaluation of Ecological Carrying Capacity

Ecological carrying capacity is defined as the maximum biomass that functional groups within an ecosystem can sustain while maintaining energy balance [68]. The methods for evaluating the ecological carrying capacity were adapted from Jiang and Gibbs (2005) [69]. The biomasses of the target functional group were incrementally increased until the ecosystem became unbalanced. The critical point immediately before ecosystem disbalance was identified as the ecological carrying capacity. No parameters other than the biomass of the target functional group were manually altered during the calculation of ecological carrying capacity.

2.5. Evaluation of Stock Enhancement Potential and Selection of Stock Enhancement Groups

All functional groups in the marine ranching were categorized into three groups based on their trophic level (TL): TL 2.0-2.5, TL 2.5-3.0, and TL 3.0-3.5. The difference between the ecological carrying capacity and current biomass of each functional group was used to evaluate the stock enhancement potential, with a larger difference indicating higher stock enhancement potential. Preferably, groups with higher stock enhancement potential within each TL category were selected as stock enhancement groups. Additionally, the maturity of related enhancement technologies, such as seedling breeding and larval releasing, was considered crucial for selecting stock enhancement groups.

2.6. Simulation of Stock Enhancement Strategies

The Ecosim model was employed to simulate the dynamics of the ecosystem food web over the next 14 years under different stock enhancement strategies. In each simulation scenario, the Ecopath model from the tenth year was extracted to represent the new state of the ecosystem. The ecosystem status was then evaluated using the indices system based on the Ecopath model’s ENA indices.

2.6.1. Introduction of the Ecosim Model

Please refer to Appendix B: Introduction of the Ecosim model.

2.6.2. Construction of Ecosim Model

The vulnerability (v) index is a critical parameter in constructing the Ecosim model. Which determines whether the trophic control between predator and prey is a top-down, bottom-up, or intermediate effect. Due to the lack of historical survey data in the marine ranching, an empirical formula was applied to calculate the v index for each functional group [70]:

$$v_i=0.1515\times {TL}_i+0.0485$$

(1)

where TL_i is the TL corresponding to functional group i. v settings ranging from 0 to 1, with 0.0–0.3 representing a bottom-up control, 0.3 representing the mixed control, and 0.3–1.0 describing a top-down impact [71]. The v_i was then transformed to derive v_new for Ecosim input, which ranged from 1 to ∞:

$$\log \left(v_{new}\right)=0.301958\times v_i+0.001051$$

(2)

Within the marine ranching ecosystem, oysters, barnacles, and mussels act as ecosystem engineers, enhancing the spatial complexity and heterogeneity of the habitat. They provide essential refuges, foraging, and reproductive spaces for organisms in the artificial reef area, which in turn influences the behavior and distribution of predators. These effects were incorporated into the simulations using mediation functions in Ecosim. Sigmoidal functions were used to modulate the predator-prey relationships through these mediation functions in this study. The specific mediation settings were determined based on methodologies outlined by Harvey (2014) and Sadchatheeswaran et al. (2020) [72,73].

2.6.3. Simulation Scenario Design

We assumed that the maximum biomass achievable by the selected stock enhancement group within the marine ranching ecosystem corresponds to the biomass at ecological carrying capacity. Therefore, setting the biomass of the stock enhancement group at ecological carrying capacity became one of the driving factors in establishing the Ecosim model. Additionally, fishing effort was also taken into account as a driving factor for model construction. Fishing data used to drive the model were obtained from the fishing efforts in the marine ranching during 2023.

Based on the selection of stock enhancement groups, three groups were chosen: mussels, large crabs, and scorpaenidae. Since the method for stock enhancement of mussels in marine ranching involved constructing artificial reefs, and both mussels and oysters would grow on the reef once deployed, these three groups, which shared similar ecological characteristics, were combined into one stock enhancement group (MOB).

Three single-group stock enhancement strategies were established for MOB, large crabs, and scorpaenidae, respectively. Additionally, four multiple-group stock enhancement strategies were set: MOB + large crabs, MOB + scorpaenidae, large crabs + scorpaenidae, and MOB + large crabs + scorpaenidae. For each stock enhancement strategy, two scenarios were simulated: with fishing and without fishing. Furthermore, one simulation scenario involving only fishing activity was also included. Therefore, a total of 15 simulation scenarios were conducted in this study (see Table 10).

3. Results

3.3. Trophic Structure

The TLs of functional groups in the marine ranching ecosystem ranged from 1 to 3.46 (Table 3, Table 4). Sparids exhibited the highest TL, followed by large and medium-sized demersal fishes and scorpaenidae (3.42 and 3.27, respectively). Additionally, cephalopods, small-sized demersal fishes, and mantis shrimps also displayed relatively high TLs (3.28, 3.22, and 3.22, respectively). Conversely, oysters and mussels had low TLs, both at 2.02. In the control ecosystem, functional groups exhibited TLs ranging from 1 to 3.63, with sparids being the highest, followed by large and medium-sized demersal fishes and cephalopods (3.35 and 3.25, respectively). Scorpaenidae and small-sized demersal fishes also had relatively high TLs (3.30 and 3.16, respectively).

The TL I comprised phytoplankton and detritus, TL II mainly consisted of shellfish, barnacles, shrimps, zooplankton, and sea urchins, while TL III mainly consisted of fish, cephalopods, mantis shrimps, and crabs in both ecosystems. Mussels had the highest biomass in TL II (80.40 t/km²), followed by barnacles (30 t/km²), while sparids had the highest biomass in TL III (0.33 t/km²), followed by large crabs and other crabs (both at 0.30 t/km²) in the marine ranching ecosystem. The total biomass distributed among TLs I, II, and III in the marine ranching ecosystem was 60.29, 153.17, and 1.92 t/km², respectively, compared to 65.67, 7.00, and 1.26 t/km², respectively, in the control ecosystem. The phytoplankton showed the highest biomass in the control ecosystem (24.64 t/km²).

3.4. Energy Flow Structure

The EE values of functional groups in the two ecosystems are presented in Table 3 and Table 4. In the marine ranching ecosystem, large crabs, mantis shrimps, and cephalopods exhibited high EE values of 0.92, 0.86, and 0.80, respectively. Conversely, sea urchins, barnacles, oysters, and mussels displayed very low EE values due to their high biomass and lack of predators. Shrimps exhibited the highest EE value (0.96) in the control ecosystem, followed by other bivalves, other crabs, M. barbata, and other benthic organisms, with EE values of 0.95, 0.92, 0.91, and 0.89, respectively. Leiognathidae, zooplankton, and detritus had very low EE values (0.01, 0.07, and 0.09, respectively).

The energy flow among TLs in the two ecosystems is depicted in Figure 3. Approximately 2152.00 and 2365.00 t/km²/a of energy flowed to TL II in the marine ranching and control ecosystems, respectively, accounting for 68.00% and 16.00% of their total primary production, respectively. The transfer efficiency between TL II and III, III and IV were both 4.00% in the marine ranching ecosystem, while in the control ecosystem, they were 4.00% and 11.00%, respectively. The average transfer efficiency among TLs II to V was 5.84% and 6.47% in the marine ranching and control ecosystems, respectively.

3.5. Ecosystem Attributes

The metrics of total system throughput (TST), total production, and total biomass—key indicators of ecological size—were 2.75, 1.40, and 5.56 times higher, respectively, in the marine ranching ecosystem compared to the control ecosystem (Table 5). In the marine ranching ecosystem, the proportion of total consumption to TST was the highest at 39.06%, followed by the flow to detritus, which accounted for 30.56% of TST. In contrast, in the control ecosystem, the highest proportion was the flow to detritus at 72.59% of TST, followed by total consumption, which accounted for 19.08%.

3.6. Ecosystem Status

The values of ENA indices of the marine ranching and control ecosystems are presented in Table 6. The evaluated results of the ecosystem status of the marine ranching and control ecosystems are presented in Table 7. The ecosystem status of the marine ranching ecosystem was rated as “relatively good”. Specifically, the FCI was classified as “good”, while the FML, TPP/TR, TPP/TB, and TB/TST were classified as “relatively good”. The CI and D/H were classified as “medium”, and the SOI, A/C, and TE were classified as “relatively poor”. In contrast, the ecosystem status of the control ecosystem was rated as “relatively poor”. Specifically, the CI and SOI were classified as “relatively good”, while D/H was classified as “medium”. Furthermore, FCI, FML, A/C, TPP/TR, and TPP/TB were all classified as “relatively poor”.

3.7. Ecological Carrying Capacity and Stock Enhancement Potential

The ecological carrying capacity of the two ecosystems is detailed in Table 8. Notably, the functional groups within the marine ranching system exhibited generally higher carrying capacities compared to those in the control ecosystem. Mussels displayed the highest carrying capacity within the marine ranching at 163 t/km², closely followed by oysters at 96 t/km². Pelagic fishes emerged as the group with the highest carrying capacity among fish species. Conversely, the control ecosystem's highest carrying capacity was observed with shrimps at 0.95 t/km², followed by large and medium-sized demersal fishes at 0.48 t/km².

The stock enhancement potential of economic functional groups within the marine ranching system was estimated based on the disparity between carrying capacity and current biomass (Table 9). Among groups with TLs of 3.0 - 3.5, gobiidae exhibited the highest stock enhancement potential, followed by scorpaenidae. In the TL range of 2.5 - 3.0, pelagic fishes demonstrated the greatest potential, followed by large crabs. For TLs between 2.0 and 2.5, mussels showcased the highest potential, followed by oysters.

Currently, there are no effective stock enhancement technologies available for gobiidae and pelagic fishes. However, seedling breeding and larval releasing technologies for large crabs (mainly composed of Portunus trituberculatus) and scorpaenidae (mainly composed of Epinephelus moara) have reached a mature stage [74,75], this study opted for large crabs and scorpaenidae as suitable candidates for stock enhancement. Moreover, the enhancement of mussels is primarily hindered by the limited availability of hard substrate within marine ranching. Therefore, constructing artificial reefs can effectively enhance mussel populations. Recognizing that mussels, oysters, and barnacles (MOB) thrive upon the deployment of artificial reefs, these three groups were amalgamated into a single stock enhancement group, with an estimated carrying capacity of 218.06 t/km².

3.8. Ecosystem Dynamics Under Different Stock Enhancement Strategies

The Ecosim model was employed to simulate the dynamics of the marine ranching ecosystem under various stock enhancement scenarios for the forthcoming 14 years. The outcomes are illustrated in Appendix F: Figures (Figure A1, Figure A2, Figure A3, Figure A4, Figure A5, Figure A6, Figure A7, Figure A8, Figure A9, Figure A10, Figure A11, Figure A12, Figure A13, Figure A14 and Figure A15). Table 10 and Table 11 present the ecosystem status of the marine ranching system at the tenth year of modeling. The results revealed that in scenarios involving fishing alone, stock enhancement of MOB + fishing, scorpaenidae + fishing, and MOB + scorpaenidae + fishing, large crabs experienced a rapid decline in biomass to zero. Similarly, in scenarios of large crabs + fishing and large crabs + scorpaenidae + fishing, the biomass of large and medium demersal fishes also decreased to zero. These observations suggest a collapse in the energy flow structure within these scenarios. Consequently, only the ecosystem status of predicted scenarios without such collapse can be effectively evaluated using the index system. The evaluated results indicate that in scenarios involving the stock enhancement of MOB, MOB + large crabs, MOB + large crabs + fishing, MOB + scorpaenidae, MOB + scorpaenidae + fishing, MOB + large crabs + scorpaenidae, and MOB + large crabs + scorpaenidae + fishing, the ecosystem status notably improved to a "good" level. Specifically, there was a significant decrease in the indices of TPP/TR and TPP/TB, accompanied by a significant increase in FCI and FML, while other indices displayed minimal changes.

4. Discussion

In China, the primary objective behind constructing marine ranching was to restore the marine environment and ensure the sustainable yield of fishery resources [76]. Following almost four years of development, the Jinggong marine ranching has experienced significant transformations in both biotic community composition and system functionality compared to the control ecosystem. The biomasses of fish, crustaceans, and mollusks within the marine ranching are 1.82, 16.99, and 33.51 times greater, respectively, than those observed in the control ecosystem. The substantial increase in mollusk biomass can be primarily attributed to the deployment of artificial reefs, which serve as crucial substrates for attached organisms like mussels and oysters [77,78,79]. The robust filter-feeding and biodeposition activities of mussels and oysters contribute significantly to nutrient enrichment in the benthic environment surrounding the artificial reefs [80]. Furthermore, the pronounced spatial heterogeneity offered by artificial shellfish reefs provides effective refuge and foraging grounds for marine organisms. This dynamic supports the growth and reproductive success of various organisms, including echinoderms, crustaceans, and fishes inhabiting the reef area [81,82].

SOI and CI serve as crucial indicators of ecosystem food web complexity. In the marine ranching and control ecosystems, CI values were 0.28 and 0.32, respectively, while SOI values were 0.13 and 0.21, respectively. Comparatively, the CI in the marine ranching system was rated at a “medium” level in relation to coastal marine ecosystems collected in this study, while the SOI was rated as “relatively poor”. Moreover, both CI and SOI values were notably lower compared to small marine ecosystems (1−10 km2) investigated by Heymans et al. (2014) [47]. CI reflects the ratio of actual links to potential links within an ecosystem’s food web, whereas SOI delineates the distribution of feeding interactions across TLs (Christensen et al., 2005; Libralato, 2008) [45,56]. SOI compensates for the limitations of CI in delineating food web complexity, especially given that CI remains constant despite variations in prey proportions within predator diets [56,83]. The low SOI observed in the marine ranching ecosystem suggests weak connections among functional groups, potentially resulting in lower energy flux and a simplified food web structure. The lower CI and SOI in marine ranching compared to the control ecosystem may stem from increased biomasses of mollusk, echinoderm, and barnacles. The simpler diets of these species contribute to a less complex food web structure. The increased biomass of these groups also contributed to increased EE values for phytoplankton and detritus within the marine ranching ecosystem, resulting in a notable enhancement of energy utilization efficiency between TLs I and II. However, despite these improvements, the TE values among TLs II to V remained notably low, averaging 5.84%. This level was classified as “relatively poor” according to the index system, falling far below the natural ecosystem average of 10% [84].

A/C serves as an indicator of ecosystem organization and efficiency. Both ecosystems in this study were classified at the “relatively poor” level, suggesting diminished organization and system efficiency. The observed low A/C value in the marine ranching ecosystem may be attributed to the fact that increasing system organization levels hinge on the succession of community structure, a process that typically unfolds over an extended period [85,86]. Additionally, external disturbances can also disrupt the self-organization process of the ecosystem [87], thereby impeding improvements in its organizational levels.

The low complexity of the food web, TE, and A/C values in marine ranching suggest that this system may be subject to significant external pressures [47,88,89,90], resulting in the majority of TST occurring at low TLs. Overfishing emerges as a prominent source of pressure among various external stressors. Indeed, overfishing has led to a severe decline, and in some cases depletion, of large carnivorous fish in China's coastal ecosystems [6,91], echoing the phenomenon of “fishing down the marine food web” observed in diverse contexts [92,93,94]. Modeled results from Ecosim in this study indicate that large crabs in the marine ranching are subject to overfishing. Chen et al. (2008) similarly observed overfishing in the Beibu Gulf [95]. Fishing activities in areas adjacent to marine ranching may adversely affect marine ranching. The migration patterns of many marine creatures within and around these ranches, particularly for type II and III fish species that only spend part of their lifecycle in the ranches, expose them to significant fishing risks during their movement between regions [14,96,97,98]. Considering the diverse habitat requirements of marine creatures at different life stages, proposals such as those by Yang and Ding (2022) are noteworthy [99]. They suggest constructing global aquatic ecological ranching systems that designate entire estuaries or bays as ranching areas. Similarly, Liang et al. (2020) propose developing marine ranching facilities designed to meet all habitat needs across different life stages of fish [98]. In the long term, these proposals hold the potential to mitigate challenges faced by marine ranching, including strong external disturbances and difficulties in maintaining a complex food web structure and high system organization levels.

The TPP/TR and TPP/TB ratios in the marine ranching and control ecosystems were 1.82 and 8.93, and 17.27 and 85.22, respectively. These values in the marine ranching were rated as “good”, whereas those in the control ecosystem were deemed “relatively poor”. Additionally, the D/H ratio in marine ranching exceeded that in the control ecosystem. These findings suggest that the establishment of marine ranching has significantly enhanced ecosystem maturity and stability [100,101]. Despite comparable total primary production in both ecosystems, the decline in TPP/TR and TPP/TB ratios can be attributed to increases in total biomass and total respiration. Notably, in marine ranching, filter-feeding bivalves accounted for the largest portion of both total biomass and total respiration. These bivalves exhibit ecological characteristics similar to those of zooplankton, but with lower turnover rates. While zooplankton can quickly respond to ecosystem disturbances, bivalves exhibit slower responses, thus creating a pathway for slow energy flow. The asynchronous dynamics of bivalves and zooplankton likely play a pivotal role in sustaining ecosystem stability [102].

The evaluation based on the index system revealed that the ecosystem status in marine ranching was “relatively good”, contrasting with the “relatively poor” status of the control ecosystem. Notably, marine ranching's maturity indices, such as TPP/TR, TPP/TB, and TB/TST, significantly outperformed those of the control ecosystem. Furthermore, indices like FCI and FML demonstrated marked improvements, indicating that marine ranching enhances system maturity, energy recycling efficiency, and food chain length. Moreover, metrics such as TST, total production, and total biomass in marine ranching were 2.75, 1.40, and 5.56 times higher, respectively, than those in the control ecosystems. The carrying capacities for fish, crustaceans, and shellfish have also significantly increased in marine ranching areas, underscoring the positive impacts of marine ranching construction efforts. However, despite these achievements, the construction efforts have yet to effectively improve metrics such as TE, the level of system organization, and food web complexity.

Stock enhancement serves as a pivotal approach to improving marine environments and achieving sustainable utilization of fishery resources [103]. However, the efficacy of stock enhancement in restoring marine resources varies across different initiatives [18,104,105,106,107]. Some studies even contend that this method is entirely ineffective [108]. Simulation scenarios conducted in this study revealed that stocking single fish or crab species did not notably optimize the ecosystem. However, the stock enhancement of the bivalve-dominated MOB group enhanced the ecosystem status from “relatively good” to “good”. This may be attributed to the higher enhancing capacity of MOB compared to crab and fish groups. Additionally, the augmentation in biomasses of mussels, barnacles, and oysters expanded shellfish reefs, providing increased refuge and food resources for crustaceans, cephalopods, fish, and other organisms [79]. This indirect effect also contributed positively to optimizing the ecosystem status.

Yang et al. (2023) demonstrated that stock enhancement of multiple species may have better ecological effects than single-species stock enhancement [109]. This finding aligns with the results of our study. For instance, in scenarios involving single-group stock enhancement of crab, fish, and the MOB group alongside fishing activities, the populations of large crabs or large and medium demersal fishes rapidly collapsed. Conversely, in scenarios combining MOB with crab and fishing, or MOB with crab, fish, and fishing, all functional groups maintained relatively stable biomasses, leading to an increase in ecosystem status to a higher level (from “relatively good” to “good”). These findings suggest that stock enhancement involving multiple species from different TLs represents a more effective strategy for optimizing ecosystem structure and function. However, it is important to note that maintaining the biomass of stock enhancement groups at the group’s carrying capacity in simulation scenarios may not fully reflect real-world conditions. This is because the carrying capacity of each functional group is dynamically influenced by external factors, such as climate change and human activities [20]. To improve ecological outcomes, it is essential to strengthen research on the dynamics of ecological carrying capacity in marine ranching and adjust stock enhancement strategies accordingly [19,110,111]. Moreover, enhancing activities aimed at boosting the biomass of stock enhancement groups to carrying capacity levels may pose ecological risks, potentially pushing the ecosystem toward a critical state of energy imbalance. Integrating the ecological roles of different organisms and implementing multi-species stock enhancement [107], or reducing the stocking density of target organisms, can help mitigate these risks.

In scenarios focused on MOB (excluding MOB + fishing and MOB + fish + fishing), the ecosystem status was elevated to a “good” level, indicating that ecosystem structure and function were effectively optimized compared to the initial state. However, this does not imply that the ecosystem has reached an ideal or perfect state. The index system we constructed in this study provides a general overview of coastal ecosystems, and achieving a 'good' level does not necessarily align with the standards of mature ecosystems as described by Odum (1969) [100]. For example, values of TE, A/C, CI, and SOI showed only slight changes in these scenarios, remaining at “relatively poor” or “medium” levels. This indicates that the simulated stock enhancement strategies had a limited impact on optimizing the food web structure, system organization level, and energy transfer efficiency. This finding aligns with the conclusions of Yang et al. (2023) [112], who reported that stock enhancement did not significantly increase ecosystem species diversity, and its restoration effects were limited. It can be inferred that stock enhancement of only a few species may not be sufficient to fully optimize system structure and function. In line with this, Deevesh et al. (2024) emphasized the importance of prioritizing protection and natural recovery to enhance ecosystem structure and function [113]. To truly optimize the structure and function of marine ranching ecosystems, it is crucial to adopt a more integrated approach that includes strengthening ecological connectivity between marine ranching and adjacent areas, improving fishing management strategies, and advancing the design and deployment of artificial reefs [22,114,115]. When implemented together, these measures could provide a more sustainable and holistic framework for ecosystem restoration and management in marine ranching [22,114,115].

5. Conclusions

In this study, we developed two Ecopath models to assess the energy flow and trophic structure of both marine ranching and control ecosystems. Additionally, we established an index system based on ENA and the fuzzy comprehensive model, incorporating ENA data from 139 coastal marine ecosystems, to evaluate the ecosystem status. The results provide valuable insights for the ongoing development of marine ranching initiatives. Importantly, this index system addresses the challenge of lacking reliable standards for evaluating marine ranching ecosystems, offering a methodological framework that can support the assessment of other marine ranching projects in China. The ecosystem status was rated as “relatively good” in the marine ranching ecosystem and “relatively poor” in the control ecosystem. TST, total production, total biomass, and ecological carrying capacity of economic groups were significantly higher in the marine ranching ecosystem compared to the control, indicating the success of the construction efforts. Based on their biomass enhancement potential, mussels, large crabs, and scorpaenidae were identified as the most suiTable Apecies for stock enhancement. Future research should focus on determining optimal stocking densities by accounting for the mortality rates and growth characteristics of stock enhancement species. The scenario simulations conducted using Ecosim revealed that stock enhancement strategies involving MOB (excluding MOB + fishing and MOB + fish + fishing) elevated the status of marine ranching to a “good” level. Furthermore, enhancing multiple species simultaneously appears to yield better ecological outcomes than focusing solely on single-species stock enhancement. In the simulation scenarios, the biomasses of stock enhancement groups were assumed to reach their carrying capacity levels, which may not fully reflect real-world conditions. Future research should focus on exploring the dynamics of ecological carrying capacity across different temporal scales and adjusting stock enhancement strategies accordingly to achieve the most effective ecological outcomes. The scenario simulations suggest that stock enhancement involving only a few species is insufficient to fully optimize ecosystem structure and function. To address this limitation, future research should adopt a more integrated approach, focusing on strengthening protection and management strategies, as well as advancing the design and deployment of artificial reefs. These measures could play a pivotal role in enhancing ecosystem structure and function, ultimately ensuring the long-term sustainability and resilience of marine ranching ecosystems.