There is a significant disconnection between the richness of data and the effectiveness of decision-making in current marine observation practices. This study argues that the root cause lies in the “ontological” bias and “methodological” mismatch in the logic that guides observation design. We first examine the “uniformity assumption” that views the ocean as a spatially homogeneous and temporally gradual system. Based on the “land-sea heterogeneity isomorphism” framework, we demonstrate that the ocean is essentially a heterogeneous landscape dominated by key dynamic processes, characterized by “resource-rich areas”, “risk-concentrated areas”, and “pulsatile events” that drive their changes. These local, transient, and nonlinear threshold processes often dominate the function and evolution of coastal and estuarine ecosystems. Traditional “mission-oriented” or “undifferentiated census” observations, which ignore this essence, lead to systematic omissions of core processes and key risk signals. Based on this, we propose a fundamental shift in the marine science observation paradigm: from “routine observation-driven alone” to “a combination of routine observation and scientific questions/management goal-driven”. We construct a logical framework of “scientific questions/management goals-key processes-observation requirements”, and systematically expound on three methodological breakthroughs to address the “fluid medium” characteristics of the ocean: building a “dynamic adaptive perception network” to replace fixed grids, establishing “full-chain process analysis” to replace isolated hypothesis testing, and achieving an intelligent closed loop of “real-time assimilation and predictive decision-making”. This study aims to provide a comprehensive theoretical pathway from ontological cognition to methodological practice to build the next generation of “precise, efficient, and applicable” marine science observation systems.

The global ocean is currently facing a severe challenge of “cryptic ecosystem degradation”, characterized by subtle but progressive declines in ecosystem integrity. Conventional water quality monitoring often indicates regulatory “compliance”, while underlying processes such as food web disruption, loss of keystone species, and functional degradation of ecosystems continue largely undetected. This management paradox arises from traditional assessment paradigms that emphasize physicochemical conditions while overlooking biological structure and ecosystem functionality. This study proposes a target-driven framework for marine ecosystem health assessment, aiming to facilitate a paradigm shift in environmental management from “comprehensive census” approaches to “precision-based diagnostic” strategies. Departing from fixed indicator frameworks, the proposed system introduces a “three-level funnel” generation logic: (i) identification of core ecological issues based on specific management objectives, (ii) selection of key biological components and associated indicators, and (iii) design of an optimal spatiotemporal observation strategy. Guided by the principles of usability, effectiveness and adequacy, the framework is supported by an integrated “space-air-sea-intelligence” technological network and enables the transformation of observational data into actionable decision-making knowledge through an intelligent diagnostic engine. Case studies demonstrate that the framework can dynamically generate customized assessment schemes tailored to diverse management objectives, including “overall health assessment, ” “aquaculture carrying capacity evaluation”, and “disaster early warning”. Among these, the classical triad-based Index of Biotic Integrity (Z/F/B-IBI), encompassing “zooplankton-fish-benthos” represents the optimal scheme generated by the framework for achieving “regional-scale ecosystem health assessment”. Overall, the proposed system aims to maximize management efficiency while minimizing observational costs, thereby providing a dynamically adaptable, “precision medicine-like” solution for ecosystem-based marine management.

Health assessments of marine ecosystems rely heavily on the real-time diagnosis of biological status and functions. However, the “cognitive lag” of traditional biological observation (i.e., the long period from sample collection to species identification and data output, which makes it difficult to reflect rapid changes in the ecosystem promptly) has become the core bottleneck for the operational application of health assessment. This article systematically analyzes the historical roots and contemporary predicaments of this bottleneck, pointing out that it stems from “technological path dependence”, “verification logic paradox”, and “observation network absence”. To break through this predicament, this article proposes a new biological observation paradigm that will enable the shift from “lagging evidence collection to real-time diagnosis”. This paradigm includes three core dimensions: at the cognitive level, it will survey the “complete species set”, track the “functional fingerprints”, and identify key ecological function signals; at the technical level, it will construct a “real-time perception-quality control verification” collaborative technical system, integrate cutting-edge measures such as environmental DNA, in situ imaging, and intelligent acoustics, and innovatively iterate the verification mode; at the network level, through the development of modular intelligent biological sensing modules, it will embed biological sensing capabilities in the global observation network in an “plug-and-play” manner, achieving automated and continuous perception of biological signals. This paradigm shift aims to upgrade biological observations to a forward-looking information infrastructure that supports real-time health diagnoses and early warnings regarding marine ecosystems.

Under the stresses of climate change and anthropogenic activities, marine ecosystems are undergoing unprecedented changes. Accurately assessing the ecological quality status of marine ecosystems and their spatiotemporal variation characteristics is an urgent scientific challenge. Within marine ecological quality assessment frameworks, macrobenthos serve as critical evaluation criteria due to their distinctive physiological and ecological traits. However, current assessment indicators based on macrobenthic communities are numerous yet singular, lacking a comprehensive integrated evaluation system. To address this gap, the present study proposes a multi-dimensional, multi-indicator evaluation framework constructed from an ecological perspective that integrates both community structure and function. This framework comprises three primary categories and nine secondary indices: (1) biotic indices (M-AMBI and BENTIX index); (2) community structure (Shannon-Wiener diversity index, species richness, taxonomic composition, abundance, and biomass); and (3) functional attributes and stability (functional diversity and stability). Weight assignments were allocated according to indicator importance to calculate the macrobenthic community ecological quality index (MCEQI) for benthic ecological status assessment. Application of MCEQI to analyze the macrobenthic community ecological status in Jiaozhou Bay demonstrated that this index exhibits higher accuracy than single-indicator approaches, providing a more comprehensive reflection of community ecological quality.

The health status assessment of marine benthic ecosystems, as an important basis for maintaining the ecological balance of the oceans, relies on the long-term dynamic monitoring of benthic communities. In this study, we systematically reviewed the history and current applications of ecological assessment indices developed based on benthic organisms, including the perspectives of future fields in marine ecosystem health assessment. Traditional biological indices—including biodiversity indices (Shannon-Wiener index, Pielou index, etc.), functional group analyses (feeding evenness index), and indices (AMBI, M-AMBI, BENTIX index, etc.) based on the proportion of pollution-tolerant/sensitive species have advanced ecosystem health assessment by quantifying the responses of community structure to environmental stressors and facilitating the shift from qualitative to quantitative assessments. Nevertheless, traditional methods are limited by cumbersome procedures for morphological characterization, the limitations of single indicators, and regional differences in applicability. In recent years, environmental DNA (eDNA) technology has made up for the shortcomings of traditional methods by rapidly obtaining biodiversity information through high-throughput sequencing, and has derived novel indices such as gAMBI, which validates its complementarity with morphological methods. Integration of Artificial Intelligence (AI) techniques such as machine learning algorithms (Random Forests, Convolutional Neural Networks, and so on) with statistical analysis has improved the ability of ecosystem assessment models to resolve nonlinear relationships and multiple stressors. Meanwhile, automatic and intelligent image recognition technology offers the possibility of accurate and rapid identification of macrofauna and their monitoring. Future research shall integrate multidimensional data and interdisciplinary techniques to construct a more universal and dynamically responsive assessment system to cope with the potential impacts of global climate change and human activities on marine ecosystems.

Marine microorganisms, characterized by their immense biomass, rapid environmental response, and crucial ecological functions, serve as sensitive indicators for assessing the health of marine ecosystems. This article provides a systematic review of recent advances in marine ecological monitoring and health assessment based on microbial communities. First, we discussed the key characteristics of marine microorganisms as bioindicators, such as their high sensitivity to environmental stress and their functional redundancy in maintaining ecosystem processes. Case studies are presented to highlight the successful use of microbial monitoring to address climate change, pollution events, and ecological disasters. Next, we reviewed the evolution of marine microbial monitoring technologies, spanning from traditional cultivation methods to modern techniques such as high-throughput sequencing and Raman spectroscopy. We also compared the advantages and limitations of these approaches in practical applications. Finally, in response to challenges such as insufficient data standardization and the lack of quantitative assessment metrics, we proposed a systematic framework for future development. This framework emphasizes the need for end-to-end standardization from sampling to data analysis, the creation of intelligent diagnostic models that integrate multi-dimensional “Raman spectroscopy-genetic-environmental” information with the establishment of a national-scale specialized monitoring network. This review aims to provide theoretical supports and technical pathways for the development of a next-generation, high-resolution, real-time microbial-based marine ecological health assessment system.

Dissolved oxygen (DO) is fundamental to maintaining the balance and stability of marine ecosystems. Under the combined influence of climate warming and intensified human activities, hypoxia in coastal waters worldwide is increasing in severity and spatial extent, making deoxygenation a major threat to coastal ecosystem health. As a typical semi-enclosed sea, the Bohai Sea is subject to multiple environmental pressures, including ocean warming, eutrophication, and rapid socio-economic development. In recent years, varying degrees of hypoxia have been reported across multiple subregions of the Bohai Sea, with hypoxic centers primarily located northeast of the Yellow River estuary and southeast of Qinhuangdao (occurring ≥3 times). This study reviews domestic and international research on hypoxia in the Bohai Sea, characterizes the spatial distribution patterns of summer hypoxia over the past two decades, and summarizes the dominant controlling processes. It further outlines the long-term development trends and their influencing factors in the Bohai Sea and proposes key directions for future research on hypoxia in the region. The study aims to provide a scientific basis and decision-making support for ecological health assessment and marine resource management in the region.

Changes in the seasonal cycle of phytoplankton are important indicators of coastal ecosystem responses to climate change and human activities. Based on historical observations from Jiaozhou Bay in 2005, 2010, 2015, and 2020, this study examined long-term variations in the seasonal cycle of phytoplankton biomass and their environmental drivers. The results showed that the seasonal distribution pattern of phytoplankton biomass in Jiaozhou Bay changed markedly during 2005~2020, shifting from a bimodal pattern with winter and summer peaks to a predominantly unimodal pattern with a summer-autumn peak. Winter peak values of net-collected phytoplankton abundance and chlorophyll a concentration decreased substantially, from 103.62×10⁶ cells/m³ and 15.29 μg/L in 2005 to 12.81×10⁶ cells/m³ and 1.13 μg/L in 2020, respectively. Community structure analysis indicated that diatoms consistently dominated the phytoplankton assemblage and that variation in diatom abundance was the primary driver of the observed shift in the seasonal cycle of phytoplankton biomass. In contrast, dinoflagellate abundance declined markedly after peaking in 2010. Correlation analyses suggested that increasing winter water temperature and decreasing phosphate concentration were the main factors underlying the attenuation and eventual disappearance of the winter phytoplankton biomass peak. These findings reveal the long-term evolution of phytoplankton seasonal structure in Jiaozhou Bay under the combined influence of climate change and nutrient adjustment, and offer scientific support for coastal ecosystem management.

As primary producers, phytoplankton constitute a vital component of marine ecosystems, and their community structure largely determines the stability of these ecosystems. Previous studies on phytoplankton community structure in Laizhou Bay have relied primarily on short-term observations or data from one or two sampling cruises. There remains a scarcity of multi-year continuous survey data, and the temporal evolution of phytoplankton community structure has rarely been systematically analyzed. Based on survey data from 2016 to 2023 summer in Laizhou Bay, we investigated the interannual variations of phytoplankton and conducted correlation analyses between phytoplankton communities and environmental factors. Results show that a total of 106 species from 3 phyla were identified over the 8-year period, predominantly diatoms, followed by dinoflagellates. Significant interannual variations in species composition were observed, and the lowest diversity (48 species) occurred in 2019 and the highest (68 species) in 2020. Phytoplankton cell abundance ranged from 39.1×10⁴ cells/m³ to 2 687.3×10⁴ cells/m³, peaked in 2017 and reached the minimum levels in 2021. In the planar distribution, the phytoplankton cell abundance, the nearshore waters of estuaries such as the Yellow River estuary and Xiaoqing River estuary are higher than that of the central part of the bay. Dominant species included Pseudo-nitzschia pungens, Chaetoceros curvisetus, Skeletonema costatum, and Cerataulina pelagica. Correlation analysis revealed that nitrate-nitrogen was the primary environmental factor on phytoplankton abundance. The overall upward trend in the interannual variations of diversity index (H'), richness index (d), and evenness index (J') indicate that phytoplankton community structure tended to improve. This study contributed to clarifying the evolutionary trend and influencing factors of phytoplankton in Laizhou Bay, and provided a scientific foundation for decision-making in marine ecological environment protection of the bay area.