A long-term spatial analysis investigating ecosystem changes across Tianjin, China, indicates that diverging trajectories in habitat quality, carbon storage, water yield, and soil conservation can guide targeted urban and coastal land rehabilitation. Writing in Frontiers in Ecology and Evolution, researchers established an analytical framework that integrates historical evaluations from 2000 through 2020 with multi-scenario projections looking ahead to 2030. Rapid socioeconomic development and extensive urban expansion across China’s coastal zones have generated pronounced conflicts between environmental protection and municipal growth. Because semi-enclosed water bodies like the Bohai Rim experience weak hydrodynamic exchange and limited self-purification capacity, managing land transformation along adjacent metropolitan sectors has become critical. The research team evaluated historical transitions and future land-use simulations across Tianjin’s 11,917 square-kilometer administrative boundary to identify specific ecological restoration priority areas, providing regional planners with an empirical foundation to balance conservation with sustained territorial development.

To evaluate baseline conditions across two decades of urbanization, the researchers examined four foundational ecosystem services using the Integrated Valuation of Ecosystem Services and Trade-offs modeling suite on a uniform one-kilometer grid. Between 2000 and 2020, the aggregate volume of soil conservation increased from 54,500 metric tons to 91,700 metric tons, registering an overall rise of 68.25 percent. In contrast, habitat quality suffered a sustained deterioration over the identical twenty-year interval, dropping by 12.34 percent from an index value of 1,701 down to 1,491. Carbon storage exhibited early gains followed by localized losses, moving from 141.1 million metric tons in 2000 up to 141.8 million metric tons in 2010 before declining to 140.5 million metric tons in 2020, representing a net total decline of 0.43 percent. Model outputs also captured a substantial surge in modeled water yield from 319 million cubic meters to 976 million cubic meters, marking a 206 percent increase that was largely concentrated around impervious municipal surfaces.

The study emphasized that interpreting high water yield figures requires careful distinction between productive hydrologic supplies and urban surface runoff dynamics. Because the water-balance mechanism in the InVEST model calculates water yield as the residual difference between precipitation and actual evapotranspiration, densely built environments with expansive impervious pavement and minimal vegetative cover naturally show reduced interception and elevated runoff. Rather than indicating an abundant ecological freshwater supply, elevated water yield in urban centers frequently signals a disruption in regional hydrological balance alongside escalated local flooding hazards. Geographically, carbon storage, soil conservation, and habitat quality displayed common patterns characterized by lower values across southern sectors and peak concentrations within northern mountainous terrain. Water yield demonstrated an opposing geographic distribution, reaching its highest values across the southeastern coastal plain, the six core urban districts, Beidagang Reservoir, and Binhai New Area, thereby establishing distinct spatial management challenges across divergent municipal zones.

Examining functional interactions among the four ecosystem services revealed pronounced and durable functional trade-offs alongside consistent synergies. Pearson and Spearman correlation analyses incorporating Dutilleul’s modified t-test to adjust for spatial autocorrelation confirmed that the relationship between water yield and habitat quality represented the most persistent regional trade-off. This inverse association intensified over the study period, shifting from a Pearson correlation of negative 0.30 in 2000 to negative 0.37 in 2020, matched by consistently negative Spearman rank values ranging between negative 0.707 and negative 0.670. Conversely, carbon storage and soil conservation sustained a stable synergistic relationship, maintaining positive Pearson correlation coefficients near 0.28 and Spearman values between 0.252 and 0.198. Relationships pairing water yield with carbon storage and soil conservation exhibited conflicting directional signs between parametric and non-parametric testing, indicating sensitivity to analytical methodology and demonstrating that regional trends may mask highly varied local interactions occurring at neighborhood scales.

To resolve local spatial complexities obscured by municipality-wide averages, the authors applied geographically weighted regression models alongside Getis-Ord spatial statistics to track the localized clustering of ecosystem functions. Hotspot calculations showed that statistically significant water yield clusters expanded outward from the urban center in direct alignment with expanding construction land, while also increasing around the Beidagang Wetland Nature Reserve and the Binhai National Marine Park. Geographically weighted regression demonstrated that while water yield and soil conservation exhibited method-sensitive regional associations, localized synergy zones expanded over time across the western ecological barrier and eastern coastal margins. Concurrently, intense functional trade-offs between water yield and soil conservation clustered near Taiping Town in southern Binhai New Area, where infrastructural works such as the Tianjin Binhai Douzhuang General Aviation Airport accelerated land-use conversions. Strong trade-offs between water yield and carbon storage also magnified across built-up areas within Ji Prefecture, Wuqing, Xiqing, Jinnan, and Jinghai districts.

Future landscape dynamics were simulated for the year 2030 using the Patch-generating Land Use Simulation platform, coupling a land expansion analysis strategy driven by random forest algorithms with cellular automata incorporating random seeds. Calibration of the simulation framework using historical data between 2000 and 2020 demonstrated strong predictive fidelity, producing a Kappa coefficient of 0.835, an overall accuracy of 88.75 percent, and a figure of merit of 0.180. The authors projected land cover under three distinct Shared Socioeconomic Pathways integrated with Representative Concentration Pathways: the sustainable SSP126 scenario, the historical-continuation SSP245 baseline, and the fossil-fuel-dependent SSP585 development trajectory. Across all modeled scenarios, agricultural and unused lands consistently shrank as construction land expanded along urban perimeters and coastal boundaries. Under the resource-intensive SSP585 pathway, cultivated farmland experienced the steepest reduction, contracting by 542 square kilometers, while construction surfaces grew by 735 square kilometers, significantly exceeding the development footprints observed under more conservative developmental trajectories.

Projected shifts in land use generated corresponding variations in future ecosystem service delivery across the municipal landscape under each alternative future scenario. Under the baseline SSP245 scenario, total water yield was projected to decline to 719 million cubic meters, reflecting a 26.3 percent decrease from 2020 baseline levels, while soil conservation fell by 33.2 percent down to 61,300 metric tons. The sustainable SSP126 scenario yielded moderate functional improvements, expanding total water yield by 80.2 percent to 1.759 billion cubic meters and soil conservation by 5.8 percent to 97,000 metric tons, while holding habitat quality losses to 3.2 percent. The highly intensive SSP585 pathway produced an anomalous surge in simulated water yield to 6.081 billion cubic meters, representing a 523.1 percent rise, accompanied by a 173.1 percent increase in calculated soil conservation to 250,400 metric tons. The researchers linked these substantial SSP585 increases to altered regional precipitation inputs and heightened runoff from extensive impervious pavement expansion rather than genuine natural conservation enhancements.

To synthesize these historical trajectories, functional conflicts, and simulated risks into an operational planning tool, the researchers formulated an integrated Ecological Restoration Priority Index. This composite index combines a historical ecosystem service loss index, an ecosystem service trade-off intensity index, and a multi-scenario future ecological risk index into a normalized spatial metric. Applying local indicators of spatial association to the resulting index values revealed 1,919 High-High spatial clusters concentrated across contiguous sectors of the central and southern municipality, marking locations characterized by simultaneous historical impairment, heightened functional conflicts, and high projected degradation risks. The analysis also categorized 3,874 Low-Low spatial units primarily across northern uplands and peripheral agricultural areas where restorative urgency remains low, alongside 1,216 Low-High transition units situated on the immediate fringes of high-priority clusters. Only 11 High-Low isolated units were detected across the entire municipality, verifying that pressing ecological restoration requirements are strongly clustered geographically rather than scattered randomly.

Drawing upon the spatial configuration of the restoration index alongside municipal territorial planning boundaries, the authors categorized Tianjin into three progressive ecological management zones. Core Protection Zones, defined explicitly as the primary restoration priority areas, encompass 1,597 square kilometers concentrated within heavily urbanized central and southeastern districts where built construction land covers roughly 85 percent of the surface. Occupying an intermediate spatial position, Synergistic Optimization Zones cover 3,336 square kilometers across transitional landscapes where agricultural, aquatic, and construction spaces interlock. These transitional territories display prominent trade-offs between water yield and carbon storage around urbanized fringes, but retain significant potential for dual functional gains through coordinated land reconfiguration and vegetative buffer installation.

The study proposed differentiated management prescriptions tailored to the specific functional characteristics and land-use compositions of each delineated ecological zone. For Core Protection Zones, the authors advocated strict boundary controls to halt municipal encroachment into remaining green spaces, coupled with the ecological retrofitting of existing urban centers through brownfield redevelopment, permeable surfaces, and expanded urban blue-green infrastructure. In Synergistic Optimization Zones, recommendations center on converting fragmented rural construction parcels into grasslands, re-establishing riparian buffer strips, and adopting low-impact agricultural techniques to strengthen connectivity between water networks and farmlands. Within Ecological Improvement Zones, recommended actions include curbing nitrogen fertilizer application, rehabilitating degraded woodlands at critical pinch points, and expanding wetland conservation. While noting limitations surrounding model parameterization, unmodeled cultural services, and the absence of localized field validation, the authors concluded that their dual-perspective framework offers a robust, transferable methodology for balancing conservation with urban expansion across coastal metropolitan regions.

Subject of Research: Spatial modeling and multi-scenario identification of urban ecological restoration priority areas in Tianjin

Article Title: A comprehensive “past-future” framework for restoration priority areas identification in a coastal city: a case study of Tianjin

Article References: Hao, J., Li, X., Feng, Y., & Qian, N. (2026). A comprehensive “past-future” framework for restoration priority areas identification in a coastal city: a case study of Tianjin. Frontiers in Ecology and Evolution, 14, Article 1961285. https://doi.org/10.3389/fevo.2026.1961285

Image Credits: AI Generated

DOI: 10.3389/fevo.2026.1961285

Keywords: Tianjin, Ecosystem Services, Restoration Priority Areas, InVEST Model, PLUS Model, Habitat Quality, Soil Conservation, Water Yield, Carbon Storage, Ecological Restoration Priority Index, Core Protection Zones, Spatial Autocorrelation, Geographically Weighted Regression, Bohai Bay, Shared Socioeconomic Pathways, Land Use Simulation