Next Article in Journal
Multi-Perspective: Research Progress of Probiotics on Waste Gas Treatment and Conversion
Previous Article in Journal
Effects of Spartina Alterniflora Invasion on Soil Organic Carbon Dynamics and Potential Sequestration Mechanisms in Coastal Wetlands, Eastern China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 17
Issue 19
10.3390/su17198641
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Exploring Nighttime Fishing and Its Impact Factors in the Northwestern South China Sea for Sustainable Fisheries

by
Zhiyong Long
1,
Gao Zuo
2,
Tao Zhang
2,* and
Jinjun Zheng
2
1
College of Meteorology and Oceanography, National University of Defense Technology, Changsha 410037, China
2
School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, China
*
Author to whom correspondence should be addressed.
Sustainability 2025, 17(19), 8641; https://doi.org/10.3390/su17198641
Submission received: 20 June 2025 / Revised: 10 September 2025 / Accepted: 25 September 2025 / Published: 25 September 2025
(This article belongs to the Section Sustainable Oceans)
Browse Figures
Versions Notes

Abstract

The South China Sea (SCS) is an important region of fishery resources. However, its fishery resources have been threatened, mainly because of overfishing. In this study, we explored the distribution of night-time fishing boats and analyzed the relationship between fishing activities and marine environmental factors in the northwestern SCS (NWSCS). Firstly, the spatiotemporal variations in nighttime fishing boats in each month of 2021 in the NWSCS were studied. Meanwhile, a fishery activity center index was used to analyze the overall fishery activity trend in the NWSCS. Finally, the spatiotemporal distribution patterns of corresponding environmental factors (i.e., Chl-a, SSS, SST, latitude, longitude) were analyzed, and the nonlinear relationship between environmental factors and fishery activities was quantitatively studied using the generalized additive model. The results showed that fishery activities were mainly distributed in the waters of Beibu Gulf and the southwest of Hainan Island. Meanwhile, there were obvious seasonal differences (i.e., trimodal distribution) in the intensity of fishery activities in the NWSCS. Chl-a was the most important impact factor with a contribution of 21.7%, followed by SSS, longitude, SST, and latitude, with contributions of 12.8%, 9.4%, 4.2%, and 0.5%, respectively. Fishery activities in the NWSCS were mainly distributed in the area with Chl-a of 0~0.35 mg/m3, SST of 21.2~26.4 °C, and SSS of 32.9~33.8 Practical Salinity Unit. This study reveals that more efforts are required to prevent IUU fishing activities for the sustainable development of marine ecosystems in the NWSCS. It is also necessary to improve remote sensing technology to support making sustainable fishing plans.

1. Introduction

The South China Sea (SCS) is one of the largest marginal seas in the Western Pacific Ocean [1]. It boasts an excellent natural environment, a rich variety of fishery ecological environment types, and considerable fishery resources. Particularly, the northern SCS carries the most frequent continental shelf fishing activities in the world [2,3,4]. Specifically, eighty percent of the catch in the SCS comes from the nearshore waters along the northern coast of the SCS [5]. In recent years, with the frequent increase in fishery activities, overfishing has brought serious negative impacts on fishery resources in the northern SCS [6,7,8,9]. For example, the average nutritional levels of its fishery resources and coastal and continental shelf fisheries have declined [10]. Meanwhile, the actual catch far exceeds the extent that natural ecological restoration can bear [11,12]. The number of fishing boats detected by nighttime lights in the open SCS increased from ~400 to ~2000 from 2012 to 2020 [13]. However, only ~46% of these satellite detections were reported by Chinese vessel monitoring systems (VMS) data. Therefore, there is an urgent need to manage and protect fishery resources to maintain the sustainability of fishery resources, the health of marine ecosystems, and human food security [14,15,16]. As the marine environment provides the survival and activity of marine fish (e.g., spawning, feeding, aggregation), it is also necessary to explore the relationship between fishing activities and marine environmental factors in the northwestern SCS (NWSCS).
Due to the over-exploitation of fishery resources, dominant fishes in the SCS have changed from traditional bottom economic fishes to small and medium-sized pelagic fishes, which have lower food chain levels, shorter life cycles, and strong fertility [13,17,18]. Therefore, light fishery, which mainly catches pelagic fish resources, plays an increasingly important role in fishery production. Light fishery refers to the fishery activities that use artificial light sources to attract fish to gather and catch according to the phototaxis of fish. Multiple studies have found that most of the night-time vessels identified by satellite low-light remote sensing images are light-emitting vessels that mainly target middle- and upper-layer fish resources for fishing [18,19,20,21,22]. That is, the distribution of these vessels can reflect the distribution of fishery resources. Accordingly, temporal and spatial distribution characteristics of fishery activities can be obtained based on nighttime low-light remote sensing data [2,13,23,24].
Nighttime low-light remote sensing technology has been widely used for exploring fishery resources, due to its advantages in wide coverage and periodic observations [2,17,19,21,23,25,26,27], especially in the SCS [13,18,22,24,28,29]. For example, Wang extracted the lighting of fishing vessels in the SCS using the VIIRS nighttime light data from 2014 to 2015 and produced the distribution maps of the production intensity of fishing boats in the SCS in different years and months [28]. The results showed that the fishery production intensity in the SCS decreased from the shallow sea to the open sea waters, and the main fishing vessel operation areas were distributed in the Beibu Gulf (BG) and the mouth of the Jiulong River. The seasonal differences were also significant, i.e., high fishery production intensity in spring and the lowest value in winter. Zhang extracted the light information of nighttime fishing vessels in the SCS based on the VIIRS DNB nighttime light data, analyzed the seasonal characteristics of nighttime fishery production in the SCS, and conducted kernel density analysis based on the identification results to extract the key fishing grounds for nighttime lights in the SCS [29]. Meanwhile, the SCS was divided into several main fishing areas, and the fishing area dataset of surrounding countries or regions was also produced for further analysis. Li et al. investigated the distribution pattern of light fishery in the SCS using the seasonal average nighttime light remote sensing data since 2016 and discovered previously unrecorded light fishery clusters [18]. However, most of the existing studies were implemented using threshold-based methods to extract nighttime vessels, which will reduce precision when the volume of remote sensing data has increased, leading to larger uncertainties for further analysis. Accordingly, deep learning-based methods were developed and showed improved accuracies [30,31]. However, these methods cannot easily detect small vessels. To overcome this issue, Zuo et al. proposed the enhanced dense nested attention network (EDNA-Net) to improve the detection of small vessel targets under low-light conditions, and then validated its effectiveness in the SCS [32].
Marine environmental factors, including biotic and abiotic factors, have complex impacts on the distribution of fishery resources [33,34]. Among abiotic factors, water temperature, chlorophyll concentration, and water depth play important roles in the migration, aggregation, and movement of fish [35,36,37,38]. Ji et al. analyzed the relationship between monthly average sea surface temperature (SST) and the monthly average catch per unit effort (CPUE) of yellowfin tuna fishing grounds in the SCS and its adjacent waters to get the most suitable SST range for predicting the spatial distribution of yellowfin tuna [35]. The relationship between key marine environmental elements (i.e., SST, chlorophyll-a concentration (Chl-a), suspended sediment concentration, wind waves and surges, seawater flow velocity) and typical fishery resources in the NWSCS based on remote sensing data of the marine environment and catch yield data of fishery resources from 2010 to 2019 has been explored by He [36]. Huang obtained and investigated the spatiotemporal distribution of light fishing vessels in the northern SCS based on night light remote sensing images, and then studied the marine time environmental drivers (i.e., SST, Chl-a, Sea Level Anomalies) [37]. However, these studies have not yet fully exploited the advantages of nighttime low-light remote sensing data and marine environmental data for mining spatiotemporal variations in nighttime fishery activities and corresponding impact factors in the NWSCS. For example, the most recent study in the NWSCS by He [36] used statistical fishery data from Chinese government to implement the analysis. This may seriously underestimate the fishing activities in the NWSCS due to the lack of statistical data from neighboring countries. Accordingly, there is still a knowledge gap in the detailed spatiotemporal distribution of nighttime fishery activities and corresponding impact factors in the NWSCS.
To fill the above-mentioned knowledge gap, we implemented the EDNA-Net algorithm proposed by Zuo et al. to identify nighttime light ships in the NWSCS for each day of 2021 [32]. Then, we analyzed the spatiotemporal patterns of nighttime light ships and corresponding impact factors in this region (divided into four subregions) based on accumulated data at a monthly scale. Accordingly, we would like to solve the following scientific problems: (1) How many nighttime fishing vessels were in the NWSCS in a complete calendar year? (2) How did nighttime fishing activities change for different months of the year? (3) How much did different marine environmental factors impact nighttime fishing activities?

2. Study Area and Data

2.1. Study Area

The study area is located at the NWSCS, including the BG fishing ground (106°30′–109°50′ E, 20°20′–21°30′ N) and the Pearl River Estuary (PRE) fishing ground (112°50′–114°20′ E, 21°08′–22°00′ N), the southeast of Hainan Island (SEHI), and the southwest of Hainan Island (SWHI) (Figure 1) [10]. The continental shelf in the NWSCS has a depth of less than 500 m; specifically, the average depth in the BG is 60 m. The continental shelf in the west and east of Guangdong is broad, and the water depth is less than 200 m. The continental shelf in the southeast of Hainan becomes narrower and steeper and slopes toward the middle of the SCS with a depth of more than 1000 m [36]. The bathymetric line runs from northeast to southwest and is roughly parallel to the coastline. This region is located in tropical and subtropical regions, where southwest wind prevails in summer and northeast wind prevails in winter, and hydrological conditions and circulation characteristics vary for different seasons [39]. The NWSCS was selected because it is a typical region with rich fishery resources, but its sustainable development of fishery is threatened by overfishing [6,7,8,11,40]. Specifically, eighty percent of the catch in the SCS comes from the nearshore waters along the northern coast of the SCS [5]. However, the fishery conflicts between surrounding countries of the SCS occurred frequently in recent years due to reduced fishery resources [7].

2.2. Data Sources

Multiple types of satellite remote sensing data in 2021 were used in this study, including nighttime light data, cloud mask, Chl-a, SST, sea surface salinity (SSS), and digital elevation model (DEM) (Table 1).
The nighttime light data used in this study is the Day/Night Band (DNB) data obtained by NPP/VIIRS. The transit time of the NPP satellite is 13:30 (local solar time) every day, and the spatial resolution of the sub-satellite point of DNB data is 750 m. Compared with DMSP/OLS data, the radiation resolution of VIIRS/DNB data has been significantly improved, and the recorded range has reached 14 bits, avoiding the defect of data supersaturation. The VIIRS/DNB sensor recording level (SDR) product (SVDNB) was selected as the data source of nighttime remote sensing images. It is a radiometric product that has undergone geometric correction and radiometric calibration. Meanwhile, to obtain the corresponding lunar phase Angle and other parameters for geographic correction, the geospatial reference data of SVDNB (VIIRS/DNB SDR Ellipsoid Geolocation product, GDNBO) were downloaded. Specifically, the VIIRS Cloud Mask product was used to perform a cloud mask on the images with poor lunar phase conditions to eliminate the influence of moonlight. The VIIRS/ DNB and corresponding data can be obtained from the Comprehensive Large Array-data Stewardship System (CLASS) released by NOAA (https://www.aev.class.noaa.gov/saa/products/welcome (accessed on 15 March 2024)).
The boat detection product (VIIRS Boat Detection, VBD) developed based on VIIRS/DNB data was used for determining the specific location of nighttime ships [23]. The VBD data contains the geographical locations of the bright spot pixels, radiation intensity, satellite passage time, satellite observation zenith angle (VZA), etc. According to the Quality Flag (QF) among them, the identified highlights can be classified into the following categories: QF = 1 indicates strong detection of the ship, QF = 2 indicates weak detection of the ship, and QF = 3 indicates fuzzy detection of the ship. In this study, the VBD data was downloaded from the NOAA Earth Observation organization EOG products (https:// eogdata.mines.edu/products/vbd/ (accessed on 25 September 2025)).
SST and Chl-a products are synthetic monthly data with spatial resolutions of 0.01° and 4 km, respectively. They were downloaded from the NASA ERDDAP data server (http://coastwatch.pfeg.noaa.gov/erddap/files/ (accessed on 9 March 2024)). The DEM data was released by NGDC ETOPO global terrain model data (https://www.ncei.noaa.gov/products/etopo-global-relief-model (accessed on 22 March 2024)), with the spatial resolution of 15 arc seconds (greater than 0.5 km). The SSS products in Global Ocean Physical Reanalysis (CMEMS) cmems_obs-mob_glo_phy-sss_my_multi_P1M was provided by the Copernicus National Centers for Environmental Information. It is monthly data with a spatial resolution of 0.125 (https://data.marine.copernicus.eu/products/ (accessed on 13 March 2024)).

3. Methodology

In this study, the Kriging method was introduced to interpolate missing values in environmental variables, used the generalized additive model (GAM) to explore the correlations between nighttime fishery activities and environmental variables, and used the geographic center index to explore the spatial distribution rules of fishery activities. Besides, we used the newly developed EDNA-Net to identify nighttime light ships to represent the nighttime fishery activities in the NWSCS in 2021 [32]. The EDNA-Net algorithm was selected because of its good performance to detect small vessels under low-light conditions, compared to other existing deep learning methods. The GAM algorithm was used because it can well demonstrate the nonlinear relationship between the response variable and explanatory variables, and this kind of method has been widely used for analyzing the relationship between fishery activities and environmental factors [33,36,38].

3.1. Kriging Interpolation

We used the ordinary kriging method to interpolate the missing values in the environmental variables used in this study. Kriging is a commonly used interpolation method, which can be used to infer the values of specific locations without observations (named unknown points) based on values from neighboring locations with observations (named known points) in geographic space. The basic principle of the method is to assume that the value of the unknown point is related to the value of surrounding known points, the distance, and the spatial relationship between these known points. The Kriging method describes the spatial autocorrelation of a random function by analyzing the spatial relationship between sampling points in detail and establishing a semi-variance function. The values of the sampling points are weighted to average, where the weights are provided by the semi-variance function, and the estimated values of the interpolation points are obtained. Commonly used semi-variance functions include the exponential function, the Gaussian function, etc. Different semi-variance functions are used according to different spatial variability models, thus suitable for interpolating various geospatial data such as soil moisture, air temperature, and precipitation [41]. It can be expressed as Equation (1).
z ^ o = ∑ i = 1 n λ i z i
where z ^ o is the point pending interpolation, zi is the observation of a known data point; λi is the weight, which can be calculated based on the spatial distance and the parameters of the semi-variance function.

3.2. Generalized Additive Model

A GAM was used to fit the correlations between fishery activities (i.e., nighttime boats identified based on nighttime light data and boat detection data listed in Table 1 using the EDNA-Net algorithm [32]) and corresponding factors (i.e., Chl-a, SST, SSS, and latitude and longitude) in the NWSCS. These factors were selected since fishery activities are highly correlated with the spatial distribution of fish, which is mainly decided by these environmental factors, and these factors have also been used in existing studies [20,22,33,36,37,38]. (1) Chl-a is highly correlated with the food of fish. Chl-a concentration is an important indicator of the biomass of marine phytoplankton. Phytoplankton synthesize organic matter through photosynthesis to provide food for zooplankton such as krill and copepods. Zooplankton then became the prey of small fish and eventually supported the entire fish community. Therefore, sea areas with high chlorophyll concentrations (such as upwelling zones and near estuaries) are usually highly productive and can attract fish to gather and forage [27,42]. (2) SST is highly correlated with the “survival threshold” and “behavioral orientation” of fish. Fish are poikilothermic animals, with their body temperature varying with the ambient temperature. Their metabolic rate, growth cycle, and reproductive activities all depend on suitable temperature conditions. Therefore, SST is a key limiting factor in determining the distribution range of fish stocks. Specifically, each type of fish has a specific “optimal temperature range” and exceeding this range can lead to metabolic disorders or even death [35]. Meanwhile, temperature changes can trigger the reproductive behaviors of fish, such as egg-laying and hatching [42]. (3) SSS is highly correlated with the “osmotic regulatory barrier” of fish. Salinity limits the distribution range of fish by affecting their osmotic balance (the exchange of water and salt in their bodies). (4) Latitude and longitude indirectly determine the macroscopic distribution pattern of fishes by correlating environmental gradients such as temperature, light, and ocean currents. Latitude mainly influences the intensity of solar radiation and temperature gradients, forming a “latitudinal environmental zone” from the equator to the poles, and thereby dividing the latitudinal distribution range of fishes. Longitude mainly affects the distribution of fish by correlating the distribution of land and sea, and ocean current systems. Specifically, the GAM was implemented based on monthly data in 2021, where the monthly nighttime boats data were accumulated from daily nighttime boats data. For each day, one boat was recorded when a pixel was recognized as a vessel pixel, and the corresponding location of this pixel was also recorded in the resulting image using the EDNA-Net algorithm. Accordingly, the number of nighttime boats for each month can be calculated based on the daily data.
The GAM is a flexible non-parametric statistical model, which can be used to fit complex nonlinear relationships between dependent variables and predictors [43]. The principle of GAM is to minimize residuals while maximizing minimalism, and it combines Generalized Linear Models (GLMs) and nonlinear smoothing techniques to allow for more flexible modeling of data. GAM consists of four parts, including the linear component, the nonlinear component, the connection function, and the interaction. Linear components are used to capture linear relationships between predictors and are usually represented as traditional linear regression terms. The nonlinear component is composed of a smoothing function fi(), which is used for nonlinear modeling of predictor variables, based on models such as the spline function, local regression function, kernel smoothing function, etc. The connection function g(·) defines the relationship between the expectation of the response variable and the linear component. Interaction refers to the interaction between different predictors.
The advantage of GAM is that it can flexibly fit a variety of complex nonlinear relationships without assuming a specific function form. At the same time, the smoothing function is used in the fitting process to have strong robustness to outliers and noise. The general form of the GAM can be expressed as Equation (2):
g ( E ( Y ) ) = β 0 + f 1 ( X 1 ) + f 2 ( X 2 ) + … + f p ( X p )
where g(·) is the connection function, which is used to associate the expected response variable Y with the combination of linear predictors; E(Y) represents the expectation of the response variable; β0 is a constant term; and fi() is a nonlinear smoothing function used to model the predictor Xi.
The GAM was defined as Equation (3) in this study:
x = s 0 ( L o n ) + s 1 ( L a t ) + s 2 ( S S T ) + s 3 ( C h l − a ) + s 4 ( S S S ) + ε
where x is the total monthly fishing activities of each grid in the study area in 2021; s0, s1, s2, s3, and s4 represent spline smoothing functions of covariates longitude, latitude, SST, Chl-a, and SSS, respectively. ε is the random error.
There are several structural parameters in the GAM, i.e., degrees of freedom (d.f.), effective degrees of freedom (edf), and deviance explained. Among them, d.f. represents the number of theoretical basis functions/the number of parameters, which can control the fitting degree of the spline function. edf reflects the degrees of freedom used by the model during the fitting process, where edf = 1 indicates a linear relationship between variables, while edf > 1 indicates a nonlinear relationship between variables. Deviance explained refers to the degree of explanation of independent variables to dependent variables in the established model, and the contribution rate of each variable is expressed as the ratio of its deviation explanation rate to the total deviation explanation rate of the model. The larger the value, the higher the interpretability and the more obvious the influence.
We divided the study area into grids with 0.5° × 0.5° spatial resolution and calculated the average values of fishery activities, Chl-a, SST, and SSS in each grid. The values of the latitude and longitude were extracted from the center point of the grid, and the remaining covariables are the monthly average within the grid. Accordingly, the GAM listed in Equation (3) was used to fit the correlations between fishery activities and corresponding factors.

3.3. Fishery Activity Center Index

To explore the spatial variations in nighttime fishery activities, we calculated the monthly spatial distribution center index of fishery activities in the NWSCS in 2021 [44]. This index can measure the degree of concentration of fishery activities [45], which is defined as Equation (4):
X ¯ = ∑ k = 1 s X k / s Y ¯ = ∑ k = 1 s Y k / s
where ( X ¯ , Y ¯ ) indicates the longitude and latitude position of the fishery activity center in each month within the study area, (Xk, Yk) is the longitude and latitude position of the kth detected fishery activity, and s represents the total number of detected fishery activities in each month.

4. Results and Discussion

4.1. Spatiotemporal Analysis of Fishing Vessels

The spatial pattern of nighttime fishery activities varied for different months in the NWSCS but showed specific patterns in different subregions (Figure 2). Fishery activities were mainly distributed in the waters of BG and the southwest of Hainan Island (SWHI), showing banded and clustered distribution patterns. Specifically, the waters of BG were clustered and distributed in clumps, and there was a continuous banded distribution in the nearshore area of Hainan Island. On the contrary, fishery activities in the southeast of Hainan Island (SEHI) and the waters of the Pearl River Estuary (PRE) were relatively low. The fishery activities in the PRE had the characteristics of multi-point dispersion, and the nearshore areas were parallel to the coastline. The fishery activities in the SEHI were dotted and distributed at multiple points. These findings are generally consistent with the existing study, that is, the nearshore waters along the northern coast of the SCS are the dominant fishery grounds [5]. Compared with the previous study at the seasonal scale based on statistical fishery data in the NWSCS [36], this study was implemented based on nighttime light remote sensing data to get fishery data, which can provide spatial details on fishery activities at a monthly scale. That is, more spatiotemporal details can be presented by this study, compared to the existing study [36].
The overall intensity of fishing activity (i.e., the total number of nighttime fishing boats) across all geographic grids in each subregion of the NWSCS was investigated (Figure 3). The intensities of fishing activities in the BG and the southwest of Hainan Island were generally higher than those in the PRE and the SEHI. Meanwhile, there were obvious seasonal differences in the intensity of fishery activities in the NWSCS, and the corresponding order was as follows: spring (38.8%) > summer (28.6%) > autumn (18.1%) > winter (14.4%). The overall distribution was trimodal, with the peak values appearing in April, August, and November, which is consistent with (same trend but different location of peaks) the study by Wang et al. in the whole SCS [28]. This may be due to the implementation of fishing moratorium policy (the fishing moratorium period was from May 1 to August 16, for three and a half months) in the SCS [46].
Specifically, there were still nighttime fishing boats in June and July in the NWSCS (Figure 3), possibly due to the existence of a large volume of illegal, unreported, and unregulated (IUU) fishing activities [7]. The fishing moratorium period was proposed and implemented for Chinese fishing vessels. As shown in Figure 2, there are a lot of fishing boats near the coastline of Hainan Island in April and August (Figure 2d,h), while there are only a few fishing boats in the same area from May to July (Figure 2e–g), indicating the effectiveness of the fishing moratorium policy. However, there were also a lot of IUU fishing activities from coastal countries surrounding the SCS, which were hard to control and often led to fishery conflicts in the sea areas far from the coastlines [4,7,8].
The number of fishing vessels in October-December-February (i.e., cold season) was significantly less than that in summer in the NWSCS (Figure 3), which may be caused by factors such as weather conditions, fish activity patterns, and fishery policies [47]. First, the SCS is affected by the northeast monsoon in the cold season, with strong winds and rough waves. The risk of going out to sea for fishing is high, and the safety of fishing boats and fishermen is difficult to guarantee. Therefore, many fishermen may choose to reduce the frequency of their sea operations. Second, due to factors such as the drop in water temperature, some fish species will migrate to deep-sea or low-latitude areas with more suitable water temperatures in the cold season. Meanwhile, the distribution of fish is relatively scattered, which increases the difficulty of fishing, leading to reduced willingness for fishing among fishermen. Third, after a period of fishing, some fishermen may have completed a certain amount of catch. Considering the cost and risk of fishing in the cold season, they may choose to reduce fishing frequencies, resulting in a decrease in the number of fishing boats.

4.2. Trend Analysis of Fishery Activities

The fisheries activity center index indicated that the fishery activity centers in the NWSCS were changing throughout the year, and most of the monthly centers were concentrated in the west bank and coastal areas of Hainan Island (Figure 4), which can be explained by the results in Figure 2 and Figure 3. From January to February, the center moved to the northeast, indicating a large growth trend of fishery activities in the PRE. From February to April, the center shifted vertically with large amplitude, due to the fishing activities in both the southwest and southeast of Hainan Island increasing rapidly. From April to May, the center moved to the southwest, indicating the intensity of fishery activity in the southwest of Hainan Island reached the highest value of the whole year. From May to July, the center moves to the northwest, indicating the reduced fishery activities in both the SWHI and the SEHI, due to the fishing moratorium policy for Chinese vessels [46]. From July to September, the center moved toward the northeast because the fishery intensity in the PRE area continued to increase, while the fishery activities in other waters gradually weakened. From October to December, the center gradually moved to the BG, and the overall intensity of fishery activity decreased, and the offshore fishery became the main fishery activity.

4.3. Analysis of Environmental Factors

4.3.1. Variations in Chl-a

The overall chlorophyll level in the NWSCS showed obvious seasonal characteristics, which were the highest in winter (1.15 mg/m3) and the lowest in spring (0.62 mg/m3) (Figure 5a). The results are different from existing studies (i.e., Chl-a level was the highest in winter but the lowest in summer) [36,42], which may be caused by increased marine aquaculture in the coastal areas in recent years (i.e., increased intensity of human activities) [48,49,50]. Among the sub-sea areas, Chl-a concentrations in the BG were the highest, followed by the PRE, which was consistent with the overall change trend of the whole study area. Chl-a concentrations in the SEHI were always the lowest among these subregions.
There was a strong correlation between the total chlorophyll level and the offshore distance in the NWSCS (Figure 6). The closer to the coastline, the higher Chl-a concentrations in the NWSCS, which is consistent with existing studies [36,42]. Among them, the areas with large changes in chlorophyll content mainly included the coastal areas of the BG, the nearshore areas of the PRE, the Qiongzhou Strait, and the coastal areas on the west side of Hainan Island. Specifically, the spatial range of chlorophyll content in the NWSCS tightened and continuously decreased from January to April. There was a regional transfer of chlorophyll content from May to August, but there were only a few overall changes. The high chlorophyll value in the offshore area gradually spread from September to December, and reached its highest, possibly due to the interaction between the decrease in water temperature and the northeast monsoon weakened the stratification phenomenon in the central sea area [42].

4.3.2. Variations in SST

The variation trend of SST in each sub-sea area was basically the same as that in the whole study area, showing a unimodal distribution, where the lowest SST appeared in January and the highest value appeared in August (Figure 5b). The monthly average SSTs in the BG and the PRE were generally lower than those in the SWHI and the SEHI in winter, while having fewer differences in summer.
There was a seasonal relationship between the distribution of fishery activities and the level of SST in the NWSCS (Figure 7). From January to March, there were more fishing boats at night in coastal areas, and the fishing activities were still mainly distributed in the waters of the BG and the SWHI when the SST was low. With the increase in SST, the number of fishing vessels at night increased in the waters around Hainan Island and in the offshore areas from April to September, and the center of fishing activities gradually moved from the coastal area to the deep-sea area. From October to December, the SST gradually decreased from offshore to the coast, and the spatial distribution of fishing boats also showed a similar change, and the distribution was redistributed in the coastal waters.
Figure 5b and Figure 7 indicate that SST also had a seasonal distribution in the NWSCS. As SST rises, fishing activities expanded from inshore to offshore areas. SST was distributed in a stepped band parallel to the coastline in winter, i.e., a distinct latitudinal gradient. The east sea area of Hainan Island was generally warmer than the west sea area in summer, i.e., a distinct longitudinal gradient. These findings are also consistent with existing studies [36,42], where the spatial distribution patterns of SST in different seasons may be caused by different seasonally reversing monsoons from different directions [42].

4.3.3. Variations in SSS

The content of SSS in the NWSCS and subregions was generally high in spring and winter (mean SSS in winter is 33.36 Practical Salinity Unit (PSU) and 33.6 PSU in spring) and low in summer and autumn (Figure 8). The highest SSS value appeared in March, and the lowest value appeared in November. SSS values in the SWHI and the BG were lower than those in the SEHI and the PRE throughout the year and were generally lower than the overall SSS level in the NWSCS. This is consistent with the existing study in the same study area [36].
Fishery activities in the NWSCS were correlated with SSS content to a large extent (Figure 8 and Figure 9). The BG and the SWHI were the two sub-regions with the highest intensity of fishery activities in the NWSCS, and their salinity levels were generally lower than those in the weak fishing areas. At the same time, fishery activities were very frequent from March to October, when the overall SSS content in the NWSCS continued to decline. SSS in the BG varied significantly throughout the year, with a wide fluctuation range, showing a gradual increase from northwest to southeast waters, which is consistent with the existing study [36].

4.3.4. Impacts of Environmental Factors

We further analyzed the effects of Chl-a, SST, SSS, latitude, and longitude on fishery activities in the NWSCS using the GAM listed in Equation (3), where the maximum AIC value (AIC = 25,312.06) was set to get the corresponding parameters (Table 2). The results showed that Chl-a, SST, longitude, latitude, and SSS had high impacts on fishery activities in the NWSCS (p < 0.05) and could explain 48.6% of the variance of fishery activities. Among them, the marine environmental factors (SSS, SST, and Chl-a) had significant impacts on the model, and the total contribution rate was 38.7%. The contribution of spatial factors (latitude and longitude) to the model was only 9.9%.
As shown in Table 2, Chl-a contributes most to fishery activities compared to SSS and SST. The reason may be that Chl-a can directly decide the food source, which is the most basic factor for the existence of fish. While SSS and SST only decide the habitat ranges of fish. If there was not enough Chl-a, no matter how suitable the temperature and salinity were, the number of fish would not be high. Specifically, SSS determines the ‘survival range’ and is a ‘prerequisite’ for the existence of fishery resources. Accordingly, it can only eliminate unsuitable areas but cannot promote high yields. SST shows a lower contribution rate compared to SSS (Table 2), possibly because it mainly influences the physiological efficiency of fish but is not as crucial as SSS.
There were nonlinear relationships between fishery activities and various factors in the study area (Figure 10). The impacts of Chl-a, SSS, SST, and longitude on fishery activities were statistically significant, while the impacts of latitude were weak.
Among marine environmental factors, fishery activities were the most contributed in areas with Chl-a of 0~0.35 mg/m3, SST of 21.2~26.4 °C, and SSS of 32.9~33.8 PSU. When the concentration of Chl-a was about 0.35 mg/m3, it had the greatest impact on fishery activities. When the concentration of Chl-a was less than this value, the increase in its concentration can promote fishery activities with a slightly increased confidence interval. On the contrary, when the concentration of Chl-a was greater than 0.35 mg/m3, the increase in Chl-a concentration can demote fishing activities. When SSS was about 33.8 PSU, it had the greatest impact on fishery activities. When SSS was less than 32.6 PSU (in the range of 32.9–33.8 PSU), the increase in SSS had a positive effect on fishery activities with a decreased (small and stable) confidence interval. When SSS was in the range of 32.6–32.9 PSU, the increase in SSS harmed fishery activities. The influence of SST on fishery activities had two extreme values (SST = 18.6 °C and 26.4 °C), indicating that the influence was the greatest when the SST was around these two values. However, the confidence interval of the extreme point (18.6 °C) was large, and the reliability was low. When SST was in the range of 21.2~26.4 °C, the increase in SST had a positive impact on fishery activities with a small and stable confidence interval.
Among the spatial factors, the contribution of longitude to the model was 9.4%. In the area with a longitude less than 108°24′ E, the increase in longitude had no obvious effect on fishery activities, while in the area with a longitude greater than 108°24′ E, the intensity of fishery activities dropped sharply with the increase in longitude, indicating that the main fishery activities were distributed in the area with a small longitude.

4.4. Implications for Sustainability

This study reveals that more efforts are required to prevent IUU fishing activities for the sustainable development of marine ecosystems in the NWSCS. Although the Chinese government has made great efforts to significantly reduce IUU fishing activities and declined catch volumes of neighboring countries in the SCS [7,51], there are still a large amount of fishing activities detected during the fishing moratorium period in the NWSCS (Figure 2 and Figure 3). This is because the fishing moratorium policy only applies to Chinese vessels, and a large volume of illegal fishing vessels from neighboring countries were still working from May to July. Therefore, more efforts are required to ban long-range fishing in the NWSCS during the fishing moratorium period.
It is also necessary to improve remote sensing technology for making sustainable fishing plans in the NWSCS. Remote sensing data are important sources for detecting fishery vessels and marine environmental factors. According to this study, marine environmental factors such as Chl-a, SSS, and SST show significant impacts on the distributions of fishery activities in the NWSCS. Accurate information on these factors can be used for predicting the best-suited waters for different kinds of fish. Accordingly, fishermen can make proper strategies to catch specific kinds of fish. Meanwhile, it is necessary to accurately obtain the distributions of fishing vessels based on remote sensing data. The IUU fishing activities can be distinguished by combining remote sensing and vessel monitoring system (VMS) data [24,52]. Accordingly, policies for sustainable fishing activities can be implemented, e.g., limiting the number of fishing boats in fishing grounds with a high risk of long-term overcapacity.

4.5. Limitations and Prospects

The study on the relationship between marine environmental factors and fishery resources is limited to only one year, lacking the long-term trend analysis. Meanwhile, the nighttime light remote sensing data are only available for a specific time of day, which may lead to the missed detection of fishing vessels. There is also an assumption that all the identified nighttime boats belong to fishing vessels without further type distinction. Future studies will further expand the data coverage from one year to multiple years to explore possible impacts from fishery management policies. Meanwhile, using nighttime light remote sensing data from different satellites may mitigate the miss-detection issue. Besides, more auxiliary information can be fused to distinguish different types of fishing boats, especially to identify the IUU fishing activities.

5. Conclusions

We explored the spatiotemporal variations in night-time fishing boats and analyzed the relationship between fishing activities and marine environmental factors in the NWSCS.
The spatial variation characteristics of night vessels and corresponding marine environmental factors (i.e., Chl-a, SSS, SST, latitude, and longitude) in each month of 2021 in the NWSCS were studied. Meanwhile, the nonlinear relationship between spatial, environmental factors and fishery activities was quantitatively studied using the generalized additive model. The results illustrated that Chl-a contributed the most to fishery activities (21.7%) in the range of 0~0.35 mg/m3, SSS in the range of 32.9~33.8 PSU, and SST in the range of 21.2~26.4 °C. This study can promote our understanding of the nighttime fishery activities and main impact factors in the NWSCS, which is important for the management and scientific management of fishery resources in this region. This study reveals that more efforts are required to prevent IUU fishing activities for the sustainable development of marine ecosystems in the NWSCS. It is also necessary to improve remote sensing technology for making sustainable fishing plans in the NWSCS. Future work can cover a longer period and combine multiple data sources to accurately distinguish different types of fishing vessels for making more scientific management policies.

Author Contributions

Z.L. implemented the research and carried out the analysis; G.Z. wrote the original manuscript; T.Z. supervised the experiments and revised the manuscript; J.Z. reviewed and revised the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Project “Meteorological Satellite Application System”, Sichuan Science and Technology Program (Grant No. 2024YFHZ0340).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

We would like to thank the editors and the three anonymous reviewers for their constructive comments and suggestions.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Hendrix, C.S.; Glaser, S.M.; Lambert, J.E.; Roberts, P.M. Global climate, El Nino, and militarized fisheries disputes in the East and South China Seas. Mar. Policy 2022, 143, 105137. [Google Scholar] [CrossRef]
  2. Zhao, X.; Li, D.; Li, X.; Zhao, L.; Wu, C. Spatial and seasonal patterns of night-time lights in global ocean derived from VIIRS DNB images. Int. J. Remote Sens. 2018, 39, 8151–8181. [Google Scholar] [CrossRef]
  3. Golden, C.D.; Allison, E.H.; Cheung, W.W.L.; Dey, M.M.; Halpern, B.S.; McCauley, D.J.; Smith, M.; Vaitla, B.; Zeller, D.; Myers, S.S. Fall in fish catch threatens human health. Nature 2016, 534, 317–320. [Google Scholar] [CrossRef]
  4. Cyranoski, D. South China Sea ruling sparks conservation fears. Nature 2016, 535, 334–335. [Google Scholar] [CrossRef]
  5. Chen, S.; Li, L.; Yang, X.; Wu, Y. The main types and utilization of fishery resources for Marine fishing and aquaculture in the South China Sea and its four coastal provinces. China Agric. Informait. 2015, 12, 15–19. [Google Scholar]
  6. Jiang, Y.; Xue, X. Building a cross-strait cooperation mechanism for the conservation and management of fi shery resources in the South China Sea. Ocean Coast. Manag. 2015, 116, 318–330. [Google Scholar] [CrossRef]
  7. Yu, C.; Chang, Y. China’s Incentives and Efforts against IUU Fishing in the South China Sea. Sustainability 2023, 15, 7255. [Google Scholar] [CrossRef]
  8. Wu, W.; Zhang, P.; Wang, Q.; Kang, L.; Su, F. Analysis of fishing intensity in the South China Sea based on automatic identification system data: A comparison between China and Vietnam. Mar. Coast. Fish. Dyn. Manag. Ecosyst. Sci. 2024, 16, 10309. [Google Scholar] [CrossRef]
  9. Halpern, B.S.; Walbridge, S.; Selkoe, K.A.; Kappel, C.V.; Micheli, F.; D’Agrosa, C.; Bruno, J.F.; Casey, K.S.; Ebert, C.; Fox, H.E.; et al. A Global Map of Human Impact on Marine Ecosystems. Science 2008, 319, 948–952. [Google Scholar] [CrossRef]
  10. Ju, H. Studies on the Decrement of Fishery Resource in the South China Sea. Southeast Asian Stud. 2012, 2012, 51–55. [Google Scholar] [CrossRef]
  11. Pang, Y.; Tian, Y.; Fu, C.; Wang, B.; Li, J.; Ren, Y. Variability of coastal cephalopods in overexploited China Seas under climate change with implications on fi sheries management. Fish. Res. 2018, 208, 22–33. [Google Scholar] [CrossRef]
  12. Arkhipkin, A.I.; Rodhouse, P.G.K.; Pierce, G.J.; Sakai, M.; Allcock, L.; Arguelles, J.; Bower, J.R.; Ceriola, L.; Chen, C.; Chen, X.; et al. World Squid Fisheries. Rev. Fish. Sci. Aquac. 2015, 23, 92–252. [Google Scholar] [CrossRef]
  13. Li, J.; Qiu, Y.; Cai, Y.; Zhang, K.; Zhang, P.; Jing, Z.; Wu, Q.; Ma, S.; Liu, H.; Chen, Z. Trend in fishing activity in the open South China Sea estimated from remote sensing of the lights used at night by fishing vessels. ICES J. Mar. Sci. 2022, 79, 230–241. [Google Scholar] [CrossRef]
  14. Guan, Y.; Zhang, X.; Gao, G.; Cao, C.; Li, Z.; Fu, S.; Liu, G. A new indicator for assessing fishing ecological pressure using multi-source data: A case study of the South China Sea. Ecol. Indic. 2025, 170, 113096. [Google Scholar] [CrossRef]
  15. Zhang, K.; Su, L.; Chen, Z.; Qiu, Y. An extensive assessment of exploitation indicators for multispecies fisheries in the South China Sea to inform more practical and precise management in China. Ecol. Indic. 2025, 173, 113363. [Google Scholar] [CrossRef]
  16. Liu, Z.; Lu, W.; Wang, T.; Zhang, Y.; He, L.; Yang, L.; Deng, L. The assessment of carrying capacity of marine fishery resources in China. Front. Mar. Sci. 2025, 11, 1518235. [Google Scholar] [CrossRef]
  17. Levin, N.; Kyba, C.C.M.; Zhang, Q.; De Miguel, A.S.; Román, M.O.; Li, X.; Portnov, B.A.; Molthan, A.L.; Jechow, A.; Miller, S.D.; et al. Remote sensing of night lights: A review and an outlook for the future. Remote Sens. Environ. 2020, 237, 111443. [Google Scholar] [CrossRef]
  18. Li, J.; Cai, Y.; Zhang, P.; Zhang, Q.; Jing, Z.; Wu, Q.; Qiu, Y.; Ma, S.; Chen, Z. Satellite observation of a newly developed light-fishing “hotspot” in the open South China Sea. Remote Sens. Environ. 2021, 256, 112312. [Google Scholar] [CrossRef]
  19. Straka, W.C.; Seaman, C.J.; Baugh, K.; Cole, K.; Stevens, E.; Miller, S.D. Utilization of the suomi national polar-orbiting partnership (npp) visible infrared imaging radiometer suite (viirs) day/night band for arctic ship tracking and fisheries management. Remote Sens. 2015, 7, 971–989. [Google Scholar] [CrossRef]
  20. Li, Y.; Song, L.; Zhao, S.; Zhao, D.; Wu, Y.; You, G.; Kong, Z.; Xi, X.; Yu, Z. Nighttime fishing vessel observation in Bohai Sea based on VIIRS fishing vessel detection product (VBD). Fish. Res. 2023, 258, 106539. [Google Scholar] [CrossRef]
  21. Tian, H.; Liu, Y.; Li, J.; Xing, Q.; Sun, H.; Tian, Y. Satellite nighttime remote sensing promotes the spatially refined monitoring and assessment of offshore fishery. Int. J. Digit. Earth 2024, 17, 2322762. [Google Scholar] [CrossRef]
  22. Li, J.; Zhang, Z.; Zhang, K.; Fan, J.; Liu, H.; Qiu, Y.; Li, X.; Chen, Z. Tracking the Development of Lit Fisheries by Using DMSP/OLS Data in the Open South China Sea. Remote Sens. 2024, 16, 3678. [Google Scholar] [CrossRef]
  23. Elvidge, C.D.; Zhizhin, M.; Baugh, K.; Hsu, F. Automatic Boat Identification System for VIIRS Low Light Imaging Data. Remote Sens. 2015, 7, 3020–3036. [Google Scholar] [CrossRef]
  24. Li, J.; Zhang, P.; Cai, Y.; Zhang, Q.; Zhang, K.; Jing, Z.; Wu, Q.; Qiu, Y.; Ma, S.; Chen, Z. Performance of VMS and nightly satellite in monitoring light fishing vessels in the open South China Sea. Fish. Res. 2021, 243, 106100. [Google Scholar] [CrossRef]
  25. Baek, W.-K.; Kim, E.; Jeon, H.-K.; Lee, K.-J.; Kim, S.-W.; Lee, Y.-K. Monitoring Maritime Ship Characteristics Using Satellite Remote Sensing Data from Different Sensors. Ocean Sci. J. 2024, 59, 8. [Google Scholar] [CrossRef]
  26. Kamaruzzaman, Y.N.; Mustapha, M.A. An overview Assessment of the Effectiveness of Satellite Images and Remote Sensing in Predicting Potential Fishing Grounds and its Applicability for Rastrelliger kanagurta in the Malaysian EEZ off the South China. Rev. Fish. Sci. Aquac. 2023, 31, 320–341. [Google Scholar] [CrossRef]
  27. Geronimo, R.C.; Franklin, E.C.; Brainard, R.E.; Elvidge, C.D.; Santos, M.D.; Venegas, R.; Mora, C. Mapping Fishing Activities and Suitable Fishing Grounds Using Nighttime Satellite Images and Maximum Entropy Modelling. Remote Sens. 2018, 10, 1604. [Google Scholar] [CrossRef]
  28. Wang, W. Spatio-Temporal Distribubtion of Fishery Production Intensity in the South China Sea Based on the Night-Time Images. Master’s Thesis, University of Chinese Academy of Sciences, Beijing, China, 2016. [Google Scholar]
  29. Zhang, S. Research on Fishing Dynamic Changes in the South China Sea Using Nighttime Light Data. Master’s Thesis, Nanjing University, Nanjing, China, 2017. [Google Scholar]
  30. Shao, J.; Yang, Q.; Luo, C.; Li, R.; Zhou, Y.; Zhang, F. Vessel Detection from Nighttime Remote Sensing Imagery Based on Deep Learning. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2021, 14, 12536–12544. [Google Scholar] [CrossRef]
  31. Tsuda, M.E.; Miller, N.A.; Saito, R.; Park, J.; Oozeki, Y. Automated VIIRS Boat Detection Based on Machine Learning and Its Application to Monitoring Fisheries in the East China Sea. Remote Sens. 2023, 15, 2911. [Google Scholar] [CrossRef]
  32. Zuo, G.; Zhou, J.; Meng, Y.; Zhang, T.; Long, Z. Night-Time Vessel Detection Based on Enhanced Dense Nested Attention Network. Remote Sens. 2024, 16, 1038. [Google Scholar] [CrossRef]
  33. Fei, Y.; Yang, S.; Fan, W.; Shi, H.; Zhang, H.; Yuan, S. Relationship between the Spatial and Temporal Distribution of Squid-Jigging Vessels Operations and Marine Environment in the North Pacific Ocean. J. Mar. Sci. Eng. 2022, 10, 550. [Google Scholar] [CrossRef]
  34. Chen, R.; Wang, Y.; Wu, X.; Gao, Z. Identifying the key factors influencing spatial and temporal variations of regional coastal fishing activities. Ocean Coast. Manag. 2024, 248, 106940. [Google Scholar] [CrossRef]
  35. Ji, S.; Zhou, W.; Wang, L.; Tang, F.; Wu, Z.; Chen, G. Relationship between temporal-spatial distribution of yellowfin tuna Thunnus albacares fishing grounds and sea surface temperature in the South China Sea and adjacent waters. Mar. Fish. 2016, 38, 9–16. [Google Scholar] [CrossRef]
  36. He, L. Research on the Relationship Between Key Environmental Factors and Typical Fishery Resources in the Northwestern South China Sea. Master’s Thesis, Guangdong Ocean University, Zhanjiang, China, 2022. [Google Scholar]
  37. Huang, W. Research on Light Fishing Grounds in the South China Sea Based on Remote Sensing Technology. Master’s Thesis, Dalian Ocean University, Dalian, China, 2018. [Google Scholar]
  38. Lu, Y.; Yu, J.; Lin, Z.; Chen, P. Environmental Influence on the Spatiotemporal Variability of Spawning Grounds in the Western Guangdong Waters, South China Sea. J. Mar. Sci. Eng. 2020, 8, 607. [Google Scholar] [CrossRef]
  39. Xie, L.; Cao, R.; Shang, Q. Progress of Study on Coastal Circulation near the Shore of Western Guangdong. J. Guangdong Ocean Univ. 2012, 32, 94–98. [Google Scholar]
  40. Tang, Q. Chinese Regional Oceanography: Fishery Oceanography; Ocean Press: Beijing, China, 2012. [Google Scholar]
  41. Wang, Y. Research of Spatial Interpolation Algorithm and its Application in Air Quality Monitoring. Master’s Thesis, Henan University, Kaifeng, China, 2010. [Google Scholar]
  42. Tang, D.; Kawamura, H.; Lee, M.; Dien, T. Van Seasonal and spatial distribution of chlorophyll- a concentrations and water conditions in the Gulf of Tonkin, South China Sea. Remote Sens. Environ. 2003, 85, 475–483. [Google Scholar] [CrossRef]
  43. Hastie, T.J. Generalized Additive Models. In Statistical Models in S; Routledge: London, UK, 2017; pp. 249–307. ISBN 9781118445112. [Google Scholar]
  44. Wu, J.; Guan, W. Study on the distribution and variation of fishing vessels in East China Sea and Yellow Sea based on the nighttime light remote sensing data from SNPP/VIIRS. J. Fish. Sci. China 2019, 26, 221–231. [Google Scholar] [CrossRef]
  45. Feng, Y.; Yu, W.; Lei, R. Spatial distribution Features and Controlling Factors of Traditional Villages in Guangdong Province. Sci. Geogr. Sinia 2017, 37, 236–243. [Google Scholar] [CrossRef]
  46. Notification of the Ministry of Agriculture and Rural Affairs on Adjusting the System of Moratorium on Marine Fishing in Midsummer. Available online: http://www.moa.gov.cn/govpublic/YYJ/201701/t20170120_5460478.htm (accessed on 17 July 2025).
  47. Chen, G.; Li, Y.; Chen, P. A study on spawning ground of blue mackerel scad (Decapterus maruadsi) in continental shelf waters of northern South China Sea. J. Trop. Oceanogr. 2003, 22, 22–28. [Google Scholar] [CrossRef]
  48. Hu, X.; Chen, C.; Yang, Z.; Liu, Z. Reliable, large-scale, and automated remote sensing mapping of coastal aquaculture ponds based on Sentinel-1/2 and ensemble learning algorithms. Expert Syst. Appl. 2025, 293, 128740. [Google Scholar] [CrossRef]
  49. Wang, Z.; Zhang, J.; Yang, X.; Huang, C.; Su, F.; Liu, X.; Liu, Y.; Zhang, Y. Global mapping of the landside clustering of aquaculture ponds from dense time-series 10 m Sentinel-2 images on Google Earth Engine. Int. J. Appl. Earth Obs. Geoinf. 2022, 115, 103100. [Google Scholar] [CrossRef]
  50. Meng, D.; Yang, X.; Wang, Z.; Liu, Y.; Zhang, J.; Liu, X.; Liu, B. Spatial Distribution and Differentiation Analysis of Coastal Aquaculture in China Based on Remote Sensing Monitoring. Remote Sens. 2024, 16, 1585. [Google Scholar] [CrossRef]
  51. Chen, W.; Ou, C.; Yang, M.; Shih, Y. China’s gray zone actions in the East China Sea, Taiwan Strait, and South China Sea: A comparative study and impact on fisheries. Mar. Policy 2024, 167, 106246. [Google Scholar] [CrossRef]
  52. Cope, S.; Tougher, B.; Zetterlind, V.; Gilfillan, L.; Aldana, A. Building a Practical Multi-Sensor Platform for Monitoring Vessel Activity near Marine Protected Areas: Case Studies from Urban and Remote Locations. Remote Sens. 2023, 15, 3216. [Google Scholar] [CrossRef]
Sustainability 17 08641 g001
Figure 1. The topography of the NWSCS. The subregions include four waters, i.e., BG: Beibu Gulf; PRE: Pearl River Estuary; SEHI: southeast of Hainan Island; SWHI: southwest of Hainan Island. The red dashed lines are the boundaries of subregions.
Figure 1. The topography of the NWSCS. The subregions include four waters, i.e., BG: Beibu Gulf; PRE: Pearl River Estuary; SEHI: southeast of Hainan Island; SWHI: southwest of Hainan Island. The red dashed lines are the boundaries of subregions.
Sustainability 17 08641 g001
Sustainability 17 08641 g002
Figure 2. Spatial distribution of fishery activities in the NWSCS in 2021.
Figure 2. Spatial distribution of fishery activities in the NWSCS in 2021.
Sustainability 17 08641 g002
Sustainability 17 08641 g003
Figure 3. Temporal variations in nighttime fishing boats in different subregions in 2021. SEHI: southeast of Hainan Island; PRE: Pearl River Estuary; SWHI: southwest of Hainan Island; BG: Beibu Gulf.
Figure 3. Temporal variations in nighttime fishing boats in different subregions in 2021. SEHI: southeast of Hainan Island; PRE: Pearl River Estuary; SWHI: southwest of Hainan Island; BG: Beibu Gulf.
Sustainability 17 08641 g003
Sustainability 17 08641 g004
Figure 4. Spatial distribution of fishery activity centers in the NWSCS in 2021. The numbers indicate the months of year 2021.
Figure 4. Spatial distribution of fishery activity centers in the NWSCS in 2021. The numbers indicate the months of year 2021.
Sustainability 17 08641 g004
Sustainability 17 08641 g005
Figure 5. Monthly variations in (a) Chl-a and (b) SST in the NWSCS in 2021. PRE: Pearl River Estuary; SEHI: southeast of Hainan Island; BG: Beibu Gulf; SWHI: southwest of Hainan Island; NWSCS: northwestern South China Sea.
Figure 5. Monthly variations in (a) Chl-a and (b) SST in the NWSCS in 2021. PRE: Pearl River Estuary; SEHI: southeast of Hainan Island; BG: Beibu Gulf; SWHI: southwest of Hainan Island; NWSCS: northwestern South China Sea.
Sustainability 17 08641 g005
Sustainability 17 08641 g006
Figure 6. Spatial distribution of fishery activities and chlorophyll a in the NWSCS in 2021.
Figure 6. Spatial distribution of fishery activities and chlorophyll a in the NWSCS in 2021.
Sustainability 17 08641 g006
Sustainability 17 08641 g007
Figure 7. Spatial distribution of fishery activities and SST in the NWSCS in 2021.
Figure 7. Spatial distribution of fishery activities and SST in the NWSCS in 2021.
Sustainability 17 08641 g007
Sustainability 17 08641 g008
Figure 8. Monthly variations in SSS in the NWSCS in 2021. PRE: Pearl River Estuary; SEHI: southeast of Hainan Island; BG: Beibu Gulf; SWHI: southwest of Hainan Island; NWSCS: northwestern South China Sea.
Figure 8. Monthly variations in SSS in the NWSCS in 2021. PRE: Pearl River Estuary; SEHI: southeast of Hainan Island; BG: Beibu Gulf; SWHI: southwest of Hainan Island; NWSCS: northwestern South China Sea.
Sustainability 17 08641 g008
Sustainability 17 08641 g009
Figure 9. Spatial distribution of fishery activities and SSS in the NWSCS in 2021.
Figure 9. Spatial distribution of fishery activities and SSS in the NWSCS in 2021.
Sustainability 17 08641 g009
Sustainability 17 08641 g010
Figure 10. Relationship between fishery activities and impact factors in the NWSCS (under 95% confidence interval). (a) Chl-a; (b) SSS; (c) latitude; (d) longitude; (e) SST. The grey shaded areas indicate the confidence interval of 95%.
Figure 10. Relationship between fishery activities and impact factors in the NWSCS (under 95% confidence interval). (a) Chl-a; (b) SSS; (c) latitude; (d) longitude; (e) SST. The grey shaded areas indicate the confidence interval of 95%.
Sustainability 17 08641 g010
Table 1. Satellite remote sensing data used in this study.
Table 1. Satellite remote sensing data used in this study.
Data TypeProduct NameSpatial ResolutionTemporal Resolution
Nighttime light dataVIIRS_SDR SVDNB750 mDaily
Cloud maskVIIRS Cloud Mask EDR750 mDaily
Boat detectionVBD-Daily
Chl-aChlor_a4 kmMonthly
SSTMURSST0.01°Monthly
SSSMOGOSSS0.125°Monthly
DEMNGDC ETOPO15 arcsec-
Table 2. Parameters of GAM-based relationships between fishery activities and corresponding spatial and environmental factors.
Table 2. Parameters of GAM-based relationships between fishery activities and corresponding spatial and environmental factors.
Environmental FactorsContribution Rate (%)Accumulated Contribution Rate (%)p-Valued.f.
s(Chl-a)21.721.73.2 × 10−14 ***8.53
s(SSS)12.834.52.6 × 10−12 ***5.52
s(Lon)9.443.90.6 × 10−3 **8.34
s(SST)4.248.15.8 × 10−8 ***6.75
s(Lat)0.548.60.024 *7.43
Note: *** indicates p < 0.001, ** indicates p < 0.01, * indicates p < 0.05, d.f. indicates the degree of freedom of the model.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Long, Z.; Zuo, G.; Zhang, T.; Zheng, J. Exploring Nighttime Fishing and Its Impact Factors in the Northwestern South China Sea for Sustainable Fisheries. Sustainability 2025, 17, 8641. https://doi.org/10.3390/su17198641

AMA Style

Long Z, Zuo G, Zhang T, Zheng J. Exploring Nighttime Fishing and Its Impact Factors in the Northwestern South China Sea for Sustainable Fisheries. Sustainability. 2025; 17(19):8641. https://doi.org/10.3390/su17198641

Chicago/Turabian Style

Long, Zhiyong, Gao Zuo, Tao Zhang, and Jinjun Zheng. 2025. "Exploring Nighttime Fishing and Its Impact Factors in the Northwestern South China Sea for Sustainable Fisheries" Sustainability 17, no. 19: 8641. https://doi.org/10.3390/su17198641

APA Style

Long, Z., Zuo, G., Zhang, T., & Zheng, J. (2025). Exploring Nighttime Fishing and Its Impact Factors in the Northwestern South China Sea for Sustainable Fisheries. Sustainability, 17(19), 8641. https://doi.org/10.3390/su17198641

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop