EN中文

Journal of Geographical Sciences >

2023 , Vol. 33 >Issue 12: 2377 - 2399

DOI: https://doi.org/10.1007/s11442-023-2181-z

Research Articles

Changes in the spatial distribution of mariculture in China over the past 20 years

  • LIU Yueming , 1 ,
  • WANG Zhihua 1, 2 ,
  • YANG Xiaomei , 1, 2, * ,
  • WANG Shaoqiang 2, 3 ,
  • LIU Xiaoliang 1, 2 ,
  • LIU Bin 1, 2 ,
  • ZHANG Junyao 1, 2 ,
  • MENG Dan 1, 2 ,
  • DING Kaimeng 1, 4 ,
  • GAO Ku 1, 2 ,
  • ZENG Xiaowei 1, 2, 5 ,
  • DING Yaxin 1, 6
Expand
  • 1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
  • 2. University of Chinese Academy of Sciences, Beijing 100049, China
  • 3. Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
  • 4. Jinling Institute of Technology, Nanjing 211169, China
  • 5. School of Geography and Information Engineering, China University of Geosciences, Wuhan 430074, China
  • 6. School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, Henan, China
* Yang Xiaomei (1970‒), PhD and Professor, specialized in coastal zone remote sensing, intelligent interpretation of remote sensing images and geologic understanding. E-mail:

Liu Yueming (1992‒), PhD, specialized in intelligent interpretation of remote sensing images, coastal zone remote sensing. E-mail:

Received date: 2022-10-25

  Accepted date: 2023-07-12

  Online published: 2023-12-14

Supported by

The National Key Research and Development Program of China(2021YFB3900501)

National Natural Science Foundation of China(42306246)

National Natural Science Foundation of China(42371473)

The Key Project of Innovation, LREIS(KPI001)

The Youth Project of Innovation, LREIS(YPI004)

Fold

Abstract

China's mariculture provides more than 60% of the world’s mariculture products and plays an important role in the world’s aquaculture and food supply. Research on changes in the spatial distribution pattern of China’s mariculture, however, remains lacking. To accurately reflect the changes in the spatial pattern of mariculture in China, in this study, we used multitemporal optical and synthetic aperture radar remote sensing images to enhance the characteristics of mariculture and extracted the spatial distribution data for mariculture in China in 2000, 2010, and 2020. Accordingly, we explored the distribution pattern and changes in mariculture in China. We found that, in 2020, China’s mariculture exhibited a distribution pattern of more in the north and less in the south. With the Yangtze River estuary as the boundary, the proportion of mariculture in northern China was 70.9%, and that in southern China was only 29.1%. This difference did not exist in 2000, but it emerged because of the rapid development of mariculture in northern China from 2010 to 2020. In addition, by superimposing the mariculture data with shoreline and water depth data, we found that more than 90% of China’s mariculture area was located in the sea area within 20 km of the shoreline and that more than 80% of the mariculture area was located in the sea area with water depths of less than 20 m. In addition, the spatial distribution of mariculture in China developed from near the shore and moved outward from shallow to deep water areas. We examined the driving factors that influence changes in the spatial distribution of mariculture in China. We argue that technological advancements in mariculture, as well as the intensive concentration of mariculture near the shore, policy constraints and incentives, and economic development, collaborate to shape the current pattern of mariculture development in China.

Key words： remote sensing; mariculture; coastal zone of China; spatiotemporal variation

Cite this article

LIU Yueming , WANG Zhihua , YANG Xiaomei , WANG Shaoqiang , LIU Xiaoliang , LIU Bin , ZHANG Junyao , MENG Dan , DING Kaimeng , GAO Ku , ZENG Xiaowei , DING Yaxin . Changes in the spatial distribution of mariculture in China over the past 20 years[J]. Journal of Geographical Sciences, 2023 , 33(12) : 2377 -2399 . DOI: 10.1007/s11442-023-2181-z

1 Introduction

Mariculture is an important means of using the sea in coastal areas. In addition to providing a large amount of healthy food for humans, mariculture also provides jobs and promotes regional development. Compared with traditional marine fishing, in mariculture, we do not need to worry about species extinction caused by overfishing. It is a green industry and is consistent with the concept of sustainable development. China has abundant mariculture, and its annual output of cultivated seafood exceeds 60% of the world’s total output, providing a large amount of seafood globally (FAO, 2018, 2022; Naylor et al., 2021). Although mariculture is a green industry, because of a lack of scientific understanding and management, China’s mariculture has led to a series of environmental problems, including eutrophication and deterioration of water quality caused by excess bait, damage to the original environment, and reduction in biodiversity (Ottinger et al., 2016). These problems not only have caused substantial economic losses but also have caused damage to the regional ecosystem, which has seriously threatened the sustainable development of the region. The key to solving these problems is to accurately determine the spatial location and scale information of mariculture areas and to better understand the development of and changes in mariculture in China.
According to the characteristics of a long time series and large area observations of remote sensing satellites, the extraction of mariculture areas can be realized (Nath et al., 2000). Previous studies, however, have focused on the extraction of mariculture areas in relatively small areas, such as the extraction of mariculture areas in some key bays or in a single province. Xue et al. (2018) and Wu et al. (2021) obtained distribution data of the multiphase mariculture in Sansha Bay, Fujian province, and they explored its spatial distribution changes. Some researchers obtained the distribution data for mariculture in a single period in Sansha Bay (Liu et al., 2019b; Lu et al., 2021; Ma et al., 2022). Wang et al. (2018) extracted raft mariculture data in Luoyuan Bay, Fujian, based on Gaofen-2 (GF-2) images. Hu et al. (2017) and Zhang et al. (2020) used Radarsat-2 images and Sentinel-1 images, respectively, to extract raft mariculture data in Changhai county, Liaoning province. Some researchers extracted raft mariculture data in Dalian, Liaoning, based on Gaofen-1 (GF-1) images and synthetic aperture radar (SAR) images, respectively (Fan et al., 2015; Fan et al., 2018; 2019; Shi et al., 2018, 2019). Some researchers obtained the distribution data of the multiphase mariculture in Shandong province based on Landsat images (Wang et al., 2019b; Wang et al., 2022). Kang et al. (2019) extracted multiphase mariculture data for Liaoning based on Landsat images. Wang et al. (2019) extracted cage mariculture data on Goqi Island in Zhejiang province based on GF-2 images, with an accuracy of 93.0%. Zheng et al. (2017) extracted raft mariculture data in Dayu Bay in Zhejiang based on GF-2 images. Cui et al. (2019) extracted raft mariculture data in Lianyungang, Jiangsu province, with an accuracy of 89.5%. Based on Radarsat-2 images, Geng et al. (2017) extracted raft mariculture data near Beidaihe, Hebei, with an accuracy of 94.9%. Stiller et al. (2019) extracted the multiphase aquaculture ponds in the Yellow River Delta and the Pearl River Delta in China based on long-time-series Landsat images and Sentinel-1 images. In addition, several studies have been dedicated to exploring the extraction methods of mariculture data, but the experimental areas in these studies were small (Wang et al., 2017a; Fu et al., 2019a, 2019b; Sui et al., 2020; Wang et al., 2022b).
Compared with small-scale studies, few studies have been conducted on the spatial distribution of mariculture at the national scale. In a previous study, we extracted the mariculture areas in China in 2018 based on Landsat images, and the interpretation accuracy reached 87.35% (Liu et al., 2020). Fu et al. (2021) extracted the mariculture areas in China in 2018 based on high-resolution and wide-range images, and the interpretation accuracy reached 95.83%. Based on Sentinel-2 images, Xu et al. (2021) used random forests and support vector machines to extract the cage and raft mariculture areas in China’s offshore waters, with an accuracy of 90.4%. Research by Liu et al. (2022) provided a new method for obtaining high-precision mariculture data. By combining time-series optical images and SAR images, China’s mariculture distribution data for 2020 were obtained at a higher accuracy than data extracted in previous studies.
Research on mariculture extraction methods is quickly expanding, and based on this research, distribution data for mariculture in China in recent years have been obtained. However, compared to numerous long-term and large-scale coastal remote sensing monitoring studies (Ding et al., 2019; Mou et al., 2021; Li et al., 2022; Xu et al., 2023), research related to marine aquaculture is still relatively scarce, primarily due to the following two reasons:
(1) The resolution of remote sensing images acquired in the earlier period was low. Because of the relatively small size of a single mariculture target, most previous studies used remote sensing images with a spatial resolution that was better than 15 m, and the earliest availability of such images generally was not earlier than 10 years ago. Because of the relatively low resolution of images acquired in earlier periods, the expression effect of mariculture targets was greatly reduced, and many automatic extraction algorithms could not achieve satisfactory results.
(2) Mariculture is dynamic. Mariculture does not exist in the sea for a long time, and it follows a specific aquaculture cycle according to different regions and species. Therefore, compared with land objects, mariculture data exhibit highly dynamic characteristics. It could take a few months in a year to observe mariculture on remote sensing images. In addition to the effect of the interference of clouds, rain, and other weather phenomena on remote sensing, it is difficult to ensure that appropriate images of mariculture can be selected nationwide.
Given that the significant volume of mariculture in China can have a profound impact on the world seafood and fishery market, it is necessary to explore its development and changes. In this study, we designed an experimental scheme to address these problems. First, we used temporal Landsat optical images and temporal Sentinel-1 SAR images to extract China’s mariculture areas in 2020. Then, using these data and in the form of time tracking, as well as multiple Landsat optical images acquired in 2010 and 2000, we used the visual interpretation method to extract the mariculture as accurately as possible. On this basis, combined with shoreline and water depth data, we explored the spatial and temporal distribution patterns and development trends of mariculture in China.

2 Study area

China is located in the eastern part of the Eurasian continent, facing the Pacific Ocean to the east and adjoining the Indian Ocean to the south. China has a vast coastline, with the mainland coastline stretching over 18,000 kilometers, island coastlines extending over 14,000 kilometers, and a total coastline length exceeding 32,000 kilometers. The coastal zone is home to over 6500 islands (Zhang, 1998). The coastal zone of China spans 22 latitude bands.
Over 40% of China's population and 60% of its GDP are concentrated in the coastal zone (Hou and Xu, 2011). The abundant natural resources, along with the favorable geographical and human environments, have laid a solid foundation for the thriving development of mariculture in China. Mariculture has contributed a substantial economic output and numerous employment opportunities to China’s coastal regions.
Based on preliminary visual interpretation, we found that mariculture in China is generally distributed within the waters within 50 kilometers from the mainland coastline. To prevent the omission of mariculture areas, we constructed a 100-kilometer buffer zone extending from the mainland coastline into the sea, using this region as the study area for our research. From north to south, the Chinese provincial-level regions (hereafter province) covered by the study area include Liaoning, Hebei, Tianjin, Shandong, Jiangsu, Shanghai, Zhejiang, Fujian, Guangdong, Guangxi, and Hainan (because of the lack of comparative statistical data, Hong Kong, Macao, and Taiwan were not included).

3 Data and methodology

3.1 Data preparation

We selected the Landsat series satellite images and Sentinel-1 satellite images as the data sources for this study.
Landsat series satellites have been observing the Earth since 1972 without interruption and with good data continuity. In addition, the spatial resolution of these satellites can ensure the extraction of mariculture areas. Therefore, we selected Landsat series satellite images to extract the mariculture areas.
The Sentinel-1 satellite is a SAR satellite, which makes up for the disadvantage that the Landsat optical satellite is affected by cloud cover. When extracting the mariculture areas in coastal areas with complex climatic conditions, a combination of optical images and SAR images obtained higher accuracy data than a single-image data source. The Sentinel-1 satellite was launched in 2014. In this study, we used Sentinel-1 images to extract the mariculture area in 2020 only.

3.1.1 Landsat series satellite multitemporal images

Landsat series satellite images mainly included multiple visible light bands with a 30 m spatial resolution, but since Landsat-7, panchromatic bands with a 15 m spatial resolution have increased. The track height is 705 km, the inclination is 98.22°, the revisit period is 16 days, and the image width is 185 km (Wulder et al., 2019).
Because the coastal zone is often cloudy and rainy, the available optical images were reduced by the influence of clouds. We assumed that the mariculture area would not change much within the time range of 1 year. Therefore, for the mariculture areas in 2000, 2010, and 2020, we used all of the remote sensing images with a cloud cover of less than 30% in 1999-2001, 2009-2011, and 2019-2021 for the extraction.
Because the Landsat-7 satellite was launched in 1999, we used the Landsat-7 satellite images to extract the mariculture areas from 1999 to 2001. On May 31, 2003, the Landsat-7 enhanced thematic mapper plus (ETM+) airborne scanning line corrector (SLC) failed, which resulted in the loss of data bands in the images obtained since then. Therefore, we used the Landsat-5 satellite images to extract the mariculture area from 2009 to 2011. The Landsat-8 satellite was launched in 2013, and the Landsat-8 satellite images were used to extract the mariculture areas from 2019 to 2021.

3.1.2 Sentinel-1 satellite multitemporal images

Sentinel-1 (including Sentinel-1A and Sentinel-1B) is a C-band SAR with a frequency of 5.405 GHz, which supports single polarization (vertical-vertical (VV) and horizontal-horizontal (HH)) and multipolarization (VV + VH and HH + HV). Its single-star revisit cycle is 12 days (the double-star revisit cycle is 6 days). The image obtained in the default interference wide working mode (IW) has a width of 250 km and a spatial resolution of 5 × 20 m (Torres et al., 2012). Because of the weak penetration of the cross-polarization VH data relative to the single polarization HH data, it was difficult to observe mariculture areas in the VH data (Zhang et al., 2020). Therefore, we used only the single-polarization VV data for the extraction of mariculture areas.
On the Google Earth Engine (GEE) platform, we selected all of the available VV Sentinel-1 images of the study area from February 4, 2020, to February 4, 2021, in IW mode with ground range-detected (GRD) format. The Sentinel-1 images in the GEE platform were preprocessed using the European Space Agency’s (ESA) Sentinel-1 toolbox, including thermal noise removal, radial calibration, and terrain correction.

3.1.3 Coastline data

Because mariculture is located in the seawater near the coastline, accurate shoreline data help remove the interference of land features and assist in the interpretation of mariculture areas. We used the 2015 global sea-land (and island) coastline data with a meter-level spatial resolution (Liu et al., 2019a). The dataset covers the whole world. Based on the detailed definition of shoreline production standards in different situations, the dataset was generated using human-computer interaction methods in combination with Google Earth images, DEM (Digital Elevation Model) data, multisource maps, and publications of different scales. After strict inspection, the data reliability was high. Because the coastal area of China covered by the study area had a large spatial scale and various land cover types, using this dataset to mask the land area improved the interpretation efficiency and reduced the misclassification of mariculture areas.

3.1.4 GEBCO data

To further explore the spatiotemporal distribution patterns of mariculture in China, we introduced water depth data for overlay analysis. The water depth data used is from the General Bathymetric Chart of the Oceans (GEBCO) gridded bathymetric dataset (Gebco Bathymetric Compilation Group, 2021). GEBCO has a spatial resolution of 15 arc-seconds, covering global marine areas. However, there is a small amount of missing data in coastal regions. To address this, we used Inverse Distance Weighting (IDW) interpolation to process the data (Lu and Wong, 2008).

3.2 Methods

3.2.1 Extraction of mariculture areas in 2020

We extracted the mariculture area in 2020 using the temporal Landsat-8 operational land imager images from 2019 to 2021 and the Sentinel-1 SAR images from 2020. The flow chart is shown in Figure 1. We used the method of Liu et al. (2022) in the extraction of the mariculture areas in 2020, which included the following five steps:
Figure 1 Workflow of the three periods of mariculture mapping
(1) Generate edge probability map
Compared with the fluctuation of the spectral features of mariculture with different growth stages, its shape and edge features were relatively more stable, and the edge information of mariculture could be effectively enhanced by generating edge probability maps.
This was helpful for the differentiation of mariculture and nonmaricultured information.
(2) Feature map segmentation
Segmentation of edge probability feature maps used an enhanced watershed algorithm.
(3) Mariculture area extraction based on features of segmented objects
Based on the segmented objects, we calculated the average gray value, area, and texture features of each object separately for the preliminary extraction of the mariculture area.
(4) Generation of SAR temporal mean images
Sentinel-1 imagery compensated for the omission of references to mariculture in optical imagery because of weather factors, such as clouds and rain, whereas time-series synthesis effectively removed the interference of transient information, such as vessels, and enhanced mariculture to increase its contrast with the seawater background.
(5) Morphological processing and artificial review
We used morphological closure operations to fill the void in the mariculture zones and to close the outer boundaries of the zones. Finally, with the assistance of time-synthesized SAR images, we extracted the missed mariculture zones by visual interpretation to obtain the final extraction results.

3.2.2 Extraction of mariculture areas in 2010 and 2000

The spatial resolution and pixel depth of the remote sensing images from 2000 and 2010 were inferior to those of the 2020 images, making it challenging to accurately obtain mariculture areas using automated methods. Therefore, we adopted a retrospective approach, using the previously extracted mariculture areas in 2020 as auxiliary data. We visually interpreted the mariculture areas in 2010 with the positions of the 2020 mariculture areas as crucial reference. Subsequently, using the same method, we visually interpreted the mariculture areas in 2000 with the 2010 mariculture areas as auxiliary data.

3.2.3 Accuracy evaluation method

We evaluated the precision of the mariculture data for 2020 using the random sample point method. We constructed a 10 km buffer zone based on the extracted mariculture data and randomly generated 1000 sample points within this range. We recorded the interpretation type of the sample points according to the extraction results of the mariculture area. Then, we interpreted the actual types of the sample points using high-resolution time-series images, such as Google Earth images and Sentinel-2 images. Because of the dynamic nature of mariculture, we used time-series images for the whole year in this process. As long as mariculture had occurred in the area where the sample point was located throughout the year, we recorded the type of the sample as a mariculture area; otherwise, we recorded it as a nonmariculture area. Finally, we used the F1 value to calculate the interpretation accuracy of the mariculture area in 2020.
The F1 value is a combination of precision and recall. Precision represents the proportion of correct identification of the extracted mariculture area. Recall represents the proportion of the actual mariculture area correctly extracted. In practice, precision and recall are often in conflict (Wang et al., 2017a); therefore, the F1 value is commonly used to comprehensively measure these two indicators. The larger the F1 value is, the better the accuracy of the extraction result is (Table 1).
$F1=2*\frac{recall*precision}{recall+precision}$,
$recall=\frac{TP}{TP+FN}$,
$precision=\frac{TP}{TP+FP}$,
where TP is the number of sample points when the type of mariculture area is correctly identified, FP is the number of sample points incorrectly identified as the mariculture area, and FN is the number of sample points in which the actual mariculture area was not identified.
Table 1 Obfuscation matrix
Interpretation result
Mariculture area Nonmariculture area
Truth Mariculture area TP FN
Nonmariculture area FP TN
For the accuracy verification of the 2010 and 2000 mariculture area data, we calculated the F1 value based on the random sample point type. Because we used the highest resolution images available to interpret the data for the mariculture areas in 2010 and 2000, it was not appropriate to use the same method used to assess the data for 2020 to distinguish the actual types of the sample points in these two periods. As an alternative, we invited five remote sensing image interpretation experts to identify the actual types of random sample points using the time-series Landsat image data used in the interpretation.

4 Results

4.1 Distribution and development of mariculture in China

Figures 2, 3, and 4 show the distribution of the mariculture zones in China in 2000, 2010, and 2020, respectively. Figure 5 shows the details of the distribution of mariculture in the three periods. The F1 values of the mariculture data in 2000, 2010, and 2020 were 96.2%, 94.3%, and 92.7%, respectively.
Figure 2 Distribution of mariculture in China in 2000
Figure 3 Distribution of mariculture in China in 2010
Figure 4 Distribution of mariculture in China in 2020
Figure 5 Details of mariculture in China in 2000, 2010, and 2020
Taking the coastal provinces of China as a unit, we counted the mariculture area in the three periods. The results are presented in Table 2.
Table 2 Distribution and area proportion of mariculture in different provinces of China in 2000, 2010, and 2020
Region Area (km2) Proportion (%)
2000 2010 2020 2000 2010 2020
China 1775.88 4480.69 12,736.98 / / /
Liaoning 214.49 833.04 1535.69 12.1 18.6 12.1
Hebei 0 0 1191.40 0.0 0.0 9.4
Tianjin 0 0 0 0.0 0.0 0.0
Shandong 556.24 675.14 4841.42 31.3 15.1 38.0
Jiangsu 244.36 1010.46 1448.44 13.8 22.6 11.4
Shanghai 0 0 0 0.0 0.0 0.0
Zhejiang 61.63 124.54 261.42 3.5 2.8 2.1
Fujian 653.82 1609.88 1738.06 36.8 35.9 13.6
Guangdong 17.21 172.35 1261.15 1.0 3.8 9.9
Guangxi 2.40 29.63 356.59 0.1 0.7 2.8
Hainan 5.72 5.55 82.58 0.3 0.1 0.6
In 2000, China’s mariculture area was 1775.88 km²: Fujian accounted for 36.8%, Shandong accounted for 31.3%, Jiangsu accounted for 13.8%, and Liaoning accounted for 12.1%. These four provinces accounted for 94% of the mariculture area in 2000. The rest of the mariculture area was distributed in Zhejiang (3.5%), Guangdong (1.0%), Hainan (0.3%), and Guangxi (0.1%). We did not identify any mariculture activities in the sea areas of Hebei, Tianjin, and Shanghai.
In 2010, China’s mariculture area was 4480.69 km². Fujian still had the largest mariculture area, accounting for 35.9%. Jiangsu surpassed Shandong to become the second largest province in terms of mariculture area in 2010, accounting for 22.6%, followed by Liaoning (18.6%), which also surpassed Shandong. Shandong ranked fourth, accounting for 15.1%. The mariculture area in Guangdong developed rapidly, but this proportion remained relatively small, accounting for 3.8%. We did not identify any mariculture activities in Hebei, Tianjin, and Shanghai, which was the same as in 2000.
In 2020, China’s mariculture area was 12,736.98 km², and Shandong’s mariculture area grew rapidly, becoming the largest province in terms of mariculture. Shandong accounted for 38.0%, followed by Fujian (13.6%), Liaoning (12.1%), and Jiangsu (11.4%). The mariculture area increased significantly in Guangdong and Hebei, accounting for 9.9% and 9.4%, respectively. This was the first time that mariculture appeared in the Hebei sea area. We still did not identify any mariculture activity in Tianjin or Shanghai.
The distribution of the mariculture area in the three periods is shown in Figure 6.
Figure 6 Distribution of the mariculture area in China in 2000, 2010, and 2020
Taking Shanghai as the dividing line between the northern and southern regions of China’s mariculture, we found that the proportions of mariculture in the north and south remained the same between 2000 and 2010, and the proportion of mariculture in the north increased significantly from 2010 to 2020.
In 2000, the mariculture area in the north was slightly higher than that in the south. The mariculture area in northern China (Liaoning, Shandong, and Jiangsu) accounted for 57.2%, and that in southern China (Zhejiang, Fujian, Guangdong, Guangxi, and Hainan) accounted for 42.8%.
In 2010, the mariculture area in the north accounted for 56.3%, and that in the south accounted for 43.7%. Furthermore, we did not observe any significant change compared with 2000. During this period, however, the proportion of the mariculture areas in the northern provinces changed significantly. The mariculture industry in Jiangsu and Liaoning developed rapidly during this period, surpassing that of Shandong.
In 2020, the mariculture area in the north accounted for 70.9%, and that in the south accounted for 29.1%. The mariculture industry in the north, represented by Shandong, developed rapidly. China’s mariculture finally formed a distribution pattern of more in the north and less in the south.

4.2 Analysis of the rate of change of the mariculture area

After the preliminary spatial distribution analysis of the three-phase data for China’s mariculture, we explored the development of and changes in mariculture in the various provinces using the land use dynamic index.
The land use dynamic index can quantitatively reflect the amplitude and rate of change of a certain feature type within a certain time range in a region. The formula is as follows:
$\text{S}=\frac{{{A}_{b}}-{{A}_{a}}}{{{A}_{a}}}\times \frac{1}{T}\times 100%$,
where S is the rate of change of the mariculture area, Aa and Ab are the mariculture area at the start and end time, respectively, and T is the period.
The changes in the mariculture area in the different regions of China from 2000 to 2010 and from 2010 to 2020 are shown in Table 3.
Table 3 Changes in the mariculture area in each region of China during the two periods
Region Rate of change (%)
2000-2010 2010-2020
China 152.3 184.3
Northern China 148.1 258.0
Southern China 162.1 90.5
Liaoning 288.4 84.3
Hebei / /
Tianjin / /
Shandong 21.4 617.1
Jiangsu 313.5 43.3
Shanghai / /
Zhejiang 102.1 109.9
Fujian 146.2 8.0
Guangdong 901.6 631.7
Guangxi 1134.2 1103.6
Hainan −3.0 1388.2
From 2000 to 2010, the rate of change of China’s mariculture area reached 152.3%. The fastest growth occurred in Guangxi, with a rate of change of 1134.2%, followed by Guangdong (901.6%). The rates of change in Jiangsu and Liaoning reached 313.5% and 288.4%, respectively. In 2000, the rate of change in Fujian, which had the largest aquaculture area, was slower than in these four provinces, but it also doubled, with a rate of change of 146.2%. Among the provinces in China, the increase in Shandong was the smallest during this period, with a rate of change of 21.4%. Hainan was the only province with negative growth, with a rate of change of −3.0%. During this period, the rate of change of the mariculture area in northern China was 148.1%, which was slightly lower than that in southern China (162.1%).
From 2010 to 2020, the rate of change of China’s mariculture area reached 184.3%, which was higher than that in the previous period. Hainan had the largest rate of change and was the only province with a negative area change in the previous period. The rate of change in Guangxi exceeded 1000%, reaching 1103.6%. The rate of change in Guangdong decreased compared with the previous period, but it still reached 631.7%. During this period, the mariculture area in Shandong increased rapidly, making Shandong the largest province in terms of the mariculture area in 2020, with a rate of change of 617.1%. The mariculture area in Zhejiang continued to grow, with a rate of change of 109.9%, which was the same as that in the previous period. The rates of change in Liaoning and Jiangsu slowed down, with rates of change of 84.3% and 43.3%, respectively. Fujian had the lowest rate of change, at only 8%, but the overall mariculture area was huge, and it was still the second largest province in terms of mariculture area in 2020. During this period, the mariculture area in northern China increased rapidly, reaching 258.0%, whereas that in southern China slowed down, dropping to 90.5%. Because of the development difference between the north and south during this period, the current distribution pattern of China’s mariculture (i.e., more in the north and less in the south) was established.

5 Discussion

5.1 Distance between mariculture and shoreline

Based on the shoreline data and mariculture data, we explored the characteristics of the offshore distance of mariculture areas. Using ArcGIS software, we set a buffer zone based on the shoreline data. The buffer zone was 1 km apart and covered a total of 50 km from the shoreline to the sea area. Because the distribution of some of the mariculture zones may have been closer to some of the islands than to the land, in addition to the continental coastline, we also added the contour lines of the islands in the construction of the buffer zone to obtain a more accurate distribution of the offshore distance of the mariculture zone.
We spatially superposed the mariculture data and the buffer zone data and counted the area of the mariculture zone within the buffer zone interval of 1 km. The statistical results are presented in Figure 7.
Figure 7 Distribution statistics of the mariculture area and shoreline distance in China in 2000, 2010, and 2020
In 2000, 90.1% of the mariculture area was distributed within 4 km of the coastline, and the maximum offshore distance of the mariculture area was 29 km.
In 2010, the proportion of the mariculture area within 4 km of the coastline decreased to 74.6%, and 90.1% of the mariculture area was distributed within 13 km of the coastline. The maximum offshore distance of the mariculture area increased to 44 km.
In 2020, the proportion of the mariculture area within 4 km of the shoreline decreased to 54.0%, the proportion of mariculture area within 13 km of the shoreline decreased to 85.1%, and 90.8% of the mariculture area was distributed within 17 km of the shoreline. The maximum offshore distance further increased to 48 km.
The mariculture area exhibited an obvious L-type distribution on the statistical map of the offshore distance. The mariculture area was larger closer to the shoreline. With increasing distance from the shoreline, the proportion of the mariculture area gradually decreased. According to the statistics, more than 90% of the mariculture area in China was located within 20 km of the coastline.
Over time, the mariculture area exhibited a trend of gradually moving away from the coastline and expanding toward the deep sea. The proportion of the inshore mariculture area gradually decreased, the center of the mariculture area shifted toward the far offshore line, and the farthest mariculture distance increased.

5.2 Relationship between mariculture and water depth

The GEBCO water depth data for China’s coastal waters were reclassified according to depth zones of 0-10 m, 10-20 m, and more than 20 m. The overlay situation of the third phase mariculture zone with water depth is shown in Figure 8.
Figure 8 Distribution of mariculture at different water depths in China
The depth of the coastal seawater in China is shallow in the north and deep in the south. The water depth in Bohai Bay in the north and the coastal areas of Jiangsu and Shanghai in the middle are relatively shallow, and the shallow water area outside the coastline is relatively wide. The coastal shallow water area in the south is concentrated in the bays in Guangdong and Guangxi, and the shallow water area is relatively narrow compared with that in the northern coastal area.
We spatially superimposed data for the mariculture area in the three periods with the water depth data to analyze the distribution of the mariculture area in different seawater depth zones. The statistical results are shown in Figure 9.
Figure 9 Distribution statistics of the mariculture area in the nearshore water depth areas in China in 2000, 2010, and 2020
In 2000, 54.4% of the mariculture area was located in areas with water depths of less than 10 m, 30.3% of the mariculture area was located in areas with depths of 10-20 m, and 15.3% of the mariculture area was located in areas with depths of greater than 20 m. During this period, the deepest seawater depth within the distribution range of the mariculture area was 129 m.
In 2010, 59.8% of the mariculture area was located in areas with water depths of less than 10 m, 26.4% of the mariculture area was located in areas with depths of 10-20 m, and 13.0% of the mariculture area was located in areas with depths of greater than 20 m. During this period, the deepest seawater depth within the distribution range of the mariculture area was 161 m.
In 2020, 50.4% of the mariculture area was located in areas with water depths of less than 10 m, 39.2% of the mariculture area was located in areas with depths of 10-20 m, and 10.8% of the mariculture area was located in areas with depths of greater than 20 m. During this period, the deepest seawater depth within the distribution range of the mariculture area was 181 m.
Similar to the distribution of mariculture in terms of the offshore distance, the mariculture zone also exhibited an L shape on the statistical chart of the seawater depth. We observed a large number of mariculture areas in the shallow water zone, and the mariculture area gradually decreased with increasing depth. Unlike the offshore distance distribution, however, the development of mariculture in the depth zones exhibited the characteristics of shallow first and deep later. From 2000 to 2010, the proportion of mariculture within a depth zone of less than 10 m increased, and the area of mariculture in a depth zone of greater than 20 m decreased. From 2010 to 2020, the proportion of mariculture within a depth zone of less than 10 m decreased, and the proportion of mariculture within a depth zone of between 10 and 20 m increased. Although China’s mariculture exhibited a trend of expanding toward the deep water area, more than 50% of the mariculture area was still distributed in the sea area within a depth zone of less than 10 m.

5.3 Analysis of the development pattern of mariculture in China

According to these results, the distribution of mariculture in China exhibited the development characteristics of moving from near to far and from shallow first to deeper later. This was likely to be caused by the following four factors:
(1) Development of mariculture technology
With the progress of science and technology and the development of the social economy, mariculture technology has also constantly developed. In recent years, with the completion of Haigeng-1, Shenlan-1, and other large-scale mariculture platforms in the deep sea, the depth of the water area that can be reached by mariculture continues to deepen, and the seawater depth is strongly correlated with the offshore distance. Therefore, the distribution of the mariculture area is expanding outward with the development of technology over time.
(2) Inshore mariculture water area gradually reaches saturation
With the continuous development of mariculture in China, the area and density of mariculture have increased rapidly. The space in the nearshore area is precious, however, and natural reserves, tourist resorts, ports and waterways, and industrial zones restrict mariculture. The nearshore area where mariculture can develop freely may have gradually reached saturation, which has prompted the mariculture area to expand toward the deep sea over time.
Although mariculture floats on the sea surface, it needs to be fixed to the sea bottom to prevent it from being washed away by the ocean current. Therefore, certain requirements must be met regarding the depth. Under the conditions suitable for the development of mariculture, the closer to the shore an area is, the shallower the water depth is, and the lower the cost of managing and maintaining the mariculture facilities is. Therefore, in the early stage of development, mariculture was concentrated primarily in shallow coastal areas.
Corresponding to the period from 2000 to 2010, given that the amount of mariculture in 2000 was relatively small and a large amount of space was still available for mariculture development in the coastal shallow water area, during this period, the mariculture area mainly expanded in the shallow water area with water depths of less than 10 m.
In the later stage of development, from 2010 to 2020, because of the increase in mariculture, the water area suitable for mariculture in the shallow water area gradually became saturated. At this time, the development of mariculture technology ensured that mariculture could expand toward deeper areas. Therefore, during this period, the proportion of mariculture in the depth range of less than 10 m decreased, and the proportion of mariculture in the depth range of 10-20 m increased.
(3) Policy restrictions and promotions
Policies have an important impact on the development of and changes in mariculture in China. Driven by policies, since the 1970s, China’s mariculture has experienced five waves of algae farming, prawn farming, shellfish farming, fish farming, and sea cucumber and abalone farming. The overall policy guidelines have been adjusted constantly in different development stages according to the resource utilization status and technological production level. From the early pursuit of productivity and output, the full use of shallow tidal flats to vigorously develop mariculture gradually shifted to the pursuit of development quality and to developing mariculture according to local conditions. Currently, with the comprehensive deepening of sustainable development thinking in policy formulation, the current mariculture-related policies pay more attention to the protection of the environment. The trend is adjusting and compressing near-shore farming and developing marine ranching and deep-sea farming.
For example, in 2013, the State Council’s Several Opinions on Promoting the Sustainable and Healthy Development of Marine Fisheries was proposed to control the density of offshore mariculture and to actively expand offshore mariculture and intensive mariculture. In 2015, the State Council’s National Marine Main Functional Zone Planning was proposed to scientifically develop mariculture, promote ecological and healthy mariculture models, encourage qualified enterprises to expand offshore mariculture and intensive mariculture, and support the development of pelagic fisheries. In 2017, the State Office of the Central Committee of the Communist Party of China issued Several Opinions on Establishing a Long-Term Mechanism for Monitoring and Early Warning of Resource and Environmental Carrying Capacity, which noted that “in the case of critical overloading of fishery resources, the conservation of marine fishery resources and habitat protection shall be strengthened and the transfer of inshore mariculture areas to offshore deep water areas shall be guided.” In 2019, Several Opinions on Accelerating the Green Development of the Aquaculture Industry clarified that in the future, China will actively support the development of green mariculture in deep- and open-sea areas and will encourage the construction of large-scale intelligent mariculture fisheries in deep- and open-sea areas. In 2022, the Ministry of Ecology and Environment’s Opinions on Strengthening the Supervision of the Ecological Environment of Mariculture required that “Coastal fishery departments at all levels, in conjunction with relevant departments, earnestly implement the planning of tidal flats for aquaculture waters at the same level, and strictly control the use of aquaculture waters and tidal flats in accordance with the plan for delineating the three areas (prohibited aquaculture areas, restricted aquaculture areas, and aquaculture areas). They should further optimize the spatial layout of mariculture, prohibit mariculture activities in prohibited areas according to law, strengthen pollution prevention and control in aquaculture areas and restricted aquaculture areas, and strengthen the protection of key aquaculture bases and important mariculture sea areas”.
The release and implementation of this series of national policies was an important factor in promoting the development of mariculture toward offshore deep-sea areas.
(4) Impact of economic development
Economic development may be another factor affecting the development of mariculture toward deep-sea areas. This is reflected in two aspects: the economic development related to mariculture and the overall economic development of society.
In terms of economic development related to mariculture, mariculture in China has developed rapidly in the past few decades, and its output value has continued to increase. In 2005, when the output value of mariculture was calculated separately for the first time, it was 9.42 billion yuan, and by 2020, it had reached 38.36 billion yuan. In recent years, however, the growth rate of the mariculture output value has slowed down significantly. The main reason for this reduction is the continuous expansion of the nearshore culture scale and the solidification of culture models, the lack of reasonable planning and layout of the current culture species, and the serious homogeneity of the culture products, which have resulted in excessive competition for products and have decreased the economic value of the products. Because of its unique natural conditions, the deep sea can expand the cultivation of high-value fish, such as large yellow croaker and tuna. Moreover, the end products of deep-sea mariculture are more natural and wild, the same output value can be increased by one or two times, and the market prospect for the development of deep-sea mariculture can broaden. Therefore, many large-scale mariculture companies have begun to gradually expand to deep-sea mariculture.
The development of the overall social economy and the advancement of China’s industrialization and urbanization process have led to a rapid increase in the demand of industrial land and urban construction land in coastal areas, and land reclamation projects have been very active in coastal provinces. Many high-quality coastal waters suitable for mariculture have been filled to create land, the mainland coastlines in many areas have been deformed or reduced, and the space for offshore mariculture has been severely decreased. To a certain extent, this urbanization has forced mariculture to develop in deep- and open-sea areas.
On this basis of this analysis, the development of mariculture toward the deep sea is the inevitable result of the joint influence of many factors. The rapid development of near-shore mariculture and the development of coastal zones against the backdrop of national economic development have gradually saturated the near-shore mariculture space. The development of science and technology has provided objective conditions for deep-sea mariculture, and policy restrictions and guidance have accelerated the development of mariculture toward the deep sea. These various conditions have prompted the development of mariculture toward the deep sea.

5.4 Analysis of the cause of the difference in the north-south distribution of mariculture in China

In 2020, the distribution of mariculture in China exhibited obvious differences, with a distribution pattern of more in the north and less in the south, which was not formed in the early stage. From 2000 to 2010, the distribution of mariculture in the northern and southern parts of China remained relatively balanced. From 2010 to 2020, however, this balance was broken. The mariculture areas in the north expanded rapidly, and the growth rate was significantly higher than that in the south. By 2020, a distribution pattern of more in the north and less in the south had formed.
This pattern may have been caused primarily by the difference in the water depths in the offshore areas in northern and southern China. According to the information presented thus far, more than 80% of the mariculture area in China was located in areas with water depths of less than 20 m, and mariculture exhibited a development trend of shallow first and deeper later, which indicated that the water depth exerted a relatively obvious restriction on the layout of the mariculture facilities. Figure 8 shows that the offshore water depth in northern China was significantly lower than that in southern China, which also indicated that northern China had more areas suitable for mariculture than southern China. In 2000 and 2010, the scale of mariculture was relatively small, and although the shallow sea area in the south was limited, it still met the development needs at that time. Therefore, the distribution difference between the northern and southern parts of China in this period had not yet developed. By 2020, with the rapid growth of mariculture, the north had more advantages in terms of water depth than the south, which contributed to the current distribution pattern.

6 Conclusions

Based on Landsat time-series remote sensing images and Sentinel-1 remote sensing images, we extracted data for the mariculture areas in 2000, 2010, and 2020. Based on the extracted data, we analyzed the distribution pattern and development of mariculture in China for the first time.
From 2000 to 2020, China’s mariculture exhibited an accelerated growth trend. In 2000, the mariculture area was 1775.88 km²; in 2010, the mariculture area was 4480.69 km²; and in 2020, the mariculture area was 12,736.98 km².
In China, more than 90% of the mariculture area was distributed in the sea area within 20 km of the shoreline, and more than 80% of the mariculture area was distributed in the sea area with water depths of less than 20 m. The development of mariculture in China exhibited a pattern of moving from near to far and from shallow first to deeper later. The rapid development of near-shore mariculture and the development of coastal zones against the backdrop of national economic development gradually saturated the near-shore mariculture space. The development of science and technology has provided objective conditions for deep-sea mariculture, and policy restrictions and guidance have accelerated the development of mariculture toward the deep sea.
Currently, mariculture in China has followed a distribution pattern of more in the north and less in the south. With the Yangtze River estuary as the boundary between the north and the south, the mariculture area in the north was significantly larger than that in the south. The reason for this distribution pattern likely was related to the water depth factor. The water depth in the coastal areas was significantly shallower in northern China than in southern China, providing more development space for mariculture.
In this study, we first extracted the spatial distribution data for mariculture in the three periods in China, and then we explored the distribution pattern and changes in mariculture in China. The following deficiencies still existed, however. In this study, we analyzed the north-south distribution difference in mariculture in China only from the perspective of the water depth, but the actual mariculture distribution was related to many other conditions, including the temperature, salinity, chlorophyll concentration, seawater flow rate, protection zone restrictions, and land conditions. These factors will be considered in a subsequent study. In addition, according to the statistical yearbook, the area of mariculture in China reached its peak in 2015 and has exhibited a slight decline each year since then. The time scale of 10 years in this study failed to reflect this trend, however, and this time scale will be further refined in the future.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Cui B G, Zhong Y, Fei D et al., 2019. Floating raft aquaculture area automatic extraction based on fully convolutional network. Journal of Coastal Research, 90(Suppl.): 86-94.

DOI

[2]
Ding X S, Shan X J, Chen Y L et al., 2019. Dynamics of shoreline and land reclamation from 1985 to 2015 in the Bohai Sea, China. Journal of Geographical Sciences, 29(12): 2031-2046.

DOI

[3]
Fan J, Chu J, Geng J et al., 2015. Floating raft aquaculture information automatic extraction based on high resolution SAR images. In: 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) (pp. 3898-3901). IEEE.

[4]
Fan J, Wang X, Wang X et al., 2019. GF-3 PolSAR marine aquaculture recognition based on complex convolutional neural networks. In: 2019 Tenth International Conference on Intelligent Control and Information Processing (ICICIP) (pp. 112-115). IEEE.

[5]
Fan J, Zhao J, Song D et al., 2018. Marine floating raft aquaculture dynamic monitoring based on multi-source GF imagery. In: 2018 7th International Conference on Agro-geoinformatics (Agro-geoinformatics), 1-4.

[6]
FAO, 2018. Fisheries and Aquaculture topics. The State of World Fisheries and Aquaculture (SOFIA). In: FAO Fisheries and Aquaculture Department [online].

[7]
FAO, 2022. Fisheries and Aquaculture topics. The State of World Fisheries and Aquaculture (SOFIA). In: FAO Fisheries and Aquaculture Department [online].

[8]
Fu Y Y, Deng J X, Wang H Q et al., 2021. A new satellite-derived dataset for marine aquaculture areas in China’s coastal region, Earth System Science Data, 13(5): 1829-1842.

DOI

[9]
Fu Y Y, Deng J X, Ye Z R et al., 2019a. Coastal aquaculture mapping from very high spatial resolution imagery by combining object-based neighbor features. Sustainability, 11(3): 637.

DOI

[10]
Fu Y Y, Ye Z R, Deng J X et al., 2019b. Finer resolution mapping of marine aquaculture areas using worldView-2 imagery and a hierarchical cascade convolutional neural network. Remote Sensing, 11(14): 1678.

DOI

[11]
Gebco Bathymetric Compilation Group, 2021. The GEBCO_2021 Grid: A Continuous Terrain Model of the Global Oceans and Land. In, edited by Gebco Bathymetric Compilation Group 2021. NERC EDS British Oceanographic Data Centre NOC.

[12]
Geng J, Fan J, Wang H, 2017. Weighted fusion-based representation classifiers for marine floating raft detection of SAR images. IEEE Geoscience and Remote Sensing Letters, 14(3): 444-448.

DOI

[13]
Hou X Y, Xu X L, 2011. Spatial patterns of land use in coastal zones of China in the early 21st century. Geographical Research, 30(8): 1370-1379. (in Chinese)

DOI

[14]
Hu Y Y, Fan J C, Wang J, 2017. Target recognition of floating raft aquaculture in SAR image based on statistical region merging. In: 2017 Seventh International Conference on information Science and Technology (ICIST) (pp. 429-432). IEEE.

[15]
Kang J M, Sui L C, Yang X M et al., 2019. Sea surface-visible aquaculture spatial-temporal distribution remote sensing: A case study in Liaoning province, China from 2000 to 2018. Sustainability, 11(24): 7186.

DOI

[16]
Li X, Liu Z H, Wang S Q et al., 2022. Spatial characteristics of the stability of mangrove ecosystems in freshwater and seawater floods in Southeast Asia. Journal of Geographical Sciences, 32(9): 1831-1846.

DOI

[17]
Liu C, Shi R X, Zhang Y H et al., 2019a. 2015: How many islands (Isles, Rocks), how large land areas and how long of shorelines in the world? Vector data based on Google Earth images. Journal of Global Change Data & Discovery, 3(2): 124-148.

[18]
Liu X L, Wang Z H, Yang X M et al., 2022. Mapping China’s offshore mariculture based on dense time-series optical and radar data. International Journal of Digital Earth, 15(1): 1326-1349.

DOI

[19]
Liu Y M, Wang Z H, Yang X M et al., 2020. Satellite-based monitoring and statistics for raft and cage aquaculture in China’s offshore waters. International Journal of Applied Earth Observation and Geoinformation, 91: 102118.

DOI

[20]
Liu Y M, Yang X M, Wang Z H et al., 2019b. Aquaculture area extraction and vulnerability assessment in Sanduao based on richer convolutional features network model. Journal of Oceanology and Limnology, 37(6): 1941-1954.

DOI

[21]
Lu G Y, Wong D W, 2008. An adaptive inverse-distance weighting spatial interpolation technique. Computers & Geosciences, 34(9): 1044-1055.

DOI

[22]
Lu Y, Shao W, Sun J, 2021. Extraction of offshore aquaculture areas from medium-resolution remote sensing images based on deep learning. Remote Sensing, 13(19): 3854.

DOI

[23]
Ma Y J, Qu X Y, Feng D J et al., 2022. Recognition and statistical analysis of coastal marine aquacultural cages based on R3Det single-stage detector: A case study of Fujian province, China. Ocean & Coastal Management, 225: 106244.

[24]
Mou N X, Wang C Y, Chen J H et al., 2021. Spatial pattern of location advantages of ports along the Maritime Silk Road. Journal of Geographical Sciences, 31(1): 149-176.

DOI

[25]
Nath S S, Bolte J P, Ross L G et al., 2000. Applications of geographical information systems (GIS) for spatial decision support in aquaculture. Aquacultural Engineering, 23(1-3): 233-278.

DOI

[26]
Naylor R L, Hardy R W, Buschmann A H et al., 2021. A 20-year retrospective review of global aquaculture. Nature, 591(7851): 551-563.

DOI

[27]
Ottinger M, Clauss K, Kuenzer C, 2016. Aquaculture: Relevance, distribution, impacts and spatial assessments: A review. Ocean & Coastal Management, 119: 244-266.

[28]
Shi T, Xu Q, Zou Z et al., 2018. Automatic raft labeling for remote sensing images via dual-scale homogeneous convolutional neural network. Remote Sensing, 10(7): 1130.

DOI

[29]
Stiller D, Ottinger M, Leinenkugel P, 2019. Spatio-temporal patterns of coastal aquaculture derived from Sentinel-1 time series data and the full Landsat archive. Remote Sensing, 11(14): 1707.

DOI

[30]
Sui B K, Jiang T, Zhang Z et al., 2020. A modeling method for automatic extraction of offshore aquaculture zones based on semantic segmentation. ISPRS International Journal of Geo-Information, 9(3): 145.

DOI

[31]
Torres R, Snoeij P, Geudtner D et al., 2012. GMES Sentinel-1 mission. Remote Sensing of Environment, 120: 9-24.

DOI

[32]
Wang J, Sui L C, Yang X M et al., 2019a. Extracting coastal raft aquaculture data from Landsat 8 OLI imagery. Sensors, 19(5): 1221.

DOI

[33]
Wang J, Yang X M, Wang Z H et al., 2022a. Monitoring marine aquaculture and implications for marine spatial planning: An example from Shandong province, China. Remote Sensing, 14(3): 732.

DOI

[34]
Wang J C, Fan J C, Wang J, 2022b. MDOAU-Net: A lightweight and robust deep learning model for SAR image segmentation in aquaculture raft monitoring. IEEE Geoscience and Remote Sensing Letters, 19: 1-5.

[35]
Wang M, Cui Q, Wang J et al., 2017a. Raft cultivation area extraction from high resolution remote sensing imagery by fusing multi-scale region-line primitive association features. ISPRS Journal of Photogrammetry and Remote Sensing, 123: 104-113.

DOI

[36]
Wang M, Li G, Liu Y et al., 2017b. Dynamic changes of mariculture areas in eastern Shandong Peninsula in recent 20 years. Journal of Applied Oceanography, 36: 319-326.

[37]
Wang T, Zhang X, Xiong Y et al., 2019b. Remote sensing monitoring and environmental pollution load assessment of coastal aquaculture area based on GF-2. In: 2019 8th International Conference on Agro-Geoinformatics (Agro-Geoinformatics), 1-6.

[38]
Wang Z H, Yang X M, Liu Y M et al., 2018. Extraction of coastal raft cultivation area with heterogeneous water background by thresholding object-based visually salient NDVI from high spatial resolution imagery. Remote Sensing Letters, 9(9): 839-846.

DOI

[39]
Wu J, Del Valle T M, Ruckelshaus M et al., 2021. Dramatic mariculture expansion and associated driving factors in Southeastern China. Landscape and Urban Planning, 214: 104190.

DOI

[40]
Wulder M A, Loveland T R, Roy D P et al., 2019. Current status of Landsat program, science, and applications. Remote Sensing of Environment, 225: 127-147.

DOI

[41]
Xu H, Jia A, Song X, 2023. Extraction and spatiotemporal evolution analysis of tidal flats in the Bohai Rim during 1984-2019 based on remote sensing. Journal of Geographical Sciences, 33(1): 76-98.

DOI

[42]
Xu Y, Wu W, Lu L, 2021. Remote sensing mapping of cage and floating-raft aquaculture in China’s offshore waters using machine learning methods and Google Earth Engine. In: 2021 9th International Conference on Agro-Geoinformatics (Agro-Geoinformatics), 2021 July 26 (pp. 1-5). IEEE

[43]
Xue M, Chen Y, Tian X et al., 2018. Detection the expansion of marine aquaculture in Sansha Bay by remote sensing. In: IGARSS 2018-2018 IEEE International Geoscience and Remote Sensing Symposium, 2018 July 22 (pp. 7866-7869). IEEE.

[44]
Zhang Q N, 1998. Resources, environment and sustainable development in Chinese coast. Journal of Hubei University.

[45]
Zhang Y, Wang C, Ji Y et al., 2020. Combining segmentation network and nonsubsampled contourlet transform for automatic marine raft aquaculture area extraction from Sentinel-1 images. Remote Sensing, 12(24): 4182.

DOI

[46]
Zheng Y H, Wu J P, Wang A Q et al., 2017. Object- and pixel-based classifications of macroalgae farming area with high spatial resolution imagery. Geocarto International, 33(10): 1048-1063.

DOI

Options
Outlines

/

〈 〉