Conservation and restoration efforts have promoted increases in shorebird populations and the area and quality of their habitat in the Yellow River Delta, China

ABSTRACT

Conservation policies have been formulated for coastal wetlands in China, and exploration of conservation effectiveness based on waterbirds and their habitat is important for guiding conservation actions. We characterized the effects of conservation efforts on shorebird diversity, habitat area and quality using long-term remote sensing data, and shorebird survey data in the Yellow River Delta. From 1997 to 2021, habitat area, quality and population number significantly declined by 49.8% (r = −0.72, p < 0.05), 13.6% (r = −0.72, p < 0.05) and 60.67% (R²= 0.77, p < 0.05). Before 2012, habitat area (decreased by 38.2%, r = −0.62, p > 0.05, slope = −0.25), quality (decreased by 10.53%, r = −0.68, p > 0.05, slope = −0.008), and population size (significantly decreased by 94.5%, r = −0.95, p < 0.05, slope = −7874.3) declined, and the decline in habitat area significantly contributed to population reductions (r = 0.79, p < 0.05). Since 2012, habitat area (increased by 14.3%, r = 0.71, p > 0.05, slope = 0.12), quality (increased by 17.12%, r = 0.83, p > 0.05, slope = 0.01), and population size (increased by 8.34%, R²= 0.29, p > 0.05) slightly increased. The coefficients of variation for habitat area and quality, and population size were smaller after 2012 than before 2012. These results suggest that conservation actions maintained the stability of waterbird populations and their habitat; additional actions are needed to mediate the conservation of other degraded habitats along coastal wetlands.

KEYWORDS:

• Shorebirds
• habitat degradation
• loss of shorebird diversity
• conservation and restoration
• Yellow River Delta

1. Introduction

Coastal wetlands form a transition zone between terrestrial and marine environments and have been greatly affected by anthropogenic disturbances, as well as sea level rise (Törnqvist et al. 2021), coastal erosion (Cellone, Carol, and Tosi 2016), and invasions of alien species (Mao et al. 2019). These wetlands are becoming increasingly threatened by rapid economic development and human population growth (Ma et al. 2010a). The Chinese coastline supports 40% of the human population in China and accounts for 58.6% of China’s Gross Domestic Product (Yu, Zhang, and Yang 2022). Approximately 21.9% of the area of coastal wetlands has been lost in China between 2003 and 2014, when the first and second national surveys of wetland resources in China were conducted, respectively (SFA 2014).

Natural wetlands along coastal regions carry out multiple important ecosystem functions; they also provide key breeding, stopover, and wintering grounds for migratory shorebird species along the East Asian–Australasian Flyway (EAAF) migratory route (Bai et al. 2015; Lei et al. 2021; Peng et al. 2021). According to previous studies, several species that visit multiple sites along the coastal wetlands of China meet the Ramsar 1% criterion (exceeding 1% of the total population in the flyway) (Xia et al. 2016). A total of 48 important habitat sites are visited by 31 shorebird species that meet the Ramsar 1% criterion (Duan et al. 2022b).

The degradation and loss of natural wetlands have posed serious threats to shorebird populations over the past few decades (Moores et al. 2016; Murray and Fuller 2015). The dual threat of tidal flat loss and the invasion of Spartina alterniflora have reduced the area and quality of shorebird habitat along coastal wetlands (Jackson et al. 2021) and increased competition for space and food among shorebirds (Wang et al. 2022). The loss of coastal habitats along the Yellow Sea coast between 1993 and 2012 has resulted in an annual decrease in the population size of seven shorebird species as high as 7% (Studds et al. 2017). Population decreases are especially pronounced for small habitat patches; a previous study has shown that the abundance of 19 shorebird species declined by 90.14% along the coast of the Yellow River Delta because of habitat loss over the past 25 years (Duan et al. 2022a).

A series of conservation and restoration actions for China’s coasts have been formulated. For example, a national strategy of ‘ecological civilization construction’ was implemented, and the State Oceanic Administration promulgated the National Marine Function Zoning Plan (2011–2020) to strengthen the management of reclamation projects and control the scale of reclamation (State Oceanic Administration 2012). The effectiveness of conservation and restoration actions was subsequently assessed by comparison of the population size of waterbirds before and after restoration. However, most previous studies have used short-term monitoring data of waterbird populations to conduct rapid assessments of restoration effectiveness (Fan et al. 2021; Kaˇcergyte et al. 2022; Li et al. 2011; Liu et al. 2023; Sun et al. 2019). Long-term data are better for evaluating the effectiveness of restoration actions because they are less likely to be affected by short-term stochastic variation in populations; these data could thus provide superior guidance for future restoration actions. In addition, some studies have assessed restoration effectiveness by comparing the area and quality of habitat before and after restoration using remote sensing images (Wang et al. 2021; Wang et al. 2023b). The relationship between restored habitat and waterbird population size has not yet been quantified.

In this study, we tested whether conservation and restoration efforts of a coastal wetland along the Yellow River Delta can improve shorebird habitat and biodiversity using long-term survey data of waterbirds and a long time series of remote sensing images (between 1997 and 2021). We also quantified the relationship between habitat change and variation in waterbird population sizes. Specifically, we evaluated two hypotheses: (1) conservation and restoration efforts mitigate decreases in the area and quality of shorebird habitat and shorebird diversity, and (2) there is a positive relationship between change in habitat area and the quality and variation in shorebird diversity following the formulation of conservation and restoration policies.

2. Materials and methods

2.1. Study area

Our study was conducted in the Yellow River Delta, which is an important stopover site for migratory shorebirds in Dongying City, Shandong Province, China (Figure 1) (Hou et al. 2021). The Yellow River Delta is a Ramsar site, and it was designated as a National Nature Reserve in 1992. Coastal wetlands in the Yellow River Delta provide habitat for multiple shorebird species, and the populations of many of these species meet the Ramsar 1% criterion (Xia et al. 2016). Land reclamation and invasions of S. alterniflora have reduced the area and quality of shorebird habitat and led to shorebird population declines (Duan et al. 2022a). Conservation and restoration policies and regulations, including ecological civilization policies and a national system for marine function zoning, were formulated in 2012 to mitigate decreases in the area of coastal wetlands (Wang et al. 2023). These actions could potentially benefit shorebird populations and their habitat.

Figure 1. The location of the study area and land use and land cover in the Yellow River Delta in 2021.

2.2. Land cover data

Land cover refers to the observed physical cover of the surface caused by natural processes or human activities; it not only is limited to vegetation on the surface but also includes various artificial covers and artificial modifications on the surface (Geist 2006). For example, land cover types include forest, grassland, and rivers. Object-oriented classification is a remote sensing image classification technology where textural and contextual/relational information is used in addition to spectral information for classifying data. It is useful for extracting features from high spatial resolution but low spectral resolution images (Bhaskaran, Paramananda, and Ramnarayan 2010; Luo and Yan 2009). Object-oriented classification is driven by the use of image objects rather than pixels. An image object comprises several homogeneous pixels and regions with similar spectral or spatial characteristics. They form building blocks that comprise abstract information, and these building blocks can be used at various image object levels (Bhaskaran, Paramananda, and Ramnarayan 2010). Image objects were extracted on the basis of various segmentation parameters using the homogeneity criterion of the multi-resolution segmentation algorithm, and measured the homogeneous or heterogeneous of an image object (Hofmann 2001).

We used the object-oriented classification method, as well as Landsat TM/ETM and Landsat 8 operational land imager (OLI) data at a scale of 1:100000, to acquire land cover data in the Yellow River Delta between 1997 and 2021 (1997, 1999, 2007, 2008, 2010, 2012, 2014, 2015, 2017, 2018, and 2021). The spatial resolution of these land cover data was 30 m × 30 m. These images were selected from the Geospatial Data Cloud (www.gscloud.cn/sources/) and the USGS Global Visualization Viewer (GloVis) (http://glovis.usgs.gov) according to the following criterion: total cloud cover < 5%. The classification system of land cover was consistent with that used in Di, Hou, and Wu (2014). We used ground-truth data from field survey photos to verify the classification results. Generally, the classification accuracy for these images between 1997 and 2021 was between 91% and 94%, and the kappa coefficients were between 0.90 and 0.95. S. alterniflora has occupied areas of shorebird habitat in the Yellow River Delta since 2010 (Jackson et al. 2021). We used object-based image analysis, support vector machine methods, and field investigations based on collected Landsat OLI data to identify the distribution of S. alterniflora between 2010 and 2021. The classification accuracy of these images was between 90% and 95%, and the kappa coefficients were between 0.85 and 0.92. The classification accuracy and kappa coefficients in this study were statistically obtained. We merged the distribution of S. alterniflora with maps of land use and land cover between 2010 and 2020 using the ‘Mosaic To New Raster’ tool in ArcGIS 10.5.

2.3. Simulation of shorebird habitat

We used the habitat module of the Integrated Valuation of Environmental Services and Tradeoffs (InVEST; version 3.7.0) model to assess the quality of shorebird habitat in the Yellow River Delta between 1997 and 2021 (1997, 1999, 2007, 2008, 2010, 2012, 2014, 2015, 2017, 2018, and 2021). The InVEST model was developed by the Natural Capital Project and has been used to assess ecosystem services (Hong et al. 2021). The habitat module can acquire a map of habitat quality based on habitat type and threat factors (Equation (1(1) $Q_{ij}=H_j\left(1-\left({\displaystyle \frac{D_{ij}^z}{D_{ij}^z+k^z}}\right)\right)$(1) )): (1) $Q_{ij}=H_j\left(1-\left({\displaystyle \frac{D_{ij}^z}{D_{ij}^z+k^z}}\right)\right)$(1) where Q_ij represents the habitat quality in raster cell i with land use and land cover j, H_j is the habitat suitability in land use and land cover j, and D_ij represents the threat intensity on raster cell i with land use and land cover j. z and k are the scaling constant and half-saturation constant, respectively. We set z and k to 2.5 and 0.5 in the InVEST model, respectively (Sharp et al. 2016).

2.3.1. Input data

The input factors included habitat type, habitat suitability, threat sources, threat characteristic, the sensitivity of habitat to threat factors, and habitat accessibility. Land use and land cover types, including farmland, reservoirs/ponds, bottomlands, tidal flats, estuarine waters, estuarine deltas, saltpans, mariculture, and unused land, were used as habitat type data. These habitat types were extracted from land cover maps of 1997, 1999, 2007, 2008, 2010, 2012, 2014, 2015, 2017, 2018, and 2021 in the Yellow River Delta. Habitat suitability was a continuous value between 0 (lowest suitability) and 1 (highest suitability), and these values were set using the method of Li et al. (2019). Habitat suitability information is provided in Supplementary Table 1.

Land type, including cities, rural settlements, industrial mining, mariculture, roads, unused land, and the presence of S. alterniflora, were used as threat factor data. These threat types were extracted from land cover data between 1997 and 2021. Threat characteristics included the relative intensity, maximum disturbance distance, and type of distance-based decay, and this information was derived from Li et al. (2019) (Supplementary Table 1). The sensitivity of habitat to threat factors represents the relative sensitivity of habitat to disturbance types. Habitat accessibility represents the relative accessibility of the protected areas to each threat source. The accessibility value was between 0 and 1, where 1 indicates complete accessibility. The Yellow River Delta National Nature Reserve was created in 1992. The reserve was divided into three parts, core area, buffer area, and experimental area, according to the level of human activity permitted. Access to the core area is prohibited with the exception of special permits; we set the value of the core area to 0.1. The buffer area is the area surrounding the core area, and scientific research (observational studies) is permitted in this area; we set the value of the buffer area to 0.5. The experimental area is the area surrounding the buffer area; several types of activities are permitted, such as scientific experiments, publicity and education, visits, investigations, tourism, economic production activities, and other activities; we set the value of the experimental area to 1.0. This means that the experimental area needs to be twice as accessible as the buffer area. The shapefile boundaries of the core area, buffer area, and experimental area were acquired from the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (http://english.igsnrr.cas.cn/).

2.3.2. InVEST model

We input the above parameters into the InVEST model to characterize habitat quality in 1997, 1999, 2007, 2008, 2010, 2012, 2014, 2015, 2017, 2018, and 2021 in the Yellow River Delta. Habitat quality values ranged from 0 to 1, with higher values indicating higher habitat quality. According to the habitat quality threshold proposed in previous studies (Duan and Yu 2022c; Li et al. 2019; Ren et al. 2014), the distribution of suitable habitat was inferred using a habitat quality threshold of 0.7. Generally, areas with habitat quality values above 0.7 are areas that are actually used by shorebirds.

2.4. Shorebird population data

Waterbird surveys in the Yellow River Delta were conducted once a month since 2007 by experienced observers from the Yellow River Delta National Nature Reserve Management Bureau and Dongying City Bird Watching Association. In addition, shorebird surveys in the same region in 1997 and 1999 were conducted by Barter, M. A. According to previous literature, shorebird population numbers and number of species in the Yellow River Delta are highest during the northward migration period (between March and May). Thus, we compared the status of shorebird populations and the numbers of species using data collected from March to May each year. Shorebird population and species number data in 1997, 1999, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2017, 2018, and 2021 were selected, given that these were the years with available data.

2.5. Changes in habitat area, habitat quality, and shorebird diversity

Conservation and restoration policies, including ecological civilization policies and a national system for marine function zoning, were formulated for coastal wetlands in 2012. According to previous studies, the area of coastal wetlands decreased before 2012 and slightly increased after 2012 (Wang et al. 2021; Wang et al. 2023). We assume that these conservation measures have had similar effects on the diversity and habitat of shorebirds. We determined the percent change in habitat area, habitat quality, and shorebird diversity (including the number of species and population size) over three periods (1997–2021, 1997–2012, and 2012–2021).

We calculated the percent change in habitat area, habitat quality, number of species, and population size in three periods (Equation (2(2) $VR={\displaystyle \frac{b-a}{a}\times 100\text{\%}}$(2) )): (2) $VR={\displaystyle \frac{b-a}{a}\times 100\text{\%}}$(2) where a and b represent the habitat area, habitat quality, number of species, and population size in the previous period and subsequent period, respectively.

We also used the coefficient of variation (CV) (Equation (3(3) $CV={\displaystyle \frac{SD}{Mean}\times 100\text{\%}}$(3) )) to assess the magnitude of fluctuations in habitat area, habitat quality, number of species, and population size between 1997 and 2012 and between 2012 and 2021. Higher CVs indicate greater fluctuations. (3) $CV={\displaystyle \frac{SD}{Mean}\times 100\text{\%}}$(3)

2.6. Statistical analyses

Interannual variation in habitat area, habitat quality, number of species, and population size in the three periods was analyzed using linear regression models or binomial regression models with a t-test at the 5% significance level using Origin 2019b (OriginLab). Additionally, we conducted Spearman’s correlation analyses and t-tests at the 5% significance level in IBM (International Business Machines Corporation) SPSS (Statistical Product and Service Solutions) Statistics 22.0 to evaluate the relationships of habitat area and habitat quality with the number of species and population size in the three periods. The IBM SPSS software platform was developed by IBM and can be used to conduct advanced statistical analyses; it also has a vast library of machine learning algorithms and can be used to conduct text analysis.

3. Results

3.1. Changes and interannual trends in habitat area and quality

Simulated shorebird habitat by the InVEST model was mainly distributed along coastal regions in the Yellow River Delta between 1997 and 2021. Similarly, habitats with high quality were distributed along coastal regions (Figure 2). The area of shorebird habitat significantly decreased in the Yellow River Delta from 1997 to 2021 (r = −0.72, p < 0.05, Figure 3) by 49.8%. The area of shorebird habitat decreased by 38.2% from 1997 to 2012 (r = −0.62, p > 0.05, slope = −0.25). Between 2013 and 2021, the area of shorebird habitat increased by 14.3% (r = 0.71, p > 0.05, slope = 0.12) (Figure 3). The coefficient of variation of the area of shorebird habitat between 1997 and 2012 (20.60%) was larger than that between 2013 and 2021 (8.83%).

Figure 2. Maps of the quality of shorebird habitat in the Yellow River Delta between 1997 and 2021.

Figure 3. Area of shorebird habitat in the Yellow River Delta between 1997 and 2021. Red and blue shading represents the 95% confidence interval.

Habitat quality significantly decreased in the Yellow River Delta between 1997 and 2021 by 13.6% (r = −0.72, p < 0.05, Figure 4). From 1997 to 2012, habitat quality decreased by 10.53% (r = −0.68, p > 0.05, slope = −0.008), and from 2013 to 2021, habitat quality increased by 17.12% (r = 0.83, p > 0.05, slope = 0.014) (Figure 4). The coefficient of variation of habitat quality was slightly greater between 1995 and 2012 (7.52%) than between 2013 and 2021 (6.90%).

Figure 4. Shorebird habitat quality in the Yellow River Delta between 1997 and 2021. Red and blue shading represents the 95% confidence interval.

3.2. Changes and interannual trends in shorebird diversity

3.2.1. Observed species

A total of 45 shorebird species in four orders and six families, including 13 globally threatened species and 32 non-threatened species, were detected in the Yellow River Delta between 1997 and 2021. Eight shorebird species were nationally protected species, including Nordmann’s Greenshank (Tringa guttifer), Eurasian Curlew (Numenius arquata), Asian Dowitcher (Limnodromus semipalmatus), Eastern Curlew (Numenius madagascariensis), Great Knot (Calidris tenuirostris), Ruddy Turnstone (Arenaria interpres), Broad-billed Sandpiper (Limicola falcinellus), and Little Curlew (Numenius minutus) (Table 1).

Table 1. Observed shorebirds and their IUCN Red List categories and conservation status throughout the 25-year survey period.

Taxonomic order	Code	Common name	Scientific name	IUCN Red List	Conservation status
Haematopodidae	1	Eurasian Oystercatcher	Haematopus ostralegus	NT	–
Recurvirostridae	2	Black-winged Stilt	Himantopus himantopus	LC	–
	3	Pied Avocet	Recurvirostra avosetta	LC	–
Charadriidae	4	Oriental Dotterel	Charadrius veredus	LC	–
	5	Northern Lapwing	Vanellus vanellus	NT	–
	6	Kentish Plover	Charadrius alexandrinus	LC	–
	7	Grey Plover	Pluvialis squatarola	LC	–
	8	Grey-headed Lapwing	Vanellus cinereus	LC	–
	9	Common Ringed Plover	Charadrius hiaticula	LC	–
	10	Pacific Golden Plover	Pluvialis fulva	LC	–
	11	Little Ringed Plover	Charadrius dubius	LC	–
	12	Lesser Sand Plover	Charadrius mongolus	LC	–
	13	Greater Sand Plover	Charadrius leschenaultii	LC	–
	14	Long-billed Plover	Charadrius placidus	LC	–
Rostratulidae	15	Greater Painted-snipe	Rostratula benghalensis	LC	–
Scolopacidae	16	Eurasian Curlew	Numenius arquata	NT	II
	17	Green Sandpiper	Tringa ochropus	LC	–
	18	Bar-tailed Godwit	Limosa lapponica	NT	–
	19	Asian Dowitcher	Limnodromus semipalmatus	NT	II
	20	Eastern Curlew	Numenius madagascariensis	EN	II
	21	Great Knot	Calidris tenuirostris	EN	II
	22	Ruddy Turnstone	Arenaria interpres	LC	II
	23	Spotted Redshank	Tringa erythropus	LC	–
	24	Dunlin	Calidris alpina	LC	–
	25	Black-tailed Godwit	Limosa limosa	NT	–
	26	Red Knot	Calidris canutus	NT	–
	27	Common Redshank	Tringa totanus	LC	–
	28	Red-necked Stint	Calidris ruficollis	NT	–
	29	Common Sandpiper	Actitis hypoleucos	LC	–
	30	Sharp-tailed Sandpiper	Calidris acuminata	VU	–
	31	Broad-billed Sandpiper	Limicola falcinellus	LC	II
	32	Wood Sandpiper	Tringa glareola	LC	–
	33	Ruff	Philomachus pugnax	LC	–
	34	Terek Sandpiper	Xenus cinereus	LC	–
	35	Temminck's Stint	Calidris temminckii	LC	–
	36	Common Greenshank	Tringa nebularia	LC	–
	37	Sanderling	Calidris alba	LC	–
	38	Common Snipe	Gallinago gallinago	LC	–
	39	Curlew Sandpiper	Calidris ferruginea	NT	–
	40	Little curlew	Numenius minutus	LC	II
	41	Nordmann's Greenshank	Tringa guttifer	EN	I
	42	Marsh Sandpiper	Tringa stagnatilis	LC	–
	43	Long-toed Stint	Calidris subminuta	LC	–
	44	Whimbrel	Numenius phaeopus	LC	–
Glareolidae	45	Oriental Pratincole	Glareola maldivarum	LC	–

3.2.2. Variation and interannual trends in the number of shorebird species and their abundances

No pronounced variation was observed in the number of shorebird species between 1997 and 2021 (p > 0.05, r = 0.34, Figure 5). There was no significant change in the number of shorebird species between 1997 and 2012 (r = 0.01, p > 0.05, slope = 2.8, Figure 5) nor between 2012 and 2021 (r = 0.02, p > 0.05, slope = 7.5, Figure 5). However, the coefficient of variation of the number of species was greater between 1997 and 2012 (21.11%) than between 2013 and 2021 (8.90%).

Figure 5. Changes in the number of shorebird species in the Yellow River Delta between 1997 and 2021. Red and blue shading represents the 95% confidence interval.

The number of populations significantly decreased by 60.67% between 1997 and 2021 in the Yellow River Delta (p < 0.05, R²= 0.77, Figure 6). The number of populations first decreased and then increased between 1997 and 2021. The number of populations significantly decreased by 94.5% from 1997 to 2011 (r = −0.95, p < 0.05, slope = −7874.3, Figure 6) and slightly increased by 8.34% from 2012 to 2021 (p > 0.05, R²= 0.29, Figure 6). The coefficient of variation of the number of species was higher between 1997 and 2012 (109.60%) than between 2013 and 2021 (19.60%).

Figure 6. Changes in the number of shorebird populations in the Yellow River Delta between 1997 and 2021. Red and blue shading represents the 95% confidence interval.

3.3. Relationship between habitat change and variation in shorebird diversity

Between 1997 and 2021, no significant relationship was observed between changes in the area of shorebird habitat and the number of species (r = −0.37, p > 0.05) (Figure 7) and populations (r = −0.033, p > 0.05). There was also no significant relationship between changes in the quality of shorebird habitat and the number of species (r = −0.26, p > 0.05) (Figure 7) and populations (r = −0.10, p > 0.05). Before 2012, there was a significant relationship between the area of shorebird habitat and population size (r = 0.79, p < 0.05) (Figure 7). There was no significant relationship between the area of shorebird habitat and the number of species (r = −0.43, p > 0.05); no significant relationship was observed between habitat quality and the number of species (r = −0.11, p > 0.05) nor between habitat quality and population size (r = 0.47, p > 0.05). After 2012, there was a slight positive relationship of variation in the area of shorebird habitat with habitat quality (r = 0.27, p > 0.05) and population size (r = 0.23, p > 0.05) (Figure 7).

Figure 7. Relationship between habitat area and species diversity (left) and between habitat quality and species diversity (right). The solid and dashed lines represent the number of species and populations, respectively. Red circles represent significant associations (p < 0.05). Gray bars indicate 95% confidence intervals.

4. Discussion

The loss and degradation of coastal wetlands and their deleterious effects on biodiversity have been documented at the global (Murray et al. 2018), country (Ma et al. 2010a), and regional scale (Murray et al. 2014) over the past few decades. Wetland restoration projects have been carried out in multiple countries. In China, several wetland restoration projects have been carried out in the Yellow Sea region. The area of coastal wetlands has increased gradually in China since 2012 when conservation and restoration measures were first implemented (Wang et al. 2021). We compared changes in shorebird habitat and diversity before and after 2012 using the habitat module of the InVEST model and long-term shorebird survey data. Shorebird populations, habitat quality, and habitat area have declined from 1997 to 2021, and these declines were particularly rapid from 1997 to 2012. Values of each of these variables increased after 2012 compared with those before 2012, and the coefficient of variation of these variables was smaller after 2012 compared with that before 2012.

Coastal wetlands in China have declined sharply since 1960 because of land reclamation and natural hazards (Murray et al. 2014). These reductions pose major threats to shorebird populations (Wang et al. 2022). According to a previous study, the loss of habitat at stopover sites in the Yellow Sea because of land reclamation has resulted in rapid population declines of seven shorebird species (Studds et al. 2017). In this study, the decline in the area of shorebird habitat before 2012 greatly contributed to reductions in shorebird populations. After 2012, the area and quality of habitat and shorebird population size slightly increased. In 2012, ecological civilization construction was adopted as a national development strategy, and a series of conservation and restoration development strategies were formulated; the aim of this strategy is to protect the environment and promote life in harmony with nature. The National Marine Function Zoning Plan (2011–2020) was formulated by the State Oceanic Administration to control the intensity of reclamation (State Oceanic Administration 2012). These conservation and restoration regulations have caused significant reductions in the area of aquaculture ponds (Wang et al. 2023a) and mitigated the intensity of the development of coastal wetlands. In addition, the Yellow River Delta was established as a Ramsar site in 2013, and this has played a critical role in slowing declines in the area of coastal wetlands. These actions might have also contributed to increases in the area and quality of coastal habitat since 2012.

The restored habitat greatly contributes to increases in the size of waterbird populations, and this has been confirmed by previous studies (Fan et al. 2021; Liu et al. 2023; Sun et al. 2019). However, in this study, there was a slightly positive relationship between the area of restored habitat and waterbird population size; potential explanations for this finding are manifold. First, although the object-based classification method can be used to extract objects at various spatial scales and rapidly analyze images based on the shape, size, and texture of images, the classification accuracy greatly depends on the segmentation results and classification rules, and errors can lead to decreases in the classification accuracy of coastal wetlands (Ma et al. 2017; Radoux et al. 2010). In addition, the habitat module of the InVEST model can be used to explore changes in habitat quality over large spatial and temporal scales, and some parameters for habitat accessibility and habitat sensitivity to potential threats were identified subjectively; consequently, error and uncertainty of the results are inevitable. These might have implications for the analysis of habitats and introduce error in estimates of the relationship between habitat change and population variation after 2012. Second, the small sample size caused by short survey period of wetland habitat and shorebird populations (between 2012 and 2021) might also explain the weak relationship. Since 2012, shorebird populations generally ‘bounced back’ in nearby natural wetlands, such as the Binzhou Beikedi Island and Wetland refuge and the beaches of the Yellow River in Shandong Province according to unpublished survey data. This could explain the population increase observed since 2012 in our study area. We also observed a sudden increase in the number of shorebird populations from 2011 to 2012. The positive effects of conservation actions would not have been immediately apparent in this period; instead, these changes are likely better explained by changes in regulations and extreme climatic events.

According to previous studies, coastal wetland conservation is one of the ‘weak spots’ of wetland conservation in China. The results of the Second National Wetland Resources Survey published by the former State Forestry Administration (SFA, now the National Forestry and Grassland Administration, or NFGA) in 2014 showed that the acreage of coastal wetlands under protection stands at only 1,390,400 hectares, accounting for only 23.99% of the total area of coastal wetlands, which is much lower than the national average of conserved wetland areas (43.5%) (Yu and Zhang 2020). Increased conservation and restoration efforts are thus needed to increase the area of coastal wetlands. On December 24, 2021, the 32nd Meeting of the 13th National People’s Congress Standing Committee (NPCSC) reviewed and adopted the Wetlands Conservation Law of the People’s Republic of China (PRC) (hereafter referred to as the Wetlands Conservation Law), which was then promulgated by Presidential Decree No. 102 and came into force as of June 1, 2022. This marks the first time that China has mainstreamed wetland conservation in law-based management. The introduction of the Wetlands Conservation Law has filled the legislative gap in wetland ecosystem conservation and enhanced the ecological civilization system in China (Yu, Zhang, and Yang 2022).

We evaluated the effectiveness of conservation and restoration actions along coastal wetlands on the basis of waterbirds and their habitat. The slightly positive relationship between habitat changes and shorebird population size changes after conservation and restoration policies were implemented and the significant relationship between the decline in habitat quality and area and reduction in population size before the related policy was introduced suggest that there is an urgent need for land development in areas with coastal wetlands to be continuously restricted. Waterbird diversity and species composition differed among habitat types because of the diverse habitat requirements of different foraging guilds (Cumming et al. 2012). In the future, conservation and restoration projects should trade-off among diverse habitat requirements of various foraging guilds (Ma et al. 2010b). Long-term monitoring of waterbird populations is also needed to robustly evaluate the effectiveness of conservation and restoration actions and improve the effectiveness of future actions. In addition, multiple coastal wetlands along the EAAF migratory route provide important breeding, stopover, and non-breeding sites for migratory waterbirds (Bai et al. 2015). Land reclamation activities have posed a serious threat to various sites, such as the Luannan Coast and salt pan in Hebei Province, the Binhai New Area in Tianjin City, Jinzhou Bay in Liaoning Province, and Ganyu Coast in Jiangsu Province (Duan et al. 2020); conservation and restoration in these areas are needed to increase the connectivity between habitat patches.

5. Conclusions

We evaluated whether conservation and restoration measures for coastal wetlands implemented in 2012 have affected shorebird diversity and their habitat in the Yellow River Delta. From 1997 to 2021, the area and quality of shorebird habitat decreased by 49.8% and 13.6%, respectively, and the number of populations decreased by 60.67%. Before 2012, the reduction in the area of shorebird habitat contributed greatly to the decline in shorebird populations. After 2012, a series of conservation and restoration policies and regulations were formulated for coastal wetlands. Since 2012, the area and quality of shorebird habitat and the population size of shorebirds have slightly increased. There was a slight positive relationship between variation in the area and quality of habitat and change in population size since 2012. In addition, the magnitude of variation in habitat area, habitat quality, and population size was smaller after 2012 than before 2012. These findings indicate that conservation and restoration measures should be continuously implemented in coastal wetlands to conserve waterbird biodiversity.

Author contributions

X. Y. and H. D. planned and designed the research; H. D. and S. X collected data; H. D. analyzed data and wrote the manuscript; and H. D. and X. Y. collaboratively revised the manuscript.

Acknowledgments

We thank Liu, Y. from the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences for discussion and providing literature that helped improve the Methods and Results sections of the manuscript. We thank Dehua Mao for providing Spartina alterniflora distribution data.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data are available from Figshare: https://doi.org/10.6084/m9.figshare.22239655.

Additional information

Funding

This research was jointly supported by the National Key Research and Development Program of China (2022YFF0802400), the Science and technology basic resources survey project (2021FY101002), and the National Natural Science Foundation of China (42101105).