Next Article in Journal
“Thrown Out in the Woods”: Fiber Farming, Translation Breakdown, and the Hollowed Supply Chain in West Virginia
Previous Article in Journal
Hybrid Organizationality: Dual Institutional Logics in a Large-Scale Agrifood Communal Trade Network
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 18
Issue 12
10.3390/su18125887
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

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

Direct and Indirect Effects of Natural and Anthropogenic Drivers on Avian Diversity in the Sanjiang Plain, Northeast China

by
Xiuli Sun
1,2,
Chenxiao Liu
1,2,
Yueyuan Li
2,3,
Yuehui Li
2,* and
Yue Li
1,*
1
School of Environment, Liaoning University, Shenyang 110036, China
2
CAS Key Laboratory of Forest Ecology and Silviculture, Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China
3
College of Urban Construction, Heze University, Heze 274015, China
*
Authors to whom correspondence should be addressed.
Sustainability 2026, 18(12), 5887; https://doi.org/10.3390/su18125887 (registering DOI)
Submission received: 15 April 2026 / Revised: 28 May 2026 / Accepted: 4 June 2026 / Published: 9 June 2026
Browse Figures
Versions Notes

Abstract

Understanding how natural and anthropogenic factors jointly influence avian diversity is essential for biodiversity conservation and the sustainable management of large-scale wetland ecosystems, yet their combined effects remain insufficiently understood. This gap is particularly evident for land birds, as most studies focus on waterbirds. Using structural equation modeling, we quantified the effects of these drivers on habitat quality and avian richness in the Sanjiang Plain, separately for waterbirds and land birds. Our results show that: (1) habitat quality is primarily controlled by natural factors, particularly soil organic carbon (SOC), normalized difference vegetation index (NDVI), and topography, whereas human activities exert weak negative effects; (2) waterbirds are primarily associated with SOC- and temperature-driven pathways, whereas land birds respond more directly to climate and human disturbance; (3) natural drivers exert stronger effects than anthropogenic factors on both waterbird and land bird diversity; and (4) the effects of natural drivers differ between bird groups, with SOC and NDVI showing stronger effects on waterbirds, and precipitation and temperature being more influential for land birds. These findings highlight the need for group-specific conservation strategies, including conserving soil carbon and maintaining hydrological conditions for waterbirds, and enhancing vegetation and mitigating human disturbance for land birds.

1. Introduction

Birds provide a wide range of ecosystem services that contribute to ecosystem stability, biodiversity conservation, and the sustainable management of wetland ecosystems [1,2]. In wetland ecosystems, waterbirds play critical roles in regulating food webs, transferring nutrients, and maintaining ecological integrity, making them sensitive indicators of environmental change [3,4]. Land birds, in contrast, are key components of forest, grassland, and agricultural ecosystems, where they regulate insect populations, disperse seeds, and support trophic interactions. However, rapid climate change and increasing anthropogenic pressures have led to widespread ecosystem degradation, resulting in substantial global declines in avian diversity over the past century [5,6]. According to The State of the World’s Birds 2022 report, nearly 49% of global bird populations are declining, and one in eight species is threatened with extinction [7]. This alarming trend highlights the urgent need to conserve and restore avian diversity worldwide, thereby underscoring the importance of understanding the mechanisms shaping avian diversity.
Species diversity is closely associated with habitat quality, reflecting the capacity of an ecosystem to provide essential resources and conditions for species survival and reproduction [8,9,10,11,12]. Environmental drivers influence avian diversity directly and indirectly via habitat quality. These drivers include natural factors such as climate, soil, and vegetation [13,14,15], as well as anthropogenic pressures such as land-use change and infrastructure development [16]. For example, soil organic carbon (SOC) and normalized difference vegetation index (NDVI) together significantly shape avian habitat quality across regions [17,18]. In heterogeneous landscapes, particularly large wetland systems that are highly sensitive to environmental change, understanding habitat quality and its driving factors is especially important [19,20,21].
Beyond their effects on habitat quality, environmental drivers also shape avian diversity in a context-dependent manner across landscapes with differing intensities of human activity [22,23]. These drivers operate through multiple pathways that collectively form a complex network linking environmental factors, habitat quality, and biodiversity. For example, climatic factors exert consistent effects on waterbird diversity in the Liaohe Estuary wetland, whereas human-induced fragmentation operates through both direct and habitat-mediated pathways [22]. Similarly, agricultural land cover reduces forest bird diversity via habitat loss in eastern Canada [23]. Within this network, natural gradients often set baseline ecological conditions, whereas anthropogenic pressures may modify or override these effects [24]. Such complexity highlights the need to disentangle these pathways and to clarify the relative roles of human activities and natural factors in shaping avian diversity [25,26]. This understanding is essential for forecasting biodiversity responses under future scenarios and guiding targeted conservation strategies that preserve key natural gradients while mitigating anthropogenic pressures [27,28].
Against this backdrop of interacting natural and anthropogenic influences, waterbirds and land birds differ fundamentally in their habitat requirements. Waterbirds are closely associated with aquatic habitats along the wetland hydrological gradient, whereas land birds comprise terrestrial and upland-associated bird species occurring within the wetland landscape mosaic [22,23]. Wetland landscapes comprise a heterogeneous mosaic of wetland patches and other land-cover types and support a range of land and ecotone-associated bird species that utilize adjacent upland habitats and transitional zones between aquatic and land environments [25,29,30]. In such wetland landscapes, understanding how land birds respond to environmental drivers is critical for explaining interactions across aquatic and land interfaces, which underpin food web dynamics, ecological interactions [26,31] and cross-boundary energy flows [32]. However, previous research has largely focused on changes in wetland area, paying little attention to the pathways linking avian diversity and habitat quality, while studies on terrestrial birds remain particularly scarce, thereby hindering a complete understanding of avian diversity patterns in these landscapes.
The Sanjiang Plain in northeastern China lies on the East Asian–Australasian Flyway and serves as a critical stopover and breeding node for migratory waterbirds. It has experienced intensive and rapidly changing human activities: large-scale agricultural reclamation since the 1950s has substantially transformed its natural landscape, while ecological restoration and land-use adjustments in recent decades have partly alleviated these impacts. These activities have shaped the region into a complex mosaic of wetland, cropland, and built-up patches. Thus, this region represents an ideal context for investigating the responses of waterbirds and land birds to natural and anthropogenic drivers. Our study aims to: (1) characterize habitat quality and identify its driving factors; (2) quantify the direct and indirect effects of natural and anthropogenic drivers on waterbird and land bird diversity; and (3) examine differences in these pathways between waterbird and land bird communities.

2. Materials and Methods

2.1. Study Area

The Sanjiang Plain (43°49′ N to 48°27′ N, 129°11′ E to 135°05′ E), located in northeastern China, is an alluvial plain formed by the Heilongjiang (Amur River), Ussuri River, and Songhua River systems (Figure 1). It encompasses a total area of approximately 109,000 km2, with an average elevation of 50–70 m. The study area experiences a temperate humid and semi-humid continental monsoon climate, with an average annual temperature ranging from 1 °C to 4 °C and an annual precipitation between 500 and 650 mm. The vegetation is dominated by forests, meadows, marshes, and aquatic vegetation, and the main soil types include Albic Luvisols, meadow soils, swamp soils, and Chernozems. Numerous rivers traverse the landscape, including the Muling River, Bira River, and Nongjiang River.
The Sanjiang Plain contains several national nature reserves, including Sanjiang National Nature Reserve, Honghe National Nature Reserve, Qixinghe National Nature Reserve, and Zhenbaodao National Nature Reserve, all of which have been designated as Ramsar wetlands of international importance. The region supports several nationally and internationally important waterbird species, including the Oriental Stork (Ciconia boyciana, Endangered), Red-crowned Crane (Grus japonensis, Endangered), White-naped Crane (Antigone vipio, Vulnerable), and Whooper Swan (Cygnus cygnus, Least Concern; however, it is protected under Chinese national legislation). Conservation status follows the IUCN Red List (version 2023-1; https://www.iucnredlist.org/). All species mentioned in the text have been checked against the latest IUCN categories. Historically known as “Bei Da Huang” before large-scale reclamation in the 1950s, the plain was once dominated by pristine natural wetlands. However, after decades of intensive development and exploitation, it has been converted to a major crop production base.

2.2. Data Sources

All datasets were downloaded in 2024. All raster datasets were resampled to 1 km resolution (bilinear for continuous variables, nearest-neighbor interpolation for categorical LULC) and reprojected to the Albers Equal-Area Conic projection (Krasovsky_1940 ellipsoid) (Table 1).

2.3. Methods

2.3.1. Habitat Quality Assessment

We used the Habitat Quality module of the InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) model to simulate and assess the spatial distribution of habitat quality in the Sanjiang Plain for the year 2018. InVEST is an open-source ecosystem service assessment tool widely used in environmental management and ecological conservation projects globally. The Habitat Quality module assesses and quantifies the spatial distribution of habitat quality by integrating land use/land cover (LULC) data and the spatial distribution of threat factors [36]. The evaluation process involves the following steps:
(1) Data input. We provided the model with required data, including a land use/land cover map, threat factor attributes (spatial distribution, influence distance, and influence weight), and habitat suitability data. Industrial and mining land, cultivated land, urban land, rural settlements, saline-alkali land, and bare land were selected as threat factors (Table 2). Suitability scores for each land use type (Table 3) were assigned based on documented habitat preferences of general biodiversity in the Northeast China wetland–agriculture landscape, following the parameterization approaches of previous studies in this region [37,38,39].
(2) Threat Factor Analysis. The threat factor parameters (maximum distance, weight, and decay type) in Table 2, as well as the land use/land cover sensitivity parameters (habitat suitability and sensitivity to each threat) in Table 3, were adopted from previous studies on habitat quality in Northeast China [40,41]. Using these parameters, the Habitat Quality module evaluates the impacts of threat factors on surrounding pixels through distance-decay functions. Exponential decay was applied to point-source threats, such as urban and industrial land, whereas linear decay was used for diffuse threats, such as cropland and saline-alkali land. These calculations were then integrated to generate a comprehensive threat impact index.
Habitat Quality module equations (simplified). The habitat quality value Qxj for pixel x in land use type j was calculated as:
Q x j = H j [ 1 − ( D x j z D x j z + k z ) ]
where Hj is the habitat suitability score of land use type j (Table 3), Dxj is the degradation level (see Supplementary Text S1 for the full equation), k is the half-saturation constant (set to 0.5), and z is the normalization constant (default value = 2.5), following the InVEST user guide [36]. The parameter settings in Table 2 and Table 3 (maximum distance, weight, decay type, and habitat suitability scores) are justified in Supplementary Text S1.
(3) Habitat Quality calculation: The comprehensive threat impact index is combined with habitat suitability values to determine the habitat quality of each pixel [42,43].

2.3.2. Spatial Analysis of Species Richness

Avian diversity is commonly represented by species richness in large-scale studies [44], as alternative metrics require additional data (e.g., abundance, location) and more complex calculations. Avian species richness refers to the number of bird species present [45]. According to the BirdLife range maps, 281 species potentially occur in the study area. Following the classification criteria of the Ramsar Convention on Wetlands, bird species that depend on wetlands or aquatic habitats (freshwater or marine/coastal) for key parts of their life cycle—such as breeding, feeding, roosting, or migratory stopovers—are categorized as waterbirds [46]. Conversely, non-waterbirds, namely land birds, were defined as all species not classified as waterbirds under the Ramsar-based criteria. This group includes forest, shrubland, farmland, and other terrestrial or upland-associated species. Based on this classification, the study area supports 198 land bird species and 83 waterbird species (see Supplementary Table S1 for the complete species list with classification and IUCN Red List categories).
Considering the actual natural ecological environment and the socioeconomic development requirements of the Sanjiang Plain, we overlaid the distribution ranges of all land birds and constructed a spatial distribution pattern of land bird species richness at a 1 km × 1 km resolution. We then applied the Jenks Natural Breaks Classification method to determine optimal classification intervals for the bird richness spatial data. This resulted in 10 distinct classification tiers, enabling detailed categorization of dataset segments to better reflect data characteristics. The same procedure was repeated for waterbirds to produce a waterbird diversity distribution map.

2.3.3. Structural Equation Modeling (SEM)

We used structural equation modeling (SEM) to quantitatively assess the influence of natural and anthropogenic factors on habitat quality and avian richness. SEM combines regression analysis, path analysis, and factor analysis, enabling the simultaneous analysis of complex causal relationships between latent variables and observed variables [47]. This method is effective in ecology for understanding how multiple interacting variables influence a given ecological phenomenon [48].
Habitat quality and avian richness are typically influenced by both natural and anthropogenic factors [49]. Considering the actual natural ecological environment and the socioeconomic development requirements of the Sanjiang Plain, we selected six natural factors and four socioeconomic factors as predictors of habitat quality. These included elevation, slope, annual precipitation, annual average temperature, NDVI, soil organic carbon content, distance to roads, population density, GDP, and nighttime light. Based on these factors, we developed a conceptual model to illustrate the spatial and temporal drivers of habitat quality and avian richness (Figure 2) [50].
First, we tested the normality of the 10 explanatory variables and the three response variables (habitat quality, waterbird diversity and land bird diversity). Variables that did not follow a normal distribution were transformed using Johnson transformation to meet the normality assumption. Additionally, all 13 variables were standardized to eliminate the influence of different measurement units. Next, based on the conceptual model (Figure 2), we formulated the SEM path model using the lavaan package in R. We then evaluated the model fit using four indices: the comparative fit index (CFI), the goodness-of-fit index (GFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). Generally, a well-fitting model is indicated by CFI and GFI values > 0.90, and RMSEA and SRMR values approaching 0 (RMSEA < 0.06 and SRMR < 0.05) [51]. Finally, we iteratively refined the model by removing non-significant paths (p > 0.05) and, guided by modification indices and ecological interpretability, adding theoretically supported paths, until we obtained the best-fitting model.
The simulation results of the final models are presented in the form of a path diagram, where each path is associated with a corresponding standardized regression coefficient, representing the strength of the causal effect between variables. A larger coefficient indicates a stronger influence. If a variable directly connects to a response variable, it represents a direct effect, with the coefficient on the path indicating its magnitude. If a variable does not directly connect to a response variable, it represents an indirect effect, where the magnitude is determined by the product of the path coefficients along all connecting paths. The total effect of each variable on the response variable is the sum of its direct and indirect effects.
The causal directions in the conceptual model (Figure 2) were specified a priori based on established ecological mechanisms and previous studies in this region [17,52]. Specifically, temperature influences soil organic carbon (SOC) through microbial decomposition; SOC promotes vegetation productivity (NDVI); and human activities (e.g., agricultural reclamation) reduce SOC storage. Topography (elevation and slope) was treated as an exogenous antecedent variable due to its static nature.

3. Results

3.1. Spatial Patterns of Avian Diversity and Habitat Quality

A total of 281 bird species, belonging to 16 orders and 67 families, were recorded in the Sanjiang Plain (Figure 3). Avian species richness ranged from 1 to 244 and exhibited pronounced spatial heterogeneity, broadly consistent with the patterns of habitat quality. Higher species richness was observed in the western, southern, and eastern regions adjacent to the Lesser Khingan and Changbai Mountains, whereas lower richness occurred in the central regions.
Specifically, a total of 83 waterbird species, representing 7 orders and 17 families, were recorded. Species richness ranged from 0 to 68, with over 81% of the region supporting more than 60 species. Hotspots were concentrated in the eastern and central lowland areas where wetlands are distributed. Meanwhile, 198 land bird species from 17 orders and 51 families were recorded, with species richness ranging from 0 to 127. Land bird richness was highest in forest-dominated regions along the western and southern margins.
In 2018, the mean habitat quality index across the Sanjiang Plain was quantified at 0.66, indicating a moderately high level of ecological integrity. In terms of the composition of habitat quality classes (Figure 4), the highest and high-quality areas made up 32.58% and 10.12%, respectively, while medium-quality areas accounted for 4.69%. Low- and very-low-quality areas represented 31.63% and 0.97%, respectively.
Spatially, habitat quality varied greatly across the study area. High- and very-high-quality habitats were primarily located in the western, eastern, and southern regions. These areas were dominated by forests and natural wetlands with relatively low levels of human disturbance. In contrast, low- and very-low-quality habitats were mainly distributed in the central and northern regions, where agricultural land and urban areas are concentrated.

3.2. Drivers of Habitat Quality

3.2.1. Waterbirds

The structural equation model (SEM) for habitat quality demonstrated strong model fit, with a goodness-of-fit index (GFI) of 0.986, comparative fit index (CFI) of 0.970, standardized root mean square residual (SRMR) of 0.038, and root mean square error of approximation (RMSEA) of 0.051 (Table 4). These indices collectively indicate that the model provides a robust fit to the data and effectively captures the key factors influencing habitat quality.
The structural equation model for habitat quality (Figure 5) showed that SOC and NDVI were the strongest drivers of habitat quality. SOC had the largest total effect (β = 0.375), including a direct effect of 0.31 and an indirect effect of 0.0646 through NDVI, while NDVI also showed a strong positive effect with a total effect coefficient of 0.34. Topographic factors followed with moderate effects, with slope showing a total effect of 0.305 (direct = 0.23; indirect = 0.075 via NDVI and human activities) and elevation showing a total effect of 0.291 (direct = 0.08; indirect effects via NDVI = 0.197, human activities = 0.0012, and SOC = 0.012). Finally, climatic factors showed only weak effects on habitat quality.
By contrast, human activities had a negative total effect on habitat quality (β = −0.1112), including a direct effect of −0.04 and indirect effects of −0.034 and −0.0372 through NDVI and SOC, respectively.

3.2.2. Land Birds

The structural equation model (SEM) for habitat quality demonstrated a strong model fit, with a goodness-of-fit index (GFI) of 0.989, comparative fit index (CFI) of 0.980, standardized root mean square residual (SRMR) of 0.040, and root mean square error of approximation (RMSEA) of 0.060 (Figure 6). These fit indices indicate that the model adequately represents the observed data and effectively captures the key drivers of habitat quality.
The structural equation model for habitat quality (Figure 6) showed that NDVI and SOC were the dominant drivers of habitat quality. NDVI had the largest direct effect (β = 0.327), while SOC also exerted a strong positive direct effect (β = 0.301). In addition, SOC had a positive indirect effect via NDVI (0.061 × 0.327 ≈ 0.020), resulting in a total effect of approximately 0.321.
Topographic factors showed moderate influences on habitat quality. Slope had a relatively strong direct effect (β = 0.227) and an additional indirect effect via NDVI (0.224 × 0.327 ≈ 0.073), yielding a total effect of approximately 0.300. Elevation exhibited a weaker direct effect (β = 0.075) but a substantial indirect effect through NDVI (0.582 × 0.327 ≈ 0.190), leading to a total effect of approximately 0.265.
Climatic factors showed relatively weak positive effects, with temperature (β = 0.044) and precipitation (β = 0.030) contributing only modestly.
By contrast, human activities had a clear negative effect on habitat quality, with a direct effect of −0.074 and an indirect effect via NDVI (−0.144 × 0.327 ≈ −0.047), resulting in a total effect of approximately −0.121.

3.3. Drivers of Avian Diversity

3.3.1. Waterbirds

The structural equation model (SEM) for waterbird richness exhibited strong model fit, as evidenced by the following goodness-of-fit indices: GFI = 0.988, CFI = 0.980, SRMR = 0.029, and RMSEA = 0.050 (Table 4). These values suggest that the model fits the data well and reliably represents the underlying relationships influencing waterbird richness.
The structural equation model for waterbird diversity (Figure 5) showed that several variables had positive effects on waterbird richness, ranked in decreasing order as SOC, NDVI, temperature, and habitat quality. SOC had the strongest positive effect (β = 0.32), including a direct effect of 0.24 and an indirect effect of 0.08 through habitat quality and NDVI. NDVI also showed a strong positive effect (β = 0.31). Temperature had a total effect of 0.2325, including a direct effect of 0.37 and indirect effects of 0.01 through habitat quality and −0.144 through SOC. Habitat quality showed a weaker positive effect, with a total effect coefficient of 0.13.
In contrast, several variables showed negative effects on waterbird richness, ranked from strongest to weakest as elevation, distance to road, human activities, and precipitation. Elevation had the strongest negative effect (β = −0.16), including a direct effect of −0.17 and indirect positive effects through habitat quality, SOC, and human activities. Distance to road had a total effect of −0.10. Human activities had a total effect of −0.059, including a direct effect of −0.03 and an indirect effect of −0.029 through SOC. The effect of precipitation was negligible, with a total effect coefficient of −0.02.

3.3.2. Land Birds

The structural equation model (SEM) for land bird richness demonstrated a good overall fit, with goodness-of-fit indices as follows: GFI = 0.989, CFI = 0.980, SRMR = 0.050, and RMSEA = 0.040 (Table 4). These values indicate that the model adequately represents the relationships among factors influencing land bird richness.
The structural equation model for land bird diversity showed that several variables had positive effects on bird richness, ranked in decreasing order as precipitation, temperature, NDVI, habitat quality, and SOC. Precipitation had the strongest positive effect (β = 0.126), including a direct effect of 0.136 and an indirect effect of −0.01. Temperature followed with a total effect of 0.089 (direct = 0.099; indirect = −0.01). NDVI had a total effect of 0.046, including a direct effect of 0.036 and an indirect effect of 0.01. Habitat quality and SOC showed similar effects (β = 0.035), with habitat quality mainly driven by its direct effect (0.034; indirect = 0.001) and SOC including a direct effect of 0.036 and a slight negative indirect effect of −0.001.
Negative effects on land bird richness were weaker overall, with variables ranked from strongest to weakest as distance to road, human activity, slope, and elevation. Distance to roads had the strongest negative effect (β = −0.058), primarily through its direct effect (−0.06) with a small indirect effect (0.002). Human activities had a total effect of −0.025 (direct = −0.015; indirect = −0.01). Slope and elevation both showed weak negative effects (β = −0.017), with slope including a direct effect of −0.027 and an indirect effect of 0.01, and elevation including a direct effect of −0.037 and an indirect effect of 0.02.

4. Discussion

4.1. Drivers of Habitat Quality

In the Sanjiang Plain, habitat quality is influenced by similar drivers for both waterbirds and land birds. Habitat quality is mainly controlled by natural environmental factors, among which soil properties and vegetation conditions play the most prominent roles, followed by topography. Climatic factors exert only a very weak positive influence, whereas human activities have a similarly weak negative effect. Notably, all of the above factors are statistically significant.

4.1.1. Soil Organic Carbon (SOC) and NDVI

Soil organic carbon (SOC) and NDVI emerge as the strongest determinants of habitat quality. SOC enhances soil structure, water retention, and nutrient availability, thereby supporting stable habitat conditions across regions [17,18,53,54]. NDVI, as an integrated proxy of vegetation productivity and biomass, reflects both food availability and habitat concealment, and is consistently associated with higher habitat quality at both community and species levels [55,56].
SOC and NDVI are tightly coupled and jointly contribute to habitat quality. High SOC promotes vegetation growth, while dense and healthy vegetation enhances SOC accumulation through litter input and root processes. This soil–vegetation coupling underpins habitat quality by linking nutrient cycling with aboveground productivity, thereby shaping food resources and habitat structure [52,57,58].

4.1.2. Topographic Factors

Topographic effects are moderate in both waterbird and land bird models, with predominantly indirect pathways. Topography influences habitat quality by regulating hydrological processes, soil properties, and vegetation distribution. In low-relief regions such as the Yellow River Delta and the Songnen Plain, microtopography shapes vegetation patterns through its influence on soil moisture and salinity, thereby generating spatial heterogeneity in habitat quality [59,60]. Topography also shapes disturbance patterns, with low, flat areas being more intensively used and therefore potentially associated with lower habitat quality [61].
Notably, the direct effects of elevation are relatively small (0.075 for land birds; 0.036 for waterbirds), suggesting that their influence is primarily mediated through indirect pathways—via NDVI for land birds and via SOC and human activities for waterbirds. This pattern may be related to the limited elevation range of the Sanjiang Plain (50–70 m), where weak vertical gradients likely reduce the direct role of elevation, while topography instead operates through human disturbance, soil processes, and hydrological connectivity.

4.1.3. Climatic Factors

Climatic factors (temperature and precipitation) show very weak positive effects on habitat quality (temperature: β = 0.044; precipitation: β = 0.030). This likely reflects the semi-humid climate of the region, where hydrothermal conditions are generally not limiting. A similar pattern has been reported in Heilongjiang Province, where habitat quality is significantly associated with elevation and net primary productivity, whereas mean annual precipitation shows no significant effect [41]. In contrast, climate plays a more important role in environmentally constrained regions, such as warm deserts, boreal forests, and the Qinghai–Tibet Plateau [62,63], where hydrothermal limitations strongly regulate ecosystem structure.

4.1.4. Human Activities

Human activities exert weak negative effects on habitat quality for both waterbirds and land birds, with total effects of −0.1112 and −0.121, respectively. The magnitude of these effects is strongly dependent on the intensity of human disturbance, consistent with studies showing that natural factors often dominate at large spatial scales [39,64,65,66]. In contrast, in highly disturbed regions, anthropogenic drivers can become dominant [10,12].
However, the negative impacts of human activities appear to have been partially mitigated in recent years in the study area. Wetland loss rates declined markedly from 209.40 km2 yr−1 during 2000–2010 to 89.91 km2 yr−1 during 2010–2015. Meanwhile, mean patch size exhibited a trend of initial decline followed by recovery, indicating reduced landscape fragmentation and increasing habitat connectivity [17,67]. These landscape changes suggest that the legacy effects of earlier reclamation have been partially buffered through improved habitat connectivity, which may have contributed to the relatively weak negative effects of human activities observed in this context.

4.2. Drivers of Avian Biodiversity

4.2.1. Pathway Structure

A key contribution of this study lies in disentangling the direct and indirect pathways through which environmental drivers influence avian diversity. Structural equation models reveal that many variables exert their effects not only directly on species richness but also indirectly via habitat quality and vegetation conditions.
Waterbird diversity is primarily associated with multi-step indirect cascades, e.g., temperature–SOC–habitat quality–waterbird richness. Previous studies have suggested that temperature may influence SOC dynamics through decomposition processes, while SOC is closely associated with vegetation productivity and benthic invertebrate availability, which can subsequently affect waterbird communities [52,57]. Along this pathway, indirect effects are stronger than direct effects. Aarif et al. (2025) similarly showed that organic carbon boosts shorebird populations via primary productivity [68]. In addition to these natural cascades, human activities also participate in the pathway network by indirectly modifying SOC and vegetation conditions through land-use change and hydrological alteration, although their direct effects on waterbird richness remain relatively weak. Thus, waterbirds strongly depend on habitat quality, and degradation of habitat conditions (e.g., altered hydrology and soil organic carbon loss) may undermine their conservation.
For land birds, precipitation and temperature exert strong direct effects, whereas indirect effects are weak or negative, suggesting a rapid response to climatic conditions with limited trophic or habitat-mediated cascades. NDVI and SOC affect land birds mainly indirectly and weakly, indicating that those factors are far less limiting than for waterbirds [69]. Road distance acts as the strongest negative factor for land birds, reflecting high sensitivity to ground-level linear disturbances [70]. Human activities, particularly road networks and land-use intensity, exert stronger direct effects on land bird diversity than on waterbirds.

4.2.2. Weaker Anthropogenic than Natural Effects

Building on the integrated pathway network described above, we further assessed the relative importance of anthropogenic and natural drivers within the structural equation framework. Habitat quality consistently mediates many of these effects across both groups.
Across both waterbird and land bird models, natural environmental factors consistently exert stronger effects than anthropogenic influences. Temperature, NDVI, and SOC generally make positive contributions to habitat quality and avian diversity, whereas anthropogenic variables such as road distance and land-use intensity exhibit comparatively weaker and predominantly negative effects. Although human activities participate in the pathway network by indirectly modifying SOC and vegetation conditions, their effects remain secondary relative to natural gradients.
This dominance of natural drivers is evident not only in avian diversity responses but also in habitat quality itself, indicating that baseline ecological conditions governed by soil, vegetation, and climate processes constitute the primary structuring forces in this large wetland landscape. Similar patterns have been reported in other large-scale ecosystems, where natural gradients override anthropogenic effects in structuring biodiversity patterns [39,41,64]. In contrast, anthropogenic impacts often act as secondary modifiers that reshape, rather than replace, underlying environmental controls [71].
The relatively weak current signal of human disturbance in the Sanjiang Plain can be interpreted in the context of its long-term land-use trajectory. Since the 1950s, large-scale agricultural reclamation has substantially transformed the original wetland-dominated landscape, whereas more recent decades have seen the implementation of ecological restoration and land-use optimization, which have partly mitigated historical pressures. As a result, present-day anthropogenic effects are weakened at the regional scale due to both recovery processes and strong spatial heterogeneity of disturbance intensity. Human disturbance is primarily concentrated in central and northern agricultural zones, while forest–wetland mosaics in the western, eastern, and southern regions function as refugia for bird communities, thereby buffering landscape-scale anthropogenic effects. Such spatial configuration leads to context-dependent responses, consistent with global biodiversity patterns [71].
Notably, these findings may imply that anthropogenic effects in the region are not necessarily static. Given ongoing wetland restoration, improved landscape connectivity, and shifts in land-use practices across the Sanjiang Plain, the influence of human activities could potentially transition from predominantly negative impacts toward neutral or even locally positive effects on avian diversity, particularly where restoration enhances habitat availability and ecological connectivity [72,73].

4.2.3. The Effects of Natural Environmental Drivers

To better understand this dominance of natural drivers, we further examined the effects of individual environmental variables on avian diversity and habitat quality. SOC is the primary determinant of waterbird diversity, whereas temperature and precipitation are more influential for land birds. Topographic effects are generally weak in both waterbirds and terrestrial birds, with elevation showing no significant influence in either group. This is likely due to the relatively flat topography of the study area, which limits spatial variation in key environmental variables such as SOC and NDVI and consequently reduces their effects on bird diversity.
Soil Organic Carbon
Soil organic carbon (SOC) is a major positive driver of habitat quality (total effect = 0.375) and plays an important role in shaping bird diversity, with pronounced differences between waterbirds and land birds.
For waterbirds, SOC shows the strongest positive effect among the variables considered (total effect = 0.32), aligning with evidence suggesting positive associations between SOC and shorebird populations [68]. The high SOC storage in the Sanjiang Plain, driven by long-term wetland–farmland interactions, may therefore help explain the high diversity of waterbirds. In contrast, SOC has a weak effect on land birds (0.035), mainly mediated through NDVI and habitat quality. This likely reflects weaker coupling between soil carbon and aboveground trophic pathways, as well as greater ecological flexibility of land birds [69].
NDVI
We identified NDVI as one of the most important positive drivers of habitat quality in the Sanjiang Plain (total effect = 0.340). Further analysis indicates that NDVI also significantly promotes bird diversity, with pronounced differences between waterbirds and land birds in both effect strength and pathways.
For waterbirds, NDVI shows a strong positive effect on species richness (total effect = 0.31), largely mediated through vegetation productivity and trophic resource availability. Higher NDVI supports vegetation productivity and trophic resources, enhancing food availability for waterbirds [74,75], consistent with patterns observed in the Xianghai Wetland in northeast China [76,77]. For land birds, the effect of NDVI is much weaker (total effect = 0.046) and mainly indirect via habitat quality, likely reflecting nonlinear and context-dependent responses. For example, NDVI exhibits threshold effects in temperate woodlands and explains limited variation across urban and tropical systems, indicating low transferability across ecosystems [78,79,80].
Precipitation
Precipitation and temperature both exert very weak effects on habitat quality. Their influences on waterbird and land bird diversity differ slightly. For waterbirds, precipitation has a negligible effect on species richness, likely because water availability is unlikely to be limiting given the extensive aquatic habitats in the Sanjiang Plain (>10% natural wetlands and ~24% paddy fields in 2018, a relatively wet year) [81]. In contrast, precipitation has a significant positive effect on land bird richness (total effect = 0.126), primarily through direct pathways. Increased precipitation enhances vegetation productivity and structural complexity, thereby providing more food resources and nesting sites [82,83].
Furthermore, although direct climatic effects are weak in the SEM, the anomalously high precipitation in 2018 may have indirectly influenced habitat quality by enhancing vegetation growth and soil moisture, effects that are partially captured through NDVI and SOC.
Temperature
Temperature has a weak effect on habitat quality (total effect < 0.05) but exerts a stronger, more direct influence on bird diversity, with clear differences between waterbirds and land birds.
For waterbirds, temperature shows a relatively strong total effect on richness (0.2325), primarily through direct pathways, with weaker indirect effects via SOC and habitat quality [52]. This likely reflects favorable thermal conditions during migration periods (5–12 °C), while higher temperatures may accelerate SOC decomposition, partly offsetting indirect positive effects [84].
For land birds, the effect is weaker (0.089), with a small direct effect (0.099) and a negligible indirect effect (−0.01). This lower sensitivity likely reflects canopy buffering of thermal extremes and greater ecological flexibility, which reduces dependence on habitat-mediated thermal effects [69,85]. Nonetheless, land birds can still be highly vulnerable to warming in certain systems, such as tropical montane forests [86], highlighting context-dependent responses.

4.3. Limitations and Uncertainties

Several limitations should be noted. First, some important drivers, such as fine-scale hydrological conditions and land management practices, were not included due to data constraints. Incorporating these factors in future studies would improve mechanistic interpretation, as biodiversity patterns are shaped by multiple interacting environmental and anthropogenic processes [71,87]. Second, our analysis was based on single-year data (2018), which likely reflects typical but not exhaustive environmental conditions during a relatively wet year. Incorporating multi-year data would help capture interannual variability in climate and bird communities and strengthen the robustness of the observed relationships [88]. Third, we could not account for seasonal dynamics due to data limitations, which may obscure seasonal variation in the distribution and habitat use of both waterbirds and land birds [89,90]. Fourth, all variables were standardized to a 1 km resolution, which may smooth fine-scale habitat heterogeneity (e.g., microtopography and vegetation structure) and introduce scale-related uncertainty [91,92].
Additionally, the InVEST habitat quality index used in this study was developed to represent general biodiversity and does not distinguish between waterbirds and land birds. Although we used SEMs to compare group-specific responses, we acknowledge that developing separate habitat quality indices tailored to different bird guilds (e.g., assigning different suitability scores to forests, wetlands, and water bodies) could improve ecological realism and model accuracy. This represents an important direction for future research.

4.4. Implications

These findings have important implications for the long-term sustainability of wetland ecosystems in the Sanjiang Plain, particularly in balancing biodiversity conservation, ecological restoration, and sustainable agricultural development.
Based on spatial patterns of habitat quality and bird diversity, we suggest the following: Priority conservation zones should include western and southern forests (land bird hotspots) and eastern and central wetlands (waterbird hotspots). Central and northern agricultural areas, which have low habitat quality, should be prioritized for restoration. In addition, roadside buffer strips (50–100 m) should be established along roads crossing forests and wetlands to mitigate the strongest negative effect of roads on land birds.
Management should be group-specific. For waterbirds, which are more strongly influenced by indirect climate- and habitat-mediated pathways, management should go beyond simple wetland expansion. Priority actions include protecting soil carbon stocks, maintaining hydrological refugia, and implementing adaptive water-level regulation. These measures are necessary because temperature and SOC influence waterbird diversity through complex indirect pathways mediated by habitat quality. For land birds, which are more directly associated with NDVI and habitat quality and less sensitive to climate variability, direct habitat-based management may be more effective. Priority actions include restoring vegetation in areas with declining NDVI and building ecological corridors to enhance connectivity in road-dense landscapes.

5. Conclusions

This study quantified the direct and indirect effects of natural and anthropogenic drivers on habitat quality and avian diversity in the Sanjiang Plain. The results show that habitat quality is primarily controlled by natural factors, especially soil organic carbon (SOC) and NDVI, while human activities exert weak negative effects. Waterbird diversity is mainly driven by SOC- and temperature-mediated indirect pathways, whereas land bird diversity responds more directly to climate and human disturbance (e.g., roads). Natural drivers consistently outweigh anthropogenic influences for both bird groups. SOC and NDVI are more important for waterbirds, while precipitation and temperature are more influential for land birds. We therefore recommend group-specific conservation: protecting soil carbon stocks and hydrological refugia for waterbirds, and restoring vegetation and mitigating road impacts for land birds. Priority zones include western/southern forests (land birds) and eastern/central wetlands (waterbirds).
Overall, this study provides a scientific basis for sustainable wetland management and ecological restoration in large-scale wetland landscapes. The identified driver pathways can inform future biodiversity conservation planning, improve ecosystem resilience, and support sustainable regional development under ongoing climate and land-use change.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/su18125887/s1.

Author Contributions

Y.L. (Yuehui Li) conceived and designed the study, provided supervision, and revised the manuscript. X.S. conceived the study, performed data collection and analysis, and wrote the first draft of the manuscript. C.L. prepared the figures and revised the manuscript. Y.L. (Yue Li) revised the manuscript. Y.L. (Yueyuan Li) performed data validation. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Key Research and Development Program of China (2022YFF1300904).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors on request.

Acknowledgments

We would like to express our gratitude for the publicly available datasets that were accessed online, which were crucial for the completion of this study. We also acknowledge the platforms and resources that made this data accessible.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Parajuli, K.; Ghimire, P. Farmland Birds and their Importance in Agro ecosystem: A Review. Int. Res. J. MMC 2024, 5, 158–168. [Google Scholar] [CrossRef]
  2. Sekercioglu, C.H. Increasing awareness of avian ecological function. Trends Ecol. Evol. 2006, 21, 464–471. [Google Scholar] [CrossRef]
  3. Qiu, J.; Zhang, Y.; Ma, J. Wetland habitats supporting waterbird diversity: Conservation perspective on biodiversity-ecosystem functioning relationship. J. Environ. Manag. 2024, 357, 120663. [Google Scholar] [CrossRef]
  4. Mao, D.; Tian, Y.; Wang, Z.; Jia, M.; Du, J.; Song, C. Wetland changes in the Amur River Basin: Differing trends and proximate causes on the Chinese and Russian sides. J. Environ. Manag. 2021, 280, 111670. [Google Scholar] [CrossRef] [PubMed]
  5. Zhang, H.; Zhang, C.; Hu, T.; Zhang, M.; Ren, X.; Hou, L. Exploration of roadway factors and habitat quality using InVEST. Transp. Res. Part D Transp. Environ. 2020, 87, 102551. [Google Scholar] [CrossRef]
  6. Rahimi, L.; Malekmohammadi, B.; Yavari, A.R. Assessing and Modeling the Impacts of Wetland Land Cover Changes on Water Provision and Habitat Quality Ecosystem Services. Nat. Resour. Res. 2020, 29, 3701–3718. [Google Scholar] [CrossRef]
  7. Wang, C.; Xia, S.; Duan, H.; Wu, Y.; Su, Y.; Wang, B.; Shen, Z.; Sun, G.; Dong, B. Study on the impact of reclamation and development of Yancheng coastal wetlands on the spatio-temporal evolution of wintering Anatidae habitat. Ecol. Indic. 2024, 160, 111892. [Google Scholar] [CrossRef]
  8. Song, Z.; Gao, S.; Leng, M.; Zhou, B.; Wu, B.; Song, Z.; Gao, S.; Leng, M.; Zhou, B.; Wu, B. Quantifying the Ecological Performance of Migratory Bird Conservation: Evidence from Poyang Lake Wetlands in China. Biology 2024, 13, 786. [Google Scholar] [CrossRef]
  9. Zhu, C.; Zhang, X.; Zhou, M.; He, S.; Gan, M.; Yang, L.; Wang, K. Impacts of urbanization and landscape pattern on habitat quality using OLS and GWR models in Hangzhou, China. Ecol. Indic. 2020, 117, 106654. [Google Scholar] [CrossRef]
  10. Bai, L.; Xiu, C.; Feng, X.; Liu, D. Influence of urbanization on regional habitat quality:a case study of Changchun City. Habitat Int. 2019, 93, 102042. [Google Scholar] [CrossRef]
  11. Wu, H.; Fang, S.; Yang, Y.; Cheng, J. Changes in habitat quality of nature reserves in depopulating areas due to anthropogenic pressure: Evidence from Northeast China, 2000–2018. Ecol. Indic. 2022, 138, 108844. [Google Scholar] [CrossRef]
  12. Zhao, Y.; Qu, Z.; Zhang, Y.; Ao, Y.; Han, L.; Kang, S.; Sun, Y. Effects of human activity intensity on habitat quality based on nighttime light remote sensing: A case study of Northern Shaanxi, China. Sci. Total Environ. 2022, 851, 158037. [Google Scholar] [CrossRef]
  13. Hodapp, D.; Meier, S.; Muijsers, F.; Badewien, T.H.; Hillebrand, H. Structural equation modeling approach to the diversity-productivity relationship of Wadden Sea phytoplankton. Mar. Ecol. Prog. Ser. 2015, 523, 31–40. [Google Scholar] [CrossRef]
  14. Zerbe, S.; Maurer, U.; Schmitz, S.; Sukopp, H. Biodiversity in Berlin and its potential for nature conservation. Landsc. Urban Plan. 2003, 62, 139–148. [Google Scholar] [CrossRef]
  15. Kapitza, S.; Van Ha, P.; Kompas, T.; Golding, N.; Cadenhead, N.C.R.; Bal, P.; Wintle, B.A.; Kapitza, S.; Van Ha, P.; Kompas, T.; et al. Assessing biophysical and socio-economic impacts of climate change on regional avian biodiversity. Sci. Rep. 2021, 11, 3304. [Google Scholar] [CrossRef] [PubMed]
  16. Roland, C.A.; Sadoti, G.; Nicklen, E.F.; McAfee, S.A.; Stehn, S.E. A structural equation model linking past and present plant diversity in Alaska: A framework for evaluating future change. Ecosphere 2019, 10, e02832. [Google Scholar] [CrossRef]
  17. Xiang, H.; Wang, Z.; Mao, D.; Zhang, J.; Xi, Y.; Du, B.; Zhang, B. What did China’s National Wetland Conservation Program Achieve? Observations of changes in land cover and ecosystem services in the Sanjiang Plain. J. Environ. Manag. 2020, 267, 110623. [Google Scholar] [CrossRef] [PubMed]
  18. Hines, C.; Duval, L.; Watts, B.; Mitchell, M. The Role of Patch Age and Agricultural Legacy on Carbon Sequestration and Bird Habitat in Salt Marshes. Authorea Prepr. 2024. [Google Scholar] [CrossRef]
  19. Sanderson, F.J.; Donald, P.F.; Pain, D.J.; Burfield, I.J.; van Bommel, F.P.J. Long-term population declines in Afro-Palearctic migrant birds. Biol. Conserv. 2006, 131, 93–105. [Google Scholar] [CrossRef]
  20. Benton, T.G.; Vickery, J.A.; Wilson, J.D. Farmland biodiversity: Is habitat heterogeneity the key? Trends Ecol. Evol. 2003, 18, 182–188. [Google Scholar] [CrossRef]
  21. Cao, Y.; Wang, C.; Su, Y.; Duan, H.; Wu, X.; Lu, R.; Su, Q.; Wu, Y.; Chu, Z.; Cao, Y.; et al. Study on Spatiotemporal Evolution and Driving Forces of Habitat Quality in the Basin along the Yangtze River in Anhui Province Based on InVEST Model. Land 2023, 12, 1092. [Google Scholar] [CrossRef]
  22. Li, X.; Anderson, C.J.; Wang, Y.; Lei, G. Waterbird diversity and abundance in response to variations in climate in the Liaohe Estuary, China. Ecol. Indic. 2021, 132, 108286. [Google Scholar] [CrossRef]
  23. Ónodi, G.; Czeglédi, I.; Erős, T. Drivers of the taxonomic and functional structuring of aquatic and terrestrial floodplain bird communities. Landsc. Ecol. 2024, 39, 174. [Google Scholar] [CrossRef] [PubMed]
  24. Freedman, A.H.; Buermann, W.; Mitchard, E.T.A.; Defries, R.S.; Smith, T.B. Human Impacts Flatten Rainforest-Savanna Gradient and Reduce Adaptive Diversity in a Rainforest Bird. PLoS ONE 2010, 5, e13088. [Google Scholar] [CrossRef]
  25. Poiani, K.A.; Richter, B.D.; Anderson, M.G.; Richter, H.E. Biodiversity Conservation at Multiple Scales: Functional Sites, Landscapes, and Networks. BioScience 2000, 50, 133–146. [Google Scholar] [CrossRef]
  26. Jin, H.; Van Leeuwen, C.H.A.; Bakker, E.S. Restoring land–water transition areas to stimulate food web development is mediated by the hydrological connectivity. Ecol. Indic. 2024, 167, 112615. [Google Scholar] [CrossRef]
  27. Büscher, M.; Lange, F.; Bröckel, M.; Höfer, S.; Stemberg, E.; Folsche, E.; Eylering, A.; Fiebelkorn, F. Quantifying behavioural impact and plasticity via expert and population surveys to prioritise bird conservation behaviour. Ecosyst. People 2025, 21, 2478555. [Google Scholar] [CrossRef]
  28. Cazalis, V.; Princé, K.; Mihoub, J.-B.; Kelly, J.; Butchart, S.H.M.; Rodrigues, A.S.L. Effectiveness of protected areas in conserving tropical forest birds. Nat. Commun. 2020, 11, 4461. [Google Scholar] [CrossRef]
  29. Weller, M.W. Wetland Birds: Habitat Resources and Conservation Implications; Cambridge University Press: Cambridge, UK, 1999. [Google Scholar]
  30. Zedler, J.; Kercher, S. Wetland resources: Status, trends, ecosystem services, and restorability. Annu. Rev. Environ. Resour. 2005, 30, 39–74. [Google Scholar] [CrossRef]
  31. Richardson, J.S.; Sato, T. Resource subsidy flows across freshwater–terrestrial boundaries and influence on processes linking adjacent ecosystems. Ecohydrology 2015, 8, 406–415. [Google Scholar] [CrossRef]
  32. Nakano, S.; Murakami, M. Reciprocal subsidies: Dynamic interdependence between terrestrial and aquatic food webs. Proc. Natl. Acad. Sci. USA 2001, 98, 166–170. [Google Scholar] [CrossRef]
  33. Peng, S. 1-km Monthly Mean Temperature Dataset for China (1901–2022); National Tibetan Plateau/Third Pole Environment Data Center: Beijing, China, 2020. [Google Scholar] [CrossRef]
  34. Peng, S. 1-km monthly precipitation dataset for China (1901–2022). Earth Syst. Sci. Data 2019, 11, 1931–1946. [Google Scholar] [CrossRef]
  35. Wu, Y.; Shi, K.; Chen, Z.; Liu, S.; Chang, Z. An Improved Time-Series DMSP-OLS-like Data (1992–2023) in China by Integrating DMSP-OLS and SNPP-VIIRS; Harvard Dataverse: Cambridge, MA, USA, 2021. [Google Scholar] [CrossRef]
  36. Broquet, M.; Campos, F.S.; Cabral, P.; David, J. Habitat quality on the edge of anthropogenic pressures: Predicting the impact of land use changes in the Brazilian Upper Paraguay river Basin. J. Clean. Prod. 2024, 459, 142546. [Google Scholar] [CrossRef]
  37. Dai, L.; Li, S.; Lewis, B.J.; Wu, J.; Yu, D.; Zhou, W.; Zhou, L.; Wu, S. The influence of land use change on the spatial–temporal variability of habitat quality between 1990 and 2010 in Northeast China. J. For. Res. 2018, 30, 2227–2236. [Google Scholar] [CrossRef]
  38. Jin, S.; Liu, X.; Yang, J.; Lv, J.; Gu, Y.; Yan, J.; Yuan, R.; Shi, Y. Spatial-temporal changes of land use/cover change and habitat quality in Sanjiang plain from 1985 to 2017. Front. Environ. Sci. 2022, 10, 1032584. [Google Scholar] [CrossRef]
  39. Sun, X.L.; Li, Y.H.; Hu, Y.M.; Li, Y. Human and natural factors affect habitat quality in ecologically fragile areas: Evidence from Songnen Plain, China. Front. Plant Sci. 2024, 15, 1444163. [Google Scholar] [CrossRef] [PubMed]
  40. Wang, Z.; Zhu, T.; Jiang, Q. Analysis of land use/cover change and habitat quality in Heilongjiang Province. J. Northeast Agric. Univ. 2022, 53, 77–86. [Google Scholar] [CrossRef]
  41. Liang, X.; Yuan, L.; Ning, L. Spatial pattern and influencing factors of habitat quality in Heilongjiang Province based on the InVEST model. J. Beijing Norm. Univ. (Nat. Sci.) 2020, 56, 864–872. [Google Scholar] [CrossRef]
  42. Sallustio, L.; Toni, A.D.; Strollo, A.; Febbraro, M.D.; Gissi, E.; Casella, L.; Geneletti, D.; Munafò, M.; Vizzarri, M.; Marchetti, M. Assessing habitat quality in relation to the spatial distribution of protected areas in Italy. J. Environ. Manag. 2017, 201, 129–137. [Google Scholar] [CrossRef]
  43. Xu, H.; Dong, B.; Gao, X.; Xu, Z.; Ren, C.; Fang, L.; Wei, Z.; Liu, X.; Lu, Z. Habitat quality assessment of wintering migratory birds in Poyang Lake National Nature Reserve based on InVEST model. Environ. Sci. Pollut. Res. 2022, 30, 28847–28862. [Google Scholar] [CrossRef]
  44. Gotelli, N.J.; Colwell, R.K. Quantifying biodiversity: Procedures and pitfalls in the measurement and comparison of species richness. Ecol. Lett. 2001, 4, 379–391. [Google Scholar] [CrossRef]
  45. Mittelbach, G.G.; Steiner, C.F.; Scheiner, S.M.; Gross, K.L.; Reynolds, H.L.; Waide, R.B.; Willig, M.R.; Dodson, S.I.; Gough, L. What Is the Observed Relationship Between Species Richness and Productivity? Ecology 2001, 82, 2381–2396. [Google Scholar] [CrossRef]
  46. Secretariat, R.C. The Ramsar Convention Manual: A guide to the Convention on Wetlands; Ramsar Convention Secretariat: Gland, Switzerland, 2013. [Google Scholar]
  47. Bedane, G.A.; Feyisa, G.L.; Senbeta, F. Spatial distribution of above ground carbon density in Harana Forest, Ethiopia. Ecol. Process. 2022, 11, 4. [Google Scholar] [CrossRef]
  48. Liu, S.; Liao, Q.; Xiao, M.; Zhao, D.; Huang, C. Spatial and Temporal Variations of Habitat Quality and Its Response of Landscape Dynamic in the Three Gorges Reservoir Area, China. Int. J. Environ. Res. Public Health 2022, 19, 3594. [Google Scholar] [CrossRef]
  49. Fritz, K.M.; Schofield, K.A.; Alexander, L.C.; McManus, M.G.; Golden, H.E.; Lane, C.R.; Kepner, W.G.; LeDuc, S.D.; DeMeester, J.E.; Pollard, A.I. Physical and Chemical Connectivity of Streams and Riparian Wetlands to Downstream Waters: A Synthesis. JAWRA J. Am. Water Resour. Assoc. 2018, 54, 323–345. [Google Scholar] [CrossRef]
  50. Hao, J.; Xu, G.; Luo, L.; Zhang, Z.; Yang, H.; Li, H. Quantifying the relative contribution of natural and human factors to vegetation coverage variation in coastal wetlands in China. Catena 2020, 188, 104429. [Google Scholar] [CrossRef]
  51. Maydeu-Olivares, A.; Shi, D.; Rosseel, Y. Assessing Fit in Structural Equation Models: A Monte-Carlo Evaluation of RMSEA Versus SRMR Confidence Intervals and Tests of Close Fit. Struct. Equ. Model. Multidiscip. J. 2018, 25, 1389611. [Google Scholar] [CrossRef]
  52. Mao, D.H.; Wang, Z.M.; Li, L.; Miao, Z.H.; Ma, W.H.; Song, C.C.; Ren, C.Y.; Jia, M.M. Soil organic carbon in the Sanjiang Plain of China: Storage, distribution and controlling factors. Biogeosciences 2015, 12, 1635–1645. [Google Scholar] [CrossRef]
  53. Xu, W.; Xu, Y.; Han, Z.; Yao, J.; Tryjanowski, P.; Wang, H. Disentangling the relative effects of semi-natural habitats within different landscape agricultural intensities on bird communities. Avian Res. 2025, 16, 100228. [Google Scholar] [CrossRef]
  54. Ahmadi Mirghaed, F.; Souri, B. Relationships between habitat quality and ecological properties across Ziarat Basin in northern Iran. Environ. Dev. Sustain. 2021, 23, 16192–16207. [Google Scholar] [CrossRef]
  55. Man, W.; Liu, M.; Wang, Z.; Mao, D.; Tian, Y.; Jia, M.; Li, X.; Ren, C.; Ouyang, L. Dynamics of waterfowl habitat suitability in the Sanjiang Plain ecological function area from 1990 to 2015. Chin. J. Appl. Ecol. 2017, 28, 4083–4091. [Google Scholar] [CrossRef]
  56. Gong, Z.; Su, S.; Du, B.; Guan, H.; Zhang, Q. Selection and dispersal of red-crowned crane breeding habitats in Zhalong Wetland. J. Nat. Resour. 2021, 36, 1964–1975. [Google Scholar] [CrossRef]
  57. Hagger, V.; Wilson, K.; England, J.R.; Dwyer, J.M. Water availability drives aboveground biomass and bird richness in forest restoration plantings to achieve carbon and biodiversity cobenefits. Ecol. Evol. 2019, 9, 14379–14393. [Google Scholar] [CrossRef]
  58. Brüggeshemke, J.; Fartmann, T. Predicting species richness and abundance of breeding birds by remote-sensing and field-survey data. Ecol. Solut. Evid. 2025, 6, e70170. [Google Scholar] [CrossRef]
  59. Li, X.; Hou, X.; Song, Y.; Shan, K.; Zhu, S.; Yu, X.; Mo, X. Assessing Changes of Habitat Quality for Shorebirds in Stopover Sites: A Case Study in Yellow River Delta, China. Wetlands 2019, 39, 67–77. [Google Scholar] [CrossRef]
  60. Yan, S.; Wang, X.; Cai, Y.; Li, C.; Yan, R.; Cui, G.; Yang, Z. An Integrated Investigation of Spatiotemporal Habitat Quality Dynamics and Driving Forces in the Upper Basin of Miyun Reservoir, North China. Sustainability 2018, 10, 4625. [Google Scholar] [CrossRef]
  61. Yang, F.; Wang, Z.; Wang, Y.; An, F.; Zhao, C. The correlation between soil water salinity and plant community distribution under micro-topography in Songnen Plain. Acta Ecol. Sin. 2013, 33, 6202–6208. [Google Scholar] [CrossRef]
  62. van den Bosch, M.; Steel, Z.L.; Costanza, J.K.; Kellner, K.F.; Peek, R.A. Increased drought drives avian community declines in the warm deserts of the United States. Biol. Conserv. 2025, 310, 111393. [Google Scholar] [CrossRef]
  63. Labadie, G.; Boulanger, Y.; Drapeau, P.; Stralberg, D.; Tremblay, J.A. Projecting bird assemblage responses to climate-driven changes in managed boreal forest landscapes of Québec. Biol. Conserv. 2025, 302, 110956. [Google Scholar] [CrossRef]
  64. Chen, S.; Jin, Y.; Huang, Y. Spatio-temporal variations of habitat quality and its underlying mechanism in the central region of Yangtze River Delta. Chin. J. Ecol. 2023, 42, 1175–1185. [Google Scholar]
  65. Kou, J.F.; Liu, J.; Wang, H.L.; Huang, S.; Zhou, Z.; Yang, C.; Huang, X.L. Spatio-temporal evolution in habitat quality and its driving factors in Han River Basin, China. Sci. Rep. 2025, 15, 26139. [Google Scholar] [CrossRef]
  66. Zhao, J.W.; Wang, L.Y.; Jia, D.; Wang, Y.W. Study on the effects of land use transformation on habitat quality and its driving mechanisms: A case study of the Qin-Mang River Basin. Front. Environ. Sci. 2025, 13, 1505178. [Google Scholar] [CrossRef]
  67. He, X.; Jia, M.; Wang, Z.; Ren, C.; Zheng, H.; Guo, Y.; Zhang, X.; Xin, X. Remote Sensing Based Evaluation of Effectiveness of Wetland Protection Project in Sanjiang Plain, China. Bull. Chin. Acad. Sci. 2017, 32, 3–10. [Google Scholar] [CrossRef]
  68. Aarif, K.M.; Nefla, A.; Rubeena, K.A.; Xu, Y.; Bouragaoui, Z.; Nasser, M.; Shifa, C.T.; Athira, T.R.; Jishnu, K.; Anand, J.; et al. Assessing environmental change and population declines of large wading birds in southwestern India. Environ. Sustain. Indic. 2025, 25, 100572. [Google Scholar] [CrossRef]
  69. Luck, G.W.; Lavorel, S.; McIntyre, S.; Lumb, K. Improving the application of vertebrate trait-based frameworks to the study of ecosystem services. J. Anim. Ecol. 2012, 81, 1065–1076. [Google Scholar] [CrossRef] [PubMed]
  70. Reijnen, R.; Foppen, R.; Meeuwsen, H. The effects of traffic on the density of breeding birds in Dutch agricultural grasslands. Biol. Conserv. 1996, 75, 255–260. [Google Scholar] [CrossRef]
  71. Newbold, T.; Hudson, L.N.; Hill, S.L.L.; Contu, S.; Lysenko, I.; Senior, R.A.; Börger, L.; Bennett, D.J.; Choimes, A.; Collen, B.; et al. Global effects of land use on local terrestrial biodiversity. Nature 2015, 520, 45–50. [Google Scholar] [CrossRef]
  72. Ling, Z.; Li, Q.; Jiang, W.; Zhang, Z.; Hou, P.; Huang, S.; Yang, Z.; Xiao, Z.; Yin, X. Multi-guild waterbird habitat suitability change and hotspots (2018–2024) in China’s international wetland cities: MaxEnt + Gi* within a DIKW framework. Ecol. Indic. 2026, 182, 114564. [Google Scholar] [CrossRef]
  73. Panjin Forestry and Wetland Conservation Bureau. Bird Species Diversity in Panjin Wetland Reaches 336; Panjin Municipal People’s Government: Panjin, China, 2026.
  74. Russell, I.A.; Randall, R.M.; Hanekom, N. Spatial and Temporal Patterns of Waterbird Assemblages in the Wilderness Lakes Complex, South Africa. Waterbirds 2014, 37, 1–18. [Google Scholar] [CrossRef]
  75. Henry, D.A.W.; Cumming, G.S. Can waterbirds with different movement, dietary and foraging functional traits occupy similar ecological niches? Landsc. Ecol. 2016, 32, 265–278. [Google Scholar] [CrossRef]
  76. Zhou, J.; Zhou, L.; Xu, W. Diversity of wintering waterbirds enhanced by restoring aquatic vegetation at Shengjin Lake, China. Sci. Total Environ. 2020, 737, 140190. [Google Scholar] [CrossRef]
  77. Guo, J.; He, C.; Jiang, H.; Sheng, L.; Li, L. The study on the relationship between vegetation productivity and waterfowl diversity in Xianghai wetland. J. Northeast Norm. Univ. (Nat. Sci. Ed.) 2012, 44, 120–124. [Google Scholar]
  78. Hunt, M.L.; Blackburn, G.A.; Siriwardena, G.M.; Carrasco, L.; Rowland, C.S. Using satellite data to assess spatial drivers of bird diversity. Remote Sens. Ecol. Conserv. 2023, 9, 483–500. [Google Scholar] [CrossRef]
  79. Macgregor-Fors, I.; Garizábal-Carmona, J.A.; García-Arroyo, M.; Fishel, E.; Nilon, C.H.; Lemoine-Rodríguez, R. Does NDVI explain patterns of urban bird diversity? Insights from temperate to tropical cities. Urban For. Urban Green. 2026, 116, 129211. [Google Scholar] [CrossRef]
  80. Awoyemi, A.G.; Alabi, T.R.; Ibáñez-Álamo, J.D. Remotely sensed spectral indicators of bird taxonomic, functional and phylogenetic diversity across Afrotropical urban and non-urban habitats. Ecol. Indic. 2025, 170, 112966. [Google Scholar] [CrossRef]
  81. Zeng, H.; Xiao, C.; Chen, X.; Ye, D. State of China’s climate in 2018. Atmos. Ocean. Sci. Lett. 2019, 12, 349–354. [Google Scholar] [CrossRef]
  82. Zhang, H.; Xiao, L.; Zhao, Z.; Yin, S.; Zhang, Y. Importance of surrounding landscape features in explaining wintering waterbird diversity and community assembly in Jiangsu’s inland lake wetlands. Avian Res. 2025, 16, 100295. [Google Scholar] [CrossRef]
  83. Wei, Z.; Zhou, L. The impact of earlier flood recession on metacommunity diversity of wintering waterbirds at shallow lakes in the middle and lower Yangtze River floodplain. Avian Res. 2023, 14, 100102. [Google Scholar] [CrossRef]
  84. Gaget, E.; Ovaskainen, O.; Bradter, U.; Haas, F.; Jonas, L.; Johnston, A.; Langendoen, T.; Lehikoinen, A.S.; Pärt, T.; Pavón-Jordán, D.; et al. Changes in waterbird occurrence and abundance at their northern range boundaries in response to climate warming: Importance of site area and protection status. Anim. Conserv. 2025, 28, 490–500. [Google Scholar] [CrossRef]
  85. Lauck, K.S.; Ke, A.; Olimpi, E.M.; Paredes, D.; Hood, K.; Phillips, T.; Anderegg, W.R.L.; Karp, D.S. Agriculture and hot temperatures interactively erode the nest success of habitat generalist birds across the United States. Science 2023, 382, 290–294. [Google Scholar] [CrossRef]
  86. Şekercioğlu, Ç.H.; Primack, R.B.; Wormworth, J. The effects of climate change on tropical birds. Biol. Conserv. 2012, 148, 1–18. [Google Scholar] [CrossRef]
  87. Duflot, R.; Daniel, H.; Aviron, S.; Alignier, A.; Beaujouan, V.; Burel, F.; Cochard, A.; Ernoult, A.; Pain, G.; Pithon, J.A. Adjacent woodlands rather than habitat connectivity influence grassland plant, carabid and bird assemblages in farmland landscapes. Biodivers. Conserv. 2018, 27, 1925–1942. [Google Scholar] [CrossRef]
  88. Herrando, S.; Titeux, N.; Brotons, L.; Anton, M.; Ubach, A.; Villero, D.; García-Barros, E.; Munguira, M.L.; Godinho, C.; Stefanescu, C. Contrasting impacts of precipitation on Mediterranean birds and butterflies. Sci. Rep. 2019, 9, 5680. [Google Scholar] [CrossRef]
  89. La Sorte, F.A.; Fink, D.; Hochachka, W.M.; Farnsworth, A.; Rodewald, A.D.; Rosenberg, K.V.; Sullivan, B.L.; Winkler, D.W.; Wood, C.; Kelling, S. The role of atmospheric conditions in the seasonal dynamics of North American migration flyways. J. Biogeogr. 2014, 41, 1685–1696. [Google Scholar] [CrossRef]
  90. Sullivan, B.L.; Aycrigg, J.L.; Barry, J.H.; Bonney, R.E.; Bruns, N.; Cooper, C.B.; Damoulas, T.; Dhondt, A.A.; Dietterich, T.; Farnsworth, A.; et al. The eBird enterprise: An integrated approach to development and application of citizen science. Biol. Conserv. 2014, 169, 31–40. [Google Scholar] [CrossRef]
  91. Wiens, J.A. Spatial scaling in ecology. Funct. Ecol. 1989, 3, 385–397. [Google Scholar] [CrossRef]
  92. Levin, S.A. The Problem of Pattern and Scale in Ecology; Springer: New York, NY, USA, 1992; pp. 277–326. [Google Scholar]
Sustainability 18 05887 g001
Figure 1. The location and digital elevation model (DEM) of the study area.
Figure 1. The location and digital elevation model (DEM) of the study area.
Sustainability 18 05887 g001
Sustainability 18 05887 g002
Figure 2. Conceptual model based on Structural Equation Modeling (SEM). Ellipses represent latent variables, and rectangles indicate observed variables. Unidirectional arrows denote causal relationships. Solid lines indicate positive effects, while gray lines represent negative effects.
Figure 2. Conceptual model based on Structural Equation Modeling (SEM). Ellipses represent latent variables, and rectangles indicate observed variables. Unidirectional arrows denote causal relationships. Solid lines indicate positive effects, while gray lines represent negative effects.
Sustainability 18 05887 g002
Sustainability 18 05887 g003
Figure 3. Spatial patterns of species richness for (A) all bird species, (B) waterbird species, and (C) land bird species in the Sanjiang Plain (2018).
Figure 3. Spatial patterns of species richness for (A) all bird species, (B) waterbird species, and (C) land bird species in the Sanjiang Plain (2018).
Sustainability 18 05887 g003
Sustainability 18 05887 g004
Figure 4. Distribution of habitat quality in the Sanjiang Plain (2018).
Figure 4. Distribution of habitat quality in the Sanjiang Plain (2018).
Sustainability 18 05887 g004
Sustainability 18 05887 g005
Figure 5. Structural equation model illustrating the factors influencing habitat quality and waterbird species richness. Ellipses: latent variables; rectangles: observed variables. Black/gray lines: positive/negative effects; solid/dashed lines: significant (p < 0.05)/non-significant (p ≥ 0.05) effects. Numbers: standardized path coefficients.
Figure 5. Structural equation model illustrating the factors influencing habitat quality and waterbird species richness. Ellipses: latent variables; rectangles: observed variables. Black/gray lines: positive/negative effects; solid/dashed lines: significant (p < 0.05)/non-significant (p ≥ 0.05) effects. Numbers: standardized path coefficients.
Sustainability 18 05887 g005
Sustainability 18 05887 g006
Figure 6. Structural equation model illustrating the factors influencing land bird habitat quality and species richness. Ellipses: latent variables; rectangles: observed variables. Black/gray lines: positive/negative effects; solid/dashed lines: significant (p < 0.05)/non-significant (p ≥ 0.05) effects. Numbers: standardized path coefficients.
Figure 6. Structural equation model illustrating the factors influencing land bird habitat quality and species richness. Ellipses: latent variables; rectangles: observed variables. Black/gray lines: positive/negative effects; solid/dashed lines: significant (p < 0.05)/non-significant (p ≥ 0.05) effects. Numbers: standardized path coefficients.
Sustainability 18 05887 g006
Table 1. Data sources, processing steps, and resampling methods.
Table 1. Data sources, processing steps, and resampling methods.
DatasetOriginal ResolutionSourceVersionProcessing StepsResampling Method (to 1 km)
Land use/land cover (LULC)30 mRESDC (http://www.resdc.cn)CLUD v1.0 (2018 version)Mosaic, projection transform, extract by maskNearest neighbor (categorical)
Digital Elevation Model (DEM)30 mASTER GDEM (https://www.gscloud.cn/)v3Slope calculation using ArcGIS 10.8Bilinear
Annual mean temperature1 kmNational Tibetan Plateau Data Center [33,34]1.0 (2020)Pixel-based annual average for 2018None
Annual precipitation1 kmNational Tibetan Plateau Data Center [33,34]1.0 (2020)Pixel-based annual total for 2018None
NDVI (annual maximum)250 mMOD13Q1 (MODIS/Terra)Collection 6.1Maximum value compositing (MVC) across all 23 MOD13Q1 16-day composites (1 January–31 December 2018)Bilinear
Soil organic carbon (SOC)1 kmNational Earth System Science Data Center (http://www.geodata.cn)v1.0Convert to g/kg using scale factor (original unit: 0.1%)None
Population density1 kmWorldPop (https://hub.worldpop.org/)UN-adjusted 2018Extract by maskNone
Nighttime light (NPP-VIIRS-like)500 mHarvard Dataverse (https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/GIYGJU accessed on 10 December 2025) [35]Cross-sensor corrected v1.0Extract by mask (data already corrected by provider)Bilinear
Gross Domestic Product (GDP)1 kmRESDC (GDP spatial distribution dataset for China, 2018)v1.0Extract by mask, unit: 104 CNY/km2None
Distance to roadsVector (OSM)Geofabrik (http://download.geofabrik.de/ accessed on 10 December 2025)2018 extractEuclidean distance raster at 1 km resolution–
Bird distribution ranges–BirdLife International/HBW (https://www.birdlife.org/)2021 versionOverlay, richness calculation at 1 km grid–
Table 2. Threat factors and parameters setting.
Table 2. Threat factors and parameters setting.
Threat FactorMaximum Distance (km)WeightDecay Type
Industrial and mining land6.000.75exponential
Arable land1.500.65linear
Urban land8.000.90exponential
Rural residential areas4.000.65exponential
Alkali land2.500.30linear
Barren land2.500.30exponential
Table 3. Habitat suitability and sensitivity to threat factors of LULC.
Table 3. Habitat suitability and sensitivity to threat factors of LULC.
LULCHabitat SuitabilityThreat Factor
Arable LandUrban LandRural Residential AreasIndustrial and Mining LandAlkali LandBarren Land
Paddy Field0.60.30.50.350.40.50.1
Arable Dry Land0.40.60.50.350.40.20.2
Forest Land10.50.850.650.60.60.2
Shrubland10.30.70.60.50.60.1
Sparse Woodland10.60.850.650.60.50.3
Other Woodland10.60.850.650.60.20.3
High Coverage Grassland0.750.40.60.40.50.250.2
Medium Coverage Grassland0.70.50.70.50.550.30.3
Low Coverage Grassland0.60.50.80.60.550.30.3
River10.50.90.70.80.250.15
Lake0.90.50.90.750.80.20.15
Marshland0.90.60.90.750.80.20.15
Urban Area0.1000000
Rural residential areas0.2000000
Other Construction Land0.1000000
Sandy Land0.1000000
Saline-alkali Land0.50.20.20.150.150.150.1
Bare Rocky Land0.650.70.50.20.20.30.3
Footnote: Habitat suitability scores represent the capacity of each land use type to support general biodiversity as defined in the InVEST framework. These scores are not guild-specific; the differential responses of waterbirds and land birds to the same habitat quality index are examined through structural equation modeling (see Section 3.3).
Table 4. Fitness test for structural equation fitting.
Table 4. Fitness test for structural equation fitting.
ModelNComparative Fit Index (CFI)Goodness of Fit Index (GFI)Standardized Root Mean Square Residual (SRMR)Root Mean Square Error of Approximation (RMSEA)
Habitat quality11,7700.9700.9860.0380.051
Waterbird11,7700.9800.9880.0290.050
Land bird11,7700.9890.9800.0400.050
Acceptance Standards->0.90>0.90<0.05<0.06
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sun, X.; Liu, C.; Li, Y.; Li, Y.; Li, Y. Direct and Indirect Effects of Natural and Anthropogenic Drivers on Avian Diversity in the Sanjiang Plain, Northeast China. Sustainability 2026, 18, 5887. https://doi.org/10.3390/su18125887

AMA Style

Sun X, Liu C, Li Y, Li Y, Li Y. Direct and Indirect Effects of Natural and Anthropogenic Drivers on Avian Diversity in the Sanjiang Plain, Northeast China. Sustainability. 2026; 18(12):5887. https://doi.org/10.3390/su18125887

Chicago/Turabian Style

Sun, Xiuli, Chenxiao Liu, Yueyuan Li, Yuehui Li, and Yue Li. 2026. "Direct and Indirect Effects of Natural and Anthropogenic Drivers on Avian Diversity in the Sanjiang Plain, Northeast China" Sustainability 18, no. 12: 5887. https://doi.org/10.3390/su18125887

APA Style

Sun, X., Liu, C., Li, Y., Li, Y., & Li, Y. (2026). Direct and Indirect Effects of Natural and Anthropogenic Drivers on Avian Diversity in the Sanjiang Plain, Northeast China. Sustainability, 18(12), 5887. https://doi.org/10.3390/su18125887

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

Article Metrics

Back to TopTop