Decoupling Patterns and Drivers of Macrozoobenthos Taxonomic and Functional Diversity to Wetland Chronosequences in Coal Mining Subsidence Areas

Abstract

Surface subsidence caused by coal mining activities generates diverse wetland ecosystems. These newly formed wetlands exhibit distinct environmental characteristics due to variations in subsidence age, resulting in divergent biological communities. While species adapt to environmental changes through specific functional trait combinations, the response of aquatic community functional diversity to environmental gradients across chronosequences of mining subsidence wetlands remains unclear. This study investigated 13 coal mining subsidence wetlands (1–18 years) of macrozoobenthos in Jining, China. Through seasonal monitoring, we analyzed functional traits along with taxonomic and functional diversity patterns. Initial-stage wetlands were dominated by medium-sized (63.9%) and tegument-respiring taxa, whereas late-stage wetlands exhibited a shift toward large-sized (43.9%) and gill-respiring groups. Both species richness and functional richness declined over time, with taxonomic diversity demonstrating greater sensitivity to subsidence age. Seasonal community variability was more pronounced in initial-stage wetlands (1–4 years post-subsidence). Despite increasing habitat heterogeneity with subsidence age, functional redundancy maintains ecosystem stability. The shared origin and developmental trajectory of these wetlands may constrain functional divergence. Current research predominantly relies on traditional taxonomic metrics, whereas our findings emphasize functional trait analysis’s importance for ecosystem assessment, which provides a theoretical framework for ecological restoration and biodiversity conservation in post-subsidence wetlands.

1. Introduction

Understanding how biodiversity responds to environmental change constitutes a central question in community ecology and conservation biology [1,2]. Wetlands, as critically important ecosystems, provide vital services to society [3,4,5], including carbon sequestration, nutrient cycling, flood and drought mitigation, water purification, and the maintenance of biodiversity [6,7,8,9,10]. Characterized by high productivity, diverse habitats, and abundant water resources, wetlands sustain unique and rich biodiversity, which in turn constitute key elements in sustaining wetland ecosystem functioning [11]. For example, some studies have shown that differences in ecosystem function among different life forms due to plant phylogeny or plant adaptations to the environment, and the complex interactions between species of different functional groups in natural ecosystems can have significant effects on the structure and function of ecosystems [12,13]. However, wetlands are also among the ecosystems most heavily impacted by human activities and are experiencing unprecedented degradation and loss [14,15]. China experienced a net loss of 60.9 × 10³ km² in wetland area from 1980 to 2020 [16]. In this context, constructed and newly formed wetlands have gained increasing attention as complementary or alternative systems to natural wetlands, given their ecological functions and conservation values [17,18,19].

Coal mining subsidence areas are depressions formed by vertical ground subsidence resulting from stress redistribution in overlying strata after underground coal extraction [20,21]. These topographically low-lying areas readily accumulate precipitation and groundwater, forming waterlogged zones that gradually develop into the newly formed wetlands through hydrological accumulation and ecological succession, characterized by inundated environments, hydric soils, and wetland vegetation communities [22,23]. For example, several end-pit lakes in the Rocky Mountain foothills of west-central Alberta, Canada [24], coal-mining ponds in southern Brazil [25], and coal mining basins within the Czech Republic [22] are all formed in this way. Notably, mining subsidence wetlands (MSWs) differ substantially from natural wetlands. First, the spatiotemporal heterogeneity of mining activities leads to significant variations in subsidence age, developmental stage, and environmental conditions [26,27]. Second, wetland stability varies with the time since subsidence: initial-stage wetlands exhibit hydrological fluctuations and unstable substrate structures, whereas late-stage wetlands demonstrate community stability comparable to natural wetlands, attributable to sediment compaction and organic matter accumulation that create stable microhabitats [28,29]. This unique formation mechanism likely results in biological communities with distinct characteristics from those in natural wetlands [27]. Critically, coal mining subsidence acts as a key driver shaping macrozoobenthos assemblages in these newly formed wetlands. The pronounced spatiotemporal heterogeneity in environmental features and biological communities across different subsidence chronosequences provides a natural laboratory for studying community responses to environmental changes [27].

Functional traits and functional diversity provide powerful analytical tools for elucidating how environmental changes drive community responses and associated ecosystem processes [30,31]. Ecological traits (e.g., feeding habitats, habit) regulate species’ adaptive capacity to habitat environmental changes [32,33], while life-history traits (e.g., voltinism, growth rate) reflect species-specific strategies shaped by long-term environmental adaptation, offering critical insights into habitat filtering and species interactions during community assembly [34]. In wetland ecosystems, macrozoobenthos functional diversity often exhibits greater resistance to disturbances than taxonomic diversity [35]. For instance, under eutrophication stress, functional redundancy enables pollution-tolerant species (e.g., chironomid larvae, tubificid worms) to maintain organic matter decomposition through similar filter-feeding strategies, thereby stabilizing ecosystem functioning during initial species loss [36,37,38]. However, when environmental pressures directly target key functional traits (e.g., drought eliminating gill-breathing bivalves), functional diversity responds more rapidly, serving as a better early-warning indicator of ecosystem degradation than species diversity [39]. This demonstrates that functional diversity simultaneously reflects ecosystem buffering capacity and sensitivity to specific stressors [39,40]. Despite these advantages, current assessments of community succession still predominantly rely on traditional taxonomic metrics, with limited application of trait-based approaches [41,42].

Macrozoobenthos constitute a vital component of freshwater wetland ecosystems. Their widespread distribution, ease of collection, and sensitivity to environmental changes make them excellent bioindicators for monitoring aquatic ecosystem health [43,44,45,46]. Occupying intermediate trophic levels, macrozoobenthos play crucial roles in energy flow and nutrient cycling, thereby maintaining critical wetland ecosystem functions [47]. The functional trait combinations of wetland macrozoobenthos typically reflect adaptations to stagnant, low-oxygen environments [48]. Body size demonstrates clear environmental filtering: small-bodied taxa (e.g., chironomid larvae) dominate hydrologically unstable temporary habitats, while large species prevail in stable water bodies [49]. Respiratory types vary from cutaneous respiration (e.g., oligochaetes, chironomids) to specialized gills or spiracles that enhance oxygen uptake under hypoxic conditions. Macrozoobenthos feeding habits can affect secondary productivity and the nutrient cycle [50]. These functional trait patterns not only reveal environmental filtering mechanisms but also provide novel quantitative metrics for wetland health assessment [49,51].

Wetland ecosystems in temperate climate zones exhibit pronounced seasonal dynamics [52]. This study investigates the temporal patterns of macrozoobenthos functional diversity during ecosystem development in MSWs of temperate regions through multi-seasonal sampling. Considering the habitat heterogeneity and water parameter variations caused by different subsidence ages, this study proposes three hypotheses: (1) habitat complexity and stability increase with the time since subsidence, leading to higher taxonomic and functional diversity in older subsidence wetlands [37,53]; (2) functional diversity indices will show greater sensitivity to subsidence age than taxonomic diversity, as functional traits directly reflect species’ environmental adaptation strategies; (3) younger subsidence wetlands (initial stage) exhibit stronger seasonal fluctuations in species composition, taxonomic diversity, and functional diversity compared to mature wetlands, owing to their greater environmental instability. This study not only analyzes taxonomic diversity, but also explores the response patterns of functional diversity of macrozoobenthos in temperate coal mining subsidence wetlands along the subsidence age gradient through multi-season functional trait analysis. This not only provides new insights into the theory of wetland ecological succession but also makes significant contributions to both fundamental research and practical applications in wetland ecology.

2. Materials and Methods

2.1. Study Area and Sampling Design

Based on field surveys and remote sensing interpretation, we selected 13 MSWs (1–18 years post-subsidence) in Jining City (35°40′80″~35°52′78″ N, 116°80′82″~116°88′34″ E; Figure 1;

Table A1. Site characteristics of the studied MSWs in Jining City, China.

Sample	Area (km2)	Subsidence Age (Years)	Subsidence Stage	Dominant Vegetation
A	0.1319	5	Middle stage	Phragmites australis, Typha angustifolia
B	0.0676	5	Middle stage	Phragmites australis, Typha angustifolia, Potamogeton crispus, Ceratophyllum demersum, Najas marina
C	0.0301	1	Initial stage	Phragmites australis, Typha angustifolia, Potamogeton crispus, Echinochloa caudata, Abutilon theophrasti
D	0.0628	2	Initial stage	Phragmites australis, Typha angustifolia, Persicaria lapathifolia, Echinochloa caudata, Cyperus rotundus, Abutilon theophrasti
E	0.0368	3	Initial stage	Nelumbo nucifera, Typha angustifolia, Persicaria lapathifolia, Echinochloa caudata, Euryale ferox
F	0.0313	1	Initial stage	Phragmites australis, Typha angustifolia, Lemna minor, Echinochloa crus-galli, Persicaria lapathifolia, Alternanthera philoxeroides
G	0.0170	6	Middle stage	Nelumbo nucifera, Phragmites australis, Cyperus glomeratus, Echinochloa caudata, Persicaria lapathifolia, Potentilla supina
H	0.5249	2	Initial stage	Phragmites australis, Typha angustifolia, Trapa bispinosa, Echinochloa caudata, Persicaria lapathifolia
I	0.0287	17	Late stage	Nelumbo nucifera, Phragmites australis, Bidens frondosa, Echinochloa caudata, Stuckenia pectinata, Cyperus glomeratus, Alternanthera philoxeroides, Potamogeton crispus
J	1.1300	18	Late stage	Phragmites australis, Echinochloa caudata, Cyperus glomeratus
K	0.5791	18	Late stage	Phragmites australis
L	0.0758	6	Middle stage	Phragmites australis, Trapa bispinosa
M	1.4200	18	Late stage	Phragmites australis

) for macrozoobenthos surveys during spring (May), summer (July), autumn (September), and winter (December) of 2024. Located in the temperate climate zone of China, Jining experiences hot rainy summers and cold dry winters, with mean annual precipitation of 600 mm and average temperatures of 13.5–15 °C [54]. These wetlands were all formed by coal mining-induced subsidence and were originally farmland prior to subsidence [27]. They exhibited significant spatial heterogeneity in area (0.0287–1.42 km²) and hydrological connectivity. Water sources primarily included precipitation and groundwater, with some wetlands connected to adjacent water systems via artificial channels. Dominant vegetation comprised emergent plants (Phragmites australis, Typha angustifolia), accompanied by floating species (Trapa bispinosa, Lemna minor). Surrounding landscapes were predominantly agricultural (corn–wheat rotation systems), with scattered villages and plantation forests [27].

Due to spatiotemporal variations in underground coal mining activities, surface subsidence in Jining’s coal mining areas occurred during different stages. Based on field conditions and subsidence age, we classified these MSWs into three distinct successional stages:

Initial stage (IS, five wetlands with a subsidence history of 1–4 years): Characterized by small wetland areas and fragile ecosystems, these wetlands exhibit strong hydrological dependence on precipitation and may experience complete drying during prolonged droughts. The aquatic vegetation is dominated by Phragmites australis and Typha angustifolia, forming relatively simple plant communities.

Middle stage (MS, four wetlands with a subsidence history of 5–10 years): These wetlands represent a transitional stage of dynamic development, where continued subsidence leads to gradual expansion of wetland area and improved environmental stability. While Phragmites australis and Typha angustifolia remain dominant, the vegetation begins to show increased structural complexity compared to IS.

Late stage (LS, four wetlands with a subsidence history of more than 15 years post-subsidence): Subsidence processes are largely complete, resulting in well-developed “subsidence lakes” with extensive shoreline vegetation. Distinct zonation patterns emerge, dominated by bands of Phragmites australis community and Typha angustifolia community. Some sites develop submerged macrophytes including Stuckenia pectinata and Potamogeton crispus. However, these wetlands often display reduced community complexity, tending toward homogeneous, single-dominant plant communities.

2.2. Sample Collection and Processing

Three sampling sites were randomly established in each newly formed wetland. At each site, three replicate samples were collected using a 1/40 Peterson grab sampler (Wildlife Supply Company, Yulee, USA) (following a shallow-to-deep gradient) and subsequently pooled into a single composite sample. The collected samples were sieved through a 250-μm mesh sieve, transferred to sealed bags, and preserved in 75% ethanol. Macrozoobenthos identification was conducted following relevant taxonomic keys and literature [55,56,57,58] in the laboratory. Morphological identification was primarily performed using stereomicroscopes and compound microscopes. Specifically, aquatic oligochaetes and chironomid larvae were identified under a compound microscope, while hirudineans, gastropods, and most aquatic insects were examined using a SOPTOP SZN stereomicroscope (SOPTOP, Ningbo, China). Specimens were classified into the lowest feasible taxonomic level (usually at genus level). Furthermore, all benthic organisms from each sampling site were enumerated, with counts recorded for each taxonomic unit. For fragmented specimens, only heads were counted while tails were disregarded. For data analysis, densities were standardized to individuals per square meter (ind./m²).

2.3. Measurement of Environmental Variables

A total of seven environmental variables were measured, including physical and water chemistry variables. Physical habitat variables comprised soil organic matter (SOM), while water physicochemical variables included pH, dissolved oxygen (DO), electrical conductivity (Cond), total nitrogen (TN), total phosphorus (TP), and chlorophyll-a (Chl-a). In situ measurements of Cond, pH, DO were conducted using a DZB-718 portable multiparameter water quality analyzer (Shanghai Yidian Scientific Instrument Co., Ltd., Shanghai, China). Prior to biological sampling, 1 L water samples were collected for laboratory analysis of TN, TP, and Chl-a, which were measured according to APHA (2012) [59].

The formation time of newly formed wetlands was determined through a visual interpretation of satellite imagery (2006–2024, sourced from the Geospatial Data Cloud platform) based on established interpretation keys using ArcGIS 10.8 software. Additionally, the surface area of subsidence wetlands in 2024 was measured using the area calculation tool in ArcGIS.

2.4. Functional Traits

A total of five functional traits with 19 trait categories were selected, and were known to be sensitive to environmental changes: body size, respiration type, functional feeding groups (FFG), shape, and habit (

Table 1. Functional traits, trait categories, scores, and their codes of macrozoobenthos.

Trait	Category	Score	Code
Body size	Small (<9 mm)	1	Size1
	Medium (9–16 mm)	2	Size2
	Large (>16 mm)	3	Size3
Respiration type	Integumentary	1	Resp1
	Branchial	2	Resp2
	Air (spiracles, tracheae, plastrons)	3	Resp3
Functional feeding	Collector-gatherers (GC)	1	FFG1
groups	Collector-filterers (FC)	2	FFG2
	Scrapers (SC)	3	FFG3
	Predators (PR)	4	FFG4
	Shredders (SH)	5	FFG5
Shape	Streamlined (flat, fusiform)	1	Shpe1
	Not streamlined (cylindrical, round or bluff)	2	Shpe2
Habit	Climbers	1	Habi1
	Burrowers	2	Habi2
	Sprawlers	3	Habi3
	Clingers	4	Habi4
	Swimmers	5	Habi5
	Divers	6	Habi6

). Following the approach of Colzani et al. [60], each trait category was assigned a binary score (0, 1, 2,…). Trait classification and scoring were primarily derived from published literature [58,61].

2.5. Data Analysis

2.5.1. Analysis of Taxonomic and Functional Composition

The functional trait composition of macrozoobenthos was characterized using the Community-Weighted Mean (CWM). CWM was calculated based on two data matrices: a community composition matrix (sites × species) and a functional trait matrix (species × traits), implemented via the “dbFD” function in the R package.

An analysis of similarities (ANOSIM) was applied to examine differences in taxonomic and functional composition among wetlands at three subsidence stages. The Global R statistic indicates the magnitude of dissimilarity, quantifying the ratio between among-group differences and within-group differences. Global R ≈ 1 signifies that among-group differences significantly exceed within-group differences, indicating a strong influence of the grouping factor on community structure. Conversely, Global R ≈ 0 suggests no significant distinction between among-group and within-group differences, implying negligible grouping effects. Permutation tests were used to assess whether among-group differences were significantly greater than within-group differences. p ≤ 0.05 indicates that the grouping factor significantly influenced community structure, while p > 0.05 suggests non-significant grouping effects. Prior to analysis, both community data and trait data (CWM) were log(x + 1)-transformed. These analyses were performed using the “anosim” function in the R (version 4.4.1) package vegan.

2.5.2. Analysis of Taxonomic and Functional Diversity

Taxonomic diversity of macrozoobenthos in MSWs was assessed using species richness, Pielou’s evenness index (J) [62], Shannon–Wiener index (H′) [63], and Simpson index. Species with relative abundance exceeding 10% were classified as dominant. These analyses were conducted in Primer6.

Functional diversity was characterized using functional richness (FRic), functional evenness index (FEve), and functional dispersion index (FDis). The CWM values represented the mean percentage contribution of each trait within a given trait category. These analyses were implemented via the “dbFD” function in the R package.

Following the framework for studying functional diversity and redundancy proposed by Ricotta et al. [64,65], we calculated the functional redundancy index. Quadratic Entropy Index (RaoQ) and Simpson index were used to measure functional diversity (DF) and taxonomic diversity (DS) within individual communities, respectively. The functional redundancy index (FRI) was computed as: FRI = 1 − DF/DS. RaoQ was implemented via the “dbFD” function in the R package and other analyses were conducted using the “gdm” package in the R package.

2.5.3. Statistical Analysis of Key Drivers

Constrained ordination analysis (CCA) was employed to examine the relationships between macrozoobenthos community taxonomic and functional composition, and environmental variables, respectively. First, to improve data normality and homoscedasticity, environmental variables (except pH), community data, and trait data were log(x + 1)-transformed. Highly correlated environmental variables (|r| > 0.70) were excluded. Correlation analyses were performed using the “Hmisc” package in the R package.

Subsequently, detrended correspondence analysis (DCA) was applied to macrozoobenthos community and trait data to determine the maximum axis length. If the maximum axis length exceeded 4, CCA was selected; if it was less than 3, redundancy analysis (RDA) was chosen; and if it fell between 3 and 4, either method was appropriate [66]. DCA was conducted using the “decorana” function, while CCA and RDA were performed using the “cca” and “rda” functions, respectively, in the R package (version 4.4.1) vegan (https://www.r-project.org/ accessed on 1 June 2024).

3. Results

3.1. Environmental Variables

The results demonstrated that the area progressively expanded with increasing duration of coal mining subsidence. Significant differences (p < 0.05, Figure 2) were observed in water chemistry parameters (Cond (Figure 2a), DO (Figure 2b), TP (Figure 2c), Chl-a (Figure 2d)), and physical factors (SOM (Figure 2e)) among wetlands with different subsidence age. Post hoc comparisons revealed that DO, TP, Chl-a and SOM exhibited lower values in IS compared to LS, whereas Cond showed an opposite trend with higher values in IS. Regarding seasonal variations, Chl-a and Cond displayed higher values in IS during spring and winter, while DO, TP, and SOM exhibited lower values in IS during summer and autumn.

3.2. Taxonomic Composition and Functional Composition

The macrozoobenthos we identified belong to 12 orders and 25 families. (

Table A2. Taxonomic list and functional traits of macrozoobenthos of three stages of mining subsidence wetlands.

	Initial Stage	Middle Stage	Late Stage	Body Size	Respiration Type	Functional Feeding Groups	Shape	Habit
Annelida
Oligochaeta
Dero	+			Small	Integumentary	GC	not streamlined	Sprawlers
Haemonais	+			Small	Integumentary	GC	not streamlined	Burrowers
Nais	+			Small	Integumentary	GC	not streamlined	Sprawlers
Branchiura sowerbyi	+	+	+	Large	Branchial	GC	not streamlined	Burrowers
Aulodrilus pluriseta	+			Large	Integumentary	GC	not streamlined	Burrowers
Limnodrilus	+			Large	Integumentary	GC	not streamlined	Burrowers
Limnodrilus grandisetosus		+	+	Large	Integumentary	GC	not streamlined	Burrowers
Limnodrilus claparedeianus	+	+	+	Large	Integumentary	GC	not streamlined	Burrowers
Limnodrilus hoffmeisteri	+	+	+	Large	Integumentary	GC	not streamlined	Burrowers
Hirudinida
Glossiphonia	+			Medium	Integumentary	PR	not streamlined	Clingers
Barbronia	+			Medium	Integumentary	PR	not streamlined	Clingers
Mollusca
Gastropoda
Mesogastropoda
Parafossarulus striatulus		+	+	Medium	Branchial	SC	not streamlined	Sprawlers
Semisulcospira ningpoensis		+		Large	Branchial	SC	not streamlined	Sprawlers
Cipangopaludina chinensis		+	+	Large	Branchial	SC	not streamlined	Sprawlers
Bellamya quadrata	+	+	+	Large	Branchial	SC	not streamlined	Sprawlers
Radix lagotis	+	+	+	Large	Branchial	SC	not streamlined	Sprawlers
Hippeutis cantori	+		+	Small	Branchial	SC	not streamlined	Sprawlers
Hippeutis umbilicalis		+		Small	Branchial	SC	not streamlined	Sprawlers
Bivalvia
Unionida
Anodonta woodiana wodiana	+		+	Large	Branchial	FC	not streamlined	Burrowers
Anodont arcaeformis			+	Large	Branchial	FC	not streamlined	Burrowers
Arthropoda
Insecta
Ephemeroptera
Baetidae	+			Small	Air	GC	streamlined	Swimmers
Trichoptera
Macrostemum		+		Small	Air	SC	not streamlined	Clingers
Odonata
Gomphidae		+		Medium	Air	PR	not streamlined	Divers
Libellula	+	+		Large	Air	PR	not streamlined	Divers
Orthetrum	+	+	+	Large	Air	PR	not streamlined	Divers
Coenagrion			+	Medium	Air	PR	streamlined	Divers
Ischnura		+		Medium	Air	PR	streamlined	Divers
Erythromma		+		Large	Air	PR	streamlined	Divers
Coleoptera
Chrysomelidae	+		+	Small	Air	PR	not streamlined	Sprawlers
Elmidae	+			Small	Air	SC	not streamlined	Burrowers
Staphylinidae	+			Large	Air	PR	not streamlined	Sprawlers
Hemiptera
Sphaerodema	+			Large	Air	PR	streamlined	Swimmers
Micronectra	+	+		Small	Air	PR	streamlined	Swimmers
Lepidoptera
Parapoynx	+			Large	Air	SH	not streamlined	Sprawlers
Diptera
Ceratopogonidae	+	+	+	Small	Air	PR	not streamlined	Sprawlers
Dolichopodidae	+			Medium	Air	PR	not streamlined	Burrowers
Ephydridae	+		+	Medium	Air	SH	not streamlined	Burrowers
Tabanidae	+		+	Medium	Air	PR	not streamlined	Sprawlers
Chironomidae
Chironomus	+	+	+	Medium	Integumentary	GC	not streamlined	Burrowers
Cladopelma	+	+	+	Small	Integumentary	GC	not streamlined	Burrowers
Cladotanytarsus	+	+	+	Small	Integumentary	FC	not streamlined	Sprawlers
Rheotanytarsus	+	+		Small	Integumentary	FC	not streamlined	Clingers
Cryptochironomus	+		+	Small	Integumentary	PR	not streamlined	Sprawlers
Cryptotendipes	+	+	+	Small	Integumentary	GC	not streamlined	Sprawlers
Dicrotendipes	+	+	+	Medium	Integumentary	GC	not streamlined	Burrowers
Enfeldia	+	+	+	Medium	Integumentary	GC	not streamlined	Burrowers
Endochironomus	+		+	Medium	Integumentary	SH	not streamlined	Clingers
Glyptotendipes	+	+	+	Medium	Integumentary	SH	not streamlined	Burrowers
Microchironomus	+			Small	Integumentary	GC	not streamlined	Burrowers
Microtendipes			+	Small	Integumentary	FC	not streamlined	Clingers
Polypedilum	+	+	+	Medium	Integumentary	SH	not streamlined	Climbers
Tanytarsus	+	+	+	Small	Integumentary	FC	not streamlined	Climbers
Cricotopus	+	+	+	Small	Integumentary	SH	not streamlined	Clingers
Propsilocerus	+	+	+	Large	Integumentary	GC	not streamlined	Sprawlers
Procladius	+	+	+	Medium	Integumentary	PR	not streamlined	Sprawlers
Tanypus	+	+	+	Medium	Integumentary	PR	not streamlined	Sprawlers

). Aquatic insects dominated the community, primarily represented by Diptera (39.29%) and Odonata (10.71%). Significant differences in species richness were observed among different subsidence stages (p < 0.05), with the highest richness in IS (44 species, 78.57%), followed by LS (33 species, 58.93%) and MS (32 species, 57.14%). Seasonal analysis revealed that MS had lower species richness than IS and LS during spring, autumn, and winter, but exhibited higher richness during summer (Figure 3a).

ANOSIM analysis showed significant differences in taxonomic and functional compositions among all newly formed wetlands during summer (Global R = 0.219, p = 0.001 for taxonomic; Global R = 0.116, p < 0.01 for function) (

Table 2. ANOSIM analysis of seasonal variations in taxonomic and functional composition of macrozoobenthos communities across each stage in MSWs (statistically significant differences are highlighted in bold).

		Global R	p	IS vs. MS		IS vs. LS		MS vs. LS
				R	p	R	p	R	p
Spring
	Taxonomic composition	0.041	0.171	0.068	0.100	0.013	0.392	0.037	0.234
	Functional composition	−0.028	0.713	−0.008	0.491	−0.094	0.961	0.033	0.233
Summer
	Taxonomic composition	0.219	0.001	0.253	0.001	0.330	0.001	0.037	0.203
	Functional composition	0.116	0.006	0.028	0.234	0.267	0.002	0.035	0.198
Autumn
	Taxonomic composition	−0.037	0.744	0.033	0.295	−0.054	0.825	−0.088	0.900
	Functional composition	−0.066	0.903	−0.116	0.933	−0.040	0.714	−0.059	0.765
Winter
	Taxonomic composition	0.236	0.001	0.292	0.001	0.245	0.001	0.138	0.028
	Functional composition	0.040	0.118	0.024	0.256	0.085	0.055	0.018	0.275
Average
	Taxonomic composition	0.043	0.139	0.070	0.111	−0.036	0.740	0.114	0.061
	Functional composition	0.086	0.026	0.147	0.017	0.061	0.131	0.044	0.211

). Significant differences in taxonomic composition were also detected in winter (Global R = 0.236, p = 0.001), while functional composition differed significantly across all seasons (Global R = 0.086, p < 0.05) (Table 2). Dominant groups shifted with subsidence age: Chironomus, Glyptotendipes, and Enfeldia dominated IS; Glyptotendipes, Polypedilum, and Propsilocerus were predominant in MS; while Chironomus, Glyptotendipes, and Propsilocerus prevailed in LS (

Table A3. Dominant groups across different subsidence stages of MSWs.

	Dominant Groups
Initial stage	Chironomus, Glyptotendipes, Enfeldia
Middle stage	Glyptotendipes, Polypedilum, Propsilocerus
Late stage	Chironomus, Glyptotendipes, Propsilocerus

).

Functional trait analysis revealed distinct patterns across three subsidence stages (Figure 4,

Table A4. Mean percentage composition of functional traits in macrozoobenthos communities across different subsidence stages of MSWs.

Trait	Category	Initial Stage	Middle Stage	Late Stage
Body size	Small	18.600	10.700	12.000
	Medium	63.900	53.100	44.100
	Large	17.500	36.200	43.900
Respiration type	Integumentary	84.700	82.110	80.600
	Branchial	7.000	9.740	16.200
	Air	8.300	8.150	3.200
Functional	Collector-gatherers (GC)	49.100	64.560	56.660
feeding groups	Collector-filterers (FC)	2.700	3.770	2.930
	Scrapers (SC)	4.800	7.350	6.830
	Predators (PR)	17.300	6.420	9.780
	Shredders (SH)	26.100	17.900	23.800
Shape	Streamlined	4.000	4.400	0.500
	Not Streamlined	96.000	95.600	99.500
Habit	Climbers	4.910	10.400	5.670
	Burrowers	56.765	52.920	63.200
	Sprawlers	29.100	28.800	29.200
	Clingers	4.550	2.210	1.190
	Swimmers	3.990	1.920	0.000
	Divers	0.685	3.750	0.740

). Body size composition showed medium-sized organisms (9–16 mm) predominated throughout all stages, though their relative abundance decreased progressively from IS (63.9%), MS (53.1%) to LS (44.1%). Conversely, large-bodied species (>16 mm) exhibited a consistent increase along the subsidence chronosequence, rising from 17.5% in IS to 43.9% in LS. Respiration types were dominated by integumentary respiration, maintaining high proportions in both IS (84.7%) and LS (80.6%). Notably, branchial respirers demonstrated significant increases during wetland maturation, expanding from 7.0% to 16.2% of the community. Functional feeding groups displayed temporal dynamics, with gatherer-collectors (GC) peaking in MS (64.56%). IS communities showed elevated proportions of both shredders (SH, 26.1%) and predators (PR, 17.3%) compared to other stages. Morphological analysis revealed near-complete dominance (>95%) of non-streamlined body shapes across all stages, with streamlined forms occurring only marginally (4%) in IS and MS. Life habit composition remained relatively stable, with burrowers consistently representing the dominant group (56.8–63.2% across stages), followed by sprawlers (29%). This persistence suggests habitat architecture may maintain consistent niche availability despite successional changes.

3.3. Taxonomic and Functional Diversity

The results revealed significant differences in the species richness and Shannon–Wiener index across three subsidence stages (p < 0.05), while no significant variation was observed in Pielou’s evenness index (Figure 3). Notably, both the species richness and Shannon–Wiener index were significantly higher in IS compared to MS (p < 0.05). Seasonal analysis demonstrated that the Shannon–Wiener index was lower in MS during spring, autumn, and winter, but exhibited an opposite pattern in summer when MS exhibited higher values than both IS and LS (Figure 3).

FEve showed significant differences among three subsidence stages of wetlands (p < 0.05), with IS displaying significantly higher values than MS. However, neither FRic nor FDiv exhibited significant differences across subsidence stages (Figure 3). Moreover, we found significant seasonal differences in FEve for IS, and significant seasonal differences in both FRic and FDiv for LS.

The FRI medians varied across the three subsidence stages. The IS and LS exhibited relatively similar median values, while the MS showed a slightly lower median. There is greater dispersion in the MS data. In contrast, the IS and LS demonstrated more clustered data distributions (Figure 5). Numerically, LS had the highest FRI, whereas the MS recorded the lowest values (

Table A5. FRI of macrozoobenthos communities in MSWs at each stage across different seasons.

	Spring	Summer	Autumn	Winter	Average
Initial stage	0.43	0.33	0.4	0.34	0.375
Middle stage	0.23	0.33	0.24	0.4	0.3
Late stage	0.33	0.55	0.3	0.43	0.4025

).

3.4. Key Drivers in Shaping Taxonomic and Functional Variation

Overall, the CCA of annual macrozoobenthos taxonomic composition revealed that environmental factors collectively explained 35.21% of the community structure variation. The first and second axes accounted for 24.75% and 20% of the variance, respectively (Figure 6e). Key environmental drivers significantly affecting taxonomic composition included: TN (F = 2.000, p < 0.01), Chl-a (F = 2.363, p < 0.01), pH (F = 2.001, p < 0.01), Cond (F = 2.584, p < 0.01), DO (F = 1.902, p < 0.01), subsidence age (F = 2.138, p < 0.01), and area (F = 2.122, p < 0.01) (Figure 6,

Table 3. CCA results for relationships between macrozoobenthos community structure and environmental factors in MSWs (only significantly influential factors shown).

	Taxonomic Composition					Functional Composition
Season	Key Factors	F	p	Axis 1	Axis 2	Key Factors	F	p	Axis 1	Axis 2
Spring	Cond	3.127	0.001	−0.892	0.083	TP	3.312	0.005	−0.647	−0.283
	DO	2.073	0.004	0.057	−0.582	Chl-a	2.454	0.029	0.027	0.515
						Cond	3.360	0.009	0.549	−0.685
Summer	TN	2.196	0.005	0.557	0.431	Subsidence age	3.265	0.005	0.590	−0.465
	Cond	2.835	0.001	0.359	−0.844
	Subsidence age	2.100	0.004	−0.614	−0.238
Autumn	TP	2.537	0.003	0.250	0.250	TP	3.051	0.004	0.322	0.467
	Chl-a	2.212	0.001	−0.073	0.035	Chl-a	3.067	0.001	0.383	−0.330
	Subsidence age	1.952	0.009	0.077	0.565	Cond	2.488	0.019	−0.915	0.040
						Subsidence age	2.019	0.044	0.315	0.240
Winter	TN	1.689	0.022	−0.414	−0.255
	DO	2.020	0.008	0.113	−0.715
	Area	1.896	0.006	−0.764	−0.145
	Subsidence age	1.360	0.104	−0.644	0.414
Average	TN	2.000	0.003	−0.487	0.072	TN	6.121	0.001	−0.677	0.295
	Chl-a	2.363	0.001	−0.075	−0.561	Chl-a	3.137	0.004	−0.341	0.622
	PH	2.001	0.004	−0.351	0.444	Cond	4.981	0.001	−0.567	−0.461
	Cond	2.584	0.001	−0.625	−0.043	DO	2.101	0.039	−0.292	0.256
	DO	1.902	0.002	−0.032	0.633	Area	2.183	0.040	0.109	0.349
	Area	2.122	0.003	0.375	0.253	Subsidence age	4.081	0.001	0.602	0.064
	Subsidence age	2.138	0.001	0.602	0.293

). For functional composition, CCA indicated environmental factors explained 45.64% of total variation, with the first two axes contributing 36.62% and 24.87% of explanatory power (Figure 7e). Three factors emerged as significant determinants: TN (F = 6.121, p = 0.001), Cond (F = 4.981, p = 0.001), and subsidence age (F = 4.081, p = 0.001) (Figure 7, Table 3).

Subsidence age significantly affected across spring, summer, and autumn, while TP, TN, and Cond showed pronounced influences in most seasons. Chl-a was an important factor during spring and autumn. Notably, functional composition was more sensitive to subsidence age and Cond than species composition, particularly in summer and autumn (Figure 6 and Figure 7, Table 3).

4. Discussion

Functional traits reflect species’ responses to environmental changes [31,67]. To understand how functional traits mediate species’ responses to these changes is crucial for predicting ecosystem dynamics and guiding conservation efforts. Therefore, this study compared macrozoobenthos community diversity across different subsidence stages of MSWs from functional traits and diversity perspectives, while investigating key influencing factors. Our findings revealed the following: (1) As the subsidence age increased, the species richness and FRic of the MSWs showed a declining trend, with biodiversity in LS being lower than IS. (2) Taxonomic diversity metrics were more sensitive to the subsidence age than functional diversity metrics. (3) Wetlands in LS exhibited significant environmental fluctuations, with seasonal variations in species composition and functional diversity being markedly greater than in wetlands in LS. This outcome validated the research hypothesis. The study elucidates the temporal evolution patterns of functional traits in macrozoobenthos in MSWs, providing a new evaluation dimension—functional diversity—for the ecological assessment of such wetlands. It holds significant theoretical value for understanding the succession mechanisms of wetland ecosystems under human-induced disturbances.

Our study reveals distinct patterns in the functional trait composition of macrozoobenthos across three subsidence stages of MSWs. The community was dominated by medium-sized organisms (9–16 mm), though the proportion of large-bodied taxa (>16 mm) increased significantly with subsidence age. Since large species typically exhibit higher individual biomass, this shift suggests enhanced secondary productivity and potential impacts on nutrient cycling through altered excretion patterns, indicating progressive ecosystem stabilization during succession [49,68,69]. Respiratory adaptations showed a clear temporal trend: while integumentary respiration remained predominant overall, branchial respirers increased markedly in LS, likely reflecting improved water quality and dissolved oxygen conditions [69]. Functional feeding groups varied significantly, with collector-gatherers peaking in MS, while shredders and predators were relatively more abundant in IS. Morphologically, non-streamlined body shapes dominated across all stages and burrowing was the prevalent life habit, probably due to stable sedimentary environments with high substrate heterogeneity [50]. Notably, the dominant functional trait combination (medium size, integumentary respiration, collector-gatherer feeding, non-streamlined shape, and burrowing habit) consistently matched the biological characteristics of Chironomidae larvae, the taxonomically dominant group across all subsidence stages. This finding aligns with the conclusion of Zhang et al. [27] regarding the persistent dominance of chironomids throughout wetlands’ development in MSWs.

Contrary to our first hypothesis, we found that both species richness and FRic of macrozoobenthos exhibited declining trends with wetland maturation (Figure 3), despite increasing ecosystem stability. This counterintuitive pattern can be attributed to several ecological processes: Chironomid larvae, the dominant taxa across all subsidence stages, possess unique life history strategies that facilitate cross-ecosystem dispersal and rapid colonization of newly formed aquatic habitats [70,71,72,73]. Functional trait analysis revealed that IS were characterized by medium-sized organisms (63.9% of community composition), dominance of integumentary respiration (84.7%—an adaptation to fluctuating water conditions), and relatively high proportions of shredders (26.1%) and predators (17.3%), likely supported by abundant organic matter inputs from surrounding agricultural lands and elevated nitrogen/conductivity levels. The high FRic (Figure 3) in IS probably resulted from both elevated habitat heterogeneity creating diverse microhabitats and resource availability supporting multiple feeding guilds. In contrast, LS exhibited environmental homogenization (stable conditions with low conductivity and high dissolved oxygen) and monodominant vegetation stands (primarily Phragmites australis and Typha angustifolia), leading to strong habitat filtering that was selected for only a few tolerant trait combinations (e.g., medium body size, burrowing habit). This successional trajectory toward reduced functional diversity despite ecosystem stabilization aligns with established theories of environmental filtering in anthropogenic wetlands [27,28,29,37], demonstrating how heterogeneous habitats can override expected biodiversity increases during ecosystem development.

Taxonomic diversity metrics exhibited greater sensitivity to subsidence age than functional diversity indices. CCA revealed that macrozoobenthos community composition was significantly influenced (p < 0.01) by multiple environmental parameters (TN, Chl-a, pH, Cond, DO, and subsidence age) (Figure 6 and Figure 7, Table 3), with their synergistic effects driving pronounced compositional changes over time. In contrast, functional diversity showed delayed responses to environmental changes, likely due to buffering effects from functional redundancy within ecosystems (Figure 5, Table A5) [38,39,74]. Although species richness declined across successional stages, dominant taxa (particularly Chironomus and Glyptotendipes) consistently maintained core functional traits including medium body size, integumentary respiration, collector-gatherer feeding and burrowing habits (Figure 4, Table A4). This functional conservatism suggests that key ecological processes may persist despite taxonomic turnover [75], demonstrating how functional redundancy buffers against short-term impacts of species loss [38]. Notably, LS showed increased proportions of large-bodied (>16 mm) and branchial-respiring taxa (Table A4), whose traits may functionally compensate for diversity declines [76,77], thereby maintaining stable FDiv. These findings collectively suggest that while taxonomic diversity reflects direct environmental filtering effects in newly formed wetlands, the relative stability of functional diversity underscores ecosystem resilience through dual mechanisms of functional redundancy and trait compensation [38,39].

Our results confirm the hypothesis that IS exhibit greater seasonal fluctuations in taxonomic diversity and functional diversity compared to mature systems (Figure 3). The initial high FRic in the newly formed wetlands was sustained by nutrient inputs from agricultural surroundings and diverse microhabitats supporting multiple functional groups. However, progressive heterogeneous habitats (e.g., Phragmites-dominated monocultures) and intensified environmental filtering (e.g., trait selection under stable low Cond/high DO conditions) gradually reduced functional space volume over time [27,28]. The elevated FEve observed in IS (Figure 3e) suggests equitable trait distribution among pioneer species under resource-abundant conditions, while its subsequent decline reflects competitive exclusion-induced functional convergence, followed by slight recovery indicating potential niche differentiation. The consistently stable FDiv (Figure 3f) likely stems from the persistence of core functional traits (medium body size, burrowing habit) in dominant taxa throughout succession, complemented by LS additions of large-bodied and branchial-respiring species [75]. IS ecosystems, characterized by unstable hydrology, pronounced microtopographic variation, and low functional redundancy, showed amplified seasonal responses due to their transient microhabitats and limited buffering capacity. In contrast, mature wetlands demonstrated dampened seasonal variability through stabilized environmental conditions, enhanced community resistance, and developed biotic interactions [11,78]. This developmental trajectory from highly dynamic pioneer systems to stable mature ecosystems illustrates how anthropogenic wetlands gradually acquire natural-like stability over decadal timescales, with the attenuation of seasonal biodiversity fluctuations serving as a key indicator of ecosystem maturation in post-subsidence wetlands.

5. Conclusions

Our study demonstrates that both coal mining subsidence and natural factors serve as key drivers shaping functional diversity patterns in macrozoobenthos communities across different stages of MSWs. During IS and MS, mining-induced disturbances and natural variability (e.g., seasonal changes) indirectly influence functional diversity by mediating dramatic fluctuations in habitat characteristics. As wetlands mature into LS, natural biological processes (particularly species life history strategies) emerge as the predominant factor governing functional diversity patterns. These findings reveal the complex interplay of anthropogenic and natural forces in structuring functional diversity within MSWs, providing novel insights into the formation and maintenance of biodiversity in these unique ecosystems. By elucidating the ecological narrative of subsidence wetlands across their developmental trajectory, our work establishes both theoretical foundations and practical guidelines for ecosystem assessment and restoration in post-subsidence wetlands, while contributing to the broader understanding of anthropogenic wetland ecology. Furthermore, we reinforce the necessity of a pluralistic approach, considering both taxonomic and functional aspects of biodiversity in ecosystem management.