Temporal variation in habitat quality shapes the distribution–abundance relationship in waterbirds at landscape scale

Abstract

Abstrakt

Vztah mezi distribucí a abundancí (DAR) je dobře známá makroekologická patrnost. Druhy, které jsou lokálně hojnější, jsou také více rozšířené. Nicméně není jasné, zda je tento vztah časově neměnný nebo jak se mohou změny projevovat. Zde zkoumáme DAR na úrovni krajiny v unikátním, avšak přehlíženém modelovém systému, tj. společenstvu vodních ptáků obývajících rybníky v rámci Biosférické rezervace Třeboňsko v České republice. Využili jsme dvanáctiletá data ze 134 intenzivně obhospodařovaných rybníků, data byla sbírána dvakrát ročně v květnu a červenci, popisujeme na nich proměnlivost ve sklonu DAR, vyjádřenou jako vztah mezi hustotou ptáků a obsazeností rybníků. Testovali jsme vliv environmentálních parametrů a vlastností ptáků na tvar tohoto vztahu, s využitím obecné mnohorozměrné regrese a průměrování modelů. Celkově jsme potvrdili očekávaný pozitivní DAR u vodních ptáků a poprvé ukazujeme konzistentní sezónní oslabování mezidruhového DAR. Domníváme se, že tento pokles sklonu vztahu je výsledkem zhoršování a homogenizace kvality prostředí později v létě, což vedlo k výraznému nedostatku vhodných rybníků pro rostoucí populace ptáků. DAR se mezi jednotlivými lety v žádném z měsíců neměnil. Také jsme zjistili, že populační trend byl nejvlivnějším prediktorem reakce jednotlivých druhů na měnící se environmentální podmínky; druhy se vzrůstající regionální početností vykazují silnější intraspecifické DAR. Naše výsledky ukazují, že sezónní zhoršování kvality prostředí významně ovlivňuje společenstva vodních ptáků tím, že mění prostorovou strukturu populací, což se odráží ve tvaru DAR. Toto zjištění má důležité praktické důsledky, zejména v lidmi využívané krajině, kde rozhodnutí o správě krajiny mohou určovat strukturu ekosystému. Podobné analytické přístupy mohou být použity k identifikaci procesů, které jsou jinak těžko zjistitelné, a tím přispět k snahám ochrany přírody.

INTRODUCTION

Species distribution patterns are a central focus of ecology and biogeography. The geographic ranges of different species are determined by their dispersal ability and the availability of environmental configurations that fit their ecological preferences (Gaston, 2003). Within range borders, some sites are more suitable than others, which leads to the uneven distribution of individuals across space (Ricklefs, 2013; Seliger et al., 2021). The inner structure of distributions can thus vary significantly and can be perceived as a function of how many localities are occupied by conspecific individuals and how densely inhabited those areas are.

The interaction between distribution and local abundance has been extensively studied (Buckley & Freckleton, 2010; Gaston et al., 2000; Hanski et al., 1993; He et al., 2002). Distribution is usually defined as geographic range or occupancy; local abundance can be denoted as population size or expressed as a unit of density. Typically, a positive correlation is observed: Species with greater local abundance tend to occupy more sites or have a wider range size relative to less locally abundant species (Gaston & Blackburn, 2000). To avoid confusion, Borregaard and Rahbek (2010) have suggested that the umbrella term “distribution–abundance relationship” (hereafter DAR) should be used for all forms of this correlation. Therefore, in this study, we use DAR to denominate the density–occupancy associations in our results and to refer to this group of related ecological phenomena.

Strong, positive DARs have been documented in a wide variety of groups, including marine fishes (Fisher & Frank, 2004), plant communities (Guedo & Lamb, 2013), and gill ectoparasites (Šimková et al., 2002). However, this pattern does not appear to be universal; in some cases, authors have found a nonsignificant or negative relationship (e.g., Freeman, 2019; Päivinen et al., 2005; Reif et al., 2006). Still, the majority of interspecific DARs are positive and statistically significant across spatial scales, taxa, habitats, and biogeographical regions (Blackburn et al., 2006).

Several theoretically diverse mechanisms have been suggested to explain the ubiquity of DARs in nature, with varying degrees of support. These include hypotheses involving resource use and availability, niche structures, metapopulation dynamics, or vital rates; no single explanation has emerged as a dominant driver (Borregaard & Rahbek, 2010; Gaston et al., 2000). Consequently, it seems likely that positive DARs are the product of multiple mechanisms acting in tandem and that their explanatory power varies depending on taxonomy, spatial scale, and other attributes—in other words, the relationship has “multiple forms” (Borregaard & Rahbek, 2010; Gaston, 1996).

Interestingly, only a limited number of studies have investigated temporal variation in the DAR, with conflicting results. Some findings reveal no significant temporal variability or trend (Blackburn et al., 1998; Caten et al., 2022; Heino & Grönroos, 2014), while others reach the opposite conclusion (Fisher & Frank, 2004; Manne & Veit, 2020; Webb et al., 2007). Notably, most of the aforementioned authors examined temporal variation in birds, so there is no clear consensus on the temporal stability of the DAR even within the same taxon. Moreover, to our knowledge, almost no studies have considered seasonal variation (but see Gaston, Blackburn, & Gregory, 1997). If the relationship's structure changes over time, and proximal causes for this variation are identified, theoretical assumptions can be assessed quantitatively.

Waterbirds form a highly specialized avian guild of particular conservation concern due to wetland loss and sensitivity to climate change (Fluet-Chouinard et al., 2023; Galewski & Devictor, 2016; Musilová et al., 2021; Pavón-Jordán et al., 2019). Although a positive DAR in waterbirds has been demonstrated in the past (Elmberg et al., 2000; Holt & Gaston, 2003), it has never been extensively studied. This is surprising, since they are a particularly fitting functional group in this context. Most waterbirds are conspicuous, which makes population sampling precise. Wetland species can occupy all sites equally well through flight, so variation in dispersal capability will not influence the structure of DARs and conflate results (Heino & Grönroos, 2014). Additionally, designating an appropriate unit of occupancy can be quite difficult due to habitat variability within grid cells or a lack of easily delineated areas (Blackburn et al., 2006; Borregaard & Rahbek, 2010). Aquatic specialists perceive lacustrine wetlands as a network of distinct patches separated by uninhabitable terrain, providing an objective and robust measure of distribution (i.e., proportion of occupied waterbodies). For these reasons, waterbird populations distributed across many enclosed waterbodies serve as a unique model system for testing the DAR.

Over the past centuries, European wetlands have undergone catastrophic changes due to anthropogenic activity. Natural wetland area has decreased by more than 50%, primarily due to land conversion for agricultural use and urbanization (Davidson, 2014; Fluet-Chouinard et al., 2023). Beyond habitat loss, high levels of eutrophication have severely damaged wetlands, harming biodiversity and disrupting ecosystem functions (Davidson et al., 2018; Finlayson et al., 2019; Verhoeven, 2014). Finally, climate change will continue to exacerbate many of these effects (Čížková et al., 2013; Finlayson et al., 2019). Artificial waterbodies have provided some relief, becoming biodiversity hotspots for taxonomically diverse species that are specialized to aquatic environments (Cantonati et al., 2020; Davidson et al., 2020; Hansson et al., 2005). For waterbirds in particular, it seems that artificial wetlands like irrigation ditches or fishponds can act as adequate alternative habitats (Čehovská et al., 2022; Pöysä et al., 2013; Sebastián-González et al., 2010).

A significant proportion of artificial wetlands in Europe are used for aquaculture (Čížková et al., 2013). In Czechia, fishponds are especially dominant, mainly for farming common carp (Cyprinus carpio; Čížková et al., 2013; Vavrečka et al., 2020). Modern intensive aquaculture practices, in conjunction with carp foraging behavior, have profound effects on fishpond ecosystems through multiple drivers tied to a well-known trophic cascade (Francová et al., 2019; Pechar, 2000; Roy et al., 2022). Nutrient loading is employed to boost fish growth rates, leading to much higher phytoplankton concentrations. Fishpond bottom sediments are dredged every year, simultaneously removing aquatic plants and destroying the habitat of invertebrate phytoplankton predators. Carp also prey upon macroinvertebrates, further reducing their abundance. Under warmer weather conditions, this increase in phytoplankton biomass, along with more pronounced carp grazing activity, leads to dramatic decreases in water quality in a matter of weeks and results in cyanobacterial blooms. Consequently, ponds used for carp farming tend to be highly eutrophic and have much lower food availability than natural waterbodies (Hátle, 2000; Kloskowski, 2011; Pechar, 2000). The intensity of these effects strongly correlates with reduced waterbird abundance at individual fishponds (Broyer & Calenge, 2010; Haas et al., 2007; Kloskowski, 2011).

The Třeboňsko region, in which we conducted our research, is the epicenter of the aquaculture industry in Czechia (Adámek et al., 2012; Hátle, 2000). At the same time, it is a protected landscape area (PLA), a Category V protected area under the IUCN, an important bird area (IBA), and (partially) a designated Ramsar site (Hátle, 2000). This dichotomy makes Třeboňsko a particularly appropriate location to examine how spatiotemporal variation in fishpond quality influences large-scale patterns of avian distributions through the lens of the DAR, given the substantial conservation value of such artificial wetland regions for both resident and migratory waterbird species.

Here, we described variation in the slope of the DAR in waterbird species in the Třeboňsko wetland complex of south Czechia and tested for the role of environmental parameters and traits of birds in determining its shape. Specifically, we investigated (1) the interspecific DAR and whether its shape changes between years and months during the breeding season, (2) how species traits affect the strength of the intraspecific DAR, and, importantly, (3) the influence of environmental parameters in shaping the interspecific relationship.

METHODS

Bird population data

We used 12 years (2005–2016) of waterbird census data from 134 fishponds (surface area 15.71 ± 36.50 ha; mean ± SD) within the Třeboňsko Biosphere Reserve, Czech Republic (GPS coordinates: 49.006, 14.763). The area covers approximately 700 km² of a wetland-forest mosaic landscape and is one of the most important sites for breeding, migration stopover, and wintering of wetland birds in Central Europe (Figure 1). Bird censuses were conducted twice a year, in May and July (henceforth the two sampling periods), the peak of the breeding season for most local wetland species. Birds were counted from a fixed number of points covering the entire water surface of each pond. Only adults were included in the count; juvenile individuals could be distinguished by differences in plumage and behavior (Madge, 2010). Survey effort varied according to total abundance, species present, and waterbody surface area, ensuring high detection probability and a complete waterbird census for each pond (Bibby, 2000; Čehovská et al., 2019; Sutherland et al., 2004). It should be noted that the term waterbird is not a taxonomic or morphological demarcation. Instead, waterbirds are typically defined as the functional group of species that are reliant on wetlands (Wetlands International, 2022). For our dataset, only nonpasserine waterbirds were included in the analysis as passerines were not extensively sampled.

[IMAGE OMITTED. SEE PDF]

We used an arbitrary but very low threshold (at least one individual on average) to remove species that were extremely rare in occurrence (see Appendix S1: Table S1). We also relied on expert opinion (PM) to retain species with high identification accuracy even if they were rarely recorded, resulting in the retention of the northern shoveler (Anas clypeata). Consequently, the subset included 31 species, representing 99.81% (n = 150,896) of total recorded individuals (Appendix S1: Table S2). To ensure this treatment did not significantly alter our results, we conducted paired t tests to compare the interspecific DAR linear model of each month-year census, using either the complete dataset or the selected subset as data input. No significant difference was found in the slope (t₂₃ = 0.951, p = 0.351) or model fit (t₂₃ = 0.745, p = 0.464) of the two sets of DAR models.

Finally, during preliminary data analysis, we observed that quite a few extremely rare species, represented by just one recorded individual in a specific year, were present in only one of the two sampling periods. This meant that their inclusion had a disproportionate impact on the interspecific DARs, given the high level of stochasticity implicit in the presence of a single individual. We therefore created a separate dataset based on the subset of species that were recorded in both May and July for that specific year of fieldwork and tested whether this had an effect on the calculated temporal variation in the DAR within years.

Species data

To elaborate on the significance of interspecific variability in the context of the DAR, we collected data on species' ecological and morphological traits, as well as their regional population trend.

Due to large gaps in national population trend estimates for the species in our dataset, we used regional estimates as a proxy for the more localized population trend. We primarily relied on the Waterbirds Populations Portal (WPP; Wetlands International, 2022) as it contains flyway-specific population trend estimates and thus is more relevant to our population data. The trend estimates are based on species-specific abundance data, compiled from waterbird population censuses, the scientific literature, and expert opinion. For species recorded in Třeboňsko that are not classified as waterbirds and thus are not found in the WPP (i.e., raptors that are aquatic specialists), we relied on data from the Pan-European Common Bird Monitoring Scheme (European Bird Census Council [EBCC], 2022) and the IUCN Red List (IUCN, 2022).

To account for functional diversity in migration and breeding, we compiled data from local sources that accurately document the behavior of waterbird species in our study area (Cepák et al., 2008; Delaney et al., 2009; Scott & Rose, 1996; Šastný & Hudec, 2016). We assigned species to three migratory categories (residents, short-distance migrants, and long-distance migrants); they were also defined as early or late breeders depending on whether most individuals start laying eggs before or after the first of May, respectively. Body mass (in grams, mean of both sexes) and the centroid latitude of each species' range were extracted from AVONET (Tobias et al., 2022). The degree of habitat specialization can also vary among species, so we calculated a habitat species specialization index (HSSI), based on data from Storchová and Hořák (2018), by adapting the SSI metric for occupied classes developed by Julliard et al. (2006). In a simplified way, SSI is calculated as the proportion of occupied habitat classes out of all possible habitats. We assumed equal densities (presence) in occupied habitats and null density in unoccupied ones (absence).

We determined each species' trophic niche based on information about the feeding behavior of waterbirds in Czechia (Kear, 2005; Musilová et al., 2021; Snow et al., 1998; Šťastný & Hudec, 2016) to accommodate possible intraspecific variation in food source preference across species ranges. For species without available localized dietary data, we used species-level feeding guild data from Elton Traits (Wilman et al., 2014). Our species assemblage covered four dietary groups (herbivores, invertivores, omnivores, and piscivores). Finally, we included a diet species specialization index (DSSI) to illuminate differences in the response of dietary specialists compared to generalists. This was done by extracting foraging data from Elton Traits and using a method developed by Morelli et al. (2021). The DSSI is computed similarly to the Gini coefficient of inequality, a measure of statistical dispersion. For example, a dietary generalist exploits several food sources to a varying degree. It could thus be said that by utilizing multiple dietary categories in a more equal fashion, the species exhibits higher dietary evenness across food types, and it would consequently obtain a low DSSI score (Morelli et al., 2019). The final traits dataset can be found in Appendix S1: Table S5.

Environmental data

In fishponds, water quality is tightly correlated with food availability for waterbirds (Fox et al., 2024; Musil, 2006). Transparency as a measure of water quality can thus act as a simplistic but reliable proxy for food resources. We recorded the water transparency in a subset of fishponds during each bird census using a 20-cm diameter Secchi disk.

Annual fluctuations in air temperature, as well as the typical seasonal increase from spring to summer, could also regulate temporal variation in the DAR. Therefore, monthly mean temperature data for the South Bohemia region was collected from the Czech Hydrometeorological Institute (Czech Hydrometeorological Institute, 2022). Additionally, the North Atlantic Oscillation (NAO) index is commonly used as a proxy for winter harshness, which can affect environmental conditions and bird behavior (Ottersen et al., 2001). We compiled the mean monthly NAO index from the USA's National Weather Service (downloaded from ).

Statistical analyses

We performed all analyses in R, version 4.2 (R Core Team, 2022). All models were computed using maximum likelihood estimations. To implement mixed-effects and generalized least squares linear models, we used the nlme package (Pinheiro et al., 2007).

We first calculated the density of each species at each fishpond (local abundance over fishpond surface area), averaged across all occupied sites, as is standard practice when defining mean local density (Borregaard & Rahbek, 2010; Gaston et al., 1998). Occupancy was simply denoted as the proportion of fishponds occupied by each species. Preliminary regression analyses resulted in non-normal residuals, so both variables were log-transformed (ln(x) + 1) in all models to abide by normality and homoscedasticity assumptions.

Given the dynamic and potentially multicausal nature of the DAR, it is unclear whether variation in occupancy drives variation in local abundance or vice versa (Blackburn et al., 2006; Borregaard & Rahbek, 2010). Therefore, to establish the statistical properties and significance of the interspecific DAR for each census, we arbitrarily assigned species occupancy as the response variable and regressed occupancy against the local mean density of each species. The gradient of the regression equation (hereafter the DAR slope) thus represents the direction and strength of the relationship. In all subsequent analyses where the DAR slope is used as the dependent variable, the models are weighted by the SE associated with each slope value to incorporate the uncertainty implicit in each slope within the model structure.

To investigate how DARs are affected by temporal variation, the slope of each relationship acted as the response variable, modeled against month, year, and their interaction (month: year). Additionally, a multivariate linear regression model was deployed, incorporating potentially important environmental variables to determine whether they significantly influence the DAR slope and how they interact with temporal variation itself. For each month-year census, we tested the effect of mean monthly temperature, mean water transparency across all ponds, and the NAO index. These three variables are likely to covary due to their seasonal associations. For example, temperatures are lower on average in May relative to July, and higher water transparency is associated with lower temperatures. To control for this nonindependence and to accurately model which predictors are correlated with variation in slope values in isolation, we included all dependent variable interactions in the model.

For this analysis, and for modeling the influence of species traits (see below), we used a model-averaging statistical approach. After the model set with all possible variable combinations is generated, each model's output (corrected Akaike information criterion [AIC_c] and Akaike weight [w], along with parameter values) is used to calculate model-averaged parameter coefficients. We excluded models that were extremely unlikely to be the best model; that is, they were not part of the ranked model set up to a cumulative w value of 0.95 (representing 95% confidence that the model set contains the best model; Burnham et al., 2011). To calculate the model-averaged variable coefficients and their associated CIs, we standardized the parameters by their partial standard deviations contained within each model (Cade, 2015). These standardized parameter estimates are then weighted by the Akaike weight of each model and averaged across models, resulting in directly comparable coefficients for numerical variables and the statistical significance for all predictors and groups within categorical variables (Cade, 2015; Galipaud et al., 2017). All model sets were analyzed with MuMIn (Bartón, 2022).

To further elucidate which factors play a significant role in the seasonal variation of the interspecific DAR (defined as the change in slope between May and July of each year), we used an additional set of linear mixed-effects models to individually test all environmental variables against month. Year and species were included as random effects to control for other sources of variation. All regressors with heteroscedastic residuals were log-transformed.

Beyond temporal differences in mean water transparency, variation in water quality among fishponds could also be a significant structural driver of the DAR. We therefore analyzed how the SD of water transparency varied temporally (month and year). As the transparency readings data only had partial cover, we designed a linear model to consider sampling-specific variables that may influence the resulting SD, namely, the number of ponds measured in each period and the mean and median surface area of those ponds. In some years, the set of ponds with recorded transparency readings differed between May and July, so we limited this analysis to fishponds that were measured in both months within each year (n = 125).

For the interspecific analysis incorporating species traits and population trends, we first calculated intraspecific DAR slopes for each species, one for each sampling period. We then modeled this slope against all species-specific data. As the directionality of the DAR is unclear (Gaston, Blackburn, & Lawton, 1997; Holt & Gaston, 2003), this approach was preferred instead of setting density or occupancy as the dependent variable a priori. One low-abundance species (the Caspian gull, Larus cachinnans) was an extreme outlier in its slope for both sampling periods, possibly because Czechia is close to the edge of its range, so we excluded it from this analysis. To ensure phylogenetic nonindependence in the traits of closely related species did not bias our results, we used the phylogenetic generalized least squares (PGLS) method, incorporating a Brownian correlation matrix to control for species relatedness. We generated 2000 trees of our phylogeny subset from (Jetz et al., 2012). We built a consensus tree using phytools (Revell, 2012) and generated the correlation matrix with ape (Paradis & Schliep, 2019). Two PGLS models were thus constructed, one for each month, with the slope of the DAR as the dependent variable and all species-specific variables as predictors. Body mass data were log-transformed to ensure normality, and all variables were mean-standardized to allow for quantitative comparison in the model-weighted output.

Even though many trait pairs are not independent (e.g., dietary guild and dietary specialization), and some predictors were significantly correlated, no variable pair exhibited strong collinearity (r > 0.6) or had a variance inflation factor (VIF) close to 10 (max. VIF = 7.161), which is considered a threshold over which collinearity affects model output (Kutner et al., 2004).

RESULTS

Overall, 134 ponds were occupied over the 12 years of census data, and 31 species from 10 avian families are represented in our dataset. Total abundance across all species was much higher in July than in May, with 42.08% (or 2184.4) more birds on average within years. For 22 out of 24 censuses, interspecific DARs were significant at α = 0.05 (Appendix S1: Table S3); mean model fit (R²) was 0.393.

Using the restricted subset of within-year common species marginally reduced the mean number of species recorded yearly (from n = 29.42 to n = 26.46) but did not significantly diminish total waterbird abundance (from 99.81% to 99.75%, n = 150,804). This data treatment was retained for all subsequent analyses to ensure results were directly comparable and allow for robust interpretations. Seasonal DAR slopes were strongly positive in both May and July (0.81 ± 0.16 and 0.54 ± 0.16, respectively; mean ± SD), and the relationship was weaker in July in almost all years (Figure 2).

[IMAGE OMITTED. SEE PDF]

Temporal models

The full temporal linear model (month + year) had rather low statistical power (adjusted R² = 0.383, p = 0.103); month is highly statistically significant (p = 0.002), with the slope decreasing in July on average (β = −1.321). Year had no influence on the model (p = 0.567 for all years, based on ANOVA), and no individual years showed a significant correlation with the DAR slope relative to 2005. Within each year of sampling, the mean number of occupied ponds per species decreased from 17.86 to 15.70, an 11.2% decrease, while the mean density in each occupied pond increased by 14.5% (from 4.14 to 4.74 individuals/ha).

Over 12 years, we found a mostly stable annual trend in the correlation between density and occupancy in both sampling periods. The difference in slope between May and July seems to be decreasing over time (Figure 2), but this trend is not statistically significant when testing for an association between DAR slope and year while controlling for sampling period (F_1,21 = 0.021, p = 0.887). The position of most species along the density–occupancy axes of the log-linear relationship changes substantially over time, with no identifiable directionality (Figure 3). However, by regressing the slope of the DAR against the density and occupancy of each species in each sampling period separately and controlling for their interaction, we could examine whether either variable is more influential on the resulting annual slope. In May, the strength of the DAR is not influenced to a disproportionate degree by either of its components as neither predictor is significantly correlated with the slope, although density is marginally nonsignificant (β = −0.137, p = 0.073). On the other hand, a strong negative correlation of density with the DAR slope is observed in July (β = −0.277, p = 0.0149), while occupancy remains nonsignificant (β = −0.023, p = 0.549). These results indicate that the density of species in fishponds has a substantially greater influence on the strength of the DAR relative to their occupancy (Table 1), and to some extent is responsible for the evident seasonal variation.

[IMAGE OMITTED. SEE PDF]

TABLE 1 The results of a linear mixed-effects model, testing for differences between sampled time periods (May and July) in variables associated with waterbird populations and occupied fishponds, for each species separately (n = 703).

Variable	β	t	p
Bird Occupancy	−0.0275	−0.6683	0.5045
Bird Density	0.0629	5.9641	<0.001
Mean Fishpond SA	−0.1227	−2.2681	0.0241
Mean Water Transparency	−0.6126	−21.2084	<0.001
Water Transparency SD	−0.3571	−13.0520	<0.001

In the interspecific linear mixed-effect models, occupancy and density have opposite relationships with monthly variation when controlling for annual fluctuations and species assemblage differences. Density is strongly and positively correlated with sampling period, whereas occupancy is negatively and insignificantly correlated (Table 1). In other words, species densities increase significantly in July compared to May, and this is not simply due to a large decrease in occupancy naturally causing increased densities in the remaining occupied fishponds. Furthermore, the mean surface area of fishponds utilized by each species is also significantly lower in July, exacerbating the higher densities (Table 1). This outcome corroborates the results of the previous analysis on the relative influence of the two variables on the interspecific slope.

Environmental models

We used the aforementioned model subsetting technique to disentangle the influence of tightly correlated environmental variables, such as water transparency and temperature, on the DAR. The model-weighted parameters of the top-ranking models show that mean water transparency was the most influential and only statistically significant variable (β = 0.554, p = 0.032), with month, temperature, and NAO being of minimal importance (Table 2). Higher water transparency is strongly and positively correlated with a greater DAR slope, suggesting that it is the primary link between variation in the relationship and fluctuating environmental conditions (Table 2).

TABLE 2 The output of the linear mixed-effects model set, testing for the influence of time period and environmental variables on the slope of the interspecific distribution–abundance relationship (DAR) for waterbird species, with model-averaged coefficient means and 95% CIs.

Variable	Coefficients
Mean	2.5%, 97.5% CI
Transparency	0.554*	0.047, 1.062
Temperature	−0.168	−0.768, 0.433
Month:Transparency	0.099	−0.221, 0.418
Month
May	Reference
July	−0.087	−0.737, 0.562
Transparency:Temperature	0.025	−0.293, 0.343
NAO index	−0.020	−0.341, 0.300

When comparing the set of fishponds occupied by each species for each sampling period, the mean and SD of water transparency are negatively correlated with month (Table 1); fishponds have lower and more homogeneous transparency in July compared to May (Figure 4). Additionally, when controlling for species assemblage and annual differences, bird density across species is significantly and positively correlated with fishpond water transparency in both May (β = 1.102, p < 0.0001) and July (β = 0.953, p < 0.0001).

[IMAGE OMITTED. SEE PDF]

The speedy reduction in fishpond quality variance between May and July is reaffirmed in the analysis of water transparency SD (Appendix S1: Table S4). We show that month (F_1,223 = 86.264, p < 0.001) remains a dominant factor in the overall variance of transparency even after accounting for sampling effort bias. On average, SD strongly decreases in July (β = −20.976) but remains constant among years (F_1,11 = 1.424, p = 0.287).

Interspecific trait models

Due to the relatively low number of species and the large degree of heterogeneity in intraspecific DARs (Blackburn et al., 2006; Holt & Gaston, 2003), we did not expect that traits would account for a sizable degree of residual variation around the relationship. We instead contrasted the relative importance of the tested functional and demographic traits between the two sampled months. In other words, we examined whether the observed environmental changes between sampling periods would differentially influence species based on their trait groupings.

Regional population trend was the most influential variable in May, as species with increasing population had significantly stronger intraspecific DARs (Appendix S1: Table S6). In July, invertivores had significantly weaker intraspecific DAR slope than other dietary guilds. HSSI was positively correlated with intraspecific slope in both sampling periods, and it was marginally nonsignificant in July. Interestingly, migratory status had a weak but positive association with slope in May, but it was negative and marginally nonsignificant in July. No other clear patterns emerged, as the remaining variables were of lesser importance given low coefficients or wide CIs (Appendix S1: Table S6). Broadly speaking, it therefore seems that different functional groupings of waterbird species are equally sensitive to high rates of environmental change in their habitat when it comes to their landscape-level spatial distribution.

DISCUSSION

We found a recurrent, positive correlation between local density and occupancy in the waterbird community inhabiting Třeboňsko fishponds. We also reveal a strong, consistent decline across time periods within a year (but not across years) in the strength of the interspecific DAR, something that, to our knowledge, has never been shown quantitatively. Due to the controlled nature of the comparative seasonal analysis, this weakened density–occupancy correlation is likely caused by decreased water transparency and homogenization of habitat quality at the landscape level. Finally, we found that regional population trend explained the greatest amount of variation in the DAR slope in May. In July, when the water quality is low, intraspecific DAR slopes were lower especially for invertivores, species that rely on highly transparent water for feeding (Fox et al., 2024). Such findings suggest that spatial resource distribution and population sizes substantially shape the relationship between occupancy and local density in waterbirds.

Temporal trends

We found a significant, positive interspecific DAR in almost all analyses across years. This result was not unexpected, as similar positive relationships are well-documented in birds in general (Manne & Veit, 2020; Snell Taylor et al., 2020; Webb et al., 2007) and in a few instances in waterbirds specifically (Elmberg et al., 2000; Holt & Gaston, 2003). The relatively weak correlation between the two components of the DAR in birds, resulting in substantial residual variation, is also a common occurrence (Caten et al., 2022; Holt & Gaston, 2003), most likely due to the wide range of ways that interspecific differences and habitat structure influence this relationship (Borregaard & Rahbek, 2010; Holt et al., 2004).

We have shown for the first time that seasonal variation in the interspecific DAR exists within a species assemblage and that this is consistent across years. In our analyses, this manifests as a significant decline in the slope of the relationship across sampling periods. If the DAR was temporally invariant, we would expect that an increase in mean bird densities would be accompanied by a corresponding increase in the number of occupied fishponds. At Třeboňsko, after controlling for annual variation, the mean density of species increases between the May and July time periods, while the occupancy remains relatively constant on average. The dramatic rise in species abundances in July, in concert with a minor decrease in the mean surface area of occupied fishponds, resulted in the observed density patterns. Therefore, static occupancy combined with genuine increases in density led to a different DAR structure as the breeding season progressed.

Given its consistency among years, the weakened interspecific correlation between local density and occupancy should logically be attributed to one or more drivers that substantially change in the same period (May to July). Our results highlight how fishpond water transparency is tightly linked to changes in the relationship. Fishpond transparency had the highest relative importance and was strongly correlated with slope; its mean and SD were significantly higher in May than in July, following the same pattern observed in DAR slope values.

Water transparency is a solid proxy variable for water quality and thus habitat quality (Arzel et al., 2020; Francová et al., 2019; Pechar, 2000). In particular, higher water turbidity is a strong signifier of the trophic cascade caused by carp farming, especially in warmer temperatures. Pechar (2000) documents this “dynamic multiple-feedback process” in Třeboňsko fishponds, showing how the regional intensity of aquaculture leads to worsening water quality. This negative effect is also reflected in our analyses: Water transparency was by far the strongest correlate of DAR slope variation. We can thus infer that the positive correlation between DAR slope and water transparency provides a causal link between the exploitation of the wetland ecosystem for fish production and seasonal variation in the relationship. The increasing number of low-quality (low-transparency) fishponds acts as a decisive constraint on the more abundant waterbird population in July, preventing species from occupying additional fishponds. This forces higher spatial aggregation in a relatively small number of ponds that waterbirds perceive as better habitats. It is reasonable to expect that waterbirds would have such a dramatic response to a decrease in food resources (in terms of quantity as well as foraging capability in more turbid water), especially in the breeding season when their energy demands are quite high (Čehovská et al., 2022; Patterson, 1976; Rotella & Ratti, 1992). The higher populations in July could have various causes. There is increased postbreeding, molting concentration of some species, for example, the common pochard (Aythya ferina; PM). Later in the season, after incubating the clutches hidden in vegetation, birds utilize water surface more frequently. Late migrants to the area could also increase the bird numbers estimated in July, even though we do not consider this explanation very influential. In the same vein, it could be a result of postbreeding or molting movements, which could include both local movements within study regions as well as movements within species flyways.

Over the 12-year period of our data collection, the DAR slope exhibits a mostly stable annual trend in both months. Across species, mean total abundance and occupancy per year remain relatively constant, with no noticeable trend in either sampling period, and mean water transparency does not vary among years. Additionally, there does not appear to be a clear pattern of species shifting their annual position on the DAR regression line. The lack of substantial annual variation in these factors indicates that the observed seasonal restructuring of the relationship is indeed caused by worsening habitat quality during each breeding season and is not the product of broader, long-term shifts in waterbird populations or environmental conditions.

Our results may explain why previous authors have reached conflicting conclusions regarding the temporal stability of the DAR. It seems likely that this disparity may be related to differences in the magnitude of environmental change or disturbance present in the areas sampled by each study. Caten et al. (2022), who conducted the most comprehensive analysis of temporal variation, found that relationships were constant for all studied taxa; they speculate this may be because they analyzed populations inhabiting protected areas with no anthropogenic pressures. In two other cases, analyses of the annual trend in the DAR in terrestrial birds also found the relationship to be temporally invariant, based on data from relatively stable ecosystems (Suhonen & Jokimäki, 2019; Zuckerberg et al., 2009). On the other hand, virtually all studies where DAR strength was found to be variable utilized datasets from areas experiencing widespread environmental change during data collection. The reported primary negative factors are diverse: agricultural intensification (farmland birds; Webb et al., 2007), postfire succession (grassland plants; Guedo & Lamb, 2013), and high fishing levels (marine fishes; Fisher & Frank, 2004). Finally, Manne and Veit (2020) found substantial changes in the DAR slope over a 40-year period in North American terrestrial birds, with opposing trends in permanent residents and long-distance migrants. They propose that this divergence is evidence that climate change and other anthropogenic impacts have significant and varying influence on interspecific relationships. In our results, the decline in the habitat suitability of fishpond habitats acts as the comparable destabilizing agent, leading to a weakened DAR in waterbirds in just a 2-month period. Even though almost all previously mentioned studies examined annual and not seasonal variation, it is quite clear that, at any temporal scale, the presence and magnitude of ecological fluctuations would determine whether the relationship varies over time.

Conceptually, our findings favor Borregaard and Rahbek's (2010) framework for how the DAR functions, where abundance and occupancy are connected at a structural level, but at the same time, each component can be independently influenced in a dynamic fashion. These structural and dynamic mechanisms are most likely always acting in concert, depending on the spatial scale being examined (Borregaard & Rahbek, 2010; Holt et al., 2002). Under normal conditions, this leads to a temporally stable DAR, even if only one of the two interdependent variables is affected by a change in the species' environment (Gaston & Blackburn, 2000; Suhonen & Jokimäki, 2019). However, when rapid and drastic changes disproportionately influence abundance or occupancy, the relationship can change shape. Therefore, DARs are not always temporally stable, and substantial ecosystem destabilization can lead to variation in the relationship inter- or intra-annually. In our findings, the decline and homogenization of habitat quality is the primary driver, creating the conditions for a pattern of variability in the relationship. Potentially, by investigating the shapes of DARs, we can detect important environmental instabilities that are not registered by other means.

Interspecific variation

The relatively small number of species in our dataset, as well as the high functional similarity among waterbird species, statistically limited the probability that large differences between traits would be detected. Additionally, in previous studies exploring similar questions in birds, most authors have reported that traits did not explain considerable residual variation around the DAR (Heino & Grönroos, 2014; Holt & Gaston, 2003; Snell Taylor et al., 2020; Verberk et al., 2010). Due to the seemingly complex landscape of interconnected drivers underlying the relationship, simple, independent predictors are not theorized to be particularly insightful (Blackburn et al., 1997). Our results support this assertion because most tested traits were of marginal importance overall. In May, regional population trend was the most influential species trait. The significantly higher DAR slope in species with increasing population trends found in May is intriguing. Presumably, species with regionally increasing populations have higher reproductive potential, which can lead to more individuals during the breeding season. Higher numbers could lead to the mass effect, that is, the spreading of birds across the region partly independently from species habitat preferences, which explains why a similarly strong signal was not captured by the specialization indices. Abundant species are also likely more capable of exploiting environments under pressure, like a decrease in wetland area or higher eutrophication (Čížková et al., 2013; Lehikoinen et al., 2016; Musilová et al., 2015). Note that our data only include adult birds, so local population increases would rather inform about the higher probability of detecting individuals outside breeding (females not hidden in the vegetation) or the presence of nonbreeding birds.

Interestingly, in July, invertivores exhibited a significantly weaker intraspecific DAR slope than other dietary guilds. The structure of aquatic invertebrate communities, and consequently their availability to waterbirds, may be influenced by the nutrient levels in the water (Fried-Petersen et al., 2020; van der Lee et al., 2021). Additionally, brownification is considered a key factor driving the decline in aquatic invertebrate abundance (Arzel et al., 2020; Lind et al., 2018). Specifically at fishponds, low-transparency waters are particularly poor in invertebrate abundance (Francová et al., 2019; Kloskowski, 2011; Musil et al., 1997; Pechar, 2000). In addition, high fish stocks limit the availability of invertebrates for breeding waterbirds via interspecific competition (Musil, 2006; Musil et al., 2011). In contrast, the effect of water quality on herbivores appears to be lower, as their populations were reported to increase in fishponds (Fox et al., 2024; Musilová et al., 2021). Considering the aforementioned findings, it is not surprising that invertivores were the most affected by homogenization and the worsening of water quality in July, which presumably led to a weaker DAR.

Dietary specializations, as estimated by DSSI, did not significantly affect the DAR. However, we observed a different situation for species habitat specializations. In both sampling periods, the HSSI correlated positively with intraspecific DAR slopes, being almost significant in July. Habitat specialization is an important determinant of DAR (Borregaard & Rahbek, 2010; Verberk et al., 2010), as the level of specialization potentially affects both occupancy and local density of species. Our results suggest that specialists show tighter links between occupancy and density, especially in July. This implies that such species are more confined to specific sites or sites of specific, perhaps higher, quality, which can maintain certain numbers of individuals. Under such a scenario, increasing local populations of specialists leads to larger occupancy of suitable habitats.

Species with more southerly ranges were negatively correlated with DAR strength in July. Such a finding is difficult to explain. We can speculate that southern species could start breeding later in the breeding season, which means their spatial distribution in July is still limited by nesting places. Interestingly, southern species in our dataset (Appendix S1: Table S5) frequently breed colonially (e.g., grey heron [Ardea cinerea], black-crowned night heron [Nycticorax nycticorax]) or represent a nonbreeding population (great egret [Ardea alba]). Such species theoretically have clustered spatial distribution (Emlen, 1952), and nonbreeding birds also likely have lower energetic requirements (Bryant & Tatner, 2008), factors that weaken the link between local abundance and occupancy.

Still, it seems that the general deterioration of the habitat between sampling periods overpowers much of the variation in the response of species with distinct dietary specializations. Our results suggest which ecological types of species are likely sensitive to temporal changes in water quality at fishponds. However, at the same time, they show that environmental changes of such an extent have the potential to influence whole waterbird assemblages independently of their ecological traits.

CONCLUSIONS

In cases where a positive DAR is observed, a decrease in distribution would likely lead to a decline in local abundance (or vice versa), resulting in rapid population declines (the double jeopardy effect; Lawton, 1994). Considering the widespread presence of the DAR in many taxa (Blackburn et al., 2006; Caten et al., 2022), efforts have been made to utilize the association between local abundance and distribution in conservation practice. For instance, the relationship forms the basis for modeling approaches that estimate abundance based on presence–absence data (Latham et al., 2014; Tovo et al., 2019). In a similar vein, if the DAR is sensitive to environmental instability, it can serve as a powerful explanatory tool in the struggle against biodiversity loss. Any change in the shape of the relationship can make evident precisely how an external driver, anthropogenic or otherwise, influences species distributions, best exemplified by Fisher and Frank (2004). This may not be apparent if only one of the two components of the relationship is measured.

In our specific example of the Třeboňsko waterbirds, species occupancy was not declining, and the number of available fishponds did not change. Only through studying the interspecific DAR, instead of occupancy and abundance separately, were the underlying effects constraining expansion in the landscape made explicit. We have identified that, as the breeding season progresses, many carp ponds become unsuitable habitat for wetland birds due to low water quality. Our results clearly indicate how fishpond management can change the spatial distribution of birds, which is then reflected in a macroecological pattern. Such intensively managed landscapes have a huge conservational potential as the environmental conditions influencing bird populations can be purposefully set up by humans. Water quality and the state of littoral stands around fishponds (Maceda-Veiga et al., 2017) are crucial determinants of prosperity for waterbird populations as they provide them with feeding and breeding biotopes (Fournier et al., 2021; Paracuellos & Tellería, 2004). Improving these ecosystem components through conservation-driven management, especially by increasing the number of fishponds that retain high water quality throughout the breeding season, would therefore prove highly beneficial for waterbird species.

The underlying spatial variation in species populations, typically not captured by low-resolution habitat suitability analyses, can provide important insights into how species occupy space (Hořák et al., 2022; Ricklefs, 2013). By measuring changes in the DAR seasonally, we were able to quantitatively assess what factors limit species distributions in a freshwater ecosystem with seemingly abundant suitable habitat but with a high degree of anthropogenic impact. The instability of DAR, grounded in macroecological theory, enhances our understanding of how the spatial distribution of resources shapes species assemblages across space and time. Its practical application can lead to more effective identification of potential conservation issues, thereby facilitating relevant management strategies.

AUTHOR CONTRIBUTIONS

Conceptualization: Constantinos Charalambous, David Hořák, and Mathilde Legoguelin. Conducting the research: Constantinos Charalambous, David Hořák, Petr Musil, and Zuzana Musilová. Data analysis: Constantinos Charalambous, Mathilde Legoguelin, Petr Musil, and Zuzana Musilová. Preparation of figures and tables: Constantinos Charalambous. Developing methods, data interpretation, writing: Constantinos Charalambous, David Hořák, Mathilde Legoguelin, Petr Musil, and Zuzana Musilová.

ACKNOWLEDGMENTS

We are very grateful to everyone who contributed to field data collection. We also thank Dr. Jeff Wesner and two anonymous reviewers for their thoughtful comments and constructive criticism of earlier versions of the manuscript. This work was supported by the Technology Agency of the Czech Republic (grant nos. SS01010280 and SS06010142).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT

Data and code (Charalambous, 2024) are available from Zenodo: .

References

Adámek, Z., O. Linhart, M. Kratochvíl, M. Flajšhans, T. Randák, T. Policar, J. Masojídek, and P. Kozák. 2012. “Aquaculture in The Czech Republic in 2012: Modern European Prosperous Sector Based on Thousand‐Year History of Pond Culture.” Aquaculture Europe 37(2): 5–14.

Arzel, C., P. Nummi, L. Arvola, H. Pöysä, A. Davranche, M. Rask, M. Olin, et al. 2020. “Invertebrates Are Declining in Boreal Aquatic Habitat: The Effect of Brownification?” Science of the Total Environment 724(July): 138199. https://doi.org/10.1016/j.scitotenv.2020.138199.

Bartón, K. 2022. “MuMIn: Multi‐Model Inference, Version 1.47.1.” https://cran.r-project.org/web/packages/MuMIn/index.html.

Bibby, C. J., ed. 2000. Bird Census Techniques, 2nd ed. London; San Diego, CA: Elsevier.

Blackburn, T. M., P. Cassey, and K. J. Gaston. 2006. “Variations on a Theme: Sources of Heterogeneity in the Form of the Interspecific Relationship between Abundance and Distribution.” Journal of Animal Ecology 75(6): 1426–1439. https://doi.org/10.1111/j.1365-2656.2006.01167.x.

Blackburn, T. M., K. J. Gaston, J. J. D. Greenwood, and R. D. Gregory. 1998. “The Anatomy of the Interspecific Abundance‐Range Size Relationship for the British Avifauna: II. Temporal Dynamics.” Ecology Letters 1(1): 47–55. https://doi.org/10.1046/j.1461-0248.1998.00005.x.

Blackburn, T. M., K. J. Gaston, and R. D. Gregory. 1997. “Abundance‐Range Size Relationships in British Birds: Is Unexplained Variation a Product of Life History?” Ecography 20(5): 466–474. https://doi.org/10.1111/j.1600-0587.1997.tb00414.x.

Borregaard, M. K., and C. Rahbek. 2010. “Causality of the Relationship between Geographic Distribution and Species Abundance.” Quarterly Review of Biology 85(1): 3–25. https://doi.org/10.1086/650265.

Broyer, J., and C. Calenge. 2010. “Influence of Fish‐Farming Management on Duck Breeding in French Fish Pond Systems.” Hydrobiologia 637(1): 173–185. https://doi.org/10.1007/s10750-009-9994-3.

Bryant, D. M., and P. Tatner. 2008. “Energetics of the Annual Cycle of Dippers Cinclus cinclus.” Ibis 130(1): 17–38. https://doi.org/10.1111/j.1474-919X.1988.tb00952.x.

Buckley, H. L., and R. P. Freckleton. 2010. “Understanding the Role of Species Dynamics in Abundance‐Occupancy Relationships.” Journal of Ecology 98(3): 645–658. https://doi.org/10.1111/j.1365-2745.2010.01650.x.

Burnham, K. P., D. R. Anderson, and K. P. Huyvaert. 2011. “AIC Model Selection and Multimodel Inference in Behavioral Ecology: Some Background, Observations, and Comparisons.” Behavioral Ecology and Sociobiology 65(1): 23–35. https://doi.org/10.1007/s00265-010-1029-6.

Cade, B. S. 2015. “Model Averaging and Muddled Multimodel Inferences.” Ecology 96(9): 2370–2382. https://doi.org/10.1890/14-1639.1.

Cantonati, M., S. Poikane, C. M. Pringle, L. E. Stevens, E. Turak, J. Heino, J. S. Richardson, et al. 2020. “Characteristics, Main Impacts, and Stewardship of Natural and Artificial Freshwater Environments: Consequences for Biodiversity Conservation.” Water 12(1): 260. https://doi.org/10.3390/w12010260.

Caten, C. T., L. Holian, and T. Dallas. 2022. “Weak but Consistent Abundance–Occupancy Relationships across Taxa, Space and Time.” Global Ecology and Biogeography 31(5): 10. https://doi.org/10.1111/geb.13472.

Čehovská, M., S. Kattainen, V.‐M. Väänänen, A. Putaala, and P. Nummi. 2022. “Compensating Freshwater Habitat Loss—Duck Productivity and Food Resources in Man‐Made Wetlands.” European Journal of Wildlife Research 68(3): 35. https://doi.org/10.1007/s10344-022-01577-8.

Čehovská, M., P. Musil, Z. Musilová, K. Poláková, and J. Zouhar. 2019. “Diving Duck Census Efficiency Based on Monitoring of Individually Marked Females: The Influence of Breeding Stage and Timing of Census.” Bird Study 66(2): 198–206. https://doi.org/10.1080/00063657.2019.1653823.

Cepák, J., P. Klvaňa, J. Škopek, L. Schröpfer, M. Jelínek, D. Hořák, J. Formánek, and J. Zárybnický. 2008. Atlas migrace ptáků České a Slovenské republiky (Czech and Slovak bird migration atlas). Praha: Aventium.

Charalambous, C. 2024. “Trebonsko DAR Analysis Data and Code.” Zenodo. https://doi.org/10.5281/zenodo.14144782.

Čížková, H., J. Květ, F. A. Comín, R. Laiho, J. Pokorný, and D. Pithart. 2013. “Actual State of European Wetlands and Their Possible Future in the Context of Global Climate Change.” Aquatic Sciences 75(1): 3–26. https://doi.org/10.1007/s00027-011-0233-4.

Czech Hydrometeorological Institute. 2022. Monthly Mean Regional Air Temperature. CHMI Portal. https://www.chmi.cz/historicka-data/pocasi/uzemni-teploty?l=en#.

Davidson, N. C., L. Dinesen, S. Fennessy, C. M. Finlayson, P. Grillas, A. Grobicki, R. J. McInnes, and D. A. Stroud. 2020. “Trends in the Ecological Character of the World's Wetlands.” Marine and Freshwater Research 71(1): 127. https://doi.org/10.1071/MF18329.

Davidson, N. C., E. Fluet‐Chouinard, and C. M. Finlayson. 2018. “Global Extent and Distribution of Wetlands: Trends and Issues.” Marine and Freshwater Research 69(4): 620. https://doi.org/10.1071/MF17019.

Davidson, N. C. 2014. “How Much Wetland Has the World Lost? Long‐Term and Recent Trends in Global Wetland Area.” Marine and Freshwater Research 65(10): 934. https://doi.org/10.1071/MF14173.

Delaney, S., D. A. Scott, D. A. Timothy Dodman, and Stroud, Wetlands International, and Wader Study Group, eds. 2009. An Atlas of Wader Populations in Africa and Western Eurasia. Wetlands International: Wageningen.

Elmberg, J., K. Sjöberg, H. Pöysä, and P. Nummi. 2000. “Abundance‐Distribution Relationships on Interacting Trophic Levels: The Case of Lake‐Nesting Waterfowl and Dytiscid Water Beetles.” Journal of Biogeography 27(4): 821–827. https://doi.org/10.1046/j.1365-2699.2000.00445.x.

Emlen, J. T. 1952. “Flocking Behavior in Birds.” The Auk 69(2): 160–170. https://doi.org/10.2307/4081266.

European Bird Census Council (EBCC). 2022. “Species Population Trends.” PanEuropean Common Bird Monitoring Scheme. 2022. https://pecbms.info/trends-and-indicators/species-trends/.

Finlayson, C. M., G. T. Davies, W. R. Moomaw, G. L. Chmura, S. M. Natali, J. E. Perry, N. Roulet, and A. E. Sutton‐Grier. 2019. “The Second Warning to Humanity – Providing a Context for Wetland Management and Policy.” Wetlands 39(1): 1–5. https://doi.org/10.1007/s13157-018-1064-z.

Fisher, J., and K. Frank. 2004. “Abundance‐Distribution Relationships and Conservation of Exploited Marine Fishes.” Marine Ecology Progress Series 279: 201–213. https://doi.org/10.3354/meps279201.

Fluet‐Chouinard, E., B. D. Stocker, Z. Zhang, A. Malhotra, J. R. Melton, B. Poulter, J. O. Kaplan, et al. 2023. “Extensive Global Wetland Loss over the Past Three Centuries.” Nature 614(7947): 281–286. https://doi.org/10.1038/s41586-022-05572-6.

Fournier, A. M. V., J. D. Lancaster, A. P. Yetter, C. S. Hine, T. Beckerman, J. Figge, A. Gioe, et al. 2021. “Nest Success and Nest Site Selection of Wetland Birds in a Restored Wetland System.” Avian Conservation and Ecology 16(1): 6. https://doi.org/10.5751/ACE-01782-160106.

Fox, A. D., H. E. Jørgensen, E. Jeppesen, T. L. Lauridsen, M. Søndergaard, K. Fugl, P. P. Myssen, et al. 2024. “Breeding Waterbird Species as Ecological Indicators of Shifts from Turbid to Clear Water Conditions in Northwest European Shallow Eutrophic Lakes.” Hydrobiologia. https://doi.org/10.1007/s10750-024-05549-8.

Francová, K., K. Šumberová, G. A. Janauer, and Z. Adámek. 2019. “Effects of Fish Farming on Macrophytes in Temperate Carp Ponds.” Aquaculture International 27(2): 413–436. https://doi.org/10.1007/s10499-018-0331-6.

Freeman, B. G. 2019. “No Evidence for a Positive Correlation between Abundance and Range Size in Birds along a New Guinean Elevational Gradient.” Emu – Austral Ornithology 119(3): 308–316. https://doi.org/10.1080/01584197.2018.1530062.

Fried‐Petersen, H. B., Y. G. Araya‐Ajoy, M. N. Futter, and D. G. Angeler. 2020. “Drivers of Long‐Term Invertebrate Community Stability in Changing Swedish Lakes.” Global Change Biology 26(3): 1259–1270. https://doi.org/10.1111/gcb.14952.

Galewski, T., and V. Devictor. 2016. “When Common Birds Became Rare: Historical Records Shed Light on Long‐Term Responses of Bird Communities to Global Change in the Largest Wetland of France.” PLoS One 11(11): e0165542. https://doi.org/10.1371/journal.pone.0165542.

Galipaud, M., M. A. F. Gillingham, and F.‐X. Dechaume‐Moncharmont. 2017. “A Farewell to the Sum of Akaike Weights: The Benefits of Alternative Metrics for Variable Importance Estimations in Model Selection.” Methods in Ecology and Evolution 8(12): 1668–1678. https://doi.org/10.1111/2041-210X.12835.

Gaston, K. J. 1996. “The Multiple Forms of the Interspecific Abundance‐Distribution Relationship.” Oikos 76(2): 211. https://doi.org/10.2307/3546192.

Gaston, K. J. 2003. The Structure and Dynamics of Geographic Ranges. Oxford Series in Ecology and Evolution. Oxford: Oxford University Press.

Gaston, K. J., and T. M. Blackburn, eds. 2000. Pattern and Process in Macroecology, 1st ed. 151–177. Oxford: Wiley. https://doi.org/10.1002/9780470999592.

Gaston, K. J., T. M. Blackburn, J. J. D. Greenwood, R. D. Gregory, R. M. Quinn, and J. H. Lawton. 2000. “Abundance‐Occupancy Relationships.” Journal of Applied Ecology 37(s1): 39–59. https://doi.org/10.1046/j.1365-2664.2000.00485.x.

Gaston, K. J., T. M. Blackburn, and R. D. Gregory. 1997. “Abundance‐Range Size Relationships of Breeding and Wintering Birds in Britain: A Comparative Analysis.” Ecography 20(6): 569–579. https://doi.org/10.1111/j.1600-0587.1997.tb00425.x.

Gaston, K. J., T. M. Blackburn, R. D. Gregory, and J. J. D. Greenwood. 1998. “The Anatomy of the Interspecific Abundance‐Range Size Relationship for the British Avifauna: I. Spatial Patterns.” Ecology Letters 1(1): 38–46. https://doi.org/10.1046/j.1461-0248.1998.00004.x.

Gaston, K. J., T. M. Blackburn, and J. H. Lawton. 1997. “Interspecific Abundance‐Range Size Relationships: An Appraisal of Mechanisms.” Journal of Animal Ecology 66(4): 579. https://doi.org/10.2307/5951.

Guedo, D. D., and E. G. Lamb. 2013. “Temporal Changes in Abundance‐Occupancy Relationships within and between Communities after Disturbance.” Journal of Vegetation Science 24(4): 607–615. https://doi.org/10.1111/jvs.12006.

Haas, K., U. Köhler, S. Diehl, P. Köhler, S. Dietrich, S. Holler, A. Jaensch, M. Niedermaier, and J. Vilsmeier. 2007. “Influence of Fish on Habitat Choice of Water Birds: A Whole System Experiment.” Ecology 88(11): 2915–2925. https://doi.org/10.1890/06-1981.1.

Hanski, I., J. Kouki, and A. Halkka. 1993. “Three Explanations of the Positive Relationship between Distribution and Abundance of Species.” In Species Diversity in Ecological Communities, edited by R. Ricklefs and D. Schluter, 108–116. Chicago, IL: University of Chicago Press.

Hansson, L.‐A., P. Christer Bronmark, A. Nilsson, and K. Abjornsson. 2005. “Conflicting Demands on Wetland Ecosystem Services: Nutrient Retention, Biodiversity or Both?” Freshwater Biology 50(4): 705–714. https://doi.org/10.1111/j.1365-2427.2005.01352.x.

Hátle, M. 2000. “Information Sheet on Ramsar Wetlands (RIS) – Třeboňské Rybníky (Třeboň Fishponds).” Ramsar Information Sheet (RIS). https://rsis.ramsar.org/ris/495.

He, F., K. Gaston, and W. Jianguo. 2002. “On Species Occupancy‐Abundance Models.” Écoscience 9(1): 119–126. https://doi.org/10.1080/11956860.2002.11682698.

Heino, J., and M. Grönroos. 2014. “Untangling the Relationships among Regional Occupancy, Species Traits, and Niche Characteristics in Stream Invertebrates.” Ecology and Evolution 4(10): 1931–1942. https://doi.org/10.1002/ece3.1076.

Holt, A. R., and K. J. Gaston. 2003. “Interspecific Abundance‐Occupancy Relationships of British Mammals and Birds: Is It Possible to Explain the Residual Variation?” Global Ecology and Biogeography 12(1): 37–46. https://doi.org/10.1046/j.1466-822X.2003.00315.x.

Holt, A. R., K. J. Gaston, and F. He. 2002. “Occupancy‐Abundance Relationships and Spatial Distribution: A Review.” Basic and Applied Ecology 3(1): 1–13. https://doi.org/10.1078/1439-1791-00083.

Holt, A. R., P. H. Warren, and K. J. Gaston. 2004. “The Importance of Habitat Heterogeneity, Biotic Interactions and Dispersal in Abundance‐Occupancy Relationships.” Journal of Animal Ecology 73(5): 841–851. https://doi.org/10.1111/j.0021-8790.2004.00862.x.

Hořák, D., J. Rivas‐Salvador, J. Farkač, and J. Reif. 2022. “Traits and Ecological Space Availability Predict Avian Densities at the Country Scale of The Czech Republic.” Ecology and Evolution 12(7): e9119. https://doi.org/10.1002/ece3.9119.

IUCN. 2022. “The IUCN Red List of Threatened Species. Version 2022‐2.” https://www.iucnredlist.org.

Jetz, W., G. H. Thomas, J. B. Joy, K. Hartmann, and A. O. Mooers. 2012. “The Global Diversity of Birds in Space and Time.” Nature 491(7424): 444–448. https://doi.org/10.1038/nature11631.

Julliard, R., J. Clavel, V. Devictor, F. Jiguet, and D. Couvet. 2006. “Spatial Segregation of Specialists and Generalists in Bird Communities.” Ecology Letters 9(11): 1237–1244. https://doi.org/10.1111/j.1461-0248.2006.00977.x.

Kear, J., ed. 2005. Ducks, Geese, and Swans. Vol. 1. 2 vols. Bird Families of the World 1. Oxford; New York: Oxford University Press.

Kloskowski, J. 2011. “Differential Effects of Age‐Structured Common Carp (Cyprinus carpio) Stocks on Pond Invertebrate Communities: Implications for Recreational and Wildlife Use of Farm Ponds.” Aquaculture International 19(6): 1151–1164. https://doi.org/10.1007/s10499-011-9435-y.

Kutner, M. H., C. Nachtsheim, J. Neter, and C. J. Nachtsheim. 2004. Applied Linear Regression Models. Operations and Decision Sciences., 4th ed. Boston, MA: McGraw‐Hill/Irwin.

Latham, M., A. D. Cecilia, M. Latham, N. F. Webb, N. A. Mccutchen, and S. Boutin. 2014. “Can Occupancy–Abundance Models Be Used to Monitor Wolf Abundance?” PLoS One 9(7): e102982. https://doi.org/10.1371/journal.pone.0102982.

Lawton, J. H. 1994. “Population Dynamic Principles.” Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 344(1307): 61–68. https://doi.org/10.1098/rstb.1994.0052.

Lehikoinen, A., J. Rintala, E. Lammi, and H. Pöysä. 2016. “Habitat‐Specific Population Trajectories in Boreal Waterbirds: Alarming Trends and Bioindicators for Wetlands.” Animal Conservation 19(1): 88–95. https://doi.org/10.1111/acv.12226.

Lind, L., M. S. Schuler, W. D. Hintz, A. B. Stoler, D. K. Jones, B. M. Mattes, and R. A. Relyea. 2018. “Salty Fertile Lakes: How Salinization and Eutrophication Alter the Structure of Freshwater Communities.” Ecosphere 9(9): e02383. https://doi.org/10.1002/ecs2.2383.

Maceda‐Veiga, A., R. López, and A. J. Green. 2017. “Dramatic Impact of Alien Carp Cyprinus carpio on Globally Threatened Diving Ducks and Other Waterbirds in Mediterranean Shallow Lakes.” Biological Conservation 212(August): 74–85. https://doi.org/10.1016/j.biocon.2017.06.002.

Madge, S. 2010. Wildfowl: An Identification Guide to the Ducks, Geese and Swans of the World. London: Christopher Helm.

Manne, L. L., and R. R. Veit. 2020. “Temporal Changes in Abundance–Occupancy Relationships over 40 Years.” Ecology and Evolution 10(2): 602–611. https://doi.org/10.1002/ece3.5505.

Morelli, F., Y. Benedetti, J. O. Hanson, and R. A. Fuller. 2021. “Global Distribution and Conservation of Avian Diet Specialization.” Conservation Letters 14(4): e12795. https://doi.org/10.1111/conl.12795.

Morelli, F., Y. Benedetti, A. P. Møller, and R. A. Fuller. 2019. “Measuring Avian Specialization.” Ecology and Evolution 9(14): 8378–8386. https://doi.org/10.1002/ece3.5419.

Musil, P. 2006. “Effect of Intensive Fish Production on Waterbird Breeding Population: Review of Current Knowledge.” In Waterbirds around the World: A Global Overview of the Conservation, Management and Research of the World's Waterbird Flyways, edited by G. C. Boere, C. A. Galbraith, and D. A. Stroud, 520–521. Edinburgh: The Stationery Office.

Musil, P., Z. Musilová, R. Fuchs, and S. Poláková. 2011. “Long‐Term Changes in Numbers and Distribution of Wintering Waterbirds in The Czech Republic, 1966–2008.” Bird Study 58(4): 450–460. https://doi.org/10.1080/00063657.2011.603289.

Musil, P., R. Pichlová, P. Veselý, and J. Cepák. 1997. “Habitat Selection by Waterfowl Broods on Intensively Managed Fishponds in South Bohemia (Czech Republic).” In Proceedings of Limnology and Waterfowl, Monitoring, Modelling and Management. Workshop, Sarród/Sopron, Hungary, 21–23 November., edited by S. Faragó and J. Karekes, 169–175. Sopron: Wetlands International Publication 43.

Musilová, Z., P. Musil, J. Zouhar, and D. Romportl. 2015. “Long‐Term Trends, Total Numbers and Species Richness of Increasing Waterbird Populations at Sites on the Edge of Their Winter Range: Cold‐Weather Refuge Sites Are More Important than Protected Sites.” Journal of Ornithology 156(4): 923–932. https://doi.org/10.1007/s10336-015-1223-4.

Musilová, Z., P. Musil, J. Zouhar, A. Šenkýřová, D. Pavón‐Jordán, and P. Nummi. 2021. “Changes in Wetland Habitat Use by Waterbirds Wintering in Czechia Are Related to Diet and Distribution Changes.” Freshwater Biology 67(2): 309–324. https://doi.org/10.1111/fwb.13842.

Ottersen, G., B. Planque, A. Belgrano, E. Post, P. C. Reid, and N. C. Stenseth. 2001. “Ecological Effects of the North Atlantic Oscillation.” Oecologia 128(1): 1–14. https://doi.org/10.1007/s004420100655.

Päivinen, J., A. Grapputo, V. Kaitala, A. Komonen, J. S. Kotiaho, K. Saarinen, and N. Wahlberg. 2005. “Negative Density‐Distribution Relationship in Butterflies.” BMC Biology 3(1): 5. https://doi.org/10.1186/1741-7007-3-5.

Paracuellos, M., and J. L. Tellería. 2004. “Factors Affecting the Distribution of a Waterbird Community: The Role of Habitat Configuration and Bird Abundance.” Waterbirds 27(4): 446–453. https://doi.org/10.1675/1524-4695(2004)027[0446:FATDOA]2.0.CO;2.

Paradis, E., and K. Schliep. 2019. “Ape 5.0: An Environment for Modern Phylogenetics and Evolutionary Analyses in R.” Bioinformatics 35(3): 526–528. https://doi.org/10.1093/bioinformatics/bty633.

Patterson, J. H. 1976. “The Role of Environmental Heterogeneity in the Regulation of Duck Populations.” Journal of Wildlife Management 40(1): 22. https://doi.org/10.2307/3800152.

Pavón‐Jordán, D., P. Clausen, M. Dagys, K. Devos, V. Encarnaçao, A. D. Fox, T. Frost, et al. 2019. “Habitat‐ and Species‐Mediated Short‐ and Long‐Term Distributional Changes in Waterbird Abundance Linked to Variation in European Winter Weather.” Diversity and Distributions 25(2): 225–239. https://doi.org/10.1111/ddi.12855.

Pechar, L. 2000. “Impacts of Long‐Term Changes in Fishery Management on the Trophic Level Water Quality in Czech Fish Ponds: Impact of Fishery Management on Water Quality.” Fisheries Management and Ecology 7(1–2): 23–31. https://doi.org/10.1046/j.1365-2400.2000.00193.x.

Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and R Core Team. 2007. “Nlme: Linear and Nonlinear Mixed Effects Models, Version 3.1.” https://www.rdocumentation.org/packages/nlme/versions/3.1-159.

Pöysä, H., J. Rintala, A. Lehikoinen, and R. A. Väisänen. 2013. “The Importance of Hunting Pressure, Habitat Preference and Life History for Population Trends of Breeding Waterbirds in Finland.” European Journal of Wildlife Research 59(2): 245–256. https://doi.org/10.1007/s10344-012-0673-8.

R Core Team. 2022. R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing. https://www.R-project.org/.

Reif, J., D. Hořák, O. Sedláček, J. Riegert, M. Pešata, Z. Hrázský, Š. Janeček, and D. Storch. 2006. “Unusual Abundance‐Range Size Relationship in an Afromontane Bird Community: The Effect of Geographical Isolation?” Journal of Biogeography 33(11): 1959–1968. https://doi.org/10.1111/j.1365-2699.2006.01547.x.

Revell, L. J. 2012. “Phytools: An R Package for Phylogenetic Comparative Biology (and Other Things): Phytools: R Package.” Methods in Ecology and Evolution 3(2): 217–223. https://doi.org/10.1111/j.2041-210X.2011.00169.x.

Ricklefs, R. E. 2013. “Habitat‐Independent Spatial Structure in Populations of Some Forest Birds in Eastern North America.” Journal of Animal Ecology 82(1): 145–154. https://doi.org/10.1111/j.1365-2656.2012.02024.x.

Rotella, J. J., and J. T. Ratti. 1992. “Mallard Brood Movements and Wetland Selection in Southwestern Manitoba.” Journal of Wildlife Management 56(3): 508. https://doi.org/10.2307/3808866.

Roy, K., J. Vrba, L. Kajgrova, and J. Mraz. 2022. “The Concept of Balanced Fish Nutrition in Temperate European Fishponds to Tackle Eutrophication.” Journal of Cleaner Production 364(September): 132584. https://doi.org/10.1016/j.jclepro.2022.132584.

Scott, D. A., and P. M. Rose, eds. 1996. Atlas of Anatidae Populations in Africa and Western Eurasia. Wetlands International Publication 41. Wetlands International: Wageningen.

Sebastián‐González, E., J. A. Sánchez‐Zapata, and F. Botella. 2010. “Agricultural Ponds as Alternative Habitat for Waterbirds: Spatial and Temporal Patterns of Abundance and Management Strategies.” European Journal of Wildlife Research 56(1): 11–20. https://doi.org/10.1007/s10344-009-0288-x.

Seliger, B. J., B. J. McGill, J.‐C. Svenning, and J. L. Gill. 2021. “Widespread Underfilling of the Potential Ranges of North American Trees.” Journal of Biogeography 48(2): 359–371. https://doi.org/10.1111/jbi.14001.

Šimková, A., D. Kadlec, M. Gelnar, and S. Morand. 2002. “Abundance‐Prevalence Relationship of Gill Congeneric Ectoparasites: Testing the Core Satellite Hypothesis and Ecological Specialisation.” Parasitology Research 88(7): 682–686. https://doi.org/10.1007/s00436-002-0650-3.

Snell Taylor, S., J. Umbanhowar, and A. H. Hurlbert. 2020. “The Relative Importance of Biotic and Abiotic Determinants of Temporal Occupancy for Avian Species in North America.” Global Ecology and Biogeography 29(4): 736–747. https://doi.org/10.1111/geb.13064.

Snow, D., C. M. Perrins, and R. Gillmor. 1998. The Birds of the Western Palearctic, Concise ed. Oxford; New York: Oxford University Press.

Šťastný, K., and K. Hudec, eds. 2016. Ptáci 1. 3., Přepracované a doplněné vydání. Fauna ČR, Svazek 31. Praha: Academia.

Storchová, L., and D. Hořák. 2018. “Life‐History Characteristics of European Birds.” Global Ecology and Biogeography 27(4): 400–406. https://doi.org/10.1111/geb.12709.

Suhonen, J., and J. Jokimäki. 2019. “Temporally Stable Species Occupancy Frequency Distribution and Abundance–Occupancy Relationship Patterns in Urban Wintering Bird Assemblages.” Frontiers in Ecology and Evolution 7(April): 129. https://doi.org/10.3389/fevo.2019.00129.

Sutherland, W. J., I. Newton, and R. Green, eds. 2004. Bird Ecology and Conservation: A Handbook of Techniques. Techniques in Ecology and Conservation Series 1. Oxford; New York: Oxford University Press.

Tobias, J. A., C. Sheard, A. L. Pigot, A. J. M. Devenish, J. Yang, F. Sayol, M. H. C. Neate‐Clegg, et al. 2022. “AVONET: Morphological, Ecological and Geographical Data for All Birds.” Ecology Letters 25(3): 581–597. https://doi.org/10.1111/ele.13898.

Tovo, A., M. Formentin, S. Suweis, S. Stivanello, S. Azaele, and A. Maritan. 2019. “Inferring Macro‐Ecological Patterns from Local Presence/Absence Data.” Oikos 128(11): 1641–1652. https://doi.org/10.1111/oik.06754.

van der Lee, G. H., J. A. Vonk, R. C. M. Verdonschot, M. H. S. Kraak, P. F. M. Verdonschot, and J. Huisman. 2021. “Eutrophication Induces Shifts in the Trophic Position of Invertebrates in Aquatic Food Webs.” Ecology 102(3): e03275. https://doi.org/10.1002/ecy.3275.

Vavrečka, A., P. Šánová, and L. Kalous. 2020. “Insight into the Economy of Aquaculture Production in Czechia: Assessment of Aquaculture Enterprises.” Aquaculture International 28(1): 199–209. https://doi.org/10.1007/s10499-019-00453-8.

Verberk, W. C. E. P., G. van der Velde, and H. Esselink. 2010. “Explaining Abundance‐Occupancy Relationships in Specialists and Generalists: A Case Study on Aquatic Macroinvertebrates in Standing Waters.” Journal of Animal Ecology 79(3): 589–601. https://doi.org/10.1111/j.1365-2656.2010.01660.x.

Verhoeven, J. T. A. 2014. “Wetlands in Europe: Perspectives for Restoration of a Lost Paradise.” Ecological Engineering 66: 6–9. https://doi.org/10.1016/j.ecoleng.2013.03.006.

Webb, T. J., D. Noble, and R. P. Freckleton. 2007. “Abundance‐Occupancy Dynamics in a Human Dominated Environment: Linking Interspecific and Intraspecific Trends in British Farmland and Woodland Birds.” Journal of Animal Ecology 76(1): 123–134. https://doi.org/10.1111/j.1365-2656.2006.01192.x.

Wetlands International. 2022. “Waterbird Populations Portal.” http://wpp.wetlands.org/.

Wilman, H., J. Belmaker, J. Simpson, C. de la Rosa, M. M. Rivadeneira, and W. Jetz. 2014. “EltonTraits 1.0: Species‐Level Foraging Attributes of the World's Birds and Mammals.” Ecology 95(7): 2027. https://doi.org/10.1890/13-1917.1.

Zuckerberg, B., W. F. Porter, and K. Corwin. 2009. “The Consistency and Stability of Abundance‐Occupancy Relationships in Large‐Scale Population Dynamics.” Journal of Animal Ecology 78(1): 172–181. https://doi.org/10.1111/j.1365-2656.2008.01463.x.