https://orcid.org/0000-0003-0151-0355
Figures
Abstract
Land-use intensification in cattle-ranching landscapes of the Colombian Amazon generates continuous gradients of tree cover that modify the structure of productive mosaics and may function as environmental filters. The effects of this gradient may vary depending on the biodiversity component and the scale of analysis considered. We evaluated how tree cover is associated with the abundance, taxonomic richness, and functional diversity of the granivorous bird guild in cattle-ranching landscapes of the Colombian Amazon. A total of 100 quadrats were sampled along the continuous tree cover gradient. Abundance and local richness were analyzed using generalized linear mixed models with a negative binomial distribution. Taxonomic diversity pooled by cover type was estimated using Hill numbers with explicit control of sampling completeness. Functional structure was assessed using multidimensional indices (FRic, FEve, FDiv, FDis, and RaoQ) and multivariate trait-based analyses (RLQ and fourth-corner tests). Tree cover had a significant effect on local abundance, whereas quadrat-level richness showed only a marginal response. Taxonomic diversity was higher in structurally more heterogeneous covers, consistent with greater spatial turnover. At the functional level, FDis and RaoQ varied among covers, whereas FRic and FEve did not differ significantly. RLQ analysis identified significant covariation between proportions of tree cover and traits associated with body size and bill and tarsus morphology in granivorous birds. Overall, the results describe a partial and scale-dependent reorganization of the granivorous assemblage along the structural gradient. Although consistent with theoretical expectations of environmental filtering, these patterns should be interpreted as observational associations rather than causal evidence or direct evaluations of ecosystem functions.
Citation: Díaz-Cháux JT, Velasquez-Valencia A, Navarro-Morales A, Casanoves F (2026) Livestock landscapes as ecological filters: Effects of the tree cover gradient on the taxonomic and functional diversity of granivorous birds in the Colombian Amazon. PLoS One 21(3): e0345283. https://doi.org/10.1371/journal.pone.0345283
Editor: Petr Heneberg, Charles University: Univerzita Karlova, CZECHIA
Received: August 2, 2025; Accepted: March 3, 2026; Published: March 20, 2026
Copyright: © 2026 Díaz-Cháux et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are contained within the manuscript and its supporting information files.
Funding: Esta investigación fue financiada por el Programa de Becas Doctorales de Excelencia del Bicentenario, del Ministerio de Ciencia, Tecnología e Innovación - Minciencias (código BPIN 2021000100036) del Gobierno de Colombia (JTD).
Competing interests: The authors have declared that there are no competing interests.
Introduction
The Colombian Amazon constitutes one of the most globally significant socioecological systems due to its high biodiversity and the magnitude of ongoing land-use transformation processes. In recent decades, more than two million hectares of natural forest cover have been lost. The expansion of the agricultural frontier, land grabbing, and the increase in areas devoted to illicit crops [1–3] are the primary drivers of this deforestation. According to the Instituto de Hidrología, Meteorología y Estudios Ambientales de Colombia (IDEAM) [2], between 2018 and 2023 national deforestation was concentrated in the Amazon region, with an average annual loss of 95,000 ha of forest. In this context, the northwestern Colombian Amazon, where the department of Caquetá is located, represents the principal hotspot of forest cover loss and degradation, accounting for more than 20% of annual Amazonian deforestation during the same period [1,2]. In Caquetá, extensive cattle ranching constitutes one of the dominant land uses in transformed landscapes [4,5]. This activity has become the primary regional productive sector, with an inventory exceeding three million head of cattle, despite the low natural suitability of Amazonian soils, which require an average of 1.4 ha of pasture per animal to sustain production [4–6]. Between 2019 and 2023, livestock systems occupied 45% of the departmental agricultural area [3].
At the Amazonian scale, regional assessments consistently identify cattle ranching as one of the principal lands uses historically associated with deforestation and forest fragmentation, although its relative contribution varies spatially and temporally [7–11]. The spatial configuration of Amazonian cattle-ranching landscapes results from the historical interaction between land-use change and local physiography [12–14]. These processes have generated structural mosaics characterized by a continuous tree cover gradient, including open areas dominated by pastures, semi-open areas with scattered trees and early secondary regeneration, and semi-closed areas with greater canopy continuity [15,16]. These configurations do not represent discrete habitat types or independent production units, but rather dynamic structural states derived from clearing, pasture management, natural succession, and partial tree retention within livestock systems [17,18]. Within this framework, tree cover functions as an integrative gradient of the indirect effects of cattle ranching on landscape structure [19], modulating resource availability, refuge provision, and functional connectivity for wildlife. The figures presented above are provided strictly for contextual purposes, acknowledging their interannual and regional variability, and are used solely to frame the relevance of the tree cover gradient evaluated in this study, without implying direct causal relationships between deforestation, livestock activity, and the ecological patterns analyzed.
From the perspective of classical ecology, the relationship between biological diversity and habitat structure has been interpreted through foundational principles such as environmental heterogeneity, carrying capacity, and the ecological trade-offs faced by species along environmental gradients [20–24]. Habitat heterogeneity theory proposes that moderate increases in structural complexity expand the number of available niches and promote species coexistence [23–25]. However, it also recognizes nonlinear responses, whereby extreme simplification or fragmentation reduces the landscape’s capacity to sustain viable populations, leading to declines in richness and abundance within certain functional guilds [21,25,26]. In productive landscapes, these patterns emerge from trade-offs among resource availability, structural connectivity, and ecological costs associated with differential use of open versus tree-covered environments.
Contemporary functional ecology approaches have integrated these principles within an explicitly mechanistic framework, establishing that functional traits predictably determine both species’ responses to disturbance gradients and their potential contributions to ecosystem functioning [27–30]. Under environmental filtering theory, landscape configuration operates as a set of hierarchical filters that select specific trait combinations, constraining species persistence in transformed environments through limitations imposed by habitat structure and resource availability [31–33]. In birds, these filters primarily act on traits related to body size, mobility, and trophic specialization, which modulate physiological tolerance to environmental stress, movement capacity across open matrices, and differential access to resources [34].
In fragmented and productive landscapes, the occupation of forest remnants by bird species depends critically on their ability to traverse open areas [28,29]. These processes are closely linked to mobility and body size, as well as landscape attributes such as fragment size, connectivity, and conservation status [35]. Structural habitat simplification tends to exclude larger-bodied species with lower dispersal capacity due to higher energetic costs and risks associated with using open matrices [32–34], while favoring small, highly mobile species with generalist diets [36]. This directional selection of traits progressively reduces the diversity of functional strategies available, promoting functional homogenization processes that decrease resource-use complementarity and limit functional redundancy, with potential consequences for the stability and resilience of biological assemblages under environmental disturbance [32,37,38].
In cattle-ranching landscapes of the Colombian Amazon, the progressive loss of tree cover generates continuous habitat structure gradients that intensify these mechanisms [14,39,40]. Avian community responses to such gradients are mediated by functional traits such as dispersal capacity, reproductive efficiency, and trophic specialization, resulting in assemblages dominated by generalist species in highly simplified environments [41–43]. Owing to their diversity in body size, diet, and movement strategies, birds are particularly sensitive to variation in landscape structure and serve as effective indicators of habitat transformation effects on biodiversity and ecosystem functioning [44–46]. Moreover, they perform ecological functions closely linked to morphological and trophic traits that determine their role in energy and matter fluxes within natural systems [36,47–50].
Within this conceptual framework, granivorous birds constitute a functional guild particularly suitable for evaluating how the tree cover gradient operates as an environmental filtering mechanism in cattle-ranching landscapes [51]. The presence and abundance of these species are determined by a functional trade-off between the availability of grass and herb seeds in open and semi-open areas and the presence of arboreal components providing refuge, perching sites, and structural connectivity [52]. In highly simplified productive matrices, environmental filtering tends to favor small-bodied, highly mobile granivorous species with flexible diets, such as Sicalis flaveola and Volatinia jacarina [41], as well as some species of the genus Sporophila [53]. In contrast, larger-bodied granivores with lower mobility and greater dependence on dense vegetation, such as the tinamous Crypturellus cinereus and Tinamus guttatus, tend to be restricted to covers with greater structural complexity [44,54]. This functional reorganization of assemblages illustrates how habitat simplification may reduce available trait space and promote more redundant communities [55], whereas greater tree cover retention may be associated with more functionally diverse assemblages. In this sense, intermediate livestock landscape configurations, such as silvopastoral systems, may represent structural arrangements that balance functional diversity and ecological stability without necessarily maximizing extreme [56,57].
Evaluating these responses requires quantitative approaches integrating multiple dimensions of functional diversity. Multidimensional indices such as functional richness (FRic), functional evenness (FEve), functional divergence (FDiv), functional dispersion (FDis), and Rao’s quadratic entropy (RaoQ) link species functional traits with community assembly processes and potential implications for ecosystem functioning [58–62]. Complementarily, multivariate approaches such as RLQ analysis allow identification of co-structuring patterns among environmental variables, functional traits, and species composition, providing quantitative evidence of trait–environment associations compatible with environmental filtering processes [63].
Accordingly, this study evaluated how the tree cover gradient generated by livestock-driven landscape transformation influences the structure, composition, and functional diversity of granivorous bird communities in cattle-ranching landscapes of the Colombian Amazon. We hypothesized that tree cover acts as an ecological filter structuring the taxonomic and functional diversity of this guild, such that trait combinations associated with structurally complex environments are linked to denser covers, whereas more open covers are associated with species possessing traits compatible with disturbed environments. The study focused on patterns of abundance, taxonomic richness, and trait-based functional structure measured directly in the field along the continuous tree cover gradient. Although functional traits are often discussed in relation to ecological processes and ecosystem services, the present analysis is restricted to empirically observed patterns; any implications regarding ecosystem functions, stability, or resilience are considered theoretical expectations derived from the literature and were not directly evaluated in this research.
Materials and methods
Study area
The research was conducted in the northwestern region of the Colombian Amazon, across eight livestock landscape mosaics located in seven municipalities within the department of Caquetá: (i) Albania; 1.302766° N – 75.909201° W, 277 m above sea level (a.s.l.), (ii) El Doncello; 1.643959° N – 75.261978° W, 359 m a.s.l., (iii) Florencia; 1.429608° N – 75.528399° W, 248 m a.s.l., (iv) La Montañita; 1.336173° N – 75.352393º W, 269 m a.s.l., (v) Milán; 1.300662° N – 75.519043º W, 222 m a.s.l., (vi) Morelia; 1.409152° N – 75.727271º W, 259 m a.s.l., and (vii) San José del Fragua; 1.270573° N – 76.007303º W, 376 m a.s.l. (Fig 1). The landscapes of the region are characterized by a geomorphology of low hills and piedmont, with floodable valleys and an average slope of 12%. The area has an average temperature of 28.62 °C, relative humidity of 86%, and annual precipitation of 4277 mm, with a unimodal distribution and a peak rainy season from April to October [64].
Location of the Caquetá department at the national (Colombia) and continental (South America) scale (A); distribution of the eight mosaics across the municipalities of Caquetá (B); classification of the three types of tree cover in the studied mosaics (C); Mosaics: TR: El Triunfo (1); VE: La Vega (2); ES: La Esmeralda (3); TE: El Tesoro (4); VM: Villa Mery (5); PO: El Porvenir (6); SR: Santa Rosa (7); BA: Batalla 13 (8). Base map and imagery: Sentinel-2A (Copernicus Programme, European Space Agency), accessed via the Copernicus Data Space Ecosystem (CDSE). Created by the authors under CC BY 4.0.
Classification of tree cover in livestock landscape mosaics and spatial data preparation
Each landscape mosaic corresponded to a square area of 1.56 km² (156 ha), subdivided into 25 symmetrical quadrants of 0.0625 km² (6.25 ha) each (Fig 2). The quadrant constituted the unit of spatial replication for all subsequent environmental and ecological analyses. For each quadrant, a thematic vegetation cover map was generated as the final cartographic product at the unit-of-analysis scale. The base layer was the official Land Cover vector dataset downloaded from the Geoportal of the Territorial Environmental Information System of the Colombian Amazon (SIAT-AC), developed under the CORINE Land Cover methodology adapted for Colombia [65], and its regional adjustment for the Amazon [66], at a nominal cartographic scale of 1:25,000. This product provides polygon-based classification of land-cover types derived from institutional remote-sensing interpretation.
Original figure created by the authors using Sentinel-2A imagery (Copernicus Programme, European Space Agency), accessed through the Copernicus Data Space Ecosystem (CDSE), under CC BY 4.0 license.
The SIAT-AC vector layer was clipped to the boundaries of each mosaic and subsequently to each quadrant, ensuring that only spatial information strictly corresponding to the sampling units was retained. Because the study required quadrat-level precision (6.25 ha resolution), an additional spatial verification and quality-control procedure was implemented to ensure consistency between institutional cartography and on-site conditions during the 2023 sampling period.
This verification process included: (i) visual inspection of multispectral Sentinel-2A imagery (Copernicus Programme, European Space Agency), accessed via the Copernicus Data Space Ecosystem (CDSE) under a CC BY 4.0 license, using images with cloud cover < 10% and interpreted at an approximate working scale of 1:10,000; and (ii) direct field observations conducted during biological sampling campaigns, aimed at confirming dominant cover classes and detecting transitions among land-cover types.
This step was considered necessary because transitional or heterogeneous covers—such as early secondary regeneration, agroforestry systems, and wetland edges—may exhibit local boundary mismatches or class assignment discrepancies, even when the original product has undergone institutional validation [67]. Such discrepancies can introduce measurement error when environmental predictors are quantified at fine spatial scales.
Following this verification, an edited vector layer (shapefile) was produced by the authors through: (i) geometric adjustment of polygon boundaries when inconsistencies with high-resolution imagery were detected, and/or (ii) punctual reclassification of polygons when field evidence indicated misclassification. All modifications strictly preserved the CORINE/SINCHI classification framework [65,66], avoiding the creation of ad hoc classes. All spatial layers were georeferenced in the MAGNA-SIRGAS 2018 coordinate system (EPSG:9377) and processed using ArcMap version 10.8 [68] and QGIS version 3.40 [69]. From the final edited thematic layer, the area (m2) of each polygon within each quadrant was calculated by land-cover class. Tree cover was then quantified a continuous variable at the quadrant scale. Woody vegetation included forest remnants, secondary regeneration patches, tree-dominated agroforestry systems, and arboreal components embedded in pastures, consistent with the structural definition of tree cover under the CORINE framework [65,66]. It is important to clarify that tree cover was not estimated by assigning mean values to discrete land-use categories. Instead, it was directly quantified from the spatially explicit area of polygons representing arboreal vegetation within each sampling quadrat. Consequently, the environmental variable corresponds to an empirical continuous gradient (observed range: 0–60% in this study), derived from polygon-level cartographic quantification and expressed as the proportion of tree cover relative to the total area of each quadrat.
These continuous tree-cover percentages were used directly as environmental predictors in multivariate trait-based analyses (RLQ and fourth-corner tests), without discretization, recoding, or compositional transformation. Because the primary environmental predictor corresponds to a single continuous variable rather than to a closed set of proportions summing to 100%, it is not subject to closed-sum constraints in the multivariate framework.
For descriptive and interpretative purposes only, quadrants were subsequently grouped into three operational structural states based on their observed tree-cover percentages: open (OP), semi-open (SO), and semi-closed (SC) (Table 1). These categories represent ranges along a continuous gradient and do not constitute discrete habitat types. They were used exclusively in analyses of abundance, richness, and taxonomic diversity to facilitate ecologically interpretable contrasts among contrasting structural configurations of livestock landscapes. The full continuous tree-cover values for all quadrants are reported in Supplementary S1 Table, ensuring transparency and enabling independent verification of within-category variability.
This two-step approach—continuous quantification for inferential modeling and categorical grouping for descriptive comparison—ensures analytical precision while maintaining ecological interpretability, and minimizes bias associated with arbitrary class midpoint assignments or coarse categorical proxies.
Bird census and functional traits
The granivorous bird assemblage was characterized by combining point counts and mist-net captures conducted between January and November 2023 under standardized protocols of sampling effort, duration, and environmental conditions. Both methods were implemented within controlled temporal windows and at fixed locations within each quadrat to minimize repeated detections and ensure consistency in sampling effort. The quadrat (6.25 ha) constituted the unit of replication for all analyses of abundance, richness, and functional diversity, directly corresponding to the spatial scale at which tree cover was quantified.
Sampling covered nearly a complete annual cycle, incorporating regional climatic variability. Because the species included correspond exclusively to resident granivorous birds, seasonal migration was not expected to significantly influence assemblage composition. Additionally, all landscape mosaics were sampled within the same temporal window, reducing potential seasonal bias and allowing observed patterns to be interpreted primarily along the tree cover gradient.
Observation points were established at the center of each quadrat and spaced 250 m apart, resulting in 25 fixed points per mosaic (Fig 2). Point counts followed standardized protocols widely used in studies of bird communities in tropical and productive landscapes [70,71]. Each count lasted 25 minutes and was restricted to a fixed radius of 50 m around the observer. Only individuals detected visually or acoustically within this radius were recorded. Movement direction, spatial position, and behavioral cues were considered to reduce the probability of recording the same individual more than once, particularly in open habitats with high mobility.
Mist-net sampling consisted of installing four 12 m nets per quadrat, strategically positioned to intercept flight paths between tree-covered areas [43,72]. Nets were operated for five morning hours (06:00–11:00) and checked every 20 minutes. Total mist-netting effort per mosaic was 100 net-hours. Each captured individual was measured for eight morphological traits used to quantify functional diversity (Fig 3): total body length (LTO), tail length (LCO), tarsus length (LTA), extended wing length (AEX), commissure length (COM), bill height (ALT), total culmen length (CTO), and body mass (PES). Measurements were obtained following standardized morphometric procedures.
CTO: total culmen length; LTA: tarsus length; LTO: total body length; AEX: extended wing length; LCO: tail length; COM: commissure; ALT: bill height; PES: body weight (g). Illustration of Ammodramus aurifrons (Passeriformes: Passerellidae) by Viviana Tello M.
For species detected but not captured, trait values were obtained from specimens housed in the Ornithological Collection of the Natural History Museum of the Universidad de la Amazonia (UAM), Colombia, and from the global AVONET database [45]. Taxonomic identification followed the nomenclature of the South American Classification Committee [73] and was verified using bird guides for Colombia [74–76]. Species were classified into seven trophic guilds (frugivores, insectivores, nectarivores, granivores, folivores, scavengers, and vertebrate consumers). However, only species assigned to the granivore guild (GRA) were included in this study.
Species-level abundance in each quadrat was estimated through the standardized integration of records obtained from point counts and mist-net captures. To prevent double counting, point-count records were incorporated only when they did not correspond to individuals previously captured in mist nets, based on spatial proximity, timing of detection, and observable characteristics. Temporal coordination between methods further reduced potential overlap. Abundance and richness values represent aggregated counts at the quadrat scale derived from effective sampling effort and confirmed detections, ensuring that each quadrat functions as an independent observational unit.
Although detectability may vary with vegetation structure, the objective of this study was to compare relative differences along a continuous tree cover gradient under a uniform sampling design, rather than to estimate detectability-corrected absolute densities. By standardizing sampling effort, duration, and environmental conditions across quadrats, any residual variation in detectability is assumed to be comparably distributed among units, allowing valid landscape-scale inference in accordance with methodological recommendations on detectability control and standardized avian surveys [70,71,77]. All detections were analyzed at the quadrat scale, ensuring direct correspondence between biological response variables and environmental predictors in subsequent statistical analyses.
Ethics statement
The specimen collections were conducted under the framework of Resolution 1015 of 2016, in accordance with the General Permit for the Collection of Specimens of Wildlife Species for Non-Commercial Scientific Research, issued to the Universidad de la Amazonia. Approval for practices involving the use of animals was granted by the Institutional Committee on Ethics and Bioethics in Research of the Universidad de la Amazonia (FO-A-APC-01–11), established through Agreement 020 of 2018 by the Superior Council.
Data analysis
All analyses were conducted using the sampling quadrat (n = 100) as the unit of replication, directly corresponding to the spatial scale at which the continuous percentage of tree cover was quantified. Quadrats were distributed among open (OP, n = 32), semi-open (SO, n = 34), and semi-closed (SC, n = 34) structural categories, which were used exclusively for comparative and descriptive purposes. The continuous percentage of tree cover per quadrat constituted the environmental variable used in the multivariate analyses.
The effect of tree cover on the abundance of individuals and species richness of granivorous birds was evaluated using Generalized Linear Mixed Models with a negative binomial distribution (NB GLMM; nbinom2). Tree cover was included as a fixed effect, and the livestock landscape mosaic in the Colombian Amazon was incorporated as a random effect to account for the hierarchical structure of the design and to avoid pseudoreplication. The adequacy of the distribution was verified through overdispersion diagnostics using DHARMa, confirming the robustness of statistical inference. Post hoc comparisons among OP, SO, and SC categories were performed using Tukey’s method (α = 0.05) in R version 4.2 [78], restricted to ecologically justified contrasts.
Alpha taxonomic diversity was estimated using Hill numbers, which express the effective number of species per quadrat and per tree cover category. Rarefaction–extrapolation curves were generated for the first three orders: q0 (species richness), q1 (Shannon diversity), and q2 (dominance) [79], using the iNEXT.3D package [80], incorporating explicit assessment of sampling completeness. Pooled (γ) diversity was estimated by tree cover type by aggregating all corresponding quadrats (OP = 32; SO = 34; SC = 34), allowing explicit comparison between local (α) and accumulated (γ) diversity patterns without formally attributing mechanisms of spatial turnover.
Functional structure was examined using complementary multivariate analyses. First, hierarchical cluster analysis was conducted using Ward’s algorithm and Gower’s similarity index to group granivorous bird species based on functional trait similarity, and a Principal Component Analysis (PCA) was performed to explore covariation patterns among traits. These procedures were conducted in InfoStat statistical software version 2020 [81]. Differences in morphometric traits among OP, SO, and SC categories, defined at the quadrat scale, were evaluated using General and Mixed Linear Models (GMLM), with traits as response variables and tree cover category as a fixed effect. Model assumptions were verified through graphical inspection of residuals. Post hoc comparisons were conducted using Fisher’s LSD (α = 0.05) only after significant global tests and restricted to ecologically motivated a priori contrasts in R version 4.2 [78], maintaining a balance between Type I error control and statistical power given the limited number of comparisons.
The relationship between environmental variables and functional traits was assessed using RLQ analysis with ade4 package, statistical significance was evaluated using fourth-corner tests (modeltype = 2) [82] in R version 4.2 [78]. The environmental matrix (R) included the proportional area of open (OP), semi-open (SO), and semi-closed (SC) tree cover per quadrat, treated as continuous environmental variables derived from spatial measurements. The L matrix corresponded to granivorous bird abundance per quadrat, and the Q matrix represented species-level functional traits. Matrices R and Q were subjected to Principal Component Analysis (PCA), while matrix L was analyzed using Correspondence Analysis (COA). A co-inertia approach was applied to maximize the covariation between R and Q through L. Statistical significance of environment–trait associations was evaluated using the fourth-corner test with 9,999 permutations, and p-values were adjusted for multiple comparisons using the False Discovery Rate (FDR) procedure [83]. Structural cover categories (OP, SO, SC) were not included as predictors in the RLQ or fourth-corner tests and were used exclusively for descriptive visualization.
Functional diversity at the quadrat scale was quantified using five multidimensional indices [84,85]: FRic (functional richness), FEve (functional evenness), FDiv (functional divergence), FDis (functional dispersion), and RaoQ (quadratic entropy), calculated with the FD package [62] in R version 4.2 [78]. Differences among tree cover categories were evaluated using GMLM, incorporating tree cover as a fixed effect and mosaic as a random effect. Fisher’s LSD (α = 0.05) was applied only following significant global tests. Additionally, a functional diversity analysis based on Area Under the Curve (AUC) was conducted using Hill numbers [86], generating rarefaction–extrapolation curves for q0, q1, and q2 [79,87] with iNEXT.3D [80,88].
Effective sample size varied among analytical blocks due to minimum species richness requirements per quadrat. GLMM analyses of abundance and richness included all 100 quadrats. For functional diversity indices, FRic, FEve, and FDiv were calculated only for quadrats with ≥ 3 species (n = 42; OP = 16, SO = 19, SC = 7), whereas FDis and RaoQ were calculated for quadrats with ≥ 2 species (n = 70; OP = 26, SO = 27, SC = 17). RLQ and fourth-corner analyses included n = 100 quadrats and S = 22 species, using complete community, environmental, and trait matrices. The total number of sampled quadrats and the number included in each index are reported in Supplementary S2 Table.
Results
A total of 560 individuals of granivorous birds were recorded, representing 22 species, four families, and three orders. The order Passeriformes exhibited the highest species richness and abundance. The species Volatinia jacarina, Patagioenas cayennensis, and Ammodramus aurifrons accounted for approximately 50% of the individuals observed. The complete functional trait matrix is provided in Supplementary S3 and S4 Tables, and the R code along with the data are available in Supplementary S1 Code and S1 Data.
Abundance and species richness were modeled at the quadrat scale (n = 100) using negative binomial generalized linear mixed models (NB GLMM; nbinom2; log link), with tree cover category included as a fixed effect and livestock landscape mosaic incorporated as a random intercept to account for the hierarchical sampling design. Zero-inflated alternatives were explicitly evaluated to assess whether excess zeros influenced model structure; however, no zero counts were observed in the quadrat-level response variables, and standard negative binomial models showed lower AIC values (ΔAIC = 2). Residual diagnostics based on DHARMa simulations indicated no evidence of overdispersion or zero inflation, supporting the adequacy of the selected error distribution for count data under the adopted aggregation scheme.
Because the absence of zero counts is uncommon in guild-level bird data, additional sensitivity analyses were conducted including quadrats with zero detections while maintaining the same model structure and random-effect specification. These alternative models produced consistent effect directions, comparable parameter estimates, and identical inferential conclusions. This confirms that the reported patterns were not driven by data aggregation procedures or by the exclusion of zero observations but rather reflect stable differences across the tree cover gradient under the standardized sampling design.
For abundance, the standard negative binomial NB GLMM was preferred over the zero-inflated alternative (AIC 540.78 vs 542.78). Tree cover had a significant overall effect (Type II Wald χ² = 6.42; df = 2; p = 0.040). Post hoc comparisons indicated that abundance under semi-open cover (SO) was significantly higher than under semi-closed cover (SC) (IRR = 1.69; p = 0.031), whereas other contrasts were not statistically significant. For species richness, the standard negative binomial NB GLMM was also preferred (AIC 380.00 vs 382.00). Tree cover did not show a significant overall effect (Type II Wald χ² = 5.21; df = 2; p = 0.074), indicating that mean species richness per quadrat did not differ significantly among OP, SO, and SC categories.
In contrast to quadrat-level patterns, pooled (γ) diversity estimated by tree cover category using Hill numbers revealed higher accumulated richness under SC. This category exhibited the highest observed richness (q0 = 18 species; n = 134 individuals) and higher values for q1 and q2 (Fig 4A). Sample completeness analysis indicated that the SO category would require approximately three times more individuals (~3×) to reach comparable levels of completeness (Fig 4B). This pattern is consistent with a higher proportion of rare species and/or greater heterogeneity among quadrats within that category, although spatial turnover was not formally partitioned in this study. These results therefore indicate a divergence between local (α) and pooled (γ) diversity patterns across the tree cover gradient.
OP: open; SO: semi-open; SC: semi-closed. A: Sample size-based sampling curve; B: Sample completeness curve.
Functional classification of species
The hierarchical cluster analysis (Ward algorithm; Gower similarity index) identified two clearly differentiated functional groups (Fig 5). The first group comprised nine large-bodied granivorous species belonging to Columbiformes and Tinamiformes. The second group consisted of thirteen smaller species, primarily Passeriformes, together with two species of the genus Columbina. Functional similarity between groups was below 30%, indicating a clear morphological separation between large and small granivorous birds. The resulting species-by-group matrix is presented in Supplementary S5 Table.
The algorithm’s distance was calculated using the default transformation of the Gower similarity index.
Principal component analysis (PCA) explained 71.8% of the total variance across the first two axes. Along PC1, larger-bodied and heavier species were associated with higher percentages of tree cover (semi-closed sites), whereas smaller species with broader and taller bills were associated with open and semi-open covers (Fig 6). The second axis (PC2) reflected variation in locomotor and trophic structures without strict separation among cover types.
OP: open; SO: semi-open; SC: semi-closed; CTO: total culmen length; PES: body weight (g); LTA: tarsus length; LTO: total body length; AEX: extended wing length; LCO: tail length; COM: commissure; ALT: bill height.
At the quadrant scale, several morphological traits showed significant differences among tree cover categories: total culmen length (CTO: F = 6.69; p = 0.0015), extended wing length (AEX: F = 3.56; p = 0.0297), total body length (LTO: F = 3.82; p = 0.0232), and body weight (PES: F = 6.52; p = 0.0017), with higher mean values observed under semi-closed cover. Other traits showed no statistical evidence of differences, indicating that morphological differentiation across the tree cover gradient was selective rather than uniform across all trait dimensions at the quadrat scale.
Trait–Cover association (RLQ and Fourth-Corner tests)
The RLQ analysis assessed the co-structure among the environmental matrix (R), composed of the proportions of tree cover in the OP, SO, and SC categories, the species abundance matrix (L), and the species-level functional trait matrix (Q), weighted by abundance. The first two canonical axes accounted for 100% of the total co-structural inertia, with RLQ1 explaining 93.42% and RLQ2 6.58% of the total covariance (Fig 7). These values represent the relative distribution of the shared variance among R, L, and Q. The first RLQ axis captured the dominant pattern of association between tree cover structure and species functional traits. The global Monte Carlo test was significant (p = 0.0245; 9,999 permutations), indicating that the observed co-structure exceeded that expected under the null model.
ALT: bill height; COM: commissure; AEX: extended wing; LTA: tarsus length; LCO: tail length; LTO: total body length; CTO: total culmen length; PES: body weight (g); OP: open; SO: semi-open; SC: semi-closed. Red squares indicate positive associations, while blue squares indicate negative associations. Lighter colors indicate weaker associations, whereas darker colors represent stronger associations. The illustration of Ammodramus aurifrons highlights the traits with the strongest associations.
These patterns were statistically confirmed using the fourth-corner test. Significant correlations were identified (p < 0.05, FDR-adjusted) between the percentage of tree cover and functional traits. In particular, a significant negative association was detected between open cover and the functional traits total culmen length (CTO; r = –0.1840; p = 0.0032), tarsus length (LTA; r = –0.1703; p = 0.0105), total body length (LTO; r = –0.1335; p = 0.0340), and body mass (PES; r = –0.1619; p = 0.0097). In contrast, commissure width was positively correlated with open cover (COM; r = 0.1274; p = 0.0251).
Accordingly, smaller-bodied and lighter granivorous birds, characterized by shorter and wider bills and shorter tarsi, were associated with areas exhibiting lower percentages of tree cover (OP). Conversely, larger-bodied granivorous birds (LTO; r = 0.1250; p = 0.0496) and heavier species (PES; r = 0.1718; p = 0.0063), with longer tarsi (LTA; r = 0.1929; p = 0.0032) and longer (CTO; r = –0.1840; p = 0.0032) and more slender bills (COM; r = –0.1238; p = 0.0278), were associated with areas exhibiting higher percentages of tree cover. These results indicate statistically significant trait–environment covariation along the continuous tree cover gradient. The R code along with the data are available in the Supplementary Material S2 Code and S2 Data.
Functional diversity of the granivorous bird assemblage
At the quadrant scale, only FDis (F = 7.12; p = 0.0013; df = 2) and RaoQ (F = 7.41; p = 0.0001; df = 2) showed significant differences among cover categories. FDis peaked under semi-open cover, whereas RaoQ was highest under semi-closed cover. Fric and FEve did not show significant differences. The R code along with the data are available in the Supplementary Material S3 Code and S3 Data
In the AUC-based functional diversity analysis, rarefaction–extrapolation curves indicated higher q0, q1, and q2 values under semi-closed cover at reduced sample sizes (Fig 8). When extrapolated to larger sampling sizes, functional diversity indices increase particularly in open and semi-closed covers. These patterns reflect scaling properties of accumulated functional diversity rather than changes in per-quadrat functional structure, indicating statistical association between structural complexity and scaling patterns of functional diversity.
OP: open; SO: semi-open; SC: semi-closed.
The results indicate that different metrics capture complementary dimensions of community assembly: local abundance reaches its maximum under intermediate tree cover, cumulative species richness is higher in semi-closed cover, and functional dispersion peaks in structurally intermediate states. These findings indicate that the observed patterns are not monotonic but instead reflect dimension-specific differentiation across the tree cover gradient.
Discussion
The results demonstrate a statistically significant effect of tree cover type on the mean abundance of the granivorous bird assemblage at the quadrat scale, whereas species richness exhibited only a marginal overall response. Consequently, variation along the tree cover gradient is primarily expressed through changes in the number of individuals rather than consistent modifications in the number of species per sampling unit. Although semi-open covers showed higher mean values for both abundance and richness, only abundance displayed statistically supported differences among cover types. Therefore, the observed trend in richness should be interpreted descriptively and not as conclusive evidence of a consistent effect of the structural gradient on local taxonomic diversity. In this context, the results suggest that structural differences in the landscape are mainly associated with variation in assemblage magnitude, expressed as abundance, rather than systematic shifts in taxonomic composition at the local scale.
The sampling design, implemented simultaneously across mosaics and extended over nearly one year, allowed integration of the annual climatic variability characteristic of the Colombian Amazon and reduced potential seasonal biases. Moreover, by focusing exclusively on resident species within the granivorous guild, it is unlikely that the observed patterns were influenced by migratory movements. These methodological characteristics strengthen the interpretation of the findings as consistent associations within an observational framework. Comparable patterns have been documented in tropical agricultural landscapes, where intermediate structural configurations are associated with higher abundances without necessarily implying increases in local richness [89–94]. In the present study, however, these parallels are presented as contextual comparisons rather than direct evidence of specific ecological mechanisms.
Species such as Volatinia jacarina, Patagioenas cayennensis, and Ammodramus aurifrons were the most abundant during sampling. These species were primarily recorded in open and semi-open covers dominated by pastures, vegetation edges, and early successional stages. The literature indicates that their functional trait combinations are associated with efficient resource exploitation in habitats with low tree cover: V. jacarina in anthropogenic grasslands [95,96]; P. cayennensis in fragmented landscapes [35,97]; and A. aurifrons in livestock pastures dominated by forage grasses [53,98]. These patterns are consistent with previous studies [41,99,100] describing associations between reduced vertical complexity and higher population densities in these species. However, such responses should be interpreted at the assemblage level—as expressions of disturbance tolerance, niche breadth, and ecological plasticity—rather than as evidence of strict or exclusive specialization to open environments [31,45,46,101–103].
Within this framework, the patterns observed along the tree cover gradient are consistent with the multivariate structure of avian dietary space described by Barnagaud et al. [104], in which granivory is frequent in structurally simplified and human-modified environments. From this perspective, the higher abundance of granivorous birds in semi-open covers may reflect environmental filtering along structural gradients, consistent with theoretical expectations. Accordingly, the RLQ analysis revealed statistically significant covariation between tree cover percentage and morphological traits, particularly body size and bill morphology, indicating functional structuring associated with the environmental gradient [32–34]. Nevertheless, these findings describe covariation patterns and do not demonstrate causal mechanisms or deterministic specialization by cover type [105,106].
The tree cover gradient was also associated with differential occurrence of certain granivorous species. Some species were detected exclusively in a single cover type, such as Zenaida auriculata in open areas and Sporophila nigricollis, S. intermedia, and S. minuta in semi-open covers. In contrast, larger-bodied species such as Crypturellus cinereus and Tinamus guttatus were detected exclusively in semi-closed covers. These patterns may reflect differences in structural habitat requirements [33–35]; however, they should be interpreted cautiously, as detectability may vary with structural complexity and population density [25,231, 42,97]. In the case of T. guttatus, classified as “Near Threatened” [107], its association with more complex covers suggests the potential importance of these remnants for local persistence [108,109], although this study did not explicitly evaluate demographic parameters or population persistence probabilities.
Hierarchical cluster analysis based on morphological traits identified two main functional groups within the granivorous guild: one composed of larger-bodied species, including P. plumbea, P. cayennensis, C. soui, and T. guttatus, and another comprising smaller-bodied species, including representatives of the genera Sporophila and Columbina. This differentiation was consistent with patterns revealed by PCA and RLQ analyses, which showed statistical associations between traits such as culmen length (CTO), tarsus length (LTA), and body mass (PES) and semi-closed covers. These associations indicate that certain trait combinations are more represented in structurally complex environments [110–115], without implying direct evidence of adaptive advantages or differential performance [50,116,117]. In contrast, open covers were associated with smaller-bodied species and relatively wider, deeper bills, linked to diets dominated by Poaceae and Cyperaceae seeds, including forage grasses such as Brachiaria sp., widely available in livestock systems [118–120]. Given that multiple studies have documented heterogeneous and context-dependent responses to disturbance gradients [121–124], these findings are presented as structural assemblage trends.
Overall, variation in tree cover within Amazonian livestock mosaics was statistically associated with changes in abundance patterns and specific dimensions of the functional structure of the granivorous guild. These associations are consistent with theoretical expectations of environmental filtering [31–33,37,41,51]; however, the study did not directly evaluate underlying mechanisms. Reduced tree cover was linked to decreases in abundance-weighted functional dispersion indices (FDis and RaoQ), whereas other indices such as FRic and FEve did not show significant differences. Collectively, the results suggest partial reorganization of functional space, compatible with patterns of functional simplification or homogenization documented in transformed systems [49,125–130], but they do not indicate a uniform contraction of total functional volume.
Functional diversity of the granivorous assemblage
From the conceptual framework of environmental filtering, the response of functional diversity to the tree cover gradient may be interpreted as consistent with the potential action of hierarchical filters associated with livestock landscape configuration [33,131–137]. In the Colombian Amazon, these landscapes are characterized by open matrices dominated by pastures, with scattered remnant vegetation patches and reduced functional connectivity [7,43,138,139], conditions associated with functional space restructuring and shifts in trait distribution [140].
The results indicate that different functional diversity indices capture complementary dimensions of the assemblage and respond non-uniformly to the structural gradient. FDis and RaoQ showed statistically significant differences among cover types, indicating changes in trait dispersion and dominance structure weighted by abundance. In contrast, FRic and FEve did not differ significantly, suggesting stability in total functional volume and trait space evenness. Semi-open covers exhibited relatively higher FDis values, reflecting greater abundance-weighted functional dispersion, whereas semi-closed covers showed descriptively higher FRic and FEve values, although without conclusive statistical support. These patterns are consistent with studies documenting functional reorganization under landscape transformation without necessarily implying expansion or contraction of total functional volume [34,44,104,141].
Open-cover mosaics displayed the lowest functional diversity values, consistent with the structural simplification characteristic of Amazonian livestock matrices [43,64,142–144]. Several studies have linked the dominance of small, opportunistic species with increased functional redundancy in simplified environments [9,29,100,145]. Previous research has also suggested that loss of ecosystem processes, including seed dispersal and vegetation regeneration, may be associated with functional homogenization [123,146–149]; however, such processes were not evaluated in this study and should be considered theoretical implications derived from the literature.
In summary, functional diversity within the granivorous guild was associated with the tree cover gradient through reorganization of trait dominance and abundance structure, consistent with structural trends observed in transformed landscapes [150–157]. These patterns should be interpreted as observational associations rather than deterministic rules. From an applied perspective, the findings suggest that landscape configurations with greater structural heterogeneity may favor higher abundance-weighted functional dispersion within the granivorous guild [151,153,154]. Nevertheless, the study did not directly assess livestock productivity, ecosystem service provision, or long-term sustainability; therefore, such implications should be regarded as hypotheses for future research derived from the present findings.
Conclusions
According to the study findings, the continuous tree cover gradient was significantly associated with the abundance of the granivorous bird assemblage at the quadrat scale, whereas taxonomic richness exhibited only marginal responses. This pattern indicates that structural variation in the landscape is expressed primarily through changes in the number of individual birds rather than through consistent modifications in the number of species per sampling unit. From a functional perspective, the reorganization of trait space was partial and non-uniform. Differences observed in certain functional diversity indices among cover types suggest changes in the dispersion and relative dominance of abundance-weighted functional traits, whereas other dimensions of the functional volume remained statistically stable. Consistently, significant multivariate covariation was detected between proportions of tree cover and morphological traits associated with body size and bill and tarsus morphology in granivorous birds. Overall, these results are consistent with theoretical expectations of environmental filtering along structural gradients. However, given the observational nature of the study, these findings should be interpreted as empirical associations rather than causal evidence or direct demonstrations of effects on ecosystem functions.
Supporting information
S1 Table. Total percentage of tree cover for each of the three classified categories within the 100 quadrants established across the eight livestock landscapes mosaics in the Colombian Amazon.
https://doi.org/10.1371/journal.pone.0345283.s001
(DOCX)
S2 Table. Effective sample size per functional diversity index.
https://doi.org/10.1371/journal.pone.0345283.s002
(DOCX)
S3 Table. List and matrix of morphological functional traits of the 22 granivorous bird species recorded across the three types of tree cover.
https://doi.org/10.1371/journal.pone.0345283.s003
(DOCX)
S4 Table. Matrix of morphological traits for each granivorous bird species recorded in the study.
https://doi.org/10.1371/journal.pone.0345283.s004
(DOCX)
S5 Table. Matrix of granivorous bird species assigned to two groups (small and large) based on the dendrogram generated by hierarchical cluster analysis using the ward.D algorithm.
https://doi.org/10.1371/journal.pone.0345283.s005
(DOCX)
S1 Code. R code of the GLMM with mixed-effects negative binomial regression model to evaluate the effect of tree cover types on the richness and abundance of granivorous birds.
https://doi.org/10.1371/journal.pone.0345283.s006
(R)
S2 Code. R Code for RLQ and Fourth Corner analysis between environmental matrices of tree cover types and functional traits of granivorous birds.
https://doi.org/10.1371/journal.pone.0345283.s007
(R)
S3 Code. R Code of the General and Mixed Linear Models (GMLM) of functional diversity indices of granivorous birds across three types of tree cover.
https://doi.org/10.1371/journal.pone.0345283.s008
(R)
S1 Data. Data matrix used in the R code for GLMM analysis with mixed-effects negative binomial regression model.
https://doi.org/10.1371/journal.pone.0345283.s009
(TXT)
S2 Data. Environmental, species, and trait matrices for the R code for RLQ and Fourth Corner analysis.
https://doi.org/10.1371/journal.pone.0345283.s010
(XLSX)
S3 Data. Data matrix used in the R code for GMLM of functional diversity indices.
https://doi.org/10.1371/journal.pone.0345283.s011
(CSV)
Acknowledgments
The authors thank the Universidad de la Amazonia (Caquetá, Colombia) and the Centro de Investigación de la Biodiversidad Andino Amazónica -INBIANAM for providing the equipment used for data collection. We also express our gratitude to the Comité Departamental de Ganaderos del Caquetá for facilitating access to the study sites, to biologist and illustrator Viviana Tello M., for the bird illustrations, and to the Semillero de Investigación Biodiversidad y Servicios Ecosistémicos -BySE for their support during field data collection.
References
- 1. Giraldo LP, Chará J, Velásquez-Valencia A, Ordoñez-García M. Vegetation structure and bird diversity in silvopastoral systems of the Amazonian piedmont of Colombia. Agrofor Syst. 2024;98:1–14.
- 2.
IDEAM (Instituto de Hidrología, M y EA). Monitoreo de la deforestación en Colombia, resultados año 2023. Bogotá: IDEAM. 2024.
- 3. Ferreira A. Amazon deforestation: Drivers, damages, and policies. Land Use Policy. 2026;160:107810.
- 4.
Freiría H, Nin Pratt A, Muñoz G. Productividad y eficiencia en la producción ganadera pastoril en América Latina: los casos de Bolivia y Colombia. Washington (DC): Inter-American Development Bank. 2021.
- 5. Rivera-Gomez V, Sánchez V, Henriquez C. Good livestock practices (BPG) in dual-purpose livestock farms in the Colombian Amazon. Rev Ambient Agua Aire Suelo. 2023;14(2):71–97.
- 6.
Torrijos-Rivera R. Cifras de contexto ganadero Caquetá 2025. Florencia (Colombia). 2025. https://consonante.org/wp-content/uploads/2025/07/CDGC-contexto-2025pl.pdf
- 7. Velasquez-Valencia A, Bonilla-Gomez MA. Influence of the configuration and heterogeneity of agroforestry and silvopastoral mosaics on bird communities in the Andean Amazon of Colombia. Rev Biol Trop. 2019;67(1):306–20.
- 8. Lezama Ahumada K, García Suabita W, Meza Elizalde MC. Degradación de bosques por pastoreo en el norte de la Amazonia colombiana. Investig Agrar. 2022;24(2):95–105.
- 9. Silveira JG da, Oliveira Neto SN de, Canto ACB do, Leite FFGD, Cordeiro FR, Assad LT, et al. Land Use, Land Cover Change and Sustainable Intensification of Agriculture and Livestock in the Amazon and the Atlantic Forest in Brazil. Sustainability. 2022;14(5):2563.
- 10. Agudelo-Hz WJ, Castillo-Barrera NC, Uriel MG. Scenarios of land use and land cover change in the Colombian Amazon to evaluate alternative post-conflict pathways. Sci Rep. 2023;13(1):2152. pmid:36750688
- 11. Hernández-Nuñez HE, Suárez JC, Andrade HJ, Acosta JRS, Núñez RD, Gutiérrez DR, et al. Interactions between climate, shade canopy characteristics and cocoa production in Colombia. Front Sustain Food Syst. 2024;8.
- 12. Cole RJ, Werden LK, Arroyo FC, Quirós KM, Cedeño GQ, Crowther TW. Forest restoration in practice across Latin America. Biol Conserv. 2024;294:110579.
- 13. Flores BM, Montoya E, Sakschewski B, Nascimento N, Staal A, Betts RA, et al. Critical transitions in the Amazon forest system. Nature. 2024;626(7999):555–64. pmid:38356065
- 14. López J, Qian Y, Murillo-Sandoval PJ, Clerici N, Eklundh L. Landscape connectivity loss after the de-escalation of armed conflict in the Colombian Amazon (2011–2021). Glob Ecol Conserv. 2024;54:e03094.
- 15. Messias CG, de Almeida CA, Silva DE, Soler LS, Maurano LE, Camilotti VL. Unaccounted non-forest vegetation loss in the Brazilian Amazon. Commun Earth Environ. 2024;5:456.
- 16. Franco MA, Rizzo LV, Teixeira MJ, Artaxo P, Azevedo T, Lelieveld J, et al. How climate change and deforestation interact in the transformation of the Amazon rainforest. Nat Commun. 2025;16(1):7944. pmid:40897684
- 17. Ribeiro AFS, Santos L, Randerson JT, Uribe MR, Alencar AAC, Macedo MN, et al. The time since land-use transition drives changes in fire activity in the Amazon-Cerrado region. Commun Earth Environ. 2024;5(1).
- 18. Pintado K, Knoke T. Spatializing land-use scenarios to inform decision-making in an Ecuadorian mountain forest landscape. J Environ Manage. 2025;388:125956. pmid:40440940
- 19. Valbuena D, Santander J, García-Estévez J, Franco N. Policy gaps in addressing drivers of tropical deforestation: a Colombian Amazon case study. Land Use Policy. 2025;158:105991.
- 20.
Tilman D. Functional diversity. 1a ed. San Diego: Encyclopedia of Biodiversity. 2001.
- 21. Böhning-Gaese K. Determinants of avian species richness at different spatial scales. J Biogeogr. 1997;24(1):49–60.
- 22. MacArthur RH, MacArthur JW. On Bird Species Diversity. Ecology. 1961;42(3):594–8.
- 23. Roth RR. Spatial Heterogeneity and Bird Species Diversity. Ecology. 1976;57(4):773–82.
- 24. Willson MF. Avian Community Organization and Habitat Structure. Ecology. 1974;55(5):1017–29.
- 25. Fahrig L. Effects of habitat fragmentation on biodiversity. Annu Rev Ecol Evol Syst. 2003;34:487–515.
- 26. Haddad NM, Brudvig LA, Clobert J, Davies KF, Gonzalez A, Holt RD, et al. Habitat fragmentation and its lasting impact on Earth’s ecosystems. Sci Adv. 2015;1(2):e1500052. pmid:26601154
- 27. Cadotte MW, Carscadden K, Mirotchnick N. Beyond species: functional diversity and ecological processes. J Appl Ecol. 2011;48(5):1079–87.
- 28. Ramos-Mosquera V, López-Delgado E, Moreno-Palacios M. Fragmented habitats, fragmented functions: habitat structure and Andean bird communities. Ecologies. 2025;6(3):52.
- 29. Peña R, Schleuning M, Miñarro M, García D. Variable relationships between trait diversity and avian ecological functions in agroecosystems. Funct Ecol. 2023;37(1):87–98.
- 30. Adelino JRP, Calsavara LC, Willrich G, Rosa GLM, Lima MR, Dos Anjos L. Ecosystem functions of birds as a tool to track restoration efficiency in Brazil. Ornithol Res. 2020;28(1):38–50.
- 31. Weeks TL, Walkden PA, Edwards DP, Lees AC, Pigot AL, Purvis A, et al. Land-use change undermines the stability of avian functional diversity. Nature. 2026;649(8096):381–7. pmid:41299180
- 32. He Q, Jiang X, Feng Y, Shao J, Yang X, Yang X. Avian phylogenetic and functional diversity in urban parks of Nanjing, China. Avian Res. 2025;16(4):34.
- 33. Das S, Srivastava A, Hore U. Agricultural land use impacts diversity and structure of farmland birds. Agric Ecosyst Environ. 2025;381:109221.
- 34. Basile M, Augustinus BA, Brockerhoff EG. Tree cover and geographic origin drive bird functional groups in urban environments. Wildl Biol. 2025.
- 35. Lees AC, Peres CA. Gap‐crossing movements predict species occupancy in Amazonian forest fragments. Oikos. 2009;118(2):280–90.
- 36. Sekercioglu CH. Increasing awareness of avian ecological function. Trends Ecol Evol. 2006;21(8):464–71. pmid:16762448
- 37. García-Llamas P, Rangel TF, Calvo L, Suárez-Seoane S. Linking functional traits and environmental filters in terrestrial vertebrates. PLoS One. 2019;14(2):e0211684.
- 38. Xu Y, Wu Z, Liu X, Han Q, Liu P, Wang D. Patch shape, connectivity and functional groups moderate bird diversity responses. Glob Ecol Conserv. 2024;55:e03204.
- 39. Alvarado Sandino CO, Barnes AP, Sepúlveda I, Garratt MPD, Thompson J, Escobar-Tello MP. Adoption of silvopastoral agroforestry in the Colombian Amazon. Scientific Reports. 2023;13:18462.
- 40. Quiroga S, Suárez C, Hernanz V, Aguiño JE, Fernández-Manjarrés JF. Post-conflict policies and socio-ecological restoration in Colombia. For Policy Econ. 2024;163:103144.
- 41. Morante-Filho JC, Faria D, Mariano-Neto E, Rhodes J. Birds in Anthropogenic Landscapes: The Responses of Ecological Groups to Forest Loss in the Brazilian Atlantic Forest. PLoS One. 2015;10(6):e0128923. pmid:26083245
- 42. Ahmed T, Khan A. Avifaunal feeding guild responses to landscape compositional heterogeneity in a forest mosaic of Uttarakhand, India. J Trop Ecol. 2022;38(4):183–93.
- 43. Díaz-Cháux JT, Velasquez-Valencia A, Martínez-Salinas A, Casanoves F. Functional Diversity and Ecosystem Services of Birds in Productive Landscapes of the Colombian Amazon. Diversity. 2025;17(5):305.
- 44. Bregman TP, Lees AC, MacGregor HEA, Darski B, de Moura NG, Aleixo A, et al. Using avian functional traits to assess the impact of land-cover change on ecosystem processes linked to resilience in tropical forests. Proc Biol Sci. 2016;283(1844):20161289. pmid:27928045
- 45. Tobias JA, Sheard C, Pigot AL, Devenish AJM, Yang J, Sayol F, et al. AVONET: morphological, ecological and geographical data for all birds. Ecol Lett. 2022;25(3):581–97. pmid:35199922
- 46. Rurangwa ML, Niyigaba P, Tobias JA, Whittaker RJ. Functional and phylogenetic diversity of an agricultural matrix avifauna: The role of habitat heterogeneity in Afrotropical farmland. Ecol Evol. 2022;12(7):e9024. pmid:35822114
- 47. Whelan CJ, Wenny DG, Marquis RJ. Ecosystem services provided by birds. Ann N Y Acad Sci. 2008;1134:25–60. pmid:18566089
- 48. Wenny DG, DeVault TL, Johnson MD, Kelly D, H. Sekercioglu C, Tomback DF, et al. The Need to Quantify Ecosystem Services Provided by Birds. The Auk. 2011;128(1):1–14.
- 49. Karp DS, Rominger AJ, Zook J, Ranganathan J, Ehrlich PR, Daily GC. Intensive agriculture erodes β-diversity at large scales. Ecol Lett. 2012;15(9):963–70. pmid:22727063
- 50. Thiel S, Tschapka M, Heymann EW, Heer K. Vertical stratification of seed-dispersing vertebrates and plant interactions in tropical forests. Biol Rev Camb Philos Soc. 2021;96(2):454–69.
- 51. Lakatos T, Báldi A, Gallé R, Korányi D, Kovács I, László Z, et al. Functional trait filtering and decline in species richness in urban parks hinder ground-breeding and insectivorous birds. Urban Forestry & Urban Greening. 2025;112:128988.
- 52. Lakatos T, Báldi A, Benkő Z, Gallé R, Korányi D, Kovács I. Landscape complexity and edge effects shape bird community composition in villages. Ecol Indic. 2025;176:112993.
- 53. Navarro AB, Bogoni JA, Moreira MZ, Silveira LF. Intraguild niche partitioning in granivorous birds from the late past. Avian Research. 2023;14:100075.
- 54. Newbold T, Scharlemann JPW, Butchart SHM, Sekercioğlu CH, Alkemade R, Booth H, et al. Ecological traits affect the response of tropical forest bird species to land-use intensity. Proc Biol Sci. 2013;280(1750):20122131. pmid:23173205
- 55. Tscharntke T, Klein AM, Kruess A, Steffan-Dewenter I, Thies C. Landscape perspectives on agricultural intensification and biodiversity. Ecol Lett. 2005;8(8):857–74.
- 56. Kortmann M, Chao A, Schaefer HM, Blüthgen N, Gelis R, Tremlett CJ. Sample coverage affects diversity measures along a recovery gradient of abandoned agriculture. J Appl Ecol. 2025;62(3):480–91.
- 57. Manzoli EC, Gaspar LP, Melo MA, Adorno BFCB, Ribeiro MC, Piratelli AJ. Environmental Conservation. 2024;51(2):95–103.
- 58. Villéger S, Mason NWH, Mouillot D. New multidimensional functional diversity indices for functional ecology. Ecology. 2008;89(8):2290–301.
- 59. Laliberté E, Legendre P. A distance-based framework for measuring functional diversity from multiple traits. Ecology. 2010;91(1):299–305. pmid:20380219
- 60. Schleuter D, Daufresne M, Massol F, Argillier C. A user’s guide to functional diversity indices. Ecol Monogr. 2010;80(3):469–84.
- 61. Mason NWH, Mouillot D. Functional Diversity Measures. Encyclopedia of Biodiversity. Elsevier. 2013:597–608.
- 62.
Laliberté E, Legendre P. Measuring functional diversity from multiple traits. Montréal: Université de Montréal. 2014.
- 63.
Wang Y, Naumann U, Eddelbuettel D, Wilshire J, Warton D, Byrnes J, et al. mvabund: statistical methods for analysing multivariate abundance data. Sydney: UNSW; 2022.
- 64. Díaz-Cháux JT, Velasquez-Valencia A, Casanoves F. Influence of landscape structure on carbon storage in agroforestry systems with cacao and silvopastoral systems in the Colombian Amazon. PLoS One. 2025;20(6):e0325477. pmid:40549732
- 65.
IDEAM (Instituto de Hidrología, M y EA). Leyenda nacional de coberturas de la tierra: metodología CORINE land cover adaptada para Colombia. Bogotá (Colombia). 2010.
- 66.
SINCHI (Instituto Amazónico de Investigaciones Científicas). Coberturas de la tierra para la región Amazónica, escala 1:25 000. Bogotá (Colombia). 2025.
- 67. Van der Werff H, Van der Meer F. Sentinel-2A MSI and Landsat 8 OLI Provide Data Continuity for Geological Remote Sensing. Remote Sensing. 2016;8(11):883.
- 68.
Esri. ArcMap version 10.8. Redlands (CA): Environmental Systems Research Institute. 2024.
- 69.
QGIS Development Team. QGIS geographic information system version 3.40. Open Source Geospatial Foundation. 2022.
- 70.
Ralph CJ, Geupel GR, Pyle P, Martin TE, DeSante DF, Milá B. Manual de métodos de campo para el monitoreo de aves terrestres. Albany (CA). 1996.
- 71. Sólymos P, Matsuoka SM, Bayne EM, Lele SR, Fontaine P, Cumming SG. Calibrating indices of avian density from non-standardized survey data. Methods Ecol Evol. 2013;4(11):1047–58.
- 72. Pliscoff P, Simonetti JA, Grez AA, Vergara PM, Barahona-Segovia RM. Defining corridors for movement of multiple species in forest–plantation landscapes. Glob Ecol Conserv. 2020;23:e01109.
- 73.
Remsen JV Jr. A classification of the bird species of South America. South American Classification Committee. 2025. https://www.museum.lsu.edu/~Remsen/SACCBaseline.htm
- 74.
Ayerbe Quiñones F. Guía de aves de Colombia. 3ra ed. Bogotá: WCS Colombia. 2023.
- 75. Echeverry-Galvis MÁ, Acevedo-Charry O, Avendaño JE, Gómez C, Stiles FG, Estela FA, et al. Ornitol Colomb. 2022;22(1):25–51.
- 76.
McMullan M. Field guide to the birds of Colombia. 3rd ed. Bogotá: McMullan Birding. 2023.
- 77. Buckland ST. Distance sampling: estimating abundance of biological populations. J R Stat Soc Ser A. 1993;41(1):241–66.
- 78.
R Core Team. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. 2022.
- 79. Chao A, Gotelli NJ, Hsieh TC, Sander EL, Ma KH, Colwell RK, et al. Rarefaction and extrapolation with Hill numbers. Ecological Monographs. 2014;84(1):45–67.
- 80. Chao A, Henderson PA, Chiu CH, Moyes F, Hu KH, Dornelas M. Measuring temporal change in alpha diversity integrating taxonomic, phylogenetic and functional diversity. Methods Ecol Evol. 2021;12(10):1926–40.
- 81.
Di Rienzo JA, Balzarini M, Gonzalez L, Casanoves F, Tablada M, Robledo CW. Infostat. Córdoba (Argentina): Universidad Nacional de Córdoba. 2020.
- 82. Dray S, Dufour AB. The ade4 package: implementing the duality diagram for ecologists. J Stat Softw. 2007;22(4):1–20.
- 83. Dray S, Legendre P. Testing the species traits-environment relationships: the fourth-corner problem revisited. Ecology. 2008;89(12):3400–12. pmid:19137946
- 84. Bonfim FCG, Dodonov P, Cazetta E. Landscape composition is the major driver of taxonomic and functional diversity of tropical frugivorous birds. Landscape Ecology. 2021;36(9):2535–47.
- 85.
Pla L, Casanoves F, Di Rienzo J. Quantifying functional biodiversity. Dordrecht: Springer. 2012.
- 86. Villéger S, Grenouillet G, Brosse S. Decomposing functional beta-diversity reveals low functional turnover in European fish assemblages. Global Ecology and Biogeography. 2013;22(6):671–81.
- 87. Chao A, Jost L. Coverage-based rarefaction and extrapolation. Ecology. 2012;93(12):2533–47.
- 88. Hsieh TC, Ma KH, Chao A. iNEXT: an R package for rarefaction and extrapolation of species diversity. Methods Ecol Evol. 2016;7(12):1451–6.
- 89. Echeverri A, Frishkoff LO, Gomez JP, Zook JR, Juárez P, Naidoo R. Precipitation and tree cover gradient’s structure avian alpha diversity in northwestern Costa Rica. Diversity and Distributions. 2019;25(8):1222–33.
- 90. Espejo N, Morales N. Variation of taxonomic and functional avian diversity across successional stages in a dry tropical forest of Colombia. Caldasia. 2019;41(1):108–23.
- 91. Htay T, Røskaft E, Ringsby TH, Ranke PS. Spatio-temporal variation in avian taxonomic, functional and phylogenetic diversity in a wetland ecosystem. Biodivers Conserv. 2023;32(8–9):2841–67.
- 92. Martínez-Núñez C, Martínez-Prentice R, García-Navas V. Land-use diversity predicts regional bird taxonomic and functional richness worldwide. Nat Commun. 2023;14(1):1320. pmid:36899001
- 93. Sarmiento-Garavito LP, García-Monroy JS, Carvajal-Cogollo JE. Taxonomic and functional diversity of birds in a rural Andean forest landscape. Neotrop Biol Conserv. 2022;17(1):39–57.
- 94. Nieto A, Kerwer L, Tinoco BA, Neuschulz EL, Schleuning M. Variation in bird diversity between elevations and land-use types. Acta Oecologica. 2026;130:103967.
- 95. Verea C. A case of partial leucism in the Blue-black Grassquit (Volatinia jacarina) from Venezuela. Huitzil Rev Mex Ornitol. 2020;21(2).
- 96. da Silva TW, Fontana CS. Success of active restoration in grasslands: a bird case study in southern Brazil. Restoration Ecology. 2020;28(3):512–8.
- 97. Riegert J. Detectability of birds under different sampling efforts during the breeding season. J Vertebr Biol. 2022;71:22027.
- 98. Baker DJ, Stillman RA, Smith BM, Bullock JM, Norris KJ. Vigilance and functional response of granivorous foragers. Funct Ecol. 2010;24(6):1281–90.
- 99. Maya-Elizarrarás E, Schondube JE. Birds, cattle and bracken ferns: community responses to cattle grazing. Biotropica. 2015;47(2):236–45.
- 100. Rodríguez-Maturino A, Martínez-Guerrero JH, Chairez-Hernández I, Pereda-Solís ME, Pinedo-Álvarez A. Grassland habitat variables associated with granivore bird density. Agroproductividad. 2017;10(5):3–9.
- 101. Pigot AL, Trisos CH, Tobias JA. Functional traits reveal niche expansion along an elevational gradient in passerines. Proc R Soc B. 2016;283(1822):20152013.
- 102. Chapman PM, Tobias JA, Edwards DP, Davies RG. Contrasting impacts of land-use change on tropical forest birds. J Appl Ecol. 2018;55(4):1604–14.
- 103. Emer C, Galetti M, Pizo MA, Guimarães PRJ, Moraes S, Piratelli A. Seed-dispersal interactions in fragmented landscapes. Ecol Lett. 2018;21(4):484–93.
- 104. Barnagaud JY, Mazet N, Munoz F, Grenié M, Denelle P, Sobral M. Functional biogeography of dietary strategies in birds. Global Ecology and Biogeography. 2019;28(7):1004–17.
- 105. Whittingham MJ, Markland HM. Substrate effects on the functional response of an avian granivore. Oecologia. 2002;130(4):637–44.
- 106. Cediel F, Ruíz-Gutiérrez V, Echeverri A, Parra JL. Differential avian responses to coffee farming lead to community homogenization. bioRxiv. 2025.
- 107.
BirdLife International. The IUCN Red List of Threatened Species: Tinamus guttatus. 2019.
- 108. Chuva Álvarez V, Torres Jiménez J, Rodríguez-Galarza F, López Shiguango E, Mamallacta Dahua A, Chimbo López M, et al. Richness and activity of terrestrial bird orders in the Ecuadorian Amazon. Cienc Amazon. 2023;11(1–2):29–42.
- 109. Lu X, Jia Y, Wang Y. Effects of landscape composition and heterogeneity on bird richness: a global meta-analysis. Landscape Ecology. 2024;39:1–16.
- 110. Birch BDJ, Mills SC, Socolar JB, Martínez-Revelo DE, Haugaasen T, Edwards DP. J Appl Ecol. 2024;61(5):940–50.
- 111. Castellanos S, Jácome-Negrete I, Rodríguez F, Gaba F, Enqueri I, Nemquimo O, et al. Richness and abundance of terrestrial birds in the Ecuadorian Amazon. Rev Acad Colomb Cienc Exactas Fis Nat. 2023;47(183):335–46.
- 112. Socolar JB, Wilcove DS. Forest-type specialization strongly predicts avian responses to tropical agriculture. Proc Biol Sci. 2019;286(1913):20191724. pmid:31640506
- 113. Rigby E, Johnson D. Effects of imperfect detection on inferences from bird surveys. ACE. 2025;20(2).
- 114. Zamora-Marín JM, Zamora-López A, Calvo JF, Oliva-Paterna FJ. Comparing detectability patterns using multi-method occupancy models. Sci Rep. 2021;11:21567.
- 115. Hasui É, Martensen AC, Uezu A, Pimentel RG, Ramos FN, Ribeiro MC, et al. Population responses to habitat loss and fragmentation across species ranges. Perspect Ecol Conserv. 2024;22(1):43–54.
- 116. Eliason CM, Proffitt JV, Clarke JA. Early diversification of avian limb morphology. Evolution. 2023;77(2):342–54.
- 117. Smith OM, Kennedy CM, Echeverri A, Karp DS, Latimer CE, Taylor JM, et al. Complex landscapes stabilize farm bird communities. J Appl Ecol. 2022;59(4):927–941.
- 118. Díaz-Cháux JT, Velasquez-Valencia A. Uso de coberturas vegetales y movilidad de aves semilleras en un paisaje urbano. Momentos Cienc. 2009;6(1):70–9.
- 119. Quilarque E, Marín G, Carvajal Y, Ferrer YH. Diet components of Sporophila spp. in a Venezuelan forest ecotone. Bol Cent Investig Biol. 2010;44(2):161–72.
- 120. Wider S, Zemp DC. Bird and hoverfly communities respond to vegetation heterogeneity in wood-pastures. Biodivers Conserv. 2025;34(6):2139–62.
- 121. Jones HH, Gabriel Colorado Z, Robinson SK. Widespread bird species show idiosyncratic responses in residual body mass to selective logging and edge effects in the Colombian Western Andes. Ornithological Applications. 2022;124(4):duac038.
- 122. Frank C, Hertzog L, Klimek S, Schwieder M, Tetteh GO, Böhner HGS, et al. J Appl Ecol. 2024;61(5):987–99.
- 123. Guerrero I, Duque D, Oñate JJ, Pärt T, Bengtsson J, Tscharntke T, et al. Agricultural intensification affects birds’ trait diversity across Europe. Basic and Applied Ecology. 2024;74:40–8.
- 124. Sayol F, Neate-Clegg M, Sheard C, Pigot AL, Tobias JA. AVONICHE: a global dataset of dietary and foraging niches for birds. Global Ecology and Biogeography. 2026;35(1):e70197.
- 125. Marcolin F, Lakatos T, Gallé R, Batáry P. Fragment connectivity shapes bird communities through functional trait filtering in grasslands. Global Ecology and Conservation. 2021;28.
- 126. Sayer CA, Bullock JM, Martin PA. Dynamics of avian species and functional diversity in secondary tropical forests. Biological Conservation. 2017;211:1–9.
- 127. Coelho MTP, Raniero M, Silva MI, Hasui É. The effects of landscape structure on functional groups of Atlantic forest birds. The Wilson Journal of Ornithology. 2016;128(3):520–34.
- 128. Velásquez-Trujillo V, Betancurt-Grisales JF, Vargas-Daza AM, Lara CE, Rivera-Páez FA, Fontúrbel FE, et al. Bird Functional Diversity in Agroecosystems and Secondary Forests of the Tropical Andes. Diversity. 2021;13(10):493.
- 129. Lee M-B, Martin JA. Avian Species and Functional Diversity in Agricultural Landscapes: Does Landscape Heterogeneity Matter? PLoS One. 2017;12(1):e0170540. pmid:28125653
- 130. Matuoka MA, Benchimol M, Almeida-Rocha JM, Morante-Filho JC. Effects of anthropogenic disturbances on bird functional diversity: a global meta-analysis. Ecol Indic. 2020;116:106476.
- 131. Borrow M, Fuller S, Doohan B. Rebound from restoration: diversity of bird assemblages in revegetated and remnant rainforest patches. Ecol Evol. 2025;15(10):e7050.
- 132. Socolar JB, Mills SC, Gilroy JJ, Martínez-Revelo DE, Medina-Uribe CA, Parra-Sanchez E, et al. Tropical biodiversity loss from land-use change is severely underestimated by local-scale assessments. Nat Ecol Evol. 2025;9(9):1643–55. pmid:40695982
- 133. Bradfer-Lawrence T, Gardner N, Dent DH. Canopy bird assemblages are less influenced by habitat age and isolation than understory assemblages. Ecol Evol. 2018;8(11):5586–97.
- 134. Jacoboski LI, Hartz SM. Functional and taxonomic diversity responses to afforestation of grasslands. Perspect Ecol Conserv. 2020;18(2):103–8.
- 135. Hua F, Wang W, Nakagawa S, Liu S, Miao X, Yu L, et al. Ecological filtering shapes the impacts of agricultural deforestation on biodiversity. Nat Ecol Evol. 2024;8(2):251–66. pmid:38182682
- 136. de Souza Leite M, Boesing AL, Metzger JP, Prado PI. Matrix quality determines strength of habitat-loss filtering on bird communities. J Appl Ecol. 2022;59(11):2790–802.
- 137. Eyster HN, Srivastava DS, Kreitzman M, Chan KMA. Functional traits and metacommunity theory reveal habitat filtering and competition. Ecography. 2022;2022(11):e06287.
- 138. Lima GR, Borges SH, Anciães M, Cornelius C. Ecomorphology and functional diversity of bird assemblages in Amazonian white-sand habitats. Acta Amazon. 2023;53(2):141–53.
- 139. Coddington CPJ, Cooper WJ, Mokross K, Luther DA. Forest structure predicts species richness and functional diversity in Amazonian mixed‐species bird flocks. Biotropica. 2023;55(2):467–79.
- 140. Jung K, Teuscher M, Böhm S, Wells K, Ayasse M, Fischer M. Supporting bird diversity and ecological function in managed systems. Frontiers in Environmental Science. 2024;12:1290846.
- 141. Kim JH, Park S, Kim SH, Kang K, Waldman B, Lee MH. Traditional agricultural landscapes shape bird functional diversity near the Korean DMZ. Ecol Evol. 2020;10(23):12973–82.
- 142. Nabi G, Wang Y, Lü L, Jiang C, Ahmad S, Wu Y, et al. Bats and birds as viral reservoirs: A physiological and ecological perspective. Sci Total Environ. 2021;754:142372. pmid:33254850
- 143. Tiang DCF, Morris A, Bell M, Gibbins CN, Azhar B, Lechner AM. Ecological connectivity in fragmented agricultural landscapes. Ecol Process. 2021;10(1):52.
- 144. Rodríguez-León CH, Roa-Fuentes LL, Sterling A, Suárez JC. Plant-Community Vulnerability in Highly Fragmented Landscapes Is Higher in Secondary Forests Than in Old Growth Forests in the Andean-Amazonian Transition. Plants (Basel). 2022;11(23):3284. pmid:36501322
- 145. Parajuli K, Ghimire P. Farmland Birds and their Importance in Agro ecosystem: A Review. Int Res Jnl MMC. 2024;5(4):158–68.
- 146. Kehoe L, Romero-Muñoz A, Polaina E, Estes L, Kreft H, Kuemmerle T. Nat Ecol Evol. 2017;1(8):1129–35.
- 147. Alvarado F, Escobar F, Williams DR, Arroyo-Rodríguez V, Escobar-Hernández F. Livestock intensification and landscape structure in tropical biodiversity. J Appl Ecol. 2018;55(1):185–94.
- 148. Blaum N, Mosner E, Schwager M, Jeltsch F. How functional is functional? Ecological groupings in animal ecology. Biodivers Conserv. 2011;20(11):2333–45.
- 149. Luck GW, Carter A, Smallbone L. Changes in bird functional diversity across land uses. PLoS One. 2013;8(5):e63942.
- 150. Schelsky WM. Research and conservation of forest-dependent tinamou species in Amazonia, Peru. Ornitol Neotrop. 2004;15:385–94.
- 151. Smith DG, Evans MJ, Scheele BC, Florance D, Siegrist A, Lang E, et al. Grazing control and revegetation increase bird biodiversity at farm dams. Biol Conserv. 2025;309:110882.
- 152. Filgueiras BKC, Peres CA, Melo FPL, Leal IR, Tabarelli M. Winner-Loser Species Replacements in Human-Modified Landscapes. Trends Ecol Evol. 2021;36(6):545–55. pmid:33685660
- 153. Brocardo CR, Rosa DCP, Castro AB, Rosa C, Torralvo K, Pequeno P. Responses of ground-dwelling birds to environmental variables and human pressure. Eur J Wildl Res. 2023;69(3):48.
- 154. Butler RW. Birds and People: A Symbiotic Relationship in Practice. Birds. 2024;5(2):328–40.
- 155. Pustkowiak S, Kwieciński Z, Lenda M, Żmihorski M, Rosin ZM, Tryjanowski P. Value of singular point elements for birds in agricultural landscapes. Biol Rev Camb Philos Soc. 2021;96(4):1386–403.
- 156. Bitani N, Cordier CP, Ehlers Smith DA, Ehlers Smith YC, Downs CT. Responses of bird functional communities to anthropogenic disturbance. For Ecol Manage. 2024;562:121004.
- 157. Smith OM, Kennedy CM, Owen JP, Northfield TD, Latimer CE, Snyder WE. Highly diversified crop-livestock farming systems reshape wild bird communities. Ecol Appl. 2020;30(2):e02031. pmid:31674710