Effects of habitat amount and fragmentation per se on mammals in a highly fragmented Colombian region

doi:10.21203/rs.3.rs-7294977/v1

Download PDF

Research Article

Effects of habitat amount and fragmentation per se on mammals in a highly fragmented Colombian region

https://doi.org/10.21203/rs.3.rs-7294977/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The effects of fragmentation on biodiversity are debated. Some studies find positive effects, while others link it to biodiversity loss. “Fragmentation per se” refers to habitat fragmentation independently of habitat loss or control of its effect, which ultimately relates more to how patches are organized (habitat configuration) on biodiversity. We tested the habitat amount hypothesis that postulates that habitat cover would be the main factor determining species diversity. We evaluated the effects of habitat amount and fragmentation per se of tropical dry forest in the Colombian Caribbean region on the taxonomic and functional species richness of medium- to large-bodied mammals. For this, we evaluated 34 landscapes with forest cover ranging from 5–90%. We calculated composition (forest cover) and configuration (forest edge density, number of patches, mean patch area) landscape structure metrics. In each landscape, we calculated total taxonomic richness, forest-dependent species richness, non-forest-dependent species richness, and functional richness using camera trap records. We found that the edge density had a positive effect for on total species richness and moderate positive effects on non-forest-dependent species, while, forest-dependent species richness was negatively affected by the number of patches and positively by forest cover and we found no significant or moderate effects on functional richness. These findings suggest that conservation efforts should focus on preserving and increasing the total amount of habitat and also take into account the configuration of the tropical dry forest in the Colombian Caribbean.

ecosystem functioning

functional diversity

landscape configuration

landscape composition

Anthropogenic activities, particularly the expansion of the agricultural frontier, are leading to continuous changes in the landscape structure, altering ecosystems functioning due to habitat loss and fragmentation (Rivas et al. 2024). As a consequence, the diversity of terrestrial mammals is negatively affected due to an increased state of vulnerability facilitated mainly by reduction of set of resources and necessary conditions to guarantee population persistence, modifying the taxonomic and functional diversity and spatial distribution patterns of terrestrial mammals (Magioli et al. 2021).

The concept of habitat fragmentation has historically been confused with habitat loss, but this is not necessarily true (Fahrig 2003). Habitat loss refers to the reduction in the amount of habitat and is considered the primary threat to terrestrial biodiversity (Díaz and Malhi 2022). On the other hand, fragmentation can be defined in two concepts. The first is as a process; it occurs when a continuous habitat is divided into smaller patches, isolated from one another by a matrix of habitat different from the original (Fahrig 2003). This definition of fragmentation is strongly correlated with habitat loss, leading to the conclusion that fragmentation in all cases has a negative effect on biodiversity (Fletcher et al. 2018). Other authors suggest that habitat fragmentation should be evaluated as a spatial pattern rather than as a process, allowing it to be distinguished from the effects of habitat loss (Smith et al. 2009). By controlling the amount of habitat, it is possible to isolate the specific effects of fragmentation (Fahrig 2003), which prevents the confusion between habitat loss and fragmentation per se (e.g., fragmentation independent of habitat amount; Fahrig 2017).

In recent years, a debate has emerged around the “habitat amount hypothesis”, which postulates that the total habitat amount is more determinant for explaining biodiversity distribution across fragmentated landscapes than the size and isolation of individual remnant patches (Fahrig 2013). Studies such as Watling et al. (2020). But other studies deny the habitat amount hypothesis, supporting the view that area, isolation, quality and patch configuration should be considered because they are important predictors of species richness (Hanski, 2015). Therefore, the debate has not only focused on the effects of fragmentation per se, but on which of the landscape elements is of greater importance and its effects on diversity patterns (Valente et al. 2023).

A key diversity metric for evaluating these effects is functional diversity, which is defined as the distribution of species in a functional space determined by functional traits, considering the presence and abundance of species in a community (de Bello et al. 2021). These functional traits are considered organism characteristics that serve as indicators of functions occurring at the individual scale, providing insights into both organismal and ecosystem functioning (Mouillot et al. 2014). For example, it has been demonstrated that species traits correlate with species sensitivity to landscape changes (Keinath et al. 2017), particularly for species that require specific conditions for their maintenance (e.g., species with large distribution areas and a preference for forest cover; Magioli et al. 2019). This is the case for mesopredators, which tend to be more sensitive to land-use changes than smaller species due to competition, reproductive strategies, and ecological niche specificity (Newbold et al. 2020). Therefore, functional diversity can be a suitable indicator of the effects of fragmentation and habitat loss on biodiversity. However, its study has received less attention than taxonomic diversity, which has primarily focused on species richness (Resende et al. 2024).

A good setting to address this problem is tropical dry forests (TDF), ecosystems threatened by land use expansion (Miles et al. 2006; Portillo-Quintero and Sánchez-Azofeifa 2010). Tropical dry forests harbor high biological diversity due to their structural complexity (Miles et al. 2006). In Colombia, near only 8% of the original TDF cover is estimated to remain due to historical anthropogenic influences (Pizano et al. 2016). The Caribbean region is where the largest TDF Colombian fragments are found, these cover up to ~ 6.000 ha, located between 0 and 650 m elevation (González-M. et al. 2020). But it was found that the transformation caused by urban growth, the maintenance of traditional land use types, and the occupation of large extensions for the establishment of large-scale agricultural systems promote the loss of natural cover (Carvajal-Cogollo 2014), and for this region, estimates suggest that only 1.5 to 8% of the original TDF cover remains (Etter et al. 2017), affected by social issues including armed conflict and conservation issues such as deforestation and the expansion of agriculture (Negret et al. 2019). Consequently, medium- to large-bodied mammals have been affected by habitat loss (Pardo et al. 2024).

Among the most threatened taxonomic groups are mammals, which are important as they provide balance to the ecosystem and are key in the trophic chain (Jorge et al. 2013; Lacher et al. 2019). In this context, efforts have sought to understand how medium- to large-bodied mammals respond to landscape transformation, mainly through studies of species richness and occurrences (Rios et al. 2021, 2022). Although it has been shown that the agricultural expansion has had negative effects on mammal species richness (Pardo et al. 2019). Also, it is recognized that small fragments in grassland matrices have favored the permanence of some species, because they take advantage of resources from heterogeneous landscapes such as agrosystems (Ferreira et al. 2018). Thus, remnant habitat patches immersed within cropland matrices can conserve a significant portion of species diversity (Beca et al. 2017), this suggests that some species could benefit from human-modified landscapes, and their responses may therefore vary depending on the landscape changes (Crooks 2002).

Few studies have focused on the effects of landscape transformation on the functional diversity of medium to large-sized mammals (Resende et al. 2024). Therefore, understanding how these species are distributed in transformed landscapes will support understanding potential changes the their taxonomic and functional richness (Meza-Joya et al. 2020). Identifying these effects can help guide biodiversity management and conservation practices (Valente et al. 2023), in particular, to maintain the provision of ecosystem services (Pizano and García 2014). The relationship between changes in the TDF landscape and mammal diversity is unclear. On the one hand, studies have focused on the patch scale (Pineda-Cendales et al. 2020), with a limited insight into the effects of habitat loss and fragmentation per se on species diversity. On the other hand, it is claimed that mammal conservation in dry forests depends on the available forest amount in anthropogenic landscapes (Pardo et al. 2024). However, no studies have been conducted to evaluate the effects of habitat loss and fragmentation per se at a landscape level on the taxonomic and functional diversity of medium- to large-bodied mammals in the TDF. Therefore, it is essential to identify how changes in the landscape structure may affect this diversity, in order to guide conservation practices adapted to the conditions of highly fragmented regions, such as the Colombian Caribbean.

In this study, we aimed to answer the following questions: how does fragmentation per se of the TDF affect the taxonomic (total species richness, forest-dependent species richness, and non-forest-dependent species richness) and functional richness of medium- to large-bodied mammals? Based on previous studies showing that forest cover is a strong predictor of mammal diversity in tropical forests (e.g., Pardo et al. 2024), we predict that both taxonomic and functional richness will be positively related to forest cover. Regarding the effects of fragmentation per se, we expect that both the number of patches and edge density will positively influence mammal richness, as they may increase habitat heterogeneity and resource availability for certain generalist and edge-adapted species (Fahrig 2017). Additionally, we anticipate that patch size will have no significant effect, since previous evidence suggests that forest amount, rather than patch size, is the main driver of species richness in fragmented landscapes (Fahrig 2013).

2.1. Study area

We conducted our study in the Colombian Caribbean region, in the north of the country. This region extends from the Gulf of Urabá in the west to the La Guajira Peninsula in the east. The Caribbean region is made up of three Biogeographic Units (Archipelago of San Andrés and Providencia, Pre-Caribbean Arid Belt, Sierra Nevada de Santa Marta), the sites were located within the Pre-Caribbean Arid Belt and north of the Sierra Nevada de Santa Marta (SNSM) where remnant patches of TDF are located (González-M. et al. 2018; Fig. 1). This region stands out for the presence of TDF with an area of approximately 417,838 ha, being the largest in Colombia (González-M. et al. 2018). The TDF in the Caribbean region is strongly fragmented, mainly due to the establishment of agricultural systems (Pizano and García 2014). Despite this, it has forest fragments ranging from 600 ha to1,530.4 ha (Pizano and García 2014). These areas are considered a biogeographic unit in the country characterized by similarity in environmental factors and their floristic composition (González-M. et al. 2018).

2.2. Data collection

We obtained the data of medium- to large-bodied mammals from the initiative “Mammals of the Colombian Caribbean: A Dataset on the Wild Mammal Assemblage in the Caribbean Region of Colombia” (Chacón-Pacheco et al. in prep.), whereby was consolidated using camera-trapping. This dataset includes information from 170 TDF sites where surveys were conducted for medium-sized (1–30 kg) and large (> 30 kg) mammals (Pineda-Munoz et al. 2016). To prevent overlap and minimize spatial autocorrelation (Beale et al. 2010), a subset of these sites was randomly selected. We selected the sites using the 'Proximity Analysis' tool in ArcGIS Pro, ensuring that each chosen site was separated by a minimum distance of at least 3 km.

2.3. Multiscale landscape structure analysis

For the calculation of landscape structure metrics, we used the supervised land use classification from ESA's Sentinel-2 Earth observation mission satellites with a spatial resolution of 10 m (Main-Knorn et al. 2017) according to the year in which the sites were sampled (2017–2023). For the digitization, we used ArcGIS 10.8 geographic information software. This classification has seven land cover and land use classes: water bodies, forest, secondary vegetation, crops, built-up areas, bare soils and grasslands. From which we extracted only the forest class, which is the most important indicator when analyzing the habitat quantity hypothesis and the effects of fragmentation per se (Fahrig 2013, 2017).

The study design was conducted in a site-landscape scale, where we measured the response variables (taxonomic and functional richness) at each site and assessed how they respond to the explanatory variables at the landscape scale (McGarigal and Cushman 2002). Sites were chosen with similar ranges of forest cover, but with different fragmentation gradient, in order to evaluate the independent effect of fragmentation per se (Fahrig 2017). In addition, we selected sites that had a sampling effort of at least 30 camera-days.

For each site we defined five buffers (i.e. sites with radius of different sizes, 0.5, 1, 2, 4, 6 km) around the centroid of the points where the cameras were installed, looking to identify the appropriate spatial scale at which the response variable is best explained according to the explanatory variables (the "scale of effect"; Martin and Fahrig 2012). We considered the landscape as a spatial area with a diameter that substantially exceeds the dispersal distance of the species of interest (Driscoll et al. 2013). It was considered that for TDF in the Colombian Caribbean region, medium- to large-bodied mammals have been reported with maximum daily distances traveled of less than 0.5 km (e.g., lesser capybara, Hydrochoerus isthmius; Chacón et al. 2013) and more than 20 km (jaguar, Panthera onca; Thompson et al. 2022). Although, jaguar populations have been extirpated from most of the Colombian Caribbean and those few areas where they are found are mainly associated with tropical rainforest (González-Maya et al. 2013). Therefore, we selected a maximum buffer of 6 km following studies that evaluate the relationship between landscape and the presence of medium- to large-bodied mammals in the Neotropics (Regolin et al. 2017; Rios et al. 2021, 2022), because so far no studies in Colombia have evaluated this relationship at multiple scales.

For each site, we calculated four explanatory variables, a) Edge density (ED) corresponding to the sum of all forest patch edges relative to the landscape area. b) Number of patches (NP), which describes the degree of forest fragmentation within the landscape (McGarigal and Cushman 2002). According to Wang et al. (2014) these metrics correlated weakly with habitat amount and allows us to distinguish from the effects of fragmentation. We included c) Forest cover (%) metric from each buffer (FC; representing the habitat amount) and used it as a fixed variable in the models as recommended by Smith et al. (2009). In addition, we added the d) Mean patch area (PS), this in the interest of encompassing questions such as the SLOSS debate (few large patches or several small ones; Fahrig, 2020). Metrics were calculated in R using the landscapemetrics package (Hesselbarth et al. 2019).

2.4. Taxonomic and functional richness

For each landscape, we obtained the total species richness by summing all the recorded species. Note that our data excluded species whose main preference is aquatic (otter, Lontra annectens) or arboreal (e.g., primates such as Cebus capucinus) environments, except Sciurus granatensis, which is an arboreal species, but frequently descends to the ground (Patton et al. 2015). We also excluded species of the genus Metachirus, because their taxonomy is poorly known in the region and photo-trapping is not the main sampling method for the study of this group. Additionally, we consider Sylvilagus sp. as a complex of species, given the low resolution at the specific level. Species were classified into two categories according to their dependence on forest. The first category includes forest-dependent species, which mainly inhabit undisturbed forest areas and are usually found in the forest interior (e.g., ñeque, Dasyprocta punctata). The second category includes non-forest-dependent species, which are more resistant to disturbances caused by human activities (e.g., common opossum, Didelphis marsupialis) or prefer open areas (e.g., giant anteater, Myrmecophaga tridactyla). For this classification we follow studies conducted on the species in the region or country (Chacón et al. 2013; Suárez-Castro and Ramírez-Chaves 2015; Rojano et al. 2016; Chacón Pacheco et al. 2017; Chacón-Pacheco et al. 2021; Zárrate-Charry et al. 2022).

To calculate functional richness, we use four functional traits:

Body mass (kg). This trait provides information on the demand for trophic resources, energy expenditure, energy flow between trophic levels (Mena 2010). Therefore, it explains the effect that landscape changes will have in terms of amount of nutrients dispersed, amount food consumed and spatial range of impact (Castillo-Figueroa & Pérez-Torres, 2021). We obtained the values for this trait from the EltonTraits database (Wilman et al. 2014).

Trophic guild. We classified species into carnivore, herbivore, insectivore and omnivore. This trait reveals information on the energy and material flow between species and has an effect on resource partitioning and habitat use (Mena 2010). Values were obtained from the PanTHERIA database (Jones et al. 2009).

Activity patterns. We classify species as nocturnal, cathemeral and diurnal (Jones et al. 2009). It reveals how mammals use space and time spent to maintain metabolic demands, which influences the spatiotemporal distribution of resources and species interactions (Pardo et al. 2021).. We obtained the information from the PanTHERIA database (Jones et al. 2009).

Home range. Defined as the size of the area (in km²) within which the daily activities of individuals or groups (of any type) are typically restricted. This feature provides information on the demand for trophic and spatial resources, predation and anti-predation strategies and territoriality (Mena 2010). Allows us to know the effect of as underlying ecological processes that affect intraspecific and interspecific variation in space use (Ofstad et al. 2016). We obtained this trait from the PanTHERIA database (Jones et al. 2009) and in scientific papers for the species, ñeque (D. punctata; (Aliaga-Rossel et al. 2008), paca (Cuniculus paca; Ulloa et al. 1999) and lesser capybara (H. isthmius; Chacón et al. 2013).

For Sylvilagus sp we used the functional traits of the eastern cottontail (Sylvilagus floridanus), whose distribution is documented in the Colombian Caribbean region (Avendaño-Maldonado et al. 2021; Chacón Pacheco et al. 2021, 2022).

To estimate functional richness (FRic), we used the mFD package (Magneville et al. 2022) in R, calculating the functional distance between pairs of species using the Gower metric, which allows weighting different types of traits (Magneville et al. 2022). These distances were then used to construct a functional space, defined as a multidimensional Euclidean space where each axis represents a synthetic trait gradient derived from Principal Coordinate Analysis (PCoA), and the position of each species reflects its trait values (Mouillot et al. 2013). Within this space, FRic corresponds to the convex hull volume occupied by the set of species in each assemblage. Since the number of axis influences functional diversity patterns, we evaluated spaces of up to 10 dimensions and selected those offering the best representation based on two criteria: (1) the mean absolute deviation (MAD), between the initial Gower distance matrix and the distances in the reduced functional space, which quantifies how well the multidimensional space preserves the original pairwise distances (i.e. lower MAD indicates better representation), and (2) the UPGMA method, a hierarchical clustering approach that merges clusters based on the average distance between all their species (Magneville et al. 2022). We then calculated correlations between traits and functional axis and visualized the dispersion of species across the 34 sites (Fig. S1; Magneville et al., 2022).

2.5. Data analysis

The determine the spatial scale at which landscape metrics (forest cover, number of patches, edge density and mean patch area) influence mammalian diversity, we assessed the relationship between explanatory variables (fragmentation per se metrics and habitat amount) and response variables (taxonomic and functional richness of medium- to large-bodied mammals) across multiple spatial extents (0.5, 1, 4, and 6 km). The explanatory variables included correspond according to (Fahrig 2003, 2013, 2017) to both habitat amount (measured as forest cover) and fragmentation per se, represented by three landscape metrics: number of patches, edge density, and mean patch area. We applied natural log transformation (ln) to all explanatory variables to normalize their distribution, reduce variance heterogeneity and improve model fit (Zuur et al. 2009). We performed and fitted generalized linear models (GLM) to accommodate residual deviations from normality (Zuur et al. 2009), using the 'multifit' function (Huais 2018), which simultaneously runs multiple statistical models for each response variable with each of the explanatory variables at the different spatial scales. We identified the lowest value of the Akaike Information Criterion (AICc), in order to choose the best scale for each of the explanatory variables (Martin and Fahrig 2012).

Then for each set of selected variables, we applied a variance inflation factor (VIF) test (Dormann et al. 2012), with the help of the 'vif' function of the car R package (Fox et al. 2012). We considered that collinearity existed between variables when VIF was ≥ 5 (Dormann et al. 2012). When we included mean patch area all variables except edge density presented high collinearity and when we excluded this variable collinearity was low in the other variables. Therefore, we did not take it into account in the subsequent analyses (Table S2). To assess the effects of fragmentation per se on taxonomic and functional richness, we first fitted GLMs containing all explanatory variables for to each response variable. We used different distribution families, depending on the nature of the response variable being evaluated and on the overdispersion with Pearson residuals in all global models. For taxonomic richness, we used the Poisson family, and for functional richness, the Gaussian family (Table S3).

Additionally, we checked for residual spatial autocorrelation using Moran's, with the help of the spdep R package (Bivand and Wong 2018). We found that spatial autocorrelation was significant in the global model residuals of non-forest-dependent species (Moran's I test = 0.39, p-value = 0.02; Table S3). To handle this, we performed a visual check of the residual dispersion using bubble plots showing Pearson residuals with respect to site coordinates (Zuur et al. 2009). From this inspection, it was concluded that there was no obvious spatial pattern in the correlations and that this could be due to unstructured random variation between sites, rather than spatial variation (Borcard et al. 2011). Therefore, we consider the autocorrelation as weak and as unlikely to introduce a systematic bias in model inference.

Subsequently, we performed a multi-model inference analysis with the overall models of each response variable (Burnham and Anderson 2002). For this, we used the 'dredge' function that compares all subsets of models, including all possible combinations of the explanatory variables together with a null model (only the intercept and residual variance). To control for sampling effort, this variable was included as an offset in the models. In total, 16 models were compared for each response variable. To determine the effect of each landscape metric on the response variables, we applied a model averaging approach to account for uncertainty in multi-model inference. We obtained the weighted mean of those of the best models (selected with a ∆AICc < 2) and the 95% confidence interval (Table S5), variables where 95% of the intercept did not intercept with zero we consider a substantial effect and where 75% did not intercept a moderate effect (Burnham and Anderson 2002). All these analyses were performed with the MuMIn package (Bartoń 2010) in R software.

We recorded 21 species of medium- to large-bodied mammals from 16 families and seven orders, from which 10 were forest dependent and 11 non-forest dependent species (Table 2). The order Carnivora had the highest richness with eight species, followed by Rodentia with four species. The orders with the lowest number of species were Didelphimorphia and Lagomorpha, each with only one species. The species recorded with the most independent records was the agouti (D. punctata, n = 319), the nine-banded armadillo (Dasypus fenestratus, n = 181) and the crab-eating raccoon (Procyon cancrivorus, n = 131). In contrast, the species with the fewest records were the puma (Puma concolor, n = 2), the giant anteater (M. tridactyla, n = 4), and the northern naked-tailed armadillo (Cabassous centralis, n = 12). Richness per site ranged from 1 to 14 species with a mean of seven across the study area. Most species were categorized as Least Concern (LC) according to the IUCN Red List (https://www.iucnredlist.org/), except for the margay (Leopardus wiedii), which is Near Threatened (NT), the giant anteater (M. tridactyla) classified as Vulnerable (VU), the northern naked-tailed armadillo (C. centralis) and the lesser capybara (H. isthmius), both classified as Data Deficient (DD) (Table 2).

Table 1
Landscape metrics measured in five buffers (0.5, 1, 2, 4, 6 km) for each of the 34 landscapes studied. The “Code” column is the function used for the calculation of the variable in *landscapemetrics* (Hesselbarth et al. 2019).
Variable	Equation	Code	Mean ± SD (Interval)
Mean patch area	\(\:{MN}_{c}=mean\:\left(area\left[{patch}_{ij}\right]\right)\)	lsm_c_area_mn	70.02 ± 121.73 (1.20-1150.45)
Forest cover	\(\:TA=\:\sum\:_{j=1}^{n}{a}_{ij}\)	lsm_c_ca	1871.98 ± 2551.06 (10.88–10179)
Edge density	\(\:ED=\frac{{\sum\:}_{k=1}^{m´}{e}_{ik}}{A}\times\:\text{10,000}\)	lsm_c_ed	34.12 ± 15.13 (0-82.06)
Number of patches	\(\:{\varvec{N}\varvec{P}}_{\varvec{C}}={\varvec{n}}_{\varvec{i}}\)	lsm_c_np	73.84 ± 120.43 (1-848)

Table 2
Medium- to large-bodied mammals recorded in the Colombian Caribbean Region, including the IUCN Red List category (https://www.iucn.org/) for each species (IUCN): Data Deficient (DD), Least Concern (LC), Near Threatened (NT), Vulnerable (VU) and their dependence on forest (FD) and if they are non-forest dependent (DT) species.
Taxa	Typo	IUCN
ARTIODACTYLA
Tayassuidae
Collared peccary (Dicotyles tajacu)	FD	LC
Cervidae
Santa Marta Corzuela (Mazama sanctaemartae)	FD	LC
White-tailed deer (Odocoileus cariacou)	DT	LC
CARNIVORA
Canidae
Crab-eating fox (Cerdocyon thous)	DT	LC
Felidae
Ocelot (Leopardus pardalis)	DT	LC
Margay (Leopardus wiedii)	FD	NT
Puma (Puma concolor)	FD	LC
Jaguarundi (Herpailurus yagouaroundi)	DT	LC
Mephitidae
Skunk (Conepatus semistriatus)	FD	LC
Mustelidae
Tayra (Eira barbara)	FD	LC
Procyonidae
Racoon (Procyon cancrivorus)	DT	LC
CINGULATA
Chlamyphoridae
Northern naked-tailed armadillo (Cabassous centralis)	FD	DD
Dasypodidae
Armadillo (Dasypus fenestratus)	DT	-
DIDELPHIMORPHIA
Didelphidae
Zarigüeya común (Didelphis marsupialis)	DT	LC
LAGOMORPHA
Leporidae
Cottontail (Sylvilagus sp)	DT	-
PILOSA
Myrmecophagidae
Giant anteater (Myrmecophaga tridactyla)	DT	VU
Tamandua (Tamandua mexicana)	FD	LC
RODENTIA
Caviidae
Lesser capybara (Hydrochoerus isthmius)	DT	DD
Cuniculidae
Paca (Cuniculus paca)	FD	LC
Dasyproctidae
Ñeque (Dasyprocta punctata)	FD	LC
Sciuridae
Squirrel (Sciurus granatensis)	DT	LC

The distribution of species in the functional space indicates that the best functional space is that using three dimensions (Fig. S1), according to the MAD index (0.044). As such, we found that changes in species positions along the first principal coordinates (PCoA1) capturing variation across all traits, suggesting an axis representing generalist strategies, PCoA2 is determined by trophic guild and home range, indicating a gradient from species with small home ranges and specialized diets (e.g., insectivores) to those with larger home ranges and more generalist feeding strategies, while PCoA3 is explained by trophic guild and activity patterns, separating species according to temporal activity (e.g., nocturnal vs. diurnal) and feeding behavior (Table S1). Mammal species located at the extremes of the functional space are those with extreme traits and different functional attributes. We found the giant anteater (M. tridactyla), puma (P. concolor), margay (L. wiedii), Santa Marta's corzuela (Passalites sanctaemartae), white-tailed deer (Odocoileus cariacou) and skunk (Conepatus semistriatus) (Fig. 2).

The model-averaged results revealed variability in the effects the explanatory variables according to the different species categories (Fig. 3, 4). For total richness, edge density had a significant positive effect (β = 0.09; SE = 0.04; 95% CI 0.004, 0.182) (Fig. 3a). Species richness of forest-dependent species showed a positive relationship with forest cover (β = 0.328; SE = 0.153; 95% CI 0.017, 0.640) (Fig. 3b, Table S4), and a negative relationship with the number of patches (β = -0.297; SE = 0.143; 95% CI -0.587, -0.008) (Fig. 3b). For non-forest-dependent species, no variable was significant but a moderate effect was found for edge density (Fig. 3c) and for functional richness no variable had a significant and moderate effect (Fig. 3d).

This study is the first evaluation of the effects of habitat fragmentation per se on the taxonomic and functional richness of mammals in the TDF of Colombia and specifically for the Colombian Caribbean region. We identified differences in the landscape structure that influenced the presence of mammals, which vary depending on the group evaluated. The response to the landscape structure is largely determined by the ability of the species to tolerate disturbances, reflecting the effect of the explanatory variables selected in each case.

The results show that the effects of fragmentation per se are generally weak and, in some cases, positive in relation to species richness (Fahrig 2017), in accordance to our hypothesis. In particular, we found that edge density was the only explanatory variable with a positive effect on total species richness (Fig. 3, 4). This finding is consistent with studies indicating that higher edge density in fragmented landscapes may offer a greater diversity of habitats, which may favor an increase in species richness (Pfeifer et al. 2017; Willmer et al. 2022). This effect is related to the concepts of landscape complementation and supplementation proposed by Dunning et al. (1992), who state that species benefit from access to different types of nearby habitats to satisfy complementary needs (e.g., foraging, reproduction and shelter) or supplement its resource acquisition. Thus, a mosaic of habitat patches composed, for example, of forest edges, open areas, and resource-rich microhabitats will allow species with different resource needs to coexist and occupy the landscape efficiently (Dunning et al. 1992). Edge density, therefore, not only increases habitat diversity but also favors complementary use of nearby areas, which increases the likelihood that functionally distinct species will exploit different habitat types, increasing richness in the landscape. Some species, especially certain generalist mammals (e.g., jaguarundi, Herpailurus yagouaroundi; raccon, P. cancrivorus), can take advantage of the new conditions to colonize degraded environments, phenomenon known as (Ries et al. 2004). For example, these species tend to be more abundant at habitat edges, transition areas between adjacent ecological systems, than in the forest interior (Sancha et al. 2023). However, this benefit to generalist species at habitat edges may come at a cost to specialist species, which are often more sensitive to edge effects and habitat degradation, potentially increasing their extinction risk (Ewers and Didham 2006).

Other studies consistently document positive effects of landscape configuration on the richness of medium- to large-bodied mammals (Regolin et al. 2020). Species responses to edge effects have been shown to vary, for example, some generalist species prefer the edge, as they can compensate for resource loss in fragmented landscapes by moving more widely (Willmer et al. 2022). This ability reduces their vulnerability to negative edge effects (Pfeifer et al. 2017). In addition, species can benefit in fragmented landscapes due to the closer proximity between different land cover types, which facilitates their movement between different land uses (Fahrig et al. 2011). In addition, the matrix can provide them with complementary and supplementary habitat resources, thereby increasing mammal species richness in modified landscapes (Brady et al. 2011; Fletcher et al. 2024). Thus, mammal movement decisions are also influenced in addition to natural vegetation, by the surrounding matrix and the degree of functional connectivity within the landscape (Russell et al. 2007; Berl et al. 2018).

As we propose in our hypothesis, forest cover was the most important variable in explaining the richness of forest-dependent species richness, reinforcing the importance of TDF landscapes with higher amounts of forest cover in the conservation of mammals sensitive to habitat transformation (Rios et al. 2021; Pardo et al. 2024). However, contrary to our expectation, we found that a greater number of patches was negatively associated with forest-dependent species richness, suggesting that greater habitat fragmentation compromises the persistence of these species in the Colombian Caribbean. This has been demonstrated by studies that have identified patch size and number of patches as critical predictors for the presence of medium- to large-bodied mammals (Magioli et al. 2015, 2021; Rios et al. 2022). For example, Rios et al. (2022) demonstrated that species such as paca (C. paca), collared peccary (Dicotyles tajacu), tayra (Eira barbara), puma (P. concolor), margay (L. wiedii), and tamandua (Tamandua tetradactyla) require large forest patches to maintain viable populations. Therefore, to effectively preserve this group of mammals, consideration should be given to preventing habitat loss and taking actions to reduce habitat fragmentation (Püttker et al. 2020; Pardo et al. 2024).

Some medium- to large-bodied mammal species are particularly sensitive to these changes changes (Crooks et al. 2011), especially interior forest species that are sensitive to forest transformation, which alters landscape patterns, modifying resources and environmental conditions (Pardini et al. 2010). Therefore, it is important to establish new protected natural areas, seeking to protect species that are sensitive to disturbances (Barlow et al. 2018; Rovero et al. 2020). In the Colombian Caribbean, although such areas are essential for maintaining mammal diversity, they exhibit low representativeness and connectivity (Departamento Nacional de Planeación 2021). Thus, it is necessary to establish new conservation areas or maintain existing ones to ensure the maintenance of mammal populations.This strategy aims to conserve areas with low levels of fragmentation (Zárrate-Charry et al. 2022).

Unlike forest interior species, non-forest-dependent species can survive in landscapes with little forest cover and take advantage of the resources provided by other types of land cover. This is likely due to the variability in adaptive capacity and behavioral plasticity to environmental modifications (Fisher et al. 2011). Examples include thecrab-eating fox (Cerdocyon thous), cottontail (Sylvilagus sp.) or the common opossum (D. marsupialis). These species can use landscapes transformed into agricultural systems and with a high degree of fragmentation and use the matrix as habitat (Beca et al. 2017; Pardo et al. 2018). Thus, matrix management should be prioritized by seeking habitat supplementation sensu (Driscoll et al. 2013), when focusing on less sensitive species to landscape transformation (Brady et al. 2011; Regolin et al. 2021).

For functional richness (FRic), no explanatory variable showed substantial or moderate effects, although forest cover showed the greatest relative importance and a positive trend. It is known that the increase in habitat loss can act as an environmental filter, due to the decrease in the quality and quantity of resources, which generates the loss of functions in assemblages, favoring the decrease of functional space in landscapes with less forest cover (Córdova-Tapia and Zambrano 2015). In other taxonomic groups, for example, in snakes and birds, FRic tends to increase in sites with greater forest cover, so the amount of habitat is considered the most significant variable (Rincón-Aranguri et al. 2023; Mariano-Neto and Santos 2023).

Habitat loss can decrease functional diversity in a number of ways, affecting both individual species and interactions among them within the ecosystem, such as control of insects (e.g., M. tridactyla; Lacher et al. 2019) or those that mobilize large volumes of plant biomass (e.g., P. sanctaemartae; Lacher et al. 2019), control of prey populations, seed dispersal and maintenance of the health and balance of their natural ecosystems (e.g., P. concolor; De Angelo et al. 2011; Jorge et al. 2013). This can destabilize species interactions and compromise ecosystem functioning and consequently its tolerance and resilience (Córdova-Tapia and Zambrano 2015). Current knowledge on the effects of fragmentation per se on functional diversity is still limited. Therefore, it is suggested to focus efforts on understanding the differential response of medium- to large-bodied mammals to landscape configuration changes, rather than only the taxonomic response, allowing to obtain information on ecosystem processes.

Our results support that active forest protection is fundamental for the conservation of medium- to large-bodied mammals and ecological processes in the TDF. Conservation strategies should not only focus on habitat quantity, but also take into account landscape configuration and patch size because they influence the presence of mammalian species and their effects can be both negative and positive depending on how sensitive species are to changes in the landscape. Therefore, not only the habitat cover should be used to support mammal conservation and management strategies as proposed by the “habitat amount hypothesis”.

Our results suggest caution in assuming that fragmentation effects per se are generally weak and positive (Fahrig 2003, 2017). Habitat fragmentation can pose a threat to mammal species that are sensitive to environmental disturbance, as each species has specific habitat requirements and different responses to landscape transformations. Therefore, studies assessing the effects of fragmentation per se should consider classifying species by their dependence on forest or resource use (e.g., guilds) rather than just assessing total species richness (Fahrig 2013). We emphasize the importance of protected areas and large forest patches which are crucial for mammal conservation. In addition, also the fundamental importance of small patches of forest submerged in an anthropogenic matrix, because they serve to maintain mammal populations.

We suggest that future research on functional diversity should not only focus on FRic, because it only takes into account extreme traits (Mason et al. 2005). Also assess the distribution of species in functional space such as functional evenness and functional divergence, seeking to have a better response to these landscape changes. Finally, we consider it important to be cautious when defining whether the effects of fragmentation are positive or negative, since not all species respond in the same way to changes in the landscape. Therefore, conducting more specific analyses at the species level allows us to obtain more precise and relevant conclusions.

Declaration of competing interest

The corresponding author confirms on behalf of all authors that there have been no involvements that might raise the question of bias in the work reported or in the conclusions, implications, or opinions stated.

CRediT authorship contribution statement

Sebastián Narváez-Barrios: Conceptualization; Methodology; Formal analysis; Writing - Original Draft. Lain E. Pardo: Methodology; Writing - Review & Editing; André Luis Regolin: Methodology; Writing - Review & Editing; Jairo Pérez-Torres: Writing - Review & Editing; Funding acquisition. Julio J. Chacón-Pacheco: Conceptualization; Methodology; Writing - Original Draft; Visualization; Supervision; Funding acquisition.

List tables and Fig.s

Funding

We acknowledge the financial supported by the project “Functional diversity and species richness of medium and large mammals in different scenarios of transformation of the tropical dry forest landscape in the Colombian Caribbean region (ID-PUJ 20702)” of the Pontificia Universidad Javeriana, Colombia.

Author Contribution

SN-B: Conceptualization; Methodology; Formal analysis; Writing - Original Draft. LEP: Methodology; Writing - Review & Editing; ALR: Methodology; Writing - Review & Editing; JP-T: Writing - Review & Editing; Funding acquisition. JJC-P: Conceptualization; Methodology; Writing - Original Draft; Visualization; Supervision; Funding acquisition.

Acknowledgement

SN-B, JP-T and JJC-P were supported by the project “Functional diversity and species richness of medium and large mammals in different scenarios of transformation of the tropical dry forest landscape in the Colombian Caribbean region (ID-PUJ 20702)” of the Pontificia Universidad Javeriana, Colombia. We thank Juliano André Bogoni for reviewing the manuscript and for his suggestions.

Aliaga-Rossel E, Kays RW, Fragoso JMV (2008) Home-range use by the Central American agouti (Dasyprocta punctata) on Barro Colorado Island, Panama. J Trop Ecol 24:367–374. https://doi.org/10.1017/s0266467408005129
Avendaño-Maldonado LJ, Camargo-Alarcón MÁ, Borja-Acuña R, Chacón-Pacheco JJ (2021) Mamíferos del departamento del Atlántico, Colombia. Biota Colombiana 22:108–126. https://doi.org/10.21068/C2021.V22N02A06
Barlow J, França F, Gardner TA et al (2018) The future of hyperdiverse tropical ecosystems. Nature 559:517–526. https://doi.org/10.1038/s41586-018-0301-1
Bartoń K (2010) MuMIn: Multi-Model Inference. 1.47.5
Beale CM, Lennon JJ, Yearsley JM et al (2010) Regression analysis of spatial data. Ecol Lett 13:246–264. https://doi.org/10.1111/j.1461-0248.2009.01422.x
Beca G, Vancine MH, Carvalho CS et al (2017) High mammal species turnover in forest patches immersed in biofuel plantations. Biol Conserv 210:352–359. https://doi.org/10.1016/j.biocon.2017.02.033
Berl JL, Kellner KF, Flaherty EA, Swihart RK (2018) Spatial Variation in Density of White-footed Mice Along Edges in Fragmented Habitat. Am Midl Nat 179:38–50. https://doi.org/10.1674/0003-0031-179.1.38
Bivand RS, Wong DWS (2018) Comparing implementations of global and local indicators of spatial association. TEST 27:716–748. https://doi.org/10.1007/s11749-018-0599-x
Borcard D, Gillet F, Legendre P (2011) Numerical Ecology with R. https://doi.org/10.1007/978-1-4419-7976-6
Brady MJ, McAlpine CA, Possingham HP et al (2011) Matrix is important for mammals in landscapes with small amounts of native forest habitat. Landscape Ecol 26:617–628. https://doi.org/10.1007/s10980-011-9602-6
Burnham KP, Anderson DR (eds) (2002) Statistical Theory and Numerical Results. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach. Springer, New York, NY, pp 352–436
Carvajal-Cogollo JE (2014) Evaluación a Múltiples Escalas De Los Efectos De La Transformación Del Paisaje Sobre Los Ensamblajes De Reptiles En Localidades De La Región Caribe Colombiana. 121
Castillo-Figueroa D, Pérez-Torres J (2021) On the development of a trait-based approach for studying neotropical bats. Papeis Avulsos de Zoologia 61:0–4. https://doi.org/10.11606/1807-0205/2021.61.24
Chacón J, Linares J, Carrascal J, Ballesteros J (2013) Área de acción del chigüiro (Hydrochoerus isthmius) en un sistema agropecuario en Córdoba, Colombia. Rev Colombiana Cienc Anim 5:270. https://doi.org/10.24188/recia.v5.n2.2013.291
Chacón Pacheco J, Figel J, Rojano C et al (2017) Actualización de la distribución e identificación de áreas prioritarias para la conservación de una especie olvidada: el hormiguero gigante en Colombia. Edentata: The Newsletter of the IUCN/SSC Anteater, Sloth and Armadillo Specialist Group. https://doi.org/10.2305/iucn.ch.2017.edentata-18-1.3.en
Chacón Pacheco JJ, Salcedo-Rivera GA, Zárrate-Charry DA (2022) Mamíferos del departamento de Sucre. Colombia Biota 23:e1022. https://doi.org/10.21068/2539200X.1022
Chacón Pacheco JJ, Sánchez-Londoño JD, Villada-Cadavid T, Ballesteros-Correa J (2021) Actualización de la lista de mamíferos silvestres del departamento de Córdoba, Colombia. Biota Colombiana 23:e966. https://doi.org/10.21068/2539200x.966
Chacón-Pacheco J, Ramos Madera C, Superina M (2021) State of knowledge and updated distribution of the northern naked-tailed armadillo Cabassous centralis Miller, 1899 (Cingulata, Chlamyphoridae). Mammalia 85:503–514. https://doi.org/10.1515/mammalia-2021-0054
Córdova-Tapia F, Zambrano L (2015) Functional diversity in community ecology. Ecosistemas 24. https://doi.org/10.7818/ecos.2015.24-3.10
Crooks KR (2002) Relative Sensitivities of Mammalian Carnivores to Habitat Fragmentation. Conserv Biol 16:488–502. https://doi.org/10.1046/j.1523-1739.2002.00386.x
Crooks KR, Burdett CL, Theobald DM et al (2011) Global patterns of fragmentation and connectivity of mammalian carnivore habitat. Philos Trans R Soc Lond B Biol Sci 366:2642–2651. https://doi.org/10.1098/rstb.2011.0120
De Angelo C, Paviolo A, Di Bitetti M (2011) Differential impact of landscape transformation on pumas (Puma concolor) and jaguars (Panthera onca) in the Upper Paraná Atlantic Forest. Divers Distrib 17:422–436. https://doi.org/10.1111/j.1472-4642.2011.00746.x
de Bello F, Carmona CP, Dias ATC et al (2021) Handbook of Trait-Based Ecology: From Theory to R Tools. Cambridge University Press
Departamento Nacional de Planeación (2021) CONPES 4050 Consejo Nacional de Política Económica y Social. República de Colombia. Departamento Nacional de Planeación, Bogotá
Díaz S, Malhi Y (2022) Biodiversity: Concepts, Patterns, Trends, and Perspectives. Annu Rev Environ Resour 47:31–63. https://doi.org/10.1146/annurev-environ-120120-054300
Dormann CF, Elith J, Bacher S et al (2012) Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36:27–46. https://doi.org/10.1111/j.1600-0587.2012.07348.x
Driscoll DA, Banks SC, Barton PS et al (2013) Conceptual domain of the matrix in fragmented landscapes. Trends Ecol Evol 28:605–613. https://doi.org/10.1016/J.TREE.2013.06.010
Dunning JB, Danielson BJ, Pulliam HR (1992) Ecological Processes That Affect Populations in Complex Landscapes. Oikos 65:169–175. https://doi.org/10.2307/3544901
Echeverría C, Newton AC, Lara A et al (2007) Impacts of forest fragmentation on species composition and forest structure in the temperate landscape of southern Chile. Glob Ecol Biogeogr. https://doi.org/10.1111/j.1466-8238.2007.00311.x
Etter A, Andrade A, Saavedra K et al (2017) Estado de los Ecosistemas Colombianos: una aplicación de la metodología de la Lista Roja de Ecosistemas (Vers2. 0). Informe Final Bogotá DC: Pontificia Universidad Javeriana y Conservación Internacional Colombia 108
Ewers RM, Didham RK (2006) Confounding factors in the detection of species responses to habitat fragmentation. Biol Rev Camb Philos Soc 81:117–142. https://doi.org/10.1017/S1464793105006949
Fahrig L (2017) Ecological Responses to Habitat Fragmentation per Se. Ann Rev Ecol Evol Syst 48:1–23. https://doi.org/10.1146/annurev-ecolsys-110316-022612
Fahrig L (2003) Effects of Habitat Fragmentation on Biodiversity. Annual Review of Ecology. Evol Syst 34:487–515. https://doi.org/10.1146/annurev.ecolsys.34.011802.132419
Fahrig L (2013) Rethinking patch size and isolation effects: the habitat amount hypothesis. J Biogeogr 40:1649–1663. https://doi.org/10.1111/JBI.12130
Fahrig L (2020) Why do several small patches hold more species than few large patches? Glob Ecol Biogeogr 29:615–628. https://doi.org/10.1111/GEB.13059
Fahrig L, Baudry J, Brotons L et al (2011) Functional landscape heterogeneity and animal biodiversity in agricultural landscapes. Ecol Lett 14:101–112. https://doi.org/10.1111/J.1461-0248.2010.01559.X
Ferreira AS, Peres CA, Bogoni JA, Cassano CR (2018) Use of agroecosystem matrix habitats by mammalian carnivores (Carnivora): a global-scale analysis. Mammal Rev 48:312–327. https://doi.org/10.1111/mam.12137
Fish US, Service, Wildlife (2016) Jaguar Draft Recovery Plan (Panthera onca). U.S. Fish and Wildlife Service. U. S. F. a. W. Service, Albuquerque, New Mexico
Fisher JT, Anholt B, Volpe JP (2011) Body mass explains characteristic scales of habitat selection in terrestrial mammals. Ecol Evol 1:517–528. https://doi.org/10.1002/ece3.45
Fletcher RJ, Didham RK, Banks-Leite C et al (2018) Is habitat fragmentation good for biodiversity? Biol Conserv 226:9–15. https://doi.org/10.1016/j.biocon.2018.07.022
Fletcher RJ, Smith TAH, Troy S et al (2024) The Prominent Role of the Matrix in Ecology, Evolution, and Conservation. Annu Rev Ecol Evol Syst 55:423–447. https://doi.org/10.1146/annurev-ecolsys-102722-025653
Fox J, Weisberg S, Adler D et al (2012) Package ‘car.’ Vienna: R Foundation for Statistical Computing 16:333
Alcázar G-MR, Avella C (2020) A El bosque seco en Colombia: estado de conocimiento y desafíos para su gestión integral. In: Moncada DM, Borda AC, Vieira-Muñoz MI, (eds) Elevando la acción colectiva empresarial para la gestión integral del bosque seco tropical en Colombia. Instituto Alexander von Humboldt, Bogotá, D. C., pp 12–25
García G-MR, Isaacs H P, et al (2018) Disentangling the environmental heterogeneity, floristic distinctiveness and current threats of tropical dry forests in Colombia. Environ Res Lett 13:045007. https://doi.org/10.1088/1748-9326/AAAD74
González-Maya JF, Zárrate-Charry DA, Cepeda AA et al (2013) Ecología y conservación de felinos y presas en el Caribe colombiano. Plan de Conservación de Felinos del Caribe colombiano: Los felinos y su papel en la planificación regional integral basada en especies clave 95–104
Gorczynski D, Beaudrot L (2021) Functional diversity and redundancy of tropical forest mammals over time. Biotropica 53:51–62. https://doi.org/10.1111/btp.12844
Hanski I (2015) Habitat fragmentation and species richness. J Biogeogr 42:989–993. https://doi.org/10.1111/jbi.12478
Herrero-Jáuregui C, Camba G, Andries DM et al (2022) Past and present effects of habitat amount and fragmentation per se on plant species richness, composition and traits in a deforestation hotspot. Biol Conserv 276:109815. https://doi.org/10.1016/J.BIOCON.2022.109815
Hesselbarth MHK, Sciaini M, With KA et al (2019) landscapemetrics: an open-source R tool to calculate landscape metrics. Ecography 42:1648–1657. https://doi.org/10.1111/ecog.04617
Huais PY (2018) multifit: an R function for multi-scale analysis in landscape ecology. Landscape Ecol 33:1023–1028. https://doi.org/10.1007/s10980-018-0657-5
Jones KE, Bielby J, Cardillo M et al (2009) PanTHERIA: a species-level database of life history, ecology, and geography of extant and recently extinct mammals. Ecology 90:2648–2648. https://doi.org/10.1890/08-1494.1
Jorge MLSP, Galetti M, Ribeiro MC, Ferraz KMPMB (2013) Mammal defaunation as surrogate of trophic cascades in a biodiversity hotspot. Biol Conserv 163:49–57. https://doi.org/10.1016/j.biocon.2013.04.018
Keinath DA, Doak DF, Hodges KE et al (2017) A global analysis of traits predicting species sensitivity to habitat fragmentation. Glob Ecol Biogeogr 26:115–127. https://doi.org/10.1111/geb.12509
Lacher TE, Davidson AD, Fleming TH et al (2019) The functional roles of mammals in ecosystems. J Mammal 100:942–964. https://doi.org/10.1093/jmammal/gyy183
Latorre JP, Jaramillo O, Corredor L (2014) Condición de las Unidades Ecobiogeograficas Continentales y Sistema Nacional de Áreas Protegidas en Colombia. (Base de Datos Geográfica Escala 1:100000
Magioli M, Ferraz KMPM, de Chiarello B AG, et al (2021) Land-use changes lead to functional loss of terrestrial mammals in a Neotropical rainforest. Perspect Ecol Conserv 19:161–170. https://doi.org/10.1016/j.pecon.2021.02.006
Magioli M, Moreira MZ, Fonseca RCB et al (2019) Human-modified landscapes alter mammal resource and habitat use and trophic structure. Proc Natl Acad Sci USA 116:18466–18472. https://doi.org/10.1073/pnas.1904384116
Magioli M, Ribeiro MC, Ferraz KMPMB, Rodrigues MG (2015) Thresholds in the relationship between functional diversity and patch size for mammals in the Brazilian Atlantic Forest. Anim Conserv 18:499–511. https://doi.org/10.1111/acv.12201
Magneville C, Loiseau N, Albouy C et al (2022) mFD: an R package to compute and illustrate the multiple facets of functional diversity. Ecography 2022:1–15. https://doi.org/10.1111/ecog.05904
Main-Knorn M, Pflug B, Louis J et al (2017) Sen2Cor for Sentinel-2. SPIE
Mariano-Neto E, Santos RAS (2023) Changes in the functional diversity of birds due to habitat loss in the Brazil Atlantic Forest. Front Forests Global Change 6. https://doi.org/10.3389/ffgc.2023.1041268
Martin AE, Fahrig L (2012) Measuring and selecting scales of effect for landscape predictors in species–habitat models. Ecol Appl 22:2277–2292. https://doi.org/10.1890/11-2224.1
Mason NWH, Mouillot D, Lee WG, Wilson JB (2005) Functional richness, functional evenness and functional divergence: The primary components of functional diversity. Oikos 111:112–118. https://doi.org/10.1111/j.0030-1299.2005.13886.x
McGarigal K, Cushman SA (2002) Comparative evaluation of experimental approaches to the study of habitat fragmentation effects. Ecol Appl 12:335–345. https://doi.org/10.1890/1051-0761(2002)012[0335:CEOEAT]2.0.CO;2
Mena JL (2010) Respuestas de los murciélagos a la fragmentación del bosque en Pozuzo, Perú. Revista Peruana de Biología 17:277–284
Meza-Joya FL, Ramos E, Cardona D (2020) Forest fragmentation erodes mammalian species richness and functional diversity in a human-dominated landscape in colombia. Mastozoologia Neotropical 27:338–348. https://doi.org/10.31687/saremMN.20.27.2.0.06
Miles L, Newton AC, DeFries RS et al (2006) A global overview of the conservation status of tropical dry forests. J Biogeogr 33:491–505
Mouillot D, Graham NAJ, Villéger S et al (2013) A functional approach reveals community responses to disturbances. Trends Ecol Evol 28:167–177. https://doi.org/10.1016/j.tree.2012.10.004
Mouillot D, Villéger S, Parravicini V et al (2014) Functional over-redundancy and high functional vulnerability in global fish faunas on tropical reefs. Proc Natl Acad Sci USA 111:13757–13762. https://doi.org/10.1073/pnas.1317625111
Negret PJ, Sonter L, Watson JEM et al (2019) Emerging evidence that armed conflict and coca cultivation influence deforestation patterns. Biol Conserv 239:108176. https://doi.org/10.1016/j.biocon.2019.07.021
Newbold T, Bentley LF, Hill SLL et al (2020) Global effects of land use on biodiversity differ among functional groups. Funct Ecol 34:684–693. https://doi.org/10.1111/1365-2435.13500
Ofstad EG, Herfindal I, Solberg EJ, Sæther BE (2016) Home ranges, habitat and body mass: simple correlates of home range size in ungulates. Proceedings of the Royal Society B: Biological Sciences 283:. https://doi.org/10.1098/RSPB.2016.1234
Pardini R, de Bueno AA, Gardner TA et al (2010) Beyond the fragmentation threshold hypothesis: Regime shifts in biodiversity across fragmented landscapes. PLoS ONE. https://doi.org/10.1371/journal.pone.0013666. 5:
Pardo LE, Campbell MJ, Cove MV et al (2019) Land management strategies can increase oil palm plantation use by some terrestrial mammals in Colombia. Sci Rep 9:1–12. https://doi.org/10.1038/s41598-019-44288-y
Pardo LE, Campbell MJ, Edwards W et al (2018) Terrestrial mammal responses to oil palm dominated landscapes in Colombia. PLoS ONE 13:1–22. https://doi.org/10.1371/journal.pone.0197539
Pardo LE, Edwards W, Campbell MJ et al (2021) Effects of oil palm and human presence on activity patterns of terrestrial mammals in the Colombian Llanos. Mammalian Biology 101:775–789. https://doi.org/10.1007/s42991-021-00153-y
Pardo LE, Gómez-Valencia B, Deere NJ et al (2024) Forest cover is more important than its integrity or landscape configuration in determining habitat use by mammals in a human-modified landscape in Colombia. Global Ecol Conserv 55:e03232. https://doi.org/10.1016/j.gecco.2024.e03232
Patton JL, Pardiñas UFJ, D’Elía G (2015) Mammals of South America, vol 2. University of Chicago Press
Pfeifer M, Lefebvre V, Peres CA et al (2017) Creation of forest edges has a global impact on forest vertebrates. Nat 2017 551:7679. https://doi.org/10.1038/nature24457
Pineda-Cendales S, Hernández-Rolong E, Carvajal-Cogollo JE (2020) Medium and large-sized mammals in dry forests of the Colombian Caribbean. Universitas Scientiarum 25:435–461. https://doi.org/10.11144/JAVERIANA.SC25-3.MALS
Pineda-Munoz S, Evans AR, Alroy J (2016) The relationship between diet and body mass in terrestrial mammals. Paleobiology 42:659–669. https://doi.org/10.1017/pab.2016.6
Pizano C, García H (2014) El bosque seco tropical en Colombia. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt (IAvH), Bogotá, D.C
Pizano C, González-M R, López R et al (2016) El bosque seco tropical en Colombia. In: Gómez MF, Moreno LA, Andrade GI, Rueda C (eds) Biodiversidad 2015. Estado y Tendencias de la Biodiversidad Continental de Colombia. Instituto Alexander von Humboldt, Bogotá, D.C.
Portillo-Quintero CA, Sánchez-Azofeifa GA (2010) Extent and conservation of tropical dry forests in the Americas. Biol Conserv 143:144–155. https://doi.org/10.1016/j.biocon.2009.09.020
Püttker T, Crouzeilles R, Almeida-Gomes M et al (2020) Indirect effects of habitat loss via habitat fragmentation: A cross-taxa analysis of forest-dependent species. Biol Conserv 241:108368. https://doi.org/10.1016/j.biocon.2019.108368
Regolin AL, Cherem JJ, Graipel ME et al (2017) Forest cover influences occurrence of mammalian carnivores within Brazilian Atlantic Forest. J Mammal 98:1721–1731. https://doi.org/10.1093/jmammal/gyx103
Regolin AL, Oliveira-Santos LG, Ribeiro MC, Bailey LL (2021) Habitat quality, not habitat amount, drives mammalian habitat use in the Brazilian Pantanal. Landscape Ecol 36:2519–2533. https://doi.org/10.1007/s10980-021-01280-0
Regolin AL, Ribeiro MC, Martello F et al (2020) Spatial heterogeneity and habitat configuration overcome habitat composition influences on alpha and beta mammal diversity. Biotropica 52:969–980. https://doi.org/10.1111/btp.12800
Resende PS, Viana-Junior AB, Bovendorp RS (2024) A meta‐analysis of the effects of anthropogenic disturbances on tropical mammal functional diversity. Mammal Rev. https://doi.org/10.1111/mam.12356
Ries L, Fletcher RJ, Battin J, Sisk TD (2004) Ecological Responses to Habitat Edges: Mechanisms, Models, and Variability Explained. Annu Rev Ecol Evol Syst 35:491–522. https://doi.org/10.1146/annurev.ecolsys.35.112202.130148
Rincón-Aranguri M, Toro-Cardona FA, Galeano SP et al (2023) Functional diversity of snakes is explained by the landscape composition at multiple areas of influence. Ecol Evol 13. https://doi.org/10.1002/ece3.10352
Rios E, Benchimol M, De Vleeschouwer K, Cazetta E (2022) Spatial predictors and species’ traits: evaluating what really matters for medium-sized and large mammals in the Atlantic Forest, Brazil. Mammal Rev 52:236–251. https://doi.org/10.1111/mam.12276
Rios E, Benchimol M, Dodonov P et al (2021) Testing the habitat amount hypothesis and fragmentation effects for medium- and large-sized mammals in a biodiversity hotspot. Landscape Ecol 36:1311–1323. https://doi.org/10.1007/s10980-021-01231-9
Rivas CA, Guerrero-Casado J, Navarro-Cerillo RM (2021) Deforestation and fragmentation trends of seasonal dry tropical forest in Ecuador: impact on conservation. For Ecosyst 8. https://doi.org/10.1186/s40663-021-00329-5
Rojano C, Chacón-Pacheco J, Polo AF (2016) El oso melero (Tamandua mexicana) en el Caribe colombiano: aportes sobre su ecología y amenazas. Edentata: The newsletter of the IUCN/SSC Anteater. Sloth Armadillo Specialist Group 17:17–24. https://doi.org/10.2305/iucn.ch.2016.edentata-17-1.4.en
Rovero F, Ahumada J, Jansen PA et al (2020) A standardized assessment of forest mammal communities reveals consistent functional composition and vulnerability across the tropics. Ecography 43:75–84. https://doi.org/10.1111/ecog.04773
Russell RE, Swihart RK, Craig BA (2007) The Effects of Matrix Structure on Movement Decisions of Meadow Voles (Microtus pennsylvanicus). J Mammal 88:573–579. https://doi.org/10.1644/06-mamm-a-080r1.1
Sancha NU, de la, González-Maya JF, Boyle SA et al (2023) Bioindicators of edge effects within Atlantic Forest remnants: Conservation implications in a threatened biodiversity hotspot. Divers Distrib. https://doi.org/10.1111/DDI.13663
Saura S (2021) The habitat amount hypothesis predicts that fragmentation poses a threat to biodiversity: A reply to Fahrig. J Biogeogr 48:1536–1540. https://doi.org/10.1111/jbi.14122
Saura S (2020) The Habitat Amount Hypothesis implies negative effects of habitat fragmentation on species richness. J Biogeogr 48:11–22. https://doi.org/10.1111/jbi.13958
Smith AC, Koper N, Francis CM, Fahrig L (2009) Confronting collinearity: comparing methods for disentangling the effects of habitat loss and fragmentation. Landscape Ecol 24:1271–1285. https://doi.org/10.1007/s10980-009-9383-3
Suárez-Castro AF, Ramírez-Chaves HE (2015) Los carnívoros terrestres y semiacuáticos continentales de Colombia. Guía de Campo. Universidad Nacional de Colombia, Bogotá, D. C
Thompson JJ, Velilla M, Cabral H et al (2022) Jaguar (Panthera onca) population density and landscape connectivity in a deforestation hotspot: The Paraguayan Dry Chaco as a case study. Perspect Ecol Conserv 20:377–385. https://doi.org/10.1016/j.pecon.2022.09.001
Ulloa L, Rodríguez D, Sánchez P (1999) Movimientos y uso del tiempo y el espacio por una guartinaja (Agouti paca) en la Sierra Nevada De Santa Marta, Colombia. Revista de Acad Colombiana de Ciencias 23:687–694
Valente JJ, Gannon DG, Hightower J et al (2023) Toward conciliation in the habitat fragmentation and biodiversity debate. Landscape Ecol 38:2717–2730. https://doi.org/10.1007/s10980-023-01708-9
Wang X, Blanchet FG, Koper N (2014) Measuring habitat fragmentation: An evaluation of landscape pattern metrics. Methods Ecol Evol 5:634–646. https://doi.org/10.1111/2041-210x.12198
Watling JI, Arroyo-Rodríguez V, Pfeifer M et al (2020) Support for the habitat amount hypothesis from a global synthesis of species density studies. Ecol Lett 23:674–681. https://doi.org/10.1111/ELE.13471
Willmer JNG, Püttker T, Prevedello JA (2022) Global impacts of edge effects on species richness. Biol Conserv 272:109654. https://doi.org/10.1016/j.biocon.2022.109654
Wilman H, J. B JS et al (2014) EltonTraits 1.0: Species-level foraging attributes of the world’s birds and mammals. Ecology 95:2027
Wolf A, Doughty CE, Malhi Y (2013) Lateral Diffusion of Nutrients by Mammalian Herbivores in Terrestrial Ecosystems. PLoS ONE 8:e71352. https://doi.org/10.1371/journal.pone.0071352
Zárrate-Charry DA, González-Maya JF, Arias-Alzate A et al (2022) Connectivity conservation at the crossroads: Protected areas versus payments for ecosystem services in conserving connectivity for Colombian carnivores. Royal Soc Open Sci 9. https://doi.org/10.1098/rsos.201154
Zimbres B, Peres CA, Machado RB (2017) Terrestrial mammal responses to habitat structure and quality of remnant riparian forests in an Amazonian cattle-ranching landscape. Biol Conserv 206:283–292. https://doi.org/10.1016/j.biocon.2016.11.033
Zuur AF, Ieno EN, Walker N et al (2009) Mixed effects models and extensions in ecology with R. Statistics for Biology and Health. https://doi.org/10.1007/978-0-387-87458-6

No competing interests reported.

Supplementarymaterial.docx

Download PDF

Editorial decision: Revision requested
15 Dec, 2025
Reviews received at journal
11 Dec, 2025
Reviews received at journal
08 Dec, 2025
Reviews received at journal
02 Dec, 2025
Reviewers agreed at journal
24 Nov, 2025
Reviewers agreed at journal
24 Nov, 2025
Reviewers agreed at journal
21 Nov, 2025
Reviewers agreed at journal
19 Nov, 2025
Reviewers agreed at journal
19 Nov, 2025
Reviewers agreed at journal
19 Nov, 2025
Reviewers agreed at journal
19 Nov, 2025
Reviewers agreed at journal
06 Nov, 2025
Reviewers invited by journal
31 Oct, 2025
Editor assigned by journal
24 Sep, 2025
Submission checks completed at journal
05 Aug, 2025
First submitted to journal
04 Aug, 2025

You are reading this latest preprint version

Effects of habitat amount and fragmentation per se on mammals in a highly fragmented Colombian region

Status:

Version 1

Abstract

Figures

1. INTRODUCTION

2. Materials and methods

2.1. Study area

2.2. Data collection

2.3. Multiscale landscape structure analysis

2.4. Taxonomic and functional richness

2.5. Data analysis

3. Results

4. Discussion

5. Conclusions

Declarations

Declaration of competing interest

CRediT authorship contribution statement

Funding

Author Contribution

Acknowledgement

References

Additional Declarations

Supplementary Files

Status:

Version 1