https://orcid.org/0000-0003-0966-0139
Figures
Abstract
Individual movements of bats are triggered by their life requirements, limited by their recognition of the environment and risks of moving, and mediated by habitat selection. Mining adds fragmentation and heterogeneity to landscapes, with poorly understood consequences to the life activities of the bats. Cave dwelling bats spend most of their life cycles within caves, and as they constantly forage in external landscapes, their contribution in the input of organic matter to the caves is of paramount importance to the subterranean biodiversity. We investigated habitat selection by cave bats in a mosaic of Amazonian forests, iron-rich savannas and iron-caves, studying the movements of the aerial insect-catcher Furipterus horrens (Furipteridae) and of the foliage gleaning bat Lonchorhina aurita (Phyllostomidae), both cave-dependent in the area. We used radio telemetry to assess their use of habitats, under the prediction that these species prefer preserved habitats for their foraging bouts, avoiding human-modified habitats. We also anticipated larger areas and commuting distances for both species when foraging in disturbed landscapes. Thirty-one bats were monitored in conserved habitats, mining sites and pastures resulting in an average range of 415.1 ± 837.4 ha and average commuting distances of 2163 ± 2453 m (F. horrens) and 681.1 ± 1101 ha and 2781.5 ± 2333 m (L. aurita). Our results suggest that F. horrens are open space foragers, frequently recorded in cangas, and L. aurita are forest foragers that use all habitats proportionally to their availability in the landscape. We detected landscape-related effects mediated by seasonal variation in the maximum commuting distance performed by both species. These are the first radiotelemetry data available for F. horrens and L. aurita, delivering original baselines to understand their movement behaviors. This information sheds light into constraints related to the optimal and adjusted biological cycles of these two species and their range shifts under scenarios of disturbance and may subsidize programs for their conservation.
Citation: Tavares VdC, Ribeiro MS, Prous X, Notini AA, Kaku-Oliveira NY, Maciel LMD, et al. (2025) Do you have enough space? Habitat selection of insectivorous cave-dwelling bats in fragmented landscapes of Eastern Amazon. PLoS ONE 20(1): e0296137. https://doi.org/10.1371/journal.pone.0296137
Editor: Janice L. Bossart, Southeastern Louisiana University, UNITED STATES OF AMERICA
Received: December 5, 2023; Accepted: October 28, 2024; Published: January 9, 2025
Copyright: © 2025 Tavares 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 and R scripts are available on GitHub (https://github.com/leonardotrevelin/Do_you_have_enough_space).
Funding: The authors received no specific funding for this work. The fieldwork was funded by Vale.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The behavior of organisms in the wild is expressed as a compromise among sets of possible responses, enabling them to cope with adverse situations, such as those arising from anthropogenic activities [1]. Movement behaviors result from the conciliation between the need of exploration of the environment to meet life requirements with preventive actions related to potential threats associated to moving [2, 3]. Those trade-offs are mediated by habitat selection that enhance the chances of individuals and species to optimize their performance and fitness regarding their use of space [4]. Understanding habitat selection can shed light on behavioral mechanisms that allow species to exploit heterogeneous environments [5] and provide insights into the influences of habitat quality on individuals and populations [1].
Cave-dwelling bats spend roughly half their life cycles roosting underground for resting, mating, and raising offspring [6, 7], and the other half searching for and gathering resources during night foraging bouts in the external landscapes [8]. The energetic requirements of bats for their daily activities and the heterogeneity of the landscapes available for their foraging movements are among the main limiting factors shaping these nightly outside hunts driven by habitat selection [9].
Although caves have usually been the focus of conservation efforts for the cave-related biota [10], the surrounding landscapes can encompass most of the home range of a bat. These interplays between the underground and the surrounding landscapes shapes overall biodiversity patterns in subterranean environments worldwide [7] and are pivotal to maintain the diversity and abundance of cave bats over time [11]. The same holds true for subterranean invertebrate assemblages, which are largely sustained by the guano deposition made by bats [8, 12].
The conservation of subterranean ecosystems is a contemporaneous challenge. Given their unique biodiversity, often concentrated in areas rich in mineral resources [13], cave-associated ecosystems are frequently vulnerable to intense and multi-scale landscape modification. The consequences to the underground biodiversity are poorly understood, and this demands concentrated efforts including the unveiling of biological patterns of these systems and to provide indicators that convey agile and reliable insights to monitoring programs focused on caves and their associated landscapes [14–16]. Understanding habitat selection patterns of cave-dwelling bats can provide a baseline to bat monitoring and to obtain timely evidence of changes in the bat movement behavior that may be associated to environmental changes [1] and subsidize conservation and management plans to cave bats and to the associated cave biodiversity [8].
Understanding how bat movements are affected by habitat loss and fragmentation [e.g.17, 18] remains challenging. Data on the use of space, habitat preferences and selection by bats have seldom been assessed in the Neotropics e.g. [19–23]. Information is even rarer for the bats with Amazonian distribution and virtually non-existent for bats in tropical cave systems. The scarce evidence available suggests that seasonality modulates Amazonian bats’ responses to landscape structure, signaling the role of spatiotemporal variation of food resources as a template for the foraging and movement behavior of bats [24, 25]. These patterns are complex, and they are likely species and ensemble-specific (e.g. foraging guilds) responses [26], and better understanding them is critical for karstic and pseudokarstic landscapes demanding management actions for habitat protection, such as the unique iron cave complexes from Eastern Amazonia.
One of the world´s most extensive iron cave systems occurs in the Eastern Amazonia region of Carajás. The heterogeneous landscapes of Carajás concentrate over 1500 iron caves in a mosaic of well-preserved Amazonian forests and iron-rich cangas with areas disturbed by iron-mining and peri-rural activities, all primarily enclosed within the Carajás National Forest, and harboring a rich bat fauna [27]. Some of these species are regionally cave-dependent, such as the phylogenetically distant and morpho-ecologically highly distinctive insectivorous species, the thumbless bat, Furipterus horrens Bonaparte, 1837 (Chiroptera: Furipteridae), and the common sword-nosed bat, Lonchorrina aurita (Chiroptera: Phylostomidae).
Thumbless bats (F. horrens) are tiny (3–4 g) and have a delicate appearance, broad wings, a long tail extending for most of their large uropatagium, and small thumbs enclosed by their wing membranes. Populations of F. horrens are generally associated with humid forested areas and limestone rocky shelters, forming groups that range from small to large colonies. These bats are distributed from Venezuela to the transition between the semideciduous forests and the Caatinga of Southern and Northeastern Brazil [28, 29]. They have been found inhabiting rock outcrops in the Xingu riverbed [30] and forming small groups, often including juveniles, in fallen trees in forests of French Guyana [31]. In Carajás, F. horrens relies heavily on the iron caves for roosting, forming large aggregations [30]. The biology of F. horrens, considered one of the rarest Neotropical bat species, is scarcely known [29]. Most of its known South American range and records are within Brazil, where it has been listed as “Vulnerable” based on rough estimates of the percentage of mature individuals per population and based on its association with karstic landscapes under the influence of mining activity, cattle ranching, habitat fragmentation, and urban expansion [32]. Lonchorhina aurita Tomes 1863 (Phyllostomidae) is a medium-sized foliage-gleaning insectivorous bat (12–16 g), categorized as “Near threatened” in Brazil [32], that possesses a singular echolocation system, which is considered an adaptation to aerial hawking [33]. This species has been found roosting in tree hollows [34], but like F. horrens, it extensively occupies iron caves in Carajás [27]. For both F. horrens and L. aurita, virtually nothing is known about their use of space and habitat selection.
Our two-fold goals in this study are to (1) describe movement data of the bat species F. horrens and L. aurita, generating baseline information on their use of space and habitat selection patterns, and (2) estimate responses of these two bat species to anthropogenic disturbance at landscape levels. We hypothesized that these two species of insect-feeding cave bats would rely mainly on well-preserved forest and canga habitats for their foraging activities and used radiotelemetry to assess their habitat selection modes. We predicted the avoidance of anthropogenic, modified habitats by F. horrens and L. aurita, such as mining and pastures while foraging [10], and that the range areas and commuting distances would be larger in disturbed landscapes for both species [20].
Methods
Study site
We studied the movements of F. horrens and L. aurita in the National Forest of Carajás (FLONA Carajás), and in the Campos Ferruginosos National Park, two contiguous Brazilian protected areas located in the Carajás mountain range of southeastern Amazonia, state of Pará, Brazil. Carajás encompasses complex subterranean ecosystems with over 1500 iron caves already prospected [35] and, given to its vast and unique iron-rich terrains, it has also been considered one of the world’s largest deposits of high-grade iron ore [36]. The region is composed of medium altitude plateaus (between 500 and 700 m) scattered along lowlands and forming a mosaic of typical Amazonian rainforests and rocky outcrops with savannah-like vegetation. These savannah-like ecosystems are however uniquely adapted to iron-rich soils, dissimilar to other Amazonian savannas, and called “cangas” [35, 37, 38]. The Carajás region currently combines pristine and recovering protected areas and areas for ore exploitation (iron, copper, manganese) and other unprotected areas under anthropogenic use such as pastures, abandoned crops and peri-urban zones [39]–Fig 1). The regional climate is defined as Aw following Köppen classification, marked by high annual rainfall with a well-defined dry season [40].
Contains Copernicus Sentinel data [2021] processed by Sentinel Hub.com.
Experimental design and data collection
Landscape data.
The composition of the landscapes surrounding each of the study areas was initially accessed within a circular 10 km-radius buffer centered around the main cave entrance, using photo-interpreted data from Sentinel 2 images with 10 meters of spatial resolution. The final buffer size was based on the farthest bat location detected by our equipment and according to our sampling protocol, and it was an empirical estimate of the maximum landscape area that we effectively sampled [20]. We discriminated six landcover classes as landscape habitat categories and quantified their availability within each buffer: four natural habitats (forest, canga, water bodies and exposed bare land) and two anthropogenic land uses (mining and pastures). Our experimental design was then composed of three categories of landscapes (treatments) based on the presence and proximity of the cave sampled to anthropogenic habitats: landscapes exclusively dominated by preserved natural habitats (forests and savannah-like “canga”, less the ~5% of anthropogenic habitats), landscapes containing mining land use (> 5% mine habitat close to the cave) and landscapes dominated by pastures (> 20% of pastures habitat) (Fig 1). All imagery analyses were conducted using QGIS 3.14. We based the final selection of caves for our study combining the presence of our target species (F. horrens and L. aurita) and the representativeness of the surrounding landscapes within our experimental design.
Capturing and marking bats.
Bats were captured during the daytime while roosting inside caves using entomological hand nets [41], as we wanted to make sure that we would be monitoring the movements of cave residents. The individuals monitored for each species were treated as replicates of their populations, and each population was sampled twice to encompass the wet and the dry seasons. We tagged selected individuals of F. horrens and L. aurita with radio transmitters with unique signaling frequencies (LB-2x –Holohil Inc., Canada). For every individual captured we first took measurements and weight, and additional information such as sex and life stage (juvenile or adult). For the females we also annotated reproductive condition (pregnant and/or lactant) [42]. We carefully trimmed patches of upper middorsal fur, close to the scapula horizontal plane (in relation to a longitudinal body axis) to attach the radio transmitters with the help of surgical adhesive (Fig 2). The weight of each transmitter was 0.31 to 0.41 g, corresponding roughly to 3% of L. aurita’s and 8% of F. horrens total body weights. The latter value is above an ideal threshold of 5% [43] but remains below the 10% threshold considered adequate for short-term studies and for the small size of the species [44, 45]. We released radio-tagged bats inside the caves in which they were originally captured and began to record radiotelemetry data departing from the next night to avoid bias resulting from the capture and handling of individuals. Transmitters fixed with surgical adhesive typically remain attached for fewer than 10 days [44], close to the minimal battery duration expected for the models we used. Thus, in our study, we retrieved some transmitters after detachment using positioning techniques (see below). However, our experimental design precluded us from retrieving most transmitters due to battery failure.
Attachment of radio transmitters, after the clipping of the fur, and release of the bats: A) Hair clipping in the back of a L. aurita individual; B) gluing of the transmitter to the back of a L. aurita individual; C) Transmitter glued on and D) hand release. E) Hair clipping in the back of a F. horrens individual; B) gluing of the transmitter to the back of a F. horrens individual; C) Transmitter glued on, and D) hand release. All images were taken by the authors.
We selected only healthy adult and non-reproductive females to receive radio transmitters, as indicated by our physical examination of the bats. We complied with the guidelines from the American Society of Mammalogists [46] for the use of wildlife in research, and this study was authorized by the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA) (ABIO N° 1000/2018).
Radiotracking.
We conducted six sampling sessions, three in the dry season (August and September 2018, July 2019) and three in the wet season (November 2018, February and November 2019). We set a sampling protocol for each marked individual, in which we divided every sampling night into three to four hours intervals, starting at 6 PM and ending at 6 AM of the next day. Our aims were to sample all intervals for all individuals in a way to allow for inferences about habitat use in situations where the number of location fixes was different among individuals (replicates) [20, 47]. Independent teams conducted signal monitoring by performing active radio signal searches and bi-angulation techniques to obtain locations [48] using a radio-receiver model TRX-1000 (Wildlife Materials Inc.) coupled to three-element Yagi antennas. We mapped successive locations and roost fixes to obtain flight routes and range areas. Based on our protocol and raw data exploration, we considered one individual as successfully monitored only when it was detected and monitored in all night intervals and sampled for more than one day and nine locations. We initially marked and tried to monitor a total of 46 individuals, and only successfully monitored 30 due to loss of signal. A total of 16 F. horrens and 14 L. aurita were successfully monitored, with similar numbers of sampled individuals between seasons (see Table 1).
Species/Individuals, Sex, Forearm (length in mm), Cave, Treatment, Date, Season, Days (sampled) and N (Number of locations estimated). MCP (ha)–Minimal Convex Polygon estimate. KF 95% (ha)–Fixed Kernel estimates of range use with 95% (ha) of samples; KF 50% (ha)–Fixed Kernel estimates of range use with 50% (ha) of samples; MC 95% (m)–Maximum linear distance travelled from cave entrance in KF 95%; MC 50% (m)–Maximum linear distance travelled from the cave entrance to core foraging areas in KF 50%. At the bottom of each species, we present the average and standard deviation of estimates for each parameter related to the use of space.
Data analysis
Range area estimates.
We used Best Bi-angulation Estimators to determine locations, after correcting for magnetic declination in all compass bearings, using LOAS™ (Ecological Software Solutions, Inc.). Location datasets determined for each individual were input into ESRI ArcGis Desktop 10.2 software and range areas were subsequently estimated using the HRT 2 extension, Home Range Tools [49].
We generated estimates of range areas for each individual following Kernohan et al. [50], as the ‘‘extent of area with a defined probability of occurrence of an animal during a specified period”. We used fixed normal kernel density estimations (KF) with 95% (KF 95%) and 50% (KF 50%) kernel isopleths to delineate the area ranges [51] implementing least-squares cross-validation (LSCV) to estimate the smoothing parameter (H) [52]. We also generated range area estimates using Minimum Convex Polygon (MCPs) estimates, uniting the most external locations. These range estimates were used in our hypotheses testing.
Habitat use/selection.
To evaluate habitat selection patterns, we estimated individual-based habitat uses by contrasting the proportional use of habitats by each individual with the estimates of habitat availability in the landscape. Since “habitat” is scale-dependent we organized our data following Johnson’s [53] hierarchical orders of habitat selection. In our framework, the second-order selection represents the proportional use of a determined habitat considering 95% kernel density estimations (KF 95%), and the third-order selection is the proportional of a determined habitat considering KF 50% kernel density estimations (KF), and both are nested within the MCPs estimated for each separated individual.
We used the non-parametric analytics as proposed by Fattorini et al. [54] based on the combination of sign tests, in which every habitat is individually tested. Following these analyses, overall statistic values for the simultaneous assessment of habitat selection in all habitat types can be obtained by combining P values from each test through permutation.
We considered all six habitat types: Forest, canga, mining land, pasture, exposed (bare) land and water bodies, and assessed whether each habitat was proportionally used according to its availability or whether it was over or under-utilized by each bat individually. All analytical procedures were conducted within the R environment [55], using package "phuassess" for R 3.3.1 [56].
Modelling use of space.
To evaluate the effects of landscape composition in the movement behavior of the bats, we modelled their use of space using four parameters: range area (KF 95% area in hectares), core-foraging area (KF 50% area in hectares) and two commuting distances, the Maximum linear distance (m) travelled from cave entrance in KF 95%, and Maximum linear distance (m) travelled from the cave entrance to core foraging areas in KF 50%. These four parameters were initially explored for normality using the Shapiro-Wilk test, and the effects of sex, and body (forearm length in mm) and sample sizes on them were accessed by the Wilcoxon rank-sum test and Spearman correlations, respectively.
We implemented linear mixed effect models (LMMs) in a maximum likelihood framework [57] using these parameters as response variables and landscape treatments (conservation/disturbance), seasonality, and their interactions as fixed effects. Caves and species were modelled as nested random effects, and data for both species, F. horrens and L. aurita, in all caves were polled for these analyses. We used multi-model inferences to evaluate competing candidate models representing all possible combinations of these variables. Our model selection was based on the Akaike Information Criteria (AIC) considering all top-ranking models with ΔAIC < 2 as having similar support [58]. We adopted a stepwise modelling approach, firstly, selecting between random effects and using Restricted Maximum likelihood, then selecting between fixed effects using standard Maximum likelihood, and finally assessing estimates of the final models [57, 59].
All analyzes were conducted within the R environment [55], using “glmmTMB”, “ggplot2” and “MuMIn” packages [60–62]. Residual diagnostics of fitted models were accessed using the DHARMa package [63] and all models evaluated achieved convergence.
Results
We successfully monitored sixteen individuals of F. horrens (length of forearm 34.9± 0.9 mm) and 14 of L. aurita (length of forearm: 48.8± 1.1 mm). Sampling lasted from three to four days on average for each individual of both species (F. horrens: 3.6 ± 1.86; L. aurita: 4.1 ± 1.17). We sampled a total of 1061 locations, and the mean number of locations obtained per individual was 42.7 ± 24.9 for F. horrens and 26.9 ± 13.6 for L. aurita. According to our initial exploratory analyses the parameters of space use were unrelated to the number of locations obtained and to the body size of bats for both species (Table 1 in S1 File). We also did not find relationships between the estimated parameters of space use and sex for both F. horrens or L. aurita (Table 1 in S1 File). Our exploratory results thus allowed us to pool individuals of both sexes of each of the two species for subsequent analyses, and results obtained for each individual sampled are summarized in Table 1.
Habitat use/selection
Our habitat selection estimates for F. horrens indicate frequent foraging activities for this species in open spaces. When contrasting the KF 95% range area with the habitat availability (second-order selection), we observed that individuals of F. horrens preferentially used non-forested habitats, and mainly the canga and pastures, compared to their proportional use of any other type of habitat. There was also a pattern of avoidance of forest and water covered habitats (p = 0.0004, Fig 3A). The use of open areas appears only subtly when contrasting the foraging core area range KF50% and the MCP habitat availability (third-order selection) (p>0.05, Fig 3B).
Landscape habitat composition (exposed bare land, forest, mining, pastures, rupestrian savannah, and water bodies) and habitat availability considering the estimate of KF 95% foraging area (a) and KF 50% foraging area (b) for Furipterus horrens, and KF 95% foraging area and KF 50% foraging area by Lonchorhina aurita (c) and (d).
On the other hand, the results for L. aurita individuals show that this species used all habitats proportionally to their availability in the landscape, except for their avoidance of water covered habitats (Second order selection, p = 0.005, Fig 3C). Furthermore, when contrasting the core foraging areas (KF50% defined) and the MCP habitat availability, we observed that L. aurita used all habitats proportionally to their availability in the landscapes (Third-order selection, p>0.05, Fig 3D). Forested habitats were the most used by this species as the rupestrian savannahs (cangas) and pastures appeared only as a complementary habitat use (Fig 3D).
Use of the space
The nested variable “Cave:Species” was selected as a random factor in the fixed slope model schemes for the adjusted models of all four response variables. Similarly, “Season” featured as a fixed effect between the most parsimonious models in all selection schemes (ΔAIC < 2), and the same happened with the Null model, hampering further interpretations. The exception was the Maximum commuting distance in KF95%, in which case the most parsimonious models included “Season” and the interaction between “Season” and “Landscape” (Table 2 in S1 File). Data on the use of disturbed landscapes (mining and pastures) indicate the accommodation of larger foraging areas and longer commuting distances compared to data from preserved landscapes in the wet season, and the inverse pattern is observable in the dry season (Fig 4). For the range area estimate models, there was no fixed effects narrowing the estimated foraging areas, either when KF 95% was considered or for KF 50% (Table 2 in S1 File).
Effects of landscape composition and seasonality for the combined estimates of the foraging areas of F. horrens and L. aurita considering fixed kernel estimates, KF 95% (a) and the KF 50% (b), and the Maximum commuting distances in KF95% (c) and in KF50% (d).
On the other hand, the commuting distance models considering a range of KF 95% were affected by seasonality (βdry = -0.9±0.32; p = 0.003), by “Landscape” (βPreserved = -1.04±0.50; p = 0.04) and by the interaction between both (βpreserved:dry = 1.45±0.71; p = 0.04) highlighting different commuting distances in preserved landscapes, and between seasons (Table 3 in S1 File). Additionally, no fixed effects were found for commuting distance when considering a range of KF 50% (Table 2 in S1 File). Overall, our data indicate the influence of seasonality and of the landscape composition in the commuting distances flown for by these species, a pattern not observed for the bats’ foraging areas.
Discussion
We found support for the prediction that landscape disturbances affect the commuting distances of F. horrens and L. aurita, and that this is mediated by seasonality. On the other hand, the estimated range areas for these two species were similar across preserved and anthropogenic landscapes. Moreover, we found support for our prediction of the differential use of habitats by the two bat species studied. We identified the selection of open areas for F. horrens and those included natural, preserved canga habitats, and the open degraded habitats represented by pastures. For L. aurita, forests were by far the most used habitats, and canga and pastures were only complementary as foraging areas.
A key point to our analysis of habitat preferences was the assumption that certain habitats may either be avoided or preferred relatively to others [56] and relative to what is available in the landscape. The sword-nosed bat L. aurita used proportionally most of what was available of forests, whereas the abundant forested habitats available for F. horrens were somewhat neglected by this species. Noteworthy results found for both species were that their use of space is markedly related to seasonality rather than to the landscape variation alone.
Data on the movement behavior of Neotropical bats in South America are scarce and include mostly estimates for phyllostomid species, such as nectar-feeding bat species monitored in the Brazilian Cerrado (Glossophaga and Lonchophylla) [21], fruit bats studied in the Brazilian Atlantic Forest (Carollia, Artibeus, and Sturnira) [20, 22], frugivores, aerial and foliage gleaning insectivores, and fish-eating bats radiotracked in the Tapajós region, Eastern Amazon´s (Artibeus, Carollia, Gardnerycteris, Tonatia, Trachops, and Noctilio) [23]. To our knowledge, we herein presented the first estimates of movement behavior for both L. aurita (Phyllostomidae), and for furipterids (Furipteridae), and that is why we cannot treat them in a broad comparative fashion.
Variations in habitat quality and availability, fine-tuned with specific foraging strategies, are expected to affect range area estimates as discussed elsewhere [19, 20, 64, 65]. Our estimates of area ranges and commuting distances for both species studied are larger, on average, than previous estimates found for other species of Neotropical bats [19–21, 23]. This is contrary to our expectations, particularly in the case of F. horrens, which is one of the tiniest Neotropical bats (3 to 4 g) as range sizes are expected to increase with body size and be limited by the energetic constraints in mammals [66, 67] including bats [68].
Our results suggest that these bats commute to forage differently in the pristine and disturbed landscapes of Carajás. They fly longer linear distances on disturbed landscapes in the wet season and shorter distances on preserved landscapes in the dry season (Fig 4). The influence of the seasonality, which modulates the bats’ responses to the landscape structures in Amazonia agrees with the general patterns previously described [24, 25, 69] primarily concerning to the spatiotemporal variability of food resources and the relative synchronicity of these fluctuations with the bats foraging activity and feeding behaviors. Seasonal reproductive and temporary environmental energetic demands require further investigation [26, 70]. Nonetheless, the two bat species studied appear to adjust the use of space to the observed changes in the landscape by accessing food resources similarly to their seasonal behavior. This may be evidenced by the range areas that remained similar between landscapes, by the larger effect size of “Season” on our models (Tables 1 and 2 in S1 File), and by the complementary use of anthropogenic habitats (Fig 3). Studies that evaluate the effects of mining on tropical bat communities are overall concurrent that natural habitat availability is a primary concern [71, 72], whereas studies on the movement of bats report their ability to adjust movements to explore fragmented landscapes [20, 21, 23]. We cannot make inferences about behavioral, ecological and physiological adjustments associated to the bats’ use of space based on the current data available, and further studies are necessary also to address these questions [1, 73].
All Neotropical bats use echolocation calls to some extent to navigate and search for food. The echolocation call patterns of F. horrens were recently described in more detail revealing that these bats emit short and extremely high frequency and broadband pulses that range from 70 to 210 kHz [74]. Based on the characteristics of the calls of F. horrens, resembling the calls of species of Old-World vespertilionid bats from the genus Kerivoula [75] Falcão and collaborators [74] suggested that they may also forage in dense vegetation. Our results appear to conflict with this suggestion because the combination of the movement patterns we found with the echolocation behavior described, is suggestive of patterns related to the Edge Space Aerial foragers guild as described by Denzinger et al. [76]. In terms of dietary preferences, the biology of Furipterus horrens is scarcely known [29, 74], and the little dietary data available recovered some evidence to the consumption of moths [29, 77]. Several species of moths have evolved sensitive ears to hearing the bats in one of the clearer examples of evolutionary predator-prey models [78]. Although along this “arms-race” some moths have evolved the capacity of hearing pulses much above 200 kHz, such as the case of greater wax moth, Galleria mellonella (Insecta: Lepidoptera) which can hear ultrasonic frequencies near 300 kHz [79] the combination of high frequency, low intensity and short calls of F. horrens may enable these bats to overcome the hearing abilities of moths [74], explaining, in part, these conflicting results.
Lonchorhina aurita, a foliage-gleaning, animalivorous phylostomid has been traditionally considered a narrow space forager and a “forest bat” [80]. As most animalivorous phyllostomids, bats from the genus Lonchorhina have been classified as passive gleaning foragers that would rely primarily on signs provided by their preys [76, 81]. However, several studies revealed refinements to our understanding of the evolution of foraging strategies and echolocation patterns of animalivorous phylostomids (e.g. Macrophyllum macrophyllum), including adaptations enabling then to increment their abilities to detect and locate prey with elaborated variations of echolocation calls, and some of them comparable to aerial insectivores from other bat families, such as Emballonuridae and Vespertilionidae [64, 82]. A recent study revealed that this is also the case for L. aurita, which has an unusual call structure among phyllostomids that deviates from the standard call design of bats of this family, and that may perhaps be related to its unique noseleaf, ear and tragus morphology. The echolocation system of L. aurita retains basic phyllostomid characteristics such as multi-harmonic FM-calls. On the other hand, as is typical in aerial insectivores, it has evolved to also include narrow-bandwidth CF-components and a terminal group [33]. Specialized foraging patterns among phylostomids have been more and more associated to the modulation of access to target resources and habitats [33].
Thus, by combining our results on habitat selection with the echolocation evidence available we can corroborate the suggestion of Gessinger et al. [33] of the evolution of an unusual foraging strategy for L. aurita consisting in a somewhat edge space aerial insectivore notably using primary forested habitats, and open habitats complementarily, accordingly to their availability. Our observations includes patterns of combined use of preserved and disturbed habitats, i.e. movements over fragmented landscapes that coincide with results obtained in other studies of insectivorous bats [64] and nectar feeding bats from the Cerrado [21] and poses further questions related to food constraints for both species, F. horrens and L. aurita. In central Amazonia, for instance, the foraging activity of the cave bat Pteronotus rubiginosus was mediated by insect availability, independent of the habitat types [83]. In a few words, a relevant point raised considering our results is the need to investigate how does the current mosaic of preserved and degraded habitats limit the food supply for these species. This is a key question because movements are costly, food sources are limiting, and all that information is important for the conservation and management programs of these bat species.
Food and roosts are key and limiting resources for bats [84] and ultimately, constraints to cave bats will affect cave ecosystems that largely depend upon the bat guano. This leads to the Brazilian policy concept “area of influence” of the cave, which would be an estimated space away from the cave perimeter, measured in kilometers needed to safeguard the biodiversity within the cave. Our incipient understanding of the mechanisms affecting the chances of Neotropical cave bats to effectively find and collect food in the surrounding landscapes, and the little knowledge about patterns of roosting selection of Neotropical cave bats prevents us from contributing with “area of influence” estimates, and that should be further investigated as it is crucial for bat and cave ecosystem conservation, and for exercises of prioritization for conservation.
Supporting information
S1 File. Results supporting analytical procedures.
Includes Table1-Exploratory analyses of movement data, Table2-Model selection for Linear Mixed-effects Models (LMMs), Table3-Selected LMM estimates for Maximum commuting distance (K95%).
https://doi.org/10.1371/journal.pone.0296137.s001
(XLSX)
Acknowledgments
We thank the health and safety teams of companies Ativo Ambiental and Vale S.A. for their help, as well as the technical team of Ativo Ambiental who actively participated in data collection. We are grateful to the contributions of the editors and two reviewers, which greatly improved our manuscript.
References
- 1. Romero LM. Physiological stress in ecology: lessons from biomedical research. Trends in Ecol and Evol. 2004 May;19(5):249–55. pmid:16701264
- 2. Gurarie E, Andrews RD, Laidre KL. A novel method for identifying behavioural changes in animal movement data. Ecology Letters. 2009 May;12:395–408. pmid:19379134
- 3. Mori E, Lovari S, Cozzi F, Gabbrielli C, Giari C, Torniai L et al. Safety or satiety? Spatiotemporal behaviour of a threatened herbivore. Mamm Biol. 2020 February;100:49–61.
- 4. Rosenzweig ML, Abramsky Z. Centrifugal community organization. Oikos. 1986 May;46:339–348.
- 5. Schick RS, Loarie SR, Colchero F, Best BD, Boustany A, Conde DA et al. Understanding movement data and movement processes: current and emerging directions. Ecol. Lett. 2008 Dec;11:1338–1350. pmid:19046362
- 6. Trajano E. Protecting caves for the bats or bats for the caves? Chiroptera Neotropical. 1995 Dec;1(2):19–22.
- 7.
Culver DC, Pipan T. The Biology of Caves and Other Subterranean Habitats. Oxford: Oxford University Pres; 2019.
- 8.
Furey NM, Racey PC. Conservation ecology of cave bats. In Voigt C, Kingston T, editors. Bats in the Anthropocene: Conservation of Bats in a Changing World. Springer International Publishing; 2016. p. 463–500.
- 9.
Speakman JR, Thomas DM. Physiological Ecology and energetics of bats. In Kunz TH, Fenton MB, editors. Bat Ecology. Chicago: University of Chicago Press, Illinois; 2003. p. 430–492.
- 10. Phelps K, Jose R, Labonite M. Correlates of cave-roosting bat diversity as an effective tool to identify priority caves. Biol. Conserv. 2016 Sept 201;201–209.
- 11. Struebig MJ, Kingston T, Zubaid A, Le Comber SC, Mohd-Adnan A, Turner A et al. Conservation importance of limestone karst outcrops for palaeotropical bats in a fragmented landscape. Biol. Conserv. 2009 Oct;142:2089–2096.
- 12. Ferreira RL, Martins RP. Trophic Structure and Natural History of Bat Guano Invertebrate Communities with Special Reference to Brazilian Caves. Tropical Zool. 1999 Jul;2:231–252.
- 13. Jaffé R, Prous X, Zampaulo R, Giannini TC, Imperatriz-Fonseca VL, Maurity C Oliveira G. Reconciling mining with the conservation of cave biodiversity: a quantitative baseline to help establish conservation priorities. PLoS One 2016 Dec; 11: e0168348. pmid:27997576
- 14. Sonter LJ, Ali SH, Watson JEM. Mining and biodiversity: key issues and research needs in conservation science. Proc. R. Soc. B. 2018 Dec;285: 20181926. pmid:30518573
- 15. Trevelin LC, Gastauer M, Prous X, Nicácio G, Zampaulo R, Brandi I et al. Biodiversity surrogates in Amazonian iron cave ecosystems. Ecol Indicators 2019 Jun;101:813–820.
- 16. Trevelin LC, Simões MH, Prous X, Pietrobon T, Brandi IV, Jaffé R. Optimizing speleological monitoring efforts: insights from long-term data for tropical iron caves. PeerJ. 2021 Apr;9: e11271. pmid:33959423
- 17. Umetsu F, Metzger JP, Pardini R. The importance of estimating matrix quality for modeling species distribution in complex tropical landscape: a test with Atlantic Forest small mammals. Ecography 2008 Jun;31:359–370.
- 18. Puttker T, Bueno AA, de Barros CD, Sommer S, Pardini R. Immigration rates in fragmented landscapes–empirical evidence for the importance of habitat amount for species persistence. PLoS ONE 2011 Nov;6:e27963. pmid:22125643
- 19. Kalko EKV, Fiernel D, Handley Jr. C, Schnitzier NV. Roosting and foraging behavior of two Neotropical gleaning bats, Tonatia silvicola and Trachops cirrhosus (Phylostomidae). Biotropica 2006 Mar;34:344–353.
- 20. Trevelin LC, Silveira M, Port-Carvalho M, Homem DH, Cruz-Neto AP. Use of space by frugivorous bats (Chiroptera: Phyllostomidae) in a restored Atlantic Forest fragment in Brazil. Forest Ecology and Management 2013 Mar;291:136–143.
- 21. Aguiar LMS, Bernard E, Machado RB. Habitat use and movements of Glossophaga soricina and Lonchophylla dekeyseri (Chiroptera: Phyllostomidae) in a Neotropical savannah. Zoologia (Curitiba) 2014 Jun;31: 223–229.
- 22. Kerches-Rogeri P, Niebuhr BB, Muylaert RL, Mello MAR. Individual specialization in the space use of frugivorous bats. Journal of Animal Ecol. 2020 Sep; 89: 2584–2595.
- 23. Bernard E, Fenton B. Bat mobility and roosts in a fragmented landscape in central Amazônia, Brazil. Biotropica 2003 Mar;35:262–277.
- 24. Klingbeil BT, Willig, MR. Seasonal differences in population‐, ensemble‐and community‐level responses of bats to landscape structure in Amazonia. Oikos 2010 Sep;119(10), 1654–1664.
- 25. Ferreira DF, Rocha R, López‐Baucells A, Farneda FZ, Carreiras J M, Palmeirim, J M, et al. Season‐modulated responses of Neotropical bats to forest fragmentation. Ecology and Evolution, 2017 May;7(11), 4059–4071. pmid:28616200
- 26. Meyer CF, Struebig MJ, Willig MR. Responses of tropical bats to habitat fragmentation, logging, and deforestation. In Voigt C, Kingston T, editors. Bats in the Anthropocene: conservation of bats in a changing world, 2016. p. 63–103.
- 27. Tavares VC, Palmuti CFS, Gregorin R, Dornas TT. Morcegos. In Martins FD, Castilho AF, Campos J, Hatano FM, Rolim SG, editores. Fauna da Floresta Nacional de Carajás: Estudos Sobre Vertebrados Terrestres, 2012. p. 161–177. Nitro Imagens, São Paulo, Brazil.
- 28. Handley CO Jr. Mammals of the Smithsonian Venezuelan project. Brigham Young University Science Bulletin, Provo, 1976; 20 (5): 1–91.
- 29. Uieda W. Ocorrência de Carollia castanea na Amazônia Brasileira (Chiroptera, Phyllostomidae). Acta Amaz. 1980; 10: 936–938.
- 30. Tavares VC, Nobre CC, Palmuti CF, Nogueira EDP.; Gomes J.D.et al. The bat fauna from southwestern Brazil and its affinities with the fauna of western Amazon. Acta Chiropterol. 2017;19: 93–106.
- 31. Simmons NB, Voss RS. The mammals of Paracou, French Guiana, a Neotropical lowland rainforest fauna. Part 1, Bats. Bull. Am. Mus. Nat. Hist. 1998;237: 1–219.
- 32.
ICMBio. Sistema de Avaliação do Risco de Extinção da Biodiversidade–SALVE. 2023. Available from: https://salve.icmbio.gov.br/. Accessed: 13 de nov. de 2023.
- 33. Gessinger G, Gonzalez-Terrazas TP, Page RA, Jung K, Tschapka, M. Unusual echolocation behaviour of the common sword nosed bat Lonchorhina aurita: An adaptation to aerial insectivory in a phyllostomid bat? Royal Society Open Science. 2019; 6: 182165.
- 34. Patterson BD. Mammals in the Royal Natural History Museum, Stockholm, collected in Brazilcand Bolivia by A.M. Olalla during 1934–1938. Fieldiana Zoology (new ser.). 1992; 66: 1–42.
- 35.
Piló LB, Auler AS, Martins FD. Carajás National Forest: Iron ore plateaus and caves in southeastern Amazon. In: Vieira BC, Salgado AAR, Santos LJC. editors. Landscapes and Landforms of Brazil. Berlin: Springer; 2015. p. 273–83.
- 36.
Poveromo JJ. Iron Ores: The Making, Shaping, and Treating of Steel: Ironmaking Volume. Pittsburg, PA: The AISE Steel Foundation. 1999; p. 547–550.
- 37. Viana PL, Mota NFO. Flora das cangas da Serra dos Carajás, Pará, Brasil: Styracaceae. Rodriguésia. 2016;67: 1477–1480.
- 38. Devecchi MF, Lovo J, Moro MF, Andrino CO, Barbosa-Silva RG, Viana PL, et al. Beyond forests in the Amazon: Biogeography and floristic relationships of the Amazonian savannas. Botanical Journal of the Linnean Society. 2020;193: 478–503.
- 39.
MapBiomas, Projeto de Mapeamento Anual da Cobertura e Uso do Solo do Brasil. 2020. Available from: http://www.mapbiomas.org. Acessed in: fevereiro 2020.
- 40. Peel MC, Finlayson BL, Mcmahon TA. Updated world map of the Köppen-Geiger climate classification. Hydrology and earth system sciences discussions. 2007;4: 439–473.
- 41.
Kunz TH, Hodgkison R, Weise CD. Methods of capturing and handling bats. In: Kunz TH.; Parsons S, eds. Ecological and behavioral methods for the study of bats. Baltimore: Johns Hopkins University Press. 2009. p. 3–35.
- 42.
Brunet-Rossini AK, Wilkinson GS. Methods for estimating age in bats. In: Kunz TH, Parsons S, editors. Ecological and behavioral methods for the study of bats. Baltimore: Johns Hopkins University Press. 2009. p. 315–326.
- 43. Brigham R. Load carrying and maneuverability in an insectivorous bat: a test of the 5% "rule" of radio-telemetry. J Mammal. 1988;69(2):379–82.
- 44. O’Mara MT, Wikelski M, Dechmann DK. 50 years of bat tracking: device attachment and future directions. Methods Ecol Evol. 2014;5(4):311–9.
- 45.
Amelon SK, Dalton DC, Millspaugh JJ, Wolf SA. Radiotelemetry; techniques and analysis. In: Kunz TH, Parsons S, editors. Ecological and behavioral methods for the study of bats. Baltimore: Johns Hopkins University Press; 2009. p. 57–77.
- 46. Sikes RS, Gannon WL. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. J Mammal. 2011;92: 235–253.
- 47. Börger L, Franconi N, de Michele G, Gantz A, Meschi F, Manica A, Lovari S. Coulson T. Effects of sampling regime on the mean and variance of home range size estimates. J. Anim. Ecol. 2006.;75: 1393–1405. pmid:17032372
- 48.
Jacob AA, Rudran R. Radiotelemetria em estudos populacionais. In: Cullen L Jr, Rudran R, Valladares-Padua C. (Eds.), Métodos de Estudos em Biologia da Conservação e Manejo da Vida Silvestre. Editora da UFPR & Fundação O Boticário de Proteção à Natureza, Curitiba, 2003. p. 285–341.
- 49. Rodgers AR, Carr AP, Beyer HL, Smith L, Kie JG. HRT: Home Range Tools for ArcGIS, 1.1. Centre for Northern Forest Ecosystem Research, Ontario Ministry of Natural Resources, Ontario. 2007.
- 50.
Kernohan BJ, Gitzen RA, Millspaugh JJ. Analysis of animal space use and movement. In: Millspaugh J. J. & Marzluff J. M. (Eds.), Radio Tracking and Animal Populations. Academic Press, San Diego, 2001. p. 126–166.
- 51.
White GC, Garrott RA. Analysis of Wildlife Radiotracking Data. Academic Press, San Diego. 1990.
- 52. Seaman DE, Powell RA. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology. 1996;77: 2075–2085.
- 53. Johnson DH. The comparison of usage and availability measurements for evaluating resource preference. Ecology. 1980;61: 65–71.
- 54. Fattorini L, Pisani C, Riga F, Zaccaroni M. A permutation-based combination of sign tests for assessing habitat selection. Env Ecol Stat. 2014; 21:161–187.
- 55.
R Core Team. R Project for Statistical Computing, 2019.
- 56. Fattorini L, Pisani C, Riga F, Zaccaroni M. The R package “phuassess” for assessing habitat selection using permutation-based combination of sign tests. Mammal Biol. 2017;83: 64–70.
- 57.
Zuur AF, Leno EN, Smith GM. Analyzing Ecological Data. Springer, New York. 2007.
- 58.
Burnham KP, Anderson DR. Model Selection and Multimodel Inference: a Practical Information-theoretical Approach, 2nd. New York: Springer. 2002.
- 59. Harrison XA, Donaldson L, Correa-Cano ME, Evans J, Fisher DN, Goodwin CED, et al. A brief introduction to mixed effects modelling and multi-model inference in ecology. PeerJ. 2018;6: e4794. pmid:29844961
- 60. Brooks ME, Kristensen K, van Benthem KJ, Magnusson A, Berg CW, Nielsen A, et al. glmmTMB Balances Speed and Flexibility Among Packages for Zero-inflated Generalized Linear Mixed Modeling. The R Journal, 2017;9(2), 378–400.
- 61.
Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, 2016. Available from: https://ggplot2.tidyverse.org.
- 62. Barton K. MuMIn: Multi-model inference. 2020. Available from: https://CRAN.R-project.org/package= MuMIn.
- 63.
Hartig F. DHARMa: residual diagnostics for hierarchical (multi-level/mixed) regression models. R package version 0.1. 0. 2016.
- 64. Meyer CFJ; Weinbeer M; Kalko EKV. Home-range size and spacing patterns of Macrophyllum macrophyllum (Phyllostomidae) foraging over water. J. Mammal. 2005;86: 587–598.
- 65. Kniowski AB, Gehrt SD. Home range and habitat selection of the Indiana bat in an agricultural landscape. The Journal of Wildlife Management. 2014;78(3), 503–512.
- 66. Lindstedt SL, Miller BJ, Buskirk SW. Home range, time, and body size in mammals. Ecology. 1986;67:413–418.
- 67. Kelt DA, Van Vuren D. Energetic constraints and the relationship between body size and home range area in mammals. Ecology. 1999;80:337–340.
- 68. Fenton MB. Science and the conservation of bats. Journal of Mammalogy 1997;78:1–14.
- 69. Willig MR, Presley SJ, Plante JL, Bloch CP, Solari S, Pacheco V, et al. Guild-level responses of bats to habitat conversion in a lowland Amazonian rainforest: species composition and biodiversity. Journal of Mammalogy. 2019; 100(1), 223–238. pmid:30846887
- 70. Appel G, López-Baucells A, Magnusson WE, Bobrowiec PED, Temperature, rainfall, and moonlight intensity effects on activity of tropical insectivorous bats, Journal of Mammalogy. 2019 dec;100(6):1889–1900.
- 71. Cory-Toussaint D, Taylor PJ. Anthropogenic light, noise, and vegetation cover differentially impact different foraging guilds of bat on an opencast mine in South Africa. Frontiers in Ecology and Evolution 2022;10:752665.
- 72. Franco-Rozo MC, Blanco-Torres A, Gómez-Valencia B, Etter A. Biodiversity responses to landscape transformations caused by open-pit coal mining: An assessment on bats and dung beetles in a Colombian tropical dry forest. Environmental and Sustainability Indicators 2024; 21:100335.
- 73. Miguel PH, Kerches-Rogeri P, Niebuhr BB, Cruz RAS, Ribeiro MC, Cruz Neto AP. Habitat amount partially affects physiological condition and stress level in Neotropical fruit-eating bats. Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology. 2019;237:110537. pmid:31400499
- 74. Falcão F, Ugarte-Núñez JA, Faria D, Caselli CB. Unravelling the calls of discrete hunters: acoustic structure of echolocation calls of furipterid bats (Chiroptera, Furipteridae), Bioacoustics: The International Journal of Animal Sound and its Recording. 2015; http://dx.doi.org/10.1080/09524622.2015.1017840.
- 75. Schmieder DA, Kingston T, Hashim R, Siemers BM. Breaking the trade-off: rainforest bats maximize bandwidth and repetition rate of as they approach prey. Biol Lett. 2010;6(5):604–609. pmid:20356884
- 76. Denzinger A, Tschapka M, Schnitzler H. The role of echolocation strategies for niche differentiation in bats. Canadian Journal of Zoology. 2018;96: 171–181.
- 77. Fenton MB, Whitaker JO Jr, Vonhof MJ, Waterman JM, Pedro WA, Aguiar L, et al. The diet of bats from Southeastern Brazil: the relation to echolocation and foraging behaviour. Revista Brasileira de Zoologia. 1999;16(4):1081–1085.
- 78. Ter Hofstede HM, Ratcliffe JM. Evolutionary escalation: the bat-moth arms race. J Exp Biol.2016; 219:1589–602. pmid:27252453.
- 79. Moir HM, Jackson JC, Windmill JFC. Extremely high frequency sensitivity in a ‘simple’ ear. Biol. Lett. 2013;92013024120130241
- 80. Lassieur S, Wilson DE. Lonchorhina aurita. Mammalian Species. 1989;347: 1–4.
- 81. Goerlitz HR, Greif S, Siemers BM. Cues for acoustic detection of prey: insect rustling sounds and the influence of walking substrate. J. Exp. Biol. 2008;211: 2799–2806. pmid:18723538
- 82. Weinbeer M, Kalko EK, Jung K. Behavioral flexibility of the trawling long-legged bat, Macrophyllum macrophyllum (Phyllostomidae). Front Physiol. 2013;4:342. pmid:24324442; PMCID: PMC3838978.
- 83. de Oliveira LQ, Marciente R, Magnusson WE, Bobrowiec PED. Activity of the insectivorous bat Pteronotus parnellii relative to insect resources and vegetation structure. Journal of Mammalogy. 2015;96(5), 1036–1044.
- 84.
Kunz TH, Lumsden LF. Ecology of cavity and foliage roosting bats. In: Kunz TH, Fenton MB(Eds) Bat Ecology. The University of Chicago Press, Chicago, 2003. p.3–87.