Daniel García-Valverde1
Luis Esteban Vargas-Castro- 1Escuela de Ciencias Exactas y Naturales, Universidad Estatal a Distancia (UNED), San Jose, Costa Rica
- 2Bat Henry Conservation Project, Sarapiqui, Costa Rica
- 3Laboratorio de Vida Silvestre y Salud, Universidad Estatal a Distancia (UNED), San Jose, Costa Rica
Understanding bat activity in areas with human disturbance is increasingly important for their conservation. Bat colonies roosting on infrastructure used for tourism may be negatively affected over time, as tourism-related activities can disrupt roosting behavior, and affect colony size and stability. In this study, we evaluated the impact of tourist load on bat activity with passive acoustic monitoring in one Costa Rican eco-lodge, which hosts the only currently known colony of thumbless bats, Furipterus horrens, in the region. Their elusive behavior, limited survey records, and fragmented geographic distribution within the Neotropics, make them one of the least studied bat species. The colony we studied roosts under a series of bungalows used for tourist accommodation. We deployed an acoustic system composed of 19 recorders distributed under six bungalows, to monitor bat acoustic activity. Each bungalow was recorded for up to 43 nights in total, within five monitoring periods between February and April 2025. In addition, we conducted visual count surveys during each acoustic monitoring period. We gathered data on bungalow occupancy and analyzed the relationship between the cumulative reservation load of three and seven consecutive nights, on bat activity, using generalized linear mixed models. Our results show that for a given bungalow, higher bat acoustic activity significantly predicts higher abundance of individuals. Furthermore, even though acoustic activity was not affected by a three-night reservation load, the seven-night reservation load negatively affected the acoustic activity of thumbless bats, suggesting that this species is sensitive to an accumulated human disturbance over this period. These findings highlight the importance of adequate management plans that integrate people and bat populations, particularly in the context of high anthropogenic loads, like tourism-related activities in eco-lodges. Understanding these impacts provides critical information for making better-informed decisions that balance ecotourism with the conservation of vulnerable wildlife.
1 Introduction
Human-caused landscape modifications and activities are forcing wildlife into closer contact with humans as suitable natural habitat is reduced. Wildlife is often negatively impacted, with lower species abundance and decreases in species richness being the most common consequences (Murphy and Romanuk, 2013; Fetene et al., 2019), but other negative consequences include changes in diel activity patterns, predator dynamics and patterns of spatial distribution due to resource depletion (Young et al., 2011; Fetene et al., 2019; Ferreira et al., 2022), endangering the long-term stability of populations.
The conservation of bats has become an important area of research as increasing anthropogenic pressure reshapes the natural habitats of these mammals. Bats play an important role in ecosystems functioning as insect population regulators and key pollinators (Kunz et al., 2011), and seed-dispersers (Reid et al., 2015; Mayta et al., 2024). However, their populations and ecological functions are being challenged in a world where bat and human activities are increasingly overlapping. In general, human-induced disturbance is expected to modify the foraging and resting behavioral patterns of bats, ultimately threatening their biodiversity (Arlettaz et al., 2000; Russo and Ancillotto, 2015; Jung and Threlfall, 2015; Chaves, 2024). Furthermore, as natural habitats become reduced and fragmented, bat species are increasingly forced to rely on artificial or human-made roosts like buildings or houses (Voigt and Kingston, 2016; Lučan et al., 2024).
Impacts of bats roosting on anthropogenic structures have been widely documented across the globe (Jackson et al., 2023; Sippola et al., 2025). Some bat species show greater synanthropic adaptability and may benefit from using anthropogenic structures for predator protection or favorable microclimates (Russo and Ancillotto, 2015). In eastern Madagascar, large molossid colonies roost in public buildings, choosing sites based on building type, height and lack of fire use (Lopez-Baucells et al., 2017). Light pollution in human-inhabited areas may benefit insectivorous bat species when gathering food, as artificial lights attract insects (Rydell, 1992; Frank et al., 2018). In contrast, the behavior of other species such as Daubenton’s bats (M. daubentonii) seems to be unaffected by the presence of light (Spoelstra et al., 2018). Lastly, in other cases, bat species have been reported to experience colony declines under high anthropogenic pressure (Atagana et al., 2021).
In addition to the use of anthropogenic structures, pressure from human-related activities such as tourism may also disrupt wildlife behavior and habitats. Depending on the nature of the relationship between humans and animals at a particular site, tourist activities may positively or negatively affect the avoidance responses, time budgets, or physiological responses of animals (Bateman and Fleming, 2017). In the case of bats, high levels of tourism-related human activity, noise and artificial lighting, can lead to changes in bat behavior and ecology (MacKinnon et al., 2003; Cardiff et al., 2012). Nonetheless, it has been argued that proper management and habituation allow some species of bats to endure tourist-related anthropogenic pressure (Cardiff et al., 2012).
The tourism sector, particularly ecotourism, has had a substantial increase in Costa Rica over the last decades (Jones and Spadafora, 2017; OECD, 2020; Moya-Calderón et al., 2025). With about 118 species of bats (Ramírez-Fernández et al., 2023), Costa Rica also possesses a relatively high diversity of bat species, where bat and human activities are often intertwined.
It is common to see ecotourism initiatives such as bat tours or night walks. The Honduran white bat (Ectophylla alba), for example, is a highly sought-after species and local biological reserves offer bat tours that provide visitors the opportunity to observe these bats, among other species (Arroyo-Solórzano and Rojas-Prendas, 2021; García-Sánchez and González-Chaverri, 2022; Fonda et al., 2025). Nonetheless, studies have shown that some sites, such as the caves of the Brunca region have experienced signs of negative effects of human proximity, mostly from unregulated tourist visitation (Deleva and Chaverri, 2018).
Thumbless bats Furipterus horrens provide an interesting case of the potential effects that tourism-related pressure may have on wildlife populations using anthropogenic structures. A F. horrens colony was found in Costa Rica in 2017, after a forty-four-year gap of no country reports for the species (LaVal, 1977; Alfaro-Lara et al., 2018). Their elusive behavior, limited survey records, and fragmented geographic distribution, make them one of the least studied bat species in the Neotropics (Alfaro-Lara et al., 2018; Monteiro et al., 2023; Moras et al., 2024; Rodriguez-Segovia and Carrera-E, 2025). Visual counts estimated 100–130 individuals in this colony, roosting within empty spaces below a series of bungalows in a rural eco-lodge (Alfaro-Lara et al., 2018). Although the bungalows are less frequently used than other hotel areas, the colony is still exposed to human-caused disturbance, such as noise, vibrations and other anthropogenic stressors due to tourist activity (Alfaro-Lara et al., 2018), yet the actual impact of tourist activities in the colony has not been assessed.
As far as we know, there are very few other reports of F. horrens colonies in anthropogenic environments. One colony has been observed roosting in close proximity to a highway in Brazil (Monteiro et al., 2023). Another colony is known to roost inside caves near iron-mining areas, also located in Brazil (Appel et al., 2025; Tavares et al., 2025). The bungalow colony, however, not only roosts near human activity but under direct contact with it. Direct visual observations by one of us (HAL) between 2021 and 2023 suggest a decline in colony size, but no formal evidence exists. Given that this is the only known colony of this species using anthropogenic roosting sites in Central America, it is important to assess the potential effects of human-caused disturbance this colony experiences in order to inform management plans.
In this study, we investigated the F. horrens colony at the bungalow sites using acoustic monitoring and visual surveys. We also compiled hotel reservation data to analyze the potential impacts of tourist occupancy on bat colony activity. We argue that human-caused disturbance related to cumulative tourist load has a negative effect on F. horrens activity and abundance. We predict that greater accumulation of tourist loads will result in lower levels of bat acoustic activity. Our objectives were: (i) to describe the spatial and temporal patterns of bat acoustic activity and bungalow use by tourists, (ii) to test if acoustic activity can be used to infer bat abundance in the colony, and (iv) to assess the cumulative effects of tourist loads on bat activity.
2 Methods
2.1 Study site
The study was conducted at the hotel Selva Verde Lodge, an eco-lodge located in Puerto Viejo de Sarapiquí, Heredia, Costa Rica (10.45053°N, 84.06950°W), at an elevation of 117 masl. In the hotel property, there are seven bungalows for tourist use of 15 m2 each. In this study bungalows were divided into six sampling units (B1-B6), as bungalows 6 and 7 share the same building structure (B6, which is about twice the area of the other bungalows).
Beneath these constructions there are open spaces ranging from 60 to 95 cm in height above ground, which serve as roosting sites for the Thumbless bat Furipterus horrens. For a map of the study area and photographs of the individuals, see Alfaro-Lara et al. (2018). Fieldwork was conducted during the dry season, between February and April 2025, at these roosting sites.
2.2 Acoustic monitoring
Passive acoustic monitoring was used to quantify bat activity across the six bungalow sites. We used a system composed of 19 acoustic recorders (AudioMoth version 1.2.0, firmware version 1.11.0; Open Acoustic Devices, Oxford, UK). We deployed three recorders on each site, except for bungalow 6 (B6), where four recorders were deployed to increase coverage due to its larger size.
Recorders were programmed to operate daily from 1600–0600 local time (UTC-6). A partial schedule allowed us to maximize sampling period length and reduce the number of maintenance routines, by lowering battery and storage consumption. Two of the three recorders per site were set to record with a sampling rate of 384 kHz for 1-minute intervals every 6 minutes (1 min recording + 5 min rest cycles), and high gain level. These recorders were designated as ‘fixed recorders’.
We programmed the third recorder to use a frequency-based trigger (Hill et al., 2018) using the same sampling rate (384 kHz), but only when detecting sounds ≥ 120 kHz. It employed a Goertzel filter window length of 32 samples, a detection threshold of 0.4%, a minimum trigger duration of 1 s, and high gain level. While detections were saved as 1 h interval recordings (59 min 55 s recording + 5 s rest cycles), the resulting files are often shorter in duration or ‘compressed files’, as only the time portions where detections were found are stored. These recorders were designated as ‘detector recorders’. The additional unit deployed in B6 was also a detector recorder.
All recorders were deployed under the bungalows (between the ground and platform) inside plastic resealable bags. We placed fixed recorders near potential bat entry/exit points, while detector recorders were placed at a close distance (≤ 1 m) from observed roosting spots. Recorders were strategically and carefully positioned to minimize potential disturbance to bats; once deployed, they operated autonomously with null intervention.
We conducted five sampling periods. In each period, recorders were deployed for an average of 9 consecutive days (range: 7–10 days), until they ran out of memory or battery, and being retrieved for maintenance. Consequently, each bungalow was recorded for up to 43 nights in total.
2.2.1 Audio processing and species call identification
We expanded detector-type recordings (to recover their original timestamp) and split the resulting audio segments to a maximum duration of 15 s using the AudioMoth Configuration App version 1.11.1 (Open Acoustic Devices, Oxford, UK), in order to facilitate call visualization and classification. Fixed-type recordings were kept at their original duration of 1 min.
Species call identification was conducted in two phases. First, we developed an automated classifier using Kaleidoscope Pro (version 5.7.0; Wildlife Acoustics Inc., Maynard, USA) to detect potential echolocation calls of the target species F. horrens. The classifier was initially tested using call parameters reported in Falcão et al. (2015) and subsequently refined with verified F. horrens recordings collected during field work, where the species was visually confirmed at the roost.
Our final classifier used the following parameters: minimum frequency of 120 kHz, maximum frequency of 180 kHz, pulse duration between 0.1 and 5 ms, a maximum inter-pulse interval of 100 ms, and a minimum of one pulse per sequence. Based on these parameters, each detection was then assigned to one of three clusters: FURHOR, FURHOR2 or FALSE (which included 14866, 12272, and 899202 detections, respectively); which were highly similar, moderately similar or dissimilar to the reference calls, respectively.
In the second phase, we manually verified the detections labeled as FURHOR and FURHOR2. Manual inspection involved visual and auditory analysis of the spectrograms to confirm the presence of F. horrens calls. Detections labeled as FURHOR or FURHOR2 were either confirmed as the target species or reclassified as FP (false positives; 26366 detections). Detections initially assigned to the FALSE cluster were also verified, partially, based on stratified sampling, as follows. For each sampling period, we randomly selected a sample of 25% of the bungalow-recorder combinations. This procedure resulted in 23 bungalow-recorder combinations (4–5 combinations per sampling period), covering all bungalows. For each bungalow-recorder combination, we randomly selected one index of the respective FALSE detection list and manually verified a sequence of 10% of the FALSE detections sorted by chronological order. For example, if the combination ‘bungalow B6-recorder 190D’ had 9086 FALSE detections, we manually verified 909 of these detections starting on detection 185 (randomly selected index). In total, 29059 FALSE detections were manually verified, in which only 3 FN (false negative) detections were found, resulting in an FN rate of 0.004. The rest of FALSE detections were assumed to be TN (true negatives). Since F. horrens calls have extremely high frequency and it is very challenging to record them, our classifier was designed to maximize recall (accepting many FP to minimize FN). Our final recall was very high, 99.6%, indicating that practically all positives (F. horrens calls) were properly classified as TP (true positives). This process resulted in a total of 772 confirmed F. horrens detections during the study.
2.3 Visual surveys
We conducted visual surveys to assess whether acoustic activity reflected bat abundance. One visual survey was conducted within each of the five sampling periods, except for one in which two surveys were conducted. All surveys were scheduled to closely align with acoustic monitoring, occurring within 1–3 days after deploying the acoustic recorders in each period. This timing allowed for direct comparisons between the level of acoustic activity and observed abundance within a short time window.
Visual surveys were conducted during daylight hours, between 07:00 and 15:00, when F. horrens bats are roosting. This timing also facilitated visual counts under natural light conditions. Each visual survey involved systematically inspecting the underside of all six bungalow sites. Artificial light was avoided as much as possible; faint lanterns were only used in cases of poor visibility, and with minimal intensity to avoid disturbing bat behavior.
2.4 Hotel reservations and tourist load
We analyzed the relationship between tourist load and bat acoustic detections to examine the potential effect of human disturbance on bat activity. We obtained hotel reservation data segregated by bungalow number, which also included the number of guests and number of nights per reservation. Reservation data was handled anonymously. Reservations for bungalows 6 and 7 were merged under B6, consistent with the data structure used in acoustic and visual monitoring.
2.5 Statistical analysis
All statistical tests were conducted using R software version 4.4.3. Unless otherwise stated, we employed a significance level of 0.05. When significant, generalized linear mixed-effects model (GLMM) tourist load predictions were generated on the response scale including random effects with the glmmTMB package. Plots were made using the ggplot2 version 3.5.2 and ggh4x version 0.3.1 packages in R.
2.5.1 Acoustic detection frequency
We tested if the frequency of acoustic detections was evenly distributed among the six bungalow sites using a goodness-of-fit chi-square test. Then we applied post-hoc chi-square tests to compare each bungalow against the remaining five, employing a Bonferroni correction for multiple comparisons, to determine which bungalows showed significantly higher or lower numbers of calls from those expected. Additionally, we examined the time of day in which bats were acoustically active by grouping detections into 1 h blocks over the whole study. We tested if the frequency of acoustic detections was evenly distributed among the 14 one-hour blocks using a goodness-of-fit chi-square test.
2.5.2 Acoustic activity and bat abundance relationship
We modeled visual survey counts as a function of acoustic detections to evaluate whether acoustic activity was a useful predictor of bat abundance. We paired visual counts with acoustic detection records of the same night and the night before of each visual survey date, grouping data by date. For example, a visual survey conducted on February 22 would be paired with total acoustic detections from February 21 (16:00–06:00) and February 22 (16:00–06:00, including the following morning). We aimed to capture a robust time window of bat activity surrounding the visual surveys.
First, we fit a generalized linear mixed-effects model (GLMM), with a Poisson error distribution and log-link model, implemented in the lme4 package in R. The response variable was number of individuals (as obtained from the visual counts) and number of acoustic detections was the fixed predictor. We specified random intercepts for bungalow code (1-6) and sampling period (1-5), to account for non-independence in the data, allowing variation in the baseline levels of the response variable across both random factors. Second, we run an overdispersion test with the DHARMa package version 0.4.7 in R. Since we confirmed overdispersion in the initial model (χ² = 7.47, df = 32, p < 0.001), we then fit another GLMM with a negative binomial distribution and log-link model using the same mixed-effects model structure with the glmmTMB package version 1.1.11 in R. Finally, we also conducted a zero-inflation test with DHARMa on the negative binomial model, which indicated no evidence of zero-inflation (ratio = 1.00, p-value = 1.00), justifying the use of the negative binomial model.
2.5.3 Reservation frequency among bungalows
We tested if the total frequency of nights (number of nights were at least one person stayed) and persons (number of persons that stayed at least one night) was evenly distributed among the six bungalow sites during the study using a goodness-of-fit chi-square test. Then we applied post-hoc chi-square tests to compare each bungalow against the remaining five, employing a Bonferroni correction for multiple comparisons, to determine which bungalows showed significantly higher or lower numbers of nights or persons from those expected.
2.5.4 Effect of tourist load on acoustic activity
We assessed the effect of human-induced disturbance on acoustic activity by modeling the effect of tourist load on the number of acoustic detections. Tourist load describes the cumulative load of bungalow reservations, considering both the number of nights and the number of persons (guests) over time, in two variables: a) 3-day load, the sum of person-nights over each acoustic monitoring day and two preceding nights, and b) 7-day load, the sum of person-nights over each acoustic monitoring day and six preceding nights. For example, in the case of the 3-day load, the number of acoustic detections on February 22 would be paired with the tourist load of February 20-22. Continuing with the example, if one person stayed on February 20 for a single night, and then two persons arrived on February 21 and stayed for two nights, this results in a 3-day load score of 5 for that particular bungalow. For this part of the analysis, we grouped data by date and bungalow.
We first fit two generalized linear mixed-effects models (GLMM), one for each level of tourist load (3-day and 7-day loads), with a Poisson error distribution and log-link model, using the lme4 package in R. The response variable was number of acoustic detections while the fixed predictor was the respective tourist load. We specified random intercepts for bungalow code (1-6) and sampling period (1-5), to account for both sources of non-independence. Similarly, we run overdispersion tests with DHARMa and confirmed overdispersion in both initial models (3-day load: χ² = 11.69, df = 260, p < 0.001; and 7-day load: χ² = 10.51, df = 260, p < 0.001), so we fit two new GLMMs with a negative binomial distribution and log-link model, using the same mixed-effects model structure and the glmmTMB package. Lastly, we tested for zero-inflation in the latter models using DHARMa, but we found no evidence of zero-inflation (3-day load: ratio = 1.00, p = 0.93; and 7-day load: ratio = 1.00, p-value= 1.00), justifying the use of the negative binomial GLMMs.
3 Results
3.1 Acoustic activity patterns
Acoustic monitoring over the five sampling periods yielded 772 detections of thumbless bats Furipterus horrens. Acoustic detection frequency varied between bungalows (χ² = 1068.9, df = 5, p < 0.001) (Figure 1A). Post-hoc tests showed significant higher acoustic activity than expected in bungalows B3 and B6 (adjusted p < 0.001 in all cases). In contrast, bungalows B1, B4 and B5 showed significantly lower acoustic activity than expected (adjusted p < 0.001 in all cases). In addition, F. horrens bats showed acoustic activity peaks at 5, 16 and 17 hours, matching the sunrise and sunset periods, respectively, with no detections recorded in between (Figure 2).
Figure 1. (A-C) Number of acoustic detections of thumbless bats Furipterus horrens (A), nights where at least one person stayed (B), and persons who stayed at least one night (C), by bungalow at a hotel located in Sarapiqui, Costa Rica.
Figure 2. Number of acoustic detections of thumbless bats Furipterus horrens according to time of day in 1 h blocks.
3.2 Acoustic activity and bat abundance relationship
We counted 38 bats on average (range: 30–46 bats) per visual survey, including all bungalows. Most individuals roost together under a given bungalow. On average, 94% of the individuals (range: 67-100%) per survey were observed roosting together. However, the preferred bungalow switched between visual surveys among B2, B3 and B6.
Out of the 772 acoustic detections, 193 were paired with visual surveys (those collected during the same night and the night before). We found a significant effect of number of acoustic detections on the number of individuals (estimate = 0.14, p < 0.05) (Table 1). While the random effect of sampling period showed negligible variance, random intercepts for bungalow (variance = 2.78, SD = 1.67) accounted for additional variability in baseline counts (Table 1). We observed that as the total number of acoustic detections increases the expected number of individuals also increases (Figure 3). The odds ratio (1.15, p < 0.05) indicates that each additional acoustic detection corresponds to an expected 15% increase in bat abundance, when holding other factors constant (Table 1).
Table 1. Results of the negative binomial GLMM testing the predictive value of the number of acoustic detections to estimate the number of thumbless bats Furipterus horrens.
Figure 3. Number of thumbless bats Furipterus horrens counted during visual surveys by number of acoustic detections collected during the same night and the night before (data points). The blue line shows a negative binomial GLM-based visual approximation that uses the dispersion parameter from the GLMM (random effects not included). The gray line represents the maximum size that has been reported for this bat colony. Square root transformation was applied to both axes.
3.3 Reservation frequency among bungalows
Chi-square tests showed that reservations were evenly distributed among bungalows. We found no significant differences in the frequency of reservations between bungalows, neither in terms of the total number of nights reserved by one or more people (χ² = 2.76, df = 5, p = 0.74) (Figure 1B), nor in terms of the number of people who stayed at least one night (χ² = 5.24, df = 5, p = 0.39) (Figure 1C).
3.4 Effect of tourist load on acoustic activity
Generalized linear mixed-effects models (GLMMs) of the effects of 3-day and 7-day tourist loads on the number of acoustic detections revealed differences in their cumulative impact. We found that a 3-day tourist load had no significant effect on the number of acoustic detections (estimate = -0.30, p = 0.28) (Table 2), indicating that short-term tourist load did not affect bat activity (Figure 4). In contrast, the 7-day tourist load was a significant predictor of the number of acoustic detections (estimate = -0.65, p < 0.01) (Table 2). Model predictions on the response scale, for this week-long cumulative effect, show that higher tourist load is associated with a significantly lower expected number of acoustic detections (Figure 5), where each additional tourist staying in a bungalow for one night predicts a 48% decrease in acoustic detections of F. horrens (odds ratio = 0.52, p < 0.01), when holding other factors constant (Table 2). In both models, random intercepts indicated low variability among sampling periods (3-day load: variance = 0.59, SD = 0.77; 7-day load: variance = 1.05, SD = 1.03), but substantial heterogeneity in baseline response between bungalows (3-day load: variance = 9.50, SD = 3.08; 7-day load: variance = 10.51, SD = 3.24) (Table 2).
Table 2. Results of two negative binomial GLMMs. These models tested the effect of cumulative tourist load (3-day load and 7-day-load, respectively) on the number of acoustic detections of thumbless bats Furipterus horrens.
Figure 4. Number of acoustic detections of thumbless bats Furipterus horrens by a 3-day tourist load at a hotel located in Sarapiqui, Costa Rica. Each panel shows one bungalow (B1-B6). Note differences in y axis scales.
Figure 5. Number of acoustic detections of thumbless bats Furipterus horrens by a 7-day tourist load at a hotel located in Sarapiqui, Costa Rica. Each panel shows one bungalow (B1-B6). The blue line shows the predicted response based on a GLMM with a negative binomial distribution. Note differences in y axis scales.
4 Discussion
We found that reservation load is associated with reduced bat activity, and this reduction in turn is indicative of lower bat abundance. These results confirmed our prediction, showing that greater accumulation of tourist loads correlates with lower levels of bat acoustic activity, when considering one-week periods, and are consistent with the idea that human disturbance may have a negative effect on Furipterus horrens bat colony size. As this is the only known colony of this species using anthropogenic roosts in the region, our results have important implications for their conservation, as well as management plans within the context of tourism and human-wildlife coexistence.
Acoustic monitoring provided a broad temporal coverage of F. horrens behavior and spatial distribution among bungalow roosting sites, capturing acoustic activity peaks and roost-use preferences. The distribution of acoustic detections between bungalows showed much higher activity in two of the bungalows, while essentially avoiding three other structures. This non-random use of bungalow structures reveals roosting preferences by the bat colony. Further work is required to determine if these bungalows are characterized by more favorable microclimates, as has been documented in previous studies (Russo and Ancillotto, 2015). Additionally, structure height has been associated with roost preference in other bat species, such as molossid colonies (Lopez-Baucells et al., 2017). Height might be a relevant feature for our study species, as one of the preferred bungalows, B6, is the tallest of the structures, suggesting it may influence selection here as well. More research is needed to identify possible favorable structure attributes.
Although this species is not commonly observed in anthropogenic or degraded environments, multiple-year persistence in the bungalows of the study site (Alfaro-Lara et al., 2018), and a report of F. horrens bats that have roosted for at least 7 years close to a mayor highway in Brazil (Monteiro et al., 2023), show site fidelity behavior despite being in these types of environments. Furthermore, our results suggest that site fidelity may not only apply towards a general area, but also perhaps to a small spatial scale, such as specific bungalows, but this requires further study.
Peaks of acoustic activity by time of day matched the sunset and sunrise periods, likely reflecting foraging departures and returns, respectively. Similar roost departure and return times have been used to capture F. horrens individuals near cave entrances in Brazil (Bobrowiec et al., 2025). Sunset and sunrise peaks have been described as part of the foraging ecology of multiple species of bats (Johnson et al., 2013, 2019), including Neotropical species such as Myotis nigricans, Molossus ater, and M. molossus (Erkert, 1978; Esbérard and Bergallo, 2010). In some Neotropical species like Peropteryx kappleri, P. macrotis, Pteronotus gymnonotus and Cormura brevirostris, these sunset-sunrise peaks may vary depending on moonlight (Gomes et al., 2020). Additionally, studies on F. horrens have found increased activity during darker nights (Appel et al., 2021), indicating that moonlight may influence its activity patterns. Acoustic activity peaks matching sunset and sunrise periods should therefore be further examined in relation to other environmental covariates when possible.
The absence of detections at any other time of day (between 1800–0500 h) suggests that none of the bungalows is being used through most of the night. One possible interpretation is that F. horrens forages far from this roosting site, returning only until sunrise. A recent radio telemetry study conducted in southeastern Amazonia, Brazil, reported that F. horrens are open space foragers that travel long distances at night (2.2 ± 2.5 km) (Tavares et al., 2025). If the bat colony we studied has similar commuting distances, this could explain why there were no activity records between the observed sunset-sunrise peaks. Alternatively, the F. horrens colony we studied may have other roosting sites besides the bungalow area, which could be used overnight. Caves or fallen logs constitute appropriate natural roosting sites for this species (Simmons and Voss, 1998; Voss et al., 2016; Alfaro-Lara et al., 2018; Bobrowiec et al., 2025). Further studies would be required to determine whether additional roosting sites, long commuting distances, or both, explain the absence of activity among bungalows throughout most of the night.
In addition, the integration of acoustic monitoring and visual surveys showed a significant positive association between acoustic detections and visual counts, confirming the predictive value of acoustic activity to infer bat abundance. Acoustic monitoring has recently provided informative density estimates of Indiana bats Myotis sodalis that responded to temporal variation in bat population size (Hoggatt et al., 2024). Additionally, it has been shown that acoustic recordings of bats flying out of caves can be used to estimate roost bat numbers (Revilla-Martín et al., 2020; Whiting et al., 2022). In this study, we effectively applied acoustic monitoring to predict the number of bats roosting under bungalow sites based on the number of acoustic detections. Since acoustic methods are non-invasive, low-cost, and record spontaneous behaviors without potential effects caused by the presence of observers, we recommend this method in long-term monitoring of F. horrens, and possibly other species roosting under anthropogenic structures.
Nonetheless, it is important to consider that visual surveys were conducted within the same dates, but outside the actual hours sampled by the acoustic recorders. This short lag between the time where the acoustic detections and visual count data were collected, potentially explains part of the variation seen in the predictive model; particularly when a relatively lower number of calls was paired with a higher individual count. Such cases may arise from situations where bats were not present in a given roosting site during the immediately preceding night but arrived by the time the visual survey was conducted, or individuals leaving the roosting site soon after the visual survey. However, our sampling protocol was useful to model the number of bats as a function of the number of calls. When possible, future bat monitoring programs in tourism-influenced habitats should consider integrating both approaches to balance the strengths and limitations of each.
Regarding reservation data, bungalow reservation frequency showed that the structures are being used evenly by the hotel regarding the allocation of guest reservations. The hotel is currently not implementing a management plan or conservation strategy related to the bungalow reservations, within the context of these structures acting as roosting sites for the F. horrens bat colony. Examples of management plans related to bat roosting sites include a study conducted in Yucatan, Mexico, to assess the needs of cave-dwelling species and make recommendations for development (Arita, 1996). Also, the implementation of cave gates to eliminate human disturbance and protect endangered gray bats (Myotis grisescens) in North America (Martin et al., 2003). Future initiatives towards an integral management plan at the bungalow roosting sites should be supported by scientific research to ensure bat colony sustainability, as well as a good understanding of the people who would be involved (e.g. tourist motivations and interests) (Pennisi et al., 2004).
Lastly, we observed that tourist load likely affects bat activity, with significant cumulative effects over one-week periods, but not for 3-day periods, suggesting that continuous human-caused disturbance may change bat behavior and colony spatial distribution, while transient disturbances caused by intermittent tourist reservations may be more tolerable. Higher tourist load can be translated into higher levels of human-caused disturbance. In our study site, as bats roost under bungalows, they are likely to experience different anthropogenic stressors, such as noise, physical impacts on the floor (roost top) and other parts of the structure (e.g. dropping items, slamming the door), as well as artificial lighting. Studies have shown that tourist proximity, noise and artificial lighting have negative impacts on bat behavior (Mann et al., 2002; Cardiff et al., 2012; Russo and Ancilloto, 2015; Stone et al., 2015). In fact, a recent meta-analysis determined that human-caused disturbance was the most frequently reported impact on bats from the use of anthropogenic structures (Sippola et al., 2025). Even though the observed negative association between tourist load and bat activity is strongly suggestive of a causal disturbance effect related to tourism activity, further information regarding other sources of variability, such as environmental covariates, is required to improve our understanding of possible causal mechanisms. Without a management plan that considers the use of bungalows as roosting sites and the cumulative effects of tourist load, the stability of this colony may be compromised under scenarios of increased tourism. Circumstances where all bungalows reach full occupancy for sustained periods, after a business model that exclusively focuses on tourist load, could cause a decline or even disappearance of this unique F. horrens bat colony.
5 Conclusions and recommendations
Our findings showed acoustic monitoring was a reliable indicator of bat abundance when paired to the visual surveys across all bungalows. We also found that bungalows are being used evenly by tourists, even though the colony showed roosting site preferences. The accumulation of tourist load reduces bat activity possibly due to human-caused disturbance, and this reduction is indicative of a lower abundance in the Furipterus horrens bat colony. Therefore, these results underscore the need for a management plan tailored to the bungalow area. Potential measures include limiting occupancy of the bungalows that bats prefer for roosting by prioritizing renting those with the lowest preference, rotating tourist load between available bungalows to distribute human-caused disturbance in a way that prevents negative cumulative effects, and implementing guidelines for guests regarding light and noise reduction over the structures. Additionally, once the features and conditions that make bats choose certain bungalows over others are identified, the construction of a new bat-dedicated structure, replicating these characteristics, constitutes another potential management measure.
This colony provides an invaluable opportunity to further investigate the effectiveness of implementing integrative conservation strategies that combine wildlife management practices with business models within the tourism context. Long-term monitoring coupled with a bat-friendly reservation operational plan can later be used to evaluate the effectiveness of the respective conservation strategies. Even though such a strategy would operate on a relatively small scale, it should have the potential to inform (and be replicated) in further instances with comparable cases of human-wildlife coexistence.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements. Ethical approval was not required for the study involving animals in accordance with the local legislation and institutional requirements because the study was non-invasive and data collection was purely observational.
Author contributions
DG-V: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Writing – original draft, Writing – review & editing. HA-L: Conceptualization, Investigation, Methodology, Resources, Writing – review & editing. LV-C: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Writing – review & editing.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Acknowledgments
We thank the staff at Selva Verde Lodge for logistic support, with special thanks to Gabriel González for his valuable assistance regarding reservation data access. Vicerrectoría de Investigación-UNED provided equipment, software and transportation. Additional thanks to Alex García and Mariela Valverde for their assistance during field work.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Alfaro-Lara H., Villalobos-Chaves D., Castillo-Gómez J. I., Zuñiga-Montero A. F., and Pineda-Lizano W. (2018). Rediscovery of Furipterus horrens (Chiroptera: Furipteridae) in Costa Rica. Mammalia. 83, pp.352–pp.356. doi: 10.1515/mammalia-2018-0066
Appel G., da Rocha P. A., Cerqueira P. V., Benethar T. C. M., Gianniani T. C., Pereira M. S. R., et al. (2025). Using active and passive methods to monitor cave-dwelling bats in Amazonian caves. Biodiversity Conserv. 34, 4429–4449. doi: 10.1007/s10531-025-03160-4
Appel G., López-Baucells A., Rocha R., Meyer C. F. J., and Bobrowiec P. E. D. (2021). Habitat disturbance trumps moonlight effects on the activity of tropical insectivorous bats. Anim. Conserv. 24, 1046–1058. doi: 10.1111/acv.12706
Arita H. T. (1996). The conservation of cave-roosting bats in Yucatan, Mexico. Biol. Conserv. 76, 176–185. doi: 10.1016/0006-3207(95)00105-0
Arlettaz R., Godat S., and Meyer H. (2000). Competition for food by expanding pipistrelle bat populations (Pipistrellus pipistrellus) might contribute to the decline of lesser horseshoe bats (Rhinolophus hipposideros). Biol. Conserv. 93, .55–.60. doi: 10.1016/s0006-3207(99)00112-3
Arroyo-Solórzano M. and Rojas-Prendas L. (2021). Potencial ecoturístico en Áreas Naturales Protegidas. Caso: Reserva Biológica Tirimbina, Sarapiquí, Provincia de Heredia, Costa Rica. Rev. Geográfica América Cent. 66, 289–312. doi: 10.15359/rgac.66-1.11
Atagana P. J., Fils E. M. B., and Kekeunou S. (2021). Responses of bat communities (Mammalia: chiroptera) to forest loss and habitat conversion in southern Cameroon. Trop. Conserv. Sci. 14, 1–13. doi: 10.1177/19400829211010360
Bateman P. W. and Fleming P. A. (2017). Are negative effects of tourist activities on wildlife over-reported? A review of assessment methods and empirical results. Biol. Conserv. 211, 10–19. doi: 10.1016/j.biocon.2017.05.003
Bobrowiec P. E. D., Trevelin L. C., Benathar T. C. M., Appel G., Prous X., Ribeiro M. S., et al. (2025). Reproductive phenology and sex ratio variation of the thumbless bat Furipterus horrens (Cuvier 1828) (Furipteridae). Acta Amazonica. 55, 1–10. doi: 10.1590/1809-4392202403972
Cardiff S. G., Ratrimomanarivo F. H., and Goodman S. M. (2012). The Effect of Tourist Visits on the Behavior of Rousettus madagascariensis (Chiroptera: Pteropodidae) in the Caves of Ankarana, Northern Madagascar. Acta Chiropterologica. 14, 479–490. doi:10.3161/150811012x661783
Chaves V. (2024). Relationship between human disturbance and bat species diversity in urban green spaces in Costa Rica. Int. J. Biol. 4, 23–34. doi: 10.47604/ijb.2804
Deleva S. and Chaverri G. (2018). Diversity and conservation of cave-dwelling bats in the brunca region of Costa Rica. Diversity. 10, 43. doi: 10.3390/d10020043
Erkert H. G. (1978). Sunset-related timing of flight activity in neotropical bats. Oecologia. 37, 59–67. doi: 10.1007/bf00349991
Esbérard C. E. L. and Bergallo H. G. (2010). Foraging activity of the free-tailed bat Molossus molossus (Chiroptera; Molossidae) in southeastern Brazil. Braz. J. Biol. 70, 1011–1014. doi: 10.1590/s1519-69842010000500013
Falcão F., Ugarte-Núñez J. A., Faria D., and Caselli C. B. (2015). Unravelling the calls of discrete hunters: acoustic structure of echolocation calls of furipterid bats (Chiroptera, Furipteridae). Bioacoustics. 24, 175–183. doi: 10.1080/09524622.2015.1017840
Ferreira G. B., Newbold T., Oliveira R., Pringle H., Pinheiro M. S., de F., et al. (2022). Limited temporal response of Cerrado mammals to anthropogenic pressure in areas under distinct levels of protection. J. Zoology. 317, 43–55. doi: 10.1111/jzo.12958
Fetene A., Yeshitela K., and Gebremariam E. (2019). The effects of anthropogenic landscape change on the abundance and habitat use of terrestrial large mammals of Nech Sar National Park. Environ. Syst. Res. 8, 19. doi: 10.1186/s40068-019-0147-z
Fonda F., Vezzani L., Mena Aguilar L.Á., Venegas-Elizondo C. A., Bolaños Brenes A., Lopez D. M., et al. (2025). Mammal diversity and tourism influence in an under-investigated region of Costa Rica. ZooKeys. 1260, 37–74. doi: 10.3897/zookeys.1260.128800
Frank T. M., Gabbert W. C., Chaves-Campos J., and LaVal R. K. (2018). Impact of artificial lights on foraging of insectivorous bats in a Costa Rican cloud forest. J. Trop. Ecol. 35, 8–17. doi: 10.1017/s0266467418000408
García-Sánchez M. and González-Chaverri P. (2022). The Tirimbina Biological Reserve, in Costa Rica: model in the conservation of the tropical rain forest, scientific research, ecotourism and environmental education. Rev. Cienc. Ambientales. 56, 242–253. doi: 10.15359/rca.56/2.13
Gomes D. G. E., Appel G., and Barber J. R. (2020). Time of night and moonlight structure vertical space use by insectivorous bats in a Neotropical rainforest: an acoustic monitoring study. PeerJ. 8, p.e10591. doi: 10.7717/peerj.10591
Hill A. P., Prince P., Piña Covarrubias E., Doncaster C. P., Snaddon J. L., and Rogers A. (2018). AudioMoth: Evaluation of a smart open acoustic device for monitoring biodiversity and the environment. Methods Ecol. Evol. 9, 1199–1211. doi: 10.1111/2041-210X.12955
Hoggatt M. L., Starbuck C. A., and O’Keefe J. M. (2024). Acoustic monitoring yields informative bat population density estimates. Ecol. Evol. 14, e11051. doi: 10.1002/ece3.11051
Jackson R. T., Webala P. W., Ogola J. G., Lunn T. J., and Forbes K. M. (2023). Roost selection by synanthropic bats in rural Kenya: implications for human–wildlife conflict and zoonotic pathogen spillover. R. Soc. Open Sci. 10, 230578. doi: 10.1098/rsos.230578
Johnson J. B., Rodriguez J. L., and Ford W. M. (2013). “Nightly and yearly bat activity before and after white-nose syndrome on the Fernow experimental forest in West Virginia,” in USDA forest service research paper NRS-24. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. Available online at: https://www.fs.usda.gov/nrs/pubs/rp/rp_nrs24.pdf. (Accessed October 4, 2025)
Johnson J. S., Treanor J. J., Slusher A. C., and Lacki M. J. (2019). Buildings provide vital habitat for little brown myotis (Myotis lucifugus) in a high-elevation landscape. Ecosphere. 10, e02925. doi: 10.1002/ecs2.2925
Jones G. and Spadafora A. (2017). Creating ecotourism in Costa Rica 1970–2000. Enterprise Soc. 18, pp.146–pp.183. doi: 10.1017/eso.2016.50
Jung K. and Threlfall C. G. (2015). “Urbanization and its effects on bats—a global meta-analysis,” in Bats in the anthropocene: conservation of bats in a changing world. Eds. Voigt C. C. and Kingston T. (Springer, Cham). doi: 10.1007/978-3-319-25220-9_2
Kunz T. H., Braun de Torrez E., Bauer D., Lobova T., and Fleming T. H. (2011). Ecosystem services provided by bats. Ann. New York Acad. Sci. 1223, pp.1–pp38. doi: 10.1111/j.1749-6632.2011.06004.x
LaVal R. K. (1977). Notes on some Costa Rican bats. Brenesia. 10-11, 77–83. Available online at: https://biblioteca.museocostarica.go.cr/articulo.aspx?id=2925 (Accessed October 4, 2025).
Lopez-Baucells A., Rocha R., Andriatafika Z., Tojosoa T., Kemp J., Forbes K. M., et al. (2017). Roost selection by synanthropic bats in rural Madagascar: what makes non-traditional structures so tempting? Hystrix Ital. J. Mammalogy. 28, pp.28–pp.35. doi: 10.4404/hystrix-28.1-12046
Lučan R. K., Jor T., Romportl D., and Morelli F. (2024). Use of synanthropic roosts by bats in Europe and North America. Mammal Rev. 55, e12380. doi: 10.1111/mam.12380
MacKinnon J. L., Hawkins C. E., and Racey P. A. (2003). “Pteropodidae, fruit bats, fanihy, angavo,” in The natural history of Madagascar. Ed. Goodman S.M. & Benstead J. P. (University of Chicago Press, Chicago), 12991302.
Mann S. L., Steidl R. J., and Dalton V. M. (2002). Effects of cave tours on breeding myotis velifer. J. Wildlife Manage. 66, p.618. doi: 10.2307/3803128
Martin K. W., Leslie D. M., Payton M. E., Puckette W. L., and Hensley S. L. (2003). Internal cave gating for protection of colonies of the endangered gray bat (Myotis grisescens). Acta Chiropterologica. 5, pp.143–pp.150. doi: 10.3161/001.005.0112
Mayta C., López C. L., Villegas M., Aguirre L. F., Hensen I., and Gallegos S. C. (2024). Bird perches and artificial bat roosts increase seed rain and seedling establishment in tropical bracken-dominated deforested areas. Restor. Ecol. 32, e14197. doi: 10.1111/rec.14197
Monteiro J. C. S., Oliveira A. C. M., Andrade F. A. G., Sousa L. S., Lima R. S., Cruz dos Reis S. M., et al. (2023). New record of Furipterus horrens (Cuvier 1828) (Chiroptera, Furipteridae) in eastern Brazilian Amazonia. Check List. 20, 29–39. doi: 10.15560/20.1.29
Moras L. M., Nobre C. C., Vasconcelos S., Garbino G. S. T., de Souza É.M.S., Santos F. R., et al. (2024). Almost two hundred years of monotypy and a poorly known life history: multiple lineages of Furipterus (Chiroptera: Furipteridae) were hidden by rocks. Mamm. Biol. 105, pp.153–pp.166. doi: 10.1007/s42991-024-00465-9
Moya-Calderón M., Carrillo-Jiménez E., Ugalde-Alvarado G., Ramírez-Ledezma A., and Barriocanal-Lozano C. (2025). The study of tourism in Costa Rica from the perspective of Recreational Ecology. UNED Res. J. 17, e5548. doi: 10.22458/urj.v17i1.5548
Murphy G. E. P. and Romanuk T. N. (2013). A meta-analysis of declines in local species richness from human disturbances. Ecol. Evol. 4, pp.91–pp103. doi: 10.1002/ece3.909
OECD (2020). Costa rica: OECD tourism trends and policies 2020. (Paris, France: OECD Publishing). Available online at: https://www.oecd.org/en/publications/oecd-tourism-trends-and-policies-2020_6b47b985-en/full-report/component-49.html (Accessed October 4, 2025).
Pennisi L. A., Holland S. M., and Stein T. V. (2004). Achieving bat conservation through tourism. J. Ecotourism 3, 195–207. doi: 10.1080/14664200508668432
Ramírez-Fernández J. D., Sánchez R., May-Collado L. J., González-Maya J. F., and Rodríguez-Herrera B. (2023). Revised checklist and conservation status of the mammals of Costa Rica. Therya. 14, 233–244. doi: 10.12933/therya-23-2142
Reid J. L., Mendenhall C. D., Zahawi R. A., and Holl K. D. (2015). Scale-dependent effects of forest restoration on Neotropical fruit bats. Restor. Ecol. 23, 681–689. doi: 10.1111/rec.12235
Revilla-Martín N., Budinski I., Puig-Montserrat X., Flaquer C., and López-Baucells A. (2020). Monitoring cave-dwelling bats using remote passive acoustic detectors: a new approach for cave monitoring. Bioacoustics. 30, 527–542. doi: 10.1080/09524622.2020.1816492
Rodriguez-Segovia M. A. and Carrera-E J. P. (2025). First record of Furipterus horrens on the western slopes of Pichincha Province, with observations on species development and distribution. Mammalia Aequatorialis. 7, pp.51–pp.65. doi: 10.59763/mam.aeq.v7i2.127
Russo D. and Ancillotto L. (2015). Sensitivity of bats to urbanization: a review. Mamm. Biol. 80, 205–212. doi: 10.1016/j.mambio.2014.10.003
Rydell J. (1992). Exploitation of insects around streetlamps by bats in Sweden. Funct. Ecol. 6, 744–750. doi: 10.5281/zenodo.14816954
Simmons N. B. and Voss R. S. (1998). Furipterus horrens. In: The mammals of Paracou, French Guiana, a Neotropical lowland rainforest fauna. Part 1, Bats. Bull. Am. Museum Natural History. 237, 1–219. doi: 10.5281/zenodo.4545052
Sippola E. A., Johnson J. S., Mammola S., Apoznański G., Ilze Brila I., Latapiat I. F., et al. (2025). Impacts of bat use of anthropogenic structures on bats and humans. Conserv. Biol., e70037. doi: 10.1111/cobi.70037
Spoelstra K., Ramakers J. J., van Dis N. E., and Visser M. E. (2018). No effect of artificial light of different colors on commuting Daubenton’s bats (Myotis daubentonii) in a choice experiment. J. Exp. Zoology Part A: Ecol. Integr. Physiol. 329, 506–510. doi: 10.1002/jez.2178
Stone E. L., Harris S., and Jones G. (2015). Impacts of artificial lighting on bats: a review of challenges and solutions. Mamm. Biol. 80, pp.213–pp.219. doi: 10.1016/j.mambio.2015.02.004
Tavares V. C., Ribeiro M. S., Prous X., Notini A. A., Kaku-Oliveira N. Y., Maciel L. M. D., et al. (2025). Do you have enough space? Habitat selection of insectivorous cave-dwelling bats in fragmented landscapes of Eastern Amazon. PloS One. 20, e0296137. doi: 10.1371/journal.pone.0296137
Voigt C. C. and Kingston T. (2016). Bats in the anthropocene: conservation of bats in a changing world (Cham: Springer). doi: 10.1007/978-3-319-25220-9
Voss R. S., Fleck D. W., Strauss R. E., Velazco P. M., and Simmons N. B. (2016). Roosting ecology of amazonian bats: evidence for guild structure in hyperdiverse mammalian communities. Am. Museum Novitates. 3870, 1–43. doi: 10.1206/3870.1
Whiting J. C., Doering B., and Aho K. (2022). Can acoustic recordings of cave-exiting bats in winter estimate bat abundance in hibernacula? Ecol. Indic. 137, 108755. doi: 10.1016/j.ecolind.2022.108755
Keywords: acoustic monitoring, anthropogenic roosts, human-caused disturbance, neotropical bats, thumbless bats, tourism, coexistence
Citation: García-Valverde D, Alfaro-Lara H and Vargas-Castro LE (2026) Roosting under pressure: cumulative tourist load negatively impacts colony activity in the rare bat Furipterus horrens. Front. Ethol. 5:1718968. doi: 10.3389/fetho.2026.1718968
Received: 05 October 2025; Accepted: 02 January 2026; Revised: 15 December 2025;
Published: 29 January 2026.
Edited by:
Emanuela Prato Previde, University of Milan, ItalyReviewed by:
Jesús Molinari, University of Los Andes (Venezuela), VenezuelaGiulliana Appel, Vale Technological Institute (ITV), Brazil
Copyright © 2026 García-Valverde, Alfaro-Lara and Vargas-Castro. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Luis Esteban Vargas-Castro, bHV2YXJnYXNAdW5lZC5hYy5jcg==