Hibernacula of bats in Mexico, the southernmost records of hibernation in North America

Abstract Although Mexico holds the southernmost hibernating bats in North America, information on winter behavior and hibernacula microclimate use of temperate Mexican bats is limited. We studied hibernating bats at high altitudes (>1,000 m a.s.l.) in northern and central Mexico during 5 consecutive winters. Our aims were to document and describe the hibernacula, winter behavior (such as abundance and roost pattern), and microclimates (estimated as adjacent substrate temperature) of cave-hibernating bats in Mexico. We found 78 hibernacula and 6,089 torpid bats of 10 vespertilionid species, increasing by over 50% the number of cave-hibernating bat species and quadrupling the number of hibernacula for Mexico. Hibernacula were at altitudes between 1,049 and 3,633 m a.s.l., located in 3 mountain ranges, mainly in oak and conifer forests. Myotis velifer was the most common species, followed by Corynorhinus townsendii and C. mexicanus. We recorded the adjacent substrate temperatures from 9 species totaling 1,106 torpid bats and found differences in microclimate use among the 3 most common species. In general, abundance of torpid bats in our region of study was similar to those in the western United States, with aggregations of tens to a few hundred individuals per cave, and was lower than in the eastern United States where a cave may hold thousands of individuals. Knowledge of bat hibernation is crucial for developing conservation and management strategies on current conditions while accommodating environmental changes and other threats such as emerging diseases.

During cold periods or food shortages, temperate vespertilionid bats in North America can exhibit torpor (Brack 2007;Weller et al. 2018;Geiser 2021), which is a marked but controlled reduction of corporal temperature, metabolic rate, and other biological functions to conserve energy (Boyles et al. 2006;Altringham 2011;Geiser 2021).Depending on duration, torpor is typically labeled as daily torpor when it lasts less than 24 h or as hibernation when it lasts an extended period of days, weeks, or months (usually during winter), and requires utilization of stored body fat (Altringham 2011;Geiser 2021).Likewise, hibernating bats usually reduce minimum body temperatures to between 0 °C and 5 °C, but during daily torpor to between 17 °C and 26 °C (Geiser 2021).Hibernating bats typically use microclimates with high humidity and cold (above-freezing) stable temperatures (Altringham 2011;Perry 2013), but such conditions can vary among and within species along temporal and geographical gradients (Brack 2007;Altringham 2011;Geiser 2021).Factors that influence microsite selection operate at different levels and scales including latitude, altitude, vegetation, and hibernacula morphology (Dunbar and Brigham 2010;Perry 2013;Piksa et al. 2013;Meierhofer et al. 2019a), but low ambient temperatures are the main criteria for hibernation in temperate regions (Altringham 2011;Geiser 2021).The temperature of the substrate adjacent to torpid bats has been widely used to characterize microclimates (Brack 2007;Kim et al. 2013;Hopkins et al. 2021).
Although other vespertilionid species in Mexico also hibernate in the United States and Canada (Perkins et al. 1990;Boyles et al. 2006;Brack 2007;Hendricks 2012;Haase et al. 2020), information on winter behavior and microclimates use by hibernating bats in temperate Mexico is limited (Ávila-Flores 2000; Ayala-Berdon and Solís-Cárdenas 2017; Marín et al. 2021).Likewise, data on hibernaculum locations are missing.In contrast to western United States where more than 1,200 hibernacula are known (Weller et al. 2018), Mexico has only 24 records (i.e., Villa 1966;López-González and Torres-Morales 2004).Therefore, clarifying which species hibernate and how abundant they are, where they hibernate, and which microclimates they use is crucial for developing conservation and management strategies.
Populations of some species of cave-hibernating bats in North America have declined dramatically since the advent of the whitenose syndrome disease (hereafter WNS), caused by the fungal pathogen Pseudogymnoascus destructans (hereafter Pd; Blehert et al. 2009;Cheng et al. 2021;WNSRT 2022).The fungus has spread throughout much of the United States, and in April 2022, it was found in caves in Texas, on the border with Mexico (WNSRT 2022).In addition, bats with WNS have been reported in several Texas counties less than 100 km from Mexico (WNSRT 2022).A recent study on the survival and viability of Pd conidia suggested that it can persist under elevated temperatures, facilitating long-distance dispersal in warmer conditions (Campbell et al. 2020).Furthermore, while suitable climatic conditions for WNS have been modeled in Mexico (Sierra-Durán 2020;Gómez-Rodríguez et al. 2022) there currently is no evidence that the fungus or the disease are present in Mexico.However, considering the proximity of infected hibernacula in the United States, the entry of Pd in Mexico is imminent, while potential impacts of the disease on bats remain uncertain.Shorter cold winter conditions with milder temperatures, possible availability of insects, and shorter periods of torpor of bats at subtropical latitudes might attenuate the impact of WNS (Bernard and McCracken 2017;Holz et al. 2019;Meierhofer et al. 2019a) on Mexican bat populations.Given that Mexico holds the southernmost hibernating bats in North America, a better understanding of the use of hibernation by Mexican bat populations must be considered in WNS response efforts.
For this work, we studied hibernating bats in northern and central Mexico, obtaining data on: species richness and abundance; adjacent substrate temperature of torpid bats; and the relationship between hibernation and fur temperature.Our aims were to document and describe features of hibernacula including winter behavior (such as abundance and roost pattern), and microclimates (estimated as adjacent substrate temperature) of cave-hibernating vespertilionid bats in temperate Mexico.

Study area.
We visited previously reported hibernacula and offered small rewards to colleagues, landowners, and people from rural communities for information that directed us to any new hibernacula.We explored 155 caves in 21 sites at high altitudes (>1,000 m a.s.l.) in the Sierra Madre Occidental (SMOc), the Sierra Madre Oriental (SMOr), and the Trans-Mexican Volcanic Belt (TMVB) mountain ranges.These sites were located in 11 states in central and northern Mexico, and each site includes 1 or more caves that are not more than 10 km apart.The SMOc runs north to south, 31° to 21°N, parallel to the Pacific coast, with 37.4% of its territory above 2,000 m a.s.l. and is the southern extension of the Rocky Mountains in the United States (INEGI 2001;Ferrrusquía-Villafranca et al. 2005;González-Elizondo et al. 2012).The SMOr runs north to south, 30° to 20°N, parallel to the Gulf of Mexico coast, with 10.5% of its territory above 2,000 m a.s.l.(Challenger 1998;INEGI 2001;Ferrrusquía-Villafranca et al. 2005).In central Mexico, where the SMOc and SMOr end, the TMVB crosses east-west (coast-to-coast) as far south as 18°N.The TMVB has the highest mountain peaks in Mexico (>5,000 m a.s.l.), with 44.4% of its territory over 2,000 m a.s.l.(Challenger 1998;INEGI 2001).It is important to note that we only explored sites in the northern part of the SMOc and that although we comparatively explored more sites along the SMOr and TMVB, there are still many mountainous areas to cover.The visited caves were in conifer forest (n = 38), oak forest (n = 48), xerophytic scrub (n = 61), and secondary vegetation modified from conifer forest (n = 8).
Our fieldwork included 5 consecutive periods in the coldest months during autumn and winter, hereafter considered as winter: (i) January 2018; (ii) January to March 2019; (iii) September 2019 to March 2020; (iv) September 2020 to March 2021; and (v) October 2021 to March 2022.Each winter we added between 1 and 7 sites to our evaluation.We visited each site once during the first 2 winters.In subsequent years some sites in the TMVB were visited 2 to 5 times each winter.On particular occasions we visited some sites for 2 consecutive days.

Data collection.
We recorded the cave coordinates and altitude using a GPS (Garmin Etrex 30x), and the surrounding vegetation type according to Rzedowski's (1978) classification, which was conifer forest, oak forest, xerophytic scrub, or secondary vegetation.We also measured the length of caves using a laser distance device (Bosch GLL30).
We recorded species and abundance of torpid bats at each visit visually and with digital photographs (Meretsky et al. 2010;Loeb et al. 2015).We determined roost pattern of bats as solitarily or in cluster (≥2 bats in direct contact).Some individuals were handled and identified as species using the field guides of Medellín et al. (2008) and Morgan et al. (2019).When species identification was impossible because bats were unreachable or in order to avoid handling during the pandemic caused by SARS-CoV-2, we identified individuals to genus level.We identified bats as torpid when they were motionless and cold, with ears rolled into a rams-horns position, and/or with condensed water droplets on fur.To document specific microclimates used by torpid bats, we measured adjacent substrate temperature (T sub ) and bat fur surface temperature (T fur ) using 2 infrared thermometers (Extech Instruments, Models IR400 and 42545, Nashua, New Hampshire; distance to spot size 50:1 and 8:1, detection range −20 °C to 332 °C, accuracy ± [2% of reading + 2 °C], resolution of 0.1 °C, operating temperature 0 °C to 50 °C) to minimize time and disturbance inside the hibernacula.We took temperature measurements between 7:00 and 18:00 h.T sub was taken between 1 and 3 cm from the bat, and T fur was taken from the back of each bat, both measurements at a maximum distance of 20 cm to ensure a diameter < 2.5 cm detection area for the thermometers.When bats were in clusters, we measured temperatures of 2 to 5 individuals located in the periphery per each cluster (the largest cluster surveyed was 200 bats).
We followed the USGS National Wildlife Health Center decontamination protocol for WNS (NWHC 2016) and the Laboratorio de Ecología y Conservación de Vertebrados Terrestres field protocol for minimizing the spread of Pd fungus and avoid disturbing hibernating bats.We used gloves and decontaminated both clothing and equipment with ethanol (70%) before and after surveying a roost.We also changed clothes between roosts located in different sites and carried out our fieldwork in central Mexico toward sites progressively closer to those with Pd confirmed, traveling from southern latitudes to northern latitudes.During the last 2 winters (between September 2020 and March 2022), we also followed recommendations of the International Union for Conservation of Nature Bat Specialist Group (Kingston et al. 2021) to reduce the risk of already low probabilities of transmission of SARS-CoV-2 from humans to bats.Fieldwork was done under a scientific collecting permit issued by the Dirección General de Vida Silvestre number SGPA/DGVS/08072/21 and was performed following Sikes et al. (2016).To avoid disturbing torpid bats, we reduced team size, working time, noise and light working levels, and bat handling during surveys (Loeb et al. 2015).Throughout our study we have collaborated with 25 organizations including ejidos, agricultural communities, universities, private and ecotourism companies, civil associations, and state and federal agencies in Mexico and the United States.

Data analyses.
As a part of the documentation of hibernacula, a Chi-square test was calculated to determine whether the frequencies of hibernacula and non-hibernacula differed among vegetation types.Given that most residuals of our data on T sub and T fur did not show normal distribution or variance homogeneity, we performed nonparametric tests using the median as the central tendency measure.We calculated median values of T sub and T fur of each bat species based on the measurement of all individuals of that species (approximation per individual).Likewise, we used a Spearman rank correlation to determine how T fur varied with T sub based on temperature data of all bats recorded.To compare T sub used by different species, we used data per species per hibernaculum during each visit (approximation per survey) to avoid pseudoreplication (Meierhofer et al. 2019a).The comparison was performed for those species with at least 25 surveys, using the Kruskal-Wallis test followed by a Dunn test with a Bonferroni adjustment of α/3.The level of statistical significance considered was below α = 0.05.For all analyses we used R 4.0.2(R Core Team 2021).
We recorded 6,089 torpid bats of at least 10 vespertilionid species (439 individuals were identified only to genus level).Of the 5,650 bats identified to species (92.8% of total torpid bats), M. velifer was the most abundant species in our study with 5,167 individuals.Myotis velifer were distributed in 27 hibernacula, from which 16 hibernacula were in conifer forests and 26 in TMVB (Table 2).Considering bat surveys when M. velifer was found, its maximum numbers per hibernaculum were typically of tens to a few thousand bats (range = 1 to 700 individuals, median = 9).Torpid bats were observed between late September and early March, with the largest For each hibernaculum we include the maximum abundance of individuals per bat species recorded between 1 and 10 visits along this study.1.  2), and with maximum numbers usually less than 10 bats per hibernacula (range = 1 to 24 individuals, median = 2), mainly hibernating solitarily.We recorded T sub and T fur of 1,106 vespertilionid bats that were identified to species (18.6% of total torpid bats) during 90 surveys across 4 winters between January 2019 and March 2022 (Table 4).We obtained T sub and T fur medians from 9 species of 4 genera.The median of the absolute difference between T fur and T sub was 0.2 °C (Q1 = 0.1, Q3 = 0.5).Only the 3 most abundant species (M.velifer, C. mexicanus, and C. townsendii) had enough surveys to be compared.Temperature measures were recorded for 863 M. velifer in 24 hibernacula visited from 1 to 7 times (60 surveys), 89 C. mexicanus in 17 hibernacula visited from 1 to 5 times (25 surveys), and 80 C. townsendii in 25 hibernacula visited 1 or 2 times (29 surveys).The other 6 bat species were found ≤5 times each and temperature data were obtained from 2 to 26 individuals (Table 4).We found a strong and positive correlation between T sub and T fur (rho = 0.96, P < 0.001; Fig. 2).All bats were hibernating at T sub between −6.8 °C and 18.6 °C, and at T fur between −7.3 °C and 20.2 °C, where more than 95% of measured individuals had T fur < 12 °C.Corynorhinus townsendii was found using higher T sub than M. velifer (D = 4.41, df = 2, P < 0.001) and C. mexicanus (D = −3.6,P < 0.001; Fig. 3).

Discussion
Although most of our visits to hibernacula covered only 1 day of evaluation, we considered that observed torpid bats during our study were likely hibernating, based on multiple lines of evidence.Considering that skin temperature may be a proxy of body temperature, and that skin and fur temperatures are relatively similar at low body temperatures (Bartonička et al. 2017;Geiser 2021), the T fur of torpid bats indicated hibernation rather than daily torpor given that they were very low (T fur < 12 °C).On some occasions we visited a cave for 2 consecutive days and we were able to record torpid bats of M. velifer, C. mexicanus, and P. subflavus in the same place inside the cave (similar to observations of Hall and Dalquest 1963).Camera traps placed in front of torpid M. velifer in hibernacula in central Mexico recorded them as inactive during several consecutive days between late September and early March (Ramos-H. et al. 2022).Finally, all torpid bat species we found have been previously reported hibernating in Mexico (Hall and Dalquest 1963;Villa 1966;López-González and Torres-Morales 2004;Aguilar-Rodríguez et al. 2021) and in the United States (Perkins et al. 1990;Brack 2007;Hendricks 2012;Meierhofer et al. 2019a;Haase et al. 2020).For M. occultus we are considering reports on its sister species M. lucifugus (Piaggio et al. 2002).
Therefore, here we provide relevant information on geographic locations of hibernacula, winter behavior, and specific microclimates for 10 temperate vespertilionid bat species in the southernmost occurrence bat hibernation in North America.We report the first hibernation record in Mexico for E. fuscus, M. ciliolabrum, M. occultus, and M. yumanensis.We also add 70 new hibernacula to those 24 previously known for the country.Thus, in 5 consecutive evaluated winters we have increased by over 50% the number of cave-hibernating bat species and have quadrupled the number of hibernacula for Mexico.
In our study, hibernacula and large numbers of torpid bats were more commonly found in caves in oak and conifer forests.These forests are usually located at higher altitudes in temperate mountain areas of Mexico, where comparatively higher humidity and lower environmental temperature are found (Rzedowski 1978;Ferrusquía-Villafranca et al. 2005).Likewise, it has been reported that bats in temperate regions, including M. velifer in central Mexico (Villa 1966;Camacho 2004), select caves at higher altitudes to hibernate because of their optimal microclimates (McGuire and Boyle 2013).Both vegetation type and internal cave temperature are strongly linked with external environmental temperature (Geiger et al. 1995;Perry 2013;Meierhofer et al. 2019a).Therefore, our observations suggest that conifer and oak forests in temperate Mexico would occur in the same altitudinal levels as caves used by bats to hibernate.It is important to note that we found more hibernacula located at higher altitudes in central Mexico compared to northern latitudes, which may be associated with decrease in environmental temperature with latitude (Montgomery 2006).Thus, vegetation types as well as thermal cave conditions may occur at different altitudes according to latitude.It has been reported that forest cover generates more stable temperatures and humidity at ground than those in open areas located at the same altitudinal level (Geiger et al. 1995); thus, vegetation may influence cave temperature (Perry Fig. 2. Relationship between fur and adjacent substrate temperatures for all cave-hibernating bats found in northern and central Mexico. 2013; Piksa et al. 2013).more studies about association of vegetation with hibernacula occurrence along latitudinal gradients are necessary.Large caves usually hold a higher richness and abundance of bats because they may provide a wide range of microclimates (Arita 1993;Briggler and Prather 2003;Altringham 2011;Perry 2013).However, we have not found this trend in our hibernation caves; higher numbers of species or individuals were in caves with different lengths.Although this may highlight the importance of caves, regardless of their length for hibernating Mexican bats, other studies that include additional large caves are needed.On the other hand, a third of the hibernacula we report are human-made caves (abandoned mines and culverts), reflecting the capacity of bats to take advantage of human-generated structures.
Although M. velifer has a wide distribution in Mexico (Castro-Campillo et al. 2014) and was reported hibernating in the United States (Humphrey and Oli 2015;Caire et al. 2018;Meierhofer et al. 2019a;Haase et al. 2020), in northern Mexico we only found 1 torpid individual in a single cave.With this exception, all hibernacula of M. velifer were in the TMVB, mainly located in conifer forests and with numbers of tens to hundreds of bats per cave.This species previously has been reported hibernating only in central Mexico (Villa 1966;López-Wilchis 1999;Ávila-Flores 2000;Ayala-Berdon and Solís-Cárdenas 2017), and regardless of its abundance in our study, its numbers were much lower than those in the southern United States, where several hibernacula hold thousands of bats (Humphrey and Oli 2015;Caire et al. 2018).
Similar to M. velifer, C. mexicanus was found mainly hibernating in caves in conifer forests across the TMVB.Moreover, we found both species frequently sharing hibernacula, as previously reported only in central Mexico (Hall and Dalquest 1963;López-Wilchis 1999;Ávila-Flores 2000;Aguilar-Rodríguez et al. 2021), and hibernating at very similar T sub .Our findings suggest that both species may use similar microclimates for hibernation in central Mexico, as reported for bats wintering together in other regions (Nagy and Postawa 2011;Piksa et al. 2013)-raising the question of why there is a lack of records of torpid M. velifer in northern Mexico, even when we found C. mexicanus hibernating in all 3 Mexican mountain ranges we explored.
Corynorhinus townsendii was the most ubiquitous species found in our study and had numbers typically less than 10 bats per hibernaculum.These observations are in accordance with data provided by other studies in the western United States (Hendricks 2012;Weller et al. 2018;Whiting et al. 2018), although those documented a few thousand individuals in some caves.Unlike M. velifer and C. mexicanus, C. townsendii seems more flexible in hibernacula use, given that we found torpid bats in several caves in all 4 vegetation types across the 3 mountain ranges.Although we occasionally found individuals of both Corynorhinus species sharing hibernacula-similar to previous studies (López-González and Torres-Morales 2004;Gómez-Ruiz et al. 2006)-we found C. townsendii more frequently in caves in xerophytic scrub vegetation, while C. mexicanus were in caves in conifer forest.Different habitat affinities have also been described for both species throughout the year, as C. townsendii commonly inhabits lower altitudes and arid environments, while C. mexicanus inhabits higher altitudes with cooler and more humid environments (Handley 1959).Likewise, we observed C. townsendii hibernating at higher T sub than C. mexicanus, suggesting that these species have different hibernacula microclimate preferences.Related bat species with morphological similarities may show differences in the use of hibernacula and microsites (Piksa et al. 2013), which may be associated with their summer habitat use (Nagy and Postawa 2011;Meierhofer et al. 2019a).
We found P. subflavus in hibernacula last visited and reported 30 to 70 years ago (Davis 1959;Hall and Dalquest 1963;Vargas 1998) and we recorded new hibernacula.Our numbers of P. subflavus seemed to be similar to those reported in caves by Briggler and Prather (2003) and be lower than the abundance noted by  Meierhofer al. (2019b), who hundreds and even thousands of bats in culverts.We also added new hibernacula for M. thysanodes and M. volans to those reported previously (Ávila-Flores 2000;Aguilar-Rodríguez et al. 2021).Although there are no previous records of torpid E. fuscus in Mexico, its hibernation was suggested by Woloszyn and Woloszyn (1982) and Segura (2010) based on significant fat accumulation in bats collected at the beginning of the winter.Although E. fuscus, M. ciliolabrum, M. occultus, M. thysanodes, M. volans, and M. yumanensis have wide distributions in central and northern Mexico (Medellín et al. 2008;Ceballos 2014), we found relatively few torpid individuals and hibernacula for these species.Like our data, studies in the western United States reported that E. fuscus and the other 5 Myotis species hibernate in small numbers, typically less than 10 bats per cave (Perkins et al. 1990;Hendricks et al. 2012;Whiting et al. 2018).
In this study, Corynorhinus species were usually less abundant than Myotis species, which contrasts with Weller et al. (2018) and Whiting et al. (2018) who reported larger numbers of C. townsendii in hibernacula across the western United States.In general, the abundance of torpid bats in our hibernacula was similar in size to those reported in the western United States (Perkins et al. 1990;Hendricks 2012;Weller et al. 2018;Whiting et al. 2018), with aggregations of tens to a few hundred individuals per cave.In contrast, our numbers of bats were lower than those observed in the eastern United States (prior to the arrival of WNS), where hibernacula may often hold thousands of individuals (Brack 2007;Storm and Boyles 2011;Frick et al. 2015;Langwig et al. 2015).Geographical variation in occurrence of hibernating bat species within Mexico (e.g., lack of torpid M. velifer in the north) and in abundances in comparison with the United States (e.g., differences in the number of torpid bats) may be due to several factors including variation in availability of areas to overwinter, bat abundances, availability and quality of roosts, use of roosts other than caves, roosts located in inaccessible areas, detectability of bats inside the roost, genetic characteristics, and historical stressors (Perkins et al. 1990;Frick et al. 2015;Weller et al. 2018;Whiting et al. 2018).
Site selection has been suggested as the main mechanism bats use to maintain a stable temperature during hibernation (Boyles et al. 2020).In addition to food availability and winter behavior, shorter and milder winters in Mexico could favor the use of microsites instead of active thermoregulation, especially if subtropical populations of hibernating bats exhibit shorter torpor bouts than northern populations (Dunbar and Brigham 2010;Geiser and Stawski 2011;Meierhofer et al. 2019a).Across all hibernacula, we found that T fur of bats was strongly and positively correlated with T sub and may be indicative of site selection of bats as a measure to reduce energy expenditure (Humpries et al. 2002;Kim et al. 2013;Boyles et al. 2020).However, because we did not measure temperature at different sections of the cave, it is not possible to distinguish whether bats selected for hibernacula conditions, microsites, or both.Similar to our study, a relationship between T sub and T fur has been reported in hibernating bats elsewhere (Meierhofer et al. 2019a), as well as very small differences of less than 1 °C between both temperatures (Storm and Boyles 2011;Kim et al. 2013).The correlation that we found in subtropical conditions is relevant in the context of the southernmost hibernation limit for Nearctic species and gives insight into thermoregulation strategies of the bats we surveyed.
Few studies have reported microclimatic temperatures used for hibernation by bat species in Mexico, and those include data on C. mexicanus, C. townsendii, M. velifer, and M. volans (López-Wilchis 1989;Ávila-Flores 2000).Although these studies used a different methodological approach, i.e., measuring ambient temperature as close as possible to the roosting surface, microclimatic temperatures provided are consistent with our findings on median T sub , ranging between 1.6 °C and 19.7 °C (López-Wilchis 1989;Ávila-Flores 2000).In general, median T sub for the 9 species we recorded were also consistent with averages reported in several hibernating bat species in the northern United States (Brack 2007;Ingersoll et al. 2010;Storm and Boyles 2011;Langwig et al. 2016;Hopkins et al. 2021), while our maximum T sub were similar to averages noted in the southern United States (Meierhofer et al. 2019a;Smith et al. 2021;Supplementary Data SD1).In contrast to our findings, we did not find studies that recorded hibernating bats at T sub below zero (Kim et al. 2013;Langwig et al. 2016;Meierhofer et al. 2019a).However, there are observations of negative body and skin temperatures (≥−2.9 °C) in hibernating mammals, including bats (Geiser 2021).This is likely an indication that measurements of T sub tend to be a few degrees lower than body temperature when ambient temperature is very cold.Finding these consistencies in T sub suggests that bats may use a range of microclimates suitable for hibernation throughout their wide geographic distribution in North America (Webb et al. 1996).
Our study shows that belowground hibernacula at high altitudes in northern and central Mexico are an important resource for temperate vespertilionid bats during a crucial time in their lives.Although we found 10 bat species, it is likely that more species will be found hibernating in Mexico given that our surveys were restricted to 1 to 5 days during the hibernation period per cave and that there are still several areas to explore.Furthermore, bats can use hibernacula during different life cycle periods including autumn swarming, mating, and maternity, increasing their importance for conservation.Although around half of the visited roosts are in areas with some level of federal and state protection, caves are not usually considered in conservation efforts (Medellín et al. 2017).Therefore, it is necessary to continue searches for hibernacula and evaluate their potential importance throughout the year (Arita 1993;Medellín et al. 2017), prioritizing conservation and management in roosts with high abundances and richness, maintaining crucial behaviors, and sheltering endemic and threatened species.
Baseline counts of bats at hibernacula help monitor abundance and species richness and may provide an estimate of regional population sizes (Loeb et al. 2015), and coupled with understanding of the temperatures used during hibernation are crucial for developing conservation and management strategies under current climatic conditions and providing a basis for evaluating responses to future environmental changes (Brack 2007;Meierhofer et al. 2019a) and other threats such as emerging diseases (Frick et al. 2019;Cheng et al. 2021).For example, considering the ongoing expansion of WNS in North America, continuous monitoring of hibernating bat populations would play an essential role in developing and implementing biosecurity protocols to prevent Pd spread and monitoring programs to detect the pathogen as soon as possible after its introduction (Holz et al. 2019;Cheng et al. 2021).
Future studies should focus on bat hibernation in Mexico at multiple temporal and spatial scales.More detailed information on duration of the hibernation period, duration and frequency of torpor and arousal bouts, diversity and species turnover within and between hibernacula, roost pattern, and microsite selection should be assessed.Furthermore, expanding studies across larger latitudinal and altitudinal gradients in Mexico will aid in understanding the ecology and physiology of hibernation, including the effects

Fig. 1 .
Fig. 1.Hibernacula located in 3 mountain ranges along northern and central Mexico.Hibernaculum identification numbers correspond to those listed in Table1.

Table 1 .
Continued concentrations (n > 100) often found between November and January (Table3).Regarding its roost pattern, M. velifer was found hibernating both solitarily and in clusters of up to 200 individuals.One hundred fifty-five C. mexicanus were observed in 23 hibernacula, mostly in conifer forest (n = 17) and in TMVB (n = 18), similar to M. velifer (Table2).Notably, these 2 species were found together in 17 hibernacula in TMVB.For C. townsendii we found 190 individuals in 39 hibernacula, but in contrast with the previous species,

Table 3 .
Abundance of Myotis velifer (left of slash) and Corynorhinus spp.(right of slash) hibernating in 5 caves located in central Mexico (states are in parentheses), which were visited multiple times between September and March in different years.Asterisk indicates abundance of M. velifer and M. occultus individuals considered together since some bats were unreachable and species identification was impossible.

Table 4 .
Adjacent substrate (T sub ) and fur (T fur ) temperatures (in °C) of 9 species of torpid bats reported in this study in central and northern Mexico.