This paper contains supplementary material that can be found online at http://journals.cambridge.org
Introduction
Bats are ecologically important as pollinators, seed dispersers and insect predators (Cleveland et al., Reference Cleveland, Betke, Federico, Frank, Hallam and Horn2006; Betke et al., Reference Betke, Hirsh, Makris, McCracken, Procopio and Hristov2008; Kalka et al., Reference Kalka, Smith and Kalko2008; Lobova et al., Reference Lobova, Geiselman and Mori2009; Boyles et al., Reference Boyles, Cryan, McCracken and Kunz2011; Kunz et al., Reference Kunz, de Torrez, Bauer, Lobova and Fleming2011). However, populations of many bat species are declining because of threats such as climate change, habitat loss and degradation, hunting and emerging diseases (Meyer et al., Reference Meyer, Aguiar, Aguirre, Baumgarten, Clarke and Cosson2010). Of these, habitat loss and degradation are the greatest threats to bats in most regions (Mickleburgh et al., Reference Mickleburgh, Hutson and Racey2002; Racey & Entwistle, Reference Racey, Entwistle, Kunz and Fenton2003; Jones et al., Reference Jones, Jacobs, Kunz, Willig and Racey2009; Kingston, Reference Kingston2010; Meyer et al., Reference Meyer, Aguiar, Aguirre, Baumgarten, Clarke and Cosson2010).
Subterranean sites, such as caves and mines, provide roosts for many bat species (Kunz, Reference Kunz and Kunz1982; Hutson et al., Reference Hutson, Mickleburgh and Racey2001), and some of the most important maternity and hibernating sites (Kunz, Reference Kunz and Kunz1982; Kunz & Lumsden, Reference Kunz, Lumsden, Kunz and Fenton2003; Murray & Kunz, Reference Murray, Kunz, Culver and White2005). Such sites are also used as mating and aggregation locations for millions of individuals (Kunz, Reference Kunz and Kunz1982; Hutson et al., Reference Hutson, Mickleburgh and Racey2001). The range and densities of bats that rely on caves are determined largely by the distribution, quantity and characteristics of available caves (Murray & Kunz, Reference Murray, Kunz, Culver and White2005; Struebig et al., Reference Struebig, Kingston, Zubaid, Le Comber, Mohd-Adnan and Turner2009). According to the IUCN Red List of Threatened Species (IUCN, 2010), 449 out of 1,132 extant bat species roost in caves and other subterranean habitats.
Disturbance of roosts threatens many bat species (Hutson et al., Reference Hutson, Mickleburgh and Racey2001), and declines of cave-dwelling bat populations were noted as early as 1952 (Mohr, Reference Mohr1972). Activities within or around caves, such as mining, quarrying, guano and bird nest collecting, caving, tourism, and deliberate or accidental disturbances can all have negative effects on bats (Hutson et al., Reference Hutson, Mickleburgh and Racey2001). In Mexico there have been dramatic declines of cave-dwelling bats caused by roost disturbance (Hutson et al., Reference Hutson, Mickleburgh and Racey2001) and in the USA roost disturbance was the cause of the decline of six threatened bat species (Elliott, Reference Elliott, Wilkens, Culver and Humphreys2000).
Currently, 121 bat species are known from China (IUCN, 2010). Although only one species (Pteropus lylei) is considered globally threatened, populations of 18 species are decreasing and the population trends of 79 species are unknown (IUCN, 2010). Sixty-two bat species are considered nationally threatened (Endangered, Vulnerable or Near Threatened; Wang & Xie, Reference Wang and Xie2004).
Despite the high proportion of nationally threatened bat species in China and the lack of information on the ecology of these species, little research has focused on bat conservation in the country (but see Niu et al., Reference Niu, Wang, Zhao and Liu2007). Based on our surveys of bats in 225 underground sites during 2003–2011 and an assessment of related social, economic and political issues, we recommend that protection of subterranean habitats should be a high priority for bat conservation in China.
Methods
We surveyed bats in 225 subterranean sites in China during March–August from 2003 to 2011 (Fig. 1). Information on roost localities was collected from the literature, Internet and local residents. Local people provided information on numbers of roost entrances, frequency of visits to the roost, identity of visitors, relative number of bats and safety issues. We entered roosts with torches, in the daytime, when the situation allowed. We recorded locations, with a global positioning system, and collected information on the freshness of bat faeces, traces of human activities, and roost and colony sizes. Colony size was assessed mainly by direct roost counts except for some inaccessible roosts, for which the colony sizes were estimated by evening emergence counts. Methods for estimation of colony size followed Kunz et al. (Reference Kunz, Betke, Hristov, Vonhof, Kunz and Parsons2009a).
Fig. 1 China, showing the locations of potential bat roosts surveyed during 2003–2011. The provinces and municipalities where we surveyed are numbered: 1, Hainan; 2, Yunnan; 3, Guangxi; 4, Guangdong; 5, Guizhou; 6, Hunan; 7, Guangxi; 8, Fujian; 9, Sichuan; 10, Chongqing; 11, Hubei; 12, Anhui; 13, Zhejiang; 14, Jiangsu; 15, Gusu; 16, Shaanxi; 17, Henan; 18, Shanxi; 19, Shandong; 20, Tianjin; 21, Beijing; 22, Neimenggu; 23, Jilin; 24, Heilongjiang.
Roost size was measured as the combined inner space size (in m3) of all chambers and passages of a cave. In most cases each chamber and passage was considered a cuboid and its volume determined as total length × mean height × width. As the resulting estimates of roost size were not precise we identified roosts as large or small based on the median (1,000 m3) of the data. Additionally, if a roost was too large to be surveyed in 15–30 minutes it was regarded as large. A roost was categorized as inaccessible if its entrances were too small to allow a person to enter. The small percentage of roosts for which size could not be determined were categorized as data deficient.
Frequency of visitation to a roost was determined by interviewing elderly residents, wherever possible farmers who spent considerable time working outside throughout the year. Using the information gathered we categorized roosts as either visited or seldom visited. Seldom-visited roosts were those for which the residents questioned had not seen visitors in the previous 12 months. Within each roost that we could enter we searched for traces of human activities (e.g. footprints, food packages, cigarette butts, scratches, notches and remains of torches). If a roost was categorized as seldom visited and we found no obvious traces of human activities it was categorized as undisturbed, and otherwise as disturbed.
The conservation value of roosts was assessed using three criteria (species richness, abundance, and the presence of species of special concern) proposed by Arita (Reference Arita1993, Reference Arita1996). Species of special concern were those categorized as Endangered or Vulnerable in China, or species endemic to China. We regard roosts that harbour ≥ 6 species or > 1,000 individuals as being of special concern.
Bat species were captured with ground-based mist nets set at entrances during emergence in the evening and sometimes in the early morning as bats returned from foraging. At least one person monitored the net, and when a bat entered it was removed with care to avoid any injury to either the bat or the handler (Kunz et al., Reference Kunz, Hodgkison, Weise, Kunz and Parsons2009b). The handling of bats conformed to guidelines for animal care and use established by the American Society of Mammalogists (Gannon & Sikes, Reference Gannon and Sikes2007). All fieldwork abided by the Law of the People's Republic of China on Protection of Wildlife.
Most species were identified based on their morphological characteristics. For species that were ambiguous we carried out DNA-based phylogenetic analysis of one 3-mm diameter biopsy punch from the wing-membrane per individual. For each captured bat we measured length of forearm, dorsal head and body length, and tail, ear, tibia and hind foot lengths (including claws, measured to the distal part of the claw) to an accuracy of 0.01 mm, with digital callipers. Body mass was measured to the nearest 0.01 g with a digital balance. All bats were released after handling except for a small number of individuals of taxonomic interest (e.g. newly recorded species in China). For species newly recorded in China we took additional external measurements in the field and cranial and dental measurements in the laboratory. These bats were euthanized by intraperitoneal injection of sodium pentobarbital (80 mg kg−1) in accordance with the American Veterinary Medical Association Guidelines on Euthanasia (AVMA, 2007), preserved in 75 or 95% ethanol and deposited in the Museum of Natural History of North-east Normal University, Jilin province, China.
To examine the potential effects of roost type (natural cave not developed for tourism, disused military fortification, active tourist cave, abandoned mine and disused tourist cave), roost size and human disturbance on species richness we used generalized linear modelling (GLM), with species richness as the dependent variable, roost type as a factor, and roost size and human disturbance as covariates. Only the main effects were considered, and the least significance difference method was used for pair-wise comparisons. Partial correlations were conducted to examine the relationship between the presence of species of special concern and roost type, roost size, human disturbance and species richness. We performed all statistical analyses with SPSS v. 19.0 (SPSS, Chicago, USA).
Results
In our surveys 57 bat species, in seven families, were found to roost in caves and other subterranean habitats. Of these, 11 species had not previously been documented to roost in caves (on either the China Species Red List or the IUCN Red List; Wang & Xie, Reference Wang and Xie2004; IUCN, 2010) and six species are endemic to China (Supplementary Table S1; for specimens collected see Supplementary Table S2). The number of species in occupied roosts was 1–15 but only c. 10% of these roosts harboured ≥ 6 species; 75% contained ≤ 4 species (Fig. 2, Supplementary Table S3).
Fig. 2 The number of roosts (see Fig. 1 for locations) in the 225 subterranean sites located and surveyed that contained from 0 to >9 species of bats.
Our literature search indicated that a further six bat species have been documented using caves for roosting in China. Combining information from the China Species Red List (Wang & Xie, Reference Wang and Xie2004), the IUCN Red List (IUCN, 2010) and bat species recorded in our survey and by other researchers, we determined that a total of 101 bat species (77%) are roosting in caves and other subterranean habitats in China (including Taiwan). This includes 30 nationally Endangered or Vulnerable species (Wang & Xie, Reference Wang and Xie2004), and nine endemic species (Supplementary Table S1).
Of the 225 underground sites we surveyed, there were 133 natural caves not developed for tourism, 34 disused military fortifications, 34 active tourist caves, 13 abandoned mines and 11 disused tourist caves. Almost 90% of the sites were disturbed and only 15% of the natural caves were free of disturbance (Fig. 3). Up to 46% of sites harboured threatened or endemic species, 30 sites harboured ≥ 6 species and 25 sites contained > 1,000 individuals, giving a total of 121 roosts that warrant special consideration for protection (Supplementary Table S3). Recreation was the most common cause of disturbance, affecting 85% of the disturbed roosts and 90% of the disturbed caves (Fig. 3).
Fig. 3 The main human activities observed at the sites of the 225 bat roosts surveyed (Fig. 1), for all roosts combined and for those in natural caves undeveloped for tourism. Storage indicates sites used to store items such as firewood. Ritual indicates sites used for ritualistic or religious purposes.
Results of the GLM revealed that species richness was positively correlated with roost size (χ2 = 22.75, df = 1, P < 0.001) but negatively correlated with human disturbance (χ2 = 7.05, df = 1, P < 0.01). Large roosts contained a mean of 3.29 species and small roosts a mean of 2.17 species (least significance difference comparison, P < 0.001). Twenty-three of the 25 roosts with > 1,000 individuals were large roosts. Disturbed roosts contained a mean of 2.59 species whereas undisturbed roosts had a mean of 3.86 species (least significance difference comparison, P < 0.05). Roost type had a strong effect on species richness (χ2 = 17.56, df = 4, P < 0.01). Caves formerly used for tourism contained a mean of 4.78 species whereas abandoned mines contained a mean of 1.83 species (least significance difference comparison, P < 0.001). However, other pairwise comparisons indicated that species richness did not differ significantly between roosts in abandoned mines, active tourist caves, disused military fortifications or natural caves (least significance differences, all P > 0.05).
Partial correlations indicated that the presence of threatened and endemic species was positively correlated with species richness (P < 0.001) and negatively correlated with roost disturbance (P < 0.05). The presence of species of special concern was not directly correlated with roost size (P = 0.30) or roost type (P = 0.68).
Discussion
Seventy-seven percent (101 of 131) of the bat species in China, including 30 species categorized nationally as Endangered or Vulnerable and nine endemic species, have been documented to roost in caves and other subterranean habitats for at least part of each year, with some roosts harbouring up to 15 species. The total number of bat species is a combination of the number of species recognized by IUCN (except Pipistrellus alaschanicus, which has been renamed Hypsugo alaschanicus, Horáček et al., Reference Horáček, Hanák, Gaisler and Woloszyn2000; and Rhinolophus rouxii, which may be the same species as Rhinolophus sinicus, Zhang et al., Reference Zhang, Jones, Zhang, Zhu, Parsons and Rossiter2009) and 12 species recorded recently (Zhang et al., Reference Zhang, Han, Jones, Lin, Zhang and Zhu2007, Reference Zhang, Gong, Zhu, Hong, Zhao and Mao2010; Feng et al., Reference Feng, Li and Wang2008; Sun et al., Reference Sun, Feng, Jiang, Ma, Zhang and Jin2008; Wu et al., Reference Wu, Motokawa and Harada2008; Kuo et al., Reference Kuo, Fang, Csorba and Lee2009; Zhou et al., Reference Zhou, Guillén Servent, Lim, Eger, Wang and Jiang2009; Jiang et al., Reference Jiang, Sun, Chou, Zhang and Feng2010; Wang et al., Reference Wang, Jiang, Sun, Wang, Tiunov and Feng2010; Wu & Thong, Reference Wu and Thong2011).
Our results showed that undisturbed roosts supported a greater number of bat species than disturbed roosts and that species of special concern were more commonly found in roosts devoid of human disturbance. This has important implications for bat conservation in China given that 77% of the bat species roost in subterranean sites but that only about 10% are free of human disturbance. Disturbance of roosts by recreational activities could cause severe problems for bats, which can be woken by human noise in the day (Mann et al., Reference Mann, Steidl and Dalton2002) and affected indirectly by changes in aspects of cave microclimate such as carbon dioxide concentration, temperature and humidity (Pulido-Bosch et al., Reference Pulido-Bosch, Martin-Rosales, López-Chicano, Rodriguez-Navarro and Vallejos1997; Gunn, Reference Gunn2004).
Research activities in caves in China are increasing. For example, there were only a few people studying bats at the end of the 20th century but there are now at least five research groups studying bats in mainland China (J. Feng, pers. comm.). Although this could improve our understanding of cave ecosystems it could also be a new threat to bats and other cave-obligate organisms if researchers are not sufficiently aware of the conservation considerations required when working in caves. Research activities can cause declines in bat colony sizes, depending on the intensity and frequency of disturbance (Hayes et al., Reference Hayes, Ober, Sherwin, Kunz and Parsons2009). As most researchers are generally interested in their own research subjects, researchers studying cave fauna other than bats can unintentionally pose potential threats to bats, and vice versa.
Based on the criteria of Arita (Reference Arita1993, Reference Arita1996), 121 of the 225 roosts examined deserve special consideration for protection. In general, larger roosts merit greater conservation concern because they support more bat species and greater numbers of individuals than smaller roosts. The absence of a correlation between roost size and the presence of species of special concern suggests that a conservation plan based solely on roost size would not be adequate for the protection of the cave bats of China. The greater number of species of bats in disused tourist caves indicates that this type of roost also deserves greater conservation attention. Improperly installed gates, for example, could cause species to lose their preferred roosts.
Prioritization is essential to minimize biodiversity loss because the location of and threats to biodiversity are distributed unevenly (Brooks et al., Reference Brooks, Mittermeier, da Fonseca, Gerlach, Hoffmann and Lamoreux2006). Our results suggest that protection of subterranean habitats should be a high priority for bat conservation in China. Up to 46% of the roosts surveyed harboured either threatened or endemic species of bats (i.e. those that are most critical indicators of conservation concern; Brooks et al., Reference Brooks, Mittermeier, da Fonseca, Gerlach, Hoffmann and Lamoreux2006; Miller et al., Reference Miller, Rodríguez, Aniskowicz-Fowler, Bambaradeniya, Boles and Eaton2006) and, although 77% of bat species in China rely on underground roosts, only 10% of these were free of human disturbance. Awareness of cave ecosystems and cave-dwelling species in China is poor, not only among the public but also among some government managers and researchers (Whitten, Reference Whitten2009; Zhang et al., Reference Zhang, Zhu, Jones and Zhang2009). An increasing number of caves in China continue to be exploited for tourism without prior environmental impact assessments (Song et al., Reference Song, Wang, Liang, Wei and Lin2004). By 2005 > 400 caves in China had been developed for tourism (Wu & Zhuang, Reference Wu and Zhuang2007) but there is no government agency or conservation NGO directly concerned with caves (Whitten, Reference Whitten2009). Any person, whatever their occupation, age or intention, may visit any cave without restriction.
Our findings indicate that prioritization of the protection of underground habitats for bats in China is warranted. For the development of management guidelines and policies, however, more ecological data on bats and caves are required, with an emphasis on population assessment and monitoring. To support this we are now actively communicating with government managers responsible for wildlife conservation and cave resource management by sharing our data with them. We recommend that bat researchers and other researchers specializing in caves should promote the conservation of cave ecosystems in China.
Acknowledgements
This study was financed by the National Natural Science Foundation of China (grants 31030011, 30870371, 30900166) and Fundamental Research Funds for the Central Universities (No. 10SSXT128). We are grateful to Limin Shi, Longru Jin, Sen Liu, Aiqing Lin, Keping Sun, Ying Liu, Genxian Ye, Xu Zhu, Shi Li and Yunjiao Zhao for their assistance in the field. We would also like to thank Tom Kunz and an anonymous reviewer for helpful comments.
Biographical sketches
Jinhong Luo studies the conservation biology and sensory ecology of bats. Currently he is investigating the impacts of climate change, noise pollution and roost disturbance on bats. Tinglei Jiang specializes in the taxonomy of Chinese bats and works on geographical variation in echolocation calls. Guanjun Lu and Lei Wang are working, respectively, on the phylogeography and echolocation behaviours of bats, and both have interests in the taxonomy of Chinese bats. Jing Wang studies the habitat selection of bats, with a special interest in understanding the effects of human activities on habitat use by horseshoe bats. Jiang Feng carried out some of the earliest research on the echolocation behaviour of bats in China. Currently, he leads a large research group studying the evolutionary biology, behaviour ecology and conservation biology of bats.
