Jump to ContentJump to Main Navigation
The Biology of Caves and Other Subterranean Habitats$

David C. Culver and Tanja Pipan

Print publication date: 2019

Print ISBN-13: 9780198820765

Published to Oxford Scholarship Online: June 2019

DOI: 10.1093/oso/9780198820765.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 06 July 2020

Conservation and Protection of Subterranean Habitats

Conservation and Protection of Subterranean Habitats

(p.226) 10 Conservation and Protection of Subterranean Habitats
The Biology of Caves and Other Subterranean Habitats

David C. Culver

Tanja Pipan

Oxford University Press

Abstract and Keywords

A critical factor of the subterranean fauna and one that increases the risk of extinction is geographical rarity. Some stygobionts and troglobionts are also numerically rare. Subterranean organisms are also at increased risk of extinction because of low reproductive rates, and in the case of bats, because of their propensity to cluster in large numbers in a few caves. Threats to the subterranean fauna are of four general kinds—alteration of the physical habitat, changes in water quality and quantity, direct changes to the subterranean fauna, and global warming. The selection of sites for preservation requires detailed inventory data, but available evidence suggests that a majority of species can be protected at least at one site and that a relatively small percentage of total land area is required. A variety of mechanisms are available for site protection, including listing as a Ramsar wetland and as a UNESCO world heritage site.

Keywords:   endemism, extinction threats, geographical rarity, Ramsar wetlands, UNESCO World Heritage sites

10.1 Introduction

Subterranean habitats and species, especially those involving cave habitats, are attracting increasing interest and concern among conservationists, cavers, and speleobiologists, and for good reason. Most stygobionts and troglobionts are highly restricted geographically and often are numerically rare, making them vulnerable to even relatively minor disturbances. An examination of the concept of rarity (Rabinowitz et al. 1986), how it applies to the subterranean fauna, and why being rare increases vulnerability are the subjects of the first part of the chapter. There are other biological factors that may put the obligate subterranean fauna and many bats that utilize caves at increased risk of extinction, including low reproductive rates, high susceptibility to environmental change, and inability to withstand disturbance. Taken together, these factors reduce the ability of subterranean species to respond to environmental stress. In light of these biological attributes, including rarity, we consider the threats to subterranean communities. These include universal threats such as global warming and groundwater pollution, which should be recognized as a universal threat too. Other important threats are more local or regional. For example, mining and quarrying are big threats to caves and aquifers in Brazil and Western Australia (Auler and Piló 2015; Hamilton-Smith and Eberhard 2000) but little threat at present to caves and aquifers in central Texas, where development poses a much larger threat (Hutchins 2017). Water diversion and control projects are major problems in Bosnia & Herzegovina but not at present in Slovenia [although they were a threat in the past (Sket 1979)]. In caves, a special threat exists—human visitors to caves, and we take up the question of the impacts of human visitation. In the fourth section, we consider what should be protected, where it should be protected, and how this might be accomplished. With the availability of increasingly accurate maps of the geography of subterranean species richness (p.227) (Chapter 8), it is possible to decide where preserves should be. Site selection and protection cannot be performed in a vacuum, and in the fifth section we review the international, national, and local agencies, both governmental and non-governmental, that have been part of protection efforts for subterranean habitats and species. In the final section, what in many ways is the most difficult problem of all is considered—how a local site is protected and managed. The time is past when it is appropriate to protect a cave or well by only protecting a few metres around the site. The vexing and controversial strategy of gating of caves as a conservation tool will be reviewed. There can be no doubt that in some cases, appropriately designed cave gates have protected populations and probably species such as Myotis grisescens and Myotis sodalis from extinction, but there can also be no doubt that inappropriately designed gates have caused the disappearance of some bat populations (Elliott 2012).

For many biologists and others interested in subterranean habitats and communities, subterranean species are worth protecting in their own right—a biocentric view of species protection. But subterranean species also provide ecosystem services—services to human populations provided by ecosystems that would otherwise have to be accomplished in some other way (Daily 1997). The ability of groundwater microorganisms to decompose organic matter provides a number of significant ecosystem services, including rendering harmless human pathogens, breaking down organic wastes, and the net result—purifying groundwater (Herman et al. 2001; Griebler and Avramov 2015). Groundwater microorganisms often play a significant role in the clean-up of groundwater contaminants, sometimes called natural attenuation. Groundwater clean-up is a multibillion dollar industry. Additionally, groundwater provides mitigation of floods and droughts and nutrient cycling (Griebler and Avramov 2015). Bats also provide important ecosystem services, including insect control, seed dispersal, and pollination (Kunz et al. 2011). Bat guano is still an important source of fertilizer in some countries. Fenolio (2016) makes the more general case that subterranean species can provide new drugs, a general and powerful argument for the protection of biodiversity. Indeed, there have been several studies of groundwater and cave microbiota in the context of new drugs (Bhullar et al. 2012).

10.2 Rarity

Rabinowitz (1981) pointed out that rarity means different things to different biologists. From a botanical perspective she suggested that rarity has three meanings—a species can be numerically rare throughout its range (numerical rarity), it can occur in a rare habitat or habitats (habitat rarity), and it can be geographically rare with a restricted range (geographical rarity). Species (p.228) may be rare along one, two, or all three of her axes, leading to her ‘seven forms of rarity’. Her classification works perfectly well with respect to the subterranean fauna.

There is little doubt that the majority of the subterranean fauna is geographically rare. In an analysis of stygobionts in Belgium, France, Italy, Portugal, Slovenia, and Spain (Michel et al. 2009), part of the PASCALIS project (Chapter 8), it was found that 464 of 1059 stygobionts (44 per cent) occurring in caves and interstitial habitats were limited to a single 12 min by 12 min (12´ x 12´) cell approximately 400 km2 in area. In a study of the calcretes of the Pilbara region of Western Australia, Eberhard et al. (2009) suggested that nearly the entire fauna is comprised of short range endemics (SREs), species with ranges less than 10 000 km2 (Harvey 2002). In an analysis of the obligate cave fauna of the United States, Culver et al. (2000) found that 463 out of 673 troglobionts (69 per cent) and 131 out of 300 stygobionts (44 per cent) were limited to a single county, with the average size of counties with stygobiotic or troglobiotic species being approximately 10 000 km2. In the case of the US troglobiotic fauna east of the Mississippi River, 211 out of 467 species (45 per cent) are known from a single cave. One of the hallmarks of cave organisms is their endemism.

All of these levels of endemism are much higher than that recorded for any surface habitat. Endemism seems higher in troglobionts than stygobionts, and troglobionts in general have smaller ranges (Lamoreaux 2004), but to date no pattern has been reported with respect to trophic position or body size (Culver et al. 2009).

The pattern with respect to habitat rarity is mixed. In the main karst areas (Fig. 1.1) and lava tube areas, caves are quite common, reaching up to 5 per km2 in Slovenia and in lava flows in Hawaii (Culver and Pipan 2014), and it is areas such as these that harbour most of the stygobionts and troglobionts found in caves. There are a few stygobionts and troglobionts found in isolated areas. An example of this is the troglobiotic beetle Choleva septentrionis holsatica, endemic to Segeberger Höhle, a cave in an isolated rock salt dome in northern Germany (Ipsen 2000). Some cave habitats, especially the cave hygropetric (Sket 2004a), are probably quite rare, so hygropetric specialists, such as the beetle Nauticiella stygivaga, are rare because of habitat rarity. As a habitat, epikarst is probably more common than caves, being more or less continuous in karst areas except in the tropics and glaciated areas (Williams 2008). Interstitial habitats, especially fluvial aquifers, are common, or at least as common as surface streams and rivers. Calcrete aquifers, while confined to very arid regions, are quite common in these regions. That most superficial of shallow subterranean habitats, the hypotelminorheic, is certainly not a common habitat but its geographical distribution is not known. In the environs of Washington, DC, where it has been best studied, it is localized to a few bands of habitat. All in all, most stygobionts and (p.229) troglobionts are not rare because of habitat rarity, but there are some exceptions.

The question of numerical rarity of stygobionts and troglobionts is a very interesting one. A considerable number of troglobionts and stygobionts from caves are known from only a handful of specimens, in some cases a single specimen. The obvious rarity of these species may be more apparent than real because it is likely that the primary habitats of many of these species are not caves but habitats such as epikarst and phreatic water, and that they are accidentals in caves. Other populations of stygobionts and troglobionts are known to be quite large. The best known example is that of the Baget ecosystem in France (see Chapter 9), where the number of individuals of stygobiotic copepod species that washed out of the system was in the thousands (Rouch 1970).

There are two general methods available to estimate population size that bring some clarity to the range of observations about population size. One of these is mark–recapture, based on recapturing in a second sample individuals that were marked in the first sample. Population size (X) and its standard error can be estimated as follows (Begon 1979):



where a is the number marked in the first sample, n is the number of individuals in the second sample, and r is the number of marked individuals in the second sample. Such studies with subterranean animals are technically difficult and standard errors are often very large because of relatively small sample size (Knapp and Fong 1999; Fong 2003) but nonetheless very informative. These estimates of the size of the local population and the geographical extent of the population being estimated are dependent on the extent of dispersal and mixing of the individuals marked. The small number of these estimates available (Culver 1982) indicate population sizes between 100 (the crayfish Aviticambarus sheltie in Shelta Cave, Alabama, USA) and 9000 (the crayfish Orconectes inermis inermis in Pless Cave, Indiana, USA). Those on the small end of the estimates range are certainly numerically rare. Populations can be quite dense—Porter and Hobbs (1997) estimated the density of the amphipod Crangonyx indianensis as 777 ± 108/m2 in Dillion Cave, Indiana.

Population geneticists use a different measure of population size, ‘effective population size’, basically the size of a randomly mating population that would result in the same levels of heterozygote frequency and same levels of genetic variation as the observed population. A population with low genetic variation or low heterozygosity has a smaller effective population size. Except for a few fish populations, there is little evidence of reduced genetic (p.230) variability in stygobionts and troglobionts (Sbordoni et al. 2000, 2012). Buhay and Crandall (2005), using mitochondrial DNA sequences, calculated an effective population size for several stygobiotic species of crayfish of between 20 000 and 80 000. Sbordoni et al. (2012) attribute the large estimates of effective population sizes of subterranean species in general to selection for heterozygotes rather than population size per se. In general it appears that subterranean populations, as with surface-dwelling populations, can be either large or small and some unknown fraction of stygobionts and troglobionts are numerically rare.

Thus, the majority of stygobionts and troglobionts are geographically rare, many are likely to be numerically rare, and a few are in rare habitats. This rarity has important conservation implications. Geographically rare species are subject to catastrophic losses as the result of relatively minor and frequent environmental insults, if for no other reason than their geographically restricted range. Numerically rare species are more likely to go extinct than common species because of genetic inbreeding, demographic stochasticity (such as the appearance of a single-sexed population in a generation), and environmental stochasticity (minor environmental insults). Taken together, conservation biologists call these phenomena the extinction vortex (Groom et al. 2005).

10.3 Other biological risk factors

Because many stygobionts and troglobionts, as well as bats, have relatively low reproductive rates (although lifetime rates may be high due to increased longevity—see Chapter 6), their rate of population growth following an environmental insult will be low, resulting in a smaller population for a longer period of time relative to surface-dwelling populations. This results in increased extinction risk because they are in the extinction vortex longer.

Stygobionts and troglobionts may also be especially sensitive to some kinds of environmental fluctuations. For example, troglobionts are often especially sensitive to changes in relative humidity as a result of exoskeleton thinning (Howarth 1980, 1983). Stygobionts may be more or less sensitive than surface relatives to heavy metals (Notenboom et al. 1994), but similarly to their surface counterparts, stygobionts, especially interstitial species, frequently must cope with heavy metal contamination (Vesper 2012). Stygobionts and troglobionts appear to be especially sensitive to non-subterranean competitors and predators that can occur in subterranean sites as a result of pollution events, especially organic pollution of streams (Sket 1977; Chapter 5).

Cave-inhabiting bats have special biological risk factors associated with cave use. For species that concentrate in large numbers in a small number (p.231) of caves, such as Myotis grisescens in the southeastern United States, the very fact of their concentration makes them vulnerable to any changes that occur in that cave. Hibernating bats are also sensitive to arousal from hibernation, an energetically expensive proposition, and after repeated arousals they may not have enough fat reserves to survive the winter (Kunz and Fenton 2003). For species that use caves as nursery roosts, with females gathering in large numbers to raise their young (usually one per year), disturbance at this critical time can have a damaging effect on breeding success and survival of the population. These are some of the reasons that white nose syndrome (Moore and Kunz 2012) has had such a devastating effect on the North American bat fauna. The biotic factors that increase the vulnerability of the subterranean fauna are summarized in Table 10.1.

Table 10.1 Biotic factors of the subterranean fauna that increase its extinction risk.

Biotic factor


Geographical rarity

Most stygobionts and troglobionts, especially troglobionts; some hibernating bats

Numerical rarity

Some stygobionts and troglobionts, especially vertebrates

Habitat rarity

Cave hygropetric, a few cave areas, hypotelminorheic?

Low reproductive rate

Bats and most stygobionts and troglobionts

Sensitivity to environmental flucations

Most stygobionts and troglobionts

Sensitivity to surface-dwelling competitors and predators

Most stygobionts and troglobionts

Strong clustering


Sensitivity to arousal from hibernation


10.4 Threats to the subterranean fauna

Threats to subterranean fauna are about as diverse as threats to surface-dwelling species, especially since most environmental disasters in surface habitats are environmental disasters for subsurface habitats as well. And there are, as we shall see, disasters limited to the subsurface. Jones et al. (2003) divide threats into three overarching categories:

  • Alteration of the physical habitat.

  • Water quality and quantity.

  • Direct changes to the subterranean fauna.

To this list the threats of global climate change to the subterranean fauna must be added (Mammola et al. 2017).

(p.232) Threats to subterranean faunas are present throughout the world, and there are regional differences. Threats in developed and developing countries may be different. There is more detailed information available about threats to subterranean habitats, especially caves, from the United States than elsewhere, and many of the examples we cite are from the United States. Whether there are actually more immediate threats in the United States is not clear. Certainly monitoring of subterranean habitats, including both caves and interstitial aquifers, is not as extensive in developing countries. However, it may also be that, relative to other developed countries, there are more environmental threats in the United States. This is not as implausible as it appears, because population growth rates in the United States are higher than most developed countries, especially Europe. Finally, the US Endangered Species Act may be more thoroughly implemented with respect to the subterranean fauna than legislation elsewhere, such as the European Union Habitats Directive.

10.4.1 Alteration of the physical habitat

Quarrying of limestone, especially for the making of cement, is the ultimate kind of threat because it completely removes the habitat. Worldwide, the area of fastest growth in limestone quarrying is southeast Asia. Annually,1.75 × 107 metric tons of limestone are quarried from southeast Asia and it is growing at a rate of 6 per cent per year, higher than any other area of the world (Clements et al. 2006). The limestone karst regions of Indonesia, Thailand, and Vietnam cover an area of 400 000 km2 and are ‘arks of biodiversity’ (Clements et al. 2006), both because they are biodiversity hotspots and because they are relatively untouched by agricultural and forestry practices because of the rugged terrain, such as tower karst and karst pinnacles (Ford and Williams 2007). Both the surface and subsurface biota are very incompletely known, and this is one of these situations where species disappear before they are even discovered, as can also happen with cryptic species in well-studied areas (Niemiller et al. 2013a). Because of the island-like nature of these karst outcrops (see Chapter 7), endemism of not only the subsurface species, but also the surface-dwelling species, is high. Vermeulen and Whitten (1999) report six land snails endemic to the Sarang karst, an area of only 0.2 km2; and 50 endemic to the Subis karst, a ‘large’ area of 15 km2, both in Borneo. Clements et al. (2006) also point out that the subsurface and surface of karst areas in southeast Asia is understudied, even relative to other habitats in the region. Quarrying is not the result of entirely local factors. Much of the cement produced is exported, and funding for the development of cement plants in southeast Asia is international.

Quarrying is a threat to cave and karst faunas elsewhere as well. Hamilton-Smith and Eberhard (2000) point out quarrying is a threat in much of Australia, including the highly diverse Cape Range area of northwest Australia (Humphreys 2004). Jones et al. (2003) report that Zink Cave, (p.233) Indiana, was almost completely quarried away, resulting in the extirpation of the stygobiotic fish Amblyopsis spelaea from the cave. A major bat hibernaculum in eastern North America is Hellhole in West Virginia (Dasher 2001), with 45 per cent of the known individuals of the US federally endangered bat Corynorhinus townsendii virginianus1 and a hibernating population of another endangered bat, Myotis sodalis. An extensive limestone quarry is adjacent to the cave and negotiations concerning where quarrying is allowed continue between the owner of the quarry and various government agencies. Access to the cave is controlled by the quarry, making both negotiations and monitoring especially difficult.

In karst areas, road construction both reveals new caves, such as Inner Space Caverns in Texas, USA (Elliott 2000), and destroys caves or portions of caves (Knez and Slabe 2016). The details of the location of highways through karst regions can make a big difference in terms of environmental impacts. For example, small changes in routing of a highway in southwestern Virginia, USA, protected Young–Fugate Cave, a cave with a hibernating colony of the federally listed M. grisescens (Hubbard and Balfour 1993; Tercafs 2001).

Especially in the past, caves themselves were directly mined, especially for guano from large bat colonies, and in North America for saltpetre, an ingredient of gunpowder (Hubbard 2012). Devastation of the fauna in these cases must have been extreme. Saltpetre mining was generally carried out in drier portions of the cave, which have few troglobionts because of low humidity.

Another type of site alteration is the development of a cave for tourist visitation. The commercial caves, such as Postojnska jama and Mammoth Cave, both of which are biodiversity hotspots (see Chapter 9) have over 1 million visitors per year. Commercialization of a cave requires physical alteration of natural passages and installation of lights, with the concomitant development of a lampenflora, the growth of plants associated with electric lighting (Aley 2004). Speleobiologists have frequently warned against excessive commercialization of caves because of these changes. In caves such as Postojnska jama and Mammoth Cave, a relatively small percentage of the total known passage has been altered and lit for tourist visitation. Those sections of caves that are commercialized, are, as a rule, depauperate with respect to stygobionts and troglobionts.

Physical alteration of a cave entrance, either by filling it in, enlarging it, or putting a gate on it, can have an impact on the fauna, especially the terrestrial fauna. Either filling in or gating can alter the movement of animals in and out of the cave, an important source of organic carbon for many troglobionts (Chapter 2). Enlarging an entrance or creating an artificial entrance can increase air flow and drying, reducing relative humidity. However, the major impact of alteration of entrances is on bat populations. Historically, (p.234) Mammoth Cave in Kentucky was an important bat hibernation site but after the Historic Entrance was modified to block incursions of cold winter air, bats abandoned the cave (Elliott 2000). Before the effect of gates on bats were fully understood and before gate design was perfected, gating of caves, ostensibly to protect bats, probably caused much of the decline of M. sodalis from the 1960s to the 1980s (MacGregor 1993). Improperly designed gates can cause bats to abandon a roost, and if the spaces in the gate are too close together, bats are forced to crawl through the gate. In this situation, the cave entrance becomes a magnet for bat predators, such as snakes and feral cats. Cave gates that are ‘bat-friendly’ have horizontally stiffened angle-irons at 15 cm, with vertical supports at least 1.2 m apart (Fig. 10.1) (Elliott 2000). Some bats, such as M. grisescens, do not tolerate complete gates and more complex gate designs are needed (Elliott 2012; Hildreth-Werker and Werker 2006). Some bat species, such as Miniopterus schreibersii in Europe, avoid caves with gates, especially during the breeding season (Mitchell-Jones et al. 2007). In such cases, other systems to prevent unauthorized access, such as fences around the entrance, are needed.

Conservation and Protection of Subterranean Habitats

Fig. 10.1 Gate at the entrance to Fisher Cave, Missouri, USA, designed to allow unimpeded access for bats. Photo by H. Hobbs III, with permission. See Plate 21.

Conservation and Protection of Subterranean Habitats

Plate 21 Gate at the entrance to Fisher Cave, Missouri, USA, designed to allow unimpeded access for bats. Photo by H. Hobbs III, with permission. See page 234 in text.

10.4.2 Water quality and quantity

Interstitial aquifers, especially shallow fluvial aquifers, are subject to most of the same environmental threats as rivers and streams. There are few places (p.235) left in the world where human activity has not had a negative impact on water quality. In surface habitats, species living in rivers are typically more at risk of extinction than species in other habitats, based on data for the US fauna (Master et al. 1998). The threats to groundwater are summarized in Table 10.2. Several environmental drivers are of critical importance. One is agriculture and the nearly universal overapplication of fertilizers and pesticides that accompanies it. For example, nitrate levels in groundwater throughout Europe continue to rise, and due to overuse of fertilizers and pesticides they are frequently found in shallow groundwater (Notenboom 2001). Another important driver is water extraction for agriculture, industry, and urban activities, especially the extensive use of irrigation in agriculture. Groundwater levels have fallen in many areas, more than 30 m in some cases (p.236) (Danielopol et al. 2003). This results in changed and reduced connections between surface and subsurface with the concomitant change in nutrients (see Chapter 4).

Table 10.2 Summary of drivers and associated pressures on groundwater.



Above-ground activities

Climate and natural processes

Influence of water discharge and recharge; input of organic and inorganic substances.

Urban activities and infrastructure

Seeping of oil products from fuel storage tanks; heavy metals, salts, PAHs (polyaromatic hydrocarbons; leakage of sewage systems. Growth of water demand.


Additional water demand, waste and sewage in sensitive areas.


Spillage of chemicals, seeping of oil products from fuel storage tanks. Local groundwater withdrawal.


Leaching of persistent pesticides and metabolic products; and of nitrate and metals from fertilizers and manure. Groundwater withdrawal for irrigation, lowering water table.

Waste treatment (illegal, improper)

Leaching of pollutants from waste-disposal sites.

Below-ground activities


Lowering groundwater levels and changing flows; changes in physicochemical conditions; pollution with mining spoils (heavy metals).

Water extraction

Overexploitation; saltwater intrusion; upwelling of mineral rich groundwater; decline of groundwater levels; changes in physicochemical conditions.

Waste injection and underground storage

Direct introduction of contaminants; changes in physicochemical conditions.

Heat and cold storage

Changes in physicochemical conditions.

Infrastructure building below ground

Lowering groundwater levels and changing flows; changes in physicochemical conditions.

Surface water infiltration

Nutrient enrichment of groundwater; infiltration of pollutants.

Gas and CO2 storage

Changes in physicochemical conditions.

Source: Modified from Notenboom (2001).

Even relatively mild alterations to aquifer recharge affect the fauna. In a shallow groundwater aquifer in the city of Lyon, France, Datry et al. (2005) found that artificial recharge of some sites with storm water changed the species composition at those sites, although this was probably the result of the increase in organic carbon rather than in recharge per se.

One very interesting and instructive example of recharge effects is the Trebišnjica river system in southern Bosnia & Herzegovina (Čučković 1983; Minanović 1990). The Trebišnjica had a surface water course of between 35 and 90 km, depending on water flow, and it sinks in Popovo Polje. Poljes are spring-fed large karst depressions with flat floors that are commonly covered by river sediments. During the heavy rains, especially in the fall, a temporary lake forms in the bottom of the polje due to flooding of the river. Along the sides of Popovo Polje are many caves, including Vjetrenica, one of the most diverse caves in the world (see Chapter 8). Many of these caves were periodically flooded and with the floods came organic material necessary for the survival of many stygobiotic species (see Chapters 2 and 4). To provide hydroelectric power and to control flooding, the government of what was at that time Yugoslavia instituted construction of a dam and a channelization of the Trebišnjica (Fig. 10.2). With the completion of construction in 1979, Trebišnjica was no longer a sinking (‘lost’) river (Čučković 1983), and the surrounding caves were starved of nutrients. A fish species, Paraphoxinus ghetaldi, endemic to the polje and especially common in ecotones between surface and subsurface water, is possibly extinct. The population of the salamander Proteus anguinus in the polje was perhaps the largest one anywhere and it is now decimated. Populations of other unique species such as the Dinaric cave clam Congeria kusceri, the only stygobiotic cave clam in the world, and the polychaete worm Marifugia cavatica have likewise been decimated as a consequence of damming of the river and the resulting hydrological changes.

Conservation and Protection of Subterranean Habitats

Fig. 10.2 Channelized Trebišnija watercourse in Popovo polje, Bosnia & Herzegovina in 2005. Photograph by M. Zagmajster, with permission. See Plate 22.

Conservation and Protection of Subterranean Habitats

Plate 22 Channelized Trebišnija watercourse in Popovo polje, Bosnia & Herzegovina in 2005. Photograph by M. Zagmajster, with permission. See page 237 in text.

Dam construction can have direct negative effects on caves. When the New Melones Reservoir on the Stanislaus River, California, USA, was constructed in the late 1970s, about 30 caves were inundated, including McClean’s Cave, one of the two known localities at the time for the troglobiotic harvestman, Banksula melones (Elliott et al. 2017). The population was transplanted to a nearby mine, which was stocked similar to a terrarium with cave soil, rocks, and rotting wood (Elliott 1981), and the population was still in the mine 20 years later (Elliott 2000). Transplanting to an artificial site was preferable to transplanting to another cave since the mine had no natural community that would have been disrupted. One positive outcome of this ecological disaster was that a thorough inventory of nearby caves revealed 16 other caves where B. melones was found.

(p.237) Excessive groundwater use, especially from karst aquifers, can also have a major negative impact on fauna. The Devil’s Hole pupfish, Cyprinodon diabolis, is known from a single sinkhole in Nevada, USA, where it lives in the ecotone between surface and groundwater. On the US Endangered Species list, it was reduced in numbers due to excessive pumping, and now groundwater extraction in the region is limited because of the fish. Similarly in the Edwards Aquifer (see Chapter 9), species living in springs (ecotones between surface and groundwater) are threatened, including two beetles, one fish, a salamander, and a species of wild rice (Elliott 2000).

The impact of organic pollution can be drastic and catastrophic, as the next examples show. All subterranean habitats, especially SSHs, are at risk from spills of toxic materials, typically from trucks on roads, but also from railroads and leaking storage tanks, especially associated with petrol stations. The number of underground storage tanks is staggering. In the mid-1980s, at the start of the USA Leaking Underground Storage Tank (LUST) Trust Fund, there were more than two million storage tanks in the USA, and by 2018, there had been over 540 000 toxic releases (U.S. EPA 2018). These spills either quickly enter SSHs or are buried in SSHs to begin with. Most storage materials are nonaqueous phase liquids (NAPLs), which includes fuels, solvents, and insulators (Loop 2012). NAPLs have, by definition, limited solubility in water and hence are difficult to remove from the subsurface especially because groundwater extraction and treatment (‘pump and treat’) (p.238) is very inefficient (Herman et al. 2001). Even the limited solubility of many NAPLs results in groundwater contamination that exceeds the maximum concentration level set by the U.S. Environmental Protection Agency (Loop 2012). NAPLs include hydrocarbons, vinyl chloride, PCE, PCBs, and many insecticides and herbicides such as atrazine. NAPLs in general come to reside in the shallow subsurface, and are moved out of this zone by precipitation, but are often held for extensive periods of time in the soil, epikarst, and regolith, and continue to contaminate the aquifer below. NAPLs can be retained in pores (epikarst and MSS), and in the sediments (soil and hypotelminorheic). Natural degradation (attenuation) is faster for hydrocarbons than chlorinated compounds such as chlorinated solvents and PCBs, but for a large spill may take many decades.

The organic pollution of Hidden River (Horse) Cave System, Kentucky, USA, is in some ways the equivalent of the hydrological alteration of Popovo polje—they were both catastrophic and instructive. A large cave near Mammoth Cave, with a large, attractive entrance in the small town of Horse Cave, Hidden River Cave was commercialized in 1916. It had a rich stygobiotic and troglobiotic fauna, including fish and crayfish (Elliott 2000). Increasing contamination of the cave water from indiscriminate sewage disposal from Cave City and Horse Cave, as well as wastes from a nearby creamery, led to the closing of the cave’s tourist operation in 1943. By the 1960s, the stench from the cave made it unpleasant to even walk by it. The stygobiotic fauna, at the least, was extirpated from the cave. By the late 1980s, the major sources of pollution had stopped and the cave stream began to recover. By about 1995, the original animal community had recolonized the formerly polluted cave stream from unpolluted upstream reaches of the caves (Lewis 1996). In Thompson Cedar Cave in Virginia, another cave with serious organic pollution (in this case from sawdust waste), the fauna recovered about 15 years after the original sources of pollution had been removed, having recolonized from unaffected upstream reaches (Culver et al. 1992). In all of these cases of organic pollution, the decline in water quality was accompanied by invasions of organisms typical of polluted waters, such as tubificid worms. Competition and predation from these invading species may actually be a bigger threat to the stygobionts than water quality itself (Sket 1977).

10.4.3 Direct changes to the subterranean fauna

A fungal disease, white nose syndrome, has wreaked havoc on the hibernating North American bat fauna. First discovered in the winter of 2006–2007 in Howe Caverns, a commercial cave in New York, bats were observed prematurely emerging from hibernation, flying erratically, and dying. The cause of death was determined to be the fungus Pseudogymnoascus destructans2. (p.239) Although estimates vary, six million bats have died from the disease since 2007. Nine species have developed WNS, including seven in the genus Myotis. Declines in affected bat populations in the northeast United States have averaged 73 per cent (Frick et al. 2010). Characteristics of the disease include (Moore and Kunz 2012):

  • Fungal infections and tissue damage.

  • Depleted fat reserves.

  • Atypical winter behaviour.

  • Changes in immune response during hibernation.

  • Wing damage.

Since its discovery it has spread extensively throughout the eastern half of the US, and cases have been reported from Washington state (www.whitenosesyndrome.org). The transmission of WNS by bats is well documented but transmission by humans is possible and even likely. To minimize this problem, decontamination procedures for cavers have been implemented. The response to WNS has been controversial. If some individuals have developed immunity, or the disease has declined in virulence, then more aggressive intervention is not needed. If on the other hand, there is no immunity or decline in virulence, then possible solutions such as captive breeding, immunization, and application of anti-fungal chemicals become more relevant.

The most direct human impact on cave faunas is that caused by human visitation. The most egregious example is that of deliberate destruction of bats. Elliott (2000) provides several examples, including a case where four men were convicting of shooting and crushing to death the endangered Indiana bat (M. sodalis) in Thornhill Cave, Kentucky. For many people, bats are vermin, nothing more than carriers of disease such as rabies and histoplasmosis, even though the threat of these diseases is very slight (Woloszyn 1998). There is no doubt that large concentrations of bats are unhealthy to be around if for no other reason than the concentration of ammonia resulting from urine in the air, but there is no reason, except for bat biologists, to be around large concentrations of bats. The combination of fear and loathing of bats has made bat protection difficult, and makes gates the primary tool of protection (see section 10.7). More benign visits to caves with hibernating bats may also have a negative impact, because the activities of cavers may awaken bats from hibernation, causing significant energy expenditure. Repeated arousals from hibernation may result in mortality, as is the case with WNS. Cave visitation may also have some negative impact on stygobiotic and troglobiotic species, primarily by compaction of terrestrial habitat and disruption of stream habitats when stream walking is necessary.

Overcollecting, especially of vertebrates, may well have negative effects, especially on single-site endemics (Elliot 2000). Most speleobiologists have (p.240) had the experience of collecting from a site, then returning to find as many animals as before, or returning to find very few. Prudence dictates that collecting be kept to the minimum necessary.

In Europe, the problem is exacerbated by a collectors’ market for some cave animals, such as beetles (Simečić 2017). A potentially specially damaging collecting technique is pitfall trapping. Pitfall traps baited with cheese or rotting meat can attract hundreds of troglobionts in a day or two.

10.4.4 Global warming

A final human activity may, in the coming years, dwarf all other impacts. This is global warming. Since subterranean habitats in general are less variable than surface habitats, organisms in subterranean habitats will experience less in the way of temperature extremes. However, the rise in average temperature will increase average temperatures of subterranean habitats since their temperature approximates the mean annual temperature. Since any organism, surface or subterranean, is rarely adapted to temperatures it never encounters, rising temperatures may result in lethal conditions for some stygobionts and troglobionts. Since stygobionts and troglobionts have very limited dispersal capability, if temperatures are rising relatively rapidly, species may go extinct. Mammola et al. (2017) modelled the impact of expected change in temperature on Troglophantes spiders in the Italian Alps, and predicted that most of the highly endemic species would go extinct.

10.5 Site selection

To protect subterranean species, their location needs to be known. The problem of sampling completeness was considered in Chapter 8, and it is as relevant here. Thorough inventories are critical if appropriate decisions about protection are to be made. Several useful inventories are available on the World Wide Web. The International Union for the Conservation of Nature (IUCN) maintains a list of species at risk that includes subterranean species (www.iucnredlist.org) and the US Fish and Wildlife Service (www.usfws.gov/endangered/) maintains a list of threatened and rare species. Both these lists are limited in that not all subterranean species, or even a significant fraction of them, have been evaluated. The non-governmental US conservation organization NatureServe (www.natureserve.org) maintains a more complete list for the United States of species of concern, which includes most subterranean species because of their small ranges.

The selection of sites for protection on a regional basis is a complex problem, one that has different solutions, depending on the criteria used. Possible (p.241) criteria that can be used in reserve design include (Izquierdo et al. 2001; Michel et al., 2009):

  1. 1. maximizing the number of species in the reserve network;

  2. 2. maximizing the number of endemic species in the reserve network;

  3. 3. maximizing the number of species in each reserve; and

  4. 4. minimizing the number of reserves while keeping total area constant.

More complex site selection criteria are possible. Borges et al. (2012) include a variety of other criteria, such as threat level, in a sophisticated analysis of site selection for lava tubes in the Azores. Phelps et al. (2016) used surface features of the landscape to predict high diversity bat caves in Bohol Island, Philippines. Using principal component analysis, they found that the third eigenvector, loading on low surface disturbance, was a good predictor of bat species richness. Christman et al. (2016) showed that it was possible to predict the presence or absence of different taxonomic groups of troglobionts and stygobionts using surface features. However, the predictors were complex and varied from group to group, limiting its utility in conservation planning.

Michel et al. (2009) have investigated the optimal reserve design for stygobionts for a large part of Europe—Belgium, France, Italy, Portugal, Slovenia, and Spain. They placed a 12´ × 12´ grid of 4675 cells over the study area and assigned 10 183 records of 1059 stygobiotic species to these cells. Of the 4675 cells, 1280 (27.4 per cent) had one or more stygobionts. It is characteristic of data on subterranean fauna that a relatively small fraction of the area actually has stygobionts or troglobionts. Using a much larger (and irregular) grid size of US counties, Culver et al. (2000) found that only 16.5 per cent of 3112 counties had a stygobiont or troglobiont, but they did not include stygobionts from interstitial aquifers, unlike Michel et al. (2009).

Michel and colleagues used three different criteria to determine a network of 128 cells (10 per cent of the cells with stygobionts and 2.7 per cent of the total cells). One preserve design was determined by finding the 128 cells that had the most stygobionts (the species richness criterion); the second preserve design was determined by finding the 128 cells that had the most single-cell endemic stygobionts (the endemics criterion), and the third preserve design was determined by finding the 128 cells that, in aggregate, had the most stygobionts represented at least once (the complementarity criterion, Fig. 10.3). The complementarity criterion included the most species, and nearly as many endemics as the endemic criterion (Table 10.3). The complementarity preserve design was more fragmented than the other two, and included a broader geographical spread. For example, a hotspot of stygobiotic richness occurs in several grid cells in southwest Slovenia. Nearly all of these grid cells are included in the richness and endemism design, but fewer are included in the complementarity design. The complementarity design (p.242) includes a cell in Sicily, a region of only moderate richness, and the other two do not (Fig. 10.3). A preserve of only 2.7 per cent of the total area of the six countries could include nearly 80 per cent of all known stygobionts. Culver et al. (2000) and Izquierdo et al. (2001) also found that reserves designed around complementarity principles required relatively little land area in the United States and Canary Islands, respectively.


Conservation and Protection of Subterranean Habitats

Fig. 10.3 Distribution of 128 cells selected by three methods to represent the groundwater fauna of Europe: A) richness hotspots; B) endemism hotspots; C) complementarity areas. See Table 10.3 for information on completeness of representation. From Michel et al. (2009). Used with permission of Blackwell Publishing.

Table 10.3 Percent of stygobionts and percent of single-cell endemics included in preserve designs with 128 2´ × 2´ cells.

Selection criteria

Percent of total species

Percent of endemics

Species richness









There were 1059 stygobionts and 464 single-cell endemics.

Source: Data from Michel et al. (2009).

10.6 Protection strategies

There are two international agreements that have been used to protect subterranean sites. The Convention on Wetlands is an intergovernmental treaty, commonly known as the Ramsar Convention. It is a global treaty on conservation and sustainable use of wetlands as natural resources. Technical and policy guidelines are available to assist countries in protecting their Ramsar wetlands (Beltram 2004). The number of karst wetlands has been growing steadily. As of February 2018 there were 130 sites listed under the wetland type ‘karst and other subterranean hydrological systems’ (https://rsis.ramsar.org). Several of the sites are rich in stygobionts and troglobionts, including Škocjanske jame in Slovenia, and the Baradala/Domica transboundary cave system in Hungary and Slovakia.

A second international list of importance for the protection of subterranean habitats is the World Heritage Site list, developed under the auspices of UNESCO. Natural heritage sites are nominated by the countries involved, who also agree to continue to protect their integrity and provide access to all peoples. In return UNESCO provides support for the restoration and protection of sites (Hamilton-Smith 2004). Among the sites are large parks with significant karst features, such as Halong Bay in Vietnam, and Rocky Mountain Parks in Canada, as well as cave-focused sites, such as Carlsbad Caverns National Park, New Mexico, USA. Among the sites with high subterranean biodiversity on the list are Gunung Mulu National Park in Malaysia, Škocjanske jame in Slovenia, Mammoth Cave in Kentucky, and Durmitor National Park in Montenegro.

At the national level, there are a variety of designations offering varying levels of protection. One of the most interesting examples of this is the Danube Floodplain National Park in Austria, created to protect the highly diverse interstitial aquifer, the Lobau wetlands (Danielopol and Pospisil 2001; Table 9.3). Non-governmental organizations, especially The Nature Conservancy in the United States, have also been active in protecting subterranean sites, either by outright purchase or purchase of development rights. Small groups of individuals interested in cave protection have also been active in (p.244) the United States and elsewhere, in some cases buying caves outright, as evidenced by the many small cave conservancies that have sprung up in the United States.

While most efforts at protection have been focused on the protection of sites, such as the European Union Habitats Directive, the US Endangered Species Act focuses on individual species rather than sites or communities. Because of the high levels of endemism, nearly the entire subterranean fauna is at risk. In practice, threats are unevenly distributed taxonomically and geographically, and because of the lengthy process of petitioning to have a species listed and the difficulty in adding species to the list, it is in areas of imminent threat where most of the species are listed. Of the five bat species and subspecies listed from the continental United States, four are ones that hibernate in caves. Species in springs, usually not stygobionts, are well represented on the list. For example, six species of pupfish (Cyprinodon), all of which occur in desert springs and sinkholes in the southwestern United States, are on the list. This reflects their vulnerability to excessive groundwater extraction and the consequent drop in the water table. Of the 30 stygobionts and troglobionts on the list, 15, mostly troglobionts, are from the state of Texas. Stygobionts tend to be found in deeper caves and especially in wells that tap into the Edwards Aquifer (see Chapter 9). This is not because endemism levels are higher in Texas, but rather because the major cave region—the Balcones Escarpment/Edwards Aquifer—faces much higher development pressure than other US karst areas. This development pressure comes from the growing urban areas of San Antonio and Austin, and the area in between. Thus, three groups of subterranean species in the United States are at high risk—species in springs (especially in arid regions), bats hibernating in caves, and troglobionts in Texas.

European countries often have a variety of legal protections of subterranean habitats and fauna (Tercafs 2001). One the leading countries in this regard is Slovenia, especially the 2004 Zakon o ohranjanju narave (Nature Conservation Act), which parallels the European Union Habitats Directive and the 2004 Zakon o varstvu podzemnih jam (Cave Protection Act), which is unique to Slovenia and specific for cave and cave fauna protection. Protections are in place for both species and habitat types, with strict controls on access and collecting. Additional protections for especially important sites, such as Postojna–Planina Cave System (PPCS), are in place at the local level as well.

10.7 Preserve design

As humans enter caves through an entrance, much of the focus of protection efforts for cave fauna has been on the protection of the entrance of the cave. There are numerous examples of caves thought to be protected when the (p.245) entire protection strategy consists of a gate to control human access, a gate that often alters the movement of animals (and other organic carbon) in and out of the cave. A gate is often critical to providing protection for hibernating and maternity colonies of bats, provided it is of the appropriate construction (Elliott 2012; Fig. 10.1) and one that allows the movement of organic carbon. A gate provides no protection against habitat destruction or alterations in water quality or quantity.

A more appropriate starting place for preserve design is to first consider what the species and communities are that are to be protected. It is often the case that more than one species or community is of conservation interest. In fluvial aquifers, the benthic community as well as the interstitial community is often in need of protection. At least in the United States, the fauna of freshwater streams and rivers is often at risk (Master et al. 1998), and protection efforts should involve both. In caves with important bat populations, there are often important populations of troglobionts and stygobionts. In addition, karst areas often harbour rare plants, ones adapted to the thin, acidic soils and exposed rock outcrops present in karst. For example, the discovery of a new species of clover endemic to a threatened karst area in Virginia made for a stronger case for land acquisition and changed the design of the preserve.

Subterranean species and communities of conservation interest fall into four broad categories, categories that dictate different preserve designs. The first category is that of streams, both surface and subsurface, and the associated terrestrial biota along the banks and flood plain. This includes fluvial interstitial aquifers, cave streams, terrestrial riparian communities along cave streams, and even the hypotelminorheic, albeit at a smaller scale (Fig. 1.15). The area of concern is the upstream part of the drainage basin. In a karst basin, this includes the drainage area of surface streams that sink into the karst (Fig. 4.1). It is rare indeed when an entire basin is completely protected but the basin must be the focus of protection.

The second category is one where the direction of movement of both water and organic carbon is vertical, and includes some interstitial aquifers, epikarst (Fig. 1.7), MSS (Fig. 1.16), and lava tubes (Fig. 1.12). Here the protection focus is the immediate area around the site, such as sink holes above a cave with an epikarst fauna of conservation interest.

The third category is that of deep aquifers, accessed either by wells or caves. For this category, the area of concern is often more nebulous, and often quite large. For example, the US threatened isopod species Antrolana lira occurs in groundwater along a linear distance of more than 200 km in the Shenandoah Valley of Virginia and West Virginia (Holsinger et al. 1994). Protection of this species requires protection of the regional aquifer. On the other hand, flow rates in deep aquifers are often in the range of centimetres per day and so contaminant plumes move very slowly (Heath 1982).

(p.246) The fourth category brings us back to cave entrances. Some of the terrestrial cave fauna is dependent on nutrients that come into the system from the entrance. The cricket–beetle interactions discussed in Chapter 5 are an example of such a system. Bats and other species that periodically enter and exit caves are also dependent on the entrance. Once these animals leave the cave, they need a place to forage. The terrestrial fauna of Robber Baron Cave in Texas is disappearing not because the cave is not protected but because there is not an adequate foraging area for crickets that form the base of the food web. In the case of crickets, Taylor et al. (2005) found that they used a foraging radius of 100 m around a cave entrance. For species dependent on the carcasses of pack rats, racoons and the like (Table 2.5), they would require a sufficient area for these mammals to be present as well.

10.8 Summary

A critical factor in the biology of the subterranean fauna and one that increases the risk of extinction is its geographical rarity. Endemism even at the scale of a single site is common. Some but not all stygobionts and troglobionts are also numerically rare. Subterranean organisms are also at increased risk of extinction because of low reproductive rates, increased sensitivity to environmental stress, and in the case of bats, because of their propensity to cluster in large numbers in a few caves. Threats to the subterranean fauna are of four general kinds—alteration of the physical habitat (such as quarrying, especially in southeast Asia), changes in water quality and quantity (a global problem), direct changes to the subterranean fauna (such as the effects of human visitation on cave faunas), and global warming. The selection of sites for preservation requires detailed inventory data, but available evidence suggests that a majority of species can be protected at least at one site and that a relatively small percentage of total land area is required. A variety of mechanisms are available for site protection, including listing as a Ramsar wetland and as a UNESCO World Heritage Site. Design of a preserve is highly dependent on the nature of the fauna being protected.


(1) Formerly Plecotus townsendii virginianus

(2) Formerly Geomyces destructans