Caves and associated subterranean voids harbor extraordinary ecosystems inhabited by equally remarkable animals. Insects and arachnids dominate terrestrial habitats, whereas crustaceans dominate aquatic systems. This article describes the subterranean biome, highlighting terrestrial systems and the insects that are obligately adapted to live permanently in underground voids.
DISCOVERY AND CHARACTERIZATION OF CAVE ARTHROPODS
Why an animal would abandon the lighted world and lose such adaptive characters as eyes, pigment, and dispersal ability to live permanently in perpetually damp, dark, barren caves has long fascinated both biologists and laymen. In fact, it is these pale, blind obligate cave species that one usually envisages under the rubric of cave animal, and it is this group that is featured in this article. However, numerous other animals live all or part of their life cycles in caves or are regular visitors. Although some cave insects were known in ancient times, the first scientific writings on this topic began in northern Italy in the mid-16th century with the discovery of blind aquatic crustaceans in cave streams. The science of cave biology (biospeleol-ogy) was founded in the mid- to late 19th century with studies of limestone caves in southern Europe by Schinner and continued into the early 20th century by Racovitza and colleagues. They devised the currently used classification scheme for cavernicoles, based on the degree of association with caves. Also, in the mid-1800s, obligate cave animals were discovered in Mammoth Cave, Kentucky, and in a few other North American limestone caves, but the study of North American cave faunas generally lagged behind that of Europe for the next century. The loss of eyes and other apparently adaptive characters led to a revival, circa the turn of the 20th century, of Lamarckian (i.e., acquired rather than inherited characteristics) theories to explain their evolution. During this period several major expeditions went to tropical regions to search for obligate cave faunas, but for a variety of reasons none were found or recognized. The apparent absence of tropical troglobites (obligate cave species) and the rel-ictual nature of temperate cave animals led to the development and general acceptance of the theory that these animals evolved only after populations were stranded in caves by changing climates that extinguished their surface relatives.
Troglobitic Adaptations
The most conspicuous aspect displayed by obligate cave arthropods is the reduction of structures normally considered adaptive (e.g., eyes, pigment, wings, and cuticle thickness). Compare the closely related surface and cave insects shown in Figs. 1A and 1B . Cave species also often lack a circadian rhythm and have relatively low metabolic and reproductive rates. A few characters are often enhanced, including modified structures such as increased hairiness, enlarged sensory organs, longer appendages, and specialized tarsi. These morphological, physiological, and behavioral changes allow the animals to maintain water balance, breathe unusual gas mixtures, disperse, reproduce, and locate food and other resources in their environment. The remarkable convergent evolution of troglomorphy (adaptations to caves) among unrelated cave species in different regions of the world indicates that selective pressures must be similar in all such environments.
Taxonomic Overview of Troglobites
TERRESTRIAL CAVE ARTHROPODS Insects, arachnids, and millipedes are the dominant terrestrial groups living in caves. Not all orders are represented, however. Among the Hexapoda, the orders Collembola, Orthoptera, Hemiptera, Coleoptera, and Diptera predominate. The springtails (Collembola) are represented by many troglophilic (facultative cave residents) and troglobitic species and are important scavengers in many caves. Most cavernicolous orthopter-ans are troglophilic or trogloxenic (roosting in caves), with the cave crickets (Rhiphidophoridae) being the best known. As more tropical caves are studied, many new species of troglobitic true crickets (Gryllidae) are being described. Among Hemiptera, both suborders occur in caves: true bugs (Heteroptera) and planthoppers and allies (Auchenorrhyncha or Homoptera of some classification systems). Thread-legged bugs (Reduviidae) are common troglophiles in warmer caves, and a few troglobitic forms are known from the tropics. Most species are cryptic, and many new cave species await discovery. Several other hemipteran families contain troglophilic species. Planthoppers, especially Cixiidae (Fig. 1B), are common in tropical caves. Ongoing surveys for cixiids indicate that each isolated cave system may harbor one or more cave-adapted species, and the group may be among the most speciose families in caves, rivaling even the carabid ground beetles in temperate caves.
Beetles (Coleoptera), especially the families Carabidae, Leiodidae, and Staphylinidae, are especially well represented in the temperate caves. For example, the endemic North American ground beetle genus Pseudanophthalmus contains at least 250 species, which, with one exception, are found only in caves. Flies (Diptera) are dominant troglophiles in both tropical and temperate caves, but only a few blind, flightless troglobitic species are known.
Troglobitic species are also found in the orders Diplura, Thysanura, Blattodea, Dermaptera, Grylloblattodea, Psocoptera, and Lepidoptera. Troglobitic bristletails (Diplura) occur mainly in temperate caves. Cockroaches (Blattodea) are well represented in tropical caves, and many new species await description. Only a few cave-adapted earwigs are known, and most are from oceanic islands. Grylloblattids are restricted to glaciated mountains in northwestern North America and eastern Asia. They characteristically inhabit caves and crevices; however, most species also venture outside to feed in damp surface
FIGURE 1 Cave and surface cixiid planthoppers. (A) Rain forest Oliarus species from Maui Island; note large eyes, dark color, and functional wings. (B) Adult female cave-adapted Oliarus polyphemus from Hawaii Island; note absence of eyes, enlarged antennae, and reduced wings and pigment.
habitats. Many moths habitually roost in caves, and some are troglo-philic scavengers or root feeders. A few are blind and flightless troglobites.
The arachnids are second only to the insects in numbers of terrestrial cave species. The spiders are common denizens of caves, with numerous troglobitic forms known from temperate and tropical caves. In many tropical caves, spiders instead of ground beetles are the top predators. Pseudoscorpions are also well represented in temperate and tropical caves, and over 300 cave-adapted species representing most families are known. Harvestmen (Opilionida) are more restricted in distribution, but most of the 26 families contain troglobitic species. Some surface species roost in caves in huge numbers. Mites (Acari) are often abundant and diverse in caves, especially species associated with guano. Most terrestrial cave species are troglophilic, but a few families, such as the Rhagidiidae, contain many troglobites. Cavernicolous species are also known among the palpigrades, schizo-mids, amblypygids, scorpions, and ricinuleids.
Myriapods are also well represented in caves. The millipedes are the third major group of cavernicolous arthropods, especially in temperate caves, where they are often the dominant scavengers in the ecosystem. The orders containing the most cave species are Julida, with numerous troglobites in Europe and North America; Chordeumatida and Polydesmida, with troglobites in Europe, North America, and Japan; and Callipodida, with troglobites in Europe and the Near East. Four other orders (Polyxenida, Glomerida, Spirobolida, and Spirostrepida) each have a few cave-adapted species. Cave millipedes from the tropics are still poorly known, and many new species undoubtedly await discovery. Many ground-inhabiting centipedes regularly enter caves. Whether they can live and reproduce underground is unknown for most species, but a few are troglophilic or troglobitic. The rock centipedes (Lithobiomorpha) are widespread and include several troglobitic species. A few troglobitic giant centipedes (Scolopendromorpha) are known from the tropics. An undescribed 8-cm-long Scutigerimorpha from North Queensland, Australia, is one of the largest terrestrial troglobites known.
Two groups of terrestrial Crustacea are found in caves. Isopods in the suborder Oniscidea have adapted to caves many times, especially in the Mediterranean region and in the tropics. Fourteen of the 34 recognized families contain cave species. In contrast, only a few terrestrial amphipods (Talitridae) are found in caves, and most are from islands.
AQUATIC CAVE ARTHROPODS
Aquatic subterranean habitats include underground lakes and streams, perched pools of water, water films, and water-filled phreatic aquifers. These aquatic habitats support diverse faunas of troglobitic (or stygobitic) arthropods. By far the dominant group is the crustaceans, with about 2700 cave-adapted species known worldwide. Water mites (Acari) are also well represented, especially in smaller interstitial habitats. Few insects have invaded subterranean aquatic habitats. The most successful group is the dytiscid diving beetles, several species of which are known from aquifers in Africa, Europe, North America, and Japan. Two troglobitic water bugs are known: a blind water scorpion (Nepidae) from a cave in Romania and a terrestrial water treader (Mesoveliidae) from Hawaii.
Zoogeography of Cave Arthropods
Until recently, obligate cave species were thought to occur mainly in temperate limestone caves, and the cave faunas of temperate Europe and North America are well characterized. Diverse cave faunas are also known from Japan, Tasmania, and New Zealand. However, in the past few decades discoveries of significant cave faunas in tropical
caves, lava tubes, and even fractured rock layers have revolutionized our understanding of cave biology. These findings suggest that troglo-bites have evolved wherever suitable subterranean voids are available for sufficient time. They are now known from most regions that have been appropriately investigated. Thus rather than being exceptional, cave adaptation must be a general and predictable process among animals adapting to exploit underground resources.
In hindsight, the early expeditions to the tropics missed troglobites for three main reasons. (1) The environment of caves—Troglobites are restricted to deeper, constantly moist passages. Because cave temperatures are usually near the mean annual surface temperature (MAST) over the cave and, in the tropics, the surface temperature rises above and falls below MAST almost every day, most tropical caves are subjected to drying winds created by the sinking cold nighttime air. (2) Accessibility—The higher solution rate of limestone in the tropics creates large open cave systems, exacerbating the effects of the daily drying winds and making the deeper moist cave passages, where the troglobites are found, beyond the limits of safe exploration using the equipment available at that time. In addition, the caves found and explored were often bird and bat roosts, and the biologists could fill their containers with new species without going deeper. (3) Systematics—Ironically, many troglobites were collected, but the species belonged to groups unrelated to the animals found in temperate caves, and in fact unrelated to anything the temperate-based taxonomists had seen, so their status in the cave went unrecognized. As in all fields in biology, evolutionary biology is only as good as the systematics research upon which it is based.
Each cave region is inhabited by representatives of the surface fauna currently or historically living over the caves. Only a few surface taxa within each region successfully invaded caves. In general, the surface ancestors possessed characters that facilitated their shift into underground environments; that is, they were already adapted to live in dark, moist rocky habitats and utilized food that was relatively common in caves. The chief ancestral habitats for terrestrial cave species include rocky margins of rivers, lakes, and seashores; leaf litter and moss in wet forests; and moist rocky terrains. Each cave system harbors relatively few species of troglobites; even the most diverse known fauna—that in the Postojna-Planina System, Slovenia— totals only 84 species. In North America, Mammoth Cave supports the most species (41). Among lava tubes, Bayliss Cave (North Queensland, Australia) contains the highest number (25). Because of the restricted distribution of each species, cave habitats are often likened to islands. Despite the few species found in each cave, the overall number of troglobites is quite large since subterranean habitats are much more extensive and widespread than is often assumed. Karst landscapes cover about 15% of the earth’s surface, and cavernous lava and fractured rock habitats have not been mapped but may cover another 5% or more. Submarine caves have barely been investigated, but the diverse fauna derived from marine ancestors found in anchialine systems along seacoasts indicates that caves and cave-like habitats below the seafloor may harbor diverse ecosystems at least in shallow coastal areas.
SUBTERRANEAN BIOME
Caves and Voids
Caves are subterranean voids large enough for humans to enter, but intermediate-sized voids (i.e., mesocaverns) smaller than caves but larger than capillary spaces are also important for terrestrial cave insects. Terrestrial animals rarely exploit capillary-sized spaces underground, but water-filled pore spaces (i.e., interstitial habitats) are often inhabited by numerous tiny species of stygobites. Caves and voids can form in three ways: solution, erosion, and volcanism. The largest and best known caves are dissolved in limestone, calcium carbonate. Limestone is structurally strong yet readily dissolves in weak acid, such as the small amounts of carbonic acid normally found in groundwater. The process is slow, but over millennia large interconnected systems of caves and voids can form in limestone exposed to weathering. Caves created by solution can also form in other soluble rocks, such as gypsum (hydrated calcium sulfate) and dolomite (magnesium calcium carbonate), but the caves formed are usually less stable than those in limestone. Erosional caves form during landslides and tectonic events, as well as by groundwater removing loose material from under a cap rock. Erosional caves are usually ephemeral but, in some areas, they are re-created continuously and so remain available for colonization. Tectonic caves are common on volcanoes, but lava tubes are more familiar cave features. Lava tubes form by the roofing over of lava channels during an eruption. Because the roof insulates the flow, lava tubes become efficient transporters of lava away from the vent, and long complex caves can be built over time by long-lived eruptions. Mesocavernous habitats are more extensive than caves and can be found in rock strata not suitable for supporting cave-size passages. Mesocaverns also occur in fractured rock strata and in cobbles deposited by rivers.
The terrestrial cave environment is strongly zonal (Fig. 2 ). Three zones are obvious: (1) the entrance zone where the surface and underground environments meet; (2) the twilight zone between the limits of vascular plants and total darkness; and (3) the dark zone. From biological and environmental perspectives, the dark zone can be subdivided into three zones: (a) the transition zone where short-term climatic events on the surface are still felt; (b) the deep cave zone where the atmosphere remains saturated with water vapor; and (c) the stagnant air zone where decomposition gases, especially carbon dioxide, can accumulate. The boundary between each zone is often dynamic and is determined by size, shape, orientation, and location of entrances in relation to the surface environment and size and shape of the cave passages, as well as to the climate on the surface and availability of water. Because air exchange is reduced in smaller spaces, the environment within most mesocaverns probably
FIGURE 2 Profile view of a representative cave showing the five environmental zones. Not shown to scale; length and depth are compressed. Key: D, deep zone; E, entrance zone; S, stagnant air zone; TR, transition zone; and TW, twilight zone.
remains in the stagnant air zone. Each zone often harbors a different community of organisms, with the obligate cave species found only in the inner two zones. The deep cave and stagnant air zones contain a harsh environment for most surface-dwelling organisms. It is a perpetually dark, wet, three-dimensional maze without many of the cues used by surface species and with often abnormally high concentrations of carbon dioxide. In many caves in temperate regions, the transition zone is evident only in winter when the outside temperature is below cave temperature.
Energy Sources and Nutrient Cycling in Caves
Unlike capillary spaces typical of soils, which act as filters capturing water and nutrients near the surface, caves and mesocaverns act as conduits for water and nutrients. In cavernous regions, a significant amount of organic material sinks or is carried into deeper underground voids where it is inaccessible to most species adapted to surface habitats. The principal mechanisms that transport material underground are sinking streams, percolating rainwater, trogloxenes, animals blundering into caves, and deeply penetrating plant roots. A few cave communities are known to rely on food energy created underground without the aid of sunlight by chemoautotrophic microbes. Sinking streams are more important in transporting food into limestone caves than in lava and other caves, because streams are important in creating and maintaining solution caves. Plants growing on barren rocky substrates such as lava and limestone often must send their roots deep into crevices and caves to obtain water and nutrients. Because higher temperatures result in higher rates of water loss from leaves and higher rates of leaching of tropical soils, and because there is a continuous growing season without a spring recharge of water, plant roots must penetrate deeper underground (sometimes in excess of 100 m) and are, therefore, generally more important in tropical caves than in temperate caves.
Most troglobites are detritivores or scavengers feeding on decaying organic matter and the associated microbes. Living tree roots provide food directly for several obligate cave insects. A relatively large percentage of troglobites are predators, attesting to the role of lost surface animals in bringing in food. It is these available food resources that enable the evolution of troglobites, which are highly specialized to exploit resources within medium-sized subterranean voids. They colonize or temporarily exploit cave-sized passages only where the physical environment is suitable. Most caves appear barren and therefore often are believed to be food-poor environments. However, food can be locally abundant, and exploiting such a patchy resource in a harsh, maze-like environment is probably more critical than paucity per se.
In addition to troglobites many other organisms enter caves. Many arthropods seek out caves for estivation or hibernation sites during periods of harsh weather. Some, such as agrotine moths and cave crickets, use caves for daytime retreats and sometimes oviposition sites and emerge at night to forage in the neighboring forest. Troglophilic arthropods enter to feed on guano and other organic material deposited or brought in by roosting bats, birds, crickets, and other trogloxenes. Parasites and other associates of trogloxenes also live in caves, and some of these, such as nycteribiid and streblid flies on bats, show some troglomorphies. Many leaf-litter and soil arthropods living in caves feed on accumulations of organic material left by sinking streams. These resources are usually more abundant near entrances and in the transition zone. Only a portion of the surface-inhabiting species in each region can cope with the environment and exploit these food resources. Some troglophiles apparently leave caves only to disperse to new sites, but most show no morphological adaptations to living in caves.
CONSERVATION OF CAVE LIFE
The fantastic adaptations displayed by obligate cave animals have long intrigued biologists. Their often narrow environmental tolerances, coupled with their island-like habitats, have reinforced the view that these animals are fragile, lead an endangered existence, and are in need of conservation. However, development of conservation programs is hampered by a severe lack of data about the species present and their status. Discoveries in the past few decades of cave ecosystems in a variety of cavernous rocks in diverse regions have revolutionized our understanding of cave life. We now believe that cave colonization and adaptation are general phenomena and occur wherever there are suitable underground voids available for evolutionary time. Most cave species remain undiscovered; in fact, the cave faunas of large areas containing caves, especially in the tropics, remain unsurveyed and unknown. Unfortunately, many cave systems are being destroyed before their faunas become known. The major anthropogenic threats to cave faunas include (1) mining of the surrounding rock, (2) changes in land use over subterranean habitats such as deforestation and urbanization, (3) alteration of groundwater flow patterns, (4) waste disposal and pollution, (5) invasion by nonindigenous species, (6) disruption of food inputs, and (7) direct human disturbance during visitation. Biological surveys are urgently needed. Also, recent systematic studies reveal that cave arthropod faunas are far more diverse than previously thought, indicating that priority should be focused on recognizing and protecting each distinct population rather than protecting a single population of each conventional species.
Conservation efforts must mitigate threats affecting the system, as well as recognize emerging threats. Generally, species extinctions result from novel perturbations, for example, new stresses with which a species has had little experience during its evolution. Ecological studies are needed that improve our understanding of the functioning ecosystem, as well as understanding of natural suc-cessional processes. However, experimental ecological studies in caves are problematic because in few other habitats are humans so dramatically intruders as in caves. Not only do researchers affect the environment of the passages they study, but also they cannot sample the medium-sized voids where the major activity usually occurs. Caves are a fragile window through which we can see and study the fauna living within cavernous rock. Protected areas must include a sustainable portion of the ecosystem as well as suitable source areas for food and water resources. This usually represents an area larger than the footprint of the known cave.
RESEARCH OPPORTUNITIES
The bizarre adaptations displayed by troglobites make them excellent animals for evolutionary research. Recent advances in phy-logenetic methods and molecular techniques provide important new tools for deciphering relationships among cave animals and their surface relatives. The discovery that close surface relatives are still extant for many tropical and island troglobites allows more appropriate comparative studies between species pairs adapted to wildly different environments. These studies should provide more critical understanding of how certain adaptations correlate with environmental parameters, as well as a better understanding of evolution in general. Some of these studies are in progress, for example, the work of Culver and colleagues on Gammarus minus in springs and caves in the eastern United States. Individual species of troglobites frequently have restricted distributions even within a given area of caves. Usually such a limited distribution indicates the existence of a barrier to subterranean dispersal, but not always. Critical morphological and behavioral studies, corroborated by modern molecular techniques, are showing that some troglobites thought to be widespread actually are composed of several more or less reproductively isolated populations. It has been assumed that cave adaptation was a dead end and that each of these populations evolved separately from the same or closely related surface ancestors that independently invaded caves. However, recent research by Hoch and colleagues on Hawaiian cixiid planthoppers suggests that some troglobites can disperse to new caves through underground voids and diverge into new species.
Caves are island-like habitats that support distinct ecosystems composed of communities of highly specialized organisms. Because the environment is discrete, rigorous, and easily defined, it provides an ideal system in which to conduct ecological studies. The number of species is usually manageable. The physical environment is rigidly constrained by the geological and environmental setting, and the environmental parameters can be determined with great precision because the habitat is surrounded and moderated by thick layers of rock. However, it is a rigorous, high-stress environment and difficult for humans to access and envision because it is so foreign to human experience. Also, one cannot enter or sample the mesocaverns where perhaps most cave animals live. These disadvantages can be overcome by comparing passages differing in the parameter of interest or by designing experiments that manipulate the parameter being studied in the natural environment. Biospeleology is still in the discovery phase. Although our understanding of cave biology has progressed substantially, results of future studies on evolution and ecology will be exciting and add significantly to our fascination with caves.
