118 High-Temperature Ecosystems individual size and solar-heated habitats Here the maximum temperatures tolerated somewhat exceed 43 1C The metazoans characteristic of hot springs are the arachnids (water mites), insects (mostly flies), small mollusks (small pulmonate snails), odonates (dragon flies), and, occasionally, ostracods These groups are neither characteristic of the cold water streams into which the thermal effluents drain, nor of running water systems in general Rather, they are related to (sometimes the same species as) the invertebrates found in the shallow muddy margins of lakes and ponds, where ambient daily maximum temperatures, due to solar heating, are of the same order as those in which the thermal inhabitants are found If these low to moderate temperatures are not set by a physicochemical threshold, then the problem is to explain why, in thermal habitats, the species have not evolved more thermal tolerance The second factor affecting distribution (mentioned earlier), competition for scarce resources, seems not to be a factor In every case studied, the grazers on the algae of thermal ecosystems are less tolerant of high temperature than the cyanobacteria on which they feed Examples are the ostracods in Hunter Hot Springs, Oregon, and the ephydrid flies of the ‘‘alkaline’’ thermal flows in Yellowstone Park This leaves the interaction of resources and selection determining fitness Under random sexual mating (with respect to temperature), the direction of the evolution of temperature tolerance will be determined by the relative numbers of surviving offspring from the upper half of the temperature range relative to those from the lower half But because of the nonlinear cooling curve of water, the size of the upper temperature part of the range will always be smaller than that of the lower half Thus under a tactic of shifting range, thermophily cannot evolve unless there is nonrandom mating or fecundity or survival is lower at low temperatures The randomness of mating with respect to temperature differs in the various groups of hot springs invertebrates, depending on life history and behavior In the flies and dragonflies, mating occurs after emergence and the animals are highly motile, so randomness with respect to temperature of development is assured The water mites are also highly dispersed as parasitic larvae, but they spend their adult lives within a relative narrow temperature range and mate there The smaller mite, Partnuniella, tends to spend most of its adult life in cooler patches of the cyanobacteria/ flexibacteria mat where it feeds on the eggs of the ephydrid flies and, as expected, has a rather low temperature tolerance The large predaceous mite, Thermacarus, lives directly in the hot water and has an upper temperature tolerance to 10 1C higher than the smaller species of mite and just lower than the most temperature tolerant metazoan, the ostracod This group has no widespread dispersal mechanism, thus nonrandom mating is predicted They also seem to be limited by food Not surprisingly, these organisms have evolved the highest tolerance to high temperature of any metazoan In summary, current evidence supports the following statements concerning the evolution of thermophily: (a) the heterotrophic procaryotic bacteria seem not to have any physicochemical limit, having evolved to inhabit geothermal systems of any temperature as long as liquid water is present; (b) the photosynthetic cyanobacteria seem to have evolved to the point (73 to 75 1C tolerance) where a physicochemical limit has been reached; and (c) eucaryotes vary greatly in temperature tolerance, depending on the group, but none has evolved tolerances equal to the maxima exhibited by the procaryotes In most cases, the observed limits can be explained on the basis of simple selection/fitness processes, without invoking unknown physicochemical mechanisms The Biodiversity of Thermal Ecosystems Species in, using, or around (living from) geothermal ecosystems throughout the world have been the subject of many studies The earliest intensive work was on the fauna of the hot springs of Iceland The algae of several hot springs in western India were cataloged in the 1960s The algae and fauna of warm springs in New Zealand have also been studied, as has the fauna of thermal springs in the Dutch East Indies The thermal effluents of Yellowstone National Park in the United States have been the subject of both the largest number and most intensive studies of thermophiles of any area in the world The original studies concentrated on the fauna but also included some observations of the algae (cyanobacteria) Beginning in 1964, detailed studies of the microorganisms living in the hot spring waters of Yellowstone Park were begun These became both the most extensive survey and intensive experimental protocol yet done on thermal microbiology At first the focus was on high temperature systems where only one or a few species were found in any given thermal ecosystem In 1967, I and my colleagues and students began studying the lower temperature thermal ecosystems with higher diversities of microorganisms and a richer food chain involving species of animal acting as scavengers, predators, and parasites Additional unpublished information has been obtained from 1983 to the present on the thermal ecosystems developing at temperatures below 45 1C Thus the ecology of the Yellowstone ecosystems can begin to be understood from the standpoints of species diversity and food-chain relationships In this final section I review what is known about the diversity of these systems, in their various manifestations of temperature and pH, and compare the results with the thermal biodiversity of the deep sea thermal vents, insofar as the latter are currently understood High temperature is the most difficult of extreme environmental conditions to which organisms can attempt to adapt There are many ways to adapt to exist in extreme cold or aridity and resume growth and activity during short seasonal changes in solar warming and precipitation High salinity can be avoided or salt excreted so extremes can still support a reasonably high diversity Indeed, many saline ecosystems may appear simple from the standpoint of vascular plant species and the food chains they support, but be relatively high in the diversity of algae and bacteria in the surface sediments Thus, in every thermal system investigated to date, the biodiversity decreases rapidly with increases in temperature At their most diverse, thermal system species are measured in the hundreds; their least diverse manifestations are monospecific Temperatures near the boiling point of water in Yellowstone thermal systems and elsewhere (85 1C–100 1C) are most common in source pools fed by deep subterranean springs of hot water These pools vary widely in chemical characteristics,