Skip to main content

Fishes from Toxic Springs Reveal Evolution at the Limits

Studies of fishes that inhabit toxic sulfide springs reveal mechanisms of natural selection

Credit:

Mark Ross

On a September afternoon in Tabasco in southern Mexico, the two of us made our way through the rain forest toward the sound of flowing water, in pursuit of a small but important fish. Iridescent blue morpho butterflies flitted by, and howler monkeys roared from the trees overhead, offering welcome distractions from the broiling heat and humidity. Soon we spotted a green kingfisher diving into the nearby creek and then returning to its perch to consume its catch. The bird had nabbed the same kind of fish we were after: an Atlantic molly (Poecilia mexicana), a member of a family of fishes called poeciliids, whose females give birth to live young and whose males have flashy colors that make them popular among aquarists around the world.

For a moment, we remembered with longing our fieldwork during the previous days, when we had studied Atlantic mollies at a locale only few kilometers away, a site we call Arroyo Cristal for its crystal-clear waters. Our research there had been a pleasure—we could sit on large stones and logs, dangling our legs in the water to cool off while our study species swam right between our feet.

Fieldwork on this particular day was bound to be different, however. Long before we actually reached our destination, the smell of rotten eggs filled the air, and the waters that now slowly came into view were hardly clear. Instead they were turbid and milky-white from the high concentration of sulfur particles suspended in them. Reaching the water's edge, we saw that all the submerged rocks were coated in slimy sulfur bacteria, and the many fish in the smelly waters were hanging out at the surface with their mouths agape, seeming to gasp for air.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


A newcomer would have found it hard to believe we were looking at the same species we had observed in Arroyo Cristal the day before. After all, the habitat here was drastically different, and these fish had far larger heads and showed that distinctive surface-breathing behavior. But there was no doubt about it, we had finally arrived at our field site for the day: El Azufre, a small creek with naturally toxic levels of hydrogen sulfide (H2S) and Atlantic mollies that have evolved to survive in them.

H2S-rich environments can kill most nonadapted organisms, including humans, within minutes or seconds. Given the toxicity of sulfide waters, it should come as no surprise that scientists have long been fascinated with the organisms that inhabit them. Indeed, researchers have been studying sulfide-adapted poeciliid fishes since the 1960s. But the past 15 years have seen a surge of scientific studies on the ecology and evolution of these creatures, thanks in large part to advances in genome-sequencing technology that have enabled investigators to see, at a molecular level, how organisms adapt to environmental challenges. By combining field observations of these fishes with analyses of their DNA, we and our collaborators have gained fascinating new insights into the inner workings of natural selection, a key mechanism of evolution. In addition to pulling back the curtain on natural selection, this research is allowing scientists to explore the limits of adaptation in fishes. Armed with that information, we might one day be able to forecast the fate of species in the face of pollution and other human-mediated habitat alterations.

Extreme Habitats

Poeciliid fishes are not the only organisms to adapt to seemingly inhospitable conditions. Our planet contains a variety of extreme environments—from scalding thermal springs and highly pressurized ocean depths to salt deserts and sunless subterranean caves—and all of them harbor life-forms. Still, sulfidic waters are especially hostile to life. Hydrogen sulfide is a widespread toxicant that can enter the environment naturally at freshwater sulfide springs, hydrothermal vents on the ocean floor, or coastal mudflats and salt marshes. Naturally occurring H2S can either stem from geologic activity that releases the toxic gas from deep within the earth—as occurs in the sulfide creeks we are studying—or from the decay of large amounts of organic matter. It can also enter aquatic environments as a pollutant from human activities, such as paper milling, leather tanning, and production of natural gas and geothermal power. Even very small amounts of H2S are acutely toxic to most animals because the compound binds freely available oxygen in the environment, which deprives them of breathable oxygen, and it blocks the activity of the hemoglobin protein, which transports oxygen in the blood. All this activity results in death by suffocation. H2S also blocks the process by which cells extract energy from food. Compounding these dangers, it freely penetrates cell membranes in the delicate gill tissues of fish, so they do not have to consume it to suffer harm. Humans also risk harm when inhaling H2S, thus we either try to keep our exposure time to H2S short, or we wear protective gear when working in the vicinity of sulfide sources for extended periods. Not surprisingly, both natural and anthropogenic pulses of H2S discharged in aquatic habitats around the world have caused mass mortalities in fishes and other organisms.

 

NATURALLY TOXIC creek in southern Mexico called El Azufre gets its milky appearance from suspended sulfur particles. Credit: Courtesy of Matthias Schulte

Yet various members of the teleost group of fishes—the dominant group in today's oceans and freshwater habitats and the group to which the poeciliids belong—have adapted to environmental H2S. Some of the most intriguing of these species, including sinuous eelpouts and flounderlike flatfish, live around hydrothermal vents and cold vents on the ocean floor. Reaching these animals is difficult and expensive, however, requiring the use of robust submarines, which precludes most of the experimental work we are interested in. Our own research has therefore focused on the more readily reachable poeciliids, particularly the more than 10 different species of mollies, guppies, swordtails and mosquitofish that have independently colonized dozens of toxic sulfide springs in small creeks and rivers across the New World.

Coping Mechanisms

Sulfide-adapted poeciliids have evolved a number of traits that enable them to thrive in their toxic environs. Some of these traits are behavioral. For instance, in response to the low oxygen availability in sulfidic water, the fishes spend a lot of time near the surface, where they can exploit the more highly oxygenated topmost layer of the water column. (Although the fishes practicing this so-called surface respiration look like they are gulping air, they cannot actually do that. Instead they are gulping oxygen-rich water.) The behavior has a price in that it limits the time left for other activities, such as foraging, but it helps them get the oxygen they need.

The limited availability of oxygen in these toxic waters has also shaped physical characteristics of these fishes. Most conspicuously, sulfide-spring populations have significantly larger heads than their counterparts from nonsulfidic habitats. This enlargement of the head stems from an expansion of the gill region, which helps the fishes take in more oxygen. One particular sulfide-adapted species, the sulfur molly (Poecilia sulphuraria), which is endemic to a few sulfide springs in Tabasco and Chiapas in Mexico, has also evolved odd-looking lower lip appendages to further aid oxygen intake. Similar protuberances are also found in nonpoeciliid fishes from various low-oxygen environments around the globe and are thought to facilitate the skimming of the uppermost water level for its oxygen by expanding the surface area of the mouth region.

In addition to evolving traits that enhance oxygen intake, fishes that dwell in sulfide springs have undergone adaptations that help them detoxify H2S. All animals make an enzyme called sulfide:quinone oxidoreductase (SQR) that allows them to detoxify very low concentrations of H2S by binding to it and forming nontoxic compounds. But once concentrations of the toxicant get too high, as occurs in sulfide springs, the enzymes cannot catch it all, and the excess starts interfering with cells' ability to produce energy. Our fishes have evolved modifications to the SQR pathway that enable detoxification at higher concentrations of H2S.

Sulfide-adapted poeciliids also give birth to much larger babies than their counterparts that inhabit nontoxic environs. Although the larger size of the babies means that the fishes have fewer offspring, the strategy makes sense for their environmental conditions. An increase in size results in a larger increase in volume, relative to a smaller increase in surface area. Thus, slightly larger offspring will have a higher volume-to-surface-area ratio. This arrangement is beneficial because it makes more body tissue available to detoxify the incoming H2S while only slightly increasing the body surface exposed to the toxin.

Perhaps the most striking thing about sulfide-adapted poeciliids is that they share many of the same adaptations. We have found that sulfide populations across various species and geographical regions have evolved the same novel traits when compared with their ancestors living in the surrounding sulfide-free waters.

The high degree of similarity among these separate lineages of sulfide-adapted fishes raises an intriguing question: Did the populations of poeciliids that have repeatedly and independently adapted to H2S undergo the same DNA changes in evolving their shared adaptive traits, or did they acquire the traits via different molecular pathways? Working with Markus Pfenninger of the Biodiversity and Climate Research Center in Frankfurt, Germany, and several other colleagues, we decided to find out. We analyzed DNA from several hundred Atlantic mollies from two population pairs—each of which consisted of one sulfide-adapted group and its nonadapted ancestors—from two parallel river drainages in southern Mexico. (The same river drainage can have both sulfide-rich and sulfide-free tributaries.) Statistical methods allowed us to infer how many variants exist for any given gene throughout the genome. They also allowed us to determine which variants showed signs of being driven to high frequency in these populations by natural selection—that is, by aiding survival and reproduction—as opposed to becoming abundant by chance.

We found that the genomic changes in one sulfur-adapted population tended to be unique to that population and not shared with the other. We then ran the genes that differed between the two sulfide-adapted populations through a database that lists the functions and interactions of various genes. It turns out that even though the exact genes that have been altered differ by population, most of them are involved in the regulation of the same so-called metabolic pathways—chemical reactions that support life. (Metabolism in this sense refers not only to how fishes burn energy from food but also to the actions different proteins in the biochemical machinery take to keep them alive, which could be involved in all kinds of adaptations.) What our data suggest, then, is that there are many genetic routes to evolving similar adaptations to an environmental stressor.

Credit: Mapping Specialists (map); Jillian Walters (illustrations); Source: “Parallel Evolution of Cox Genes in H2S-Tolerant Fish as Key Adaptation to a Toxic Environment,” by Markus Pfenninger et al., in Nature Communications, Vol. 5, Article No. 3873. Published Online May 12, 2014

A recent study by Joanna Kelley of Washington State University, Michael Tobler of Kansas State University and their colleagues further supports this notion. They found that patterns of gene expression—the use of genes to make proteins and certain other molecules—differed from population to population among the sulfide-tolerant Atlantic mollies of southern Mexico. But expressions of genes involved in regulating metabolic pathways were all elevated to roughly the same degree across the board. This pattern of gene activity mirrors what we see in the gene sequences themselves: the fish have followed different molecular paths to the same solutions to the problem of living in toxic waters.

Studies of sulfur-adapted poeciliids also bear on another fundamental question in evolutionary biology. Whereas some specialists have argued that populations exposed to the same stressors should undergo fairly similar evolutionary changes, others have contended that the exact sequence of evolution could affect the outcome. The thinking behind the latter idea is that if certain, randomly arising mutations represent key adaptations, they should spread fast in the respective sulfide-adapted population. Different initial key adaptations could then affect the subsequent evolutionary trajectory of a given population by altering the selective advantage of mutations arising at a later stage. Our results support this notion. We looked at three populations of sulfide-adapted Atlantic mollies and found that, in two of them, sulfide resistance had evolved in a gene that makes a key protein called cytochrome c-oxidase (COX) and is involved in generating the cell's main energy source. The third population did not acquire this initial key adaptation and had to come up with another evolutionary solution to protect its energy-making process from the toxic sulfide.

Hotbeds of Evolution

The challenges of colonizing such a hellacious environment are formidable enough that one might reasonably wonder why on earth natural selection would ever favor such an undertaking. But there is a major upside: an absence of most other species, including other fish predators and competitors for food. In sulfidic creeks in southern Mexico, for example, only specialized poeciliids can be found—none of the numerous other fishes from surrounding waters are present.

In fact, as forbidding as these sulfidic waters are, they may actually foster, rather than stifle, the evolution of new life-forms. In the conventional view of speciation, prolonged separation of formerly connected populations by geographical barriers allows the populations to evolve along their own evolutionary trajectories until they become so different from one another that they qualify as different species. But biologists are uncovering increasing evidence that adaptation to divergent ecological conditions can promote speciation even in the absence of such barriers. Observations of sulfide-adapted poeciliids bolster this scenario.

We have found that adaptation to one habitat type, be it sulfidic or nonsulfidic, limits the potential for fishes to move freely into the other habitat type. This form of natural selection essentially leads to sulfide-adapted fishes occurring only in sulfidic sites, and vice versa, even when the two habitats are separated by only a few dozen to a few hundred meters.

Other factors are important in creating and maintaining reproductive isolation in these systems, such as predation (maladapted individuals fall victim more easily to predators). If migration between habitats does occur or if members of the different populations, known as ecotypes, meet in mixing zones between sulfidic and nonsulfidic habitats, the fishes will not interbreed. Mate choice experiments have revealed that females in nonsulfidic waters prefer to mate with males of their own ecotype. Whether or not (and to what degree) females exhibit such a preference seems to depend on the strength of natural selection: we found that females' preferences for their own ecotype were stronger when natural selection against migrating sulfide-adapted males was weak. It seems that when females had a higher likelihood of encountering alien males—and thus of producing unfit hybrid offspring—they evolved a stronger aversion to the outsiders. In contrast, when natural selection against migrants was strong, and the likelihood of encountering them was therefore low, the females were unlikely to evolve an aversion to the alien males.

The exact number of poeciliids that evolved new species while adapting to sulfide springs is uncertain because, in many cases, we do not yet know how far genetic differentiation has progressed or if interbreeding with neighboring populations still occurs. But some of the sulfide-adapted lineages that show all these adaptations are approximately 100,000 years old—quite young in evolutionary terms. That they have evolved their distinguishing characteristics and achieved a certain degree of reproductive isolation from their neighbors in nontoxic waters in a relatively short time hints that the extreme conditions of the sulfidic waters may actually hasten speciation. A recent study of ours supports this notion. We found that the degree of reproductive isolation across sulfide-adapted poeciliids directly correlates with the concentration of H2S toxicity in each ecosystem.

If the poeciliids have been able to quickly evolve adaptations to a natural toxicant, are they equipped to adapt to toxic pollution from human activities? A study published last year in Science by Noah Reid, now at the University of Connecticut, and his colleagues found that killifish (which belong to a different family of fishes than the poeciliids, albeit a related one) from polluted sites in North America were capable of repeated rapid evolutionary adaptation to toxic pollution from industrial complexes. The authors suggest this might be the result of a large amount of genetic variation in killifish, which gave them a lot of preexisting genetic tools to “choose” from in adapting to new selective pressures from pollution. Whether or not the poeciliids are similarly equipped is not yet fully understood, although our research suggests that new DNA mutations are more important to these fishes than standing genetic variation is. But, taken together, the killifish research and our own findings do seem to indicate that at least a few relatively small and fast-lived fishes that produce several generations a year might, under certain conditions, be able to adapt even to some of the drastic environmental change stemming from human activities.

Many questions remain. For example, we do not yet understand why the presence of H2S has led to predictable adaptations and reproductive isolation in some ecosystems but not others. But techniques for DNA sequencing are rapidly improving, and costs are steadily decreasing. Given these trends, along with the recent publications of the genomes of several poeciliid species, we expect to soon make great gains in our understanding of the genetic mechanisms governing the shared and unique patterns of evolution in these deadly waters.

MORE TO EXPLORE

Colonisation of Toxic Environments Drives Predictable Life-History Evolution in Livebearing Fishes (Poeciliidae). Rüdiger Riesch et al. in Ecology Letters, Vol. 17, No. 1, pages 65–71; January 2014.

Parallel Evolution of COX Genes in H2S-Tolerant Fish as Key Adaptation to a Toxic Environment. Markus Pfenninger et al. in Nature Communications, Vol. 5, Article No. 3873. Published online May 12, 2014.

Unique Evolutionary Trajectories in Repeated Adaptation to Hydrogen Sulphide-Toxic Habitats of a Neotropical Fish (Poecilia mexicana). Markus Pfenninger et al. in Molecular Ecology, Vol. 24, No. 21, pages 5446–5459; November 2015.

FROM OUR ARCHIVES

Extreme Evolution. Axel Meyer; April 2015.

About Rüdiger Riesch

Rüdiger Riesch is a lecturer in evolutionary biology at Royal Holloway, University of London. His research focuses on the mechanisms that create, maintain and constrain biodiversity, with a special emphasis on speciation that occurs as a result of a population exploiting a new ecological niche.

More by Rüdiger Riesch

Martin Plath is a professor of basic and applied zoology at Northwest A&F University in Yangling, China. He studies behavior, behavioral evolution, local adaptation and speciation.

More by Martin Plath
Scientific American Magazine Vol 316 Issue 4This article was originally published with the title “Evolution at the Limits” in Scientific American Magazine Vol. 316 No. 4 (), p. 54
doi:10.1038/scientificamerican0417-54