Rewild Ibera Part II: What We Have and What We Have Lost

            South America once had the highest diversity of megafauna (defined here as animal species which regularly exceed 1,000 kilograms in weight) on earth. The Mesopotamian regions of Paraguay and Argentina, including the area we now know as the humid Chaco and the Ibera wetlands, had the greatest diversity of megafauna in South America. The disappearance of large animals in these areas corresponds with the initial arrival and subsequent spread of modern humans (Homo sapiens) into the continent, beginning around 14,000 years ago. Even in pre-colonial times, only a couple dozen species with an average weight greater than 10 kilograms survived in the region, representing roughly a third of the original number, which exceeded sixty. This loss in diversity has had serious implications for the ecological functionality of the region. Larger animals are often keystone species, and their absence can lead to serious imbalances in the ecosystem as a whole. This can be seen in modern times with the extirpation of wolves, elephants, and other important large species from areas where they were once plentiful. It is my hope that the large animal diversity in the humid Chaco can be expanded past its historical number to over thirty species, which would result in a more diverse, productive, and functional ecosystem, with the added bonus of increased ecotourism potential.

            Obviously the return of species which were historically driven out of Ibera should be a priority, but further down the line it might be worth considering the use of a prehistoric baseline. By that I mean that we could try to restore not just the ecosystem functionality we have lost since the arrival of Europeans, but also some that was lost in the original colonization of the continent. This is not so straightforward as it might be in North America or Europe, where many of the extinct megafauna have close living relatives, but there are some potential candidates that I believe might warrant consideration. Obviously this is a controversial issue, and I would stress again that I have no affiliation with the current humid Chaco rewilding projects, like Proyecto Ibera. I also understand that my choice of candidate species will probably be seen as too conservative or possibly inappropriate by those already on board with the prehistoric baseline concept, as that is typically how the conversation of “Pleistocene Rewilding” devolves. I am assuming for this article that full “de-extinction” of most South American megafauna will never be practical, due to their existence in mostly warm, wet environments, which are hostile to DNA preservation, and their lack of similarly-sized close relatives.

            Probably the best place to start is with species which had conspecific relatives living in the area during prehistoric times, of which there are two. The first of these is the wild llama, or guanaco (Lama guanicoe), a medium-sized camelid still found in many areas of the continent, but which has lost considerable portions of its distribution, especially in humid areas. During the Pleistocene and early Holocene, guanacos were widespread throughout the entirety of southern South America. Today they are relegated to the more mountainous and steppe-like habitats of the south and west, having been extirpated from the east and northeast. A small population survives in the dry Chaco to the west, but the humid Chaco, Cerrado, and Caatinga ecoregions have lost their populations of guanacos. The Pampas ecoregion has lost of most of its guanacos as well. Well-suited to many different types of environment, guanacos could be a promising candidate for reintroduction, given their previous presence in Argentine Mesopotamia, alongside related extinct genera such as Hemiauchenia and Palaeolama. Camelids are important herbivores due to their general willingness to eat just about anything, especially shrubs and spiny plants not eaten by other herbivores, helping to open up grasslands for use by other species. Population management would likely be unnecessary so long as healthy populations of jaguar (Panthera onca) and puma (Puma concolor) are present, as both these species will regularly prey on ungulates of this size. Aside from the pampas deer (Ozotoceros bezoarticus) and the greater rhea (Rhea americana), the humid Chaco is lacking in specialist grassland herbivores, and the introduction of the guanaco would be a step towards rounding out this guild. 

The second candidate species extirpated in prehistoric times is the wild horse (Equus caballus). The subgenus Amerhippus, to which many extinct native North and South American horses were previously assigned, is now known to have been indistinct from living wild and domestic caballine horses. This includes the extinct variety known from the open habitats of eastern and southern South America, Equus (Amerhippus) neogeus, now perhaps more appropriately referred to as Equus caballus neogeus or Equus ferus neogeus, depending on one’s preference in nomenclature. Knowing this, the widescale rewilding of living horses, already an ongoing process in areas of Argentina, is simply a case of subspecific substitution. There would, however, need to be some discussion as to what variety of horse would be best.

There are already numerous herds of feral horses of the Criollo type scattered across the country. Criollos are an old breed, descended from primitive Spanish horses brought by the colonials. They do well in any number of environments, including wetlands, and are known for their hardiness. Their general appearance is variable, but the traits most commonly associated with primitive ancestral horses are all represented, including an acceptable and practical shoulder height of 12-15 hands (~120-150 cm), a robust and athletic build, and a stout head with a strong jawline. The coat colour possessed by extinct American horses is unclear, but the bay dun pattern exemplified by wild Asiatic horses (also known as takhis, Equus caballus przewalski) and many domestic breeds is likely ancestral, based on its occurrence in the Asinus equids, and was probably possessed by extinct horses as well. This coat colour is common and actively bred for in Criollo horses, as are the associated primitive markings such as the dorsal line, mealy mouth, and leg stripes. If new feral populations were to be formed, one strategy might be to select more primitive-looking animals from existing herds as founding stock. In this way a more phenotypically consistent form could be acquired, with a less heterogenous appearance and minimal domestic markings, such as white patches. This is not necessary for the horses to be ecologically effective, but it would be desirable for them to appear “wild” considering their future role in ecotourism-based ventures. It might also be advisable to cross other breeds, such as the Brazilian Pantaneiro, into these primarily Criollo herds in order to promote genetic diversity and reinforce desired phenotypes. An extreme example of this might be to incorporate genetic material from Asiatic wild horses via artificial insemination, since these have never been domesticated and retain behavioural and physiological traits that were likely shared by extinct wild horses. Using purebred horses of this type would be impractical, due to their adaptations to more arid habitats and their general rarity. By using crossbreeds we can combine the local adaptations of the Criollo with the robustness and untamed nature of truly wild horses. Such a breeding program would theoretically culminate in the creation of a local breed well-suited for unassisted living and ecotourism.

There are some obstacles to establishing horses as a part of a well-balanced natural system. The first is that, as domestic animals they might be entitled to veterinary care or extra feeding, and this is rather counterintuitive to the general rewilding ethos, i.e., minimal intervention. The other obstacle is that, upon their introduction, horses would become the largest prey species in the humid Chaco, with an average weight slightly higher than that of the current largest animal in the region, the lowland tapir (Tapirus terrestris). This is an obstacle because horses reproduce quickly, and without predators capable of regularly taking down an animal of this size, their numbers might soon rise to ecologically hazardous levels. In some areas of the Americas, pumas are effective predators of foals, to the point that they effectively slow population growth by limiting the number of animals which reach adulthood. Attacks on horses by jaguars have also been recorded, though there is not much geographic overlap between jaguars and existing feral horse populations, so these instances are likely from attacks on tame animals. There are reports that jaguars have a specialized predation method for horse however, which is noteworthy. A healthy population of pumas and jaguars would likely go a long way towards controlling a local population of horses, with foal predation by smaller predators contributing, but if not then either horses should not be used or other control options must be pursued.

            There are certainly benefits to returning horses to the landscape. Horses have a unique grazing dynamic compared to other herbivores, being able to digest longer and less nutritionally rich grasses than ruminant herbivores. Their method of clipping grasses instead of pulling them up also results in the creation of mosaic landscapes where there are grasses of different lengths, allowing for higher biodiversity in a small area. Much like tapirs, their inefficient digestive systems make them excellent seed dispersers. There were browsing equids present in South America as well during the Pleistocene-Holocene (Hippidion principale), but they leave no close living relatives or ecological equivalents. Much like the guanacos, horses will help to open up grasslands for use by other herbivores, and we may begin to see a system where the guanacos and horses enter new areas first and clear them of shrubs and dead grasses, allowing new grasses to grow and be eaten by deer and rhea.

            Speaking of deer, they warrant some consideration here as well, for although there are now four species of deer living in the humid Chaco there were, until recently, seven. Three genera: Antifer, Epieurycerus, and Morenelaphus were present in prehistoric times before their extinction. With seven species of deer, all descended from similar North American ancestors, niche separation would have to have been very well enforced. We see this in the living four species, with the marsh deer (Blastocerus dichotomus) being specialized for marshlands, the red brocket (Mazama americana) for woodlands, the pampas deer (Ozotoceros bezoarticus) for grasslands, and the grey brocket (Mazama gouazoubira) for shrublands. As far as I can determine, the extinct species seem also to have occupied these habitats, but were generally of larger sizes than those they would have been sympatric with. Epieurycerus, like the grey brocket, occupied shrublands, but was more comparable in size to the larger marsh deer, and so presumably had access to, or preference for, different forage. Antifer and Morenelaphus generally were found in more open habitats, like the pampas deer, but again there was separation based on size, with Ozotoceros typically weighing around 30 kilograms, Morenelaphus around 50 kilograms, and Antifer around 125 kilograms. 

All these deer would have belonged to the Odocoilinae subfamily of the deer group which, in addition to the living South American species, still has members outside of the continent. Three of these, Capreolus, Hydropotes, and Odocoileus, are found in similar climates. Capreolus, the roe deer, is likely similar in role and habitat to the red brocket, and consequently not a desirable introduction. Water deer, Hydropotes, are adapted to similar habitats as marsh deer, but are much smaller. I wouldn’t consider it as a purposeful introduction, but it is worth noting that it has already been introduced to several areas of Argentina, as have various non-Odocoiline deer such as the chital (Axis axis). The chital is not closely related to any of the extinct deer species mentioned, but its weight, antler conformation, and habitat preference are perhaps not dissimilar from Morenelaphus, and their potential equivalency should be investigated. Though theoretically an invasive species, a native population of large cats would help to control chital numbers, and there is no real need for targeted extermination. 

Another angle is the genus Odocoileus, which includes mule deer and whitetail deer. Whitetail deer (Odocoileus virginianus) are found across North and Central America, as well as in northern South America, where they coexist with red and grey brockets. A large, generalist species suited to humid subtropical climates and not unlike the basal Pliocene Odocoileus from which South American deer descend, the whitetail could stand in as a sort of combined proxy for several of the larger extinct species. Eventually this species might specialize into similar roles, being already of a similar maximum size to Antifer and Epieurycerus. Of course much research and experimentation would be necessary to better understand these connections, and the current deer diversity is certainly not low. Deer are the only ruminant herbivores in the ecoregion, discounting pseudo-ruminants like peccaries and guanacos, and will be filling a role similar to antelope in African megafaunal systems. That being considered, a cervid diversity at least comparable, if not equivalent, to that of the prehistoric Chaco might be desirable for repairing grazing dynamics. The correct course of action in this regard remains unclear, and further research and discussion would be productive. 

            The reintroduction of collared peccaries (Pecari tajacu) and white-lipped peccaries (Tayassu pecari) to Ibera will be an important part of reinstating seed-dispersal and herbivory regimes, but there may yet have been a third species of peccary native to the humid Chaco, Catagonus stenocephalus. The Chacoan peccary, Catagonus wagneri, is the only member of this genus that survives. It was previously more widespread, with fossils found in the Pampas region of Argentina and Uruguay, in addition to the dry Chaco where they still occur. Catagonus stenocephalus seems to have had an even wider distribution, being also found in the Cerrado ecoregion of Brazil. The extinct species seems to have been significantly larger than the living one, and they were seemingly sympatric in many environments. Knowing this, Catagonus stenocephalus cannot quite be replaced by Catagonus wagneri, as their roles must have differed in some way. However, it is conceivable, based on its fossil distribution, that Chacoan peccaries might actually have occurred in the humid Chaco at one time, making them a potential reintroduction candidate. They were extirpated from most of the dry Chaco in prehistoric times, save the northern part, and are notable for being only known from fossils before the 1970s, when they were discovered alive for the first time. While range expansion would certainly be a boon to this species—which is currently experiencing a drop in numbers from habitat destruction—the humid Chaco may not be suitable. The Chacoan peccary seems highly adapted to arid environments, with a diet of mostly cactus and little need for drinking water, and they may find themselves unable to thrive in wetter environments. Probably the best move is to establish corridors between dry and humid Chaco habitats, allowing exchange between the two habitat types for those species capable of living in both. The potential equivalency of feral pigs (Sus scrofa), numerous in Ibera and the rest of the humid Chaco, to some of the larger extinct peccaries might also be worth investigating, though peccaries are typically more herbivorous than pigs, and are not as closely related as their appearance would suggest.

Aside from the guanacos, there are other living camelids which might warrant tentative consideration, given the previously high diversity of the group in the region. Wild alpacas, or vicunas (Vicugna vicugna), are found in the highland regions of western South America. In prehistoric times, they were also common in many of the same arid grassland habitats where guanacos can be found, including the dry Chaco and the Pampas. This suggests a greater adaptability than is generally ascribed to this species, but it is unclear if they ranged into the more humid habitats to the northeast. Hemiauchenia and Palaeolama certainly occurred there, but these were much larger than either the vicuna or guanaco, reaching weights of over 300 kilograms. The largest living camelids are Old World camels, and these reach even larger weights, with the largest Bactrian camels (Camelus bactrianus) weighing as much as 1000 kilograms. Though heavier than either Hemiauchenia or Paleolama, Dromedaries (Camelus dromedarius) might play a similar role as a large, open-country browser. Much like Vicugna and Catagonus, however, their ability to thrive in wet environments is questionable. They seem like ideal candidates for rewilding in the dry Chaco, and both they and the vicunas might benefit from the corridor model I suggested earlier. In addition to being an ecological substitute for the larger New World camelids, there might be some niche overlap with the larger of the extinct Litoptern ungulates, such as Macrauchenia and Xenorhinotherium. These species were similarly sized and potentially similar ecologically, likely being mixed-feeders/browsers adapted to open habitats. As domestic livestock, dromedaries could probably be sourced from within Argentina, but like horses they would have many of the same legal obstacles to establishing feral herds, as well as many of the same issues surrounding population management by predators.

            In my last article, I discussed the possibility of creating a population of red-footed tortoises (Chelonoidis carbonaria) in Ibera. What I neglected to mention was that Pleistocene-era fossils have been found in the humid Chaco (Corrientes) of a much larger species, Chelonoidis lutzae, comparable in mass to the related and extant Galapagos giant tortoises (Chelonoidis nigra). Tortoises of this size, which can reach weights of over 400 kilograms, were once common on every continent but Australia (where Meiolanid land turtles filled a similar role) and Antarctica. These species disappeared following the expansion of hominids from Africa, with some extinctions actually coinciding with the arrival of pre-modern humans such as Homo erectus. Whereas most large animals require some sort of skill and technology to hunt, tortoises only need to be caught, carried, and popped in the fire. When they are too large to carry, you can just build the fire around them. In recorded history, truly giant tortoises are only known from island chains which were never settled by humans, such as the Galapagos, Mascarenes, and Seychelles, with the largest extant mainland tortoise being only a quarter the size. The possibility of rewilding giant tortoises in areas that they previously occupied is not a new concept. Extinct island varieties in the Indian Ocean have had their niches approximated by introduced extant species, and similar plans have been proposed for Madagascar. Captive tortoises of the Centrochelys and Stigmochelys genera, the two largest mainland genera, are being used as trophic surrogates for a completely unrelated but ecologically similar variety of extinct flightless duck on the island of Kauai. Islands represent very low-risk experimental sites for trophic rewilding, since we generally have pretty good records of what their complete faunal complement was, and which species filled what niche. Tortoises are generally pretty good candidates for rewilding because they are easy to manage, are unlikely to become invasive, and are important grazers and seed-dispersers.

What complicates the issue is that, while we know a lot about the ecology and population dynamics of giant tortoises from islands, we know very little about their extinct relatives. For example, one of the current issues with giant tortoise conservation is predation on eggs and juveniles by invasive mammals, especially rats, which descend from continental ancestors. I am inclined to believe that this is partially due to the unnaturally high density of rats on islands, where otherwise they would be controlled through predation and competition. I think it is also likely that Chelonoidis lutzae simply did not live in the same high numbers as giant tortoises on islands do/did, due to most of the eggs and hatchlings being eaten. Once they reach a certain size, giant tortoises are basically immune to predation by non-humans, and can live for well over a century, giving them multiple opportunities to reproduce. Extant mainland tortoises theoretically experience all the same problems with predation, and consequently are not so densely distributed as island varieties. I am unaware of any behavioural differences between island and mainland tortoises, and living giant tortoises seem to retain all of the morphological adaptations to predator avoidance. I think if we wanted to investigate the possibility of using existing giant tortoises, for example Chelonoidis nigra, as an ecological proxy for Chelonoidis lutzae, the best strategy would simply be to establish an ex situ breeding population in Argentina, where the dietary choices and behaviours of giant tortoises in a humid Chaco biome could be observed in a managed, predator-controlled environment. This would of course be secondary to the preliminary introduction of red-footed tortoises, and dietary overlap between the two varieties would be a necessary research subject. Probably most overlap would be between juvenile giant tortoises and adult red-footed tortoises, as adult giant tortoises would be as much as twenty times larger than the smaller species, giving them access to higher-reaching forage and larger fruits.

Currently there are only two native predators in the humid Chaco were are capable of regularly taking down large game: the puma and the jaguar. In prehistoric times, however, the large predator guild of the area was much more diverse. The largest member of this group, and possibly one of the largest mammalian land carnivores of all time, was Arctotherium angustidens, a giant and hyper-carnivorous bear that could weigh over 2,000 kilograms—over twice as massive as the largest recorded polar bears (Ursus maritimus), the current heaviest land carnivore. While this species has no living equivalents, its smaller relative, Arctotherium wingei, might. The smaller species was only about 200 kilograms and was mostly herbivorous, strikingly similar to its closest living relative and currently the only South American bear species, the spectacled bear (Tremarctos ornatus). A very adaptable species, the spectacled bear is only found in the mountainous regions of western South America, but within that region it occupies deserts, shrublands, cloud forests, alpine grasslands, tropical rainforests, and many other habitats. Its likely forebear (pardon the pun), Tremarctos floridanus, was common in the southeastern USA, an area climatically very similar to the humid Chaco, before the Pleistocene-Holocene extinctions. Spectacled bears would likely thrive in similar habitats if given the chance, serving as both a seed-distributing herbivore and an occasional predator of small and large game species. While mostly eating fruit, bamboo, nuts, and invertebrates, spectacled bears are also known to attack and kill deer, guanacos, and even horses and tapirs, in addition to smaller animals like birds and rodents. Predation on large species is rare but potentially significant. They will also exploit many tough, fibrous vegetation sources that other herbivores will not, as well as competing with other carnivores for carrion sources. Arctotherium is now known from genetic studies to have descended from Tremarctos, likely from an earlier colonization event. Expanding the range of the extant species would perhaps pave the road for iterative evolution, wherein two separate colonization events by a related lineage result in the evolution of the same form twice over, through the action of similar evolutionary processes.

            In addition to bears and large cats, the Pleistocene humid Chaco hosted two genera of large, pack-hunting canids, Protocyon and Theriodictis, which were closely related to the extant bush dog (Speothos venaticus) and maned wolf (Chrysocyon brachyurus). Both of these living species are found in the humid Chaco, the former only in the northern and eastern margins, but neither is particularly similar ecologically to the two extinct species. Neither are the two additional, less closely-related canids in the area, Lycalopex and Cerdocyon, the South American “foxes”. Bush dogs are pack-hunters, but are significantly smaller than the extinct forms, and poorly suited to open habitats, being more comfortable hunting large rodents and peccaries in more heavily wooded environments. Maned wolves are large and well-adapted to open grasslands and savannahs, but are also solitary and omnivorous, enjoying fruit and tubers in addition to game. The “foxes” are diminutive in size and fill a more peripheral hunting and scavenging role, as foxes and jackals do elsewhere.

Consequently, if we did want to reintroduce large, pack-hunting canids to humid Chaco ecosystems, we would need to consider other options. One angle is to simply allow feral dogs (Canis lupus familiaris), already common in Argentina, to fulfil a similar function, much as dingoes (Canis lupus dingo) do in Australia. The problem with this is that, much like feral cats (Felis lybica catus), the behaviour of feral dogs is in many ways dissimilar to that of their wild counterparts. Feral dogs often exploit additional (usually anthropogenic) resources on top of what they can obtain in the wild, allowing them to maintain a higher population density than their potential prey can tolerate. Their neotenous behaviours, resulting from millennia of selection for more infantile animals, often leads to rather unnatural activities in adults. An example of this would be the act of chasing prey for fun rather than to kill, the frequent occurrence of which leads to overly stressed prey. The familiarity of feral dogs with humans also makes them much more comfortable approaching humans and human settlements, and consequently more likely to prey on livestock than their wild equivalents. All this considered, I would like to propose that wild canids be sourced from elsewhere.

Wolves seem an obvious candidate, being truly wild dogs, though with this option one would have to be careful about which subspecies was chosen. Most living wolf populations are from temperate or arctic climates, rather than subtropical ones, where they have largely been extirpated. If they could be obtained, wolves from the south-eastern USA, eastern China, or northern India might be suitable, but these populations have been largely eliminated. Of course with wolves there would also be the concern that they might mate with feral dogs, especially if the smaller, more primitive subspecies were used. I would suggest instead that either African wild dogs (Lycaon pictus) or Asian wild dogs (Cuon alpinus) might be more suitable, since both are still common in humid subtropical environments, and both share some particular dental morphologies with bush dogs and the two extinct wild dogs, seemingly through convergent evolution. They also both seem better able to tolerate the presence of large cats than wolves. Where wolves and Asiatic wild dogs, or dholes, are sympatric, wolves tend to dominate, unless tigers and/or leopards (Panthera tigris and Panthera pardus) are prevalent, in which case dholes fair far better than wolves.  Both are endangered would benefit from a range expansion, but I would argue that dholes would probably benefit more, since African wild dogs have more opportunities within Africa, with its many well-established parks and reserves.

            Dholes would theoretically help to round out the existing predator guild by filling the niche of Protocyon and Theriodictis. Knowing there were two large canids in the area we could investigate using more than one species. The conservative angle of using one is perhaps preferable, considering that in most places the two extinct species were not sympatric—the Chaco region being the exception. Dhole fossils have been found in Mexico, and it is conceivable that if this species had not been extirpated from North America during the Pleistocene-Holocene extinctions, it might have eventually colonized South America as well. I think wolves are less likely to have done so, despite also occurring in North America, due to their lower tolerance for tropical climates. Dholes are generally of lower impact to humans and livestock than are wolves or feral dogs. They would also be the only primarily cursorial predators in an ecosystem dominated by ambush predators. This tendency would potentially make them more effective predators of fast, open-country game like guanacos, pampas deer, and rhea. Introduction of a new large predator would obviously be a controversial step, but it wouldn’t be done without extensive modelling and experimentation. The assumption is that South American game should retain behavioural adaptations to similar predators, with which they previously existed and which has occasionally been reinforced by extant canids, but this may prove false.

Probably the extinctions that had the most profound ecological effects are that of the giant megaherbivores, by which I mean large plant-eaters that regularly weigh more than 1,000 kilograms. The humid Chaco ecoregion was previously home to numerous varieties of ground sloth, grazing armadillos, the afore-mentioned litopterns, rhinoceros-like toxodons, and elephant-like gomphotheres. All of these were likely keystone species—like living rhinos and elephants—playing vital roles in herbivory and disturbance regimes. Many South American plants are considered “evolutionary anachronisms”, due to the fact they retain features adapted to the presence of these megaherbivores, such as thorns that grow out of reach of living herbivores, or fruits with seeds too large to be swallowed by living herbivores. To my knowledge, a specific analysis of potential anachronistic plants has not been conducted for the humid Chaco. Considering the previously high density of native megaherbivores in the region, and the generally high number of anachronisms found elsewhere on the continent, I would be very surprised if the ecoregion did not have at least a few to be identified. Many small animals may also have directly suffered from the extinction of the megaherbivores, such as dung beetles, cowbirds, and carrion eaters like vultures and caracaras.

            If we wanted to reintroduce megaherbivores to the humid Chaco, where would we start? There are seven living genera of terrestrial mammals which regularly exceed weights of 1,000 kilograms. These include the giraffe (Giraffa), hippopotamus (Hippopotomus), rhinoceroses (Ceratotherium, Diceros, and Rhinoceros, but not Dicerorhinus), and the elephants (Elephas and Loxodonta). For the purpose of this thought-experiment, let us consider two of these groups as potential candidates for trophic rewilding in the humid Chaco: the greater one-horned rhinoceros (Rhinoceros unicornis) and the Asian elephant (Elephas maximus). These two species were chosen for multiple reasons, the most important of which are 1) they are both numerous or semi-numerous in captivity, 2) they are both found in similar climates to those found in the humid Chaco, i.e., humid subtropical climates as defined by the Koppen-Geiger classification system, and 3) they are non-African taxa which are generally afforded fewer protections in some of the areas where they occur, with both species being more threatened than their African equivalents. The rhinoceros and elephant would be brought in as rough equivalents for the toxodons (Toxodon platensis) and gomphotheres (Notiomastodon platensis) respectively, but likely their effects would not be exactly equivalent in either case and may overlap with other groups of extinct megafauna, such as the many species of grazing armadillos (Glyptodontidae and Pampatheriidae) and ground sloths (Megatheriidae and Mylodontidae). Elephants are in the same superfamily as the gomphotheres, and morphological similarity suggests ecological similarity in this case, especially since the global pattern of proboscidean biogeography in the past was that modern elephants were replacing the more primitive gomphotheres as they migrated out of Africa and across Asia and the Americas. Columbian mammoths (Mammuthus columbi), closely related to living Asian elephants, had colonized much of Central America before their extinction, and it is conceivable that, if not for human intervention, they might eventually have colonized South America, displacing gomphotheres in the process. Rhinoceroses are less closely related to toxodons, but odd-toed ungulates (Perissodactyla), which includes rhinoceroses, are known to have been a sister group to the extinct South American ungulates (Notoungulata), which included the toxodons. Through convergent evolution, the two taxa developed similar morphologies, and were likely similar in general ecological function as well. Whether or not toxodons were amphibious or fully terrestrial is a contested point, but one-horned rhinoceroses are somewhere in between, and so are perhaps more appropriate than fully terrestrial rhinoceros species. Another reason these two species were chosen is that they might both be regarded as mixed feeders, which engage in both grazing and browsing, making their effects on the ecosystem potentially broader. There may have been as many as twelve megafaunal species living in this area at one time, and it is hoped that a fair share of their ecological effects might be duplicated by just two stand-ins.

            Much like the giant tortoises, probably the best way to study the effects of introduced megaherbivores would simply be to set up open-air captive breeding facilities on large, fenced stretches of land within Ibera or elsewhere in the humid Chaco. Much like tortoises, elephants and rhinoceroses can be effectively fenced in while still allowing the free movement of other species, though the fencing required for such an endeavour would be considerably more expensive. Such an attraction would be a great tourist draw, in addition to being good for conservation of the species, and would allow for the potential trophic consequences of megaherbivore presence to be observed. If, after a time, the effects of these species result in a greater net biodiversity, productivity, or some other measure of ecosystem functionality, then less controlled releases might become more practical. Alternatively, something akin to the South African reserve system might be more appropriate, wherein the large, potentially more dangerous fauna are more restricted in their movements than smaller species. All of this would also have to take into account the expenses associated with the prevention of poaching, and this would have to be weighed against potential revenue from ecotourism. Some of this might be mitigated by using Asian elephants and rhinoceroses, since these offer less ivory and horn material than their African counterparts, and if the money from these enterprises is being properly channelled back into the community, then locals would find the elimination of an elephant or rhinoceros to be counter-intuitive to their well-being. As we know from places where megaherbivores still occur, this relationship is probably an oversimplification. Regardless, this is a hypothetical exercise and the global poaching situation may have changed by the time any of this is implemented. There is far more work to be done with historic-baseline trophic rewilding before we ever get to that point.

There was one more large mammalian carnivore native to the area before/during human arrival, and that is the sabretooth cat (Smilodon populator). I have left discussing this taxa until after going through the megaherbivores, rather than lumping them in with the other carnivores, due to its importance in the predation of megaherbivores. Sabre-toothed cats are thought to have been specialized hunters of juvenile elephants, ground sloths, giant armadillos, and so on. If similar species like rhinoceroses and elephants became an integrated part of South American ecosystems, the ability of existing predators to prey on calves would have to be evaluated. I have no doubt that a jaguar or puma could take down a juvenile elephant or rhino, since these would be comparable in size to a capybara or tapir. However, when attacking a calf you typically also have to deal with its mother or, in the elephant’s case, the whole herd, making predation of this sort a rare event. Other large cats like tigers or lions (Panthera leo) might engage in this activity more often than other carnivores, and the presence of large cats like Smilodon and possibly the American lion (Panthera [leo?] atrox) in prehistoric South America might justify the introduction of these species. Based on the lack of information about how these large cats might interact with jaguars, I would hesitate to introduce lions or tigers without extensive preliminary research. Some might assume that jaguars are ecologically similar to leopards, which coexist with both lions and tigers, but there seems to be some distinct differences. For example, jaguars are more aquatic and less arboreal than leopards, a trait they share with tigers. The closer analogue to leopards in the area would probably be the puma. Jaguars coexisted with lions in North America, and there is growing evidence that lions also colonized South America, with fossils initially identified as giant jaguars (Panthera “onca” mesembrina) now thought to have belonged instead to lions. However, these findings have only been found in colder steppe environments, like Patagonia, and there is no evidence as of yet that lions ever ranged into the more sub-tropical/tropical open habitats. If I were to recommend the introduction of lions or tigers, it would be more likely as a way to control horses and camels than rhinos or elephants, which in the wild are controlled more through self-regulation, resource availability, or through human management by sterilization, translocation, or, in extreme cases, culling. For now, the existing large felid diversity is sufficient for control of the prey species that remain and is probably sufficient for a more diverse guild of rewilded herbivores as well.

            There are just two more large mammals that I would like to discuss: the Atlantic manatee (Trichechus manatus) and the La Plata river dolphin (Pontoporia blainvillei). Both are native South American mammals, but they do not range quite into the Chaco region; some evidence suggests that they or their relatives did at one point. We might consider these two technically marine mammals because they both inhabit freshwater river systems in addition to saltwater environments, with some populations of both species spending most of their lives in rivers. While manatees do not currently range into the Parana and Uruguay rivers, which run around Ibera and into the humid Chaco, their relatives did during warm prehistoric periods. Fossils of related, possibly ancestral species like Ribodon limbatus, have been found in the upper Parana river, dating to the Pliocene era. Pleistocene fossils from Ohio in North America suggest that during very warm periods, manatees travelled very far inland via river systems. Amazonian manatees (Trichechus inunguis) and African manatees (Trichechus senegalensis) still do this, though they live in much warmer climates. In North America, Atlantic manatees still range well up the coast as far as the Missouri river basin, of a similar latitude to the Uruguay and Parana river basin, but the water temperatures in these regions differ due to differences in ocean current activity. Manatees are very sensitive to cold temperatures, being subject to “cold stress” if they spend too much time in water colder than 20 degrees Celsius. Consequently, manatees have not ranged as far south as they have north during modern times, perhaps also impaired by human activity. However, it is possible that, with ongoing climate change, the path to the Uruguay and Parana might open up again, allowing manatees to recolonize these river systems after a long absence. Manatees are important large herbivores, just like elephants or rhinos, grazing on a variety of both freshwater and saltwater vegetation. They also eat fish and invertebrates, but this is less common. I would not be recommending an actual introduction in this case, but the past biogeography of the species is fascinating and I felt it worth discussing. Manatees are common in Florida wetlands and river systems—which are not dissimilar climatically and ecologically from the humid Chaco—and have become quite a draw for tourism, with their pleasantly content behaviour and chubby appearance affording them the affectionate nickname of “sea cow”.

            Unlike manatees, river dolphins do occupy the Uruguay and Parana rivers, though it is unclear just how far downstream they occur. Theoretically, they should have unobscured access through both rivers, but there may be constraints to their dispersal from human obstruction or differences in water temperature or prey availability. Like most dolphins, the La Plata dolphin feeds on fish, and for this purpose it has evolved the longest snout relative to its body of any living cetacean. It is a rather small dolphin, weighing around 50 kilograms, and like other river dolphins it has an unfused neck, giving it greater flexibility to manoeuvre itself around underwater obstacles. If the health and biodiversity of these river systems improves, and the dolphin populations recover from their current vulnerable status, we might see them expand further north. Much like manatees, these are a very charismatic and appealing species for tourists, given their small size and playful, intelligent demeanour. This would again be a case of expansion rather than introduction.

As I have said, all of these (re)introductions would necessarily occur after historically extirpated or threatened species have been returned and established in the region. In addition to some of the species I mentioned in my last article, there are other medium-large animals I may have missed, such as the green iguana (Iguana iguana), the tayra (Eira barbara), and the brown-throated sloth (Bradypus variegatus). I have also learned that at least one species native to the region has gone extinct in historic times, the glaucous macaw (Anodorhynchus glaucus). Much like the green-winged macaw (Ara chloropterus), which is currently in the process of reintroduction, the glaucous macaw was native to the humid Chaco, and was an important part of its ecology as a seed-distributor. Also like the green-winged macaw, it was hunted widely for its feathers and captured for the pet trade. Unfortunately, unlike the green-winged macaw, the glaucous macaw was not found in many other regions and seems to have gone extinct in the 1870s. It is possible that it survives in remote, unsurveyed areas of Bolivia which, if true, would mean that a captive population could be started and the animals potentially reintroduced. However, in the likely event that they are truly extinct, it might be possible to bring in their relatives as a proxy. Two other members of the genus survive: the hyacinth macaw (Anodorhynchus hyacinthinus) and indigo macaw (Anodorhynchus leari). Both are ecologically similar to the glaucous macaw, but the indigo macaw is much closer in general size and appearance, so much so that some taxonomists consider them disjunct populations of the same species. Knowing this, once green-winged macaws have been established, experiments with introduced indigo macaws might also be possible. It would certainly be helpful for the conservation of the species, as indigo macaws are endangered and split between only two populations in Brazil with no animals in captivity. Interestingly, members of the Anodorhynchus genus may represent another evolutionary anachronism, with both the hyacinth and indigo macaw engaging in a peculiar behaviour wherein they will follow herds of cattle and eat partially-digested seeds out of their dung. These seeds are often from species that the same birds will not eat off the trees they grow on, and it has been suggested that this behaviour evolved from millions of years of coexistence with megaherbivores. Similar observations have been made in Africa, where grey parrots will eat seeds from the dung of elephants. 

            Of all the potential prehistoric-baseline introductions discussed here, introduction may only be practical for six or seven of them. Experiments with rewilding guanacos and horses in the humid Chaco could start tomorrow if somebody were interested (and had the funding), though it can certainly wait if necessary. Spectacled bears and possibly dholes might be possible in the future, once ungulate populations are higher. Managed conservation populations of giant tortoises, rhinos, and elephants in South American natural areas is something I hope to see in my lifetime. The beginnings of such projects might be seen in large, multi-hectare centres for retired circus and zoo elephants, the first of which in South America was opened in Brazil a few years ago. Refuges like this are very important for the ongoing abolishment of elephants in the entertainment industry, and secondarily as opportunities to study how elephants interact with South American ecosystems in a non-invasive way. With refugee animals, it is important to respect their privacy and give them adequate space to recover from a lifetime of living in unsuitable conditions and being forced to give rides, do tricks, and so on. 

There remain many possibilities for rewilding in this region of the world. Argentina has taken some very progressive steps towards rewilding their most precious natural areas, and for that they should be very proud. But, as always, there is more work to be done, as we strive for a more vibrant and diverse world than the one we born into.