Why Ratnapura is the centre of the story

Most famous Sri Lankan megafaunal fossils come from the Sabaragamuwa Basin around Ratnapura and Kuruwita. Gem mining cuts into river-laid gravels known locally as illama. Those sediments trap resistant objects: quartz, gems, mineralised teeth and fragments of bone. For palaeontologists, the pits are accidental windows into the Quaternary, the interval from about 2.58 million years ago to the present.

The collective name Ratnapura Fauna is useful, but it can mislead. It sounds like a single ecological community frozen at one instant. It is not. River erosion can remove a fossil from an older deposit, transport it and bury it beside a much younger bone. P. E. P. Deraniyagala demonstrated this problem as early as 1955 when uranium tests indicated that hippo and Asian-elephant fossils found in the gem sands were not contemporaneous. A modern palaeoart study likewise stresses that extensive redeposition has disturbed original stratigraphy and still prevents a secure chronology for many specimens.

That means a hippo tooth, rhino molar and elephant fragment from the same gravel layer do not automatically prove a hippo, rhino and elephant met on the same riverbank. The safest reconstruction is a changing sequence of faunas moving through Sri Lanka across repeated climatic and sea-level cycles.

The scientists and miners behind the discoveries

Paulus Edward Pieris Deraniyagala (1900–1976), zoologist, palaeontologist and later director of the National Museum of Ceylon, established much of the vocabulary still used for the Ratnapura fauna. From the 1930s through the 1960s he described fossil rhinoceroses, elephants, hippopotamuses, crocodiles and the tooth he named Panthera leo sinhaleyus. His work was pioneering, but it was done with small samples and before CT scanning, modern stratigraphic recording and large digital comparison sets. Some of his species names remain useful hypotheses rather than settled modern taxonomy.

The story did not end with one researcher. In 2005 Kelum Manamendra-Arachchi, Rohan Pethiyagoda, Rajith Dissanayake and Madhava Meegaskumbura remeasured Sri Lanka's large-cat fossils and published the tentative tiger identification. In 2023 Noel Amano and an international Sri Lankan-led archaeological team reported the Early Holocene rhino fragment and analysed animal diets through stable carbon isotopes. In 2026 Jason Kennedy and Aravinda Ravibhanu Sumanarathna produced evidence-explicit 3D reconstructions while documenting which anatomical decisions were measured and which remained speculative.

Gem miners, not professional fossil crews, have exposed a large share of the material. Shafts and pits reach the gem-bearing gravel; earth is lifted, washed and sorted, and unusually heavy teeth or bones may appear among the stones. This creates opportunity and loss at the same time. A miner can rescue a fossil that would otherwise remain buried, but its exact depth, orientation and sedimentary relationship may never be recorded. Researchers therefore need the specimen and its discovery context: mine location, layer, depth, photographs and associated sediment samples.

An island repeatedly connected to India

During glacial periods, water locked in continental ice lowered global sea level. At the Last Glacial Maximum, around 20,000 years ago, sea level was roughly 120–130 metres below the modern level. The shallow Palk Strait became a broad terrestrial connection or chain of connected lowlands between southern India and northern Sri Lanka. The most recent research on Sri Lanka's coastal prehistory describes this exposed land bridge directly.

That pathway explains how large land mammals could reach the island. It does not mean every low-sea-level interval was equally usable. Rainfall, vegetation and freshwater availability determined whether a corridor was suitable for a wet-forest tiger, an open-country grazer or a river-dependent hippo. Genetic research on living Sri Lankan animals also shows that physical connection did not always produce unlimited movement: ecological barriers could isolate wet-zone lineages even while dry land existed to the north.

The rhinoceroses: the strongest surprise, with a taxonomic warning

There is no serious doubt that rhinoceroses lived in prehistoric Sri Lanka. The uncertainty concerns exactly which species were present and how the old names map onto modern rhinoceros taxonomy. Historical literature recognises two forms, Rhinoceros sinhaleyus and Rhinoceros kagavena, mainly from teeth. The first has relatively low-crowned, squarish molars; the second was described with higher-crowned teeth. Such differences might reflect separate species and diets, but small samples, incomplete provenance and missing specimens make reassessment difficult.

The genus Rhinoceros contains the living greater one-horned and Javan rhinoceroses. Both have one nasal horn, not the two-horned silhouette familiar from African rhinos. The fossil Sri Lankan form was described as somewhat smaller than the living greater one-horned rhinoceros. Its body covering, exact horn size and detailed proportions are not preserved, so any life reconstruction must borrow cautiously from living and better-known fossil relatives.

The most revealing new evidence is much younger than the classic Ratnapura finds. A 2023 stable-isotope study reported a broken upper rhino tooth from Early Holocene deposits at Fa-Hien Lena. It was not directly dated, but it lay between archaeological layers bracketed by radiocarbon ages of approximately 12,575–12,150 and 8,595–7,791 calibrated years before present. Its carbon-isotope value suggests feeding in a mixed forest/open environment. This is currently the youngest reported rhinoceros occurrence in Sri Lanka and places people and rhinos in the same broad landscape long after the peak of the last ice age.

Importantly, the Fa-Hien fragment cannot confidently be assigned to the historically named R. sinhaleyus. The correct conclusion is that a rhinoceros survived into the Early Holocene—not that a particular named species has a precisely known extinction date.

The six-incisor hippo: an Asian river giant

Hippos are now restricted naturally to Africa, but hippopotamids once ranged widely across Asia. Sri Lanka's fossils belong to the Asian lineage generally placed in Hexaprotodon, a name meaning “six front teeth.” Literature variously calls the local form Hexaprotodon sinhaleyus, H. sivalensis sinhaleyus or simply H. sivalensis. These changing combinations reflect the history of classification, not evidence for several different Sri Lankan hippos.

Teeth are especially valuable because their enamel survives conditions that destroy most skeletons. Hippo remains from Ratnapura imply permanent rivers, floodplains, ponds or marshes capable of supporting a large semiaquatic herbivore. They do not imply that the entire island was a swamp. Palaeoenvironmental evidence instead points to a mosaic of humid forest, wetlands, forest edges and open patches.

The age of the hippo fossils is hard to pin down because of gravel redeposition. Older literature placed much of the Ratnapura fauna in broad Middle or Late Pleistocene bins, but those labels should not be converted into an exact extinction year. The honest statement is that hippos formed part of Sri Lanka's Pleistocene megafauna and disappeared before the securely documented late prehistoric fauna of the island.

More than one kind of elephant

Sri Lanka's elephant story is not simply the unbroken presence of the modern Sri Lankan Asian elephant. Fossil workers have reported several forms from Ratnapura deposits, using names that changed as elephant systematics developed. The most spectacular is the straight-tusked elephant Palaeoloxodon namadicus, widespread in Pleistocene South Asia. Sri Lankan isolated molars were named the subspecies P. n. sinhaleyus because they appeared smaller than mainland specimens.

That subspecies is plausible but not beyond dispute. The defining material is sparse, some historically described fossils are now lost, and tooth size can vary with sex, age and individual variation. There is no complete Sri Lankan skull or skeleton from which to reconstruct a unique island form. A 2026 peer-reviewed reconstruction therefore used Sri Lankan measurements where possible but relied heavily on mainland P. namadicus and other Palaeoloxodon anatomy. Its proposed moderate island size reduction is a testable hypothesis, not a measured fact.

Other Ratnapura elephant material has been assigned to archaic elephant forms and to Elephas maximus, the Asian elephant. The crucial lesson from the 1955 dating work is that an Asian-elephant fossil found beside hippo material may be younger and redeposited. Sri Lanka's river gravels have blended several chapters of elephant history into one deposit.

The “Sri Lankan lion”: real tooth, cautious identity

In 1939 P. E. P. Deraniyagala named Panthera leo sinhaleyus from large-cat teeth recovered near Kuruwita. The holotype—the specimen that anchors the name—is a left lower carnassial, the blade-like shearing tooth of a carnivore. A damaged canine was also referred to the form but is not diagnostic enough to carry the case.

Distinguishing lion and tiger lower carnassials is difficult. A 2005 morphometric comparison concluded that the Kuruwita tooth groups with lions rather than tigers, and a 2025 re-examination again interpreted its crown proportions as lion-like. That makes a lion identification a defensible scientific hypothesis, not merely folklore inspired by Sri Lanka's lion symbolism.

But the popular reconstruction often races far ahead of the fossil. From one usable tooth, scientists cannot know mane development, coat colour, social system, total geographic range or whether it was larger than every modern Asiatic lion. Nor does the evidence securely demonstrate that it deserves subspecies rank. It is best described as a fossil lion record historically named Panthera leo sinhaleyus, with the taxonomic status still open to better material and modern analysis.

The tiger: younger, better dated—and still tentative

The tiger case rests on two specimens analysed in 2005: a complete left lower carnassial found in Ratnapura alluvium in 1962 and a right middle toe bone recovered from a prehistoric midden at Batadomba-lena in 1982. Measurements separated them from Sri Lankan leopards and aligned them more closely with tigers than lions. The authors nevertheless wrote “tentatively considered to be a tiger,” wording that should be preserved whenever the claim is repeated.

The Batadomba-lena bone is important because its archaeological context was radiocarbon dated to about 16,500 years before present. It therefore provides a much firmer time marker than most alluvial Ratnapura fossils. A bone in a human midden does not automatically prove that hunters killed the animal at the cave; it could have been scavenged, traded or carried as a curiosity. Sri Lankan archaeological syntheses explicitly mention such “manuport” possibilities.

If the identification is correct, tigers crossed from India during a period when low sea level exposed the Palk connection and suitable habitat formed a viable corridor. They then disappeared from Sri Lanka while surviving on the mainland. The fossil record cannot yet tell us whether the island population was large, long-lived or genetically distinctive.

Did lions and tigers live together?

Possibly, but no Sri Lankan site yet proves direct coexistence. The 2005 big-cat study argued that the lion disappeared before the tiger and probably before culturally modern humans occupied the island's wet-zone caves. It also corrected earlier claims of lion remains in the Batadomba-lena midden: the identifiable big-cat toe bone from that context was assigned to tiger, not lion.

Because the lion tooth comes from disturbed alluvium without a secure direct date, a clean timeline is impossible. A responsible illustration should not stage a lion–tiger confrontation as if it were an observed event. The strongest current statement is that Sri Lanka preserves evidence interpreted as two extinct large Panthera records in addition to the leopard, with the tiger demonstrably present by the Late Pleistocene and the lion's chronology much less secure.

Gaur, buffalo, deer, wild cattle—and reptiles

The lost fauna extended beyond headline megafauna. Fossils have been attributed to gaur (Bos gaurus or the historically named B. g. sinhaleyus), wild buffalo, wild boar, several deer and extinct bovines named from fragmentary remains. Gaur still live in India and Southeast Asia but no longer occur naturally in Sri Lanka. These large herbivores would have occupied different parts of a habitat mosaic and supplied prey or carrion for big cats, crocodiles and humans.

Historical Ratnapura lists also include an extinct crocodile, Crocodylus sinhaleyus, and named fossil terrapins. As with the mammals, older subspecies and species names should not be treated as equally secure merely because they appear in a catalogue. Many were erected in an era of small comparative samples. Modern revision may synonymise some, uphold others or leave them unresolved.

What prehistoric people actually encountered

Sri Lanka has some of South Asia's most important evidence for rainforest-adapted Homo sapiens. Fa-Hien Lena, Batadomba-lena and Kitulgala Beli-lena preserve occupation extending tens of thousands of years into the Late Pleistocene. Their food remains are dominated not by megafauna but by monkeys, giant squirrels, porcupines, mouse-deer, civets, pangolins, reptiles, fish and molluscs. People were skilled broad-spectrum rainforest foragers.

The giant animals appear rarely in those cave assemblages. The possible tiger phalanx at Batadomba-lena and Early Holocene rhino tooth at Fa-Hien Lena are exceptional. Their rarity may reflect true scarcity, transport costs—the difficulty of carrying large carcass parts to a cave—or human choices. It prevents any simple picture of “Balangoda Man hunting the megafauna to extinction.”

Why did the giants disappear?

There is no single demonstrated Sri Lankan kill mechanism or precisely dated extinction pulse. Several pressures are plausible:

1. Rising seas ended repeated immigration. After the last ice age, inundation of the Palk lowlands isolated Sri Lankan populations from mainland source populations.
2. Climate reorganised habitats. Shifts in monsoon strength, rainfall seasonality, rivers and vegetation could shrink the particular wetland or open-forest combinations required by large specialists.
3. Island populations are vulnerable. Small, isolated populations suffer more from droughts, disease, demographic bad luck and loss of genetic exchange.
4. Humans may have added pressure. Hunting, scavenging competition, burning or disturbance could matter, especially for slow-breeding megafauna. Yet direct Sri Lankan evidence is presently too sparse to quantify the role.

These factors could act at different times on different species. Hippos may have vanished long before the last rhinos; the possible lion may have disappeared before the dated tiger; some elephant lineages died out while Elephas maximus survived. “The extinction of the Ratnapura Fauna” was probably not one event.

How certain is each claim?

The next fossil could rewrite the story

Sri Lanka's fossil record is biased toward places where people dig. Sabaragamuwa is rich partly because gem miners expose deep gravels; other regions have not received comparable palaeontological sampling. Many historic specimens lack modern stratigraphic recording, and tropical acidity destroys bone. Absence elsewhere is therefore not proof that an animal never lived there.

The highest research priorities are systematic excavations with recorded sediment layers; direct radiometric dating of teeth and bones; high-resolution CT scans; renewed comparison of museum specimens; stable-isotope work on diet and habitat; and, where preservation permits, ancient proteins or DNA. Even a securely dated jaw or limb bone could settle debates that a century of isolated teeth cannot.

The wonder of prehistoric Sri Lanka does not need exaggeration. The defensible story is already remarkable: across the Quaternary, changing land bridges and monsoon landscapes admitted a succession of Asian megafauna. Rhinos lasted surprisingly late. A tiger entered the rainforest world of Late Pleistocene people. A lone carnassial keeps the possibility of a Sri Lankan lion alive. The island known today is not a diminished museum of one vanished golden age, but the latest frame in a long, dynamic history.