Vertebrate Palaeontology


Evolution and Extinction of Mammals on Islands

[The Channel Islands mammoth (Mammuthus exilis) of the Late Pleisocene of Santa Barbara compared to its mainland relative (Mammuthus colombi). Reconstruction by George Lyras, based on casts, after the original in Santa Barbara Natural History Museum and]

Evolution of Island Mammals: Adaptation and Extinction of Placental Mammals on Islands. Wiley-Blackwell (Oxford, UK), 2010 (ISBN-13 978-1-4051-9009-1, 479 pp., index, figs, 26 full-colour plates)

More information at Wiley-Blackwell

[Order with price reduction]

Evolution on islands differs essentially from evolution on mainlands. Especially islands of the past are uniquely intriguing. Due to millions of years of isolation, exceptional and sometimes bizarre mammals evolved, such as pig-sized elephants and hippos, giant rats and gorilla-sized lemurs, formidable to their mainland ancestors. This timely and innovative book is the first of its kind to offer a much-needed synthesis of recent advances in the exciting field of the evolution and extinction of fossil insular placental mammals. It provides a comprehensive overview of current knowledge on fossil island mammals worldwide, ranging from the Oligocene to the onset of the Holocene. This book addresses evolutionary processes and key aspects of insular mammal biology, exemplified by a variety of fossil species. The authors discuss the human factor in past extinction events and loss of insular biodiversity. This accessible and richly illustrated textbook is written for graduate level students and professional researchers in evolutionary biology, palaeontology, biogeography, zoology, and ecology.

This material is based upon work published by Wiley-Blackwell, 2010 Wiley-Blackwell. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the publisher.

Dispersal to Islands

[Dwarf hippo's are among the most common islanders of the Pleistocene. Hippopotamus creutzburgi, Crete, Middle Pleistocene against a profile of the African hippo of today. Copyright Alexis Vlachos]

I do not deny that there are many and grave difficulties in understanding how several of the inhabitants of the more remote islands, whether still retaining the same specific form or modified since their arrival, could have reached their present homes. (Charles Darwin 1859: 396.)
Island faunas are influenced by several factors. These are, amongst others, the various types of islands, the ways of dispersal to the island, the distance to the mainland and island area, the faunal composition as a whole and characteristics of its elements, and physiography of the island. Naturally, these factors are interwoven and subject to changes over time, influencing each other constantly. For example, the physical geography of the island together with local and global climate put a heavy stamp on island size in terms of suitable surface area.
There are several ways for vertebrates, including mammals, to reach an island and maintain a viable population. Needless to say, the different taxa may disperse differently, so not all dispersal types apply to all taxa. Roughly speaking, three main types of dispersal routes can be distinguished, respectively over land, over water and through the air. The first type has no restriction, except for strictly aquatic taxa. The second type is open only for taxa that can swim, float or raft on a floating mass across wide water barriers, while the last type is restricted to bats. For invertebrate taxa, hitchhiking on a host during the journey has been recorded, for example freshwater snails sticking to the feathers of a bird, but this kind of dispersal is irrelevant for mammals. The total distance may be broken up into smaller units, for which the popular term island hopping is sometimes used.
The first type of dispersal—over land— is on its turn artificially divided into two routes, being corridor dispersal and filter dispersal. The corridor route applies to cases in which faunal interchange between two areas is possible while this is probable for some animals but improbable for others in the case of a filter route or filter bridge. The former route is by definition over land (land bridge), whereas the latter may include (very) short distances over water (‘step stones’, not to be confused with island hopping). Land bridges are supposed to provide the only possible way of dispersal for groups that are intolerant to salt water, such as freshwater fishes and amphibians.
Continental islands are typically colonised by a normal mainland fauna through corridors (at the start) or filters (in a later stage). During further disconnection and isolation, the filter becomes stronger and eventually no new direct influx takes place anymore. At this point the taxa are confined to the island and may undergo so-called vicariance, the effect of being separated from the rest of the original group by a geographic barrier, in this case a body of water. This separation often results in a differentiation into new varieties or species. In case an array of new species arises, the term adaptive radiation is applied.
An example of such a vicariance phenomenon is provided by the marsupials of Australia and perhaps by the ground sloths of the West Indies. Isolation not necessarily leads to speciation. In some rare cases, hardly or no change takes place at all and some sort of living fossil, or relic taxon, is found on the island, often amidst truly endemic taxa that did evolve and radiate. This is the case with the monotremata of Australia, but also with the primitive Amami rabbit (Pentalagus furnessi) of the central Ryukyu Islands. Paradoxically, islands function as laboratories for speciation, sometimes to the absurd, and at the same time as sanctuaries for the preservation of rare and primitive taxa. The ecological meltdown that took place in Barro Colorado, Central America—which became an island when in the early 20th century Charges River was dammed for the construction of the Panama Canal—provides a present-day example of the extinction and colonisation patterns on a ‘suddenly’ isolated, continental island.
The opposite situation takes place when an island gradually becomes connected to the mainland, and a mainland fauna arrives, first through filter dispersal, later through corridor dispersal. This is seen on Japan where during the Late Pleistocene a mainland fauna with large Chinese deer (Sinomegaceros) met the endemic Elephas naumanni fauna, resulting in the latter’s extinction.
For dispersals over water, read our book.


[The distribution of fossil insular species, ranging from the Miocene to the early Holocene; map and copyright G. Lyras]

Four main types of fossil insular faunas are recognised. These are referred to as balanced mainland faunas, balanced impoverished faunas, unbalanced impoverished faunas and unbalanced endemic faunas.
A typical feature of an impoverished fauna is that taxonomic diversity is poor, with many major groups absent. At the same time, though, some of the insular genera or species have undergone radiations to fill in the gaps. The lack of diversity on higher taxonomic levels is balanced as it were by higher than normal diversity on lower taxonomic levels, as compared to the mainland. A further typical feature of an impoverished fauna is the occupation of some endotherm niches by ectotherms.
The terms balanced and unbalanced are applied in relation to the number of carnivore species in relation to herbivore species. A typical mainland fauna is considered balanced, whereas typical insular faunas are devoid of terrestrial mammalian carnivores and are thus named unbalanced. This lack of mammalian carnivores leaves the niche of top predator open, which is subsequently filled by birds of prey, such as the giant eagles Garganoaetus freudenthali on Gargano and Haast’s eagle (Harpagornis moorei) in New Zealand, or reptiles, such as the Komodo dragon (Varanus komodensis) on Flores, Rinca and Komodo.
There is a difference in the understanding of the term balanced between palaeontology and ecology. The way the term is used by most palaeontologists only means that large carnivores are present and that thus all major trophic niches are occupied. In ecology a balanced fauna means it is ecologically stable, for instance the trophic niches are not only occupied, but occupied in ratios which do not encourage transitions.

Human Impact

[The impact of humans on islands: habitat fragmentation and depletion. Seychelles, Bird Island; photo M. de Loos and S. van der Geer]

The timing and cause of coordinated extinction of endemic fossil faunas, termed faunal turnovers, or of a single taxon, termed species extinction, is often poorly known, due to the incomplete fossil record. The subject is, however, of great importance, mainly because of the extensive and ongoing Holocene extinctions of islanders caused by humans, directly as well as indirectly, and climatic changes.
There are two main classes of causes for extinctions of insular endemics. Firstly, causes which form part of the natural dynamics of the island system: These include competition with new species—new arrivals or locally evolved species—, natural changes of habitat—e.g. due to climate change—, natural disasters and the end of the island—either by total submersion or the connection to the mainland. Evidence of these factors is extremely limited in the fossil record. Secondly, causes that are related to human activity: The impact of humans is manifold (hunting, active and passive introductions of alien taxa, spread of pathogens, habitat destruction etc.) and acts on levels from the individual taxon to entire fauna. The colonisation of islands by Palaeolithic and Neolithic humans is relatively well documented by archaeological findings (e.g. Flores, Ryukyu Islands, Cyprus, Caribbean islands) but much remains to be understood about the degree of its impact. Most is known, of course, about the impact of colonisations during historical times. This, of course, does not provide a proxy for the impact of Palaeolithic nor of Neolithic humans. It is merely a rough indication of the potential dangers that humans can bring to an endemic island fauna.
Anthropogenic factors are relatively well-documented in the fossil record compared to the ‘natural’ factors. Extinction events, however, are often caused by a combination of factors, and rarely by a single factor alone. In addition, natural factors and anthropogenic factors are often difficult to separate, for example, whether a competitive species came on its own to the island or was introduced by humans, the eventual outcome is the same. Finally, the combination of natural factors and anthropogenic factors may give the fatal blow to a species, for example climate change and hunting.
The dramatic and devastating impact of hunting by first settlers is generally referred to as Blitzkrieg extinction: fast, overwhelming and with no change for defence. The theory of Blitzkrieg extinction has been used to explain the extinction of the megafauna of New Zealand, Australia, Madagascar and the Americas practically immediately after the arrival of humans. The basic assumption here is that the endemic fauna is naive by nature and thus easy to hunt down. According to Steve Wroe and colleagues in 2004, the model is an oversimplification. First of all, the fauna was not that naive, and contained large predatory animals. Secondly, the theory entirely overlooks the effects of climate change. Australia’s climate became hotter and drier and large parts lost their original vegetation cover. It might be that the model works for some extinction events, such as that of the moas of New Zealand, but not for all. The theory that the moas went extinct practically overnight after the Polynesians set food on the island in the 13th century is widely accepted, though MacPhee in 2009 puts a question mark here. As the moas became rarer, it took considerably more time to find them, and after a while, with just a few birds left, the hunt would not have been worth the effort anymore, and the Polynesians would have turned to easier game, like seals and seafood. The giant birds would have survived in low numbers in remote or inaccessible places, as did the still extant kiwi, which has an equally low reproduction rate. Therefore, even in the moa case, another factor must have contributed to their extinction, apart from hunting.
A problem in most extinction studies is the absence of reliable data concerning the timing of the extinctions. For example, extinctions in the West Indies started after the arrival of the first Amerindians, some 6000 years ago. However, the Amerindians had no devastating impact on the fauna and many taxa continued to exist after their arrival. Thus, in the case of West Indies, the Blitzkrieg scenario is not supported by facts. On the contrary, the extinction caused by the Amerindians and later by the Europeans presumably took place within a rather broad time frame, sometimes referred to as Sitzkrieg, as opposed to Blitzkrieg. Extinction of West Indian mammalian taxa was likely caused by the culminating effects of centuries of human-induced distress.

Speciation Through Time

[Cynotherium sardous of Sardinia-Corsica (Late Pleistocene) compared to its mainland ancestory (Xenocyon). Drawing G. Lyras]

Evolutionary changes in island mammals follow certain patterns. In the third part of the book, observed anatomical and morphological patterns or trends are discussed, and the drive behind them explained. Well-known trends are a change in body size (size reduction in larger mammals, size increase in small mammals), the development of a stable and sure-footed but rather slow locomotion (short limbs, bone fusions) in artiodactyls and proboscideans, an increase in hypsodonty in herbivores. Much less known trends (because they are observed in fewer taxa) are the development of ever-growing rodent-like incisors in some goats, the loss of dental elements, the simplification of antler morphology, adaptive radiation (larger size and form variation than on the mainland), a more frontal view, and a change in dental eruption sequence.
Terms like sympatric and allopatric speciation, metapopulations, species flock are explained in this context and discussed with examples. Island arches and archipelagos are described, and the speed of evolution of islanders is illustrated with an example of a long-term canid lineage, being that of Cynotherium, spanning the entire Pleistocene of Sardinia-Corsica.

Which Animals?

[The Cretan dwarf deer (Candiacervus sp. 2, Late Pleistocene); collection University of Athens, Greece]

The majority of fossil insular mammals are ruminants (deer, bovids), hippos, proboscideans and rodents. Less frequently found are otters, pigs, lagomorphs, insectivores and sloths. Rare insular mammals are canids, hyenas, felids, rhinos and primates. The rarest are perhaps bats, due to the fragmentary nature of their fossil record.
This is of course only true on a global scale. Per island the picture might differ essentially. For example, primates are not rare at all on Madagascar, where they form half the macro-biodiversity. In the remote past, there were even more of them, including gorilla-sized species, but they got extinct when humans arrived. On Ibiza, there is a time-slice with only bats and birds. For some unknown reason, most likely strong currents, the island could not be reached but by air. Sloths are rather common in the West Indies, but totally lacking elsewhere. The same is true for the caviomorph rodents. On the other hand, endemic hamsters are found only on Miocene Italian islands.

Fossil islanders can be roughly divided into two classes: those that arrrived either by swimming or by rafting (sweepstake or 'by chance' dispersal), and those that became isolated after the region got separated from the mainland (vicariance effect). Deer (e.g. Candiacervus of Crete, see photograph), hippos, proboscideans, murids belong to the first class; bovids, lagomorphs, primates belong to the second class. The book describes and discusses the effects of insularity on these taxa, based on what is known from 89 or more endemic insular taxa, which lived or still live on at least 30 islands all over the world. Endemic faunas are by nature impoverished and unbalanced, and comprise only the large herbivores (deer, bovids, proboscideans, hippopotamuses), a range of micromammals (e.g. rats and mice, insectivores, tenrecs, hutias, lagomorphs), a very limited number of carnivores (dogs, grey fox, lion, but mostly mustelids) and primates, of which the most spectacular is the recently discovered so-called hobbit Homo floresiensis. The discussed islands are found all over the world, and include those of the Mediterranean Sea (e.g. Crete, Sicily-Malta, Cyprus, Dodecanesos, Cycladic Islands), Southeast Asia (e.g. Java, Flores, Philippines, Japan), the Pacific Ocean (Galapagos, Channel Islands), the Atlantic Ocean (Falkland Islands, West Indies), and the Indian Ocean (Madagascar).

Example of an Endemic

[The five-horned deer (Hoplitomeryx matthei, cast of holotype) of the Gargano, South Italy (Late Miocene). Photo Eelco Kruidenier; copyright Naturalis, Leiden, the Netherlands]


Acknowledgments for our fieldwork

[Paul Yves Sondaar excavating at Mavro Muri, Crete, Greece, in the 70s. Photo John de Vos]

First of all we like to thank the late Dr Paul Yves Sondaar (1934-2003), from the University of Utrecht, the Netherlands, who was the motor of island research. He was the one who brought us to all those exotic places and came up with all kinds of hypothesis and theories. It is therefore to him to whom this book is dedicated.
The data and results of 40 years of island research, which form the basis for this book, could not have been possible without the help and assistant of a lot of people as well as in the field, in the lab as in the institutions and museums.
For the help in Greece and its many islands, especially Crete, we like to thank Prof. Dr Nikos Symeonidis of the Geological and Paleontological Institute of the University of Athens for the permissions; for the help in the field Dr Jef Leinders, Dr Peter Weesie (who did a PhD as student of Paul Sondaar on the fossil birds of Crete), Reinier van Zelst, Dr George Theodorou, Dr Johanna de Visser and all the students who were involved.
For the Gargano we like to thank dr. Mathijs Freudenthal, who invited us to participate in his excavations. For help in the field we thank Dr Jef Leinders who did a PhD as student of Paul Sondaar on the artiodactyl Hoplitomeryx.
For Majorca we like to thank Dr Bill (William) Waldren, who showed us the material and gave us the opportunity to excavate in the Son Muleta cave. Secondly, we thank Joseph Alcover for giving Paul Sondaar the opportunity to stay a year in Palma de Mallorca to further investigate the fossil Majorcan fauna.
For Sardinia we like to thank Mario Sangez of the Archaeological Survey of Nuoro for his decade-long help, support and interest. Secondly, our thanks are due to Dr Gerard Klein Hofmeijer, PhD student of Paul Sondaar and organizer and effective leader of the excavations at Corbeddu Cave. Thirdly, we thank Fulvia Lo Schavo and Maria Fadda, of the Superintendency of Sassari. For help in the field and the lab we thank Marisa Arca, Catarinella Tuveri, Pascuolino Catte and many others who assisted us in various ways during our extensive excavations at Corbeddu Cave and later at Monte Tuttavista.
For Indonesia we like to thank first of all Dr Fachroel Aziz of the Geological Research and development Centre (GRDC) and the Museum of Bandung for organizing the fieldwork and all his help. Secondly, we like to thank Dr Gert Van den Bergh of Wollogong University, Australia, who did his field research on the islands of Indonesia as a PhD student of Paul Sondaar, and who came up with a lot of new data, concerning the geology and palaeontology of that area. The general supervision of the Indonesian projects was the responsibility of the Mines and Energy ministry of the Republic of Indonesia. Our thanks go out to Dr Rab Sukamto, former director of the GRDC, and all the other officials of the GRDC. Furthermore, without the approval of regional and local governmental authorities in Java, Sumatra, Sulawesi, Flores and Timor, the projects would have been impossible. Further a lot of people helped in the field during excavations and made the study of the material in Indonesia possible. These are, in random order, Ungkap L.M. Batu, Darwis Fallah, Ani Krisnawati,Iwan Kurniawan, Ijep Seafudin, Slamat Sudjarwadi, Suyatman, Reinier van Zelst, Torkis Sihombingand, Hans Brinkerink, M. Anwar Akib, Dr Paul Storm, Dr Johanna de Visser and Tony Djubiantono.
For our work on the Philippines we like to thank Dr Angel Bautista of the National Museum of the Philippines, Manila, for the organization of the fieldtrips. The excavations and study of the material was made possible by Dr Fr. Gabriel S. Casal, director of the Museum. For the permissions are thanked Dr Eusebio Z. Dizon (director), and Wilfredo P. Ronquillo, head of the department of Archaeology. For taking part in the excavations we like to thank: Dr Angel Bautista, Ernesto B. Toribio (called June), Willem Renema, Mrs. Rosemarie A. Fernandez, Gerardo Navarro, Diosdado A. Manaloto, Arie and Edith Janssen, Augustin dela Cruz, Dr George Lyras, Alejandro Regina and Dr Wilhelm Solheim II.
For our work in Japan we like to thank Prof Dr Hisao Baba, curator and head of the department of Anthropology of the National Science Museum of Tokyo, for the invitation to visit the Museum. We further thank Dr Huaro Otsuka for inviting us to the University of Kagoshima to study the material from the Ryu Kyu islands, Prof Dr Hasagawa and Dr Nagasagi for showing us the material in the Gunma Museum, Prof Dr Hidetoshi Kamiya, Dr Takahashi and Dr Haruo Saegusa for inviting us to the University of Kyoto and showing us the proboscidean material and organizing an excursion to lake Biwa.

This site has been designed and developed by Gregory Lyras