Thursday, 20 December 2012

A Word on Methodology: The Sporormiella Proxy

In my previous post, I mentioned the Sporormiella proxy used to determine abundance of Pleistocene megafauna in Madagascar. This is an analytical technique that has recently gained prominence in the study of the late Pleistocene megafauna extinction. There is an interesting paper by Feranec et al (2011) about the Sporormiella proxy and the problems associated with using it.

The Sporormiella Proxy
Sporormiella is a fungus that is present on the dung of herbivores. Sporormiella sporulating on dung release spores which adhere to nearby objects (usually plant matter). Herbivores then eat this plant matter and the spores, which pass through their digestive tracts, are released in their dung. The spores of this fungus are preserved readily in lake sediments, and stratigraphic changes in the abundance of this fungus in Pleistocene and Holocene sediment sequences have been used as a proxy to define megafaunal presence, decline and extinction globally.

Sporormiella Spores 
The presence of Sporormiella is not exclusive to large herbivore dung and has been found in the dung of small herbivores as well, such as hares. Thus, it is difficult to use Sporormiella as a sole and direct proxy for megafauna abundance unless specific species of Sporormiella associated only with large herbivores can be identified.

A stratigraphic decline in Sporormiella does not necessarily indicate a decline in megafauna. For example, Sporormiella is more abundant near lake shores than in the middle of lakes, so a decrease could simply mean a rise in the lake level. Sporormiella may also be preserved to varying degrees depending on type of lake sediment, lake levels, etc. A related point is that the absence of Sporormiella does not indicate the absence of herbivores – some modern day sites with abundant livestock have been shown not to contain Sporormiella in Davis and Shafer’s (2006) study. Thus, Sporormiella needs to be calibrated to other indicators of large herbivore population and is non-conclusive on its own.

Some academic papers must be viewed with some scepticism due to methodological over-reliance on this particular proxy. For example, in a Gill et al (2009) paper, a decline in Sporormiella in a Lake Appleman core in Indiania which starts from 14,800 years ago and which pre-dates a major change in the pollen assemblage is used to conclude that the late Pleistocene megafauna extinction was not caused by (usually climate-linked) vegetation changes. They also show that charcoal frequency increased at that site, indicating that human factors (like vegetation burning) were probably behind the extinctions. However, the tail end of the Sporormiella decline is also associated with a change in lake sediment size, which may reflect changes in the sediment input and hence catchment area of the Sporormiella source, rather than megafauna decline.

While this analytical technique is certainly promising in contributing to research on Pleistocene megafauna extinction, it still needs to be refined. What is also important is to avoid complete reliance on just one proxy; the conclusions drawn from using this proxy should be calibrated to other indicators of megafauna abundance.


Davis, O. K. and Shafer, D. S. (2006) ‘Sporormiella fungal spores, a palynological means of detecting herbivore density’, Palaeogeography, Palaeoclimatology, Palaeoecology237, 1, pp. 40-50.

Feranec, R. S. et al (2011) ‘The Sporormiella proxy and end-Pleistocene megafaunal extinction: A perspective’, Quarternary International, 245, 2, pp. 333-338

Gill, J. L. et al (2009) ‘Pleistocene megafaunal collapse, novel plant communities, and enhanced fire regimes in North America’, Science, 326, pp. 1100-1103

Sunday, 16 December 2012

Island Extinctions and the Case of Madagascar

Island Extinctions
Islands which were only inhabited by humans after the late Pleistocene megafauna extinction event offer interesting ‘control experiments’. A compelling argument for human factors driving extinction is the time lag between continent extinctions and those of nearby islands (Martin and Steadman 1999). For example, New Zealand’s moas lasted 30,000 years longer than Australia’s extinct giant bird, the mihirung, which went extinct during a Australia’s wave of rapid megafauna extinction (before the late Pleistocene).

The argument for prey naiveté on early contact with human hunters finds support also in the remarkable tameness of wild birds in remote islands which were undiscovered by prehistoric explorers. These include the Galapagos, Christmas Islands, etc. Galapagos’ avifauna were unafraid of humans, as depicted in historic accounts from 17th century sailors who discovered the island (Martin and Steadman 1999).

In this blog post I look closer at the island of Madagascar.


This paper by Burney et al (2004) discusses how islands can be used to better understand megafauna extinction. Madagascar is interesting because it is the last place on Earth where megafauna went extinct prehistorically – extinctions in which humans had some part to play in most other parts of the world occurred much earlier, during the late Pleistocene or even earlier in Australia. Madagascar offers a relatively fresh record of paleoecological change, since humans only arrived in the late Holocene, about 2,000 radiocarbon years ago. Very little is known about how or why this group of Iron Age people came to Madagascar. Evidence of the first humans can be shown by human-modified megafauna bones, such as cuts on the fossilized bones which show removal of flesh from bone by a sharp object.

Very little is known about the late Pleistocene biota in Madagascar. The amount of data increases greatly for megafauna in the mid-Holocene, where conditions for fossil formation probably became more favourable. For example, most lakes and swamps along the coastline formed only after around 5,000 radiocarbon years ago. Nevertheless, there were major climatic changes in in the late Pleistocene and pre-human Holocene, most of which were survived by most of the megafauna. Although there have been range shrinkages, there were no extinctions. Of the 9 genera of extinct lemurs dated, only one is not securely dated to the human period. Some examples of major climate change are as follows: 
  •      20,000 radiocarbon years ago (LGM): widespread dessication occurred. Lake Alaotra, a large lake in humid eastern Madagascar, was dramatically reduced in area if not completely dry during that period. 
  •      10,000 calendar years BP: At another site called Trtrivakely, pollen evidence shows the nearly complete replacement of heath vegetation with wooded grassland.
A drastic decline in megafauna, as shown by a huge decrease in Sporormiella in sediments at 1700 radiocarbon years BP (within a few centuries of first human contact), was observed. Sporomiella is a fungus that grows in the dung of large plant-eating mammals, and it releases spores which are preserved in sediments. The presence of these spores is used as a proxy for the presence of megafauna. Humans could have hunted these megafauna or altered their habitat. Before humans arrived, these herbivores had very few predators other than large crocodiles. Although there is very scant evidence for direct human hunting of megafauna, as in many other continents where Pleistocene megafauna extinction has occurred, another way in which humans could have contributed to the decline is through altering the existing fire regime by further increasing fire incidence through burning for settlement and agriculture and through hunting of plant-eating megafauna. The decline of large herbivores such as giant hippos caused ground litter to accumulate, feeding more fires. This can be shown by charcoal peaks above background values, first occurring in the South West where humans first settled, and then spreading outward over Madagascar. Nevertheless, the extinction pattern on Madagascar does not support a Blitzkrieg hypothesis. There is an overlap of around 2,000 years from earliest human evidence to the last occurrence of extinct megafauna.

The chart (Burney et al 2004) shows a summary of events in Madagascar:

My Thoughts
The evidence from Madagascar is indeed intriguing and I feel it does make the argument for human factors in the extinction of megafauna more compelling. Madagascar’s physical geography and vegetation is very similar to Africa such that it is referred to as an ‘Africa in miniature’, and it is probably safe to assume it went through similar climate changes and vegetation responses as Africa. The megafauna on this island certainly survived all these before the humans came, after which they experienced dramatic decline and finally, extinction. The fact that it is an island is important; in a previous post I mentioned the reason for why Africa still has such a large diversity of megafauna left is that it is larger and probably provided more refugia for megafauna. Madagascar probably provided more limited refugia for the stressed populations of megafauna.     

Burney, D. A. et al (2004) ‘A chronology for late prehistoric Madagascar’, Journal of Human Evolution, 47, pp. 25-63.

Martin, P. S. and Steadman, D. W. (1999) ‘Prehistoric extinctions on islands and continents’ in MacPhee, R. D. E. (ed.) Extinctions in Near Time: Causes, Contexts and Consequences, New York: Kluwer Academic/Plenum, pp. 17-50

Friday, 7 December 2012

Dissecting the Hyperdisease Hypothesis

I decided to do a post on the disease hypothesis after Josh from suggested an interesting paper by Rothschild and Laub (2006). Here is the link to his post on this topic specifically. The hyperdisease hypothesis proposes that humans and their domesticates introduced novel hyperdisease to vulnerable populations of Pleistocene megafauna who had never encountered such diseases before and whose bodies were therefore unable to cope. Since migrations of animals from Europe to North America were not uncommon before the period we are studying, it is more likely that humans and their domesticates were the disease vectors (Lyons et al (2004).

Tuberculosis and the American Mastodon 
Rothschild and Laub (2006) have suggested that new evidence for the hyperdisease theory has surfaced in the form of bone alterations from infectious tuberculosis found in just over half of 113 mastodon skeletons in the Western Hemisphere. Since not all animals infected with tuberculosis develop this bone alteration, it must follow that probably almost all of the mastodon population must have been infected with tuberculosis. The disease thus qualifies as a pandemic in the sense that it had an extremely high infection rate. Besides, it has a persistent presence in the fossil record from around 34,000 – 10,000 years BP, establishing that it must have been present in the late Pleistocene period. 

However, there is a difference between infection and mortality – the disease was not necessarily fatal. Rothschild and Laub (2006) hypothesize that this disease may have weakened mastodons in the face of climate change and human impacts in the late Pleistocene, further stressing their populations. While the disease could have remained latent, the environmental stresses of that period could have resulted in a loss of latency, increasing mortality. However, it is unlikely that the hyperdisease could have been a major factor in the extinction event. 

The Modern Day West Nile Virus: A Proxy for the Mystery Hyperdisease?

I also found another paper by Lyons et al (2004) which proposes some criteria for the hyperdisease theory to be plausible. 

  1. It must be able to survive in a carrier state in a ‘reservoir’ species when there are no susceptible hosts to infect.
  2. It must have a very high infection rate.
  3. It must be extremely deadly with a 50-75% mortality rate
  4. It must be able to infect multiple host species without infecting humans

Lyons et al (2004) use the West Nile Virus in birds, a disease which has seen recent introduction and spread in North America’s bird population, as a proxy to test this hypothesis as it appears to fulfil all of the above criteria of a hyperdisease.

One of the unique features of the late Pleistocene megafauna extinction event was its size-selectivity – smaller and medium-sized animals were largely unaffected. Thus Lyons et al (2004) have tried to test if West Nile virus causes such size-selective infections in birds. It can be shown that it does not, as infection rate increases positively with body size (Fig. 1) and infection occurs across a range of body sizes. This contrasts with the pattern shown by late Pleistocene mammal extinctions. The x-axis of the graph shows the size category of the bird species infected by the West Nile virus and those of the mammals which went extinct during the late Pleistocene. It has been re-scaled for mammals since they contain a much larger range of body masses. Each filled square shows the percentage of species pool in each size category infected by the virus or that went extinct. 

Fig 1 (Lyons et al 2004)
Some have argued that large body size makes species inherently vulnerable to extinction because of life history factors, e.g. low reproduction rates which make it harder for populations to recover from mortality caused by disease. However, Lyons et al (2004) counter-argue that if this is true, then larger species should have high extinction rates relative to smaller species over evolutionary time, which is not the case. 

The Verdict?

I find the hyperdisease hypothesis unconvincing so far and I think it is only considered seriously as a factor in the extinction event because of the general lack of evidence surrounding even the exhaustively-researched hypotheses of climate change and human hunting (e.g. lack of kill sites). However, the even more severe lack of evidence in the hyperdisease hypothesis is even more disturbing. The only known modern disease which fulfils the 4 criteria of a hyperdisease capable of wiping out megafauna during Pleistocene times, the West Nile virus, itself cannot be shown to cause the size-selective extinction pattern in modern day bird populations in North America. Besides, the paper by Rothschild and Laub (2006) only shows a pandemic-scale disease in one type of animal, the American mastodon. It is difficult to find equivalent disease explanations for all other megafauna species killed during the late Pleistocene (mammoth, for example, were not affected by this disease and were close cousins of the mastodon). Therefore, I conclude that hyperdisease is an unlikely explanation for the megafauna extinction event we are studying here.


Lyons, S. K. et al (2004) ‘Was a hyperdisease responsible for the late Pleistocene megafaunal extinction?’, Ecology Letters, 7, pp. 859-868

Rothschild, B. M. and Laub, R. (2006) ‘Hyperdisease in the late Pleistocene: validation of an early 20th century hypothesis’, Naturwissenschaften, 93, 11, pp. 557-564

Wednesday, 5 December 2012

The African Anomaly

Africa is the ‘anomaly’ in the Pleistocene megafauna extinction debate because it only lost 21% of its megafauna species, in contrast to the high levels of extinction in other continents. Today, it also contains a much higher level of megafauna diversity than other continents, including some species which have survived from Pleistocene times, such as the Cape Buffalo. 

Climate Change, Refugia and Disease

Climate change in Africa has been associated with dramatic species extinctions in the whole Pleistocene, not only the late Pleistocene. 59% of the Pleistocene megafauna extinctions occurred in the early Pleistocene, 21% in the middle Pleistocene and 20% in the late Pleistocene. All of the late Pleistocene extinctions happened during the late Pleistocene/Holocene transition (Graham and Lundelius 1984). 

The present-day megafauna has been called a ‘living Pleistocene fauna’ (Graham and Lundelius 1984: 240) because of their diversity is almost similar to the diversity of the extinct Pleistocene megafauna. Graham and Lundelius (1984) argue that perhaps the rate and magnitude of climate change was slower in Africa than in other continents, thus explaining the lower rates of megafauna extinction. Savannah environments survived into the Holocene in Africa while they disappeared in other parts of the world. For example in South America, savannah environments were abundant during the late Pleistocene but is today restricted to only a few areas. Other explanations for the African Anomaly have been put forward, although many of these tend to be speculative as Africa is the least studied continent with regards to late Pleistocene megafauna extinction. One possibility is that Africa has a great variety of habitat types which may offer better refugia for megafauna pressured by human activities (Heller 2012). Another explanation could be the existence of diseases that prevented humans or livestock from living in certain areas. This is still true today, as exemplified by locally endemic livestock diseases making large tracts of attractive pasture in Africa unavailable for human settlement. This is a phenomenon unique to Africa (Heller 2012). 

Human Impacts?

Graham and Lundelius (1984) claim that it is unlikely that humans have had much ecological impact on Africa’s megafauna because they have been known to coexist with them for a much longer time than on other continents. Martin (1984) even attributes the lower extinction rates to lower prey naiveté as a result of adapting to the hunting styles of humans. However, the human impact cannot be underestimated. Klein (1984) points to archaeological evidence that the humans of the late Pleistocene/Holocene transition were much more competent hunters than earlier humans. For example, studies of archaeological sites of earlier humans have found that eland (a type of ungulate) remains occur more frequently. Within the archaeological sites of humans who lived in the late Pleistocene/Holocene period (but under similar environmental conditions), remains of wild pigs, which were more dangerous to hunt and therefore required more sophisticated hunting techniques such as traps for example, were more prevalent than those of eland. 

African Wild Pig

The Cape Buffalo Study

A recent paper by Heller et al (2012) using genetic sequencing of African Cape buffalo, a species which has survived from the late Pleistocene period to the present-day, has found evidence of benign human – megafauna interaction during the late Pleistocene. Cape buffalo began a population expansion from 80,000 radiocarbon years ago and reached a peak at 8,000 radiocarbon years ago, which shows that humans and climate change had relatively little impact on the population. This study provides further evidence of benign human-megafauna co-existence during the late Pleistocene. To the extent that Cape buffalo is representative of the ecological dynamics facing other African megafauna, this new research also supports the Graham and Lundelius’ (1984) finding that most of African Pleistocene extinctions occurred in the early Pleistocene. If Klein is correct, this was at a time when human hunters were very poorly technologically developed. Thus, climate change would be the larger factor impacting megafauna populations in Africa. 

Cape Buffalo

Klein, R. G. (1984) ‘Mammalian extinctions and Stone Age people in Africa’ in Martin, P. S. and Klein, R.G. (eds.) Quaternary Extinctions, Arizona: Arizona University Press, pp. 553 – 573

Graham, R. W. and Lundelius, E. L. (1984) ‘Coevolutionary disequilibrium and Pleistocene extinctions’ in Martin, P. S. and Klein, R.G. (eds.) Quaternary Extinctions, Arizona: Arizona University Press pp. 223 – 249

Heller, R. et al (2012) ‘Cape buffalo mitogenomics reveals a Holocene shift in the African human–megafauna dynamics’, Molecular Ecology, 21, pp. 3947–3959