A great start to the conference today. What a dream to have all these great speakers and thinkers in the same room for three days. Lots of good discussion and some bold thinking,
Adrian Lister: the Eurasian extinctions have a particularly good record. I don’t feel we have adequate data, even in Europe, to solve the megafauna/climate debate.
Felisa Smith: we are mixing apples and oranges. There is a background extinction rate. The Pleistocene extinctions in northern Eurasia are not that different from background rates, but elsewhere they are way above background rates Kate Lyons: the extinctions were highly size biased. These extinctions would not have happened without the addition of humans to the system. Todd Surovell: spatial distribution of elephant kill sites and decline globally can only be explained by humans Blaire Van Valkenburgh: what is lacking in our models is thinking about the importance of the biotic interactions that are going on. David Nogues-Bravo: Extinction events were rarely caused by a single driver. Debates about climate vs humans may not be useful. Jens-Chrstian Svenning: we should exploit the repeated climate changes of the whole Quaternary. From that perspective it is even more striking that we have such a size bias in the recent extinction but little plant extinctions. We don’t expect large bodied animal species to be the most sensitive to climate change because they have ranges and are relatively mobile. Felisa Smith: large body size does not make species prone to extinction in the fossil record. We only see this in the late Pleistocene extinctions. And we see it in recent marine extinctions. We have to be careful to infer the past from highly modified ecosystems of today Frans Vera: many megafauna became extinct or effectively extinct in Eurasia in the last few centuries. We can look at modern animals to know what happened in the past. Tony Stuart: There is a tendency for North Americans to play down northern Eurasia, but the extinctions there were significant. They were not just background extinction rates. The super-mega mammals go out as they did elsewhere in the world. Felisa Smith: we need to be careful about cherry-picking taxa. Todd Surovell: I don’t see Eurasia as anomalous. Hominids and megafauna occupied in different parts of the continent. Before the LGM little evidence of humans in high Arctic, but abundant megafauna. Like to see more explicit hypothesis testing as to what is affected. Adrian Lister: From our data and the pattern of staggered extinction, we can predict based on distinct ecology of species (e.g. Mammoth was grassland species, giant deer was forest edge species). Don’t expect range of species of different ecologies to all go extinct at the same climate signal. In N America, if we find all species disappear at the same time then we could be convinced of human drivers of extinction, but we don’t have the data. John Terborgh: Several interesting cases of refugia where megafaunal species persisted into Holocene (elephants in Mediterranean, giant sloths in Caribbean etc). They survive until humans show up on the islands. Abby Swann: there has been a lack of discussion about the uncertainty in climate modelling of past climate conditions. David Nogues-Bravo: one way to get at climate uncertainty is through using multiple climate models. Tony Stuart: very few people doubt that extinctions on islands re caused by humans. Adrian Lister: no-one has really shown a clear mechanism as to why African species persisted within humans Todd Surovell: in archaeology Pleistocene overkill is very controversial. He is happy to be in good company here. Kate Lyons/Blaire Van Valkenburgh: in Africa we have been looking at the wrong timescale. There is evidence of major extinctions around 2-2.5 million years ago around emerging of Homo erectus. Much of Africa’s megafaunal extinctions happened then. Early carnivores would have been quite disturbed by early hominids. Felisa Smith: Shame Paul Martin is no longer with us. He would have loved this meeting. Melissa Pardi: in order to test the climate hypothesis we are looking at the long time. We need to look into the deeper past. Josh Donlan: how important is attribution of Pleistocene extinction to moving forward? Do we need evidence that humans played a major role in order to move forward with rewilding etc., or is it irrelevant. Kate Lyons: it depends on what is going to motivate people to change policy. Some think that human-caused extinctions mean we’re powerful agents of change and can fix it; others feel hopeless. Adrian Lister: justifying for rewilding has to be made on basis of the ecological situation now, not on Pleistocene arguments Frans Vera: disagree, we can learn from the past overkill Adrian Lister: We can see the importance of synergy of forces, how climate and human drivers can interact. Maan Barua: naive to consider modern human drivers as equivalent to Pleistocene human drivers Adam Wolf; Ancient food webs are important. How much do we understand them. We have two main questions: why did the losses occur, and what were the ecological consequences
His answer is that the answers are deeply intertwined. He looks at a marine megafaunal species: Steller’s sea cow. A dugongoid serenian radiated into N Pacific with onset of last ice age. It winked out at end of Pleistocene, with the exception of the Commander Island where they persisted until 1768 (off east coast of Russia). This was the only island group not occupied by aboriginal peoples. Russians trappers discovered the Commander Islands in the early 18th century. Why did they go extinct? Direct overkill? Or sea otter-kelp cascade? When sea otters are abundant, sea urchins are rare, and kelp is abundant as it is nor browsed by urchins. There are two states: a kelp state and an urchin state, with rapid transition between the states. The threshold otter density seems top be about 6 otters/km2. Sea cows are browsers, urchins are much more destructive as browsers Turvey and Risley (Biol. Lett. Royal Society 2006) concluded human kill rates are sufficient to explain their extinction. Sea otter-kelp hypothesis not needed. Jim Estes disagrees with this conclusion. The last sea otter was recorded on the island over 1743-1754, well before the extinction of the sea cow. Did sea otter loss drive the sea cow to extinction? We have no records of how 18th century kelp changed, but the collapse has mirrors with what happened with sea otter collapse in the late 20th century. Overall there is a 12-fold decline in kelp caused by a loss of sea otters (Estes et al. 1998). This decline in kelp would cause a 90% decline in food supply for sea cows. How might this have influenced the sea cow population? Create demographical model of trajectory of sea cow population assuming that not a single sea cow was hunted. The annual sea cow starvation rate because of a loss of sea grasses would be 0.74-0.91. If the sea cow population was around 1500 at the start, how long would it take to drive them to extinction. He argues the changes in the ecosystem alone were sufficient to cause the extinction even with zero hunting. Conclusion: Food webs are deeply interconnected things. When we talk about extinction we have to thing about ecological interactions. Many megafaunal extinctions may be associated with loss of keystone predators - more complex and subtle than a simple model of direct hunting. Estimated loss in species, phylogenetic and functional diversity as a consequence of Late Pleistocene and Holocene range collapse and extinctions
He tries to work out what the modern world animal distributions would have looked like if there had been no Pleistocene extinctions or human influenced range reductions. Some very striking maps show that megafaunal diversity in Americas and Eurasia would have been as high as Africa. Africa would have been nothing special. Highest megafaunal diversity in the world would have been in southern South America (around southern Brazil and Uruguay)! How different the modern biosphere looks from this world without the megafaunal extinctions. What are the drivers of this loss? Similar story as previous talks. First human contact dominates in Americas and Australia, .and insular E Asia Japan, Madagascar. Climate drives northern N America and Eurasia. Elevation range change affects biodiversity loss in high mountains (Andes, Tibet etc.) Ripple et al (2014) showed that in many places outside Africa there are only three for four large carnivores today. In contrast, in N America in the late Pleistocene there were at 10-15 large carnivores (lions, cheetah-like cats, wolves, sabre-tooth cats etc). Such numbers of hyper carnivores have been typical for millions of years
E Africa 2.5 myr ago had 17 hyper carnivores (but only 13 now), Yellowstone has around 9. There was an earlier extinction event in Africa, around 2 million years ago. But North America in the Pleistocene was richer in hypercarnivores than East Africa today. Why was Pleistocene N America so diverse in carnivores? There had been immigration of lion, hyena and brown bear via Siberia. Bison immigration also may have supported high carnivore diversity. It is very difficult to get animal abundance data from the fossil record. She looked at tooth wear and fracture. Percent of broken teeth is a good metric of amount of prey in the diet. More broken teeth indicate rare prey. Nice ground testing on modern wolf studies in N America and Sweden. She looked at Pleistocene predators. Much higher proportional breakage of teeth than most model carnivores. This is even true for carnivores that still persist (e.g. coyote and wolf). This suggests that food was limited for large predators and interspecific competition was more intense than on average for modern ecosystems. Prey abundance was tightly constrained by this abundance of hyper carnivores. But despite this, these populations were quite stable for millions of years. In many cases, prey densities were limited by the predators. This may have made the ecosystem vulnerable by an unusually flexible predator - humans. Humans are unusual members of the large predator community that are less affected by competition and predation than other species. They have a flexible diet (can switch between foods), they were accompanied by dogs (additional hunting benefit), complex social behaviour, weapons, fire and language. A super predator outside the guild yet having a huge impact. (Ripple and Van Valkenburgh 2012, BioScience). Human hunting may have triggered a cascade that led to extinctions, forcing other carnivores to eat down the food chain. Humans did not need to kill all the herbivores themselves to trigger Pleistocene overkill. Analogy with sea in 18th and 19th centuries observed trophic cascade after removal of whales. Great wolves - harbour seals - otters - kelp forests. What can we learn from the Pleistocene? It was a resilient ecosystem because there was carnivore redundancy, carnivores could move. If we add a new predator to an ecosystem there will be consequences for predators and also other prey. Felisa Smith: recalibrating the Anthropocene: humans, megafauna and global biogeochemical cycles3/18/2014 #oxmegafauna
Megafauna and methane emissions through their burps It doesn’t take much to drive a big animal to extinction. All you do is need to add a little extra mortality and the species is heading to extinction (Zuo et al (2013) The American Naturalist) Who went extinct? In North America, 78 species went extinct, In South America 71 species. Most of these were large-bodied herbivores There were a 100 million animals that were lost from North America and similar or more from South America. That’s a lot of microbial vats! Questions: 1. Did the arrival of humans into the Americas 13.4 kyr and the subsequent extinction coincide with a decrease in atmospheric methane 2. Could the megafaunal extinction have caused the is 3. Is there a detectable climate effect Answer 1. There was a large drop in methane around the time of Clovis arrival in the Americas. This drop is much more rapid and distinct compared to the previous 700,000 years The isotopic signature of methane was also different from previous blips Answer 2. Estimate emissions from megfauna related to body size. Animal population density well predicted by body size. Geographic range was harder to estimate. Maximum and median range can be reasonably well-predicted from body size. To estimate methane emissions per animal use IPCC approaches Significant influence of methane emissions on body size but also on strategy: hind gut (horses and elephants) or ruminants (cattle etc. higher emissions). Calculate that megafauna loss of methane source 9.6 Tg/year. This can explain anything from the entire drop to 13% of the drop (the assumption of residence time is critical) Could human extinction of megafauna have contributed to the onset of the Younger Dryas cold event. The Anthropocene could be said to have started at 13.4 kyr BP Rinderpest epidemics in Africa in 1880s. Aroind 90% of cattle and grazing wild animals died. This would have caused a decline of 6.5 Tg year-1 in methane emissions Conclusions: Wild animals influence climate. Herbivores are “walking microbial vats” and we lost 100s of millions of them in the American megafaunal extinctions. Global integrative assessment of extinction. Split globe into zines: map extinction. Nice global pie chart of degree of megafaunal abundance and extinction worldwide.
Model suggests that human arrival is the biggest factor. Climate is 1/4 to 1/3 of megafaunal extinction. Geographically specific examples: Europe: early human arrival caused some extinction early , climate change a major influence at end of ice age Madagascar, New Zealand, Tasmania humans explain most of extinction North and South America: similar patterns to small islands, with humans biggest factor and climate a minor factor. Indo-Malaya and Central Asia: the model fails. It predicts extinctions that do not occur. Not clear why Asia is different. The models strongly implicate human colonisation. Central and Southern Asia is the next frontier - we cannot explain relatively low levels of extinction there. Identified 177 large mammal species that went extinct or continentally extinct over the period 132 kyr to 1000 yrs BP
Hypothesis 1: Climate change severity is positively correlated with magnitude of megafaunal extinction Made global maps of climate change anomaly and velocity based on difference between large glacial maximum and Holocene Hypothesis 2: Megafaunal extinctions greatest where co-evolution with hominins is least Takes into account pre-Homo sapiens. Longest record is Africa, intermediate in Eurasia, less and less as you go further north. Other places were only colonised by a single hominin, Home sapiens 13 megafauna went extinct in Britain, Kenya 5, Uruguay 31, Texas 33 (the highest number) Proportion of extinct large mammals (Uruguay around 77%, Britain 63%) His model comes to a clear conclusion that human arrival into human-naive regions is the main driver of megafaunal extinction in most regions, with the exception of northern Eurasia. Striking comparison between South America and Africa. Similar climate variability but much more extinction in South America North America suffered a major megafaunal mass extinction around the time of human arrival and climate change. Human arrival: the Clovis complex (13,33-12,600 kyr) is clear in the archaeological extinction. More controversial and sparse evidence for pre-Colvis settlement
Looking globally, there is “deadly syncopation” (Paul Martin) between human first contact and megafaunal extinction. Boulanger and Lyman (2014, Quaternary Science Reviews) argue that megafaunal decline preceded human arrival and were not caused by humans . Others say the same evidence would drive him to completely the opposite conclusion.Gill et al (2009) see similar evidence but leave the door open to causes. If an overkill happened, what would the records look like. He develops a simple model of human colonisation and animal populations. In one version of the simulation, slow human population growth (0.5% per year), hunting rate 0.75 animals per person per year. This causes extinction in around 1000 years. You would not find much overlap in radiocarbon dates of humans and megafauna. Extinctions can happen quickly but human population growth occurs slowly. Hence it should be much easier to date extinction events than colonisation events . Hypothesis: it is easier to detect the ecological impact of a small population of humans than it is to detect the humans themselves in the archaeological record. Prediction1. Declines in megafauna should precede human archaeological sites when site sampling is sparse. Prediction 2: Decline should happen first in Alaska, then in rest of North America, then in South America He does a data synthesis to see when these declines first happened (not extinction date). In Alaska (Guthrie dataset), initial decline occurs around 15 kyr BP. Oldest reliable archaeological site is 14-14.3 kyr, as expected. In rest of North America, decline begins at 14.2-14.5 kyr. Best pre-Clovis at 14.6 kyr, best Clovis at 13-13.6 In South America, initial decline around 13.6 kyr BP. S America archaeological sites around 13 kr BP. One anomaly is Monte Verde site in Chile, at 15 kyr BP, oldest site in Americas. Hence the initial declines in megafauna we see probably indicates the first colonisation by humans, prior to the archaeological records. Lewis Bartlett: what would it take to falsify overkill? Adrian Lister: human population increase of 0.5% /year is a huge compound increase over 1000 years. Is there any evidence for this. Michael Borregard: human debris (stones) should be more detectable than animal bones. His team uses multiple data types (fossil record, bioclimatic simulations) to generate a genetic hypothesis that can be tested with DNA data. Tested on American Bison (Metcalf et al 2014) - best hypothesis for current genetic make-up of bison was a combination of climate combined with historical human impacts. The next step is to synthesise across biomes and taxa. This is work in progress. He feels that both climate change and humans played a role in Eurasia. Declining species paradigm with factors contributing to the general decline of species before their populations become rare The small population paradigm - genetic drift and stochastic extinction of small populations Humans and climate change sit on a spectrum of influences. For some species humans may be able to collapse continent-wide ranges. For others climate change may have caused the broad decline and human finished off refugial populations. Difficult to to this in areas with poor fossilisation (e.g. much of tropics). |