Blog Posts

Elizabeth Jeffers: Differential effects of megafaunal loss on nutrient availability

3/19/2014

Eurasian species respond to climatic changes were very individual. There was a global decline in terrestrial N availability over Holocene due to N being locked up in tree biomass (McLauchlan et al 2013)

There were bottom up drivers of change during the LGM -Holocene period through changing climate and atmosphere, which then affected vegetation and then megafauna. There were top-down controls via megafaunal hunting. Can we separate these?

Looks are British and Irish ecosystem dynamics from two lakes, during the period of woody encroachment at the glacial-Holocene transition

Lough Nadourcan: heathlands transition to birch woodland

Megaherbivore biomass increased (dung fungi), increase in biomass, decline in N availability. Birch consumed more by large herbivores and susceptible to burning, but does end up locking up N. For birch biomass, climate was most important, for the heathlands herbivore biomass was most important.

Quidenham Mere: grasslands to deciduous woodlands

Herbivore biomass decreased, decrease in fire. N availability also increased. The abiotic factors were most important but herbivore biomass contribution to that interaction.

Any evidence of direct impact of large herbvores on N availability? Negative relationship in herbivores, were site was dominated by cervods In English site there was a positive relationship, and there was a much more diverse assemblage of herbivores.

Conclusions

Guild-specific response of herbivores to woody plant encroachment

Climate and fire were direct drivers of tree expansion, herbivores had indirect effects

Impacts of herbivores on N availability depends on herbivore assemblage, soil properties (a finding similar to Elisabeth Bakker earlier). No simple generic relationship between herbivory and nutrient supply.

Mathias Pires: The reconfiguration of seed dispersal networks after megafaunal extinction

3/19/2014

In South America, 80% of South American large mammal species went extinct, including three orders. This could have affected not only megafaunal-dependent plants, but also others plant species through interactions. How were seed dispersal networks disrupted by megafaunal extinction?

Looks at the Pantanal seed dispersal network. There are modules I the network - groups of tightly connected species. Looked at Pleistocene megafauna that probably included fruit in their diets. They reconstructed networks for

1. Pleistocene

2. Early Holocene humans (interactions with fruits currently used by indigenous peoples)

3. Colonial period (livestock and humans)

4. Moderns (extant species and livestock, but less human interaction)

Are their modules in the network, and do the modules change over time?

In the Pleistocene there are 5 modules (large birds, small birds, megafauna, mesofauna and fish dispersal)

In early Holocene, the mammal-dominated modules merged

In the colonial period, the mammal-dominated modules separate again.

In the modern, there are 5 modules again.

The large mammal module includes species that were functionally small in the Pleistocene (e.g. Tapirs).

The livestock helped in partially recover of seed dispersal networks in the colonial period.

Plants dispersed by small animals may have been favoured by megafaunal decline ,e.g. This may have favoured expansion of grasses and decline in fruiting trees.

Ahimsa: can seed dispersal networks assist in guiding rewilding experiments?

Carlos Peres: there is little anthropological evidence that humans are dispersing seeds

John Terborgh: he has seen plenty of Amazonians carrying fruit and throwing away the seeds after eating?

Mauro Galetti: Seed dispersal by Pleistocene megafauna in South America

3/19/2014

Janzen and Martin (1982) first pointed out that there seem to be a lot of megafauna-dispered fruit in South America - “the fruit the gomphotheres ate”. Howe (1985) made a critique - how do these plants survive 10,000 years without dispersers? Lack consistent criteria. Guimaraes et al (2008) came up with a more consistent evaluation, using an operational definition of “elephant fruits” from Africa to Amazonian fruits. 50% of savanna trees of Brazil have megafaunal fruit.

Megafauna fruits have much larger seeds that non-megafauna fruit of the same size. What is the value of large fruits? They have larger seedlings, can establish in less favourable habitats, can survive being partially eating.

How did these trees survive 10,000 years without megafauna?
He describes his research site in the Brazilian pantanal, include exotic megafauna like livestock and feral pigs. Collected lots of dung and scats and looked for fruiting, and also did camera trapping to same who came to feed on these fruits. They show that agoutis can also disperse a lot of megafaunal fruits.

At least 25% of megafaunal fruit can resprout and reproduce vegetatively

Many megafaunal fruits are dispersed by humans (e.g cacao). The dominant dispersers were first megafauna, then pre-Colombian humans, then fire and exotic megafauna. A succession of dispersers

Megafauna gut retention time is a function of body mass. They are using this to model seed dispersal kernels to simulate seed dispersal. He compares the Pleistocene dispersal kernel with that of the Holocene (where it is dominated by tapirs). The dispersal kernel did drop but not hugely (by a factor of about 1.6)

What are the consequences of reducing dispersal distance. The extinction of megafauna dispersers means that gene flow is occurring mainly by pollen movement (Collevatti et al. 2003 Molecular Ecology). They are compiling genetic differentiation of megafauna fruits with other zoochoric fruits. We can detect the “megafauna footprint” in the genetic structure of plants.

Yadvinder: there is some debate as to whether megafauna were abundant within the Amazon forest biome, as opposed to in the savanna. He has not studied this specifically but he things that megafaunal fruit will be abundant in the rainforest.

Adam Wolf: are there any long-term consequences because of the genetic shifts?

No evidence of any megafaunal dispersed seeds having gone extinct. Humans played a major role in their dispersal..

Ahimsa Campos-Arceiz: The rapid decline of Asia’s megafauna and its consequences for seed dispersal

3/19/2014

Shows a picture of durian fruit, which draws a few laughs and memories. Wallace: “to eat a Durian is a new sensation worth a visit to the East”. Durian is trying to attract dispersers such as elephants.

Seed dispersal efficiency is a product go quality, quality, diversity of seeds and seed size.

In the Anthropocene there has been a downgrading of body rise (Hansen and Galetti 2009). Being a megafaunal syndrome plant is risky.

Tropical asian megafauna include elephants, rhinos, bovids and tapirs. Shows massive declines in range (from Mahmood et al in prep). Forest elephants have lost 95.1% of range, Indian rhinos 99.8%, Javan rhinos 100%, Sumatran rhinos 99.9%, Malaysan tapirs 98.0%. All monogastric megaherbivores have lost >95% of historical ranges!

Are there megafaunal specialisations in Asia? Yes. Lots of evidence from Africa. He has compiled a database of plant traits using Guimaraes et al (2008) criteria. How are megafaunal plants doing? We have no baseline data. Only in Singapore. 50% of megafaunal plants in the forest are critically endangered. Campos-Arceiz and Blake had a paper on elephants as dispersers of seeds.

Human-mediated changes in behaviour. Elephants seek human-made gaps such as roads because food is more available (more grasslands). Elephants near the road eat many more grasses and smaller home ranges.

Very large intact seeds of Borassus are much more common in male dung than female. Males (they are larger) are more important for seed dispersal but also more vulnerable to hunting (they only have tusks) and more likely to experience human-wildlife conflict

Rhinos are too endangered to be studied - the research would attract too much attention

Tapirs and sun bears can do medium sized seedsand have some functional overlap with megafauna. Rodents remove a substantial proportion of seeds in traps. What happens to them?

Conclusions: tropical Asia has a rich megafaunal community but this is rapidly declining and “we are moving to a Neotropicalization of the region”. Many tropical Asian plants show adaptations to megafaunal dispersal. Some but not a lot of functional redundancy with smaller dispersers.

Question: do elephants belong in Borneo. Reply: There is SE Asia islands are an oddity, they have only been islands for only a small fraction of the last few million years. There are plenty of megafaunal plants in Borneo. The question is not ecologically relevant.

“The urgent things don’t let us see the important things. In conservation we are always fighting fires With the pressure of loss from deforestation donors are not so interested by seed dispersal, which would take centuries for the consequences to play out”

Jens-Christian Svenning: Pre-extinction ecosystem structure in Europe

3/19/2014

Megafauna were standard components of most ecosystems (Smith et al 2010 show how body six increased over time). So the current megafauna-poor world in an anomaly. We had elephants everywhere in the late Pleistocene except Australia and Antarctica. He shows a very striking map.

How do megafauna-rich ecosystems function in the temperate zone? Shows artist’s impression of Norfolk in Pleistocene with abundant megafauna. He shows megafauna of the last interglacial in Germany, including lions, elephants, rhinos, hyenas, wild horses, leopards

There has been a big discussion about whether the temperate forest biome was a closed forest, a mosaic, a pasture or a no-analogue system. The paleorecord says that the meso and small biota were largely the same as modern. But the megafauna are gone.

If you try and make a synthesis of the interglacial paleorecord, it says there were many trees in the landscape, but also evidence of more open vegetation (meadow/grassland pollen), plant macro fossils in small forest hollows. Clear signs of open vegetation near rivers, and on poor soil. We had a forest ecosystem but it might have been more heterogeneous than we normally associate with the forest. In the vertebrate fauna you see species of mixed habitats, such as horses, fallow deer and grazing rhino (Dicerorhinus), and grassland small mammals (field voles) and birds as well as woodland species.

There are some outstanding questions:

What were the densities of large herbivores?

How much of the vegetation was closed forest, and how much was open or semi-open?

Recent study by Sandom et al (2014, PNAS) shows herbivores had a positive influence on diversity. Used beetles as indicators - useful for ecosystem reconstruction as many beetles are very specialised in habitat needs. The UK has by far the best record on fossil beetles! This analysis is focussed on the UK.

They look at four periods.

Last interglacial (132-100 kyr) - no evidence of any humans, and certainly before modern humans

Last ice age (50-15 kyr)

Early Holocene (10-5 kyr)

Late Holocene cultural landscape (2-0 kyr)

The look and dung beetles. Pasture dung levels at LIG were as high as cultural landscape (this says high densities of herbivores in LIG)

Woodland dung is high in early Holocene, but lower in LIG

Hence there were often high herbivore densities and mosaic landscape during the LIG. More open vegetation in LIG, more closed forest in the early Holocene. Clearly shows very high species diversity and mosaic landscapes in the LIG. The study is biased towards lowland sites near floodplains

What is the relevancy beyond Europe.

Asia also had high megafaunal diversity in the temperate forest zone (e.g. Japanese elephant and giant deer). Same story in temperate forest zone of North America, and the “temperate” high Andes of Ecuador! (Sanchez et al 2013 PPP). Dung beetles are an underexploited resource for paleoecological studies.

David Nogues-Bravo: could today’s European forests support high densities of megafauna.

Jens-Christian Svenning: I don’t see why not.

Liesbett Bakker: Effect of herbivores on terrestrial and aquatic ecosystems

3/19/2014

Herbivores affect structural diversity, plant diversity, nutrient cycling. To understand these effects we need a common framework. She shows a cross-site study on grasslands in North America and the Netherlands. They have large herbivores (cows, bison) and also smaller herbivores (deer, pronghorns). They do exclosure studies on these grassland sites and look at effects.

1. Impact on plant diversity (Bakker et al 2006 Ecology Letters). Change in diversity is proportional to primary productiction. In more productive systems the herbivores increase diversity. In less productive systems they decrease species richness. Why is this?

In a low productivity system (e.g. Short grass step) the above ground biomass is low (below ground biomass is high) so herbivores have little influence on the vegetation. The small herbivores can get at plants and remove some species

In a high productivity system (e.g. Flood plain grassland), the herbivores prevent domination by a few plant species. The large herbivores facilitate small herbivores such as rabbits. Rabbit burrowing (1% bare soil coverage) provides a regeneration niche.

2. Impact on nitrogen availability

In N poor systems, herbivores have a positive effect on nitrogen availability by transforming biomass into dung. In N rich systems however, herbivores can have a slight negative effect on nitrogen availability She shows a conceptual pathways by Schrama et al (2013). This may be an effect of soil compaction and trampling having an effect of slowing down the nitrogen cycle.

She shows that clay fraction is a good proxy for soil compactability. Under intermediate soil texture and moisture conditions herbivores increase N availability, but under high clay or high soil moisture conditions thet can inhibit N cycling.

Should we introduce livestock into nature reserves? Are they the best proxy we have for extinct fauna?

In productive grasslands, there is good quality food for herbivores, and well-defended shrub species. In Netherlands, thorny shrub is much moore common on rich soils. This is a type of habitat where grazers have been influential, and would be suitable for livestock grazing.

She did an exclusion experiment in a shrub-grassland ecotone ecotone. There was woody encroachment, but actually the most important influence was from rabbit grazing, not from livestock.

In aquatic systems, similar principles may apply. She did exclosures in ten different wetland areas. There was expansion of vegetation inside exclosures, driven by muskrats as “aquatic rabbits”

Aquatic megafauna role? These include hippos, moose and capybaras. Very little known about their roles. Meso fauna such a beaver very important too.

Conclusions

There are some general rules of herbivore impact. They increase plant species in productive ecosystems, facilitate smaller herbivores.

Chris Doughty: Are nutrient limitations a consequence of the Pleistocene megafauna extinctions?

3/19/2014

We know very little directly about Pleistocene megafauna behaviour other than their body size. But by looking at contemporary megafauna when can strong correlations between body size and day range, gut residence time, population densities.

They developed a model to examine lateral diffusion by megafauna, with phosphorus as an example mineral. The lateral diffusivity is a power law scaling function of body mass.

They model the nutrient diffusion from the floodplains of the Amazon to the wider Amazon basin. Argue that nutrient diffusion now is still not in equilibrium after the megafaunal extinctions. Doughty Wolf and Malhi (2013) Nature Geoscience.

Next look at sodium, which plants try their best to get rid off but animals are desperate for. Many continental interiors far from the sea are sodium limited, because sea salt aerosol input is important. They now apply the same model for sodium diffusion. The model suggests that 15,000 years ago the megafauna diffused sodium inland and also reduced concentrations in plants near the coast.

How to test these predictions? Look at transects of sodium in Ghana and Gabon

Did the extinctions affect tree diversity. Apply the same diffusion model to seed dispersal and see how tree ranges differ. This is work in progress.

Did the Pleistocene extinctions affect land cover reflectivity (albedo) and climate? Albedo can be more important than carbon as a tree effect. Animals can be affected by climate, but can also affect climate. Doughty Wolf and Field (2010), Geophysical Research Letters

Jim Estes: there is a whole parallel literature developing in the marine literature, with deep diving megafauna playing a key role in nutrient dispersal in the oceans.

Travis Knowles: Temperate mountain grasslands as ghosts of megafauna past

3/19/2014

Appalachian balds - mountain top grassland ecosystems that are anomalous under current climate regimes. They cause has been a long debate. Some have been created or modified by humans, but he argues that many of these have a megafaunal origin.

Many of the plants found there are Pleistocene relics (e.g. Greenland stitchwort). Lots of endemic plants also that are found only in these ecosystems. Many of these are light demanding plants that do not survive under trees - a botanical signature of bald antiquity

Variety of ideas to cause of balds (climate, insect attack, soils, fire) but none of these factors are different on other surrounding peaks. Human causes are also unlikely given their antiquity
Now there is a lot of woody encroachment of these bales.

Additional case studies of Oregon Coast Range Balds, Poloninas (or poloninys) in East Carpathians.
There are several features in common: they are sub-alpine, grass dominated communities, richly diverse shade intolerant communities, south of glacial maximum extent, and all had history of pastoral grazing after human settlement.

Some work shows that these sites corresponding to a climate-imposed treeline during the peak ice age. This is the origin proposed. But then he argues that megafaunal browsing maintained these grasslands after their origination

Owen-Smith’s “keystone megaherbivirore” hypothesis. Abundant evidence of effects of contemporary herbivores today in altering vegetation structure. Elephants breaking trees.

How were the balds maintained after megafaunal extinction? There was a chain of herbivores. The successors were bison, elk, deer that maintained these until they were taken out by European hunters, then maintained by pastoral grazing. Today the only bales persisting are those actively managed. They are missing their natural disturbance regime of grazing animals.

These balds represent ghost ecosystems; remnants of larger open expanses in a disturbance-maintained non-equilibrium state.

Weigl and Knowles Biological Reviews summarises this argument - is available online.

Frans Vera: why are the successors of a megafauna not able to keep areas open. You don’t need elephants to keep out a tree, just a deer to take out the seedlings.

Jacquelyn Gill: A natural experiment in keystone herbivore removal

3/19/2014

The Pleistocene extinctions removed a whole functional guild of browsers and grazers. The ecological role of these large animals has been relatively unknown. Modern megaherbivores have a wide range ef effects, including browsing, grazing, seed dispersal.

What drives vegetation dynamics during the Qauternary. At large timescales it is climate, but at fine scales the pattern is less clear. There is vegetation change, but also climate warming and cooling at Pleistocene-Holocene transition.

We do know that during deglaciation, novel vegetation associations were widespread across eastern North America. Plant communities were very different from today - swathes of spruce/ash parklands (no modern analogue). What has driven the formation of these novel associations? Climate change?

Niche theory predicts that novel communities can arise in novel environments. Establishing the order of environment events is challenging. She uses dung fungus Sporomiella as a proxy for megafauna abundance. The analysis combines new pollen analyses from classic no-analogue communities.

Applemean Lake, Indiana. 12 kyr transition from spruce to pine to oak, matches dung fungus decline. Ecosystem novelty is high in the period immediately AFTER the decline in herbivores - allowed deciduous taxa such as ash and ironwood/hophornbeam to thrive. See a spike in charcoal immediately after megafaunal decline (maybe indicator fs loss of herbivory causing build up of biomass and fire fuel)

Now she shows a bunch of data from six sites in the great lakes region. Across all these sites there is ecosystem novelty immediately following the collapse of the megaherbivores.

She then does a massive ordination of fossil and modern pollen assemblages across North America. Axis 1 picks up composition (decid-evergreen) and axis 2 is structure (prairie to forest). Fascinating description of sites jumping from sone stable state to another following megafaunal collapse, with novel and unstable ecosystems in the transition. These sites become ecologically novel, but they are all becoming novel in their own way, and possibly less resilient in their own way. This makes it difficult to predict impacts of extinction or rewilding. Then climate change takes over and ecosystems move to their current states.

Clearly shows widespread impact of megafauna and their loss

When you see evidence for a vegetation driven shift driving herbivore decline, try flipping that in your mind and see if the reverse is true.

Chris Johnson: Australia’s megafaunal extinctions: causes and impacts

3/19/2014

Australia is important to the debate of megafaunal extinction because it is very different (“a separate creation” - Charles Darwin) and humans arrived much earlier 40-50 kyr BP, well before the end of the ice age, so there is less confusion with climate change. It is a valuable independent test case.

It has been said there are not enough data; Chris disagrees. He presents the radiocarbon extinction chronology of megafauna in Australia. Shows clearly they all go around 40-5 kyr. There is archaeological evidence that humans were widespread by 45 kyr. Sporomielia at Lynch’s crater also decline at the same time. Pretty convincing there was a megafaunal mass extinction.

Could climate play a role? He shows a climate-niche model for megafaunal species for the last 120 kyr using Hadley Centre GCM. They matched location records to climate at that time and place. The model shows climate was improving and range sizes would have peaked at around 30 kyr. There is no signal at all of a climate shift that could have caused megafaunal range reduction at the time they went extinct.

The extinctions were severe: nothing above 40 kg survived - more severe than even the Americas. Plots risk of extinction as body mass. He replots this as extinction risk vs fecundity and the American and Australian curves collapse into one curve. If you were an animal with less than one offspring per year you were highlight likely to go extinct. Exactly the same story in Australia and the Americas. The best predictor of megafaunal extinction is reproduction rate. “Mega” is a red herring - size did not matter - what mattered is fecundity and rate of population growth. The sustainable harvest rate for kangaroos is 40%, for diprotodons it would have been around 5%.

Pretty convincing evidence that the Australian megafaunal collapse was driven by human arrival. What about the ecological impacts of the extinction?

Evidence from Lynch’s crater in Queensland. There were rainforest angiosperms and gymnosperms until around 40 kyr, then shift to fire-prone sclerophyll forest (eucalyptus). That shift has always been blamed on human arrival and fire use. How do animals like Diprotodon fit into that time sequence? They look at dung spores as an indicator of megafauna. The Sporomiella drops around 40 kyr, at the same time when charcoal appears in the record, with a charcoal spike around 500 years after the arrival, then increase in grasslands. He shows the velocity of change in each of these proxies (100 years sampling resolution). High inverse correlation with abundance of megafauna and total tree influx.

Summary: started with mixed sparse rainforest and sclerophyll forest, little fire (no modern analogue to this ecosystem). Changed to uniform dense sclerophyll forest; this area changed to wet rainforest in the Holocene.

Where would we expect the most change in vegetation? At the broadest scale vegetation is controlled by climate, but there is a sliver of climate space where you cannot predicted vegetation type. William Bond argues in those climates vegetate type is controlled by the removal of biomass by herbivory and fire. This climate space is a huge part of the Earth’s land surface.

Lynch’s Crater is within this uncertain zone. Their second site (Caledonia Fen) is outside this uncertain zone, and there is little evidence of vegetation change after megafaunal loss.

To summarise:

Human impact, not climate, caused megafaunal extinction

Extinction was closely followed by vegetation change and increased fire in some places but not others, depending on the climate space.

Sandra Diaz: analogy with kangaroo survival in Australia and guanaco/vicuna survival in South America. Both have reasonably high fecundity.