Oxford Megafauna Meeting
  • Home
  • Conference Blog
  • Programme
  • Contact
  • Meal
  • Map & directions
  • Accommodation
  • Gallery

Sergey Zimov: lessons from a Pleistocene rewilding in Siberia

3/20/2014

2 Comments

 
Unusual sediment (yedoma) covers all of northern Siberia. This is the fossil soil of the mammoth ecosystem. Te diversity was close to that of the African savanna. In high attic, if there is grazing or nutrient introduction highly productive grassland can appear.

In Wrangel Island it is very cold but mammoths managed to live here because they managed their pastures themselves.  In cold or dry climate decomposition can be very slow, but in Pleistocene ecosystems decomposition was independent form climate because in a mammoth stomach temperature is around 36 C all year round.

In the 1930s Gorodkov estimated Wrangel Island could support 300-400 reindeers. In 20000, the number renderer exceeded 16,000 and mud ox 1000. By far the high biomass of all Russian national parks, in the coldest climate

Estimated mammoth density in Pleistocene was 1 mammoth per square km, 5 bison, 15 reindeer, 6 horses. Total biomass around 10 t/km2


Planet, herbivore and predator productivity was high in far north.

In Pleistocene park in winter only excrement covers the soil. Recycling of plant matter is rapid. Introduced animals converted boggy wetland to dry meadow.  Horses, bison and musk-ox plus three species of deer.

Mammoth biome was the biggest biome of our planet. Walter et al (2006)

Various pictures show how adding grazing causes grasses to appear, covering bare soils. Grasses have co-evolved with herbivores, Grasses grow only with herbivores.

Photos of national park in Netherlands. Deer in forest kill one tree per day.

What is the total biomass of megafauna in our planet before humans expanded? Use methane as a proxy of megafauna biomass. In Pleistoce methane production from herbivores estimated to be 100 Tg methane per year.

Prior to appearance of grasslands around 20 million years,  plants responded t herbivory through chemical defences, decomposition and recycling were slow. Then with grasslands a new animal-vegetation symbiosis  appeared, with rapid recyclng that maintained high animal biomass systems.

Humans knocked out half of this symbiosis (the megafauna), caused grasses to retreat into forests. Humans lost their traditional food source, but this opened the way for artificial grasslands (farmlands and pastures). Our cattle and horses are Pleistocene fauna. But even now we have not reached the biomass of animal biomass in the Pleistocene.

He suggests that Pleistocene predators shepherd “their” herds in a similar way to You don’t need brains to maintain herds - leaf cutter ants could do this. Dogs can maintain herds from their genomic memory.

He has a new rewilding park experiment south of Moscow. He advocates new rewilding experiments in fertile regions with good climate, in contrast to current national parks.


2 Comments

Frans Vera:  30 years of rewilding in the Netherlands

3/20/2014

2 Comments

 
A 6000 ha reserve was created on land reclaimed from the sea. Greylag geese arrived and settled the new land. Through their grazing the facilitated many other species - they were keystone species. The geese and dynamic water table drive succession (not respond to succession)

The geese’s complete biotope is the marsh and grassland combined. 

Then he introduced wild cattle to maintain the grassland. Critics said wild ungulates create a closed canopy reserve. Two large herbivores went extinct  auroch 1627 and tarpan 1850. Grazing guild completely vanished. Other herbivores became functionally extinct. 

The closed canopy forest is an incorrect baseline of the natural vegetation in Europe. But cattle prevent the regeneration of trees in a forest. The conclusion is that a wood pasture is a degraded ecosystem. If nature is to be restored you have to remove the cattle. And extant ungulates need to be culled to low levels. All the wood pastures changed into closed canopy forests. These closed canopy forests of Germany are called the new primeval forests of the future. In the former wood pastures oak and other light demanding tree species disappeared within decades. He shows forest structure graphs for 4 species in a national park in Denmark. In the park, 50% of all vascular plants disappeared within 40 years (the light demanding species)

In the wood pasture we see a mantle and fringe vegetation of sloe/blackthorn (Prunus spinosa). Within it there are everywhere young oaks within this fringe. The oaks are planted by the jay hoarding acorns. The regeneration of trees in a wood pasture is in the mantle fringe. When you have hawthorn they also nurse an oak. Hazel is a a light demansing shrub planted in the sloe by the nut hatch. The end result is climbs of trees surrounded by a scrub; a grove. A very diverse landscape driven by grazing animals (Bakker et al 2004).

In a closed canopy forest the sloe is disappearing. There is no shrub layer. Forest-grown trees with narrow crowns because they compete for light. When trees dies the slow occupies the gaps and the grove retreats.

You don’t need elephants to open up a forest. The wood pasture is a non-linear cyclical succession

He argues that cycling wood pastures are the natural early Holocene ecosystem of Europe. Sloe is insect-pollinated so not visible in record. Also, grazed grasses rarely flower so no pollen. If it does flower it is in the mantle vegetation so is not wind-spread. Only tree pollen appears in the long term records, giving a false impression of closed-canopy forest. 

Saproxylic beetles demonstrate that there were large, open grown trees

Back to the Netherlands hypothesis. The herbivores introduced were regulated bottom up by the amount of food. There is no such thing as a carrying capacity - every year is different. 

The false baseline is that herbivores must have been in very low densities otherwise there could not be a forest. So bison and red deer numbers are kept at too low a density. A wrong theory can cause an overkill. Bison are an IUCN red list species yet are still shot to maintain low densities.




Sprint shrubs became established at high deer densities




In the Holocene large herbivores are keystone species and drivers of the natural succession.




Vera’s hypothesis explained here:




http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2745.2004.00964.x/abstract




http://www.knepp.co.uk/Other_docs/Frans%20Vera/Birks%20over%20Vera%20in%20TREE1.pdf


2 Comments

Carlos Peres:  Ecosystem effects of contemporary defaunation in Amazonia

3/20/2014

0 Comments

 
He shows the extent of game harvesting in Amazonia. The pattern of harvest varies with village size (rodents and possums in depleted state). The vulnerable, low fecundity species start dropping out at a human density of 1 person per km2 (Peres 2011 Conservation Biology)

They have censused vertebrate populations throughout the Amazon.  Map hunting pressure throughout the Amazon (all sites are structurally intact). Did lots of paired comparisons.

The harvest-sensitive species are largely frugivores. Frugivory is the unifying guild. In non-hunted areas large bodied primates dominate community animal biomass. In hunted areas they are largely gone. The build a demographic model of source-sink dynamics.

Over 3-5 years the community skews towards the smaller bodied species. There is density compensation and a lot of the smaller-bodied species come back.

Cascade; hunting pressure increases -> frugivore populations decline -> dispersal imitation -> plant fitness reduced of large seeded species -> some density compensation -> shifts in community structure and biomass.

Janzen-Connell effects inhibit recruitment and survival of sapling near the mother trees (pathogen load).

Woolly monkeys alone are responsible for scattering a million seeds km2

One fruit syndrome common in Amazonia: large seeded, indehiscent (don’t break open, need a passage through a gut) fruits are particularly vulnerable to frugivore loss.




They do a model simulation where they replace random trees by non fruit-dispersed trees. Many frugivore trees tend to be heavy wooded. This suggests that forest biomass will reduce over time as frugivore-dispersed trees drop out and are replaced by lower wood density species.




Adam Wolf:: at what time scale do you imagine this loss of biomass to take place




Carlos Peres: one tree generation time.


0 Comments

John Terborgh: Africa rainforests "the odd man out" - the effects of forest elephants on tree diversity and structure

3/20/2014

0 Comments

 
PW Richards (http://ind.pn/OzMlyf ), the father of tropical ecology, proposed that Africa is the odd man out.

Tree diversity in equatorial Africa is much lower than in the Amazon.  50-75/ha vs >200 Equatorial SA. Parmentier et al (2007) and others discuss that area, drought, fire, contraction through history are primary reasons for low species diversity

He will present biotic arguments rather than abiotic arguments for the low diversity.

Small islands such as the Solomons have twice the tree species diversity of Central Africa

The total flora of Gabon is quite rich (> 6000 species). In a tropical forest flora about a third of the flora are trees (around 2000 tree species). A hectare contains about 1/10 of regional (Gamma) diversity, so we would expect about twice as many species per hectare than we see.\

He plots species accumulation curves for various tropical forests. Cameroon is no different to other tropical forests with similar climates. The problem is the per hectare diversity is much lower. Why?

He compares Gabon with the Peruvian Amazon. Lots more trees per hectare in Peru (mean 618) than Gabon (around 420). When you look at forest structure, 

Much much lower diversity in small and large saplings, somewhat lower than small trees. Most diversity in African trees is in the large trees (at similar levels of diversity to the Amazon).  The Amazon has most diversity in small trees. Gabonese forests have no middle storey and understorey.

Gabon has megafauna. Forest elephants are different from savanna elephants. They don’t knock over trees, they snap them to feed them. “In Africa you don’t need a machete to go through the forest, but you do need to be very careful”. Where there are no elephants the number of small saplings is indistuishable from the Amazon

How could he find an elephant free refuge to compare structure. They compared flat ground forests on slopes and tops of hills. But they found that the elephants go everywhere. Only one slope was so steep (36 degrees) that there was no elephant sign.

The peak height for trees pushed over by forests is 1-2 metres. Elephants are major filters on forest structure and diversity.

Yadvinder Malhi: the large trees are more abundant and live longer. Is there some sort of interaction where “weeding” of small plants enhances the lifetime of large trees?

There are two responses to herbivory: defence or tolerance (where you can bounce back)

John Terborgh: African forests are the most amazing and wonderful forest I have ever seen (and that is from a many who knows the Amazon forest better than anyone).

Andrew Balmford: How does the range contraction of Asian elephants (and forests in other parts of Africa).


0 Comments

Greg Asner: Effect of soaring elephant numbers on savanna community structure

3/20/2014

1 Comment

 
South African park managers ask the question: how to we manage for ecosystem heterogeneity in a landscape of fences. 

Elephant numbers in Kruger. Culling ended in 1994 (numbers around 9000). Now numbers are 16,000 and continuing to soar. Are these elephants making the landscape more homogeneous or heterogeneous?

They “tag” and monitor millions of trees, using airborne lidar (3D tree structure) and spectroscopy (tree species)

Stunning 3D fly through over area with elephants (few trees) and areas with elephants excluded (many trees). Browsers have it tough in one regime, grazers in the other.

There is a strong abiotic template. High fertility basalts on east, low fertility granites on the west. There is also a finer-scale catena linked to micro topography. More woody vegetation high on crests (Levick et al 2010 Nature Comm). Regional precipitation gradient from very dry in the north and fairly wet in the south. Woody vegetation is in low areas in the north, in uplands in the south. So need to be careful of abiotic heterogeneity.




Various types of exclude. Either all animals large than a hare, or only elephants and giraffes.




Background tree fall rate in exclosures increases with rainfall. Tree fall rates change by up to TWO ORDERS OF MAGNITUDE in the presence of elephants. (Asner and Levick 2012 Ecology Letters). IN these exclusres, tree number loss is strongest in the height range 5-8 metres - this is an “elephant trap”




Shows map of processes. Woody encroachment in background as well and elephant knock-down. Massive loss of Acacias. Park-wide changes in canopy height. There is some growth but it is trumped by losses. They assess if the change in fire regimes would affect tree fall rates. Less than 10% of tree fall in the park can be attributed to fire. The fire programme has had little impact on woody vegetation.




Question: there is a strong focus on outcomes and states in Kruger rather than processes (as pointed by Monbiot). Kruger is fenced. One key process in movement. How much of what we see is a result of confinement by fences or fear of poachers? Is this an unnatural system?


1 Comment

Megafauna conference slides and audio available

3/20/2014

1 Comment

 
Slides and audio from the conference are available for download at:
http://www.eci.ox.ac.uk/news/events/2014/megafauna/index.html
1 Comment

Rodolfo Dirzo: How extinction drives cascading effects on ecological processes

3/20/2014

0 Comments

 
Defaunation is a massive contemporary phenomenon. Lots of drivers and synergies amongst drivers. He shows a graph of bushmeat extraction per hectare in various tropical countries. In the Brazilian Amazon 16.6-23.5 million animals removed per year (Robinson and Redford, Peres 2000)

The characteristic extinction time (CET) - what is the time required in the future to put 50% of the species in a group into extinction. The timescale is around a few centuries

Map of density of endangered mammal species. Hotspots in tropics, especially SE Asia.

The defaunation gradient. Plots biomass/abundance of animals vs impact of defaunation drivers. Large vulnerable species decline quickly. Medium animals may increase in the short term (loss of predators) but decrease in the long term. Small, fecund, robust species increase in density and biomass over time, In the extreme case we can talk about the phenomenon of RODENTIZATION, or in Spanish, RATIFICACION.. 

Two cases.

1. The rodentation-defaunation cascade. Increased rodent abundance -> increased predation on small sets -> decline of small seeded species. Large wide life decline -> reduced predation of large seeds -> reduced loss herbivory of large seeded plants.  End result is reduction of small seeded trees and increase of large seeded trees.

In Los Tuxtlas, Mexico, understorey ends up dominated by large-seeded Astrocaryum mexicanum palm (Martinez-Ramos et al., in prep). As palm density increases, there is an decrease in sapling density of other species. This leads to impoverishment of future tree diversity.

2. African savanna example. Replicated experimental exclosures at Mpala Research Station, Kenya.. A cascade of effects resulting from experimental defaunation. 

Large wildlife removal -> cascade to other consumers -> effects on function and services -> effects on plant-animal interactions

High rodents increase ectoparasites which increase pathogens. The increase in number of rodents increases the density of infected fleas. 

Conclusions: The mesofauna of ecosystem are important functional components of the system and relevant for the well-being of humans

Andrew Balmford: situation is more nuanced when we consider introduction of livestock. Deforestation and fire also add to complexity.



John Terborgh: his results from the Peru forests is the opposite. Release of small seeded trees. The rodent community is different

Boo Maisels: the response in Central Africa is also the opposite. Good to compare across sites



0 Comments

Final day of Oxford megafauna conference

3/19/2014

0 Comments

 
Today is Day 3 of the conference. Yesterday evening was topped by a wonderfully evocative and passionate talk by writer George Monbiot, making the case for megafaunal rewilding of Europe. It managed to combine being scientifically informed, challenging policy at a fundamental level and having many moments of poetic beauty. Unfortunately, there was two much there and no powerpoint props to slow it down, so it could not be adequately blogged.


All blog posts thus far provided by Yadvinder Malhi. Please email me your thoughts and reflections if you would like them posted up to the blog - it would be great to have multiple views and perspectives.
0 Comments

Panel discussion: what are the major impacts of megafaunal extinction?

3/19/2014

0 Comments

 
Jens-Christian Svenning: the world today is divided into regions of ongoing megafauna attrition and megafauna rebounce in other parts of the world


Jacquelyn Gill: really impressive combination of paleo specialists with contemporary ecologists

Chris Johnson: in the modern world we are always being surprised by sudden changes. The paleo world is full of dramatic shocks. If we want to understand how rapidly the world can change we can look at the past.

Mauro Galetti: in Brazil most researchers work in a abiotic world view

Yadvinder Malhi: what is the most important ecosystem effect of megafaunal extinction?

Jacquelyn Gill: Cascading effects on small mammal communities and invertebrates. Impacts on aquatic systems, 

Mathias: co-extinction (e.g. dung beetles and mites, giant scavenger birds) and competitive release of other species

Chris Johnson: in Australia the biggest effect was the loss of predators, and resulting increase in abundance of mesopredators. The effect was disastrous for Australia’s ecosystems. Predators are important.

Frans Vera: many herbivores are still considered destructive elements of the ecosystem because we do not understand what is natural. We cull them to homeopathic densities

Sandra Diaz: the megafaunal extinction is yesterday in evolutionary time. With climate change we learn that evolution can be past. Have the megafaunal extinctions affected plant evolution?

Jens-Christian Svenning: we see microevolutionary responses but there are constraints in range of response.
Jens-Christian Svenning: pre-extinction carnivore guilds were hyper diverse, and there is really no modern analogue

Andrew Balmford: in marine ecosystems the megafaunal loss is happening in recent times.Any lessons between the terrestrial and marine worlds?

Jim Estes: something crystallised  for him today. When we talking about rewilding today it really is about bringing back carnivores. Now he realised that they were very different systems with big herbivores sitting up on the top. The Pleistocene was a very different system and a very different food web.

John Terborgh. What trophic downgrading means is successive loss of trophic levels. In African forests the herbivore biomass is more than an order of magnitude than it is in South America, In the contemporary world the predators are able to take any of the herbivores and control their populations,

Large animals move nutrients laterally and subsidise poor habitats, but they also hugely accelerate the mobilisation of plant biomass. The recycling greatly short-circuited nutrient cycling. In a post-megafaunal world this is left to detritivores and is much smaller.

Adrian Lister: we heard several examples today of environmental change following megafaunal loss. In these cases one is need of a null, e.g. run a vegetation model responding to climate change

Abby Swann: the models are not up to that yet. The relevant processes aren’t written in.

David Nogoue-Bravo: how far are we are from understanding consequences of shifts in fauna?

Elizabeth Jeffers: we can do that now, even if not very reliable for decision-making at this stage.

Jim Estes: can the oceans be an interesting comparison to terrestrial systems? There are important differences: there has been almost no megafaunal extinction in the ocean (Steller’s sea cow and Caribbean seal). Ocoan trophic structure is very different - the autotrophs turn over so quickly. The ocean sciences community are so focussed on physical forcing that they really consider what what role animals play. The coastal sees may be a useful comparison system.

Blaire Van Valkenburgh: there are lessons to be learnt from marine ecosystems in terms of rebounding effects, importance of redundancy, hysteresis.

Jim Estes: there is still the possibility of megafaunal rebound in the oceans.

Frans Vera: we have a lot of theories and models that should be tested in reality. He has seen that 80% of young die anyway without predation. Let’s practically test the effects. 

Rodolfo Dirzo: Janzen’s concept of the living dead. Species lingering on but functionally extinct.

Jacquelyn Gill: we actually do a lot about about the behaviour of extinct creatures

David Nogues-Bravo: what are the possible negative consequences of rewilding?

Jacquelyn Gill: driving species extinctions caused novel ecosystems, reintroducing them will also cause novel ecosystems

Jens-Christian Svenning: it would be easy to undo megafauna introductions if we needed to




Chris Doughty: Human-wildlife conflict is a major issue

0 Comments

Adam Wolf: Modelling lateral transport of nutrients by herbivores

3/19/2014

0 Comments

 
Maths as a tool to get insights into the past. He starts with a quote by Hairston, Smith & Slobodkin ("The World is Green”) indicating that herbivores must not matter in determining the effect of vegetation on climate. If it does not affect albedo or CO2 it does not matter.

Example of vegetation cover and convective clouds around the rabbit-proof fence in Australia. Their model suggests loss of megafauna in Eurasia caused warming as dark trees encroached into snow-covered grasslands.

Animals matter because they move and can transport nutrients and genes. This can be represented as a diffusivity in a random walk. How can we get numbers for the diffusivity?


Diffusion of animals is a function of body size. The diffusivity of materials is a function of how many animals there are, how much material there is in the biomass, how much they eat, how far they walk and their gut residence time.

They develop the model and evaluate it in Kruger National Park. He argues that it would be impossible to reconcile the nutrient budget of Kruger National Park without taking megafaunal nutrient diffusion into account.

Presents global picture of nutrient dispersal. Africa and South Asia are hotspots, everywhere else is cool and dark. How different this would have been without the megafaunal extinctions.

Andrew Balmford: IUCN maps are areas where they recently occupied. Where they actually exist is greatly reduced and restricted compared to IUCN maps

Some questions say that now we have to take  feral exotics into account into modern—day maps they are performing functions that extinct megafauna once did.



The papers he mentions are available here


Doughty, C.E., Wolf A., and Malhi Y. (2013) The legacy of the Pleistocene megafauna extinctions on nutrient availability in Amazonia, Nature Geoscience,Advance Online Publication. Supplementary Information.



Wolf, A., Doughty, C.E., and Malhi, Y. (2013) Lateral Diffusion of Nutrients by Mammalian Herbivores in Terrestrial Ecosystems. PLoS ONE 8(8): e71352. doi:10.1371/journal.pone.0071352. Supplementary Information.
0 Comments
<<Previous
Forward>>

    RSS Feed

Powered by Create your own unique website with customizable templates.