These are my notes from IBS 2015, as a sort of not-so-live blog. I’m still figuring this scicomm thing out and I’ll probably have a livelier way of putting up notes next time, but I feel like the Twitters don’t give me a chance to document the conference the way I’d like. All smart stuff can be attributed to the various smarties presenting, but all the mistakes are mine. Also, when I have an aside, I’ll put it in bold. So, my dead blog of #IBS2015:

Day 1: Friday, January 9, 2015.

No notes on the first talks because I was too jet lagged. Note: Germany is nine hours ahead of Eugene, so it was midnight at 9 AM…

11:00ish: Scheiter et al. Modeling functional trait diversity. aDGVM (adaptive Digital Global Vegetation Model) does very well simulating African savannas. Does it do well with Australian ones? Not at any useful scale, so they re-parameterized for Australia. This means that e.g. stem diameter doesn’t relate to canopy diameter & height the same way in these two systems. I would suspect phylogenetic constraint…. 

Today, many models use PFTs – plant functional types – to parameterize; lump species by shared functional traits to keep system simple. aDGVM2 is a trait-based dynamic veg model, uses an individual-based approach, so it doesn’t lump species by PFTs. Models natural selection, mutation and cross-over to optimize for productivity. Cool. The model produces emergent coexisting life history strategies: these are not programmed in, but arise from the interaction of the digital plants and the physical environmental parameters.

11:23: Ruiz-Benito et al. Trying to understand how 21st century climate change will affect functional diversity of European floras. The current consensus seems to be that Europe will lose tree diversity and overall biomass, producing a carbon source. These authors want to know whether higher functional diversity might protect some forests from climate change. FD had an effect on growth and regeneration, but not mortality, but Functional Identity had an effect on mortality. These effects were different in evergreen, broadleaf, and Mediterranean forests, so those cannot be expected to respond to climate change in parallel ways.

11:35: Hawkins et al. Begins with an apology: there will be no biology in this talk. I do a silent happy dance. Also, he says that his papers are sets of maps with filler text, which seems a good idea to me. Now, onto the science: begins with a division of drivers of community structure into extrinsic and intrinsic variables. Traits are intrinsic variables. He now suggests intrinsic drivers may not even show a correlated pattern, let alone mistaking correlation for causation. I’m not sure I understand this point. He shows that cherry-picking random ‘traits’ can produce patterns that fit any particular hypothesis. I think the point of this talk is that dumping trait data into biogeographic models (of, say, richness) you can get very high R-squared: many of these methods seem to overfit as a consequence of the spatial patterning of the species. Species aren’t randomly distributed, so models can see patterns where there are none. This problem also appears when modeling random traits to predict random macroecological values (modeling, e.g. Functional Diversity).  An issue I see here: the apparent strength of random predictors means that even if you’re not statistically fishing, but testing a hypothesis, you cannot be confident that you’re actually seeing a biologically meaningful effect. What’s the answer? We don’t know yet, but we’re a smart community of scientists and we can work on this.

It was in this talk that he gave what became the tagline for the meeting “Don’t trust maps!”

In the Q&A, one of the questioners (Pedro ?) suggested the problem has already been solved in publications…

The jet lag got me really bad for the rest of Day 1, so I spent the time until the poster sessions chatting with colleagues in the hallway and procrastinating working on grant proposal and classwork.

Day 2: Saturday, Jan 10, 2015

This is the paleobiogeography session! I’m in my element.

8:30: Stigall gives an introduction and overview of the principles of paleontology for the neontologists in the audience. I’m a jetlagged zombie, so I’m zoned out through most of it. She has a critical part at the end where she distinguishes between completeness and adequacy in the fossil record, beginning with Darwin’s apologia for the incomplete fossil record and then reviewing 1) Phylogenetic completeness, 2) Stratigraphic completeness, and 3) Geographic coverage. As it happens, the fossil record is known adequately in these three ways to test many, many paleobiogeographic hypotheses.

8:52: Servais and Harper on the origin of biogeography. They are from a different world, not digging in Quaternary mud, but doing real, hard fieldwork in real rocks… Mentions Wegener and Continental Drift, but Servais and Harper are working before Pangea. How can they reconstruct the shape of Earth’s geography before Pangea? Let’s find out… In the end, you have to pull together all of the lines of evidence, from structural geology, sedimentology, etc. etc. (Note: Pangea looks an awful lot like Pac-Man. Coincidence?). Then you make a hypothesis and test it with new data. Science!

Now he’s linking Sepkoski’s paleodiversity curve to break up and unification of supercontinents: Times of lower diversity have been supercontinent times (Rodinia and Pangea) and times of higher diversity have had distributed continents (like now!). Servais et al. (2009) GSA Today.

This example is followed by several others that I won’t record because I’m feeling lazy.

9:15 Kiessling and Kocsis: Marine benthos over last 300 million years… Starting out with Valentine’s seminal 1970 work on species richness over geologic time. He explained this pattern with an increase in provincialism after K-T. However, Miller et al. 2009 used the PaleoDB to show no significant change in provincialism through the paleo record.

Kiessling points out that the PaleoBioDB is millions of records freely available to everyone. It’s a great resource.

Then a little dig at Servais’ talk, pointing out that you cannot use paleocontinental reconstructions from before Pangea because the researchers are still moving the continents around. So, he’s focusing on the Pangea to today part of the record.

Climate context: Warming and cooling in two cycles from Permian to today.

He’s fine-tuned a system to diagnose biogeographic provinces in marine inverts over time, producing a movie of province changes through time. He has good evidence that the number of provinces is driven strongly by sample size, so must standardize for sampling. Once you do, the variation in number of provinces through time disappears. The only interval that seems to stand out is a dramatic loss of provincialism in the wake of the P-Tr extinction.

9:39: Denk, on whether or not we need fossils for historical biogeography. Shifting to Cenozoic terrestrial plants. Then he uses quotes to show that over the last 20 years, biogeographers have shifted from shunning fossils as unnecessary in a Bayesian phylogeographic context to part of a continuous paleo- and neo-biological dataset. I can dig it.

Now a plot of angiosperm diversity through time from Willis and McElwain 2013 (The Evolution of Plants)

Example 1) Smilax, greenbriar. Reminds me of Smilex gas from the old Tim Burton Batman movie. Good fossil record, good phylogeny. Four clades identified in recent phylogenies. Three new-old world disjunctions in these four clades. Mid-Miocene fossils of Smilax in Iceland suggest the disjunction of some clades might be because of a bird-accessible North Atlantic crossing. Maybe not quite a continuous land bridge, but a convenient bird corridor. Plus, fossils show that the one clade that currently has no new-old world disjunction did, back in the Miocene. Really, everything was just better in the Miocene.

Example 2) Quercus, the oaks. They have a rich fossil record and a strong, well supported (almost oak-like?) evolutionary tree. See what I did there? The surface texture of the pollen grains is distinct among the deepest-diverging branches of the oak tree. Heh. Oak tree. The white oak tree seems to germinate in the high arctic of the Eocene and then spreads its leafy boughs over the temperate zone with climate cooling through the Tertiary.

Example 3) Dracaena, dragon trees. There’s no phylogeny for this group: sad because there’s no Dragon tree tree. Modern distribution: disjunct distribution on coast of western, eastern, and southern Africa. Middle Miocene fossils from Turkey! Fossils suggest the ancestors of modern disjunct species had distinctly different ecology from modern groups. Convergent ecological divergence among these lineages? If you had a tea tray with the phylogeny of Dracaena on it, it would be a dragon tree tree tray. And.. if someone was giving away carts made from Dracaena, you could get a dragon tree wagon free. If someone carved the phylogeny of dragons onto one of these trees, it would be a dragon tree dragon tree.

10:33 Sanchez Meseguer et al. The present isn’t always the key to the past, when there is sampling bias (e.g. extinctions in phylogenetic trees of extant taxa). This problem can have a major effect on reconstruction of ancestral areas. For a moment, clouds part and sunlight obscures the screen, but then the gray day returns and the lecture continues.

Hypericum example. Originates in the palearctic and doesn’t spread to the rest of the world until the Pliocene. Modern phylogeny says that it originated in Africa, Europe, and North America with previous clades in Africa – disagreement! So, they added ancestral inferences from fossils. Next, they fit niche models to fossil distributions and paleoclimate simulations to get climate tolerances for the earliest members of the genus. In this way, they could see climate barriers to distribution in Beringia and equatorial regions. These climate data explain the palearctic limitation of the clade until Pliocene cooling. More detailed analysis also suggests bias in extinction with more eastern palearctic lineages going extinct, producing the erroneous pattern in the phylogenetic reconstruction of ancestral regions.

10:48 Badgley – Continental gateways and mammalian faunal change. Four-part biogeographic states: permeable/impermeable barriers and changing/stable climate. Four unique macroecological predictions for the four states.

Example 1) Siwaliks of Pakistan. Focus on mammals >1kg. Floral turnover from C3 to C4 grasslands through this sequence (told from carbon isotopes). Most of the sequence has endemic species in genera shared with other Asian sites: so regionally similar on the broad scale, but different in details. Three of the four biogeographic states recognized in this sequence. 14 out of 16 predictions upheld by observations, suggesting support to this model of climate/barrier interaction.

Example 2) Miocene mammals of Iberian Peninsula. Three analysis intervals, with two of the four states. Environments reconstructed from paleobotanical remains + stable isotopes from mammal teeth. Domingo et al. 2014 Paleobiology. 10/12 predictions upheld by evidence.

Example 3) Quaternary mammals of Cape Fold Belt of South Africa. (Faith and Behrensmeyer 2013, Paleobiology). Area of ecosystems and connections fluctuate with sea level change during glacial/interglacial cycling. At Pleistocene->Holocene, ¾ predictions met, and the last one, too, but in a nuanced way. In particular: grazers hardest-hit by the climate change.

11:17: announcements, etc. etc.

11:25: MacArthur and Wilson Award presentation and lecture. Given to Dan Rabosky, U Michigan. “Speciation, extinction, and the geography of species richness.” (As an aside, this award has only been given twice, both to men, but the sample size is too small to be concerning the way the male-domination of the Paleo Society young scientist award is.)

He’s asking us all to visualize a heatmap of global species richness of our favorite group… Then he goes to the Fine et al. 2006 (book chapter: no link) plot of tree richness and unlogs the diversity axis: the diversity of non-tropics is like a thin green veneer compared to the diversity of the tropics.

Now squamate reptile richness: snake richness peaks in the tropics. The other ‘lizards’ add a peak in Australia (in desert scrubland: a desert ‘coral reef’, he explains). Other environmentally comparable deserts have nothing like this diversity of lizards. An example of a non-latitudinal diversity anomaly.

What processes can cause these gradients? Mittelbach et al. (2007) Ecology Letters framework: Ternary diagram of Ecological/equilibrium vs. diversification rate vs. time.

Today’s focus: “Speciation rate” is not a single hypothesis.

Rhode (1992) Oikos: higher metabolic rates in the tropics.

Biogeographic species pumps: Mosaic habitats of Australian arid zone. Pianka (1972) Copeia: Habitat fluctuations lead to speciation.

These all predict that speciation rates should be higher in hotspots. Does speciation rate explain major gradients in vertebrate diversity? (I like that he has an actual question with a ‘?’ here).

Diversification rates on phylogenetic trees: need a modeling process that can deal with changing evolutionary dynamics through time and space, shaping the resulting evolutionary tree. BAMM: Bayesian Analysis of Macroevolutionary Mixtures. Machine learning method to optimize complexity of model to explain rate structure on the tree.

Now an example from Australian sphenomorphine skinks. Most diverse terrestrial vertebrate radiation in Australia. Big, big tree. Rabosky et al. (2014) Sys. Bio. Creates Bayesian 95% credible set of rate shift configurations.

What about extinction? He’s not going to interpret extinction rates from molecular phylogenies. One would need to integrate fossils to get any kind of confidence. Davis (no relation to me!) et al. (2013) BMC EvoBio. Testing effectiveness of reconstructing speciation and extinction rates from model phylogenies. Any phylogenetic tree is compatible with a range of extinction events from zero to infinity.

Nee et al. (1994) Phil Trans B: phylogenies contain max recoverable information about species rates towards the present.

Example: Speciation and geography of marine fish diversity. Collaboration with a bunch of people, whose names I didn’t get because I’m a fail slow typer.

Building a spatial database for living fishes from OBIS, VertNet, FishNet2, GBIF. The records are biased towards onshore fishing areas. Need to standardize for sampling, so using, as a start, rarefaction. Once the data are rarefied, they make more sense…

Is the Indo-Pacific a species pump for fishes? BAMM analysis across whole living fish tree, translated to the map using their spatial database. Then a map of mean speciation rate for all available grid cells. Surprise: the highest rates are in the Arctic and Antarctic! It’s like the speciation rate is the inverse of the latitudinal richness gradient. So, instead take the latitude-midpoint of each species, but it still produces a smile-shaped plot with highest rates at the high latitudes. Then he goes through a bunch of specific examples: Arctic/Antarctic clades have high rates of speciation. Famously diverse tropical fishes have low speciation rates.

What’s driving this pattern? “I have no idea.” But he can tell us what it isn’t.

Not driven by

  • 1 hemisphere
  • well-sampled clades
  • shallow-water clades
  • poorly-known tropical faunas

Short on time, so one more story in 4 quick slides. Whew! This talk is a wringer.

NW land birds dominated by tropical sub-Oscine clade. Now plot rates on using BAMM. Lots of clade-specific variation in speciation, but no relationship with latitude. Tropical and temperate rate distributions are identical. See also Jets et al. 2012 and Kennedy et al. 2014.


  • We can make high-res macroevolutionary rates.
  • Need the fossils to do extinction, and he’s extended BAMM to work with total evidence trees that include fossils (yay!).
  • No matching of speciation rate with high species richness. Evidence against a broad class of mechanisms that have been invoked for creating species diversity hotspots. Not a lot of evidence for species pumps, at least for vertebrates. Plants, maybe different.

Lunch was good, chatting with Catherine Badgley and Kaitlin Maguire about neoendemism projects.

I had talks picked out for after lunch, but I’m really feeling the jet lag now, so I’m taking a mental health break instead. Looks like the next talk for me is at 14:30. I’m really looking forward to the evening’s activity: banquet at the steam train museum!

2:48 Renner volunteers to not give her talk to get the session back on schedule, but there is a general outcry from the audience, so she goes ahead! So, how to assign fossils with too few characters to place them precisely on a phylogeny? Don’t want to throw out those data. My battery is dying, but this talk is derived from a 2014 paper, so you can check it out later. Heath et al (2014) PNAS.

Day 3, Sunday Jan. 11, 2015

8:38: Rodrigues-Sanchez introduces the session on using past climate to inform predictions of response to future climate change. Cites Clark et al. (2001) Science: Need a better ability to make ecological forecasts. Emphasizes 1) climate is changing and we need to anticipate the future and 2) climate has always been changing, so we can use past data to test models of ecological responses.

What can we really learn from the past? What? E.g. species response to past climate variations. How likely? How often have species had consistent responses to climate change? Why? What processes are actually controlling range change, plastic response, and extinction?

8:47 Steve Jackson, begins by pointing out that the last IBS he went to was the first, and his talk there was his last one using 35mm slides! Times have changed quickly.

Sources of knowledge about biotic responses to climate change:

  • Direct observation
  • Models
  • Geohistorical records

Now he’s going to give some for-instances of how paleoecology can help conservation biogeography.

  1. How sensitive are natural populations and communities to climate change?
    1. Silver Lake example of stasis, ecological persistence through changing climate
      1. How are they able to persist?
    2. Long-term population persistence – example populations >10k yr in one place.
      1. Range of environmental variation
      2. Physiological tolerance?
  2. What are the biodiversity costs of climate change?
    1. Examples, including Tony’s Irish Elk study.
    2. Needed: a systematic inventory of Quaternary casualties
      1. Fossil evidence
      2. Genetic diversity
      3. aDNA studies
      4. Phenotypic plasticity?
  3. How and how far do species move in response to climate change?
    1. Note from me: He’s using ‘migration’ for this movement, but I would not.
      1. Migration is a regular, cyclic range change as part of a species’ natural history.
      2. This sort of range change might be called immigration or emigration, but it is one-time response to climate change over multiple generations, so not migration.
    2. How to discriminate between real ‘migration’ and apparent ‘migration’.
      1. Example Williams et al. 2004 Ecol. Monogr. Fagus grandifolia moves north, but McLachlan et al 2005 Ecology DNA shows that they were already in the north.
    3. What are the relative roles of environment, biotic interactions, and chance on ecological and biogeography outcomes of climate change?
      1. Lesser’s studies on Bighorn basin ponderosa pine
      2. Gill’s work on no-analog forests in Upper midwest
      3. Figure from Jackson and Blois 2015 PNAS (in press). Extent by duration in log-log space with processes shown at different scales. I need to find this figure.
      4. Need to carefully consider the complex interactions among these factors.

Wraps up with a call for better interdisciplinary interaction, better consideration of taphonomy, and overall epistemological rigor. And don’t forget natural history: these are real individuals in real species, real ecosystems.

Know what’s at stake for decision makers: what are the consequences of errors? Scientists have relatively low consequences for being wrong. Not so for conservation planners.

9:13: Morlon on past effects of climate changes on biodiversity.

Current geographic gradients suggest climate has a control on diversity, morphology. Similarly when you look at species richness/traits versus temperature in the fossil record (example from Ceonozoic pollen, brachiopods). Gilooly et al.2001 Science: temperature affects metabolic rates, body size.

So, she presents some hypotheses about the processes that would link temperature changes on speciation rates. Overall hypothesis is that speciation rates should be positively correlated with temperature. Similarly, extinction rates should be negatively correlated with temperature, and rates of phenotypic evolution should be positively correlated with temperature.

How to test? Traditionally, using modern gradients and the fossil record. E.g. Ezard et al 2011 Science. However, looking only at the modern ignores the contingent evolutionary history. Now she just suggests that some groups (birds, insects?) have fossil records too poor to be used for this type of study. That seems too defeatist to me (referring back to Stigall’s adequacy vs. completeness bit from yesterday).

Now she’s onto the idea of using genetically-based phylogeny to make tests of these hypotheses using comparative methods. E.g. Nee et al. PTRSB 1994. Morlon et al. 2011 PNAS, Morlon et al. 2014 ELE.

Condamine et al. 2013 ELE. –> Incorporate external knowledge of temperature, sea level change, etc. into modeling of rates on phylogeny. Interesting. New in prep paper tests 21 different models of the way temperature affects rates. Results for mammals: support positive correlation between temperature and diversification rates, as expected from hypothesis.

Cantalapiedra et al 2013 PRSB: similar results for ruminants.

So: could decreasing temperatures in the Cenozoic explain the widespread observation of declining speciation rates over time through many comparative studies?

One of her collaborators is working on incorporating environmental dependence in Brownian model of trait evolution. Interesting: signs point to an inverse relationship between size evolution and temperature (faster trait evolution in cold intervals). C.f. Cooper and Purvis 2010 Am Nat.

She hopes to convince us that phylogenies can be useful for understanding the interaction between evolution and climate change. Quest completed.

The Q&A focuses on two things: the disconnection between speciation rate and mass evolution rate and the problem with using global climate to model a local process (natural selection/speciation).

9:40, Orsini

How natural populations respond to environmental change: move, go extinct, or evolve.

Need to understand feedbacks between phenotype, environment, and genotype. Problems: phenotype-genotype link is only established in model species, but with little knowledge of their link to environment. Similarly, ecological model species are not known in genotype-phenotype linkage. She would like to close this loop.

Complexity of natural systems:

  1. Interaction of environmental gradients
  2. Interaction of neutral and adaptive processes
  3. Interactions of past and present (we need to know the past for its effects, evolutionary baggage)

Need accurate measurements of effects of environmental change on natural populations. But we have no long-enough-term datasets. How can we do this without investing multiple human careers? Using omics on short-generation time species in the lab, like yeast. Also, we can use space for time, e.g. Gillespie et al 2013.

But here is her new proposal: Orsini et al 2013 TREE. Evolutionary time machine: sample at different spatial scales with multiple time series. It looks like she’s integrating across multiple sediment cores taken from a single lake. Yes, she is looking at waterflea microevolution over space and time. Aha! She can re-activate resting stage waterfleas trapped in the sediment to see what different time steps do in “common garden” experiments. She has several citations here: Orsini et al 2011, 2012, Pfrender et al. 2012, and others. This seems a very elegant use of natural experiments, isolating sediment cores that have one or another factor controlled so she can see what each one does to the local lineage. They found a clear genomic signature of natural selection and repeatable patterns of local adaptation.

This project shows the prospect of a new field of paleo “omics”.

Now for the coffee break, so I’m off to put my presentation on the computer. Except… There were no people or computers in the room, so I’ll get to put my talk on the computer at the last minute. As one does.

10:37, Fordham: Extinction risk from anthropogenic change, fossils and molecular log books.

Moving beyond predictions: using predictions to improve reserve design, plan translocations.

Wants to use the late Quaternary as a natural experiment to improve predictive models for policy makers. Again, emphasizing the need for integrated approaches across paleo-, neo-, et al. This is a refrain of the morning’s presentations: if we want to actually save any bits of biodiversity or natural ecosystem function, we need to all work together.

By combining aDNA, climate, and conventional paleoecology data, one can create a model of the demographic changes that led to extinctions in the past, providing an empirical basis for extinction risk given current or expected demographic shifts.

Is our current high forecast biodiversity loss a real thing or a consequence of model bias? He makes a good point here: testing climate biodiversity models using 20th century data can’t capture the range of expected climate change into the 22nd century. You MUST go back to the Quaternary fossil record.

Nouges-Bravo Plos Biology. Sorry I got distracted by his plot from this study, so I looked it up on PLoS and skimmed it. Link:

OK, back to the talk: Mechanistic underpinnings of climatic refugia. Blois et al. 2011 Quat Sci Rev. Williams et al. 2000 Journal of Biogeography.

Handy link to the paper underlying this talk, from the Twitters:

11:05, Jansson et al. Effects of stability on bird diversification.

Dynesius & Jansson 2014 Evolution: does change or stability foster diversification more? Reminds me of the old Plus ca change model of Sheldon 1996 Palaeo3.

74 sister groups of birds from New World. Used niche models to calculate areas in LGM and today to measure persistence. They’re using genera and one of the characters seems to be species richness within the genus. They’re using sister genera so they can do PIC on the pairs.

Climatic stability was significant to diversification only for tropical genera. Interesting.

11:21, de Boer and Hooghiemstra: Paleoecology of Mauritius. Settled by humans very late (1600s), so excellent natural lab to distinguish human from natural ecological perturbations.

Cool: mass-death assemblage records 100 years of dying during global megadrought 4200 years ago. Sediment cores indicate the blooming of toxic cyanobacteria during this interval, killing the vertebrates using the waterhole.

Pollen record from the central crater shows stability until the start of the Holocene, at which time there is a revolution in the island’s ecosystem.

Lowlands and uplands have different disturbance regimes, do not respond the same. Take home: be sure you have analogous systems before extrapolating from one to another.

The Q&A seems centered on whether or not he can call his bone assemblage built up over 200 years a ‘mass death’. In the end, it is a terminological concern, but I’m not sure the two sides are understanding one another.

11:38, Nogues-Bravo: summarizing. “The unknown future of biological diversity lies in the past.” Nice. Examples of what we need:

  1. aDNA to get at demographics through time.
  2. Models based upon these data to forecast range dynamics using current genetic diversity data.
  3. Modeling niche space through time, instead of as a static thing from modern time slice.

Another call for integrating data and methods across disciplines! And he wrapped fast to keep us on time.

I led the paleobiogeography student discussion table at lunch. We had a lively chat, ranging over topics from the climate zones of Oregon to how to use niche modeling techniques to predict fossil localities.

1:01, Hoogiemstra et al.: Islands and sky islands.

Sky islands in the N Andes. Most of these islands have existed in this way for <5Ma. Defining sky islands by their tree lines. How much time is available for connectivity among these islands? There is a lot of fluctuation in the tree line, with the highest expression only 2000 years ago. The modern high tree line and divvied-up sky islands are the exception, only about 10% of the time in the Pleistocene. You must test ecological theory against glacial settings, more like 60% of the time. He’s suggesting that these sky islands are not obeying the MacArthur and Wilson island rules, because they have not lost species richness, despite losing 96% of their expressed area since the glacial maximum.

Now ocean islands. Where is the species pool for the change at the onset of the Holocene in Mauritius (see earlier talk)? He suggests the gallery forest. Also, the forest is down to 2% of pre-human levels on the island, but they still have high species richness… The unanswered question: how long can it last?

1:15. Cool. I gave my talk and folks liked it. No tomatoes. I was on time, too…

1:33 Gutierrez-Garcia et al. Opens with a Jurassic Park reference: when it came out she thought there was little chance of getting any DNA from any fossil. No longer. However, there are no current tropical aDNA sites. Loltun cave: Late Pleistocene to Holocene fossils. Ototylomys phyllotiys, an extant central american mouse, is found in the cave. Nice! They made a custom 3D scanning system just to get a surface model of the tiny, tiny jaw they were going to destroy for aDNA. Wow! The cave sample made a new, basal clade outside the entire crown group of the species. (Gutierrez-Garcia et al. Biology Letters vol 10).

Q&A points out that they could not micro-CT the specimens because it would cook the aDNA. Hence the super fancy structure-from-motion micro scanner.

1:44, Smith et al.

Trophic downgrading and loss of megafauna is a big deal today, and there is a lot of public support for protecting the big animals. Cites Estes et al. 2011 Science paper about trophic downgrading in the wake of large animal loss. Felisa suggests we can use the wake of the paleoextinction to model what to expect in our neo-extinction future. She’s going to tell us what happened to the survivors in the Hall’s Cave, Texas, fossil record. Pre-extincntion, the cave record has a very large number of species, many of which are megafauna. Only 3 of the megafauna species survive. The body size distribution changes in the wake of the extinction, and not just in the loss of megafauna. There’s also a bigger turnover in species composition than the change in mass distribution. Details of PAIRS results wash over me in rapid succession. Take home: more significant segregations than associations. Many pairs of species were pushing one another out.

2:03, Milan Chytry et al. A refugium in southern Siberian mountains.

European full-glacial fossil sites often record tundra, taiga, and steppe organisms together for both plants and animals. A no-analog community? He asks whether there is a potential modern analog. Mammal and plant distributions of tundra, taiga, and steppe organisms coincide in southern Siberia. The pollen record confirms the hypothesis that this area matches the unusual full-glacial sites. The snails also agree. So do the mammals.

2:18, Willner et al. Florsitic legacies of Quaternary climate change in European Beech Forest.

He’s looking at the distribution of the understory plants from beech forests to see if they have a pattern of distributions that matches the hypothesized refugia. In the end, they do.

I skipped the next session of talks to chat with folks and be generally zoned out.

Now up: the Wallace Award, awarded to Simberloff this year.

Character displacement and release in the small Indian mongoose and the stone marten.

Links his topic to Wallace through Wallace’s work on the effect of introduced predatory mammals on islands.

1872 introduction of Mongoose to Jamaica –> there to kill rats, it became one of the worst 10 invasive species. It was subsequently introduced to many islands worldwide, and it has always been a disaster. In its native range, it’s the smallest of three mongooses, but in all introduced areas, it was put into a carnivore-free system. So: it should make a good test of Brown & Wilson’s 1956 ideas about character displacement and release. Would it get bigger in the absence of competition?

Measured two traits: skull length as proxy for body size and upper canine length.

Simberloff et al. 2000: first publication on this project. In most islands, males got bigger and females did not, so sexual dimporhism also went up.

Mediterranean islands off Croatia some have stone martens as well… Hypothesized to defeat male body size increase. Yep it seems to work.

The problem is in the exceptions: Several islands do not fit the trends.

Now details about Stone Martin distribution. Interestingly, the invasive island forms are smaller.

The mongoose won’t cross water without help, so rivers are important boundaries for them.

The data are broadly consistent with the hypothesis of character release.

So ends my notes of #IBS2015. I look forward to the next meeting, January 2017 in Brazil!