Wednesday, August 22, 2012

Justifying assumptions: tests of seed size/mass tradeoffs

When ecologists develop theory and models, we generally need to make assumptions. The nicest definition of an assumption is that they are the framework we use to capture our beliefs about a system. All future analyses will treat these assumptions as true, and so ultimately the validity of a model is tied to the validity of its assumptions. As Joseph Connell said: “Ecological theory does not establish or show anything about nature. It simply lays out the consequences of certain assumptions. Only a study of nature itself can tell us whether these assumptions and consequences are true.” Often times the most interesting advances in ecology come when we questions popular assumptions, such as that species are ecologically different, that interspecific differences are more important than intraspecific differences, or that ecological interactions occur much more rapidly then evolutionary changes. 

Assumptions in models and theory can often serve as a sort of shorthand for ideas that there is some general evidence for, but for which comprehensive data may be lacking. Community ecology is full of assumptions about functional tradeoffs that mediate coexistence between species. Various assumptions about plant species coexistence include that species experience tradeoffs between competition and colonization, growth versus reproduction, or seed size versus seed number. A simplistic explanation for such tradeoffs is that you can’t do everything well: a strong competitor can’t be a good colonizer too, which creates opportunities for strong colonizers but poor competitors, etc.

Tests of these functional tradeoffs are lacking, or lag behind the theory that relies on them. For example, the idea that there should be a tradeoff between seed size and seed number has long been proposed to explain why plants have highly variable seed sizes. Plants with small seeds should produce more offspring, and these seeds should be more successful at reaching empty sites. Large seeded species should be more competitive in the seedling stage or more tolerant of difficult conditions, and so have higher survival. Theoretical models that rely on such a tradeoff suggest that many species could co-exist and that the resulting community would exhibit a wide variety of seed sizes. 

But though many studies and theories depend on this assumed tradeoff, a comprehensive experimental test was lacking. Ben-Hur et al. have finally provided such an experiment, testing the basic prediction that a negative correlation between seed size and seed number should increase species richness. They also tested whether small-seeded species were more likely to remain in the community when this tradeoff existed, increasing the amount of among-species variation in seed size. To do so, the authors created 3 ‘community treatments’ of 15 plant species. The abundance of each species in the starting seed mix was manipulated to create either (1) positive correlation between seed mass and seed number; (2) negative correlation between seed mass and seed number or (3) random allocation of the 15 species regardless of seed size.

From Ben-Hur et al. 2012. Ecology Letters. a) Final number of species in the community, when the correlation between seed size and seed number is negative, random, or positive. b) Seed mass distribution in community under positive correlation between seed size and seed number (left), and negative correlation (right). 

Ben-Hur et al.’s results strongly suggest that a seed size/seed mass functional tradeoff can increase species richness (figure, a). Further, when there is such a tradeoff, the variation in seed size represented in a community increases, again in agreement with predictions (figure, b). The results are particularly convincing because the authors used experimental manipulation of the strength of the correlation (i.e. from negative to positive) to test its importance. The authors suggest that the tradeoff they simulated did not involve competitive differences (i.e. was not a competition-colonisation tradeoff), and more likely reflects a trade-off in establishment probability and colonisation (Dalling and Hubbell 2002; Muller-Landau 2010).

Of course, these results represent relatively short-term coexistence, and community richness may have changed had the experiment been allowed to continue for longer. But as a starting point, this suggests that theories that rely on functional tradeoffs in seed characteristics to explain coexistence are capturing a mechanism that has some experimental support. 

Friday, August 10, 2012

Things I've learned at ESA this year

1. It's more useful to talk to people than it is to be an audience member.
2. A successful talk is one that produces interactions with people.
3. The grass is not always greener- the talk you missed was probably not as great as everyone is saying anyways (actually it probably was, but it's too late now...)
4. Picking only specific talks and people to hear can be a good strategy for avoiding talk burnout. Symptoms of talk burnout include napping in conference centre hallways, feelings of annoyance when you hear the same concept re-explained for the 10th time (which is probably because you're in the 7th Community Patterns and Dynamics session), and a desire to yell 'but what is your hypothesis?!' during talks (this may just be me). The only cure for this is to go have a drink.
5. Conversely, sitting through entire sessions can lead to important discoveries.
6. There are more areas of research in ecology than you can list: by bringing these researchers together, ESA is helping to foster continued growth in our field. Integrating all these bodies of knowledge is important if ecology is to be a healthy, mature discipline.

08/10/6:50, edited for clarity. #7 could be 'it's better not to blog while tired'.

ESA day 3: a meeting of meetings

Wednesday was a crazy day, bouncing between talks and one-on-one meetings. This is what ESA is about: connecting with friends and colleagues, and seeing exciting new science. There were a bunch of fun talks that introduced new ideas and concepts, or made connections between different approaches. Some of these talks included Dylan Craven, who linked plant functional traits to performance in secondary successional forests in Panama. In Nicholas Gotelli’s talk, he tried to reconcile thousands of museum ant records with ecological surveys to estimate abundance, distribution and numbers of ant species in the north east USA. Sam Scheiner discussed a new approach to combine phylogeny and traits at the community level.

There were also some talks that seemed to really resonate with me, and the audiences attending them. Katherine Richgels gave a very interesting talk on trematode metacommunities, where the primary patches (snails) live in other patches (ponds). The primary patches have unique dynamics, including movement. The environment and host abundance seem to strongly determine trematode community patterns.

Bruce Menge astounded his audience with a new hypothesis: the ‘intermittent upwelling hypothesis’ which states that ecological process rates should be maximized at intermittent upwelling coastal zones. He ran experiments on coasts around the world and showed that recruitment, herbivory and predation rates were all maximized when upwelling was intermittent.

Cecil Albert showed how a model can predict the effects of global change and landscape alteration. She used a ‘sandwich’ modeling approach, where vegetation structure is sandwiched between climate change influences at large scale and landscape change at smaller scales. The resulting vegetation changes can be used to predict responses from specific indicator species or ecosystem function. She then showed how different scenarios of landuse change (random habitat removal, zoning and protecting corridors) can result in different responses in indicator species.

Finally, Caroline Tucker* gave a great talk on the effects of global warming on changes in flowering time in competitive communities.  Most people assume that plants will flower earlier in a warmer world, but these predictions ignore competitive effects. Using a set of linked growth and phenology models, she showed that indeed plants increase growth and flower earlier with warming in the absence of competition. However, once you allow the species to compete, the advance in flowering time is unequal. Early species, which are generally released from competition will flower earlier. So too will late species which tend to be good competitors. However, intermediate species do not advance their flowering due to competition.

*Yes this is our Caroline Tucker.
**Caroline has been on me to post my Wed. talk summary for two days.

Thursday, August 9, 2012

ESA 2012: Day 4

 For some reason, Day 4 had many talks I wanted to see, just when the effects of late nights and over-caffeination were starting to peak. The reward to remaining awake through a day of talks was that I got to hear some excellent ecology.

At 8:20 (yes, 8:20) in the Biodiversity III session, Xubing Liu spoke about some of the work his research group is producing to expand our understanding of the Janzen-Connell effect. (For a good example of this work, see The Janzen-Connell effect is a density-dependent mechanism in which proximity to individuals of the same species increases their chance of encountering species-specific predators or diseases, and therefore reduces their chance of survival. This is hypothesized to produce coexistence by maintaining lower abundances and higher diversity. In this talk, Liu explained how intraspecific variation could similarly be maintained via a Janzen-Connell effect. He showed experimentally that decreasing the degree of relationship between two individuals of the same species (increasing intraspecific diversity) increased their odds of surviving fungal infection. Such a mechanism could help explain how intraspecific variation is maintained, which is a hot topic these days.

A talk I found particularly interesting, perhaps because it was so different in content and style from my own work was that by Robert Beschta from Oregon State University. He convinced me, without statistics or plots, that the outcome of a natural experiment – the removal of apex predators from America’s park system – was highly detrimental to those ecosystems. Removal of wolves and cougars from National Parks such as Yellowstone and Olympia have produced many changes in community structure and function – the understory disappeared as deer and elk browsed all young greenery, river edges eroded without shrubbery, and forests aged. Yellowstone provided an additional validation to this conclusion; re-introducing wolves appears to be producing gradual reversion to more diverse and functional habitat.

Diane Srivastava further provided the type of perspective only gained from years of research. She also illustrated that the contribution of a body of work is often more than the sum of its parts. Diane has spent 15 years of studying a bromeliad system in which multiple invertebrates live in the water collected in the plants, forming a complex ecosystem with multiple trophic levels. The data collected over this time allowed her to perform a meta-analysis which shed more light on the dynamics of this system than any individual study allowed.

There were multiple talks from students of Peter Chesson, an eminent theoretician, and all shed light on mechanisms of coexistence. Although perhaps too complicated to explore in a short summary, they covered topics in keeping with other work from the lab, especially the role of temporal and spatial variability in driving fluctuations in recruitment and ultimately coexistence, and in understanding how mechanisms will scale with space. His students were well informed on the intricacies of Chessonian theory and the talks certainly created lots to think about.

Finally, two talks discussed the growing problem of reconciling trait- and phylogenetic-based community ecology. Rebecca Best presented the results of a amphipod competition experiment, in which she examined whether feeding traits or phylogenetic distances were a better explanation for the resulting diversity and abundances. She found, as is not uncommon, that traits were by far more useful in understanding the amphipod community. She didn’t stop there, however, and tested further how the phylogeny and trait values actually related – it turned out that traits and phylogeny were not correlated, and represented different mechanisms at play in the species' ecologies. Though she found that phylogenies could not predict the outcome of her community experiment, she concluded that this didn’t mean that phylogenies were not important, only that they were important at different scales or in different mechanisms then she had been focusing on.

Finally, a talk directly relevant to Best’s work came from the EEB & Flow’s Marc Cadotte. Since it was a well-received and interesting talk, I feel like giving his talk a plug here isn’t too biased. Cadotte presented a metric meant to incorporate both trait and phylogenetic information, and further to incorporate them in a meaningful way. Name FPDist (for functional phylogenetic distance), this metric incorporates an additional axis (functional diversity): this can be represented with a phylogenetic tree in which the x-axis represents trait distance and the y-axis phylogenetic distance. This allows you to visualize trait divergence and convergence in a way that traditional trees cannot. Further, the metric he presented is a function of both traits and phylogeny, combined in such a way that the relative importance of each can be captured and recognized. This allows us to more fully investigate both traits and phylogeny contribute to community diversity. No doubt an interesting paper will follow soon.

Off to survive one more night and one more morning.

Wednesday, August 8, 2012

ESA Portland day 2: march of the phylogeny

The day phylogenies took over. This is how I would describe the talks I attended on day two. There was a palpable collective enthusiasm for what phylogenies can bring to understanding ecological patterns. It seemed like every session I went to there were several talks that test for phylogenetic patterns and it will be interesting to see where this all goes in the future. For me, this phylogenetic onslaught was heralded by the very first talk I went to by Jeannine Cavender-Bares. She spoke about how phylogenetic relatedness and species traits can provide important insights into community patterns and ecosystem function. She ask some of the most pertinent questions such as: how do evolutionary processes affect ecological processes; and how deep in the phylogeny is the evolutionary signal in community assembly. This last question is interesting because it can potentially tell us about past environments when certain lineages evolved. Her talk was divided into three parts. In the first part, she discussed how certain plant traits, like specific leaf area (SLA), were correlated with fire frequency. At extremely low and high fire frequencies, there is a strong trait pattern associated with communtiy membership, and with a strong phylogenetic pattern as well. But this wasn't the case with intermediate fire frequencies. In the second part, she discussed plant community patterns across an urban to natural gradient. There were important trait differences, with species having smaller seeds and higher specific leaf area in urban areas. There were more species in urban areas, but they represented less phylogenetic diversity than in natural areas -meaning that there is an environmental filter selecting for similar species. In the third part, she investigated oak adaptive radiations in North America and the resulting biogeographical patterns. There we differences in diversity across latitude, with high diversity regions also have more close relatives.

The were a number of other very interesting talks, and I spent the day fluttering from room to room, like a confused butterfly in search of sweet rewards. And rewarded I was. There were handful of very memorable talks. By both young graduate students and established researchers. Christina Lamanna gave a nice talk about phylogenetic and functional diversity (PD and FD, respectively) across an elevation gradient, which in part she used to highlight a new measure of species functional overlap. Richness and FD peak at intermediate elevation. She also examined the turnover in FD and PD and that both of these show decreasing turnover at higher elevations. At high elevations, PD was found to be overdispersed as were some of the traits, but other traits appeared underdispersed, indicating the combination of traits under very different selective regimes.

In a session on ecosystem function Jane Cowles told us how diversity and warming interact to shape patterns of ecosystem function. The experiment was great, and they overlaid warming arrays on some of the plots at the classic biodiversity experiment at Cedar Creek, Minnesota. The arrays warmed 1.5 and 3 degrees on 1, 4 and 16 spp plots and they measured aboveground and belowground biomass. More aboveground biomass was observed with warming, but not for belowground, except for deeper roots. Dominant species increased the most in aboveground biomass, seeming to respond to large pools of nitrogen available in early spring. 

One of the two best talks I saw today was given by Amelia Wolf. She constructed a biodiversity-ecosystem function experiment based on realistic scenarios of species loss. Whereas most experiments randomly assemble species together, realistic species loss selects species with certain traits, and once they are lost, those species are not part of the system at lower diversity. She used 20 years of observational data to select those species most susceptible to extinction and then created a series of plots where diversity was based on removing susceptible species. These plots were nested in that when a species was excluded from say the highest to next highest treatment, it could not be included in a lower diversity treatment. She compared this to random diversity treatments and found that the realistic species loss had a stronger effect on ecosystem function. But she suggested that this could be due to the nested structure and not the realistic scenario. So, to cover all her bases, she created 32 different nested loss regimes that were not the realistic one, and found that they were no different than random. Thus species identity and susceptibility really matter for ecosystem function decline with species extinction, as most susceptible species are often from the same functional group.

The other superb talk was from Jay Stachowicz on the influence of eelgrass genotypic richness, relatedness and trait diversity on productivity. Genotypes interact though a number of mechanisms including competition, cooperation, interbreeding, and so there are complex possibilities for the influence of genotype on productivity. From experimental combinations, he found that, counter to his expectations, plots with closer relatives had higher productivity. Further these plots with close relatives also had greater trait diversity, highlighting the complex nature of species interactions and differentiation.

Andrew Siefert gave a talk on disentangling multiple drivers on species turnover in space. Betadiversity is driven by both niche based decay of environmental similarity and stochastic due to dispersal limitation. Both generate similar patterns. But if one uses functional traits, then you can see higher or lower functional turnover than expected from chance, which indicates niche based turnover. He reported the results from 1500 forest plots across eastern USA,  with climate data and data on four functional traits. He found high turnover in soils and species, lower for climate and functional diversity. Both taxonomic and functional betadiversity best explained by climate. Close sites have high taxonomic turnover, but low functional turnover, thus climate filtering.

Finally, Elizabeth Boyle exmined arthropod phylogenetic community patterns in near arctic aquatic systems (ponds, streams, rivers, etc.). These habitats harbor an amazing diversity of insects and Elizabeth collected data from dozens of habitats over a large area, for hundreds of species and constructed a molecular phylogeny based on her own genetic work. An amazing effort for a masters project! She resampled the habitats through the summer and found that many of the habitats started off as phylogenetically clustered but became overdispersed through time. But not all habitats showed the same response, and she found that some environmental variables seemed to be strongly correlated with relatedness patterns. She also questioned whether the emergence of adults caused some of these patterns as the timing of emergence is phylogenetically nonrandom, which to me is a new explanation of potential phylogenetic patterns.

- Posted using BlogPress from my iPad

Tuesday, August 7, 2012

ESA Portland: Day 1

Today proved a typical first day of ESA, with delayed flights, hotel difficulties and luggage to carry around. Then you head to the convention centre and are reminded all over again just how big ESA actually is. The benefit of the crowds is that the sessions take on a specificity and quantity that you can't find anywhere else. The bad news is that you will have to make choices.

Today's choices weren’t too difficult - I moved between the Community Assembly and Neutral I and Community Pattern and Dynamics I sessions to start. Some common themes emerged, especially that people are quite interested in the relationship between diversity and phylogeny, and then phylogeny and traits, and also in patterns of beta and alpha-diversity along environmental gradients. 

A few talks stood out: Emma Moran spoke about identifying the processes of deterministic assembly and stochasticity that drive community diversity. In agreement with previous work from Jonathan Chase, her co-author, she spoke about how using null models allows scientists to differentiate between these two processes by observing temporal and spatial patterns of species diversity and comparing them to those patterns expected by chance alone. By looking at both temporal and spatial patterns of diversity, it is possible to differentiate between stochastic arrival of species at a site, followed by deterministic interactions (spatial stochasticity + temporal determinism) and purely stochastic assembly (both temporal and spatial patterns of diversity random), for example. Using both simulations and empirical data, she demonstrated the patterns of diversity that might be expected and tested for under these scenarios. When she used a null model that controlled for random expectations, her conclusions about which processes were important were dramatically different from those arrived at without a null model.

In the Population Dynamics session, Emanuel Fronhofer asked 'Why are metapopulations so rare' and came to the possibly controversial conclusion that they are rare because there aren't many conditions that should result in metapopulations. Metapopulations are a common concept in ecology, based on the idea that population dynamics in different patches are linked via dispersal between those patches. However, it's unclear how common metapopulations really are in nature. Fronhofer used individual based models (IBMs) to explore the range of dispersal values, environmental stochasticity, or reproductive system type, for example, that would result in a metapopulation. In particular, he looked at the most strict definition of a metapopulation: occupancy of patches less than 1 and more than 0, turnover through time, and FST values such that populations are genetically differentiated. What he found agreed with the nay-sayers: only quite narrow values of parameters like dispersal resulted in true metapopulations. Does this mean metapopulation ecology is a highly specialized field? Difficult to say, although it maybe that a particularly stringent definition of a metapopulation (with occupancy between 0 and 1, for example) is not necessary to describe the movement of individuals and alleles between patches in a way that is consistent with metapopulation dynamics. 

Finally, Geoff Legault discussed spatial synchrony among populations, in which population cycles in different spatial patches become synchronized. In particular, using a protist predator-prey microcosm, he showed that in agreement with some theory, there is a dispersal threshold after which synchrony is achieved between the two populations. This leads to interesting questions about what determines the level of dispersal required to produce synchrony, and how factors such as population growth rates alter this threshold, something which a microcosm is particularly useful to address. 

Wednesday, August 1, 2012

EEB & Flow Portland bound

Just a heads up that Marc Cadotte and I will be live blogging the Ecological Society of America's Annual Meeting in Portland, from Aug. 6-10th. As always, this is a great chance for ecologists to hear about great science and run into old and new friends. If you see Marc Cadotte there, be sure to harass him to post on time, as he claims to be 'busy' ;)