Contrasting changes in taxonomic, phylogenetic and functional diversity during a long-term succession: insights into assembly processes

Purschke, O., Schmid, B. C., Sykes, M. T., Poschlod, P., Michalski, S. G., Durka, W., Kühn, I., Winter, M., Prentice, H. C. (2013), Contrasting changes in taxonomic, phylogenetic and functional diversity during a long-term succession: insights into assembly processes. Journal of Ecology, 101: 857–866. doi: 10.1111/1365-2745.12098


Stora Alvaret on southeast of Öland with Eketorp Fortress in background (from Wikipedia)

HarisSL pic

Haris Saslis-Lagoudakis

Why this paper? I just saw it advertised on the facebook page of the International Biogeography Society by one of its authors. It’s not really my field, but the title seemed cool. Message to self (and all of us): be shameless to put your work out there.

What’s the paper about? The idea is pretty straightforward: after an ecosystem is disturbed, its biodiversity changes and goes through several steps before it reaches some sort of “equilibrium” again. This process is called succession. The paper I’m discussing today applies a comparative approach to determine whether succession is ruled by deterministic or stochastic processes, and whether these processes change during succession.

What did they do? Their study site is a 4.5 x 4.5 km landscape on the Baltic Island of Öland, Sweden. They assessed taxonomic, phylogenetic and functional (alpha and beta) diversity at different successional stages, as well as turnover between stages covering a more than 270-year-long succession from arable-to semi-natural grassland. Taxonomic diversity was measured based on presence-absence data from across plots. Functional diversity was measured from 11 functional traits associated with response to and/or tolerance of disturbance, scored for all plant species. Last but not least, phylogenetic diversity was calculated on a phylogenetic tree extracted from a published supertree of Central European vascular plant species. First, the authors quantified taxonomic, phylogenetic and functional diversity within grassland communities (alpha diversity), and turnover of these measures between communities (beta diversity), at four successional time steps. Grassland age (step) was assigned using a GIS overlay analysis of land-use maps from different times in the past, and they sampled 55 units for each successional stage. Then, they assessed whether, for each of the four successional time steps, species co-occurring within sites were phylogenetically or functionally more (or less) similar than expected, given the taxonomic diversity. Finally, they measured phylogenetic and functional turnover between successional stages.

What did they find? In summary, the main finding is that – perhaps unsurprisingly – the three aspects of biodiversity studied show different patterns during succession. During early and early-mid successional stages, species richness increases, in contrast with functional or phylogenetic diversity. Functional similarity between species within plots of early stages was higher than expected, a sign of abiotic filtering, which encourages co-occurrence of species with similar traits. During mid-late and late successional stages: Species richness does not increase further, but functional diversity within sites increases significantly, and closely related species are replaced by phylogenetically more distinct species. Between the earlier and late successional stages, functional turnover was higher than the within-stage turnover, suggesting that different environmental filtering processes govern community assembly at different successional stages. Throughout succession, species co-occurring within sites were functionally more similar than expected by chance, indicating that community assembly is deterministic with respect to species traits. However, functional turnover between stages was higher than predicted, and higher than within-stage turnover, indicating that different assembly processes act at different successional stages, possibly caused by differential environmental filtering, during the course of succession.

What do I think? I actually quite liked reading this paper, although I felt at times there were too many results for my brain to cope with. Nevertheless, I found their findings pretty cool. Nothing shook my world, but I like reading articles that convince me that what I imagined was true is actually true (hey, a scientist has so few chances for an ego stroke). One finding that stands out for me is that “phylogenetic turnover did not differ significantly from random expectations, either within or between successional stages, and provided no insights into the temporal dynamics of the processes underlying community assembly”. Phylogenetic diversity is often used as a proxy of functional diversity in this type of studies, often based on the assumption of trait conservatism: that closely related species are ecologically similar. As the authors suggest, the mismatch between functional and phylogenetic diversity could indicate two things. First, that ecological traits might not always be phylogenetically conserved. Second, that phylogenetic diversity represents a more inclusive measure of ecological similarity than measures of functional diversity based on a limited set of traits. I suppose I am slightly biased towards the phylogenetic approach, so it’s useful to know that other measures can provide better resolved results. If anything, I guess this brings to the surface the need for more interdisciplinary and integrative approaches.

What do you think?

Lynsey McInnes

Lynsey McInnes

I enjoyed this paper a lot. I was entertained by the idea of a chronosequence (sampling in the present day but classifying plots based on, in this case, grassland age). Neat! I’m sure this isn’t as good as having data at different timepoints for a single plot, but it surely must help in uncovering some facets of the temporal trajectory of grassland diversity.

The authors throw a whole suite of analyses at their chronosequences, looking at taxonomic, functional and phylogenetic diversity and turnover within and between timepoints for each facet of diversity. They find that each changes through time, but differently, that phylogeny doesn’t add much to the story, but function does, suggesting that there is a deterministic element to community assembly – you either have the right traits at the right time or you don’t. Seems reasonable.

Again, I’m no community assembly pro (though perhaps I should just succumb given how many community-y papers we cover here), but I appreciated the authors thorough approach here and I thought this was a good study on a limited spatial scale whose methods could get rolled out to a more macro scale with appropriate tweaks for less comprehensive trait data availability.

Some thoughts that occurred to me. Perhaps phylogenetic diversity came out as kinda uninformative in this context as the taxonomic scope was too broad…basically around 200 species widely distributed across the plant phylogeny. At this scale, can many traits show conservatism (there must have been rampant trait convergence, esp. if we are only looking at grassland-relevant traits)? Perhaps phylogeny should have been assessed at narrower taxonomic scales, e.g. within grasses, within eudicots, what have you… Then again, the gains made by doing this, might already be covered by the functional approach.

I didn’t look into this in detail, but I wasn’t sure what the authors were after with all the dispersal related traits? Were they included as a measure of species’ ability to get from “elsewhere” (the elusive regional species pool perhaps?) into the plots? So, nothing to do with crossing plots (now I think about it, probably not).

Was there any consideration of the proximity, location, layout of the plots? I guess this concern stems from using a chronosequence rather than an actual timeseries. The plots classified in different successional stages are limited in the diversity (all three kinds) given what came before. The authors might need to correct for within-stage variation in their interpretation of across-stage variation. I think.

But that was a pedantic grumble, I really do like the idea of chronosequences. I wonder in what other contexts this space for time substitution is useful (for instance, its been used a bunch in climate change related studies) and when does it really breakdown?

In conclusion, a surprisingly (for me) thought-provoking study. Thanks Haris!

Will Pearse

Will Pearse

I liked this paper, and it wins the Will Pearse “thank you for thinking about your methods” award because the methods are wonderfully executed, and give the authors exactly the answer they need. I just want to chime in (briefly!) about the same sentence that Haris picked up on:  “phylogenetic turnover… provided no insights into the temporal dynamics of the processes underlying community assembly”. I too am biased towards phylogenetic approaches, and I’ve actually been finding similar patterns (no change in phylogenetic structure over time, no significant phylogenetic structure, etc.) but I interpret them completely differently.

When I find that, despite changes in species composition, or changes in the functional trait structure of a community, that the phylogenetic structure of that community stays the same, it provides really quite profound insights into the community as a whole. Why is it that the trait that, apparently, determine community memberships are seemingly free from phylogenetic constraint? What does that say about how that pool of species, which have been evolving and interacting (potentially) for millions of years? These are the kinds of questions I want to be asking, because I’m really not interested in whether a particular trait shows phylogenetic signal or not, rather I’m interested in what divergence between traits and phylogeny can tell me.

When I find that phylogenetic turnover is minimal, but species (or taxonomic, if you like to call it that) turnover is quite strong, that suggests to me that close-relatives are replacing one-another. Does that mean, rather than being uninformative about ecological interactions, that focusing on species as our unit of measurement, is not giving us the most information about that system? I’ve been examining phylogenetic homogenisation of species recently, and it’s a very different process to species/taxonomic homogenisation – and figuring out why has helped me better-understand my urban ecosystems.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: