Are there general laws in ecology?

John Lawton. Oikos 84(2): 177-192. Are there general laws in ecology?

Dunbar harbour (Will Pearse, 2006)

Dunbar harbour (Will Pearse, 2006)

Will Pearse

Will Pearse

This week Lynsey and I are trying something different, and we’re covering the single most-important paper I’ve read, because it made me study ecology. Hence the picture above these words is one I took a few years beforehand, during an internship where I was doing a lot of marine biology.

While natural history was always interesting to me, ecology seemed very dull. Everything depended on very specific details (“this leaf miner eats this particular plant when there has been that much rain in November”) and with so little generality I never saw the point. When Lawton almost gleefully noted that he “can summarise twenty years of work on bracken at Skipwith in a short paragraph”, it was a breath of fresh air. Apparently there were these people called macroecologists who asked questions so simple I had always naively assumed we already knew the answer. I decided, almost as soon as I finished the paper, that I wanted to be one of those people who asked extremely (hopefully embarassingly) simple questions, and as such started to understand generalities among systems. I saw Lawton’s focus on contingency as almost a call to arms – if things are contingent, figure out why they’re contingent, and how the rules of that contingency operate.

Reading it again now, I notice that, part-way down page 180, Lawton notes that examining “taxonomic relatives” might provide “some progress” in dealing with this overwhelming sense of contingency. I try and work phylogeny into ecology precisely because I feel it is the only way of generalising across, and finding links among, profoundly different systems with no species in common. I am always struck by the work on network motifs, and have a great hope that one day I will uncover some grand unifying repeating motif in the phylogenetic structure of assemblages. Indeed, if community phylogenetics is an attempt to link wider source pool dynamics (how did those species evolve and migrate to form a source pool) into community ecology, then perhaps the field is a response to Lawton’s call for more people to listen to Rickleffs. I’m reading ‘Maximum Entropy and Ecology’ at the moment, and I’m struck by how much of the book seeks to find a mechanistic underpinning for the nine ‘general laws’ that Lawton laws out at the end of the article. Long may the search for an understanding of these laws continue.

Lynsey McInnes

Lynsey McInnes

Will and I decided we were getting a bit lax with our PEGE commentaries so we thought we’d shake things up a bit with some classics. Will wasted no time in choosing this paper by John Lawton, so off we go.I never thought of myself as an ecologist until I joined a department of population geneticists and realised I definitely wasn’t one of them….yet. So I’ve spent the past 6 months honing my ecology skills, hoping to merge them into some population genetics as and when…I enjoyed this paper a lot. I hadn’t read it before, so it was a fun read. I took too major points away from it, one, community ecology really is a bit dodgy and two, why macroecology can be exciting (but doesn’t appear to have moved on much since 1999).

The enthusiasm Lawton has for ecology and ecological patterns and processes is infectious and I felt like I had to say thank you to him for really driving home the message that general patterns are exciting when found, mostly because they are damn hard to find, especially in that mesoscale of community ecology, less so in the fuzzy, broad brush scale of macroecology.

I’ve always been dubious of community ecology, especially of its latest incarnation meshed with phylogenetics. Conclusions always seem to depend on how you circumscribe the question and few communities appear genuinely comparable. So, if patterns are hard to detect, processes are even harder to ascribe to these shadowy patterns.

And then you have macroecology which is all about patterns. Satisfying ones too, most clades follow latitudinal diversity gradients, Bergmann’s rule, Rapoport’s rule, exceptions make sense, its all great. But all these patterns were already known when Lawton wrote his paper, and 14 years later it doesn’t seem like macroecology has moved on much. We are still finding more and more patterns, the hypotheses on the underlying processes are proliferating, the methods to detect these things are getting more and more jazzy, but it doesn’t feel like we know anything much more than we did then. Why is that?

I kind of feel like macroecology, perhaps just like community ecology, doesn’t work because it’s all at one scale. Patterns don’t just come about at a single macro scale. They are built up from individuals, populations, even communities. Lawton seemed really keen on revelling in the really clear patterns he could see at the macro scale, but if we want to get at process maybe we have to delve down into that murky world where general laws and even general patterns most certainly don’t exist. I know for sure this is not a new idea! The Razgour et al. paper we covered a couple of weeks ago is just one example of a paper that brings together all sorts of data from the population to the macro level to understand the fate of a single species. The Ferri et al paper from a few weeks earlier took a comparative look at niche evolution by splitting up multiple species’ ranges into cold and hot extremes. This is definitely the way I think advances can be made, go below the species level to find out stuff emergent above the species level.

I still reckon Lawton’s conclusions will remain unchanged. General laws are damn hard to find, their pursuit can be self-defeating (but also a lot of fun) and they are mega exciting to stumble upon.


Clade-specific consequences of climate change to amphibians in Atlantic Forest protected areas

Loyola et al. Ecography 36: 1-8. DOI:10.1111/j.1600-0587.2013.00396.x. Clade-specific consequences of climate change to amphibians in Atlantic Forest protected areas


Amphibian species recent now (left) and 2080 (right). A message… from the future! From Loyola et al.

Will Pearse

Will Pearse

Lynsey gave me a wide selection of papers this week, and I picked this one because I’m thinking a lot of about the phylogenetic structure of ecological communities and biogeography. The authors predict amphibians’ distirbutions in Brasil in 2080 on the basis of their present distributions, and then examine the implications for protected areas.

I found this paper quite hard to interpret, and I’d appreciate your input! I found figure 2c, which shows phylogenetic diversity in protected areas now and in 2080, hard to read, as did the authors: “phylogenetic diversity increased under future climatic conditions, albeit such increase was not clear (Fig. 2C)”. The magnitude of the change (0.04 is the biggest I could see) seems very small, but it looks to me like those areas with the greatest diversity now have lesser diversity in 2080, and vice-versa, so could there be some kind of interaction going on? I’d love to write about the within-clade results, but I simply don’t understand figure 3 where they’re presented. The authors seem to have made a matrix that represents clade composition in the protected areas, and then plot principal components of that matrix, but figure 3 shows present and future climatic conditions plotted on the same axes as monophyletic clades, and I don’t know how that’s possible. Help!

However, the general approach of examining variation in response among clades is interesting. I think making predictions about protected areas’ future phylogenetic diversity is particularly useful if we want to understand ecosystem function, and I think approaches like these have the potential to be of conservation importance.

Lynsey McInnes

Lynsey McInnes

Ouf, this was a strange paper. I do think the premise was well-intentioned, but the execution was really quite confusing. As I understand it, the authors were interested in seeing how climate change will affect amphibian richness, diversity and phylogenetic diversity in protected areas within the Atlantic Forest of Brazil. A fair enough intention, protected areas are, debatably, the best hope for the persistence of endangered species, and amphibians, more so than other vertebrate groups are undoubtedly endangered. Furthermore, it is of interest whether protected areas protect certain amphibian species better than others, so the phylogenetic perspective could be a valid and important one in order to predict what kinds of amphibians we will be left with in the future.

But this is kind of where I lost the plot. I really struggled with the authors’ approach to defining phylogenetic diversity and similar to Will struggled to interpret figure 3. I also thought they altered between considering the traits of vulnerable species vs. their basal vs. derived status in the phylogeny. Phylogenetic diversity and its maintenance is of interest, but ultimately, I’m more interested in maintaining a diverse amphibian fauna (and I think the authors are too) so they could have devoted more discussion to the traits that help amphibians persist in warmer, drier areas. They highlight that the increase in phylogenetic diversity that they predict should be interpreted with caution given that it comes hand in hand with a decrease in species richness and the paper could have used this interesting result as a springboard into a more philosophical discussion of whether or not this is an acceptable tradeoff.

The authors are also admirably open as to the deficiencies in the methods they employ, namely that their species’ distribution models do not incorporate realistic dispersal parameters or the effects of sustained or disrupted biotic interactions. These problems plague many similar studies and it seems like the field is changing so fast that soon these omissions might make publishingsuch studies harder and harder. On the one hand, this is good, we should as a research community be tough on ourselves, and on the other, is a shame, as this study, even with its flaws and difficulties, can provoke valid discussion on how to go about conservation with limited funds.

One last grumble, I am deeply sceptical on ensemble forecasting. My understanding is, all models are quite uncertain so we should take an average of all such uncertain models and this will give us an average model with much LESS uncertainty? Hm.

%d bloggers like this: