For ESA’s centennial year, they are running a pretty cool series called “The Paper Trail”. A variety of ecologists write about the particular paper or papers that catalyzed their research path. Sometimes the papers are valuable for bringing up particular questions, sometimes they facilitated the connection of particular ideas.
William Reiner provides some insight into the value of this exercise: “What are some of the generalizations one can deduce from this paper trail? For me there are five. First, in ecology one cannot take too large a view of the problem one is addressing. Second, it is useful to step out of one's science into others to gain useful new ways of addressing questions. Collaboration with others outside ones field facilitates this complementarity. Third, teaching provides a useful forum for developing one's ideas. Fourth, there is no literature that is too old to have no value for current issues. And fifth, one must take time to read to be a thoughtful, creative scholar.”
In general, people are writing about papers that either specifically related to their own research at the time and opened their eyes to something new, or else broadly inspired or fascinated them at a critical time. (For Lee Frelich, this was reading The Vegetation of Wisconsin, an Ordination of Plant Communities at 12 years old.) I probably fall into the second group. My undergrad degree was in general biology and math, so although I had taken a couple of ecology courses, I knew essentially nothing about the fundamentals of ecological literature. So I was an impressionable PhD student, and I read a lot of papers. When I started, my plan was to do something related to macroecology, and the first paper I remember being excited about was James H. Brown’s 1984 “On the Relationship between Abundance and Distribution of Species”. It is everything a big idea paper should be – confident, persuasive, suggesting that simple tradeoffs may allow us to predict broad ecological patterns. And while with time I feel that some of the logic in the paper is flawed or at least unsupported, it definitely is a reminder of how exciting thinking big can be (and 1870 citations suggests others agree).
The next paper was R.H. Whittaker’s “Gradient analysis of vegetation” (1967). There is a lot of recommend in Whittaker’s work, in particular the fact that it straddles so well modern ecology and traditional ecology. He introduces early multidimensional analyses of plant ecology and asks what an ecological community is, while also having such a clear passion for natural history.
Finally, and perhaps not surprisingly, the biggest influence was probably Chesson (2000) “Mechanisms of the maintenance of species diversity”. The value of the ideas in this paper is that they can (and have) be applied to many modern ecological questions. In many ways, this felt like the most important advance in ecological theory in some time. It is also the sort of paper that you can read many times (and probably have to) and still something new every time.
Of course, there are many other papers that could be on this list, and I’ve probably overlooked something. Also, makes me miss having free time to read lots of papers :)
Tuesday, July 15, 2014
Monday, July 7, 2014
*Sorry for the low frequency of posts these days – I seem to be insanely busy this summer ☺
Ecology is hard in part because of the things we can’t (at least easily) measure: fitness, interaction strengths, and the niche, all fundamental ecological concepts. Since we are unable to measure these concepts directly, ecologists have come up with proxies and correlates. Take Darwin’s hypothesis that competition should be greater between closely related species. It relies a chain of assumptions about proxy relationships – first that relatedness should correlate with greater similarity of traits, secondly that similar traits should correlate with greater niche overlap. The true concept of interest, the niche, is un-measurable (if it is an n-dimensional hypervolume) so instead shared evolutionary history provides possible insight into species coexistence.
Ecophylogenetic studies have adopted Darwin's hypothesis as an example of how molecular phylogenies may provide information about evolutionary history which in turn informs current ecological interactions. Phylogenies ideally capture feature diversity, and so (all things being equal) should provide information about similarity between species based on their relationship. Despite this, studies have been mixed in terms of finding the relationship predicted by Darwin between phylogenetic relatedness and competition. It is not clear whether this mixed result suggests problems with the phylogenetic approaches being used, or non-generality of Darwin’s hypothesis.
Oscar Godoy, Nathan Kraft, and Jonathan Levine attempt to explore this question once again, but through the lens of Chesson’s coexistence framework (2000). Chesson’s framework describes competitive differences between species not as a single quantity, but instead the outcome of both stabilizing niche differences and equalizing fitness differences between species. This framework predicts that competitive differences should be greatest when species have similar niches (low stabilizing niche differences) and/or when they have large differences in fitness. This divisions alters the predictions from Darwin's hypothesis: if closely related species have similar niches, they should compete more strongly, but on the other hand, if closely related species have similar fitnesses, they should compete less strongly. Darwin’s hypothesis as it has been tested may be too simplistic.
The authors used an experiment involving 18 California grassland species to look at first, whether competitive ability is conserved, and more generally to explore whether phylogenetic distance predicts “the niche differences that stabilize coexistence and the fitness differences that drive competitive exclusion?” Further, can this information be used to predict the relationship between phylogeny and competitive outcomes? To determine this, they quantified germination, fecundity, seed survival, and interaction coefficients for the 18 species based on competition with different competitors (both by identity and density), and quantified the strength of stabilizing and equalizing forces (as in previous works). With this information, they calculated for each species the average fitness and ranked species in a competitive hierarchy using a fully parameterized annual plant population model. Species’ competitive rank did in fact show a phylogenetic signal (Figure 1), and the strongest competitors were clustered in the Asteraceae and its sister node.
|Fig 1. Relationship between competitive rank among the 18 CA grassland species.|
|Fig. 2. Relationships between fitness differences and phylogenetic distance.|
|Fig 3. Solid line - observed niche distances as a function of phylogenetic distance. Dashed line, size of distances actually needed to assure coexistence.|
How should we interpret these results more broadly? Is this reinforcement of the use of phylogenetic information to answer ecological questions, provided the questions are asked correctly? One of the most interesting contributions of this paper is their discussion of the oft-seen, but poorly incorporated, increase in variation in a trait (here fitness differences) as phylogenetic distances increase. This uneven variance often leads to phylogenetic-trait correlations being labelled non-significant, since it violates the assumptions of linear models. In contrast, here the authors suggest that this uneven variance is important. “For example, even if on average, both niche and fitness differences increase with phylogenetic distance, the increasing variance in these relationships means that only distant relatives are likely combine large competitive asymmetries with small niche differences (rapid competitive exclusion), or large niche differences with small competitive asymmetries (highly stable coexistence). Overall, our results suggest that increasing variance in niche or fitness differences with phylogenetic distance may play a central role in determining the phylogenetic relatedness of coexisting species.”
This discussion is important for questions about phylogenetic relatedness and coexistence – variability is part of the answer, not evidence against the existence of such relationships. However, a few caveats seem important: Because fitness differences and niche differences as defined in the Chesson framework may not be easily associated with traits (since a single trait might contribute to both components), it seems that it will be a little difficult to expand these analyses to less rigourous experimental settings. This might also be important to hypothesize how fitness or niche differences per se become associated with phylogenetic differences, since traits/genes are actually under selection. But the paper definitely provides an interesting direction forward.
Chesson, P. 2000. Mechanisms of maintenance of species diversity. Annual Review of Ecology and Systematics 31:343-366.