Research

 

I like to think that the strengths of the work we do in our lab is the integration of theory, experimental and observational approaches to some of the major questions at the interface of community and ecosystems ecology. Our work usually incorporates a strong food-web approach involving interactions among animals and plants (and sometimes even decomposers!).

 

Integration of 3 Approaches:

 

The goal is to

test (using experiments),

logical (using theory), and

relevant (based on observations)

hypotheses.

 

This means looking

at the interactions among

approaches (blue arrows)

as much as perfecting any

given approach (red arrows)

 

The field of ecology, and especially the fields of community and ecosystems ecology, are in a state of tremendous and exciting change. In part this is due to improved methods and theories that are revitalizing the field from an academic perspective. But this is also largely driven by changes in the motivations and goals of scientists who are increasingly interested in questions that have implications (or even applications) in solving pressing environmental issues. Because of this, it is important to do work that will help develop conceptual insights that will solidify the field of community ecology into a more quantitative and predictive science.

(For more on this perspective, read this paper by Jane Lubchenco: http://sciencemag.org/cgi/content/full/279/5350/491
 

We have worked primarily in freshwater aquatic systems such as lakes and ponds because of their convenience as model ecosystems. However, we hope our findings are more general and we look for ways to see if they are. Below are some of the recent projects directed by Mathew Leibold but you might also check the web pages of others in the lab and of my past collaborators to see what else is going on (assuming they get their web pages updated!).

 

Mathew Leibold’s Research:

 

I have been involved in a number of different lines of research since my arrival at the University of Chicago. All are oriented at understanding how the outcomes of species interactions, involving especially predation and interspecific competition, depend on the environmental context in which they occur. My approach in general, is to use experimental and observational methods in the field to test and refine theories about underlying causal mechanisms. Almost all my work is done on aquatic (especially planktonic) organisms because they are very convenient "model systems"; however because most of my studies are closely linked to general theoretical models they can provide valuable insights into the operation of other, less tractable, systems.

 

Trophic and food-web interactions in ponds and lakes: One focus of my work has been to understand how the structure of communities affects the responses of entire trophic levels to environmental change. To this end, I have and tested some simple models of trophic interactions in food webs to explore how competition among plants is affected by herbivores that differ in their ability to feed on herbivore-resistant plants. My experiments have demonstrated that food-web structure (as described by the occurrence of specific taxa) can be qualitatively important in regulating responses of organisms to environmental change. For example, whether or not tadpoles are present in a pond can determine whether enhanced nutrient levels will favor plants or herbivores. Theory on indirect interactions of species in complex food webs predicts this kind of context dependence, however the challenge is to identify the particular mechanisms that are responsible for such effects.

 

This graph shows the results of an experiment in which I manipulated

nutrients (high vs low) in four different food webs (depending on the

presence/absence of Daphnia and Rana grazers). The arrows point from

the low nutrient to the high nutrient conditions. They show that when

Rana was present plant biomass responded very positively to nutrients

whereas herbivores responded weakly. In contrast when Rana was

absent, plant biomass decreased and herbivore biomass increased very

strongly. The effects of Daphnia were weaker but also significant.

See Leibold and Wilbur, 1992 for more details….

 

 

 

 

This type of result is important because it violates the common assumption that ecosystem phenomena (e.g. biomass accrual of plants and herbivores) are independent of community structure (the composition and diversity of taxa). These results are also important because they were only poorly predicted by previous models, including my own. I am currently working to refine these models, but find that explaining the results of my experiments requires models that go beyond the more conventional (i.e., niche theoretic) forms of community ecology, to include to process that better account for the ways species from a broader regional biota sort into local communities.

 

Here is a graphical model of “keystone predation”

using ZNIGs (“zero net growth isoclines”) that includes

species sorting to produce joint biomass responses among

organisms at adjacent trophic levels to enhanced potential

productivity. See this paper for details that will make

more sense (hopefully!) of this: Leibold, MA. A graphical

model of keystone predators in food webs: trophic regulation

of abundance, incidence and diversity patterns in communities.

Am. Nat. 147:784-812.

 

Role of “species sorting” in metacommunities along environmental gradients: The model described above assumes that there is a process of extinctions (due to local interactions) and invasions (facilitated by these interactions but dependent on colonization from nearby sites), so that different species are found at different sites along gradients in productivity. I have found that this sorting process is critical in explaining why both plant and herbivore biomass increase jointly with productivity. In a meta-analysis of previous experiments (in which there were no/few opportunities for sorting), the addition of nutrients led to either strong plant or strong herbivore responses but not both (see Leibold et al. 1997):

 

Here is the most important result from the paper with my students (Leibold et al. 1997). We surveyed papers in which nutrients were added to lake communities either in replicated bags (solid symbols) or in whole-lake manipulations (X’s) where scientists measured both zooplankton (herbivores) and phytoplankton (plants) biomass. This graph shows the proportional plant response to nutrient manipulations (difference in biomass between low and high nutrient conditions divided by the low nutrient conditions) plotted against the proportional herbivore response. In natural systems, correlational studies would lead us to expect these values to scale so that the herbivore response is about .4 times the plant response (inside the “wedge” in the graph). We found that none of these studies gave this result because about half had herbivore responses that were much stronger than the plants (points above the wedge) and half had plant responses that were much stronger than the herbivores (below the wedge). We hypothesized that most of these experiments just didn’t last long enough to allow for species sorting in which invasions by new species might have altered the response.

 

However in an experiment in which I manipulated the opportunities for sorting (comparing high sorting with low), I found that plant and herbivore biomass jointly responded to nutrients only when sorting was high and that the response was unpredictable when sorting was low.

 

Each symbol here shows the results from a single artificial pond (created in cattle watering troughs). I manipulated the species pool by adding plankton (algae and zoops) from single ponds in Michigan (red dots) or by pooling from at least 6 ponds with different nutrient levels (blue stars). I also subjected each of these to two different nutrient levels,“low” (near the origin) and “high” (away from the origin). These results show no consistent effect of nutrients when the species pool was local; sometime the algae responded strongly and sometimes the zoops did. But they do show that both responded strongly, jointly and consistently when there was a big species pool that allowed for sorting (Leibold and Smith in prep).

 

I think that many responses of communities to environmental change (including factors besides productivity and in ecosystems besides ponds) will turn out to be much more predictable and stable when there is sorting than when there is not. The implications are important because they imply that biodiversity can be important in buffering ecosystems to environmental change at a regional level (where it regulates how many species are “in the wings” available for sorting) as well as at a local level (i.e.“on stage”).

 

Ecological genetics of competing species: In my dissertation work I documented how two species of Daphnia zooplankton segregate by habitat in stratified lakes. In subsequent collaboration with Alan Tessier (Mich. St. Univ.), I found a substantial amount of genetic divergence among different lake populations of both species in life-history traits, behavior, and in ecological traits such as the ability to exploit resources and avoid predators. There is a substantial amount of natural selection on such traits in natural populations that mimics, in direction and magnitude, the selection occurring in our experimental enclosures. These data indicate that the basic tradeoffs associated with ecological traits, are different in lake populations that have diverged (via natural selection in lakes with different levels of predation). Our experiments show that these genetic effects can substantially modify the strength of interspecific competition, and therefore characteristics of species coexistence, as well as modify the role that the Daphnia play in indirectly mediating the effect of predacious fish on the Daphnia's algal resources. Such closely linked studies of ecological and evolutionary phenomena are important because they illustrate how the evolutionary context of species interactions can alter how species participate in the broader community.

 

Species diversity and ecosystem productivity: I have also become interested in how mechanistic approaches to community ecology can be used to understand patterns involving entire communities of interacting organisms. Most such models have focused on a small number of species (usually between 2 and 6) and I have worked to extend the results of these models to address issues of biodiversity of entire trophic levels. I have particularly focused on how ecosystem productivity can regulate the overall diversity of coexisting species of plants and herbivores in a local assemblage. Until recently, it was thought that the relationship between species diversity and productivity was monotonically increasing, but there is mounting evidence that it is often a unimodal curve. Although there are several published models that make this prediction some are inconsistent with other common observations on biomass accrual and nutrient availabilities. I am particularly interested in determining if unimodal diversity-productivity relations might result from an interaction between the effects of interspecific competition for resources occurring within trophic levels, and the regulation of such interactions by predators between trophic levels. I have found that models based solely on resource competition are not likely explanations for observed positive inter-correlations among nutrient availabilities and overall nutrient levels in lakes. I have also documented that patterns of variation in species diversity in pond algae are closely linked to patterns of variation in density, composition, and diversity of zooplankton grazers.

Here is a landsat pic of one of the areas we study. This area is called the Lux-Arbor Research Reserve at the Kellogg Biological Station. Vegetation shows up red on here whereas ponds and lakes are dark. We mostly sample the smaller ponds you can in the central peninsula as well as ponds in other sites in southern and central Michigan. These ponds are fishless and have a biota that is relatively distinct from the other nearby fish-containing ponds. Together they form what might be called a “meta-community”.

 

 

 

Here is some data from a survey I did in natural ponds in Michigan (Leibold 1999) including some of the ponds shown in the figure above. These data show that both plants and herbivores show unimodal (“humped” ) relations between diversity and potential ecosystem productivity (here using total nitrogen as a proxy). However there is a lot of variability because this increase in diversity does not always occur at intermediate productivity levels. Right now, I’m trying to understand why this might be so….

Further, in microcosm experiments, variation in diversity and composition of algae in response to manipulations of nutrients only occurred in the presence of herbivores and did not occur when herbivores were absent. Finally, preliminary data indicate that such patterns involving plants and herbivores are also linked to patterns in functional diversity (the ability to use different substrates) of microbial decomposers. On a broad scale these experiments are most consistent with a model based on the “keystone-predator effect” but synthesizing other known effects of resource heterogeneity and disturbance are still important outstanding questions. My comparative work has further identified some important analytical-descriptive problems in comparing community composition. In particular, I am determining how to quantify and discriminate among major patterns of geographic distributions of species such as nested subsets, checkerboard distributions, and gradient replacements.

 

Mechanistic approaches to niche relations: The insights arising from my research on trophic interactions and, especially, from my recent work on biodiversity has led me to a different perspective on community ecology than that resulting from previous theory (as embodied in "niche-theoretic" models of species interactions, see Leibold 1995). I have, for example, used mechanistic models like those described above to show that “community-wide character displacement” is unlikely in situations where both regional and local communities are near equilibrium (Leibold 1998). To the extent that these different views are supported by empirical observations they may imply a substantial change in how we view variation in the outcome of species interactions in ecological communities with respect to biodiversity, patterns of similarity among coexisting species, and effects of biodiversity on ecosystem responses to environmental change. My ultimate goal is to continue to work toward developing these ideas into a more coherent theory of community ecology. I am especially excited about the work on biodiversity because environmental problems associated with biodiversity and the integrity of ecosystems are an increasingly large societal concern. In this area in particular, progress on understanding community ecology is vital from an applied perspective as well as for purely academic reasons.

 

Future work:

My current work (funded by NSF) takes a new approach towards testing community models. I will be using new statistical methods (Legendre et al. 1998, see this web page for more on this method: http://www.fas.umontreal.ca/biol/casgrain/en/labo/4th_corner.html), to document how trade-offs in species traits are expressed across environmental gradients and how the distribution of these traits is correlated with ecosystem properties.

 

The idea that trade-offs among species are a crucial component for coexistence is an old one but it has previously been applied only (?) at the local level. Here, my collaborators (V. Smith at Kansas and S. Pinca, a post-doc in my lab) and I will use the idea to examine the role of trade-offs in regulating how species coexist in larger regional-scaled communities and how this alters the roles that species play in local communities within a larger region. We will also begin working on theoretical aspects of this question which is, to date, very poorly developed compared to the role of trade-offs at the local scale.

 

There usually are numerous competing hypotheses about the distribution of species along gradients. One of the most important assumptions in models of these hypotheses is about the distribution of traits among species and the expression of correlations between these traits’ benefits and costs along the gradients. We will conduct short-term “selection” experiments to document trade-offs among ecological traits such as minimum resource requirements, tolerance of predators, and maximum growth rates as well as measure other traits such as stoichiometry and morphology for a large and representative array of plant and herbivorous planktonic species. We will evaluate how these traits are correlated with environmental gradients, especially related to nutrients but also other water quality measures and the distribution of predators, in natural ponds to establish what the pattern is in unmanipulated systems. Finally, we will examine how this pattern might be determined by conducting experiments in mesocosms that selectively manipulate potential factors (including different nutrients, light, and predators) one at a time, to see which manipulations are crucial in generating the pattern we see in nature.

 

If successful, this approach could be important in refining our ability to understand distributional patterns in less tractable systems, especially those involving sensitive areas where experiments may not be practical or in patterns such as paleontogical ones where experiments are out of the question.