Paul Keddy. Competition -- an introduction and annotated bibliography for plant communities.

Paul Keddy Signature
Biography Research Papers Conservation Books Commentary Teaching Research in Progress Contact



a guide and annotated reading list


Paul A. Keddy




General Overviews



Along Natural Gradients

Competitive Hierarchies and Asymmetric Competition

Roots and Shoots


Suggested citation: Keddy, Paul A. 2012. Competition in Plant Communities. Viewed online at, date. Adapted from: Oxford Bibliographies Online: Ecology. Ed. David Gibson. New York: Oxford University Press.  


Competition is generally understood to mean the negative effects caused by the presence of neighbors, usually by reducing the availability of resources. Competition is one of the most important factors controlling plant communities, along with resources, disturbance, grazing, and mutualism. Since all plants require a few basic elements, the resource is generally light, water, nitrogen, or phosphorus, depending upon the species and the location. The effects of competition are widespread, and easily observed in mixtures of crops and managed forests.  That is why weeding and thinning are practiced.  Competition is also widespread in native habitats, from deserts to wetlands, and is known to have important—indeed crucial—effects upon recruitment, growth, and reproduction. In the late 1800s Darwin wrote extensively about the importance of competition in nature, particularly its role in driving natural selection. Thereafter, interest in the phenomenon grew. There were many experiments with crops and with wild species, most now overlooked.  Models of competitive interactions were also constructed, with the number and size of the models increasing rapidly with the advent of computers in the 1970s.  Hence, simple, early models are now often overlooked.  Because the word “competition” has a common usage in English, it is frequently taken for granted and therefore misunderstood. Care must be taken in using or interpreting the word “competition” without specifying what kind of competition is being investigated. For example, competition may be looked at from the perspective of an individual, a population, or a species, and it may be asymmetric or involve multiple species simultaneously. Experimental design carries its own assumptions; often these are not stated in published articles.  One of the most difficult tasks in reading the literature is sifting through large numbers of experiments where investigators have haphazardly selected (a pair of) species and grown them in mixture, without adequately justifying their choice of species and study design.  Another difficult task is distinguishing between models that, at least in principle, have measurable inputs and/or make measurable predictions, and those that do not and cannot be tested.  Overall, the very ease of growing plants in mixture, and the ease of making models, has made people careless, with the result that basic questions are still being overlooked.  Ongoing issues of importance include mechanisms of competition, types of competition, and the intensity of competition under different conditions.



Darwin’s Origin of Species contains a good deal about competition, usually competition between species operating as the force of natural selection. Of course, Darwin was greatly influenced by the English economist, Malthus, who wrote about resources and population growth, including a famous essay on the Principle of Population The first major work of the 20th century is Weaver and Clements 1938.  This is a frequently-overlooked volume with a wealth of competition experiments.  Perhaps the book is overlooked because of its extensive discussion of succession, and the many new terms introduced under this topic—try reading the book as a treatise on competition and skip the other parts.  Although plant ecologists continued with studies of competition, by the 1970s work by zoologists was ascendant, as illustrated by MacArthur 1972. Models such as the Lotka-Volterra model had become standard tools. The book also illustrates the concern with how species might escape competition by using different resources (“resource partitioning”), although there is still disagreement about how applicable this concept is to plants, which share a common set of resources.  Harper’s 1977 book can be considered very influential for re-focusing attention upon plant populations and plant life cycles. It summarizes a vast number of studies on plant populations, including studies in agriculture and forestry.  The emphasis on populations, and agricultural systems, was challenged in the 1970s by Grime 1979, which emphasized that habitats lacking in natural resources (stressed habitats) are very different from the relatively fertile sites favored by agricultural researchers. It introduced the CSR model, in which plants respond to two basic gradients, stress and disturbance.  It is a landmark, shifting attention away from populations and back to plant traits and environmental gradients. By 1990 the situation had become confused, as illustrated by the many views and ideas and data sets in Grace and Tilman 1990.  As one referee noted, the book shows primarily how little agreement there was about what the word competition meant, how it should be measured, and how even common experimental designs should be interpreted.  Keddy 2001 (originally published in 1989) emphasized that some of the confusion was the result of there being many different components of competition, intraspecific/interspecific, symmetrical/asymmetrical, and diffuse/monopolistic competition, to name just three. This book explicitly addresses some of the work done since the era of Clements, as well as the advances, such as the different types of competition and the different types of experiments necessary to measure each type; the 2001 second edition includes additional and more recent examples compared to the original 1989 version, while a chapter in Keddy 2007 provides a shorter introduction to the field.  

Grace, James B., and David Tilman, eds. 1990. Perspectives on Plant Competition. San Diego, CA: Academic Press.

It is instructive to read the book for this historical snapshot. Weaver and Clements, for example, are largely forgotten.  Grime is beginning to challenge the status quo—but does not appear as a contributor. It is also useful to compare and contrast the definitions of competition and the sources of evidence used in each chapter.

Grime, J. Philip. 1979. Plant Strategies and Vegetation Processes. Chichester: John Wiley.

This book played an important role by challenging ecologists to rethink their understanding of competition.  It emphasizes an active role for competition, by suggesting that plants must invest resources and forage to obtain resources.  It describes how other (environmental) factors like stress and disturbance interact with competition. It stimulated decades of research and debate.

Harper, John. L. 1977. Population Biology of Plants. London: Academic Press.

The key chapters are 6 to 11 and these are must reading for someone planning to study plant populations and competition.  However, most other chapters are relevant, since competition may occur at different stages in life history.

Keddy, Paul. A. 2001. Competition. Dordrecht: Kluwer.

Chapters 1 (read here)  and 2 introduce kinds of competition as measured by different kinds of experiments.  Originally published in 1989 by Chapman and Hall, many more examples were added to this second edition.  Chapter 9 (read here) in the second edition introduces many models, including models by Skellam and Pielou, for competition among patches, and along gradients, respectively.

Keddy, Paul A. 2007. Plants and Vegetation. Cambridge: Cambridge University Press. 

Chapter 6 provides a contemporary overview of plant competition, including newer examples, set within the context of other forces in plant communities such as stress, herbivory, and mutualism. Link here.

MacArthur, Robert. H. 1972. Geographical Ecology. New York: Harper and Row.

This book begins with general issues affecting plant and animal distributions. It has a lucid description of the two-species Lotka-Volterra model.

Weaver, John E., and Frederick E. Clements. 1938. Plant Ecology. 2nd ed. New York: McGraw-Hill Book Company.

A classic that all plant ecologists should own, and read, before planning their own work. Inexpensive copies of this book can still be found second hand. For the study of competition, it is best to ignore the other main theme of this book, succession, which is an entire different topic. This book also introduces the use of phytometers, that is, using an easily-grown species to compare habitats from a plant’s perspective.



Because the word “competition” has a common usage, it is tempting to assume one knows what it means.  In fact, there are many kinds of field and lab experiments that can be used to measure competition, each of which measures a slightly different aspect of competitionGoing back at least to Darwin, it is thought that that competition is most intense between similar species, raising an issue refereed to as limiting similarity.  But in plant communities, species of very different sizes and life history compete, so the importance of limiting similarity remains unclear.  Moreover, most plants grow surrounded by a large number of species. The tendency to do a pair-wise competition study of two haphazardly selected plant species is one of the greater obstacles to progress. It is necessary to carefully think about the distinction between competition effect and competitive response, between asymmetric and symmetric outcomes, and between monopolistic and diffuse interactions.  Each of these distinctions will affect the type of question asked and the type of experiment used. A good example is the classic de Wit replacement series, (Harper 1977), which assumes— indeed requires—that plants be of equal size to avoid changing total biomass among the different mixtures of the two species.  Of course, if the plants are nearly identical, it is impossible to study the majority of plant interactions, which are among species of very different sizes. Harper and McNaughton 1962 is an example of the De Wit replacement series in grassland species, Goldsmith 1978 shows a multispecies application to sea cliff plants, while Wilson and Keddy 1986a use the design to construct a multi-species competitive hierarchy.  Fonteyn and Mahall 1981 illustrate a two-species removal experiment, while Silander and Antonovics 1982 show how to extend this to many more species. Goldberg 1990 raises the distinction between competitive response and competitive effects; Keddy et al. 1998 illustrates how to assess competitive response across many species. The total effects of all neighbors (which has been called competition intensity and diffuse competition) can be assessed using removal plots with selected species introduced to clearings and intact vegetation (Wilson and Keddy 1986b).  This list of types of competition is not exhaustive (for a short introduction to even more possibilities, see Chapter 1 in Keddy 2001 (read here), particularly Figure 1.5. More types of competition are discussed in the sections on Competitive Hierarchies and Asymmetric Competition and Roots and Shoots.


Fonteyn, Paul J., and Bruce E. Mahall. 1981. An experimental analysis of structure in a desert plant community. Journal of Ecology 69: 883-896.

A removal experiment, as opposed to a pot experiment, used to measure intra- and interspecific competition in the field.  Some further interpretations of this experiment and found in Keddy 2001 (p. 21-23), which is cited under General Overviews.

Goldberg, Deborah E. 1990. “Components of resource competition in plant communities.” In Perspectives on Plant Competition. Edited by James B. Grace and David Tilman, 27-49. San Diego, CA: Academic Press.

Draws the useful distinction between “effect” and “response” in competition experiments.  These terms are now well-established in the field.

Goldsmith, F. Barrie. 1978. Interaction (competition) studies as a step towards the synthesis of sea-cliff vegetation. Journal of Ecology 66: 921-931.

Competition between a pair of species with different morphologies, and along an experimentally imposed gradient of salinity.

Harper, John L., and I. H. McNaughton. 1962. The comparative biology of closely related species living in the same area. VII. Interference between individuals in pure and mixed populations of Papaver species. New Phytologist 61: 175-188.

Competition studied among very similar species in old fields, using the classic de Wit design, and illustrating as well the problem of drawing general conclusions.

Keddy, Paul A., Lauchlan. H. Fraser, and Irene C. Wisheu. 1998. A comparative approach to examine competitive response of 48 wetland plant species. Journal of Vegetation Science 9: 777-786.

Exposing very different types of plants to experimentally created canopies allows one to measure competitive response as opposed to effects in a large set of plant species.

Silander, John A., and Janis Antonovics. 1982. Analysis of interspecific interactions in a coastal plant community—a perturbation approach. Nature 298: 557–560.

A removal experiment using a larger number of species.  Unfortunately, this paper also shows that the larger the number of species, the less likely one is to find significant effects.  This is a good reminder that large sample sizes are needed for perturbation studies in the field.

Wilson, Scott D., and Paul A. Keddy. 1986a. Species competitive ability and position along a natural stress/disturbance gradient.  Ecology 67:1236-1242.

Illustrates how adding a larger number of species to a de Wit style design allows one to experimentally create a competitive hierarchy for a set of species.

Wilson Scott D., and Paul A. Keddy. 1986b. Measuring diffuse competition along an environmental gradient:  Results from a shoreline plant community.  The American Naturalist 127:  862?869.

One of the first attempts to measure diffuse competition in the field, and along a natural gradient.  This design might now be thought of as measuring monopolistic rather than diffuse competition, but the difference is not widely appreciated or written about (see Keddy 2001 under General Overviews).



Are there traits that consistently are associated with higher competitive ability in plants? And if so, what are they?  We can use the word “trait” in the broadest possible sense, from morphological and physiological traits to life history traits. There is a very long list of candidate traits (Grime 1977, Weiher et al. 1999, Westoby and Wright 2006), and some discretion is needed in selecting them. There are vastly more studies using morphological traits (e.g. height) than using physiological traits (e.g. relative growth rate), not because the former are more important, but because they are easier to measure. There is also a good deal of confusion about the fundamental distinction between functional traits and phylogenetic traits—much of the current work on phylogeny, while important for understanding plant evolution, does not contribute to the understanding of plant functional traits that affect competition (Keddy 1990). In general, height seems to be the most important trait, since taller plants can intercept light for shorter plants—and indeed, woody plants have evolved multiple times in response to this pressure to grow higher (see Keddy 2001, cited under General Overviews). Lycopods, gymnosperms, as well as both major groups of angiosperms have woody clades. The other traits are less obvious, and it is not clear if they vary with habitat.  Below-ground competition, and the traits associated with it, are particularly poorly-understood, except for the general principle that plants best able to lower resources to neighbors are likely to be superior in the long run.  Hence, in reading the literature, there are two issues to keep in mind.  First, at the larger scale, height is likely of over-riding importance, reflecting the general observation that most plants shade one another.  Indeed, back in 1933, Clements 1933 summarized the many field experiments completed by this time (most of which are now forgotten), noting that taller plants have “a decisive advantage” over shorter ones. There are also more recent studies that specifically examine height as a predictor of competitive ability (Givnish 1982, Gaudet and Keddy 1988).  One general trait which is difficult to measure, but of possible importance, is the ability to lower the resource supply to neighbors (Tilman 1982).  Below-ground interactions are demonstrably important in some cases, but the traits that provide this competitive success are difficult to measure, except perhaps efficiency of nutrient uptake (see also Roots and Shoots). There are many other possible traits that may become important under certain conditions, including litter accumulation, shoot thrust, rhizome aerenchyma, foraging ability, and even mycorrhizae. Large data sets of plant traits provide an emerging tool for linking plant traits to ecological interactions and ecosystem processes (Westoby and Wright 2006).


Clements, Frederick E. 1933. Competition in plant societies. News Service Bulletin, Carnegie Institution of Washington, 2 April.

Summarizes the many field experiments completed by this time (most of which are now forgotten), noting that taller plants have “a decisive advantage” over shorter ones.

Gaudet, Connie L., and Paul A. Keddy.  1988.  Predicting competitive ability from plant traits:  a comparative approach. Nature 334: 242-243.

A comparison of competitive effect among 44 different plant species shows that height and above ground biomass are among the best predictors of competitive effects. Read here.

Givnish, Thomas J. 1982. On the adaptive significance of leaf height in forest herbs. The American Naturalist 120: 353–81. 

A model showing how taller plants can exclude shorter plants, thereby explaining field distribution patterns. The data are from forest herbs, but the general principle may extend to many other habitats.

Grime, J. Philip. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. The American Naturalist 111: 1169–1194.

One important view of the main traits associated with competitive performance in plants.

Keddy, Paul A.  1990.  “The use of functional as opposed to phylogenetic systematics:  a first step in predictive community ecology.” In Biological Approaches and Evolutionary Trends in Plants. Edited by Shoichi Kawano, p. 387-406. London: Academic Press.

There is a fundamental distinction between functional traits and phylogenetic traits, and the two are often confused.  Functional traits are generally those of greatest interest in the study of competition, and the construction of predictive models in ecology. Read here.

Tilman, David. 1982. Resource Competition and Community Structure. Princeton, NJ: Princeton University Press.

The impact of plants upon neighbors can be summarized primarily as the ability to reduce resources to those neighbors.  Introduces the concept of resource isoclines as a tool for investigating competition in multispecies communities, as well as above and below ground effects.

Weiher, Evan, Adrie van der Werf, Ken Thompson, Michael Roderick, Eric Garnier, and Ove Eriksson. 1999. Challenging Theophrastus: a common core list of plant traits for functional ecology. Journal of Vegetation Science 10: 609-620.

An overview of the role of traits, including traits that affect competitive ability, and a good starting point for thinking about the issue.

Westoby, Mark and Ian J. Wright. 2006. Land-plant ecology on the basis of functional traits. Trends in Ecology and Evolution 21: 261-268.

It is pointless to talk about ecology, competition, or plant communities, without reference to traits, since all plants possess them. A tree is not the same as a cactus, and neither of them looks like a water-lily. Growing data bases of traits provide world-wide relationships in traits such as leaf characteristics.  Linking such traits to plant competition is more conceptually difficult, but requires information on traits as its foundation.



As Chris Pielou once observed in her characteristic acerbic tone, ecologists spend too much time looking for mythical homogeneous environments.  The alternative, as she vividly illustrated with her zonation studies (e.g., Pielou 1975), was to accept that natural gradients exist and use them as research tools. Natural environmental gradients provide many advantages – not unlike a prism, they make patterns in plant communities obvious. Gradients therefore allow for quantitative tests for pattern against null models.  They allow changes in environment factors to be quantified, making it possible to construct testable hypotheses as to causation and mechanism.  They provide the opportunity for removal experiments, Here is a set of studies that illustrate how you can take a gradient and use it to your advantage to study competition.  A general overview to the opportunities provided by gradients as a research tool, which could be treated as an amplification of Pielou’s observation, is found in Keddy 1991. Possibility the earliest experiment on competition in native species (Tansley 1917); although it uses only two environments, it clearly addresses a natural pH gradient in the English environment. More recent experimental work taking advantage of natural gradients includes studies of competition using water depth (Grace and Wetzel 1981), elevation (Wilson 1993), salinity (Pennings and Callaway 1992), stress/disturbance (Wilson and Keddy 1986) and nutrients (Gaudet and Keddy 1995). At a larger scale of time and space, gradients of nutrients and productivity also likely are associated with changes in herbivory driven in part by changes in predation rates (Oksanen 1990), suggesting additional factors that might be manipulated in large field experiments. As humans continue to modify natural gradients of moisture, nutrients and even temperature, studies of existing natural gradients will merge naturally with studies of human induced changes in competitive interactions (Niu and Wan 2008) .


Gaudet, Connie L., and Paul A. Keddy. 1995. Competitive performance and species distribution in shoreline plant communities: a comparative approach. Ecology 76: 280-291.

Relative competitive performance measured in pots allows one to predict the field distribution of plants along natural gradients.

Grace, James B., and Robert G. Wetzel. 1981.  Habitat partitioning and competitive displacement in cattails (Typha): Experimental field studies.  The American Naturalist 118: 463-474.

Competition between a pair of similar species along an experimental gradient of water depth shows that one species displaces the other into deeper water. An extension of the Tansley paper to a full natural gradient, and a similar design to Goldsmith 1978 (cited under Types) with a different gradient.

Keddy, Paul A.  1991.  “Working with heterogeneity:  an operator's guide to environmental gradients.” In Ecological Heterogeneity. Edited by Jurek Kolasa and Steward T.A. Pickett, p. 181-201. New York: Springer Verlag.

            An overview of the many opportunities offered by natural environmental gradients. Read here.

Oksanen, Lauri. 1990. Predation, herbivory, and plant strategies along gradients of primary production. In Perspectives on Plant Competition. Edited by James B. Grace and David Tilman, p. 445-474.  San Diego, CA: Academic Press.

The presence of absence of predators upon herbivores may determine the intensity of competition among plants. This paper, and other related ones, is a reminder of the big picture, that plant-plant interactions take place in a larger context involving herbivores, gradients and the predators upon those herbivores.

Niu, Shuli and Shiqiang Wan. 2008. Warming changes plant competitive hierarchy in a temperate steppe in northern China. Journal of Plant Ecology 1: 103-110.

Fifteen two species mixtures were grown under ambient and elevated temperatures. The resulting competitive hierarchy changed with experimental warming

Pennings, Steven C., and Ragan M. Callaway. 1992. Salt marsh plant zonation: the relative importance of competition and physical factors. Ecology 73: 681-690.

The effects of competition depend upon the presence of other factors, in this case salinity.

Tansley, Arthur. G. 1917. On competition between Galium saxatile L. (G. hercynicum Weig.) and Galium sylvestre Poll. (G. asperum Schreb.) on different types of soil.  Journal of Ecology 5: 173-179.

One of the earliest plant competition experiments, showing a simple design, and also the early focus upon similar species.

Wilson, Scott D. 1993. Competition and resource availability in heath and grassland in the Snowy Mountains of Australia. Journal of Ecology 81: 445–451.

Although ecologists have observed vegetation zonation on mountainsides for many years, there are rather few experiments that manipulate resources to seek causes.



Interspecific pairwise interactions fall along a continuum, from symmetric to asymmetric.  In the former, each population (or individual) has an equal effect upon the other, whereas in the latter, one population (or individual) is dominant over the other (see Figure 1.6 in Keddy 2001, cited under General Overviews). Many studies look at only a pair of species, often very similar ones, hoping that this will demonstrate something useful about competition and coexistence. Even so, it is usually found that one of the species is dominant over the other, that is, that competition is asymmetric in some way (Weiner 1990; Keddy 2001, cited under General Overviews).  Even studies of intraspecific competition—which you might think of as the limiting case for similarity—show that size hierarchies arise (Weiner 1985). It is also possible that natural selection may cause even more similarity (Aarssen 1983)—a situation which we can think of as being the limiting case along a gradient from symmetrical to asymmetrical competition. At the other extreme, deliberate studies of asymmetry and hierarchies, one has to look hard for studies that explicitly examine asymmetrical interactions among species differing in traits, and involving large numbers of species. Studies using only a pair of congeneric species, or small groups of morphologically similar species like grasses, are least likely to provide insight or understanding of this general phenomenon. Studies using very different species, or even functional groups of different species, may be more likely to find asymmetry (Johansson and Keddy 1991). Existing published studies showed that competitive hierarchies are relatively widespread (Keddy and Shipley 1989). One can use large pairwise designs to find hierarchies, although as the number of species increases, these designs become cumbersome.  An alternative approach to detecting hierarchies is to screen for competitive performance using one (or preferably more) phytometer species (see Traits). Important questions remain. How does asymmetry change with species traits, or environmental conditions? Does the ranking in a hierarchy depend upon the phytometer chosen or vary across environments?  What traits might predict position in the hierarchy, or the degree of asymmetry in an interaction?  It seems probable that above ground competition inherently more asymmetric (Shipley and Keddy 1989, Cahill and Casper 2000)?  A final caution—the term “intensity” of competition is independent from the asymmetry of competition—a distinction that needs to be borne in mind whether one is reading an older experiment or planning a new one.


Aarssen, Lonnie W. 1983. Ecological combining ability and competitive combining ability in plants: towards a general evolutionary theory of coexistence in systems of competition. The American Naturalist 122: 707-731.

This paper describes the extreme case, where species are so similar that asymmetry is unimportant.  Of course, depending upon the situation, asymmetry may range from legible to overwhelming; studies of the former tend to overwhelm the latter.

Cahill, James F., Jr. and Brenda B. Casper, B. B. 2000. Investigating the relationship between neighbor root biomass and belowground competition: field evidence for symmetric competition belowground. – Oikos 90: 311–320.

An experimental study of below ground competition shows that below competition is symmetric.  Small areas with low root biomass may provide patches within which weaker competitors may establish.

Johansson, Mats E., and Paul A. Keddy. 1991. Intensity and asymmetry of competition between plant pairs of different degrees of similarity: an experimental study on two guilds of wetland plants. Oikos 60: 27–34.

One can measure asymmetry between functional groups  as well as species.  Also illustrates how to measure both asymmetry and intensity as independent aspects of a competition experiment.

Keddy, Paul A., and Bill Shipley.  1989.  Competitive hierarchies in herbaceous plant communities.  Oikos 54: 234-241.

Competitive hierarchies can be found in previously published studies.  This paper also introduces a test for detecting hierarchies in data sets containing many pair-wise interactions.

Keddy, Paul, Lisa Twolan-Strutt, and Bill Shipley. 1997. Experimental evidence that interspecific competitive asymmetry increases with soil productivity. Oikos 80: 253-256.

Eighteen species of wetland plants were grown in all possible pairwise combinations in three different environments. Asymmetry was measured for each interaction and each environment, and increased with availability of below ground resources. This experiment could easily be repeated using other groups of species and other kinds of environmental factors.

Weiner, Jacob. 1985. Size hierarchies in experimental populations of annual plants. Ecology 66: 743–752.

Within individuals of a single species, size hierarchies rapidly arise.  This is a reminder that plants are plastic, and that both interspecific and intraspecific competition may be marked by competition upon individuals differing greatly in size.

Weiner, Jacob. 1990. Asymmetric competition in plant populations. Trends in Ecology and Evolution 5: 360-364.

A general overview of the concept of asymmetry in its many manifestations.



Plants must forage simultaneously in two completely different kinds of environments— above and below ground. The obvious differences in morphology of roots and shoots illustrate the different selective pressures and ecological forces between these two environments. Below-ground interactions tend to be less studied, partly because they are “out of sight, out of mind”, although data from many published experiments suggests that root competition is often greater than shoot competition (Wilson 1988, Casper and Jackson 1997).  But below-ground effects are also less studied simply because they are harder to investigate—experiments are hard to do, unintended effects multiply, and traits may be nearly impossible to measure. As a further complication, many studies use crop plants, which provide limited insight into natural communities, since crop soils tend to be fertile and regularly tilled.  A clarifying and frequently overlooked exmple is the use of guy wires and trenches to manipulate above- and below-ground competition from trees on pine seedlings (Putz 1992).  In this situation, below-ground competition clearly matters a great deal.  Another approach is to manipulate resource availability to see, if, say, adding water or nutrients shifts species composition, or, better still, the intensity of interactions between pairs of species (Tilman 1986).  Another approach is to use naturally occurring gradients to measure above- and below-ground competition in a variety of conditions (Twolan-Strutt and Keddy 1996). With enough studies, it is possible to draw generalizations, such as the relative importance of below-ground competition in forests as opposed to prairies (Wilson 1993). Mycorrhizae in natural communities link plants in unknown ways, further complicating below-ground interactions (Woodward and Kelly 1997). Experiments run with and without mycorrhizal fungi show that competitive interactions are affected by mycorrhizae (Watkinson and Freckleton 1997). When it comes to interpreting the results of root and shoot studies, some authors assume that competition occurs everywhere at about the same intensity, and that only differences in above- and below-ground competition matter, while other authors assume that competition is dependent upon circumstances. It is, of course, in principle, a measurable issue.  One experiment asked this question explicitly, and showed that total competition increased along a fertility gradient, that root competition was more or less constant along the gradient, and that (consequently) the relative importance of shoot competition increased with fertility (Twolan-Strutt and Keddy 1996; see also Along Natural Gradients).  Until more such studies are done, generalization will remain elusive (Casper and Jackson 1997).


Capser, Brenda B. and Robert. B. Jackson. 1997. Plant competition underground. Annual Review of Ecology and Systematics 28: 545-570.

Much of the competition among plants takes place underground, and in contrast to above ground competition for light, there are many different below ground resources. This review of 144 papers concludes that below-ground competition is often greater than above ground competition, particularly in arid areas.

Putz, Francis E. 1992. Reduction of root competition increases growth of slash pine seedlings on a cutover site in Florida. Southern Journal of Applied Forestry 16: 193-197.

Guy wires and trenches used to study competition over two years between pine seedlings and mature trees. Below-ground competition was significant, while above-ground was not.  The fact that this forest paper is so rarely cited by plant ecologists illustrates how much good work is being overlooked—this is one paper you can use as a probe of an author’s citations to see whether an author has done his/her homework.

Tilman, David. 1986. “Evolution and differentiation in terrestrial plant communities: The importance of the soil resource: light gradient.” In Community Ecology. Edited by Jared Diamond and Ted J. Case, p. 359–380. New York: Harper and Row.

A useful general overview of above- and below-ground resources in plant communities and some possible trade-offs in acquiring them.

Twolan-Strutt, Lisa and Paul A. Keddy. 1996. Above- and below-ground competition intensity in two contrasting wetland plant communities. Ecology 77: 259-270.

A large field experiment showing that both competition intensity and root to shoot ratios change along a natural environmental gradient.

Watkinson, Andrew. R. and Rob P. Freckleton. 1997. Quantifying the impact of arbuscular mycorrhizae on plant competition. Journal of Ecology 85: 541–545.

An example of a competition experiment with and without the presence of mycorrhizae; overall, the mycorrhizae led to increased plant weight, and therefore to more intense competition.

Wilson, J. Bastow. 1988. Shoot competition and root competition. Journal of Applied Ecology 25: 279-296.

A review of documented above and below ground effects that actually begins with reference to Clements and summarizes the results of 47 cases collected from the literature.  In 33 of these cases, root competition was greater than shoot competition.

Wilson, Scott D. 1993. Belowground competition in forest and prairie.  Oikos 68: 146-150.

The measurement of below-ground competition in field experiments in herbaceous vegetation, set in the context of two contrasting vegetation types, woody and herbaceous.  In this case, below-ground competition was greater in the prairie than the forest.

Woodward, F. Ian, and Colleen K. Kelly. 1997. “Plant functional types: towards a definition by environmental constraints.” In Plant Functional Types. Edited by Thomas M. Smith, Herman H. Shugart, and F. Ian Woodward, p. 47-65. Cambridge: Cambridge University Press.

In unproductive habitats (or, more generally, “stressful habitats”) there may be strong selection in favor of traits that allow higher rates of nutrient uptake.  This is more of a paper about evolution and nutrition by mycorrhizae than competition per se, but it challenges us to think beyond obvious traits like height of shoots and short term soil nutrient depletion by roots.



There are many, many published models of competition, and they continue to proliferate. Models may be constructed to (1) define problems; (2) organize thoughts; (3) understand data; (4) communicate principles; (5)  test understanding; (6) make predictions. Too often writers and readers get these confused. Models constructed for one purpose may be quite unsuitable for another. The last category is likely the most important. Predictive models are designed to predict the future states of systems based on relationships specified within the model between measurable predictor (independent) variables and measurable predicted (dependent) variables. Keddy 2001, cited under General Overviews, explores many different kinds of competition models, and summarizes some twenty of them, from which this account is largely drawn. These provide a foundation to newer models, which often do not even acknowledge the presence and value of the previous models. New is not necessarily better. The following selection tends to reflect models with demonstrable general application to entire plant communities.   It begins with Skellam’s 1951 patch dynamics model and Pielou’s 1975 gradient model, both of which often have been frequently overlooked. One could argue that by addressing patches (Skellam) and gradients (Pielou), these two models are foundational. An early attempt to model interactions among much larger numbers of plants based on competition coefficients is found in Yodzis 1978. A resource consumption model  is found in Tilman 1982, cited under Traits, while another model of root/shoot allocation can be found in Tilman 1988.  There is ongoing debate over whether these latter models are testable, much less predictive. Large forest models such as JABOWA (Botkin 1993) include competition among trees for access to units of land.. These pragmatic models tend to be general and powerful, but are often not even cited, apparently in order to direct attention to studies using few species of grasses or agricultural weeds rather than entire plant communities. Single species stands (monocultures) are rare in nature, but exhibit strong quantitative relationships between plant size (or reproductive output) and density (Watkinson 1985).  Another distinctive class of models considers the biomechanical aspects of photosynthetic structures (Givnish 1982, cited under Traits); clearly there are costs to growing taller.  Another class of descriptive models explores the effects of competition upon plant diversity (the humped back model, Grime 1973,) and at larger scales, the centrifugal model (Wisheu and Keddy 1992).


Botkin, Daniel B. 1993. Forest Dynamics, Oxford University Press, Oxford.

The Jabowa model simulates the behavior of a forest by simulating the growth of individual trees on small forest plots.  It does so by exploring competitive interactions among trees in small patches of land. It is very different in nature from the sort of work more typically cited by plant ecologists, but is also more powerful in its predictions.

Grime, J. Philip. 1973. Control of species density in herbaceous vegetation. Journalof Environmental Management 1: 151–167.

A simple descriptive model (the “humped back model”, perhaps better called “the intermediate biomass model”) relates plant diversity to biomass, stress, and disturbance. This article is an early version of the principles, now better known through Figures 47 and 58 of Plant Strategies and Vegetation Processes (Grime 1979 cited under General Overviews).  The importance of such work for conservation management is still routinely overlooked.

Pielou, E. Christine. 1975. Ecological Diversity, New York: Wiley.

The section Modelling competition along a gradient (p. 90–99) is particularly useful.  This frequently-overlooked gradient model examines how two species interacting along an environmental gradient can produce different patterns of zonation.

Skellam, John G. 1951. Random dispersal in theoretical populations. Biometrika 38: 196–218.

A more accessible treatment is found in Pielou 1975 (p. 121-126), along with comments about how this reduces the possible significance of the “competitive exclusion principle” in plant communities. The model is also summarized on p. 148-150 in Keddy, Paul A. 2010. Wetland Ecology: Principles and Conservation, Cambridge: Cambridge University Press.

Tilman, David. 1988. Plant Strategies and the Dynamics and Structure of Plant Communities. Princeton, NJ: Princeton University Press.

A large computer model exploring assumptions about above-ground and below-ground allocation and their possible effects on behavior of plant communities.  Very selective referencing, but food for thought about allocation patterns. More often cited than read.

Wisheu, Irene C., and Paul A. Keddy. 1992. Competition and centrifugal organization of ecological communities: theory and tests. Journal of Vegetation Science 3: 147-156.

Multiple environmental gradients can be combined to produce a core habitat where competition predominates, contrasting with multiple kinds of peripheral habitats where environmental constraints reduce impacts of competition. The implications for biodiversity and competitive exclusion are discussed.

Yodzis, Peter. 1978. Competition for Space and the Structure of Ecological Communities. Berlin: Springer-Verlag.

An exploration of patterns that may arise out of matarices of pair-wise competition coefficients, one of the early examples of physicists entering the study of ecological phenomena. Also discusses the effects of disturbance and founder control on competitive outcomes.

Watkinson, Andrew R. 1985. “Plant responses to crowding.” In Studies in Plant Demography: A Festschrift for John L. Harper. Edited by James White, p. 275–289. London: Academic Press.

Explores the quantitative relationships between reproductive output and density in monocultures, using a standard equation.



The diversity of definitions, views, and methods present in Grace and Tilman (1990 cited under General Overviews) continues to manifest. Overall, the rate at which important ideas about competition can be tested is slowed inherently by the nature of competition experiments, which need to run for several years, if not more. Moreover, plant communities contain so many species, and there are so many potential mechanisms of competition, and so many other interactions such as mycorrhizal linking.  How, then, is one to intelligently simplify or seek generalizations?  One simplistic approach that avoids the question is to pick a couple of common species on a field station, and automatically apply reductionism.  But such habitual behaviour by ecologists provides no guarantee of progress in understanding how competition structures ecological communities and affects ecosystem functions. The range of current work illustrates work at many scales.  At the smallest scales, there are models of nutrient movement in soil (Craine et al. 2005) and experiments on infra-specific variation in competitive ability (Fridley et al. 2007). At the larger scale, there are studies of competitive interactions along stress gradients (Pennings et al. 2005, Maestre et al. 2009).  An unexpected consequence of work along gradients has been the detection of mutualism as well as competition.  There is growing evidence that as stress increases, the importance of mutualism increases (Maestre et al. 2009). At an even large scale in terms of time and space and number of species, newer phylogenetic methods provide tools for asking about longer term trends in plant competition, offering real data to test Darwin’s proposition that more similar species have stronger competitive interactions (Cahill et al. 2008). Since competition appears to be so prevalent in so many plant communities, the question remains how weaker species can survive at all.  Studies of disturbance continue to explore the factors that create gaps in plant communities, the nature of these gaps, and the processes by which species disperse to fill these gaps.  The rate of disturbance, and the rate of competitive exclusion, can achieve some sort of balance, determining, in turn, the relate abundance of competitive dominants and weaker competitors, that is, patch-dependent species (Brewer 1998, Cahill and Casper 2000, Pennings et al 2005).

Brewer, J. S. 1998. Effects of competition and litter on a carnivorous plant, Drosera capillaris (Droseraceae). American Journal of Botany 85: 1592–1596.

The production of persistent litter is a trait that suppresses neighbours, even in pine savannahs were there are low nutrient levels and frequent fires. This field study shows how such litter suppresses a carnivorous plant.

Cahill, James F. Jr. 1999. Fertilization effects on interactions between above- and belowground competition in an old field. Ecology 80: 466–480

Above ground and below ground competition may interact, complicating measures of total competition intensity. In general, competition in infertile plots is likely to be symmetric.  Figure 4 offers a general model for changes in competition along a productivity gradient.

Cahill, James F. Jr., Steven W. Kembel, Eric G. Lamb and Paul A. Keddy. 2008. Does phylogenetic relatedness influence the strength of competition among vascular plants? Perspectives in Plant Ecology, Evolution and Systematics 10: 41-50.

Phylogenetic relationships were assessed for 50 vascular plant species grown against 92 competitor species. There was no relationship between relatedness and competition for eudicots competing with other eudicots, while monocots did compete more intensely with closely related monocots than with distantly related monocots. Overall, functional traits appear more important that relatedness.

Craine, Joseph M., Joseph Fargione and Shinya Sugita. 2005. Supply pre-emption, not concentration reduction, is the mechanism of competition for nutrients. New Phytologist 166: 933-940.

A fine scale model of nutrient concentration gradients between the roots of plants shows that competitive dominants may pre-empt nutrient supplies from competitive subordinants, thereby falsifying a central assumption of competition models based upon R* (such as Tilman 1982 cited under Traits) that focus on concentration reduction rather than supply pre-emption.

Fridley, Jason D., J. Philip Grime and Mark Bilton. 2007. Genetic identity of interspecific neighbours mediates plant responses to competition and environmental variation in a species-rich grassland. Journal of Ecology 2007.

Different genotypes of both a dominant grass and a dominant sedge have different competitive abilities, and these are influenced by the genotype of neighbouring individuals.

Maestre, Fernando T., Ragan M. Callaway, Fernando Valladares and Christopher J. Lortie. 2009. Refining the stress-gradient hypothesis for competition and facilitation in plant communities. Journal of Ecology 97: 199–205.

The relative important of facilitation and competition may vary inversely across abiotic stress gradients. It is not only the number of such interactions, but their strength, that needs to be evaluated.

Pennings, Steven C., Mary-Bestor Grant and Mark D. Bertness. 2005. Plant zonation in low-latitude salt marshes: disentangling the roles of flooding, salinity and competition. Journal of Ecology 93: 159–167.

Lower limits of species appear to be set by intolerance to stress, while upper limits appear to be set by intolerance to competition. Since salinity and flooding vary with latitude, there are likely latitudinal gradients in the relative importance of stress and competition in wetlands.


(--the end--)