Photo Gallery

Red-eyed Treefrog Metamorphs

(all rights reserved, used by permission)

click on a photo below for a larger image

(Lower two pairs) Ecological character displacement such as that between omnivores and carnivores is only possible in ponds where both resources are abundant (upper left photo). In ponds in which detritus is rare (upper right photo), S. multiplicata is absent, and in ponds in which shrimp are rare (lower photos), S. bombifrons is absent. Thus, ecological character displacement, and therefore coexistence of close competitors, is only possible when diverse resources are available.

The sun sets along the road westward to Winton, in Australia’s northern rangelands. Like most rangelands, those in Australia are characterized by a high degree of resource variation, in both time and space. Before Europeans colonized these landscapes, seminomadic humans buffered resource variation through complex social institutions.

Pastoralism and private-property rights have since fragmented and disconnected Australian landscapes. But modern Australian pastoral systems are developing their own informal institutions, which are restoring rangeland connectivity over massive spatial scales.

A male Lark Bunting (Calamospiza melanocorys) is walking through vegetation after delivering food to his recently fledged young. Tracking radio-marked adults allowed us to track fledglings without radio transmitters.

        
Santa Monica Mountains National Recreation Area, Los Angeles County, California

Fuel break construction and maintenance methods can disturb soils and remove native plant cover, making fuel breaks susceptible to invasion by nonnative plants. Many fuel breaks have significantly higher cover, richness, and density of nonnative species than adjacent wildlands, particularly in areas that have been subject to recurrent fires or grazing.

 A marine iguana foraging in the low intertidal, Puerto Egas, Santiago Island.

The study shows that the outcome of plant–herbivore interactions on the diversity of rocky shore communities was drastically affected by changes in the in situ algal productivity (bottom-up effect) associated with the 1997–1998 El Niño and the posterior La Niña events, i.e., large-scale temporal fluctuations in temperature and nutrient levels. In non-El Niño years, herbivores tend to decrease the diversity of sessile organisms to a few grazing-resistant forms (top-down effects), but the El Niño impact on these consumers was subordinated to the local algal response to the spatiotemporal differences and fluctuations in the regional oceanographic conditions. These results provide new insight into the regulation of tropical rocky shores and the likely consequences of large-scale climate change on marine communities.

Introduction

The stability of natural ecosystems, including their resilience and resistance to perturbations, rests to a large degree on interactions and components of biotic communities (May 1973, 1974, De Angelis 1992, Holling 1996, Gunderson and Pritchard 2002, Ludwig et al. 2002). Interspecific interactions of communities, including, for example, the commonly studied predator–prey and competitive interactions, form the basis for ecological communities. Initial mathematical investigations into the stability of ecosystems revealed that interspecific interactions like those between predator and prey and competing species act to stabilize communities. In contrast, interspecific mutualistic interactions (those that benefit both interacting partner species through increases in per capita demographic rates of survival and/or reproduction [Holland 2002]) destabilize communities due to the inherent positive feedback associated with the interactions (May 1973, 1974). Initial mathematical models incorporating interspecific mutualism into communities, and analyzing for stability (May 1973, 1974), indicated that mutualisms must be unimportant and uncommon in nature, as they destabilize populations within the community. However, in nature, we know mutualism to be a common feature of natural communities worldwide, and empirical studies demonstrate that positive feedback associated with these interactions does not result in unbounded population growth (Heithaus et al. 1980, Thompson 1982, Roubik 1992:327–354). Thus, prior theory has not stood up to the empirical test. Despite our limited understanding of pairwise mutualistic interactions compared to pairwise predator–prey and competitive interactions, today these interactions are increasingly recognized as fundamental to the structure and dynamics of natural communities worldwide. Examples include pollinators and seed dispersers in tropical forests; nitrogen-fixing bacteria in deserts and agroecosystems; mycorrhizal fungi in grasslands; lichens in tundras; corals in marine systems; and microbes in deep-sea vents. Moreover, the ecosystem services mutualists provide are leading them to be increasingly considered a conservation priority (Buchman and Nabhan 1996, Costanza et al. 1997, Wall and Moore 1999).

There are two philosophically distinct classes of models that have been developed to address the effects of mutualist–nonmutualist interactions on community stability: (1) models whose predictions of community dynamics are sensitive to specific biological details about that community, and (2) models whose predictions of community dynamics are robust despite changes in the biological details of the system. Within each class of models are three ecologically distinct types of mathematical models: (a) population-dynamic models; (b) multispecies resource-based population dynamic models; and (c) multispecies spatiotemporal models.

The specific goals of this review are threefold. First, I present a brief history of the development of theory for exploring the population dynamics of interspecific mutualism. I then present theoretical approaches that are being used to address the effects of interactions between mutualists and exploiters on community stability. For each approach, I detail results and highlight models that have the potential to be operational in multiple empirical settings. This review can be thought of as preemptive: although the theory is in its infancy, there is a clear distinction between operational and nonoperational models. Without being too bold, I hope this paper leads theoreticians and empiricists toward models that are robust in their predictions, given only a general biological framework within which to work, as opposed to models whose predictions are only relevant for a specific set of a priori assumptions about the biological details of a specific system.

1. Theory development: the effects of intrinsic constraints on the stability of pairwise mutualistic interactions

Historically, mathematical analysis of communities has begun with mathematical models that explore the population dynamics of predator–prey and competitive interactions. The most commonly used models have been those that compare prey population dynamics in the presence or absence of a predator. For example, May (1973, 1974) used the classical continuous-time Lotka-Volterra ecological model (Lotka 1925, Volterra 1926) for n species (hereafter, Eq. 1) as a template for exploring the effect of changing the number of species in a predator–prey interaction on population dynamics. The model is described as follows:

             (1)

where N_i is the number of individuals, and R_i is the intrinsic rate of growth of species i. The parameter aij describes the negative density-dependent effect of species j on species i. The goal of such models is to track changes in species’ abundances over time. Predictions generated from these types of models are unrealistic, however, because they ignore nonlinearities in per capita growth rates, which when incorporated into the model, give rise to complicated system dynamics far from equilibrium (e.g., unbounded population growth).

May (1973, 1974) demonstrated that changes in the population dynamics of pairwise mutualistic interactions could be tracked by changing the sign in Eq. 1 from negative to positive. However, because this model relied on linear functional responses, positive feedback resulted in unbounded population growth of all populations, unless interaction coefficients were small, so that mutualism had little influence on interacting species. Numerous theoretical works have demonstrated that the interaction between two mutualist species can be stabilized by incorporating intrinsic constraints into Eq. 1. For example, the introduction of resource-handling times or metabolic costs incurred by one or both mutualists helps to stabilize the interaction (Vandermeer and Boucher 1978, May 1981, Addicott 1986). Recent mathematical models of population change in pairwise mutualistic interactions have shown that the benefits of mutualism that leads to positive feedback (i.e., saturation of benefits to per capita demographic rates of survival and/or reproduction) do not last forever. For example, by incorporating asymptotically or unimodal saturating functional responses into Model 1, Holland et al. (2002) were able to explain how mutualism has positive feedback, but not indefinitely. Specifically, their model showed that fundamental differences in population dynamics can occur when net effects to that population change linearly, unimodally, or in a saturating fashion.

2. Class 1 models

Multispecies population-dynamic models

Interestingly, mathematical models for exploring the population dynamics of mutualism in the presence of a third, nonmutualistic species are sparse, and have only recently started to grow in number. This can be attributed to two factors: (1) theoretical ecologists have only recently developed theory for understanding intrinsic mechanisms that stabilize pairwise mutualistic interactions, and (2) the introduction of a third nonmutualistic species into the basic Lotka-Volterra model for mutualism makes interpretation of qualitative dynamics and determination of stability criteria difficult . Although theory has been outpaced by empirical observation, models have been developed to explain the ecology of mutualism in a multispecies community context. Several studies have demonstrated that the incorporation of extrinsic constraints like a third species, either a predator or competitor, acts to dampen mutualist growth (Heithaus et al. 1980, Freedman and Rai 1987, 1988, Freedman et al. 1987). For example, Ringel et al. (1996) built a four-species community model by generalizing Eq. 1 to explore the effects of two additional species, a predator and prey, on the per capita growth of two mutualist species. The effects of each interaction type on the per capita growth of each population was either negative (self and predator effects), neutral (two species that did not interact), positive (prey and mutualism effects), or indeterminate (positive effects for mutualism, negative effects for nectar theft). Results of the model demonstrated that community interactions stabilized the four-species community. However, this type of model did not include intrinsic constraints on the inherent positive feedback between mutualists, which we know to be true in nature.

With this in mind, the model of Bacher and Fiedli (2002) is more appropriate for analyzing the dynamics of community stability because their model takes into account interspecific feedback between mutualists. The authors developed deterministic difference equation Lotka-Volterra models to ask the question: How was the well-studied mutualism between the shoot-base boring weevil Apion onopordi and the rust fungus Puccinia punctiformis influenced by the dynamics of their shared host plant Cirsium arvense, and vice versa (Watson and Keogh 1980, Thomas et al. 1994, Fiedli and Bacher 2001). In one version of the model, the rust pathogen was introduced into a thistle population that had reached carrying capacity, and the pathogen was then introduced into a thistle population well below carrying capacity. In the second version of the model, the authors allowed unbounded thistle population growth in order to evaluate the influence of host-plant dynamics on the qualitative dynamics of the three-species system. For both models, intrinsic constraints on the inherent positive feedback between mutualists included a trade-off in benefits of the mutualism for the weevil: ovipositing in infected shoots created an optimal food source for the first generation of offspring, but did not spread the rust, thereby leaving the second generation with an inadequate food supply. Growth of the weevil population increased when feeding on rust-infected plant tissue, which increased byproduct benefits received by the rust, which resulted in infection of healthy thistle in the year following weevil attack. The authors hypothesized that differences in the density of the weevil’s egg load between healthy and rust-infected thistles would regulate population abundances of the mutualists and their host plant. In contrast to the expected hypothesis, both models showed that the population dynamics of the mutualists were largely determined by the dynamics of their host plant. Analysis of the first model showed that neither the starting thistle population size nor its initial carrying capacity had an influence on the size of the thistle population at equilibrium. The second model demonstrated that changes in certain parameter values resulted in unbounded thistle population growth, which resulted in all three species either growing unbounded or going extinct. Hence, three-species coexistence was never reached, which led the authors to conclude that the mutualism between the weevil and the rust alone could not create or maintain system stability. In contrast, the population dynamics of the resource largely determined the stability of the host-plant mutualist.

Empirical evidence has demonstrated that exploiters of pollinating seed-parasite mutualisms are present in some communities (Kurdelhue et al. 2000, Pellmyr and Leebens-Mack 2000). Guided by these empirical findings, Morris et al. (2003) developed a mathematical model to explore conditions for stability of an obligate mutualism between a pollinating seed parasite and its plant host, and an exploiter (nonpollinating seed parasite) of the host. The parameters in their model were based on the general biological properties of a well-studied natural system consisting of flowering plants, pollinating seed parasites, and nonpollinating exploiters (Pellmyr and Huth 1994, Kurdelhue and Rasplus 1996, Pellmyr et al. 1996, Pellmyr and Lubens-Mack 2000, Marr et al. 2001). The goals of the model were twofold: (1) One goal was to explore how the rate of interaction between obligate plant-pollinating seed parasite mutualists and nonpollinating exploiters and the type of interspecific and intraspecific competition affected whether the exploiter could invade the plant–pollinator mutualism. (2) The authors also wanted to know whether three-species coexistence was stable. The model was based on the following simplifying assumptions: each fertilized ovule could support at most, one pollinator or one exploiter larva to pupation, and larval survival of exploiter and pollinating seed parasites was reduced by intraspecific competition when more then one egg was oviposited into a single ovule. Thus, the inherent positive feedback between mutualists was constrained by competition. Analysis of the model demonstrated that when competition among and between pollinators and exploiters was weak, the exploiter could invade and the three-species system could persist at a stable equilibrium. However, weak intraspecific competition resulted in a competitive advantage of pollinators over exploiters, which made invasion of the exploiter population and three-way coexistence impossible.

3. Class 2 models

Multispecies resource-based population dynamic models

Nearly all mutualisms, as is the case with predator–prey and competitive interactions, involve consumer–resource interactions (i.e., the exchange of nutrients between species [Holland et al. 2005]). The chemostat is both a laboratory apparatus and a theoretical construct that is used to investigate the ecological dynamics of consumer–resource interactions in continuous culture. The ecological dynamics of Eq. 1 can be examined within a theoretical chemostat environment, in which the effects of nutrient flow between species can be used to infer consequences of changes in population stability of pairwise predator–prey, competitive, and mutualistic interactions on the stability of communities.

In a theoretical chemostat environment, the culture vessel in which the species interact is mixed continuously, which results in a spatially homogenous distribution of nutrients, organisms, and byproducts. Thus, spatial interactions within the vessel can be neglected (Smith and Waltman 1995). Historically, classic continuous-time Lotka-Volterra differential equations have been the mathematical foundation for investigating interactions between predator–prey and competing species in a chemostat. However, the linear functional responses inherent in these equations give rise to unbounded solutions, whereby population sizes of interacting species grow indefinitely (May 1981, Murray 1989, Kot 2001). One solution to this problem is to include asymptotically saturating functional responses in the general Lotka-Volterra model. This allows the largely phenomenological Lotka-Volterra models to include mechanism. Such functional responses are useful because they render unbounded solutions impossible (Smith and Waltman 1995, Kooi et al. 2004) and allow for explicit modeling of nutrient flow through communities (Smith and Waltman 1995, Kooi et al. 2004). Surprisingly, only a limited number of studies have used the chemostat as a theoretical environment for exploring the effects of interactions between mutualists and exploiters on community stability. Kooi et al. (2004) developed a chemostat model of a well-mixed community of two mutualistic prey species with substitutable symbiosis coexisting on a single nutrient, and a predator population that consumed one (specialist) or both (generalist) prey populations. Both prey species were coprophagic mutualists (i.e., each species mutually benefited the other by feeding on the other’s waste products). The authors constrained the inherent positive feedback between mutualists by using a mass-balance model formulation. The model demonstrated that coexistence between a generalist predator feeding on two mutualist populations could only occur if exploitation of the mutualists by the prey was below a critical value. If exploitation was low enough, the predator was able to invade through the boundary equilibrium of two-mutualist coexistence. If exploitation was too great, the community collapsed and all species went extinct. Qualitative behavior of the dynamics of the model were robust to changes in intrinsic species parameters (e.g., intrinsic rate of growth, per capita death rate) and to parameters describing movement of waste products through the system.

Spatial–temporal models

The current trend in the theoretical study of the effects of mutualist-exploiter interactions on community stability is the use of simulation models incorporating spatial structure. The goal of these types of models is to understand how community stability is driven by the spatial distribution of populations. Empirical studies have demonstrated that there are interspecific mutualisms whose dynamics are regulated by their distribution among habitat patches (e.g., Thompson 1994, Yong et al. 1997, Yu and Davidson 1997, Parker 1999). Empirical studies have also demonstrated that attacks of mutualisms by exploiters in spatially heterogeneous environments can occur in a random pattern over a wide range of spatial scales (e.g., Hawkins and Compton 1992, Bronstein and Hossaert-McKey 1996, Yu et al. 2001). Wilson et al. (2003) used general biological properties derived from these empirical studies to develop an individual-based simulation model that incorporated stochastic interactions and spatial structure. Their model system was composed of two obligate mutualist species and an exploiter species of the mutualism. Like past models, the parameters in their model were based on the general biology of the well-studied interactions between a single species of plant, its pollinator insect, and an exploiter of the plant. However, their model was unique because it incorporated spatial heterogeneity in the distribution of the mutualist and exploiter species. The authors first constructed and analyzed a model of individual-scale interactions between the two obligate mutualist species and an exploiter species. Inherent feedback between mutualists was constrained by adding density-dependent effects in plant reproduction and by adding reduction in seed number as pollinator density became too high. Simulation results demonstrated that incorporation of spatial heterogeneity into the individual-based, nonspatial model resulted in the stable coexistence of all three populations. Analysis of the model demonstrated patterns of temporally constant, fixed-habitat patches. These patterns were robust to changes in the intrinsic parameters of mutualist and exploiter species (e.g., mutualist/exploiter density and life history traits).

Bronstein et al. (2003) used a spatial-stochastic simulation model with spatial heterogeneity to explore the population dynamics of obligate, species-specific mutualisms. They based their model on the general biological properties of the well-studied pollinating seed parasite exhibited by fig/fig wasp and yucca/yucca moth interactions (e.g., Dufay and Anstett 2003). Their model system consisted of plants and pollinating seed parasite insects in the presence of one of two obligate exploiter species: florivores or exploiter insects that parasitized seeds but failed to pollinate. Positive feedback inherent to the mutualism was constrained by extrinsic effects of exploitation, rather than the intrinsic effects generated by the mutualist interaction. Simulation results showed that mutualist populations could persist over ecological time in the presence of exploiter specialists, but temporal and spatial dynamics of the mutualist populations were altered in the presence of the exploiter. These patterns were robust to changes in the intrinsic parameters of exploiter species (e.g., exploiter density and life history traits).

Conclusions

This review has highlighted recent theoretical progress in exploring the effects of mutualist–exploiter interactions on community stability. There are two philosophically distinct classes of models for exploring these effects: (1) bottom-up or individual-case models that are constructed from the biological details of well-studied communities, and (2) top-down or system-based models that are constructed with limited a priori knowledge of the biology of the system. There are many weaknesses in individual case models, among them the fact that they are usually only operational in one particular biological setting, and the fact that their predictions are sensitive to changes in biological details. An additional weakness of the class 1 population dynamic models described in this paper is their use of simplifying assumptions about the biology of communities to make predictions of the model more tractable. This can result in spurious results, because these simplifying assumptions often do not match biological reality. For example, the assumption that only one ovule is produced per plant per season is unrealistic. Most importantly, unlike the predictions of class 1 models, which rely on the inductive method of hypothesis testing, the predications of class 2 models do not rely heavily on a priori knowledge of the biology of a system. Hence, class 2 models are robust in their predictions over a broad range of parameter space, and have the potential to be operational in multiple empirical settings.

Acknowledgments

Michael R. Golinski is supported by NSF grant DEB-0129630 to Dr. William J. Boecklen.

Literature cited

Addicott, J. F. 1986. On the population consequences of mutualism. Pages 425–436 in J. Diamond and T. J. Case, editors. Community ecology. Harper and Row, New York, New York, USA.

Atmar, W., and B. Patterson. 1993. The measure of order and disorder in the distribution of species in fragmented habitat. Oecologia 96:373–382.

Bacher, S., and J. Friedli. 2002. Dynamics of mutualism in a multi-species context. Proceedings of the Royal Society of London B 269:1517–1522.

Bronstein, J. L., and M. Hossaert-McKey. 1996. Variation in reproductive success within a subtropical fig/pollinator mutualism. Journal of Biogeography 23:433–446.

Bronstein, J. L., W. G. Wilson, and W. F. Morris. 2003. Ecological dynamics of mutualist/antagonist communities. American Naturalist 162:1–20.

Buchman, S. L., and G. P. Nabhan. 1996. The forgotten pollinators. Island Press, Washington, D.C., USA.

Costanza, R., R. d’Arge, R. de Groot, R. Farber, S. Grasso, M. Hannon, B. Limburg, K. Naeem, S. O’Neill, R. V. Paruelo, J. Raskin, P. Sutton, and M. van den Belt. 1997. The allure of the world’s ecosystem services and natural capital. Nature 387:253–260.

DeAngelis, D. L. 1992. Dynamics of nutrient cycling and food webs. Chapman and Hall, London, UK.
Dufay, M., and M. C. Anstett. 2003. Conflicts between plants and pollinators that reproduce within inflorescences: evolutionary variations on a theme. Oikos 100:3–14.

Freedman, H. I., J. F. Addicot, and B. Rai. 1987. Obligate mutualism with a predator: stability and persistence of three-species models. Theoretical Population Biology 32:157–175.

Freedman, H. I., and B. Rai. 1987. Uniform persistence and global stability in models involving mutualism 1. Predator-prey-mutualistic systems. In J. B. Shukla, T. G. Hallam, and V. Capasso, editors. Elsevier, Amsterdam, The Netherlands.

Freedman, H. I., and B. Rai. 1988. Uniform persistence and global stability in models involving mutualism 2. Competitor–competitor-mutualist systems. Indian Journal of Mathematics 30:175–186.

Friedli, J., and S. Bacher. 2001. Mutualistic interaction between a weevil and a rust fungus, two parasites of the weed Cirsium arvense. Oecologia 129:571–576.

Gunderson, L. H., and L. Pritchard, Jr. 2002. Resilience and the behaviour of large-scale systems. Island Press, Washington, D.C., USA.

Hawkins, B. A., and S. G. Compton. 1992. African fig wasp communities: undersaturation and latitudinal gradients in species richness. Journal of Animal Ecology 61:361–372.

Heithaus, E. R., D. C. Culver, and A. J. Beattie. 1980. Models of some ant–plant mutualisms. American Naturalist 116:347–361.

Holland, J. N. 2002. Benefits and costs of mutualism: demographic consequences in a pollinating seed–consumer interaction. Proceedings of the Royal Society of London B 269:1405–1412.

Holland, J. N., and D. L. DeAngelis. 2001. Population dynamics and the stability of obligate pollination mutualisms. Oecologia 126:575–586.

Holland, J. N., and D. L. DeAngelis. 2002. Ecological and evolutionary conditions for fruit abortion to regulate pollinating seed-eaters and increase plant reproduction. Theoretical Population Biology 61:251–263.

Holland, J. N., D. L. DeAngelis, and J. L. Bronstein. 2002. Population dynamics and mutualism: functional responses of benefits and costs. American Naturalist 159:231–244.

Holland, J. N., J. H. Ness, A. L. Boyle, and J. L. Bronstein. 2005. Mutualisms as consumer–resource interactions. In P. Barbosa and I. Castellanos, editors. Ecology of predator–prey interactions. Oxford University Press, Oxford, UK.

Holling, C. S. 1996. Surprise for science, resilience for ecosystems, and incentives for people. Ecological Applications 6:733–735.

Kooi, B. W., L. D. J. Kuijper, and S. A. L. M. Kooijman. 2004. Consequences of symbiosis for food web dynamics. Journal of Mathematical Biology 49:227–271.

Kot, M. 2001. Elements of mathematical ecology. Cambridge University Press, Cambridge, UK.
Kumar, R., and H. I. Freedman. 1989. A mathematical model of facultative mutualism with populations interacting in a food chain. Mathematical Biology 97:235–261.

Kurdelhue, C., J. P. Rossi, and J. Y. Rasplus. 1996. Nonpollinating Afrotropical fig wasps affect the fig-pollinator mutualism in Ficus within the subgenus Sycamorus. Oikos 75:3–14.

Kurdelhue, C., J. P. Rossi, and J. Y. Rasplus. 2000. Comparative community ecology studies on Old World figs and fig wasps. Ecology 81:2832–2849.

Lotka, A. J. 1925. Elements of physical biology. Williams and Wilkins, Baltimore, Maryland, USA.
Ludwig, D., B. H. Walker, and C. S. Holling. 2002. Models and metaphors of sustainability, stability and resilience. Chapter 2 in L. H. Gunderson, editor. Resilience and the behavior of large-scale systems. Island Press, Washington, D.C., USA.

Marr, D. L., M. T. Brock, and O. Pellmyr. 2001. Coexistence of mutualists and antagonists: exploring the impact of cheaters on the yucca–yucca moth mutualism. Oecologia 128:454–463.

May, R. M. 1973. Stability and connectance in model ecosystems. Monographs in Population Biology Volume 6. Princeton University Press, Princeton, New Jersey, USA.

May, R. M. 1974. Stability and complexity in model ecosystems. Princeton University Press, Princeton, New Jersey, USA.

May, R. M. 1981. Theoretical ecology: principles and applications. Sinauer Associates, Sunderland, Massachusetts, USA.

Morris, W. F., J. L. Bronstein, and W. G. Wilson. 2003. Three-way coexistence in obligate mutualist–exploiter interactions: the potential role of competition. American Naturalist 161(6):860–875.

Parker, M. A. 1999. Mutualism in metapopulations of legumes and rhizobia. American Naturalist 153:S48–S60.

Pellmyr, O., and C. J. Huth. 1994. Evolutionary stability of mutualism between yuccas and yucca moths. Nature 372:257–260.

Pellmyr, O., and J. Leebens-Mack. 2000. Reversal of mutualism as a mechanism for adaptive radiation in yucca moths. American Naturalist 156:S45–S61.

Pellmyr, O., J. Leebens-Mack, and C. J. Huth. 1996. Non-mutualist yucca moths and their evolutionary consequences. Nature 380:155–156.

Ringel, M. S., H. H. Hu, and G. Anderson. 1996. The stability and persistence of mutualisms embedded in community interactions. Theoretical Population Biology 50:281–297.

Roubik, D. W. 1992. Loose niches in tropical communities: why are there so few bees and so many trees? Pages 327–354 in M. D. Hunter, T. Ohgushi, and P. W. Price, editors. Effects of resource distribution on animal–plant interactions. Academic Press, San Diego, California, USA.

Smith, H. L., and P. Waltman. 1995. The theory of the chemostat. Cambridge University Press, Cambridge, UK.

Thomas, R. F., T. J. Tworkoski, R. C. French, and G. R. Leather. 1994. Puccinia puntiformis affects growth and reproduction of Canada thistle (Cirsium arvense). Weed Technology 8:488–493.

Thompson, J. N. 1982. Interaction and coevolution. Wiley, New York, New York, USA.

Thompson, J. N. 1994. The coevolutionary process. University of Chicago Press, Chicago, Illinois, USA.

Vandermeer, J. H., and D. H. Boucher. 1978. Varieties of mutualistic interactions in population models. Journal of Theoretical Biology 74:549–558.

Volterra, V. 1926. Variations and fluctuations of the number of individuals in animal species living together. Pages 409–448 (translated) in R. N. Chapman. Animal ecology. McGraw-Hill, New York, New York, USA.

Wall, D. H., and J. C. Moore. 1999. Interactions underground. BioScience 49:109–117.

Watson, A. K., and W. J. Keogh. 1980. Mortality of Canada thistle due to Puccinia punctiformis. Pages 325–332 in E. S. Del Fosse, editor. Proceedings of the Fifth International Symposium on Biological Control of Weeds. CSIRO, Brisbane, Australia.

Wilson, W. G., W. F. Morris, and J. L. Bronstein. 2003. Coexistence of mutualists and exploiters on spatial landscapes. Ecological Monographs 73:397–413.

Yong, T. P., C. H. Stubblefield, and L. A. Isbell. 1997. Ants on swollen-thorn acacias: species coexistence in a simple system. Oecologia 109:98–107.

Yu, D. W., and D. W. Davidson. 1997. Experimental studies of species-specificity in Cecropia–ant relationships. Ecological Monographs 67:273–294.

Yu, D. W., H. Wilson, and N. E. Pierce. 2001. An empirical model of species coexistence in a spatially structured environment. Ecology 82:1761–1771.

Michael R. Golinski
Department of Biology
New Mexico State University
Las Cruces, New Mexico 88003 USA
(505) 646-5770
Fax: (505) 646-5665
E-mail: mgolinsk@nmsu.edu

Scientists and academic institutions widely use Impact Factors ‹http://wos.mimas.ac.uk/› to evaluate the relative importance of journals. Although sometimes considered controversial, publishing in relatively high Impact Factor journals has been broadly applied as a stamp of approval for hiring and promotions, to rate the accomplishments of academic departments, and the importance of particular disciplines. Both authors and publishers strive to publish high impact journal articles, and the pressure to do so has apparently led to an insidious abuse in how some publishers correspond with authors of nearly accepted manuscripts. At or before the time of acceptance, several journals’ editors are requesting that authors cite additional papers published in that same journal. Some of these requests are general such as ‘We would also appreciate it if you would consider citing relevant past papers [from our journal] in your manuscript’, whereas others are more specific, with journal editors indicating one to several recent (often unpublished) citations.

Although the extent of this practice is unknown, at least four major journals in the area of ecology and evolutionary biology routinely encourage such self citation. Because Impact Factors are calculated by dividing the number of citations in the current year (e.g. in 2004) by the total number of articles published in the two previous years (i.e. in 2003 and 2002), citation of articles relatively hot off the press will increase the Impact Factor of a journal.

A gentle nudge by an editor to cite additional papers if relevant is all too easy to be uncritically accepted by most authors who are simply overjoyed with the news that their paper has been accepted. To maintain the integrity of objective scientific research, this questionable policy that essentially results in the ‘businessification’ of science must be stopped. Publishers should be embarrassed and authors should not comply.

[Reprinted from Trends in Ecology and Evolution, Volume 20, No. 4, Anurag A. Agrawal,
“Corruption of journal impact factors,” page 157, Copyright (2005), with permission from
Elsevier.]

Anurag A. Agrawal
Department of Ecology and Evolutionary Biology
Cornell University
Ithaca, NY 14853 USA

On 12 October 2005, the National Academy of Sciences of the United States of America released one of the most comprehensive looks at these concerns in a massive document called Rising above the Gathering Storm. The link to the executive summary is ‹http://www.nap.edu/execsumm_pdf/11463.pdf›, and the complete report can be found at ‹http://www.nap.edu/books/0309100399/html/1.html›. The blue ribbon committee that produced The Gathering Storm and the focus groups that assembled the in-depth information provide 10 recommendations.

Although The Gathering Storm is a United States document, probably many of the issues and suggestions are applicable in most countries. However, as ecologists, this document offers us another opportunity to remind the public and policy makers that the innovation that will drive the world’s economies in the 21st century will require considerably better understanding of the environment and ecology than has been true in the past. Some of the innovations in past centuries have created the environmental problems of today.

Rising above the Gathering Storm identifies two important challenges: creating high-quality jobs for Americans, and responding to the nation’s need for clean, affordable, and reliable energy. Certainly both of these challenges are not restricted to the United States; creating high-quality employment and finding environmentally and ecologically sound methods for producing energy to run our economies must be the goal of all countries. Again, ecologists might add to these two key characteristics, the fact that science and engineering must incorporate ecosystem services into their innovative developments. Together with the development of new energy sources, the addition of ecosystem services will be one of the biggest innovations since agriculture.

E. A. Johnson
Editor-in-Chief