2. What mechanisms cause positive interactions
3. Can we predict when positive or negative interactions may be important?
4. What do positive interactions mean for community theory?
Current plant community ecology, as presented in most textbooks, often promotes the perspective that communities are produced only by the traits of populations and that assemblages of different plant species exist primarily because each shares adaptations to particular abiotic conditions. To some degree, this perspective leads to the conclusion that plant communities are simply a handy typological construct. However, a large body of research accruing during the last 30 years demonstrates that many if not most plant communities have fascinating interdependent characteristics, and although they are not “organic entities,” it is clearthat many species create conditions that are crucial for the occurrence and abundance of other species. This research is the focus of this chapter.
continuum. A distribution of many species along a gradient in which each species appears to be distributed randomly with respect to other species
facilitation. The positive effect of one species on another
holistic communities. The idea that species within a community are highly interdependent, forming organism-like units
hydraulic lift. The process by which some plant species passively move water from deep in the soil profile, where water potentials are high, to more shallow regions where water potentials are low
indirect interactions. Interactions between two species that are modified by a third species
individualistic communities. The idea that communities are fundamentally groups of populations that occur together primarily because they share adaptations to the same abiotic environment; communities do not have organism-like qualities
niche complementarity. The condition in which different niches result in variation in the utilization of resources or space
As the discipline of ecology emerged from its biogeographic origins in the early 1900s, two strikingly polar views on the nature of plant communities vied for recognition, and the conflict established a precedent for ecological thought today. Initially, the view of Frederic Clements was ascendant with most ecologists accepting the idea that
[T]he community is an organic entity. As an organism the community arises, grows, and dies. Furthermore, each community is able to reproduce itself, repeating with essential fidelity the stages of its development… comparable in its chief features with the life history of an individual plant. (Clements, F. E. 1916. Plant Succession. Washington, DC: The Carnegie Institution, Publication 242)
This holistic perspective, however, was replaced in the middle of the 1900s by new ideas promoted by Henry Gleason. In this new individualistic world view, the community “is merely the resultant of two factors, the fluctuating and fortuitous immigration of plants and an equally fluctuating and variable environment… not an organism, scarcely even a vegetational unit, but merely a coincidence” (Gleason, H. A. 1917. The structure and development of the plant association. Bulletin of the Torrey Botanical Club 44: 463–481). It would be hard to dream up two more diametrically contrasting perspectives for how species are organized into groups.
Texts have a strong individualistic flavor, but most ecologists are fully aware that the nature of plant communities is more nuanced than the hyperdichotomy of individualistic versus organismal communities. Perceiving such nuances is important, but the dominant individualistic perception of plant community organization has probably left lingering but strong effects on the way we conduct research, leading to a great deal of information on negative interactions such as predation, competition for resources, and allelopathy. However, this dominant perception has probably impeded the progress of empirical research on facilitation and indirect interactions among plants.
Understanding the nature of communities is not just an academic issue. Whether or not communities have weak or strong tendencies toward independent or interdependent assembly has strong implications for conservation. For example, the view that plant species are fully individualistic and interchangeable in communities has been used to advocate active human involvement in “shaping and synthesizing new ecosystems, even in the ‘natural’ environment” (italics mine; Johnson and Mayeux, 1992). This may be reasonable if maintaining functional plant communities is simply a matter of finding a suite of populations that can grow in a particular set of conditions. But if interactions among plants are more complex and interdependent, as suggested by research on facilitation, indirect effects of herbivores and mycorrhizae, and networks of direct and indirect interactions within the plant community, shaping and synthesizing new communities may not work. Conservationists typically assume a high degree of interdependence in communities when they argue for the preservation of natural systems and biological diversity. The Ecological Society of America recommends the following for conservation priorities: Does the species play an especially important role in the ecosystem in which it lives? Do other species depend on it for their survival? Will its loss substantially alter the functioning of the ecosystem? These priorities assume interdependence.
In this chapter, I focus on several general questions: What mechanisms cause positive interactions? Can we predict when positive or negative interactions may be important? What do positive interactions mean for community theory?
Positive interactions can be direct, the effect of one species on one other species, or positive interactions can be indirect, requiring an intermediate species in order to occur (see chapter III.5). There are many direct and indirect facilitative mechanisms, probably far more than mechanisms for resource competition, and they can be difficult to separate experimentally or conceptually. I present a brief overview of mechanisms here, but for more detail on mechanisms see Callaway (2007).
Shade from other species can keep plant tissues below lethal temperatures, decrease respiration costs and transpiration loss, reduce ultraviolet irradiation, and increase soil moisture. Shade is one of nature’s most important facilitative mechanisms. In the Sonoran Desert, the grouping of saguaro (Carnegia gigantean) seedlings under other desert perennials has been studied intensively and coined the “nurse plant” syndrome. Raymond Turner and colleagues (1966) studied several mechanisms with the potential to cause the associations between young saguaro cacti and nurse trees by experimentally transplanting young saguaros in factorial treatments of shade, supplemental water, and protection from herbivores. Turner and colleagues found that that predation was important, but all nonshaded saguaro seedlings died regardless of water addition.
In nature, the trade-offs between the facilitative and competitive effects of shade are complex because many species reach their maximum photosynthetic rates at light levels far below the natural maximum (~2000·μmol·m–2·s–1). These species may benefit from the effects of shade from neighbors without any cost of decreased carbon gain. For example, Arnica cordifolia is a perennial herb that is common in conifer understories in the northern Rocky Mountains. Donald Young and Bill Smith found that a 30% decrease in light on the forest floor during cloudy days in the Medicine Bow Mountains of Wyoming resulted in a 37% increase in carbon gain for Arnica and an 84% reduction in transpiration (Young, 1983). They found that the photosynthetic rates of Arnica remained near saturation even on very cloudy days. In other words, Arnica gained from the lower transpiration rates associated with increased shade without an accompanying cost of lower photosynthetic rates.
Water relations of plants can also be facilitated in many different ways. Facilitators can build up litter, decrease soil bulk density, intercept rain or fog by canopies, snow accumulation, or hydraulic lift. Soil beneath canopies is commonly wetter than that in nearby open areas, and the difference in soil moisture has been correlated with facilitative effects in many systems.
Vegetation is a fundamental driver of soil development, and shrubs and trees add nutrients to the soil in ways that favor some species over others. In the 1950s, J. D. Ovington noted that “whilst the trees cannot alter primary site factors such as bedrock or topography, they may modify some secondary factors. Nutrients are removed from the soil and are returned in part as litter fall so that the trees influence those soil processes which affect the physical and chemical condition of the soil. Ovington also noted that deeply rooted perennials take up nutrients that are unavailable to more shallowly rooted understory plants and deposit them on the soil surface via litter fall and throughfall. Understory plants may eventually acquire these nutrients after the litter fall from the overstory decays.
Wind can damage plants by decreasing tissue temperatures, increasing vapor pressure differences between leaf and air, or simply by damaging plant parts. In environments where wind is extreme, many plants may be facilitated by sheltering beneath or behind other plants. This form of facilitation creates subalpine tree islands and ribbon forests with seedling regeneration restricted to the leeward side of the islands or ribbons. As another example, Pinus flexilis (limber pine) shades other species and protects them from high winds at the ecotone of the Rocky Mountains and the Great Plains.
Other mechanisms that have been shown to play important roles in facilitation include soil oxygenation, substrate building, protection from disturbance, and forms of chemical communication among plants. Most examples of chemical communication among plants involve herbivores and thus are indirect interactions requiring intermediate species, such as herbivores, pollinators, mycorrhizal fungi, soil microbes, or other competing plant species in order to occur.
The seminal paper on indirect defense interactions among plants was published by Peter Attsat and Dennis O’Dowd in 1976. They argued that many plant species were “functionally interdependent with respect to their herbivores.” Soon afterward, Sam McNaughton published a paper demonstrating lower mortality rates of the highly palatable grass Themeda triandra when associated with unpalatable species. Since then many other studies have shown similar processes of indirect facilitation involving herbivores. Other indirect interactions can involve reproductive feedback determined by the density of individuals, enhanced sharing of pollinators or dispersers, feedback involving soil microbial communities, shared mycorrhizal networks, and intriguing indirect interactions among competing plant species.
Competitive interactions between two species can be altered by simultaneous competitive interactions with additional species or through cumulative diffuse effects that occur when numerous species have different kinds of direct effects that act on a single species. In all cases the facilitative effect is produced by something analogous to an alliance—an enemy of my enemy is my friend. This has been understood theoretically for a long time, but the first experiment designed explicitly to quantify indirect effects among interacting plants was conducted by Tom Miller. Miller found that direct and indirect effects were common and strong and that interactions sorted themselves out so direct negative effects among particular species were balanced by positive indirect effects. Strong direct inhibition by dominant competitors was consistently ameliorated by the presence of other competitors. For example, Miller found that Ambrosia artemisiifolia reduced the biomass of Chenopodium album by 94–98% in two-species experiments. In multispecies communities, however, Ambrosia reduced the biomass of Chenopodium by only 17%. Ecologists have tried to understand coexistence and species diversity in the context of niche partitioning, variation in particular resource requirements and uptake, shifts in competitive hierarchies in different microenvironments, and non-equilibrium processes; however, Miller’s results suggest that the balance of competitive interactions may create facilitative effects that are crucial for sustaining coexistence among species in communities.
In the 1950s, ecologists from the Intermountain Research Station in Utah found that herbaceous species were much smaller when grown under Populus tremuloides (quaking aspen) canopies than in open meadows near the trees, indicating that the trees had competitive effects. However, when they also trenched plots to exclude P. tremuloides root systems under canopies, the biomass of some of the herbs was greater than that in the open, demonstrating that strong facilitative and competitive effects were functioning at the same time in their system.
Since the 1950s a great deal of other research has also shown that facilitative and competitive interactions often operate in balance. For example, Quercus douglasii (blue oak) in California facilitates the growth of understory species through nutrients in litter and throughfall but often competes with the same species at the same time through its lateral root system.
Such co-occurring positive and negative interactions were given a strong element of predictability in the early 1990s when Mark Bertness experimentally demonstrated competition among salt marsh plants in relatively moderate abiotic conditions, showed facilitation in abiotically stressful conditions, and then eliminated facilitation by experimentally eliminating the abiotic stress. Bertness’s field studies led to a general conceptual model for the relationship between stress and the relative importance of competition and facilitation, the “stress gradient hypothesis” proposed by Bertness and Callaway in 1994. In part derived from J. P. Grime’s hypotheses about the relative importance of competition in plant communities, they postulated that competitive interactions would be most important to the organization of plant communities when abiotic stress does not strongly limit the ability of plants to acquire and exploit resources. Under relatively benign abiotic conditions that permit rapid resource acquisition, competition can be intense. However, if severe physical conditions restrict resource acquisition, amelioration of severe stress by a neighbor may be more likely to favor growth than competition with the same neighbor is to reduce growth (figure 1).
Strong support for the stress gradient hypothesis comes from many studies, but two groups in particular have conducted experiments in many places and under a wide range of abiotic conditions. Lorena Gómez-Aparicio and colleagues at the University of Granada in Spain conducted the largest-scale study to date of facilitation and competition in semiarid environments using 18,000 replicates of 11 different potential beneficiary species with 18 different species of potential nurse shrubs. The work was carried out over 4 years at many different sites. These results convincingly demonstrated that pioneer shrubs facilitate the establishment of woody, late-successional Mediterranean species and that nurse shrubs had a stronger facilitative effect on seedling survival and growth at low altitudes and sunny, drier slopes than at high altitudes or shady, wetter slopes. The second study was a series of experiments conducted in subalpine and alpine plant communities with 115 species in 11 different mountain ranges around the world (Callaway et al., 2002). Callaway and colleagues found that competition generally dominated interactions at lower elevations where productivity was higher and abiotic conditions are less physically stressful. In contrast, at high elevations where abiotic stress is high, the interactions among plants were predominantly positive.
Figure 1. Conceptual model for variation in the importance of competition and facilitation in communities along gradients of abiotic stress and herbivory. (Redrawn from Bertness and Callaway, 1994)
Not all tests of the stress gradient hypothesis have supported it, but the idea is rooted in an effort to understand in what conditions we might expect to find strong facilitative effects. Facilitation virtually always occurs through the alleviation of some kind of stress experienced by a plant; thus, common sense suggests facilitation would be more common in relatively stressful conditions. Without some kind of stress there is nothing for a neighbor to facilitate.
Facilitation has important implications for several key concepts in ecology. Jon Bruno proposed that incorporating facilitation into ecological theory “will fundamentally challenge some of our most cherished paradigms” and “that current theory emphasizing competition or predation paints an incomplete, and in some cases misleading picture of our understanding of the structure and organization of ecological systems.” Facilitation suggests new perspectives on the realized niche, diversity-community attribute relationships, the role of interactions in evolution, and, as noted at the beginning of the chapter, the nature of communities.
Implicit in the process of facilitation is the idea that the realized niche can be increased by other species. It has been assumed that the performance of a species along a set of relevant environmental variables is sufficient to explain its fundamental niche, and competitive and consumer interactions have been incorporated into definitions of the realized niche. Discrepancies between realized and fundamental niches are virtually always attributed to resource competition. However, research on facilitation clearly demonstrates that both positive and negative interactions must be incorporated into the concept of the niche. Competition and facilitation may act somewhat symmetrically at different margins of a species’ distribution with competition as a limiting factor at one extreme and facilitation an expanding factor at the other extreme.
A substantial amount of research has found a positive relationship between the number of species in a community and aspects of ecosystem function of the community (see chapter III.14). In general, the effects of facilitation have been combined conceptually with niche differentiation within the broad idea of complementarity as possible drivers of the effects of species richness in diversity-ecosystem function theory. This is an effective operational approach because the two processes are quite difficult. For example, legumes facilitate by fertilizing soil though N fixation, but the use of atmospheric N can also be considered a form of niche differentiation. Complementarity describes the divergence in niche space among species in a community that allows for an increase in total utilization of resources, something different from facilitation. Complementarity occurs when neighbors do not substantially infringe on each other’s resource requirements; in other words, there is reduced competition. Facilitation occurs when a species benefits from the presence of a neighbor.
In a substitutive experimental design, the total number of individuals is constant. If interspecific competition is weaker than intraspecific competition because of niche differentiation, individuals on average will experience weaker competitive effects and should perform better. On the other hand, superior performance of individuals in species-rich mixtures compared to monocultures suggests the possibility of facilitation. As in many other studies, Maria Caldeira and colleagues found that productivity in species-rich plots was significantly higher than that in monocultures. However, species performed better as a community than as individuals. Caldeira found that some individuals performed better in mixtures. Plants were sown into Caldeira’s plots, and thus the experiment was not clearly substitutive, but measurements of soil moisture and stable carbon isotope ratios in the leaves of several species indicated that plants growing in species-rich mixtures improved their water relations, which suggests that either complementarity or facilitation may have been responsible for the increase in productivity with diversity. Future experiments that separate these different conceptual mechanisms may allow a better understanding of the role of diversity on community and ecosystem functioning.
Indirect facilitation may also affect diversity-ecosystem function relationships. Johannes Knops and colleagues quantified foliar disease severity in plots varying in species richness. They found that for each of the four target plant species, foliar disease was significantly negatively correlated with plant species richness. However, they found that disease severity was also dependent on host plant density. Although not yet investigated in the context of diversity-ecosystem function, negative feedback that has been documented between plants and soil microbes could also enhance community diversity by increasing species turnover rates. As the number of species increases, turnover among species in a particular place may be much greater. Alternatively, high species richness may reduce strong negative feedbacks in general.
Thymus vulgaris is one of the most widespread plant species in Europe, and the species is composed of several different chemotypes that differ in the biochemical composition of the essential oils produced in the leaves. Bodil Ehlers and John Thompson found that six different Thymus chemotypes could be identified using the dominant monoterpene in the essential oil, which is either phenolic or nonphenolic. The grass, Bromus erectus, is often spatially associated with all of the different Thymus chemotypes, suggesting facilitation. Ehlers and Thompson found that B. erectus from nonphenolic Thymus patches performed significantly better on its home soil than on soil from a different nonphenolic or phenolic Thymus patch. This superior performance of matched local B. erectus to familiar Thymus chemotypes was observed only for soil collected directly underneath Thymus plants and not on soil collected away from Thymus. These results suggest that B. erectus may be genetically adapting to soil modifications mediated by different Thymus chemotypes, and importantly, this may occur only because Thymus facilitates the growth of B. erectus in the chemically modified environment.
The scenario described by Ehlers and Thompson is unique in the ecological literature for plants, but it describes a reasonable way that facilitation may drive evolutionary changes. By pulling other species into an expanded niche (figure 2), benefactors may expose beneficiaries to new abiotic or biotic environments to which they may adapt.
In a similar example, Manuel Figueroa and colleagues examined facilitation and evolution in the Odiel Marshes in southwest Spain. They found that a Sarcocornia hybrid grows on the raised centers of Spartina martima patches. These patches are invaded by Sarcocornia perennis, a species common to lower parts of the marsh; however, once established in Spartina patches, an opportunity is provided for hybridization with Sarcocornia fruticosa, a species that occurs in higher parts of the marsh. Hybrids occur only on Spartina patches with S. perennis. Figueroa called this scenario with the hybrid Sarcocornia genetic facilitation and suggested that succession might be facilitated genetically through the establishment of conditions leading to hybridization.
Figure 2. Illustration of how facilitation might create novel evolutionary opportunities for beneficiaries (e.g., Ehlers, B. K., and J. Thompson. 2004. Do co-occurring plant species adapt to one another? The response of Bromus erectus to the presence of different Thymus vulgaris chemotypes. Oecologia 141: 511–518.) Peak A represents the niche of a beneficiary species (e.g., Bromus erectus) in the absence of a benefactor species (e.g., Thymus vulgaris). Peak B represents the niche of the benefactor species. Peak C represents the realized niche of the beneficiary species in the presence of the benefactor, with the shaded area representing the new environment to which the beneficiary can now evolve. (Reprinted from Callaway, 2007)
Recently, Alfonso Valiente-Banuet and colleagues explored the facilitative effects of plant taxa that evolved during the drying climate of the more recent Quaternary (within the last 2 million years) period on more ancient plant taxa that evolved during the wetter Tertiary (~60 million years ago). Most global deserts and semiarid environments developed during the Quaternary. The development of desert corresponded with the evolution of new species, but, interestingly, many mesic-adapted Tertiary species did not become extinct in the drier climate. Valiente-Banuet found that “modern” species, those that arose during the Quaternary, currently facilitate ancient Tertiary species. In fact, very few ancient species recruited in any microhabitat other than beneath the canopies of other species. In other words, species that rapidly evolved to new stressful abiotic conditions appeared to be pulling ancient species, which were not adapted to xeric conditions, into modern communities by creating appropriate regeneration niches. These results have profound implications for the processes that sustain global biodiversity and for the nature of plant communities. Communities such as these are clearly not individualistic.
As discussed at the beginning of this chapter, most ecologists probably do not perceive plant communities as fully individualistic. However, the presentation of communities as individualistic is almost the rule in general and specialized textbooks. Moreover, this classic and often artificial historical dichotomy in viewpoints is one with lingering impacts on the way we think and conduct research.
There are many arguments for the individualistic paradigm, but descriptions of continuous distributions of species along environmental gradients have been a central component of the argument for individualistic communities since the idea was first articulated by Robert Whittaker in 1951. Most gradient analyses show a continuum of apparently randomly overlapping species, and this is used to argue for individualistic communities. If species depended on each other would they not always occur together? However, it is now clear that species can facilitate each other in some conditions but compete with each other in different conditions. Furthermore, most gradient analyses do not quantify spatial relationships at a scale appropriate to detect positive associations, and some gradient analyses are at odds with the continuum. For example, the presence of Prosopis velutina on desert and terrace landforms and Olneya tesota throughout the Sonoran Desert is strongly associated with particular understory communities. The distributions of species on these gradients and many others are not continuous but grouped into nodes, and these nodes suggest facilitation and some degree of interdependence among species.
Robert Whittaker died in 1979, but in a story with an ironic twist, he posthumously published a paper in 1981 in which virtually perfect correlations occur between understory communities and the presence of different desert or chaparral shrubs—the nodal distribution of species along gradients indicative of holistic communities (figure 3). It is hard to imagine tighter correlations among the distributions of different plant species along gradients, yet even with these “strongly differentiated patterns,” and despite the lack of evidence for a continuum in their results, there was no discussion of how these findings might be reconciled with Whittaker’s paradigm; nor was there any discussion of rejecting the continuum as universal in plant communities.
Chris Lortie argued that recent experimental efforts to understand the relative importance of positive or negative interactions in many different communities allows ecologists “to explicitly reconsider what most ecologists appear to have done implicitly: our formal conceptual theory of the fundamental nature of communities.” Lortie also proposed a conceptual model for the nature of plant communities termed the “integrated community concept.” The integrated community concept is based on evidence that community composition is determined by (1) stochastic processes, (2) species-specific tolerances to local abiotic conditions, (3) positive and negative direct and indirect interactions among plants, and (4) direct interactions with other organisms. If communities are determined by complex interactions among all of these processes, including facilitation, communities should vary from those that act very much as collections of independent individual species to others that act as highly interdependent groups of species. As stated by Lortie, “communities (and even a single community) will encompass a range of different dependencies among species—or degrees of integration—determined by the relative importance, and variation in space and time, of each of the processes we proposed.”
Figure 3. Stylized representation of graphic results presented by Schmida and Whittaker demonstrating strong nodality among species in plant communities. In the chaparral, high ordination scores for understory herbs correlate strongly with the presence of shrubs, and in the Mojave Desert high ordination scores correlate strongly with the presence of Larrea or Ambrosia. (Reprinted from Callaway, 2007)
The integrated community concept, incorporating facilitation, offers a different and more mechanistically inclusive understanding of the organization of plant communities than the individualistic paradigm. Community composition, biological diversity, and coexistence are determined not only by competitive and consumer interactions but also by powerful and ubiquitous facilitative effects. Because facilitative effects suggest a substantial component of interdependence in plant communities, communities cannot be understood only by studying populations.
Archer, S., C. Scifres, and C. R. Bassham. 1988. Autogenic succession in a subtropical savanna: Conversion of grassland to thorn woodland. Ecological Monographs 58: 111–127.
Atsatt, P. R., and D. O’Dowd. 1976. Plant defense guilds. Science 193: 24–29.
Bertness, M. D., and R. M. Callaway. 1994. Positive interactions in communities. Trends in Ecology and Evolution 9: 191–193.
Brooker, R., Z. Kikvidze, F. I. Pugnaire, R. M. Callaway, P. Choler, C. Lortie, and M. Michalet. 2005. The importance of importance. Oikos 109: 63–70.
Bruno, J. F., J. J. Stachowitcz, and M. E. Bertness. 2003. Inclusion of facilitation into general ecological theory. Trends in Ecology and Evolution 18: 119–125.
Callaway, R. M. 2007. Positive Interactions and Interdependence in Plant Communities. Dordrecht: Springer.
Holmgren, M., M. Scheffer, and M. A. Huston. 1997. The interplay of facilitation and competition in plant communities. Ecology 78: 1966–1975.
Miller, T. E. 1994. Direct and indirect species interactions in an early old-field plant community. American Naturalist 143: 1007–1025.
Turner, R. M., S. M. Alcorn, G. Olin, and J. A. Booth. 1966. The influence of shade, soil, and water on saguaro seedling establishment. Botanical Gazette 127: 95–102.
Valiente-Banuet, A., A. V. Rumebe, M. Verdú, and R. M. Callaway. 2006. Modern Quaternary plant lineages promote diversity through facilitation of ancient Tertiary lineages. Proceedings of the National Academy of Sciences U.S.A. 103: 16812–16817.
