Species diversity may affect ecosystem function, and different species may have disproportional influences upon ecosystem processes. To start off with, the null hypothesis for the relationship between biodiversity and ecosystem processes would state: Species diversity does not affect ecosystem function. Given the varied plant assemblages and ecosystems on Earth, however, diversity probably influences processes. Several alternative hypotheses have been proposed in order to elucidate the extent to which ecosystem function depends on diversity. Depending on the interpretation followed, the redundancy hypothesis implies that under existing conditions species richness is irrelevant or that a minimum level of diversity is necessary for proper system functioning, but most species overlap in their functional roles. A different view asserts that each and every species contributes to ecosystem functioning. Both the “Type 1” linear relationship and the “rivet” hypothesis are based on this idea of unique species contributions. The order of species deletions or additions does not matter in any of these hypotheses.
According to the Global Biodiversity Assessment (GBA), "There is no evidence that each and every species plays a unique role such that its absence would immediately result in a dramatic change in the functioning of ecological system". Moreover, in most ecosystems, species diversity is higher than the level required for efficient biogeochemical and trophic function. Also, some researchers assert that richness is not important per se, but a higher level of richness increases the likelihood of including a productive species. Scientists acknowledge that the relationship between diversity and function most likely sits between the two extreme hypotheses listed above. Instead, a few species probably dominate processes and species richness does not matter past a certain threshold. This differs from the redundancy hypothesis in that the order of deletions or additions matters. The idiosyncratic response hypothesis summarizes the relationship: ecosystem function changes when diversity changes, but the magnitude and direction are unpredictable since the roles of individuals are complex and varied. The remainder of this article will consider the concepts invoked for the differential roles species may have in ecosystems. The focus will remain on autotrophs.
Species can individually affect ecosystem processes through unique traits. Examples of these unique traits include nitrogen (N) fixation, water redistribution, trace gas emission, high growth rates or unique secondary chemistry. A few species usually have values for traits quite different from the rest. These types of effects are often studied in the context of invasive species but can also occur in non-invasive situations. These functions may be predictable and may have profound effects if the species is added or deleted.
One framework for understanding species traits is that species modify available resources either through consumption or supply. The resource supply rate can also be affected by species. In particular, differences in tissue quality, which control litter decomposition, occur. Other environmental factors can be affected, including soil acidity and microenvironment (through evapotranspiration and insulation). These, of course, also will affect community processes. For example, in Asia Nepalese alder increases nitrogen (N) inputs while bamboo retains weathered potassium (K). Two processes affected by species traits- water redistribution and nutrient cycling- will be explored further below.
Hydraulic lift denotes the phenomenon of water redistribution from deep, moist soil horizons to dry, shallow layers. Besides aiding the "hydraulic lifter", water can also be provided to plants in the vicinity of the hydraulic lifter. The process involves passive movement of water from roots to soil water when the soil water potential is less than the xylem water potential (or, when one part of the soil has lower soil water potential than another part). One study found that groundwater lifted by sugar maple (Acer saccharum) was used by plants up to 2.5 meters (m) away from the tree, but no effect was seen further than 5 m from the tree. All neighbors used some fraction of the lifted water (3-60%), and differential plant effects were noted. Since plant water deficit limits production, the existence of a hydraulic lifter in a community may be a crucial presence. This also may affect biogeochemical conditions for helping mineral availability, nutrient acquisition, or microbial processes. To date, the phenomenon has been demonstrated in about 27 species of different life forms– grasses, shrubs, trees, and herbs– and scientists expect it to be more widespread.
Inverse hydraulic lift has also been demonstrated. Hydraulic lift may provide a significant amount of water needed for evapotranspiration and so contribute to this process at an ecosystem scale. On the other hand, demonstration of inverse lift may prevent neighbors/competitors from gaining access to water. These coarse-scale effects, in addition to the facilitation of neighboring plants, make hydraulic lift a community- and ecosystem-level process. It also may be a general process, as additional studies of different taxa and ecosystems indicate.
In general, individual plant species can play an important role in soil fertility. Usually they create positive feedbacks to plant persistence. Plants from low-nutrient systems use nutrients efficiently and grow slowly, so the cycling of nutrients is slow; the opposite is true for high-nutrient sites. Sometimes the vegetation is more important than abiotic effects in soil process control. Species traits elucidate whether a species loss or addition will affect nutrient cycling. Plants affect the nitrogen cycle through litter quality and its subsequent influence on transformation rates. Competitive displacement may occur after a change in nitrogen (N) supply due to deposition from anthropogenic sources. Community composition in moist meadow tundra was shown to have a potential effect on the nutrient cycling of the community. Deschampsia caespitosa had a positive influence on the rates of net soil N transformations, whereas Acomastylis rossii did not respond to an increased N supply and had a negative influence on the cycle. Therefore, the replacement of Acomastylis rossii by Deschampsia caespitosa had positive feedbacks, and the litter could reinforce the patchiness of species composition and available N.
Particular species do not always affect processes, though. In semi-arid grassland, researchers found that plant presence, rather than plant species identification, probably had the most effect on ecosystem processes. They do acknowledge that specific plant characteristics, such as life span, biomass allocation, and tissue chemical allocation can affect nutrient dynamics and soil organic matter. The traits that best predict resource consumption are height (as related to an individual’s ability to capture light and exploit a large soil volume) or biomass per individual and relative growth rate (as related to individual ability to capture carbon and nutrients). These traits vary continuously among organisms.
Some species dominate ecosystem functioning by virtue of their abundance in a system. The most abundant species in a system usually have the most influence in terms of productivity, transpiration, decomposition, and nutrient cycling. While many rare species may contribute to higher species richness, they might not contribute much to ecosystem function. Accordingly, the deletion of a dominant species would have a greater impact on processes than the deletion of a rare one. The mass ratio hypothesis derives from theoretical and experimental evidence that shows the extent to which a plant species affects ecosystem functions is likely to be predictable from its contribution to the total plant biomass.
Accordingly, the four standard models explained at the beginning of this article only work if the species are deleted in order of rarest to most abundant. If the most abundant species are deleted first, then the response of ecosystem function is wholly different. Although dominance is identified by a measure of biomass, the impact of dominance in a system is not due to biomass itself. Instead, the dominant species affects the ecosystem process most by being best adapted to the environmental conditions present in the system. Deleting a dominant species that comprises 40% of the biomass of a system would have a greater impact upon the system than taking out 40% of the biomass distributed proportionately among all the species present in the system. Since order of removal matters, the relationship between diversity and ecosystem process could be modeled many ways, depending on the order in which species are subtracted or deleted. This also supports criticism of experimental work in this field: species richness affects ecosystem properties only insofar as it increases the likelihood that a dominant species will be included in the assemblage.
Given that dominants dominate processes, it is important to figure out whether less common species play any role in ecosystem function. Although “there is no a priori reason to suspect that such minor contributors must influence ecosystem functioning", Grime proposes considering how subordinates and transients affect ecosystem processes. Species are classified as belonging to one of three groups: dominant, subordinate, or transient. Dominants occupy a high proportion of biomass as well as usually have an expansive morphology and stature, and number only a few individuals. Dominants also occur in particular vegetation types. Subordinates usually co-occur with dominants, comprise less biomass, have smaller stature, and are more numerous. Transients, usually ignored by ecologists, have a heterogeneous distribution and no particular association. They vary in number and functional traits, make a small contribution to biomass and have few individuals. They usually exist only as seedlings and juveniles. Ecosystem processes may be controlled by one dominant. Alternatively, there may be functional differences between “co-dominants”. Ecosystem processes may benefit from this redundancy and complementarity. The association between dominants and subordinates might represent a fuller exploitation of resources; this may be a relationship of complementarity. Overall, changes in subordinates and transients will have less effect on ecosystem properties.
It is important to note that the response of a system after perturbation to species richness will depend on the rate of the process and the time at which response is studied. The longer a response to a deletion is measured, the lesser the magnitude will be due to the compensatory response by other species.
Whereas the mass ratio hypothesis tackles immediate effects on processes, some effects become apparent on longer time scales. This may be especially true of less abundant species. Grime terms longer-term effects “filter and founder effects”. Often following disturbance, subordinates of the previous community establish first. Therefore, subordinates can influence the regeneration of dominants after disturbance events, thereby controlling the identity, functional diversity, and relative abundance of dominants. Transients may indicate the effectiveness with which potential dominants are dispersed across the landscape into suitable ecosystems. A high presence of transients may also indicate low competition and a rich assortment of colonizers. This could allow the ingress of different plant functional types, some of which are capable of exploiting the new conditions created by a disturbance. So, whereas the mass ratio hypothesis argues against the importance of species richness for ecosystem function, the filter and founder effects argue for a need for diversity in a disturbance-laden world. In addition, species with a small effect on ecosystem processes can have large effect if they affect the abundance of other species with large ecosystem effect (e.g. a seed disperser or pollinator might be essential for persistence of canopy species with greater ecosystem impacts).
The gain or loss of one or a few certain species sometimes seem to have amplifying effects on both community and ecosystem processes. Those species with this effect, but unpredictable traits, are termed “keystone species”. First coined by Paine in 1966 to explain the trophic dynamics in a rocky intertidal pool, the concept has been extended to many relationships. The removal of the keystone species causes major changes in the rest of the community dynamics and ecosystem processes; usually, the species affected are functionally related. Keystones are distinguished from dominants in that they have a greater effect than would be expected given their abundance in the ecosystem. Typically this is still applied to trophic dynamics (e.g. the classic example of the cascade of effects after the removal of the sea otter from its environs), but it has also been applied to autotrophic situations. Since keystones do not have a set of specific traits but are defined upon their impacts, they are considered to be unpredictable.
Keystones have been organized into five categories: predators, herbivores, competitors, pathogens, and mutualists. An example of a keystone plant species is a dominant tree outcompeting others during succession. Dominants, though, can be excluded by “keystone weeds” that suppress the establishment of seedlings. Plants are also described as keystone mutualists, which provide critical support to pollinators and dispersers. Keystones can be a relative concept: a keystone from one area may not play such a crucial role in another due to the complexity of associations that occur.
Ecosystem engineers are defined as species which directly or indirectly modulate the availability of resources (apart from themselves) to other species. Ecosystem engineers cause physical state changes in biotic or abiotic materials and thereby modify, maintain and/or create habitats and resources for other organisms. To some degree, the other organisms which make use of the changes are dependent upon the modulated structure. Autogenic engineers modify the ecosystem via their own physical structures (e.g. living and dead tissue of a tree). Allogenic engineers transform living or nonliving materials from one state to another. These activities differ from the direct provision of resources an organism may provide through living or dead tissue.
The supporters of the ecosystem engineer concept assert that every system is affected by engineering. An obvious example of an allogenic engineer is the beaver and its impact on riparian systems through the construction of dams. Earthworms have also been cited for moving soil, making casts, and burying seeds. Trees and coral reefs represent autogenic engineers not because of the resources they provide through their tissues, but rather due to the effects their physical structures have on modulating the environment, such as altering hydrology, nutrient cycles, soil stability, and microclimate. According to the theory, early successional stages are “plant-engineered environments". Proponents of this theory incorporate other well-known processes. For example, the kelp which is disrupted by the loss of sea otters in the keystone example is considered itself to be an autogenic engineer. According to proponents of the ecosystem engineer story, the example should be retold: otters change the effect of an allogenic engineer – urchins – upon an autogenic engineer – kelp beds; the change in kelp bed structure causes the rest of the cascade of effects. The impact of engineers is scaled by: lifetime per capita activity of individual organisms; population density; spatial distribution of population (local and regional); the length of time the population is present on a site; the durability of the engineer's constructs; artifacts in absence of the engineer; the number and type of resource flows modulated by the engineer; and the number of other species dependent upon these flows.
The link between species diversity and ecosystem function seem to be explained by components of all of the hypotheses and conceptual models; none of them can be considered alone. Most likely, the critical components of diversity are the approximate number of species in the community; their relative abundance; how strongly a given species differs from other species in the community; the impact of particular species traits on community and ecosystem processes; and indirect effects that a species has on other species in the community. Accordingly, as Mooney notes, “the gain or loss of a species will have its greatest impact on ecosystem processes when there are few species in the community, when the species gained or lost is a dominant species, and/or when the species differs strongly from other species in the community”. The continuous distribution of traits means that species differ quantitatively rather than qualitatively in their effects on processes. The source of the change in biodiversity may also be important; grazing may affect a system differently than successional processes. According to the Global Biodiversity Assessment (GBA), “An understanding of unique species' traits, overlap among species, and the possible functional significance of low or high numbers of species, apart from how they differ in traits, is clearly immediately relevant to understanding the conditions under which ‘species matter’”.
- Aarssen, L. W. 1997. High productivity in grassland ecosystems: effected by species diversity or productive species? Oikos 80:183-184.
- Bond, W. J. 1994. Keystone Species. P. 237-253 In: Biodiversity and Ecosystem Function. E. D. Schulze and H. A. Mooney, eds. New York: Springer Verlag. ISBN: 0387581030
- Burgess, S. S. O., M. A. Adams, N. C. Turner, C. K. Ong. 1998. The redistribution of soil water by tree root systems. Oecologia 115:306-311.
- Caldwell, M. M, T. E. Dawson, J. H. Richards. 1998. Hydraulic lift: consequences of water efflux from the roots of plants. Oecologia 113: 151-161.
- Dawson, T. E. 1993. Hydraulic lift and water use by plants: implications for water balance, performance and plant-plant interactions. Oecologia 95:565-574.
- Grime, J. P. 1997. Biodiversity and Ecosystem Function: The debate deepens. Science 277:1260-1261.
- Grime, J. P. 1998. Benefits of plant diversity to ecosystems: immediate, filter, and founder effects. Journal of Ecology 86:902-910.
- Hobbie, S. E. 1992. Effects of Plant Species on nutrient cycling. Trends in Ecology and Evolution 336-339.
- Holling, C. S. 1992. Cross-scale morphology, geometry, and dynamics of ecosystems. Ecological Monographs 62:447-502.
- Jones, C. G., J. H. Lawton, and M. Shachak. 1994. Organisms as ecosystem engineers. Oikos 69:373-386.
- Jones, C. G., J. H. Lawton, and M. Schachak. 1997. Positive and negative effects of organisms as physical ecosystem engineers. Ecology 78:1946-1957.
- Lawton, J. H. 1994. What do species do in ecosystems? Oikos 71:367-374.
- Mooney, H. A., J. H. Cushman, E. Medina, O. E. Sala, E. D. Schulze, eds. 1996. Functional Roles of Biodiversity: A Global Perspective. New York: John Wiliey and Sons. ISBN: 0471956015
- Mooney, H. A., J. Lunbchenco, R. Dirzo, O. E. Sala, eds. 1995. Biodiversity and Ecosystem Functioning: Basic Principles. In: Global Biodiversity Assessment. V. H. Heywood, ed. New York: Cambridge University Press. ISBN: 0521564816
- Power, M. E. and L. S. Mills. 1995. The Keystone cops meet in Hilo. Trends in Ecology and Evolution 10:182-184.
- Sala, O. E., W. K. Lauenroth, S. J. McNaughton, G. Rusch, and Xinshi Zhang. 1996. Biodiversity and Ecosystem Functioning in Grasslands. p. 129-149 In: Functional Roles of Biodiversity: A Global Perspective. H. A. Mooney, J. H. Cushman, E. Medina, O. E. Sala, E. D. Schulze, eds. New York: John Wiley and Sons. ISBN: 0471956015
- Solbrig, O. T. 1992. The IUBS-SCOPE-UNESCO program of research in biodiversity. Ecological Applications 2:131-138.
- Steltzer, H. and W. D. Bowman. 1998. Differential influence of plant species on soil nitrogen transformations within moist meadow alpine tundra. Ecosystems 1:464-474.
- Tilman, D., J. Knops, D. Wedin, P. Reich, M. Ritchie, and E. Siemann. 1997. The influence of functional diversity and composition on ecosystem processes. Science 277:1300-1302.
- Vinton, M. A. and I. C. Burke. 1995. Interactions between individual plant species and soil nutrient status in shortgrass steppe. Ecology 76:1116-1133.
- Wardle, D. A., O. Zacriksson, G. Hornberg, C. Gallet. 1997. The influence of island area on ecosystem properties. Science 277:1296-1299.