Effects of Global Changes on Above- and Belowground Biodiversity in Terrestrial Ecosystems: Implications for Ecosystem Functioning

Abstract

Above- and belowground organisms are critical for the biogeochemical cycles that sustain the Earth, but there is limited knowledge on the extent to which the biota below ground and the functions they perform are dependent on the biota above ground, and vice versa. Hooper et al. (2000) provide a synthesis of the patterns and mechanisms linking above- and belowground biodiversity. The close relationship between vegetation change and soil carbon (C) dynamics (Jobbágy and Jackson 2000) suggests that any disruption of the coupling between plants and soil organisms as a result of global change may have deleterious consequences for functioning of terrestrial ecosystems. However, most of the scientific evidence supporting this hypothesis comes from correlative approaches. The complexity of the numerous interactions between various environmental factors controlling both plants and soil organisms has hampered formal tests of causation. Moreover, the agents of global change (such as changes in land use, climate, and atmospheric inputs), acting either individually or in combination, may lead to nonlinear changes in above- and belowground relationships, with effects varying in distribution and intensity around the globe (Heal 1997). This places serious constraints on predictions regarding the effects of global change on above- and belowground interactions. Interpretation of future global change scenarios and the consequences to ecosystems globally depends on understanding the primary factors affecting above- and belowground interactions (Ingram and Gregory 1996). In this article, we synthesize the current knowledge of the potential effects of global change on terrestrial ecosystem processes via changes in above- and belowground biodiversity. We identify the types of interactions between vascular plants and soil biota, describe how each interaction type is sensitive to changes in species composition, apply this framework to evaluate the potential consequences of different global change drivers on ecosystem processes via changes in species composition, and suggest scientific priorities. Plants and soil biota influence each other as engineers, as providers of nutritional resources, and as direct interactors (Table 1; see also Hooper et al. 2000). The engineers are organisms that modulate the flux of resources to other organisms through physical modification of the environment, through their physical presence, or through catalytic activities (Jones et al. 1994, Anderson 1995). Providers of nutritional resources modify the availability of mineral nutrients or organic molecules to other organisms by their metabolic activities. Direct interactors are organisms that affect each other in mutualistic or antagonistic relationships. The three interaction types are not mutually exclusive. Plant transfer of nutrients to symbionts, and vice versa, is a direct interaction occurring through the provision of nutritional resources. Mycorrhizal fungi increase the transfer of phosphorus (P) and other nutrients to plants and may also affect soil aggregation (Douglas 1995). The leaf litter produced within a stand controls the availability of energy and nutrients to the soil biota and also affects the protection of soil from erosion (Swift and Anderson 1993). Similarly, engineers such as earthworms influence not only soil structure, but are linked as well to plant nutrient availability through changes in soil microflora (Edwards and Bohlen 1996). Chapin et al. (1996) suggested that a system's tolerance for change is related to the degree of continuity of biological traits among species. Changes in species with traits that are broadly distributed across the community should be minimally recognized at the ecosystem level because more resistant species may compensate for such changes (Schimel 1995). By contrast, changes affecting species with unique traits cannot be compensated for by other species. Invasion or extinction of these species is predicted to be maximized at the level of an ecosystem process if the availability of soil resources to plants is altered, food web interactions are fundamentally changed, or the normal disturbance regime of the ecosystem is strongly modified (Vitousek 1990). This concept serves as a starting point to examine how linked interactions between above- and belowground biota and accompanying ecosystem processes may be sensitive to environmental change. Invasion or extinction of freeliving and mutualistic soil biota with unique traits may alter ecosystem processes by modifying the availability of soil resources to plants (Figure 1). Changes in the composition of the vegetation involving species with unique traits may have large-scale impacts by affecting the utilization of resources by free-living and mutualistic soil biota. This occurs through alterations in the amount, quality and distribution of litter and exudates produced. Finally, changes in engineering activities or in antagonistic effects may alter ecosystem processes by modifying the disturbance regime experienced by those organisms on one side of the above-and belowground interface. Soil organisms appear to have a high level of functional redundancy (Andrén et al. 1995). This suggests that traits involving the transfer of nutrients through the soil food web are broadly distributed across the belowground community. Changes in the taxonomic diversity of the soil food web would thus have little impact on ecosystem processes because there would be little change in the availability of soil resources to plants (Wolters 1998). Functions carried out by species with unique traits, however, are an exception. The most notable example is provided by the few genera of bacteria performing certain transformations within the nitrogen (N) cycle (e.g., nitrification, denitrification; Swift et al. 1998, Hooper 2000). Changes in the species composition of this component of the belowground community may significantly alter the N transfer to plants. Similarly, the ability to degrade recalcitrant organic substrates such as lignin or humic substances is confined to a few microbial genera (Wolters 2000). A greater buildup of litter on the soil surface due to the loss of organisms able to break down recalcitrant materials may drastically alter the availability of nutrients to plants (Schimel and Gulledge 1998). Mutualistic soil organisms may display traits related to the supply of soil resources to plants that are unique to a few species or conditions. Alternately, other mutualistic soil organisms are broadly distributed among the community and/or expressed under a variety of environmental conditions (Chanaway et al. 1991). For example, associations between plants and N-fixers depend on a few microbial taxa, whereas mycorrhizal fungi have high diversity (Swift et al. 1998). Changes in species composition are thus expected to alter the availability of soil resources to plants in the former but not the latter case. This conclusion contradicts the recent finding that the coexistence of plants can be determined by the diversity of arbuscular mycorrhizae (Van der Heijden et al. 1998). A probable resolution of this paradox is that there may be different effects upon plant performance depending on the number of fungal genotypes that are present; that is, there is no fungal species specificity but there may be functional specificity (Read 1998). Litter diversity per se does not seem to have a predictable effect on decomposition processes (Wardle et al. 1997). Moderate changes in species richness of the vegetation are thus unlikely to have an ecologically important role unless the modifications are associated with distinct changes in litter quality. Strong effects on ecosystem processes are most likely when plant functional types differing in ecological strategies invade or become extinct, because ecological strategies of plants and physicochemical properties of the litter produced are closely related (Hobbie 1992). For example, unpalatable litter produced by the stress-tolerant species Vaccinium myrtillus during maturation of spruce forests shifted the humus type from mull to mor by eliminating soil-dwelling organisms from the surface layer (Bernier et al. 1993). The impoverishment of the decomposer community reduced the N availability to the vegetation, which in turn increased the competitive ability of stress tolerant plant species adapted to acidic infertile soils (Van Breemen and Finzi 1998). Most plant symbionts have a broad to intermediate host range, especially in natural vegetation (Douglas 1995). Variations in the composition of plant communities thus should rarely affect the utilization of resources by symbiotic soil biota. This conclusion is not valid, however, when plant symbionts do not function equally with all hosts, or on occasions when associations are host specific. For example, coniferous forests in North America and eucalyptus forests in Australia support large guilds of highly specific ectomycorrhizal fungi (Molina et al. 1992). When such close associations exist, traits controlling consumption rates of nutritional resources by mutualistic soil organisms are unique to a few plant species. Shifts in composition, resulting in deletions or additions of new plant species, of the vegetation may have significant effects on ecosystem processes including changes in P or in C:N ratios. Changes in soil structure may dramatically affect plants by altering the physical properties that control plant growth, i.e., resistance to root penetration, water availability, and aeration (Angers and Caron 1998). However, the ability to stabilize soil structure is broadly distributed across the microbial community (see Table 1; Ladd et al. 1996). Changes in the composition of the soil microflora may have little effect on plants because there would be few changes in soil structure stabilization. In sharp contrast, biological formation of soil macrostructure is limited to the physical artifacts created by a comparatively small number of soil-moving macroinvertebrates (earthworms, termites, ants, etc.) and mammals (prairie dogs, gophers, moles, etc.). Invasion or extinction of burrowing macroinvertebrates and mammals is thus expected to have strong effects on plants in systems where soil engineers play a role (Dangerfield et al. 1998, Anderson 2000). Rhizophagous animals and soil-borne pathogens can strongly modify the disturbance regime to which the vegetation is subjected (Vitousek 1990). Root feeding by invertebrates decreases flowering and seed production, alters patterns of seedling recruitment, reduces plant species richness, and modifies the competitive balance between different plant life-history groupings (Brown 1990). However, alterations in species richness of rhizophages are unlikely to have a strong effect on the disturbance regime encountered by the natural vegetation, because root-feeding animals are comparatively diverse and are usually generalist feeders. By contrast, associations between plant pathogenic nematodes, fungi, and viruses may be highly species specific. For example, Mills and Bever (1998) reported that even co-occurring plant species show differential susceptibilities to the soil pathogens of their neighbors. Because soil pathogens may influence interspecific plant competition, they may also influence plant community dynamics and possibly plant species diversity as well (Van der Putten and Peters 1997). Alteration of the disturbance regime to plants by soil-borne pathogens should therefore have a strong effect on ecosystem processes. Changes in engineering activities of plants can have a dramatic effect on the disturbance regime to soil organisms, i.e., by altering the fire regime, by triggering biological invasions, or by altering structural features of the soil habitat. Here we confine the discussion to the alteration of structural features. Plant engineering of the soil habitat results primarily from individual and plant community attributes such as root architecture and vegetation cover (Table 1). Alterations to the disturbance regime for soil biota through changes in plant engineering should be confined to situations in which plant life forms (trees, shrubs, grasses) differing from the original vegetation become extinct or dramatically change in dominance. For example, when invade they modify processes such as the of soil materials by erosion and the of water and 1998). most soil et al. and ecosystem functions become under shrubs, the become of Invasion by to in the disturbance regime of the soil habitat and changes the distribution of soil resources from a in to a distribution in the of and soils flux of and a surface may to global and 1998). changes in plant community structure alter both engineering and provision of food resources to soil biota. This is because plant species in a variety of attributes (e.g., and that alter soil physicochemical of soil organic and nutrient This is by of plant effects on the of et al. (1998) the of and from soils of communities by or a of the and plants in of growth, water and and N The of and affect all of these because microbial processes to the of are by interactions between soil and soil and the availability of However, in the effects of and plants on at a not on et al. (1998) under certain environmental when factors such as and are not plant composition can belowground processes that modify This that the mechanisms strongly depend on and conditions. We suggest that plant to soil biota ground to below can be as a of control with attributes of life forms engineering activities at the ecosystem physicochemical properties of plant functional modifying the provision of nutritional resources at the community and biological properties of individual species controlling direct interactions at the level (Figure We that plant effects on soil biota at the depend on traits and are thus likely to be to changes at the plant species Changes in ecosystem functions created by alterations in the disturbance regime and consumption rates soil organisms should thus be confined to situations where traits of the vegetation are drastically a change is most likely when the of environmental change all other factors controlling plant structure, when plants with invade or become extinct, and when are these even a species may be the of a vegetation and 1992). A control in the soil with attributes of burrowing macroinvertebrates and mammals different for soil at the ecosystem physicochemical properties of affecting the vegetation at the community and biological properties of individual species direct interactions at the level (Figure Because plants more of the in most land terrestrial food and control of the affecting soil biota, the impacts of global change on belowground organisms are likely to be through Alterations in species composition of the belowground community may significantly environmental changes to ecosystem processes when the disturbance regime is by changes in the species composition of or specific and when the supply of soil resources is modified either by or by organisms performing specific transformations within decomposition processes and nutrient around the globe and from to various forms of and of primary habitat and changes in land and influence above- and belowground organisms by altering three of physical and biological all other factors controlling plant community of land is thus the change in of the in the (Ingram and Gregory et al. 2000). to a few plants alters the attributes plant engineering activities. alterations in and soil conditions strongly affect the structure and function of the belowground community and 1993). Moreover, the decomposer habitat is through alterations in the and of plant of and also dramatically alters provision of food resources to soil biota because the and the quality of litter and exudates produced by plants from those produced by the a functional of the soil are and and species adapted to are (Swift and Anderson 1993). The of soil through the results of change on ecosystem processes by altering the disturbance regime to plants. The for changes in the supply of soil resources to plants from alterations in the metabolic and effects of soil of an and by for example, the species and of the The of the soil surface in a to the formation of a surface with of soil in as and conditions in the soil the The that of carbon per in three as and through to the potential of to atmospheric change via alterations of above- and belowground et al. also direct interactions. plants for for reduces the of (Van et al. 1998). This effect is by the of natural and a the disturbance regime and the provision of nutritional resources are The for control and and functional between plants and soil biota are change and alters all of above- and belowground interactions and on a large global increase of the is predicted as a of atmospheric of 1996). not only in but also in and the of are The close of vegetation types with that global change lead to a in the of ecological systems et al. 1992). evidence suggests that alterations in the composition and structure of plant communities drastically change disturbance and availability to the soil biota. The predicted of by due to global et al. is one example of potential of alterations in plant physicochemical traits to the vegetation through changes in the provision of resources to soil biota. forests a quality litter and have a primary the original availability of organic resources is expected to increase the of the soil microflora and of the food web by both control and litter quality (Wardle et al. 1998). of soil organisms and associated changes in decomposition rates significantly alter the availability of soil resources to the Soil biota are also by conditions. In for example, different of produced by different fungal may significantly affect fungal for organic resources and thus nutrient provision to plants under and et al. The potential impact of changes on soil organisms has by various (e.g., and of soil Changes in water supply and are Alterations in ecosystem functions should be when the soil communities that are by change organisms that out functions by few other and Gulledge (1998) that in where and of soil associated with change more of and fungi may be resulting in a in litter decomposition greater would be predicted by only the changes in soil and litter In a distinct litter result in greater and plant as nutrients in In a greater increase the or intensity of with impacts on the of a of (Schimel and Gulledge 1998). important of change is that direct species interactions be because of different rates of symbionts, and invertebrates et al. 1993). For plant species may be soils with root-feeding invertebrates which they have no (Van et al. 1998). Moreover, of cycles may alter patterns of both and and as well as mutualistic interactions between plants and soil biota. as with alterations in on a large and all of above- and belowground interactions are likely to be In to however, change occurs more and is strongly dependent upon conditions. results are thus expected to be and between especially in the few The influence of change on above- and belowground interactions be by the mechanisms associated with changes in species composition related to plant control of both disturbance regime of and provision of food resources to soil biota. Changes in ecosystem function should be strong in systems communities because traits controlling above- and belowground are distributed at each level of the ecological and strong alterations of the disturbance regime, soil resources and and consumption rates of resources. This is the for example, in high and where environmental conditions lead to communities with a diversity and species with and et al. 1992). species are to change and more to to is to be expected from changes in species of current for global through changes in and energy et al. The and global atmospheric and of impact terrestrial ecosystems. a of organic and is not to and changes on a large Alterations in the atmospheric composition of in the for example, are to those in high in and and Breemen 1997). Here we confine the discussion to the impact of increased current atmospheric is predicted to by the of the that of the A of a increase of in the of a large number of plant species with to high et al. 1995). primary production, litter and in plant structure resulting from may all have large effects on global vegetation patterns et al. 1993). A of is that on to the increase in is with different for different species and Moreover, the depend on the and ecological of the organisms For species are to et al. 1993). is that soil biota to increased of atmospheric because of high in changes are by a may soil and and root architecture may change The provision of food resources to soil biota is by in the of energy as a result of more plant by alterations in the composition of and by alterations in mycorrhizal to a of the effects of on suggested by Swift et al. a increase of primary to decomposition is to be Soil organic is likely to of carbon under Changes in the provision of organic resources by plants may significant functional within the soil community. Alterations in rates by plants due to sensitive of mutualistic symbionts (e.g., to high may lead to effects between global change and nutrient important role for is by the that alterations of ecosystem functions seem to be strongly by changes in between changes in plant species composition, and changes in and of litter (Swift et al. 1998). The and of effects are highly species specific and depend on other environmental such as soil nutrient and plant conditions et al. 1997). For example, the increased C:N in plant litter may decomposition rates and thus the supply of soil the other increased also increase N availability by the in soil or by altering plant engineering of soil as a result of Moreover, nutrient through changes in the performance of the belowground community under are likely to be more important in systems or in systems with of N et al. (1996) suggested that may a competitive under increased of by N provision by belowground or decreases N availability depends on the ecosystem of and on the organisms In the for the functional of changes in above- and belowground interactions also apply to alterations in atmospheric that is, changes be dependent and of atmospheric are and show strong and (Wolters and knowledge suggests that of resources (e.g., to changes in above- and belowground The can for different resources and a effects of increased nutrient on species richness to be (Heal Changes are with and in but are at high and in soils Changes in atmospheric are thus not expected to strong and large-scale alterations in above- and belowground interactions in the the ecological of atmospheric may increase through interactions with other drivers of global change. For example, and (2000) reported that increased N a mycorrhizal community one that is by This may increase the of the vegetation to environmental because species a balance on the host that is of associations and 2000). to changes in above- and belowground interactions are to because interactions are and direct and effects of ecosystem engineers, food web and and antagonistic among the soil biota and plants. factors on above- and belowground organisms to alter their and the environmental change. However, conditions have in this under which alterations of above- and belowground interactions are expected to drastically alter ecosystem because traits affecting the supply of soil resources, consumption rates of resources, and disturbance regime are unique to a few species or traits changes plants strongly differing in life and ecological alteration of the performance of soil modifications to communities with a high degree of functional and of highly specific Moreover, different life and for plants and soil organisms lead to the of the above- and belowground resulting from or disturbance Most of the vegetation change occurring is likely to a in plant life and ecological because alters the of and plants et al. 2000). This in and in and with and and in and systems where there is of The of above- and belowground to changes in species composition has important for future and et al. 2000). are of land and by changes in both and atmospheric affecting the of communities at their ecological In both all of the ecological the interactions among species at both of the are either strongly or sensitive to environmental change. in these not only the for the most alterations in ecosystem functions but also a for new strategies of under a variety of environmental conditions. alterations in ecosystem functions by changes in above- and belowground of specific interactions in and and large and and 1998). is a for and for new in but important (e.g., high should environmental to provide changes and potential of and ecological complexity of these In systems most to global the and of and species both above ground and below ground provision of food resources, and direct interactions should be in A understanding of both interactions between above- and belowground organisms and between and that may through in vegetation types to more global change scenarios of ecosystem functioning that soil biota and soil processes (Schimel and Gulledge 1998). We would to Chapin and for We the on Soil and and and for the between Above- and and for and an and the of and the The for to host the in The also to the of a in for Table of interaction between plants and soil biota and of affecting these interactions of global alterations in ecosystem functioning via changes in above- and belowground interactions of plant effects on soil biota and vice