Descriptions of soil food webs are resolved at a functional level, with taxa aggregated into trophic groups (Fig. 1). This is by necessity due to the high taxonomic richness associated with most soil food webs and lack of knowledge regarding the specific feeding behavior of many of these taxa (Hunt et al., 1987). Here, I provide a general description of the current model of soil food web organization, adapted from Colemanet al. (2004), Wardle (2002), and various other sources.
As in terrestrial aboveground food webs, plants are the dominant pri- mary producers in soil. Resources enter the soil food web either via living plant material (roots and other underground structures) or via detritus (litter, dead roots, sloughed root cells, root exudates, and organic matter originally derived from flora and fauna). The distinction between living plant material and detritus is significant; abundance of each of these resources differs both spatially and temporally, which goes on to influence the abundance and activities of the organisms utilizing these resources (Bardgett et al., 2005a;
De Deyn and Van der Putten, 2005). Thus, trophic interactions and energy flow in the soil food web tend to cluster into a ‘‘herbivore food web’’ and a
‘‘detritivore food web’’ (Wardle, 2002). Algae and other photosynthetic protists are additional producers occurring in soil, but these represent significant sources of productivity only where plants are absent or sparse (e.g., in the dry valleys of Antarctica; Adamset al., 2006).
Symbiotic microorganisms and root-feeding invertebrates represent first- order consumers of living plant material. Some symbionts engage in mutualis- tic interactions with the host plant, providing access to some limiting resource or protection from antagonists in exchange for photosynthate. For example,
Shoots Roots-
feeding nematodes
Collem bolans
Mites I
Mites II
Fungus- feeding nematodes
Flagellates
Bacteria- feeding nematodes
Amebae Omni vorous nematodes
Predaceous nematodes
Nematode- feeding
mites Predaceous
mites
Mycorrhizae
Saprophytic fungi
Bacteria Roots
Inorganic N
Labile substrates
Resistant substrates
Figure 1 Connectivity web indicating trophic relationships among various functional groups of soil biota and their substrates. Reprinted from Hunt and Wall (2002), with permission from Blackwell Publishing.
mycorrhizal fungi can increase nutrient uptake for most plant species, and rhizobial bacteria fix atmospheric nitrogen for leguminous plant species. Other symbionts extract photosynthate to the detriment of the plant host. These parasitic symbionts include a diverse assemblage of endophytic bacterial, fungal, and nematode species. Several invertebrate species graze on plant roots, including a variety of nematodes and arthropods. Nematodes, as a group, exhibit highly variable feeding strategies, ranging from sedentary endoparasitism to migratory endoparasitism to ectoparasites (Yeates et al., 1993).
Second-order consumers in the herbivore food web are spatially and temporally distributed based on their feeding behaviors. Predators are active in the rhizosphere and bulk soil, where they encounter herbivores in search of food (consumed by, e.g., nematode-trapping fungi, predatory nema- todes, collembolans, and mites) and the external hyphae of symbiotic fungi (consumed by, e.g., fungal-feeding nematodes and collembolans).
A variety of parasites and pathogens are not only active in the rhizosphere and bulk soil (e.g.,Bacillusspp.,Pasteuriaspp.) but also encounter herbivores at the feeding site (e.g., various fungal parasites of nematode eggs).
The detritivore food web is active in the rhizosphere and extends into the soil where litter and organic matter are present. Saprotrophic bacteria and fungi represent first-order consumers of detritus. Some invertebrates (e.g., collembolans, enchytraeids) also feed directly on detritus and make resources available to other saprotrophs. For instance, the size of litter and structural barriers within it may prevent bacteria and fungi from accessing the nutrients contained within; these barriers are removed following com- minution and digestion of the litter. Invertebrates that engage in this activity are called ‘‘litter transformers.’’
A variety of bacterial predators, including some nematodes and protists, and fungal grazers, including some nematodes and microarthropods, represent second-order consumers in the detritivore food web. Among second-order consumers in both the herbivore and detritivore food webs, morphological characteristics of consumers and their resources indicate general patterns of consumption. Protozoan predators and bacterial-feeding nematodes consume their prey whole, while fungal-feeding mites, collembolans, and nematodes have mouthparts that are specialized for chewing or piercing. The distinction between consumers of bacteria and fungi also turns out to be important as bacterial and fungal feeders are spatially and temporally separated in terms of their activities in the soil: bacterial predators forage primarily in water-filled soil pores and water films adhering to the surfaces of soil particles where bacteria occur, while fungal grazers can occur in water films (nematodes) and in the humid, air-filled soil pores (various microarthropods) through which fungal hyphae pass (Colemanet al., 1983). Thus, energy flow in the detritivore food web is compartmented further into a ‘‘bacterial pathway’’ and a ‘‘fungal pathway.’’ Rates of production and turnover also differ between
these two pathways (Colemanet al., 1983). Production in the ‘‘fast’’ bacterial pathway rapidly increases in response to labile resource inputs and then falls off as resources are depleted. The ‘‘slow’’ fungal pathway is active over longer timescales, breaking down more recalcitrant substrates that are less variable through time. The fungal pathway is also less constrained by water avail- ability than the bacterial pathway. Such generalizations are useful in practice, even though variation in resource consumption and environmental constraints exist within these broad taxonomic categories. For example, zygomycete
‘‘sugar fungi’’ are early colonizers of plant litter, disappearing early during fungal succession (Garrett, 1981).
Predatory nematodes, collembolans, mites, and larger arthropods repre- sent higher-order consumers. These consumers tend to prey on consumers from both the herbivore and detritivore food webs and from the bacterial and fungal pathways, linking these energy flows. Earthworms greatly influ- ence soil structure and trophic interactions via their feeding and migratory activities. However, in general, earthworms are not included as a compo- nent of the soil food web even though they feed on most trophic groups within the web, albeit indirectly while digesting litter and soil organic matter. Small soil-dwelling mammals, such as moles and ground squirrels, also feed on larger soil invertebrates but are usually not included in soil food web models. These deletions illustrate that further descriptive research on soil food webs is necessary.
The environment in which these interactions occur represents an addi- tional player in the soil food web. Soil is a complex, three-dimensional matrix with hierarchical levels of structure at particulate, micro-, and mac- roaggregate levels (Rillig and Mummey, 2006). Soil texture modifies bacte- rial population dynamics at fine spatial scales by mediating interactions with predators (Elliott et al., 1980). Bacteria gain access to particulate organic matter sequestered within microaggregates via narrow-necked pores; bacte- rial feeders require pores with neck size greater than 3, 20, and 30mm for flagellates, nematodes, and ciliates, respectively (Brussaard, 1998). In addi- tion, soil texture and structure influence trophic interactions indirectly by affecting water potential (Brady and Weil, 2002). Thus, spatial patterns of soil food web dynamics depend, to a certain extent, on fine-scale patterns of soil structure.
Soil biota are important drivers of soil structure. Earthworms and plant root systems have strong effects on soil structure and texture (Brady and Weil, 2002). Direct effects of microorganisms on aggregate formation are believed to be active at different scales: fungal activity influences the forma- tion of macroaggregates while bacteria and archaea are thought to be more important at the microaggregate level (Rillig and Mummey, 2006). Indirect effects may arise due to interactions among microorganisms and other soil biota; for example, microbiota associated with arbuscular mycorrhizal (AM) fungi had differential effects on soil aggregate stability depending on the
identity of the fungal isolate they were associated with (Rilliget al., 2005).
Trophic interactions in soil have received little study with regard to their effects on soil structure. One hypothesis is that alterations in grazing intensity on fungi might influence soil aggregation via physical (effects on mycelial structure) and chemical (altered exudation patterns from grazed hyphae) mechanisms (Rillig and Mummey, 2006).