Fossorial arthropods alter the structure of the soil by redistributing soil and organic ma- terial and increasing soil porosity (J. Anderson 1988). Porosity determines the depth to which air and water can penetrate the substrate. A variety of substrate-nesting verte- brates, colonial arthropods, and detritivorous arthropods and earthworms affect the spa- tial and temporal patterns of substrate structure, organic matter content, and infiltration in terrestrial and aquatic systems.
441 II. EffECTS of DETrITIvory AnD BUrrowIng
Defecation by a larval caddisfly, Sericostoma personatum, increases subsurface organic content in a stream ecosystem by 75–185% (R. Wagner 1991). The caddisfly feeds on detritus on the surface of the stream bed at night and burrows into the stream bed dur- ing the day, trapping organic matter in burrows. Frouz et al. (2004) found higher survival of chironomid larvae in benthic substrates composed primarily of sand and/or accumulated chironomid fecal pellet aggregates (particle diameter >0.25 mm), compared to substrates composed of fine organic sediment (<0.25mm). Accumulated fecal pellets increased mean substrate particle size, leading to more extensive larval tunneling and higher dissolved oxygen levels also seen in sand substrates.
Ants and termites are particularly important soil engineers (Dangerfield et al. 1998, Jouquet et al. 2006, MacMahon et al. 2000). Colonies of these insects often occur at high densities and introduce cavities into large volumes of substrate. Eldridge (1993) reported that densities of funnel ant, Aphaenogaster barbigula, nest entrances could reach 37 m−2, equivalent to 9% of the surface area over portions of the eastern Australian landscape.
Nests of leaf-cutting ants, Atta vollenweideri, reach depths of >3 m in pastures in west- ern Paraguay (Jonkman 1978). J. Moser (2006) excavated a leaf-cutting ant, Atta texana, nest in northern Louisiana, U.S.A., and found 97 fungus-garden chambers, 27 dormancy chambers, 45 detritus chambers (for disposal of depleted foliage substrate) and a central cavity at 4 m depth in which the ants and fungus overwinter (Fig. 14.2). The nest extended over an area of 12 × 17 m on the surface and was at least 4 m deep. The bottom of the colony could not be reached, but vertical tunnels extended to at least 7.5 m and might have extended to the water table at 32 m. Whitford et al. (1976) excavated nests of desert harvester ants, Pogonomyrmex spp., in New Mexico, U.S., and mapped their 3-dimensional structure (Fig. 14.7). Colony densities were 21–23 ha−1 at four sites, and each colony consisted of 12–15 interconnected galleries (each about 0.035 m3) within a 1.1 m3 volume (1.5 m diameter × 2 m deep) of soil, equivalent to about 10 m3 ha−1 of cavity space (Fig. 14.7). These colonies frequently penetrated the calcified hardpan (caliche) layer, 1.7–1.8 m below the surface.
The infusion of large soil volumes with galleries and tunnels greatly alters soil structure and chemistry. Termite and ant nests typically represent sites of concentrated organic matter and nutrients (Ackerman et al. 2007, J. Anderson 1988, Culver and Beattie 1983, Herzog et al. 1976, Holdo and McDowell 2004, J. Jones 1990, Jurgensen et al. 2008, Le- sica and Konnowski 1998, MacMahon et al. 2000, Mahaney et al. 1999, A. Risch et al.
2005, Salick et al. 1983, D. Wagner 1997, D. Wagner and Jones 2004, D. Wagner et al.
1997). Nests may have concentrations of macronutrients 2–3 times higher than surround- ing soil (Fig. 14.8). J. Jones (1990) and Salick et al. (1983) noted that soils outside termite nest zones become relatively depleted of organic matter and nutrients. L. Parker et al.
(1982) reported that experimental exclusion of termites for 4 yrs increased soil nitrogen concentration by 11%. Ant nests also have been found to have higher rates of microbial activity, and carbon and nitrogen mineralization than do surrounding soils (Dauber and Wolters 2000, Lenoir et al. 2001, D. Wagner and Jones 2004) and represent sites of concentrated CO2 efflux (Domisch et al. 2006, Jurgensen et al. 2008, A. Risch et al. 2005).
Nest pH often differs from that of surrounding soil. Mahaney et al. (1999) found significantly higher pH in termite mounds than in surrounding soils. Jonkman (1978) noted that soil within leaf-cutter ant, Atta spp., nests tended to have higher pH than did soil outside the nest. However, D. Wagner et al. (1997) measured significantly lower pH (6.1) in nests of harvester ants, Pogonomyrmex barbatus, than in reference soil (6.4).
Lenoir et al. (2001) reported that Formica rufa nests had a higher pH than did surrounding soil at one site and a lower pH than did surrounding soil at a second site, both in Sweden.
Ant mounds in Germany did not differ in pH from surrounding soils (Dauber and Wolters 2000).
Termites and ants also transport large amounts of soil from lower horizons to the sur- face and above for the construction of nests (Fig. 14.9), gallery tunnels, and “carton”, the soil deposited around litter material by termites for protection and to retain moisture during feeding above ground (Fig. 14.10) (Whitford 1986). Whitford et al. (1982) reported that termites brought 10–27 g m−2 of fine-textured soil material (35% coarse sand, 45%
fIG. 14.7 Vertical structure of a harvester ant, Pogonomyrmex rugosus, nest in southern New Mexico. From Whitford, et al. (1976) with kind permission of the authors and Springer Science + Business Media.
443 II. EffECTS of DETrITIvory AnD BUrrowIng
medium fine sand, and 21% very fine sand, clay and silt) to the surface and deposited 6–20 g of soil carton per gram of litter removed (Fig. 14.4). Herrick and Lal (1996) found that termites deposited an average of 2.0 g of soil at the surface for every gram of dung removed. Mahaney et al. (1999) reported that termite mound soil contained significantly more (20%) clay than did surrounding soils.
A variety of vertebrate species in Africa have been observed to selectively ingest termite mound soil (Holdo and McDowell 2004, Mahaney et al. 1999). Mahaney et al.
(1999) suggested that the higher clay content of these mounds, along with their higher pH and nutrient concentrations, could mitigate gastrointestinal ailments and explain the consumption of mound soil by chimpanzees. Termite mound soils, as well as surrounding soils, had high concentrations of metahalloysite, used pharmaceutically, and other clay fIG. 14.8 Concentrations of major nutrients from bog soil (Grnd), hummocks (Hum) and Formica nests (Ant) in bogs in Montana, U.S. Vertical bars represent 1 SE. Means with different letters are significantly different at P < 0.05. From Lesica and Kannowski (1998) with permission from The University of Notre Dame.
minerals that showed mean binding capacities of 74–95% for four tested alkaloids. Chim- panzees could bind most of the dietary toxins present in 1–10 g of leaves by eating 100 mg of termite mound soil.
A number of studies have demonstrated effects of soil animals on soil moisture (Fig. 14.11). Experimental reduction or removal of litter from the soil surface increases soil temperature and evaporation, and reduces infiltration of water. Burrowing and redistribution of soil and litter by animals increase soil porosity, water infiltration, and the stability of soil aggregates that control water- and nutrient-holding capacity. Conversely, the dense pavement over mound-building termite nests restricts water infiltration but increases moisture in the runoff zone surrounding the mound (I. Ackerman et al. 2007, Eldridge 1993, 1994).
Ant and termite nests have particularly important effects on soil moisture because of the large substrate surface areas and volumes affected (MacMahon et al. 2000). D. Wagner (1997) reported that soil near ant nests had higher moisture content than did more distant soil. Elkins et al. (1986) compared runoff and water infiltration in plots with termites fIG. 14.9 Termite castle in northern Australian woodland. Dimensions are approximately 3 m height and 1.5 m diameter.
445 II. EffECTS of DETrITIvory AnD BUrrowIng
present or excluded during the previous four years in New Mexico, U.S. Plots with
< 10% plant cover had higher infiltration rates when termites were present (88 mm hr−1) than when termites were absent (51 mm hr−1); runoff volumes were twice as high in the termite-free plots with low plant cover (40 mm) as in untreated plots (20 mm). Infiltra- tion and runoff volumes did not differ between shrub-dominated plots (higher vegetation cover) with or without termites.
Eldridge (1993, 1994) measured effects of funnel ants and subterranean harvester termites, Drepanotermes spp., on the infiltration of water in semi-arid eastern Australia.
He found that infiltration rates in soils with ant nest entrances were 4–10-fold higher (1030–1380 mm hr−1) than in soils without nest entrances (120–340 mm hr−1). Infiltration rate was correlated positively with nest entrance diameter. However, infiltration rate on the sub-circular pavements covering the surface over termite nests was an order of magni- tude lower than in the annular zone surrounding the pavement or in inter-pavement soils (Fig. 14.12). The cemented surface of the pavement redistributed water and nutrients fIG. 14.10 Termite gallery carton on stems of dead creosotebush. Soil particles are
cemented together to provide protection and moisture control during termite feeding on detrital material.
from the pavement to the surrounding annular zone. Ant and termite control of infiltra- tion creates wetter microsites in moisture-limited environments.
c. primary production and vegetation dynamics
Through control of decomposition, mineralization and pedogenesis, detritivorous and fos- sorial arthropods have the capacity to control nutrient availability for, and perhaps uptake by, plants (Crossley 1977, Setọlọ and Huhta 1991). In particular, the release of nitrogen and phosphorus from decaying organic matter often is correlated with plant productivity (Vi- tousek 1982, T.E. Wood et al. 2009). Some species, especially ants, directly alter the vegetation around their nests by concentrating harvested seeds in nests and/or by clipping plants that surround nests (MacMahon et al. 2000, see Chapter 13). However, relatively few studies have measured the effect of detritivores and burrowers on plant growth or vegetation dynamics.
C. Edwards and Lofty (1978) compared seedling emergence and shoot and root growth of barley between pots of intact, sterilized soil (from fields in which seed had been either drilled into the soil or planted during plowing) with microarthropods or earthworms either absent or reintroduced. Percent seedling emergence, plant height, and root weight were higher in plowed soil and direct-drilled soil with animals, compared to sterile direct- drilled soil, suggesting important effects of soil animals on mineralization, soil porosity and infiltration that support primary production.
R. Ingham et al. (1985) inoculated microcosms of blue grama grass, Bouteloua gracilis, in sandy loam soil low in inorganic nitrogen, with bacteria or fungi. Half of each microflora fIG. 14.11 Effects of soil invertebrates on soil water balance. Reprinted from
J. Anderson, (1988) with permission from Elsevier.
447 II. EffECTS of DETrITIvory AnD BUrrowIng
treatment was inoculated with microbivorous nematodes. Plants growing in soil with bac- teria and bacteriophagous nematodes grew faster and acquired more nitrogen initially than did plants in soil with bacteria only. Addition of mycophagous nematodes did not increase plant growth. These differences in plant growth resulted from greater nitrogen mineralization by bacteria (compared to fungi), excretion of NH4+-N by bacteriophagous (but not mycophagous) nematodes, and rapid uptake of available nitrogen by plants. My- cophagous nematodes did not increase plant growth or nitrogen uptake over fungi alone, because these nematodes excreted less NH4+-N, and the fungus alone mineralized suffi- cient nitrogen for plant growth.
In a unique, definitive study, Setọlọ and Huhta (1991) created laboratory microcosms with birch seedlings, Betula pendula, that had been planted in partially sterilized soil that was reinoculated with soil microorganisms only or with soil microorganisms and a diverse soil fauna. During two growing periods, the presence of soil fauna increased birch leaf, stem and root biomass by 70%, 53% and 38%, respectively, and increased foliar nitro- gen and phosphorus contents 3-fold and 1.5-fold, respectively, compared to controls with microorganisms only (Fig. 14.13). More recently, Laakso and Setọlọ (1999) found that experimental removal of microbe- or detritus-feeding soil fauna, especially the microbi- detritivorous enchytraeid worm, Cognettia sphagnetorum, reduced plant biomass and up- take of nitrogen.
Soil arthropods could influence plant growth by inoculating roots with mycorrhizal fungi. Rabatin and Stinner (1988) reported that 28–97% of soil animals contained mycor- rhizal spores or hyphae in their guts. Conversely, fungivore grazing on mycorrhizae could inhibit plant growth by interfering with nutrient uptake.
fIG. 14.12 Effect of termite colony structure on infiltration of water under ponded condi- tions (brown) and under tension (yellow). Vertical lines indicate 1 standard error of the mean.
Reprinted from Eldridge (1994) with permission from Elsevier.
Soil animals also influence community dynamics. Although ant and termite nests may represent relatively minor components of total soil carbon and nutrient pools, they sub- stantially increase spatial heterogeneity of soil water and nutrient availability (Domisch et al. 2006, Jurgensen et al. 2008, MacMahon et al. 2000, A. Risch et al. 2005), thereby influencing patterns of community development. Zaragoza et al. (2007) reported that distinct protozoan communities were associated with ant nest mounds, compared to refer- ence soils 5 m away from ant nests. Several studies have demonstrated that ant and ter- mite mounds typically support distinct plant communities, compared to surrounding soil (Fig. 14.14, Brody et al. 2010, Garrettson et al. 1998, Q. Guo 1998, Holdo and McDowell 2004, T. King 1977a, M. Schütz et al. 2008), but the effect on vegetation development may differ between active and abandoned mounds (Lesica and Kannowski 1998, F. Smith and Yeaton 1998).
Lesica and Kannowski (1998) reported that wood ants, Formica podzolica, were responsible for mound formation in peat bogs in Montana, U.S. Mounds provided an elevated habitat that was warmer, better aerated, and had higher nutrient content than did the surrounding peat surfaces (Fig. 14.8). Although active mounds supported only a few species of grasses, abandoned nests supported shrubs, as well as plant species that fIG. 14.13 Biomass production (left of break in horizontal axis) and nitrogen accumulation (right of break in horizontal axis) of birch, Betula pendula, seedlings. Bars above the horizontal axis are stems (orange) and leaves (yellow); bars below the horizontal axis are roots in humus (blue) and roots in mineral soil (yellow). C = fauna removed; F = refaunated. Vertical lines represent 1 standard deviation for all data (except nitrogen at week 45, where vertical lines represent minimum and maximum values). For C vs. F, * = P < 0.05; *** = P < 0.001. Stem nitrogen was not measured week 10. From Setọlọ and Huhta (1991) with permission from the Ecological Society of America.
449 II. EffECTS of DETrITIvory AnD BUrrowIng
could not grow in the saturated peat surface. The ants foraged primarily on honeydew from aphids that were tended on shrubs, indicating a positive feedback relationship.
Abandoned nests of leaf-cutter ants, A. vollenweideri, serve as sites of accelerated succession in Paraguayan pastures (Jonkman 1978). Collapse of the nest chamber forms a depression that holds water and facilitates the development of woody vegetation. At high nest densities, these oases coalesce, greatly increasing forest area. Brenner and Silva (1995) found that active nests of Atta laevigata were associated more frequently with groves of trees, and the size of nests increased with grove size and the abundance of forest tree species in Venezuelan savanna, suggesting that active nests both facilitated and were facilitated by formation of groves.
L. Parker et al. (1982) demonstrated that termite exclusion significantly reduced the biomass of four annual plant species and significantly increased the biomass of one annual plant species. They observed an overall trend toward increased biomass of annual plants in plots with termites excluded. These results probably reflected increased nitrogen availability in plots from which termites were excluded, compared to plots with fIG. 14.14 Vegetation composition along 46 transects from Formica exsecta mound centers into surrounding alpine grassland in Switzerland. Box plots indicate median, 25th and 75th percen- tiles, black lines are 10th and 90th percentiles, and filled circles are all values below 10th and above 90th percentiles. Different letters indicate significant differences (P < 0.05). Note different scales for y axes. From M. Schütz et al. (2008) with permission from the authors and John Wiley & Sons.
unmanipulated termite abundance. I. Ackerman et al. (2007) found that active termite mounds in abandoned farmland in central Amazonia inhibited water retention and re- stricted secondary succession. By contrast, Bloesch (2008) reported that Macrotermes spp.
mounds provided spatially-distinct elevated sites for development of wooded thickets in seasonally flooded savanna landscapes in East Africa.
The high nutrient concentrations of termite and ant nests are incorporated by plants growing on or near nests and become available to higher trophic levels (Fig. 14.15).
fIG. 14.15 Vertebrate dung density at increasing distance from termite, Odontotermes spp., mound centers into surrounding savanna in north-central Kenya. A) for eland and steinbuck, a linear regression explained 40% and 19% of variance, respectively. B) for Grant’s gazelle, zebra and cow, a second-order polynomial explained 26%, 73% and 54% of variance, respectively. In all cases, the first term was not significant, but the second term was highly significant (P < 0.01). From Brody et al. (2010) with permission from the Ecological Society of America.
451 III. SUMMAry
Holdo and McDowell (2004) reported that trees growing on termite mounds had higher concen trations of all nutrients tested, except sodium and crude protein, than did trees from the surrounding woodland matrix in Zimbabwe. Trees on mounds were also subjected to more intense feeding by elephants. Termite and ant nests thereby affect food availability and feeding patterns for herbivores, providing indirect positive feedback for herbivore effects on litter quality and availability for detritivores. Brody et al. (2010) found that Acacia drepanolobium trees growing at the edge of termite, Odonototermes spp., mounds in Kenya showed twice the foliar nitrogen content and seed production as trees growing away from mounds, due to enhanced soil N and P in mounds. Fox-Dobbs et al. (2010) further reported that A. drepanolobium trees off mounds acquired a higher percentage of N via fixation (55–80%) than did trees near mounds (40–50%), reflecting the higher availability and use of soil-based N near termite mounds. Exclusion of vertebrate herbivores, which preferentially used mound vegetation, as evidenced by patterns of dung deposition (Fig. 14.15), did not affect these results, demonstrating that soil enrichment by termites, rather than dung and urine deposition by vertebrates, was responsible for the vegetation responses (Brody et al. 2010).
III. SummARy
Decomposition and pedogenesis are major ecosystem processes that affect biogeochemi- cal cycling, trace gas fluxes, soil fertility and primary production. Decomposition of organic matter involves four component processes: photo-oxidation, leaching, comminution, and mineralization. Arthropods are key factors influencing comminution and mineralization.
The functional groups that are involved in decomposition include coarse comminuters that fragment large materials and fine comminuters that fragment smaller materials, of- ten those produced by large comminuters. In aquatic ecosystems, scrapers and shredders represent coarse comminuters, whereas gatherers and filterers represent fine comminut- ers. Xylophages represent a specialized group of comminuters that fragment woody litter in terrestrial and aquatic ecosystems. Carrion feeders reduce carcasses, and coprophages feed on animal excrement. Fungivores and bacteriovores fragment detrital material while grazing on microflora. Fossorial functional groups include subterranean nesters that ex- cavate simple burrows, gatherers that return detrital or other organic materials to nesting areas, and fossorial feeders that consume organic material and/or soil and mix biotic and abiotic materials in their wake.
Evaluation of the effects of detritivores and burrowers on decomposition and pedo- genesis requires appropriate methods for measuring animal abundances and process rates. Abundances of detritivores or burrowers can be manipulated using exclusion and microcosm techniques, and detritivory can be measured as the product of detritivore abundance and individual consumption rate or as the rate of disappearance of substrate.
Decomposition is measured most commonly as respiration rate, as the ratio of litter input to litter standing crop, or as the rate of litter disappearance. Isotopic tracers also provide data on decomposition rate.
Decomposition rates typically are higher in mesic than in arid ecosystems. Differ- ent functional groups dominate different ecosystems, depending on the availability and quality of detrital resources. For example, shredders dominate headwater streams where coarse detrital inputs are the primary resource, whereas filterers dominate larger streams with greater availability of suspended fine organic material. Xylophages occur only in ecosystems with woody residues. Decomposition generally can be modeled as a multiple