The importance of plant litter decomposers to soil formation is unquestioned (Odum and Biever, 1984; Vi-tousek and Sanford, 1986; Whitford, 1986; Swift and Anderson, 1989; Meadows, 1991). Soils in turn provide an array of ecosystem services that are so fundamental to life that their total value could only be expressed as infinite (Daily et al., 1997). Detailing the contribution of cockroaches relative to other decomposers, however, is difficult. First, information is scarce. For any given ecosystem, it is the decomposers that receive the least detailed attention. Second, like most decomposers, cockroaches are so adaptable that they often do not have well defined ecological roles; functional redundancy among detriti-vores is high (Scheu and Setala, 2002). Third, because of the intricate synergistic and antagonistic interactions among diverse bacteria, fungi, and invertebrates, decomposition is manifested in scales of space and time not easily observed or quantified. Decomposition occurs both internal and external to the gut, and at microscopic spatial scales. It operates via the creation of physical artifacts, like burrows and fecal pellets, which accumulate and continue to function in the absence of their creators. Effects can be localized and short term, or wide ranging and extended in time; wood decomposition in particular is a very long-term stabilizing force in forest ecosystems (Anderson et al., 1982; Anderson, 1983; Swift and Anderson, 1989; Wolters and Ekschmitt, 1997; Wardle, 2002).
Other problems in attempting to quantify the role of arthropods in decompositional processes are related to sampling bias; no one method works best for all groups and all soils (Wolters and Ekschmitt, 1997). The results of pitfall trapping, for example, can be difficult to interpret. No cockroaches were taken in unbaited pitfall traps in four habitats in Tennessee, but traps attracted quite a number of blattellids when bait (cornmeal, cantaloupe, fish) was added (Walker, 1957). Surface-collecting methodology such as soil and litter cores may not account for cockroach species that are only active after seasonal precipitation or those that shelter under bark, under stones, or in other concealed locations during the day. Sampling techniques for canopy arthropods also have methodological biases with regard to a given taxon, particularly those species in suspended soils and those that are seasonally present. Diurnal, seasonal, and spatial aggregation fur ther complicate the proper estimation of abundance (Basset, 2001).
Members of the blattoid stem group undoubtedly played a major role in plant decomposition during the Paleozoic (Shear and Kukalová-Peck, 1990). The ecological significance of extant cockroaches, however, is usually assumed to be negligible (Kevan, 1993) because of their often low numbers during surveys (e.g., some Australian studies—Postle, 1985;Tanton et al., 1985;Greenslade and Greenslade, 1989). If considered in terms of biomass, however, their importance is magnified because of large individual body size relative to many other detritivores such as mites and Collembola. Basset (2001), in a review of studies conducted worldwide, concluded that cockroaches dominated in canopies, comprising an astonishing 24.3% of the invertebrate biomass (discussed in Chapter 3). The clumped distribution and social tendencies of many species also tends to increase their ecological impact. Cockroaches that aggregate in tree hollows, for example, directly benefit their host plant, as defecation steadily fertilizes the soil at the base of the tree (Janzen, 1976). Large, subsocial or gregarious woodfeeding cockroaches may be able to pulverize logs on a time scale comparable to, if not better than, termites. In this regard, several studies in montane environments report that cockroach population levels in plant litter are negatively correlated with the presence of termites, a group that strongly and predominantly influences the pattern of decomposition processes and whose ecological importance is clear. Surveys on Mt. Mulu in Sarawak, Borneo, indicate that the density of soil- and litterdwelling termites declines with altitude (Collins, 1980). Cockroaches were present in low numbers at all altitudes, but individuals were larger and more numerous in upper montane forests, where they constituted 40% of the total macrofauna biomass. Rhabdoblatta was the most common genus at upper altitudes, found in all plots from 1130 m upward, but not below. The Cryptocercus punctu-latus species complex dominates the saproxylic guild in the Southern Appalachian Mountains, and occupies the same niche as does the subterranean termite Reticuli-termes at lower elevations (Nalepa et al., 2002). The same altitudinal trend was evident in soil and litter core samples taken on Volcán Barva in Costa Rica; the biomass of cockroaches fluctuated, but generally increased with altitude. Termites were not found above 1500 m, but cockroaches made up 61% of the biomass at that altitude (Atkin and Proctor, 1988). On Gunung Silam, a small mountain in Sabah, the altitudinal associations were reversed. At 280 m, cockroaches were 84% of the invertebrate biomass and termites were not found; at 870 m, termites were 25% of the biomass, while cockroaches were
< 1% (Leakey, 1987, Table 3). The reasons for these alti-tudinal changes in distribution were not causally related to measured changes in other site properties such as forest structure and soil organic matter in the Costa Rican study (Atkin and Proctor, 1988).
Was this article helpful?