Problems Of Concern In Soil Biodiversity Studies

An alternative to the functional approaches just discussed is taken by André et al. (2002), who note that most investigators use inadequate sampling designs or sample too shallowly in the soil profile to get a complete sample of microarthropods to provide the information used in the models noted previously. In an extensive survey of the worldwide literature on microarthropods, they claim that, on average, at most 10% of the soil microarthropod populations have been explored and 10% of the species described, due to the use of inefficient extraction procedures. This is supported by Walter and Proctor (2000), who suggest that perhaps only 5% of the species of mites worldwide are described so far. André et al. (2002) make the very valid point that ecologists need to be aware of the numerous pitfalls and possible flaws inherent in many extraction procedures; that is, none of them are 100% efficient. In the section on field studies and laboratory analyses, we explore some of these concerns more extensively.

There is an understandable concern that some quantifiable relationship be given to the relationship between ecosystem function and diversity. This is portrayed in Figure 7.5 (Bengtsson, 1998), which contrasts two curves of ecosystem function as a function of increasing numbers of species. Type 1, a continually ascending curve, represents the hypothesis that all species are important for ecosystem function. Type 2, initially convex and then flat, represents the species redundancy hypothesis. Bengtsson (1998) argues that it is more informative to consider specific functions in ecosystems, namely decomposition, nutrient mineralization, or primary production, thus focusing on phenomena that are more amenable to scientific inquiry. Bengtsson (1998) argues strongly that diversity does not play a role in ecosystem function. He goes so far as to assert that: "correlations between diversity and ecosystem functions— which may very well exist—will be mainly non-causal correlations only." As we are trying to show in this chapter, the truth may well lie in some midpoint between these extremes. The fact that certain functions may be linked to just a few genera or species, such as autotrophic and het-erotrophic nitrifiers, for example, means that this might well be a "pressure point" for concern about long-term ecosystem function. The "natural insurance capital" concept of Folke et al. (1996), also discussed in detail by Bolger (2001), suggests that it is essential to retain as much species richness as possible to ensure that complete ecosystem services exist as human needs or environmental changes occur.

As in all areas of ecology, there is a spatial dimension to the biodiversity of soil organisms. It is essential to know not only which species are present, but also where the counted species occur in relation to one another. Do species occur together at every microsite, or do they occur mostly individually in separate sites? This has an important bearing on competition and other interactions, with functional consequences for the ecosystem. Ettema and Yeates (2003) measured patterns of small (centimeter) and intermediate (meter [m]) scales in nematode communities in a forest compared to a pasture system on a similar soil type in New Zealand. Using geostatistical techniques and mathematical calcu-

FIGURE 7.5. A hypothetical example of an attempt to quantify the form of the relationship between ecosystem function and diversity. The Type 1 curve represents the hypothesis that all species are important for ecosystem function, while Type 2 is the species redundancy hypothesis. The bars indicate the range of responses as different numbers of species are randomly drawn from a source pool of species, given that species' effects on ecosystem function are mainly species-specific (idiosyncratic) and not related to diversity. Note that although an average response may be observed, it neither allows a distinction between the two different hypotheses (Type 1 and Type 2), nor does this average response allow any useful prediction of what will happen in individual cases of species deletions (from Bengtsson, 1998).

FIGURE 7.5. A hypothetical example of an attempt to quantify the form of the relationship between ecosystem function and diversity. The Type 1 curve represents the hypothesis that all species are important for ecosystem function, while Type 2 is the species redundancy hypothesis. The bars indicate the range of responses as different numbers of species are randomly drawn from a source pool of species, given that species' effects on ecosystem function are mainly species-specific (idiosyncratic) and not related to diversity. Note that although an average response may be observed, it neither allows a distinction between the two different hypotheses (Type 1 and Type 2), nor does this average response allow any useful prediction of what will happen in individual cases of species deletions (from Bengtsson, 1998).

lations of species turnover, they compared nematode genera in forestland, then in pasture. The forestland was assumed to have greater variation in vegetation and hence belowground inputs, on small and intermediate scales, than in the pasture. Thus they hypothesized that nematode genera are more strongly aggregated (occurring in "hot spots") in the mixed forest than in the ryegrass/white clover pasture. Applying an optimization method for sampling in geostatistical studies called spatial simulated annealing (SSA) developed by Van Groenigen and Stein (1998), Ettema and Yeates (2003) sampled along 40-m-long transects for the meter scale, with distance classes of 3 m, reflecting the scale of tree spacing. The centimeter scale transect was one-tenth of the large scale, or 4 m. The total number of nematodes per soil core volume was more than five times higher in the pasture (2800 ± 1234) than in the forest (430 ± 252), but the average number of genera in the forest (23.7 ± 3.3) was higher in the forest than in the pasture (19.1 ± 2.5). Also, many more nematode genera occurred in the forest (53) than in the pasture (37). Dissimilarity analysis showed that generic turnover was significantly greater in the forest than in the pasture, both at the small and intermediate scales (Fig. 7.6) (Ettema and Yeates, 2003). Because increasing distance in the forest led to increasing dissimilarity between communities, and no plateau was reached, it is possible that there is additional species turnover on scales larger than those explicitly sam-

(a) Short Transects

(b) Long Transects

0 0

Post a comment