**Acknowledgements**

that determine litter decomposition is complex, and several pathways may occur. The effects of litter diversity on decomposition rates are dependent on several features, and the experi‐ mental control of these factors is absolutely necessary to clarify their relative roles and

Therefore, studies controlling, simultaneously, litter and edaphic fauna diversities, may give evidence to their joined effects on decomposition rates, allowing a better understanding of the relationship between these two factors. Furthermore, since several elements external to litter diversity (such as soil use, habitat fragmentation and others) may affect the effect of the discussed factors on litter decomposition, the incorporation of such elements would add information to the process. To do this it is necessary to design studies with manipulative control, which would render more informative studies than those that limit to compare

There are some alternative for these studies, and each alternative may achieve an answer that would clarify one or more of the above discussed points. One proposal it is the use of several plant species, among which different number of species are drawn, giving an idea of the relationship between the number of species in litter mixture and decomposition. In the array of litter mixtures other parameters may also be tested, such as arthropod species richness and abundance, as well as microbial biomass and activity. This approach tends to create situations more close to the actual environments studied, since usually the plant species used are the same as those occurring in the study site. On the other hand, this approach may increase very much the variability among repetitions, creating a noise that may impair data interpretation. The alternative approach would be the use of the same plant species in all repetitions, decreasing variability among them and giving more interpretable data, although decreasing their realism towards the studied environments. These studies are particularly common in habitats involving crops, associated or not to other plants, using decomposition litterbags

containing leaves of the cultivated plant, alone and together with other plant species.

In order to study the relative role of the trophic web components, manipulative studies are the most usual approach. In these studies, detritivore species are added or removed, and their effect on other species and/or on decomposition rates are observed. This approach suffers from the same problem of the manipulative studies involving litter mixtures: the distancing from biological reality. Furthermore, these studies are virtually impossible to carry out in hyper diversity habitats (as it is usual in most tropical biomes), and when the knowledge regarding taxonomic identity and/or biology of the involved species is weak. In such cases, experiments excluding litter fauna may be carried out through methods as fumigation, or by the use or other biocides, such as naphthalene. These substances, however, may cause impacts in the decomposer community, modifying decomposition process through more than one pathway of the activity of organisms, impairing once more data interpretation. Another option to exclude fauna is decomposition litterbags with different sized mesh, which may exclude selectively the fauna and modify their species composition. Nevertheless, most options to exclude fauna deal with species richness of the edaphic trophic web, and not on its species composition, weakening the interpretation of the effects of the functional and phylogenetic

interactions.

44 Biodiversity - The Dynamic Balance of the Planet

decomposition rates in different litter mixtures.

diversity on litter decomposition process.

We thank Sabrina Almeida, Júlio Louzada, Ricardo Ildefonso Campos and Elizabeth Nichols for contributions in manuscript idealization. Financial support of Fundação de Amparo à Pesquisa do Estado de Minas Gerais – FAPEMIG and Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq are gratefully acknowledged.
