Two-phase mosaics, or striped vegetation patterns (densely vegetated bands alternating regularly with bare areas), have been reported in arid and semiarid zones. They can occur provided (a) total rainfall is not sufficient t o maintain a dense cover; and (b) sufficient and uniform sheet flow can compensate, at least partly, for the lack of water. 2 Recent studies demonstrate that vegetation bands follow a successional model, in which bare areas are colonized by a pioneer front. Local studies in South West Niger suggest that transitions could occur between the various patterns observed within the same ecological region, but with varying average rainfall and surface features. 3 Because these transitions could not be followed in the-field without long term studies, a simple model has been elaborated to simulate the different structures observed. This model, based on cellular automata, is derived from the 'ganze of life' and depends only on two hypotheses which reflect conzpetition and synergy: the establishment, growth and survival of a given plant will be affected negatively by the influence of plants situated up-slope and positively by lateral and down-slope plants. 4 A matrix with 9 x 3 elements is applied to the same initial tree distribution grid, with three values of the a coefficient reflecting up-slope resource competition and two values of b reflecting lateral synergies. 5 The results demonstrate that almost all the structures observed in the field can be generated by this simple model, by varying only the two parameters a and 2, and the number of iterations. This result is independent of the initial tree density, showing that observed structures could equally well be derived from more or less dense vegetation patterns.

Research interest in microbial biodiversity over the past 25 years has increased markedly as microbiologists have become interested in the significance of biodiversity for ecological processes and as the industrial, medical, and agricultural applications of this diversity have evolved. One major challenge for studies of microbial habitats is how to account for the diversity of extremely large and heterogeneous populations with samples that represent only a very small fraction of these populations. This review presents an analysis of the way in which the field of microbial biodiversity has exploited sampling, experimental design, and the process of hypothesis testing to meet this challenge. This review is based on a systematic analysis of 753 publications randomly sampled from the primary scientific literature from 1975 to 1999 concerning the microbial biodiversity of eight habitats related to water, soil, plants, and food. These publications illustrate a dominant and growing interest in questions concerning the effect of specific environmental factors on microbial biodiversity, the spatial and temporal heterogeneity of this biodiversity, and quantitative measures of population structure for most of the habitats covered here. Nevertheless, our analysis reveals that descriptions of sampling strategies or other information concerning the representativeness of the sample are often missing from publications, that there is very limited use of statistical tests of hypotheses, and that only a very few publications report the results of multiple independent tests of hypotheses. Examples are cited of different approaches and constraints to experimental design and hypothesis testing in studies of microbial biodiversity. To prompt a more rigorous approach to unambiguous evaluation of the impact of microbial biodiversity on ecological processes, we present guidelines for reporting information about experimental design, sampling strategies, and analyses of results in publications concerning microbial biodiversity

No abstract

No abstract