Earthworms play an important role is soil functioning because they affect soil physical structure, soil matter dynamics and interact with other soil dwellers (Lavelle and Spain 2001). Assessing their functional role implies establishing both time and space scales at which they constitute the primary determinant of soil processes. A common strategy to tackle this question is to adopt a two steps approach involving a global spatiotemporal analysis of the community followed by an analysis of the co-structures shared by the community a the set of environmental variables conveying the processes at hand (Rossi 2003a).

It is beyond the scope of the present paper to discuss the statistical approaches that allow to perform such data processing. Various tools are available and the Partial Triadic Analysis (PTA) has proved to be effective on several occasions (Rossi 2003a, Jiménez et al. 2006). It provides useful summaries of the space-time community structure that can be compared to soil environmental descriptors. In a nutshell, the PTA searches for the community spatial structure common to all sampling dates (referred to the “compromise” in the jargon of the PTA). A complementary step (referred to as the intra-structure analysis) examines the discrepancies between the observed pattern recorded at each sampling occasions and the model common to all dates. Readers are referred to Rossi (2003a) and Rossi et al. (2014) for a presentation of the PTA.

Because space-time analyses are still not familiar to many biologists, we believe that an educational effort is welcome. One reason why the dissemination of such methods is still limited while computer resources are available (Chessel et al. 2004) is that ad hoc field data are not always easily available. The purpose of the present paper is to fill this gap by providing the original raw data set used in Rossi (2003a). This will facilitate the percolation of space-time multivariate methods in community ecology and help biologist who learn the PTA and related methods. In addition, we hope that making our raw data widely available will allow re-analyses by means of new tools and give rise to meaningful discussions (e.g. MEM analysis, see Jiménez et al. (2014)). Note also that part of the present data set was used to illustrate the SADIE method in Rossi (2003b).

We thank Dr R. Vuattoux, Director of the Station d’Ecologie de Lamto at the time of the work was carried out, for his assistance and interest. We are indebted to R. Zouzou Bi Danko for technical assistance in the field.

JPR and PL designed and performed the experiments and wrote the present note.