Spatial associations and patterns of perennial vegetation ina semi-arid steppe: A multivariate geostatistics approach

Fernando T. Maestre*, Francisco Rodríguez, Susana Bautista, Jordi Cortina, Juan Bellot

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

42 Scopus citations

Abstract

We investigated the spatial patterns of perennial species (Stipa tenacissima, Anthyllis cytisoides Globularia alypum, Brachypodium retusum and chamaephytes) in a 50 m × 50 m semi-arid steppe by using the combination of a linear model of coregionalization (LMC) and sampling units of varying size (1.25 m × 1.25 m, 2.5 m × 2.5 m, and 5 m × 5 m). The data-adjusted LMC showed the patchy structure of the vegetation, which was especially evident with the highest resolution grid. It also detected a periodic pattern in the distribution of S. tenacissima, as well as autocorrelation at two spatial scales for A. cytisoides and G. alypum. The latter species was negatively associated with the other species at both short and long distances. These negative associations were consistent for all sampling grids and suggest the presence of interference between G. alypum and the rest of the evaluated species. Despite species-specific differences, the LMC was fitted satisfactorily to all of them. This suggests a common variation pattern for all the species, which may be caused by an underlying environmental property driving the patterns of all the species or, alternatively, by the dominance of some species' spatial pattern, or another kind of species association, over the rest. The spatial patterns found were profoundly affected by the observational scale. Our results reveal that the multivariate geostatistical approach introduced in this paper is a suitable technique for the spatial analysis of semi-arid plant communities. It allows plant ecologists to evaluate if the species forming the plant community of interest share a common spatial pattern, and to assess the spatial covariation between the species forming a plant community at different spatial scales independently.

Original languageEnglish (US)
Pages (from-to)133-147
Number of pages15
JournalPlant Ecology
Volume179
Issue number2
DOIs
StatePublished - Aug 2005

Keywords

  • Linear model of coregionalization
  • Mediterranean
  • Scale
  • Size of sampling units
  • Spatial pattern
  • Stipa tenacissima

ASJC Scopus subject areas

  • Ecology
  • Plant Science

Fingerprint

Dive into the research topics of 'Spatial associations and patterns of perennial vegetation ina semi-arid steppe: A multivariate geostatistics approach'. Together they form a unique fingerprint.

Cite this