MEPS 242:1-14 (2002)  -  doi:10.3354/meps242001

Spatial pattern quantification of Antarctic benthic communities using landscape indices

N. Teixidó1,*, J. Garrabou2, W. E. Arntz1

1Alfred Wegener Institut für Polar- und Meeresforschung, Columbusstraße, 27568 Bremerhaven, Germany
2Station Marine d¹Endoume, Centre d¹Oceanologie de Marseille Station Marine d¹Endoume, rue Batterie des Lions, 13007 Marseille, France

ABSTRACT: Antarctic benthos exhibits highly complex communities showing a wide array of spatial patterns at several scales which are poorly quantified. In this study, we introduce the use of methods borrowed from landscape ecology to study quantitatively spatial patterns in the Antarctic megaepibenthic communities. This discipline focuses on the notion that communities can be observed as a patch mosaic at any scale. From this perspective we investigated spatial patterns in an Antartcic benthic assemblage across different stations based on landscape indices, and we chose the optimal subset for describing Antarctic benthic patterns. For this purpose, 42 photographs (1 m2 each) corresponding to 6 stations from the Weddell Sea shelf were investigated. Canonical variate analysis (CVA) showed the arrangement of photographic records along a patch size and diversity gradient on the first axis and a heterogeneity pattern gradient (cover area, interspersion and juxtaposition, landscape shape indices) on the second axis. Based on a forward stepwise selection, mean patch size (MPS), patch size coefficient of deviation (PSCV), patch richness (PR), interspersion and juxtaposition index (IJI), mean shape index (MSI), Shannon¹s evenness (SHEI), and periarea index (PERIAREA) were chosen as the adequate subset of indices to describe the Antarctic benthos. Principal Component Analysis (PCA) was used to identify relationships among them. The resulting 3 factors were interpreted as (1) a heterogeneity pattern (related to patch size, form, diversity, and interspersion indices), (2) an equitability pattern (represented by the evenness index), and (3) a perimeter-area pattern (characterised by the periarea index). Analysis of variance (ANOVA) was carried out to detect differences among the stations based on the subset of indices. Overall, the results showed large differences in patch characteristics (mean and its coefficient of variation, and shape indices), diversity, and interspersion. The successful description of Antarctic benthic communities through landscape pattern indices provides a useful tool for the characterisation and comparison of spatial patterns in these diverse marine benthic habitats, which gives insights in their organisation.

KEY WORDS: Antarctic · Benthic communities · Landscape indices · Multivariate ordination · Underwater photographs · GIS

Full text in pdf format