Application of ordinal correspondence analysis for submerged aquatic vegetation monitoring

作者：Mante Claude^*; Bernard Guillaume; Bonhomme Patrick; Nerini David

来源：Journal of Applied Statistics, 2013, 40(8): 1619-1638.

DOI：10.1080/02664763.2013.789494

• ordinal correspondence analysis
• tests
• Chi-squared partitioning
• ecological monitoring
• macrophytes
• Berre lagoon

The European Water Framework states that macrophyte communities (seaweeds and seagrass) are key indicators of the ecological health of lagoons. Furthermore, the restoration of these communities, especially the Zostera meadows, is one of the main objectives of the Berre lagoon restoration plan. Consequently, a monitoring programme of the main macrophyte species still present in the lagoon was initiated in 1996. This monitoring resulted in a sequence of 11 spatially structured annual tables consisting of the observed density of these species. These tables are processed in this study. First, we specify the principles of Beh%26apos;s ordinal correspondence analysis (OCA), designed for ordered row/column categories, and compare this method to classical correspondence analysis (CA). Then, we show that OCA is straightforwardly adaptable for processing a sequence of ordered contingency tables like ours. Both OCA and CA are afterwards used to reveal and test the main patterns of spatio-temporal changes of two macrophyte species in the Berre lagoon: Ulva and Zostera. The results we obtained are compared and discussed.

• 出版日期2013-8-1