Scaling hierarchy of factors controlling riparian vegetation patterns of the Fynbos Biome at the Western Cape, South Africa

• Published in: Journal of vegetation science Vol. 20; no. 1; pp. 17 - 26
• Main Authors: Sieben, E.J.J, Mucina, L, Boucher, C
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.02.2009
• Subjects: Canonical Correspondence Analysis; climatic zones; ecosystems; geographical variation; Gradient analysis; Hierarchical scaling; mountains; riparian areas; species diversity; Variation partitioning; vegetation; South Africa
• ISSN: 1100-9233; 1654-1103
• DOI: 10.1111/j.1654-1103.2009.03989.x

Question: How can vegetation gradients be described in riparian zones located in a species‐rich mountain range and how do these gradients explain the variation found in the vegetation? Location: Hottentots Holland Mountains, Western Cape, South Africa. Methods: Three gradients (geographic, longitudinal, lateral) were defined to describe the complex vegetation patterns found here. The gradients are related as follows: (1) the geographic gradient: related to the climatic and topographical changes across the entire mountain range; (2) the longitudinal gradient related to the changes along the different river reaches; (3) the lateral gradient related to the processes along the profile of the riverbed. These three gradients operated on three different hierarchical levels. Partial Canonical Correspondence Analysis (pCCA) was used to determine the amount of variation that is explained on each of the hierarchical levels. Results: The geographical gradient explained the highest fraction (more than 50%) of the total variation explained. This can be ascribed to the high species turnover across landscapes in the Fynbos Biome; this is most likely an outcome that is specific for this region. The second most important gradient was the lateral gradient, which reflects stream power and inundation frequencies of the river. This gradient is represented by ca. 48% of the explained variation and this gradient explains the major disturbances occurring in a riverine ecosystem. The longitudinal gradient was the least important of the gradients and shows overlap with the geographical gradient. Conclusions: In the species‐rich environment of the Fynbos Biome geographical factors do not only account for variation in zonal vegetation but also for variation within azonal vegetation, like riparian corridors.