| Zusammenfassung: | The variation within the Melaleuca quinquenervia dominated forested wetlands of the Moreton Region of South-East Queensland (Australia) was analysed in each of four vegetation 'segments' (stems, and canopy, shrub, and ground layers) in order to relate this variation to environmental attributes. Thirty sites were chosen for examination. Vegetation variables assessed include the species composition of stems, their diameters, number, and average height and the foliage projective cover of each species in the tree, shrub, and ground layers. Environmental variables assessed include landform, soil texture, ground and surface water levels, soil conductivity, soil pH, soil carbon content, and fire incidence. Data were analysed using semi-strong hybrid multidimensional scaling and the unweighted pair-groups method using arithmetic averages (UPGMA) classification method, supplemented by principal axis correlations and Monte Carlo methods to assign significance of correlations. Association matrices were correlated with one another using the Pearson correlation coefficient and subjected to a Mantel permutation (k = 1000) to obtain a degree of significance of the correlation. The descriptions and analyses presented show that there is considerable variation in the species composition and structure of M. quinquenervia dominated forested wetlands of the Moreton Region. However, the variation in one vegetation segment is a relatively poor indicator for the variation in another segment, as shown by the poor fidelities of vegetation Groups towards one another and the low correlations between the covariance matrices. In terms of species composition and quantities, the greatest predictive power exists between the composition of the canopy layer foliage and the composition of stems. The ground layer had the strongest correlations with environmental variables, largely due to relatively high species diversity. It suggests that the characteristics of the ground layer could suitably be used as indicators of wetland functioning. It also implies that though the assessment of ground layer vegetation is more expensive and difficult to undertake than canopy layer assessments, in order to understand how the landscape functions, how to manage it, and how to take advantage of the useful features of the landscape, it is clearly necessary to collect and analyse those data.
|
|---|