This case study illustrates some applications of the methods explained in the preceding chapters. It is a 'real world' example assessing ecological and also methodological questions. The vegetation data originate from a survey across Switzerland (Wohlgemuth et al. 2008) which aimed to reveal relationships between vegetation composition and the growth rate of tree species. Just as in an experimental approach, questions were posed prior to the investigation and a sampling design is developed below to obtain the answers. But as is often the case in large surveys the order of steps is partly reversed, as in some examples the data appear at the outset of the exercise and the questions that they could potentially answer are identified later. This imposes restrictions on the analytical methods, but it also allows exploration of the variable set far beyond its primary scope.
The analyses shown below concentrate on species composition and species spatial distribution. An alternative would be to focus on species richness, as done by Wohlgemuth et al. (2008), who found that there is a high correlation between diversity (the number of species per plot) and the canopy cover of trees (a surrogate for light availability inside the forest stands). However, they also detected that the spatial resolution of the survey is probably not the best considering biodiversity, as correlations become higher after clumping the sampling units into landscape patches of about 100 km2 in size.
Hence, the scale as well as the variables chosen in sampling design determine the potentials and restrictions of application. As will be shown below the sampling area is the territory of Switzerland (~41000km2), restricted to its forested area (~30% of the surface). The strong elevation gradient is the main cause of spatial variation in climate, and climatic relationships are therefore an issue. Concentrating the study on forests means that human influence is weak and controlled, a consequence of the strong regulations imposed on forest management by public law. The question of the strength of human disturbance will be raised below. Due to the extent of the study area, traces of the post-glacial history of the vegetation can be expected, as well as patterns related to the diversity in geology; that is, the parent material for soil formation. There is no explicit temporal information included in this data set. Yet, the fact that young trees (seedlings and shrubs) are distinguished from grown-up trees opens a window on change in time. Variables describing the soil conditions and the geological pattern are as yet scarce. The statistical analysis should reveal whether these are needed to reveal the relationship between vegetation and site factors.
Was this article helpful?