Sensitivity analysis of the habitat models used in the Heureka planning system
Abstract: Forest management is getting more and more complex. New important considerations like biological diversity, environmental protection and social benefits have to be taken into account. Many novel approaches and initiatives have been prepared and applied in recent times. The Heureka Forestry Decision Support System is one example of this trend being a relatively new analysis program for forest management. One of the purposes with developing this program has been to be able to add new values into the traditional forest management. Issues like biodiversity and recreation are dimensions that traditionally haven’t been able to take in consideration in forest management planning systems. This project focus on the Heureka habitat models where six different indicator species and their ecological requirements act as surrogates for biodiversity. I have, by using sensitivity analysis, tested the performances of three out of six different habitat models used in Heureka with different levels of parameter values. These were the habitat models for Red Squirrel Sciurius vulgaris, Lesser Spotted Woodpecker Dendrocopos minor and a lichen Witch's Hair Alectoria sarmentosa. The first two models worked correctly while it seems like the model for A. sarmentosa had some bug that made my result not useful. In order to study the potential impact of regional characteristics I performed the sensitivity analysis in four regions of Sweden representing the north-south gradient in broad forest types. I have chosen to present the result in Relative Areal Points (RAP). This means that the suitable habitat area according to the model with parameters representing the lowest demands for particular species and particular region always represented 100%. With increasing requirements the habitat areas decreases and are put in relation to the value representing 100% The sensitivity analyses indicate that changes in parameter value may have great consequences in resulting effective habitat area for examined species. In particular change of mean stand age has big effect. Some regional differences can be observed. For S. vulgaris one can see a slight increase of Relative Areal Points the further north one get. For D. minor one could suppose an increasing level of suitable habitat the further southward one got as deciduous proportion is an important factor and one can expect more deciduous component in the forests of south of Sweden than further north. Strangely I observed more of effective habitat area in Västernorrland than in both Örebro and Gävleborg. Generally one can see a more dramatic outcome of changes of parameter values where effective habitat area is low when default values are used. This might indicate that one has to be careful when using the models on small areas. As far as this degree project can indicate the habitat module in Heureka is a reliable tool to try to identify the amount of suitable habitats for the indicator species included but habitat module should be used with understanding of possible effects of the applied parameter values.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)