In conservation biology and conservation planning there is a great diversity of GIS-based species distribution, habitat or population models (Blaschke, 1997; Blaschke, 2003; Taeger, 2010; Guisan & Zimmermann, 2000; Gontier, 2007; Gontier et al., 2010; Pietsch et al., 2007; Amler et al., 1999). Habitat suitability models based on empirical data versus models based on expert knowledge can be distinguished. On the other hand the level of detail (e. g. individuals, populations, species occurrences or species communities) is another way to describe the different models (Gontier et al. 2010).
There has been a lot of discussion about the possibilities to implement habitat suitability analysis (HSI) in environmental and landscape planning (Kleyer et al., 1999/2000; Schroder, 2000; Blaschke, 1999 and 2003; Rudner et al., 2003). They are established in environmental – and bio-science but because of the data requirements and the time – and cost-consuming modeling used only in a few planning examples (Joofi, 2003 and 2005; Pietsch et al., 2007; Gontier, 2007; Rudner et al., 2004; Schroder, 2000).
GIS-based models based on expert knowledge normally use presence datasets of specific species. Using the knowledge about the habitat preferences it’s possible to analyze the suitability. Actual land use maps or other thematic information about habitats and specific structures or qualities (e. g. hydrological situation, soils, water quality) are used to evaluate the actual situation (Blaschke, 1997; Joofi, 2003; Taeger 2010). In contrast to ecological models using statistical methods models based on expert knowledge have a great potential to be used in landscape planning (see table 1). They are not as precise as ecological models but easier to interprete and applicable in larger areas.
|
ecological model |
"planning" model |
|
|
method |
– statistical approach |
– expert knowledge – prediction model |
|
datasets |
– actual presence data (based on field work) – (knowledge about habitat preferences) |
– knowledge about habitat preferences |
|
criteria / habitat information |
– very detailed – for a small area |
– more general – based on existing planning information – depends on the expert knowledge |
|
quality/validity |
– precise models, but only usable for a small area – scientific models |
– fuzzy models, but for larger areas |
|
validation |
Easy to validate |
Validation based on control samples |
|
usability |
– precise model for one specific species – maximum quality |
– much more general, but useful for planning tasks – cumulative models for several species – scoping for species research |
|
Table 1. Types of habitat models (adapted from Joofi, 2003) |
They can be used for several species using existing species information. Based on the prediction model future conditions and different scenarios can be simulated to evaluate the impact of land use changes in the planning process (Gontier et al., 2010; Taeger, 2010; Pietsch et al., 2007; Blaschke, 1997). The visualization of future habitat suitability is possible. GIS offers the capability to create models (Fig. 7) based on existing datasets (species, land use, previous impact, structures, qualities) to analyze the actual and future suitability.
It’s possible to create evaluation models, to create scenarios to improve the situation for one specific or several species, to develop measures to reduce negative impacts or create new habitats (Hunger, 2002; Hennig & Bogel, 2004) and they are useful to evaluate the negative impact of a plan, project or program in the context of an SEA or EIA (Gontier, 2007; Pietsch et al, 2007; Lang & Blaschke, 2007; Blaschke, 1999). In combination with connectivity analysis (see chapter 3.2.3) physical and functional links in ecological networks can be examined.
|
Fig. 7. Example for a habitat suitable model (Schmidt, 2007) |
