Page 67 - A bird’s-eye view of recreation - Rogier Pouwels

Relating local impacts to regional conservation targets
al., 2011). To determine the location and size of parking areas GIS-data, hiking maps and Google Earth maps were used. During a three-day field trip we validated the input maps by ground truthing. Parking areas near sport fields, company sites and highways were left out as these are not likely to be used as starting point for walking and hiking. To distribute visitors over the parking areas we used the rule-based spatial model FORVISITS. In this model proximity of parking areas and their size are the most influential parameters (De Vries and Goossen, 2002). The number of visitors for each parking area was used as input for the rule-based spatial model MASOOR-SCAN (Jochem et al., 2008). MASOOR-SCAN distributes visitors over the area using a negative exponential algorithm within road compartments, based on the observation that visitors rarely cross asphalt roads (Jochem and Van Marwijk, 2008). This algorithm has been derived from GPS data that revealed that 90 percent of all GPS locations of the visitor presence are within a 3 km radius of the parking area. The output of the model is used as a proxy value for disturbance by recreation: the higher the predicted density of visitors the larger the disturbance.
We used two versions of outputs: one representing overall disturbance in an area and 4 one representing disturbance only around the trail network. For the first version we
used a kernel density method (‘oil spill’) where visitor densities are distributed across
the area regardless of the trail network and for the second version visitors were limited
to a trail network and are assumed to cause disturbance in a zone on both sides of the trail. As visibility and detectability are expected to be important factors regarding disturbance (Fernández-Juricic et al., 2005) we used a disturbance zone of 200 m wide in open landscapes of 100 m wide in woodlands. These disturbance zones are based on flush distances for Woodlark and Stonechat (Krijgsveld et al., 2008). For the second version we also added non-official parking areas as the field trip revealed that crossings of asphalt roads and dirt roads were used as small parking areas (one or two cars).
4.3.2 Determining local impact of visitor densities on birds
4.3.2.1 Bird data
We focus on three heathland breeding birds because they are prone to disturbance: Woodlark (Lullula arborea), Stonechat (Saxicola torquata) and Nightjar (Caprimulgus europaeus). For all these species (f.e. Vos and Peltzer, 1987, Liley and Clarke, 2003, Bijlsma, 2006, Mallord et al., 2007) there is evidence that recreation has a negative effect on breeding success or breeding densities. Bird data for these species are
65