Page 35 - A bird’s-eye view of recreation - Rogier Pouwels

Developing tools and rules of thumb for managers
times higher in areas on the same side of the road as the car park than on the opposite
side of the road. Managers could use this rule of thumb to change visitor densities
by relocating car parks. These interventions might reduce visitor densities by 80% in
areas that are sensitive to disturbance without restricting visitor use completely. The
presence of dogs might even be reduced by 90%. The second rule of thumb concerns
the interaction between tarmac roads and openness. In woodlands the impact of 2 tarmac roads on visitor densities is less distinct (around 78% reduction), while in open
landscapes, like heathlands, the impact is larger (around 95% reduction). Combining both rules of thumb shows that managers might be able to reduce visitor densities by up to 95% by relocating a car park from one side of the road to another in open landscapes.
Results from the exploratory data analysis and the random forest model give a reliable estimate of how visitor densities decline with increasing distance from the car park. We used the frequency distribution of GPS locations, the single data points (Fig. 3), to derive a simple algorithm that estimates the number of visitor groups at a specific path segment. First, we chose an algorithm that describes the sigmoid declining curve and fitted the parameters for the correlation of single data points. This curve represents the probability that a visitor group is present at a specific distance (Function 1; Fig. 7). Next, we multiplied this by the number of visitor groups starting at a specific car park, taking into account that visitor groups will be present at a specific distance twice: when they enter the area and when they return to the car park. Finally, the number of visitor groups was divided by the number of paths segments at a specific distance class to account for a potential unevenness in path density over distance (Function 2). Managers can use Function 2 to acquire a first estimate of the number of visitors at a specific location (Nv). The parameters needed are quite easy to collect and are (1) the distance to the car park of interest, (2) the number of visitors that use the car park, and (3) the density of the path network around the car park. For locations within 5 km of more than one car park, the algorithm should be applied for each car park separately and the number of visitors per path segment should be summed.
Function 1
fsdp fraction of single data point at distance d
d distance to car park (m)
H parameter at which visitor presence is 50% (m); 965 in Fig. 7
α parameter determining the rate at which visitor presence declines; 2.80 in Fig. 7
33