Page 36 - A bird’s-eye view of recreation - Rogier Pouwels

A bird's-eye view of recreation
Function 2
Nv Predicted number of visitors present at a path segment at distance d (per day or per year) d distance to car park (m)
Vcp number of visitors starting at a specific car park (per day or per year)
pd number of path (segments) at a specific distance class
1.0 0.8 0.6 0.4 0.2 0.0
0 1,000 2,000 3,000
Distance from car park; as crow flies (m)
Figure 7. Fraction of single data points at a certain distance from a cark park. The parameter values of Function 1 are 2.80 for α and 965 m for H.
2.5 Discussion and implications for recreation management
2.5.1 Practical implications
In this paper we show that random forest models are suitable for modelling the complex interaction between different landscape and environmental features to explain visitor densities in nature areas. A random forest model was used as a tool to assess the impact of potential interventions on visitor densities and consequently on a population of a target species, the Nightjar, in the New Forest, UK. We focused on reallocating visitors, but interventions such as changes to path type or vegetation type could also be assessed. Although the GPS data only covered one third of all car parks, we believe that the data are representative of all car parks in the area and so the model predicts visitor densities for the whole area (Fig. 5). Random forest models based on GPS monitoring data are particularly useful in areas where managers need tools to estimate visitor densities and relate them to social or ecological thresholds. Managers could use these tools in decision-making processes with stakeholders to discuss and find support for potential interventions.
34
GPS data Function 1
4,000 5,000
Fraction of single data points