Description

This map represents a classification of forest types based on the influence of the edge effect (distance to the forest edge) and elevation effect (temperature and area) on tree community richness.

• The edge effect influences tree diversity through an environmental aridity filter. In New Caledonia, the maximum temperature recorded at the forest edge is 41°C in February, while it never exceeds 24°C beyond 100 meters from the edge. This temperature difference induces a selection for species that tolerate the most arid conditions, leading to a reduction in the biological richness of tree communities (Ibanez et al., 2017; Birnbaum et al., 2022; Blanchard et al., 2023).
• Altitude also affects tree diversity due to temperature variation and available area (Ibanez et al., 2014; Birnbaum et al., 2015; Pouteau et al., 2015; Ibanez et al., 2016; Ibanez et al., 2018). In New Caledonia, observed tree community richness ranges from 35 to 121 species per hectare within the NC-PIPPN network, peaking at mid-altitude ranges (refer to figure  'amap_elevation_richness.png'). Potential richness was assessed using the S-SDM model, with the 80th percentile used as a threshold to distinguish low and high potential richness across three elevation classes: [0 - 400m[, [400 - 900m[, and [900 - 1628m[.

The classification of forest types combines distance from the forest edge and potential richness by elevation into three major categories, as illustrated in the figure 'amap_forest_types_nc.png':

1. Edge Forest: Parts of the forest located less than 100 meters from the forest edge.
2. Mature Forest: Parts of the forest located beyond 100 meters from the edge with a lower potential richness of tree communities.
3. Core Forest: Parts of the forest located more than 300 meters from the edge with a higher potential richness of tree communities.

Content

The map is computed from the Forest Map of New Caledonia (v2024) and the Potential Tree Species Richness in the Forests of New Caledonia (v2024). This dataset was produced, analyzed, and verified using a combination of open-source software, including QGIS, PostgreSQL, PostGIS, Python, R, and the GDAL library, all running on Linux. 

• amap_forest_types_nc.png is a picture illustrating the forest type classification 
• amap_forest_types_nc.zip is a compressed file contains the six essential files for an ESRI-format GIS system, using the WGS84 international coordinate system, and can be uploaded to a spatial database such as PostgreSQL/PostGIS. Each row of the attribute table represents a forest type (a multi-polygon) with associated fields :

Limitations

We caution users that the distinction between the three classes is based on an ecological interpretation and does not reflect directly perceptible breaks in the forest. The ecological transition from the edge to the core of the forest follows multiple gradient modulated by environmental conditions.

Moreover, this classification is based on local observations and measurements, which are complex to generalize and extrapolate across a territory as environmentally diverse as New Caledonia. Nevertheless, it allows us to address the impact of fragmentation at the scale of New Caledonia.