Figure 1.
Location and geographic context of villages sampled in primary and secondary surveys in Indonesian and Malaysian Borneo.
(a) All villages surveyed, overlaid with elevation information (DEM at 1 km2 resolution). (b) Villages with the most reliable information, overlaid with the 2010 land cover classes used in spatial analyses.
Table 1.
Outline of the eight land use/land cover classes used in the spatial analysis.
Figure 2.
Perceptions on the Importance of Forest for Health and for Cultural and Spiritual Benefits expressed as percentages of the total of respondents’ answers.
(a) Importance of Forest for Health. (b) Cultural and Spiritual Benefits. Analyses are based on the full data set.
Table 2.
Dominant items in the forest perception indices, based on Analysis of Variance (for further details see Text S1).
Table 3.
Pearson correlations (r) between indices, with moderately large values in bold,: (1) Direct economic uses; (2) Other forest uses; (3) Environmental health benefits; (4) Direct health benefits; (5) Ecosystem services (sum of indices 3,4 and Cultural and spiritual importance); (6) Advantages of small-scale clearing; (7) Advantages of large-scale clearing; (8) Disadvantages of large-scale clearing.
Table 4.
The ten most important explanatory variables for the Direct economic uses, Other forest uses, Cultural spiritual importance, Importance for health, and Environmental health benefits indices in the BRT analysis, showing the explanatory variables in order down each column with their relative importance.
Table 5.
The ten most important explanatory variables for the Direct health benefits, Ecosystem services, Advantages small scale clearing, Advantages large scale clearing, and Disadvantages large scale clearing indices in the BRT analysis, showing the explanatory variables in order down each column with their relative importance.
Figure 3.
Observed versus predicted values for four indices, based on BRT analyses.
Figure 4.
Regression trees for Direct economic uses and Other forest uses, based on CART analyses.
(a) Direct economic uses. (b) Other forest uses. Each classification shows the value or threshold for taking the branch on the left side of the split below it. Values at the tips of the trees indicate the mean value of the index for that group of responses. The classifications are: Christian (% Christians in the village), Province (a = Central Kalimantan, b = East Kalimantan, c = Sabah, d = West Kalimantan), Rice and Hunting (a = no, b = yes as a predominant activity of the village), FT (index of time spent in forest); Population (population of the village).
Figure 5.
Regression tree for advantages and disadvantages of large-scale clearing, based on CART analyses.
(a) Advantages of large-scale clearing. (b) Disadvantages of large-scale clearing. Classifications are explained in Figure 3 caption. Other classifications are oil palm3 (% area in a 3 km radius of the village that is oil palm), similarly for agroreg3 (agro-forests, forest regrowth, 3 km radius) and agroreg (10 km radius).
Figure 6.
Variation in index scores by geographic region (Province or State).
Regions are Central Kalimantan (light blue), East Kalimantan (red), Sabah (green), West Kalimantan (purple). Indices are (1) Direct economic uses, (2) Other forest uses, (3) Cultural and spiritual importance, (4) Importance for health, (5) Environmental health benefits, (6) Direct health benefits, (7) Ecosystem services, (8) Advantages of small-scale clearing, (9) Advantages of large-scale clearing, (10) Disadvantages of large-scale clearing.