|
| This study assesses the factors that influence land-cover diversity, including the specific
contributions of environmental and anthropogenic forces in determining landscape
diversity (spatial variability in climate, lithological variations and human management).
The proposed model was tested in Navarra (northern Spain), a region with a long history
of human settlement and distinct management practices, ranging from mountain communities
in the Pyrenees to Mediterranean lowland cropland systems. Variance in landscape
diversity was divided into environmental- and human-influenced fractions, and generalized
additive models within a GIS framework were used to evaluate the effects of environmental
and anthropogenic factors. We also assessed the influence on the results of the number
of land-cover classes by employing contrasting thematic resolutions of 220 and 24 classes.
The model that includes only environmental factors, using 24 classes, accounted for 65
per cent of the total variance (P < 0.005). The residuals obtained from this model were then
regressed against human variables (distance to settlements, accessibility, total municipality
population and productive specialization). Residuals and productive specialization,
which varies between areas devoted to grazing and forest exploitation in the Pyrenees
mountains to lowland Mediterranean croplands, showed a strong correlation (r
2
= 0.40;
P < 0.005), but weak correlations were found when 220 land-cover classes were used.
Results based on the sign of residuals suggest that human activities have resulted in an
increase in land-cover diversity in mountainous areas and have acted to homogenize
land cover within the Mediterranean agricultural landscape. In summary, the model that
uses the map of 24 classes as dependent variables and includes human and environmental
factors explains 80.75 per cent of total land-cover diversity variance using only three
variables: climatic spatial variability, lithological diversity and productive specialization. | |
|