The final model explains less of the total variance
than the climatic factor when regressed separately
against land-cover diversity. This result is due to the
use of penalized splines to reduce the complexity
of the curves and control the estimated degrees of
freedom. To obtain statistical significance, curve
complexity was reduced and the estimated degrees of freedom in the climatic variability variable and
lithological diversity were 1.007 and 1.664, respectively.
Nevertheless, separate regression of the climatic
factor against land-cover diversity produced
an estimated 2.9 degrees of freedom. Accordingly,
the spline is more complex and a greater percentage
of the observed variance is explained by the model