Skip to main content
. 2013 Mar 18;8(3):e58878. doi: 10.1371/journal.pone.0058878

Table 3. Multiple regressions predicting intraspecific variation in six plant functional traits from five environmental variables in four Nothofagus species.

Trait Species MAT Elevation Latitude MAR Soil P
Wood density N. solandri –ve –ve –ve NS NS
N. menziesii NS NS –ve NS NS
N. fusca NS NS NS +ve NS
N. truncata +ve +ve +ve NS –ve
Leaf size N. solandri +ve +ve +ve NS NS
N. menziesii +ve NS NS –ve NS
N. fusca NS NS –ve NS +ve
N. truncata –ve –ve –ve +ve +ve
LMA N. solandri –ve NS –ve +ve NS
N. menziesii NS +ve NS NS –ve
N. fusca NS NS NS NS –ve
N. truncata +ve NS +ve NS NS
LDMC N. solandri –ve NS –ve +ve NS
N. menziesii NS +ve –ve NS NS
N. fusca NS NS NS NS –ve
N. truncata +ve NS +ve NS NS
Leaf thickness N. solandri NS +ve NS +ve +ve
N. menziesii +ve +ve +ve NS –ve
N. fusca –ve –ve –ve –ve –ve
N. truncata NS –ve –ve NS NS
Leaf density N. solandri –ve –ve –ve NS –ve
N. menziesii –ve NS –ve NS +ve
N. fusca +ve +ve +ve +ve NS
N. truncata +ve +ve +ve NS NS

A full regression model with all five environmental variables was run for each trait and species. This model was reduced to significant terms through backwards selection. The direction of significant terms in each model is shown. Non-significant terms, removed from each model, are shown with NS. All trait data were corrected for variation in tree size (see Methods) and log10-transformed