Contrasting leaf trait scaling relationships in tropical and temperate wet forest species

We investigated whether plants adapted to thermally contrasting environments (e.g. tropical-temperate habitats) exhibit inherent differences in leaf trait scaling relationships. Thirteen tropical and 12 temperate species (all characteristic of wet forests) were grown in a glasshouse (25/20 degrees C day/night). A range of leaf traits were quantified, including mass-based leaf nitrogen [N], mass per unit area (LMA), light-saturated photosynthesis (A) and respiration (Rdark). Average area- and mass-based rates of net CO2 exchange were higher in the temperate species, compared to their tropical counterparts. Average leaf [N] and LMA values were also higher in temperate species than in their tropical counterparts. The higher LMA in the metabolically more active temperate species was the most striking contrast to the patterns and predictions of the GLOPNET leaf trait data base, and was associated with different elevations (i.e. y-axis intercepts) but not slopes of bivariate trait scaling relationships. As expected, mass-based rates of A and Rdark scaled positively with increasing [N] and negatively with increasing LMA in both tropical and temperate species. No differences were found between temperate and tropical species groups in terms of log-log scaling relationships linking A and Rdark to N. However, at any given LMA, mass-based values of [N], A and Rdark were all higher in the temperate species than in their tropical counterparts. Underpinning higher A in temperate species was a higher capacity for carboxylation (Vcmax) and RuBP regeneration (Jmax), with Jmax:Vcmax being greater in temperate species. In conclusion, our results suggest that as a consequence of greater overall N investment as well as greater proportional N investment in metabolic capacity, cool-adapted temperate wet forest species exhibit higher photosynthetic and respiration rates than their warm-adapted tropical counterparts when compared in a common environment