What determines interspecific variation in relative growth rate of Eucalyptus seedlings?

The present study examines relative growth rate (RGR) and its determinants in seedlings of nine Eucalyptus species. Species were selected from mesic (1,800 mm a(-1) rainfall) through to semi-arid habitats (300 mm a(-1)), and thus, notionally vary in "stress" tolerance. Seedlings were grown in a glasshouse during early summer and received between 33 mol and 41 mol PAR m(-2) day(-1). The mean RGR varied among species-from a minimum of 66 mg g(-1) day(-1) in E. hypochlamydea to a maximum of 106 mg g(-1) day(-1) in E. delegatensis. RGR was positively related to rainfall at the sites of seed collection. Neither specific leaf area (SLA) nor net assimilation rate was related to rainfall or RGR. While the absence of relationships with SLA and net assimilation rate contrasts with other studies and species, we cannot rule out the effects of sample size (n=9 species) and modest ranges in SLA and RGR. The ratio of leaf mass to total mass (LMR) varied from 0.49 +/- 10.07 g g(-1) in E. socialis to 0.74 +/- 0.04 g g(-1) in E. delegatensis and was strongly positively related with rainfall (r(2)=0.77). Interspecific differences in RGR were strongly related to LMR (positive relationship, r(2)=0.50) and the rate of dry matter production per mol of leaf nitrogen (positive relationship, r(2)=0.64). Hence, the slow RGR of low-rainfall species was functionally related to a lower growth rate per mol of leaf nitrogen than high-rainfall species. Furthermore, slow RGR of low-rainfall species was related to greater allocation to roots at the expense of leaves. Increasing allocation to roots versus leaves is likely an adaptation to soil and atmospheric water deficits, but one that comes at the expense of a slow RGR.