Leaf attributes and tree growth in a tropical dry forest

Questions: How are leaf attributes and relative growth rate (RGR) of the dominant tree species of tropical deciduous forest (TDF) affected by seasonal changes in soil moisture content (SMC)? What is the relationship of functional attributes with each other? Can leaf attributes singly or in combination predict the growth rate of tree species of TDF? Location: Sonebhadra district of Uttar Pradesh, India. Methods: Eight leaf attributes, specific leaf area (SLA); leaf carbon concentration (LCC); leaf nitrogen concentration (LNC); leaf phosphorus concentration (LPC); chlorophyll concentration (Chl), mass-based stomatal conductance (Gs(mass)); mass based photosynthetic rate (A(mass)); intrinsic water use efficiency (WUEi); and relative growth rate (RGR), of six dominant tree species of a dry tropical forest on four sites were analysed for species, site and season effects over a 2-year period. Step-wise multiple regression was performed for predicting RGR from mean values of SMC and leaf attributes. Path analysis was used to determine which leaf attributes influence RGR directly and which indirectly. Results: Species differed significantly in terms of all leaf attributes and RGR. The response of species varied across sites and seasons. The attributes were positively interrelated, except for WUEi, which was negatively related to all other attributes. The positive correlation was strongest between Gs(mass) and A(mass) and the negative correlation was strongest between Gs(mass) and WUEi. Differences in RGR due to site were not significant when soil moisture was controlled, but differences due to season remained significant. The attributes showed plasticity across moisture gradients, which differed among attributes and species. Gs(mass) was the most plastic attribute. Among the six species, Terminalia tomentosa exhibited the greatest plasticity in six functional attributes. In the step-wise multiple regression, A(mass), SLA and Chl among leaf attributes and SMC among environmental factors influenced the RGR of tree species. Path analysis indicated the importance of SLA, LNC, Chl and A(mass) in determining RGR. Conclusion: A(mass), SMC, SLA and Chl in combination can be used to predict RGR but could explain only three-quarters of the variability in RGR, indicating that other traits/factors, not studied here, are also important in modulating growth of tropical trees. RGR of tree species in the dry tropical environment is determined by soil moisture, whereas the response of mature trees of different species is modulated by alterations in key functional attributes such as SLA, LNC and Chl.