The structural and compositional basis of leaf traits and their influence on plant water use and drought tolerance

Plants provide invaluable ecosystem services to urban centers. Yet, these services cannot be decoupled from the cost to maintain landscaping plants. Costs are particularly difficult to quantify in urban ecosystems, like Los Angeles, where a combination of mild climate and a legacy of landscape design driven by social and economic factors has produced unprecedented species diversity. As these plants subsist entirely on irrigation water, the resulting ‘global common garden’ is ideally suited for testing anatomical and physiological scaling hypotheses across many diverse species to inform generalized predictive models for other species and ecosystems. My dissertation seeks to uncover relationships between leaf anatomy and function to build models estimating water use and drought tolerance in Los Angeles.I first analyzed allometric slopes describing relationships between the dimensions of leaf cells and tissues and found cell sizes, cell wall thickness and leaf thickness scale together across tissues and species. I then created a mathematically explicit model that predicted leaf mass per area (LMA) from its constitutive anatomy and composition with unprecedented explanatory power. Cell size, the number of mesophyll cell layers, and cell mass density principally drove species differences in LMA and had strong impacts on the Leaf Economics Spectrum. I linked leaf structure with physiology by resurrecting a classic method for quantifying loss of rehydration capacity in dehydrating leaves, placed associated thresholds in the sequence of drought response (e.g. stomatal closure, turgor loss), and tested their relationship with structure and composition. I found leaves to be considerably more vulnerable than previously believed. This prompted a study in which I tested canopy stomatal conductance across minor shifts in vapor pressure deficit (VPD) modeled from plant traits against empirical sap flux and leaf area data collected in urban trees. Finally, I found enormous variation in functional traits across Los Angeles species, influenced by climates of origin, growth forms, leaf habit and canopy position (i.e., sun versus shade leaves). The vast trait diversity in Los Angeles urban trees demonstrates the need for integration of species information into predictive models in urban ecosystems and highlights Los Angeles as an invaluable resource for future ecophysiological study