ORIGINAL ARTICLE doi:10.1111/evo.12768 Evolution of the leaf economics spectrum in herbs: Evidence from environmental divergences in leaf physiology across Helianthus (Asteraceae) Chase M. Mason1,2 and Lisa A. Donovan1 1Department of Plant Biology, University of Georgia, Athens, Georgia, 30602 2E-mail: [email protected] Received January 31, 2015 Accepted August 23, 2015 The leaf economics spectrum (LES) describes a major axis of plant functional trait variation worldwide, defining suites of leaf traits aligned with resource-acquisitive to resource-conservative ecological strategies. The LES has been interpreted to arise from leaf-level trade-offs among ecophysiological traits common to all plants. However, it has been suggested that the defining leaf- level trade-offs of the LES may not hold within specific functional groups (e.g., herbs) nor within many groups of closely related species, which challenges the usefulness of the LES paradigm across evolutionary scales. Here, we examine the evolution of the LES across 28 species of the diverse herbaceous genus Helianthus (the sunflowers), which occupies a wide range of habitats and climate variation across North America. Using a phylogenetic comparative approach, we find repeated evolution of more resource-acquisitive LES strategies in cooler, drier, and more fertile environments. We also find macroevolutionary correlations among LES traits that recapitulate aspects of the global LES, but with one major difference: leaf mass per area is uncorrelated with leaf lifespan. This indicates that whole-plant processes likely drive variation in leaf lifespan across Helianthus, rather than leaf-level trade-offs. These results suggest that LES patterns do not reflect universal physiological trade-offs at small evolutionary scales. KEY WORDS: Climate, leaf lifespan, photosynthesis, soil fertility, sunflower. A central question in evolutionary biology concerns why organ- and stress tolerance (Westoby et al., 2002; Diaz et al., 2004; isms have evolved certain forms rather than others, and why cer- Reich et al., 2014). One such spectrum that has received much at- tain regions of phenotypic space are crowded while others remain tention over the past two decades is the leaf economics spectrum unfilled (Raup et al., 1966; Pigliucci et al., 2007). Most expla- (LES), characterizing global covariation in key leaf ecophysio- nations include some combination of the effects of selection, ge- logical traits (Reich et al., 1997; Wright et al., 2004). Built upon netic constraints, and trade-offs among traits that generate the similar concepts dating back much further (e.g., Grime et al., repeated evolution of phenotypes with specific trait combinations 1977; Chapin et al., 1980; Bloom et al., 1985; Bloom et al., 1986; (Pigliucci et al., 2007; Donovan et al., 2011). With respect to Chapin et al., 1987), variation in the LES has been shown to functional traits (those directly related to growth, survival, and govern resource acquisition and loss (Eckstein et al., 1999; Mc- fitness), phenotypes are often found to be distributed along spec- Murtrie et al., 2011), growth rate (Poorter et al., 1990; Reich et al., tra that define variation in ecological and life-history strategies 1997; Poorter and Garnier, 1999; Sterk et al., 2006), stress toler- (Reich, 2003; Laughlin et al., 2014). Land plants exhibit func- ance (Hallik et al., 2009; Nardini et al., 2012), ecosystem services tional trait differentiation across many such spectra, including like productivity and decomposition (Reich et al., 2012; Bakker overall plant size, reproductive timing and output, and growth et al., 2011; Freschet et al., 2012), and to provide key links to C 2015 The Author(s). 1 Evolution C. M. MASON AND L. A. DONOVAN environmental gradients (Reich, 1999; Wright et al., 2005; in a range of trait space such that key LES relationships do not Ordonez et al., 2009; Ordonez et al., 2010). apply, in particular the central relationship between LMA and The LES is defined by six leaf traits: leaf lifespan, leaf mass leaf lifespan (Diemer et al., 1992). These early concerns about per area (LMA), rates of photosynthesis and respiration, and con- the relevance of LES trade-offs to herbs were largely dismissed, centrations of nitrogen and phosphorus (Reich et al., 1997; Wright with appeals to consider the full global range of variation in LES et al., 2004). These leaf traits strongly covary along a single ma- traits as the stage upon which fundamental trade-offs play out jor axis at global scales, forming a spectrum of trait combinations (Reich et al., 1993). While this view works conceptually when that runs from species with high photosynthetic and respiration thinking about how the leaves of vascular plants have diversified rates, high leaf nutrient contents, low LMA, and short leaf lifes- in trait space over the past 400 million years, broad divergences pan, to species with low photosynthetic and respiration rates, low between phenotypically disparate clades dominate such analy- leaf nutrient contents, high LMA, and long leaf lifespan. This ses. This means that the trait correlations that hold across global spectrum represents a continuum of strategies of leaf carbon in- datasets may not be particularly useful nor predictive for our un- vestment and return, ranging from resource-acquisitive strategies derstanding of leaf trait evolution at many scales to which the that produce low-investment, high-productivity leaves that persist LES paradigm has been increasingly applied—evolution among for only a short period, to resource-conservative strategies that closely related species (e.g., Dunbar-Co et al., 2009; Santiago produce high-investment, low-productivity leaves that persist for and Kim, 2009; Palow et al., 2012; Edwards et al., 2014; Muir much longer. Strategies at both ends of the spectrum result in et al., 2014), among populations and genotypes within species net-positive carbon return on investment over the lifespan of the (e.g., Martin et al., 2007; Grady et al., 2013; Brouillette et al., leaf, but reflect differentiation between high and low resource 2014; Niinemets, 2015), predicting genetic correlations and the use and tissue turnover rates (Chabot and Hicks, 1982; Reich response to selection among traits (e.g., Donovan et al., 2011; et al., 2003). As such, the LES largely represents a leaf-level Vasseur et al., 2012; McKown et al., 2014), understanding do- trait-based approach to quantifying and explaining whole-plant mestication and crop evolution (e.g., Cornwell and Cornelissen, differences in growth rate and stress-tolerance strategies identi- 2013; Garcia-Palacios et al, 2013), and predicting responses to fied many decades ago (e.g., Grime et al., 1977; Chapin et al., climate change (e.g., Gornish and Prather, 2014), all scales in 1980). which leaf trait evolution is unlikely to play out over the full The covariation of traits along the LES has been interpreted to global range of LES variation. be driven by leaf-level physiological dependencies and trade-offs Herbs constitute a functional group with huge ecological between productivity and persistence (Reich et al., 1997; Reich importance, diversifying into niches in all biomes and forming et al., 1999; Wright et al., 2004). For instance, high photosyn- the foundation of many ecosystems. Growth rate and life-history thetic rate has been inferred to require high nutrient concentrations strategies vary widely within herbaceous plants, and recent and high respiration rates (to support RuBisCo, chlorophyll, and phylogenetically informed estimates indicate that between 52 general metabolism), as well as low LMA (because thick/dense and 55% of vascular plant species are herbaceous (FitzJohn et al., leaves are prone to within-leaf shading of chloroplasts and limita- 2014). If the LES can be called “universal” or “fundamental” tions on internal CO2 diffusion) (Field and Mooney, 1986; Reich to vascular plants, it must apply within herbaceous plants. et al., 1997). At the other extreme, long leaf lifespan has been Additionally, if the LES truly reflects universal leaf-level inferred to require high LMA (as higher structural support is trade-offs, LES relationships should hold at small evolutionary needed to defend against biotic and abiotic damage), and high scales, such as among species within genera adapted to diverse LMA constrains photosynthetic rate and nutrient concentrations environments. (Reich et al., 1999; Wright et al., 2004). These trade-offs have Here, we examine the evolution of the LES under broad envi- been demonstrated to exist among plants within communities ronmental divergence in the diverse herbaceous genus Helianthus worldwide (Reich et al., 1999), across plants of diverse growth (the sunflowers). Members of the genus occur across North Amer- forms and life histories (Reich et al., 1997; Wright et al, 2004), ica, with the highest concentration of species found within the and within well-studied functional groups like evergreen trees and contiguous United States (Heiser et al., 1969). Sunflowers have shrubs (Wright et al., 2002; Wright and Westoby, 2003). They are radiated into a wide variety of habitats, including deserts, wet- considered so core to plant variation as to be referred to as “uni- lands, grasslands, forests, coastal dunes, and rock outcrops, and versal” and “fundamental” (Wright et al., 2004; Shipley et al., possess a concomitant broad diversity in growth form, life history, 2006). and leaf traits (Fig. 1). The broad environmental and trait diversity However, early work on LES traits suggested that herbaceous present within the genus makes Helianthus an excellent system plants might not experience all of the same leaf-level trade-offs in which to study leaf trait evolution in herbs. Using a phyloge- as woody taxa, and might operate ecologically and evolutionarily netic comparative approach, here we test whether predicted LES 2 EVOLUTION 2015 EVOLUTION OF LEAF ECONOMICS IN DIVERSE SUNFLOWERS Figure 1. A selection of leaf and growth form diversity within Helianthus. Top row: H. radula, H. salicifolius,andH. divaricatus. Second row: H. maximiliani, H. grosseserratus, H. porteri. Third row: H.