Plant, Cell and Environment (2016) 39,2123–2132 doi: 10.1111/pce.12761 Original Article Hydraulic conductance and the maintenance of water balance in flowers Adam B. Roddy1,2, Craig R. Brodersen2 & Todd E. Dawson1 1Department of Integrative Biology, University of California, Berkeley, CA 94720, USA and 2School of Forestry & Environmental Studies, Yale University, New Haven, CT 06511, USA ABSTRACT been relatively ignored, even though non-pollinator agents of selection, such as physiological costs, often oppose the effects Flowers face desiccating conditions, yet little is known about of pollinators (Galen 1999; Galen et al. 1999; Lambrecht their ability to transport water. We quantified variability in 2013; Strauss and Whittall 2006). floral hydraulic conductance (Kfl )for20speciesfrom10 ower Like leaves, flowers are terminal structures often located in families and related it to traits hypothesized to be associated the hottest, driest parts of the plant canopy, but flowers and with liquid and vapour phase water transport. Basal angio- leaves have experienced different selective forces. Leaves are sperm flowers had trait values associated with higher water a critical component in the plant hydraulic pathway and have and carbon costs than monocot and eudicot flowers. Kfl ower alargeinfluence on the terrestrial water cycle (Hetherington was coordinated with water supply (vein length per area, and Woodward 2003). Because carbon assimilation is mecha- VLA) and loss (minimum epidermal conductance, g ) traits min nistically linked to plant transpiration (Brodribb and Feild among the magnoliids, but was insensitive to variation in these 2000; Brodribb et al. 2007; Sack and Holbrook 2006), moving traits among the monocots and eudicots. Phylogenetic inde- water from the roots to the leaves requires coordination in pendent contrast (PIC) correlations revealed that few traits the structural traits governing water flow through each compo- had undergone coordinated evolution. However, VLA and nent in this pathway to prevent declines in water content that the desiccation time (T ), the quotient of water content and des could irreversibly prohibit hydraulic function (e.g. cavitation; g ,hadsignificant trait and PIC correlations. The near min Drake et al. 2015; Skelton et al. 2015). Angiosperm leaves are absence of stomata from monocot and eudicot flowers may particularly capable of efficiently transporting water because have been critical in minimizing water loss rates among these of high leaf vein densities (vein length per area, VLA) that clades. Early divergent, basal angiosperm flowers maintain move water closer to the sites of evaporation (Brodribb et al. higher Kfl because of traits associated with high rates water ower 2007; Brodribb et al. 2010; Buckley 2015; Feild and Brodribb loss and water supply, while monocot and eudicot flowers 2013), minimize changes in leaf water content (Noblin et al. employ a more conservative strategy of limiting water loss 2008; Zwieniecki and Boyce 2014) and, ultimately, increase and may rely on stored water to maintain turgor and delay growth rates (Berendse and Scheffer 2009). The fundamental desiccation. constraint of maintaining water balance by matching liquid wa- ter supply to vapour loss determines the range of possible leaf Key-words: Angiosperms; flower; hydraulic conductance; vein designs, from the organization of cells within leaves to the density; water balance. shapes and sizes of leaves themselves (John et al. 2013; Li et al. 2013; Sack et al. 2008; Sack et al. 2012). fl INTRODUCTION Unlike leaves, owers are relatively ephemeral and assimi- late little carbon (but see Galen et al. 1993 for an important ex- Flowers are often considered the hallmark of angiosperm evo- ception) although they still may transpire significant amounts lution. Their appearance enabled the evolution of more inti- of water (Roddy and Dawson 2012; Teixido and Valladares mate, specialized interactions with animal pollinators than 2014). Instead, they promote pollen dispersal with animal- had previously been possible (Crepet and Niklas 2009; Darwin and wind-pollinated flowers differentiating along a spectrum 1888; Fenster et al. 2004; Kölreuter 1761; Sprengel 1793). De- of carbon and water investment. Because flowers often face spite the influence of flowers on plant-pollinator networks desiccating conditions that would lead to wilting and prevent (Memmott and Waser 2002), the generation and maintenance successful pollination, they must maintain water balance and of biodiversity (Dodd et al. 1999; Stebbins 1970), and ecosys- turgor throughout anthesis to attract pollinators. At least part tem services (Costanza et al. 1997), the fundamental relation- of this hydraulic pathway may need to remain functional after ships between floral structure and physiological function have pollination to facilitate proper fruit and seed development. Understanding how flowers do this is fundamentally important Correspondence: A. B. Roddy. Tel: +1 510.224.4432; e-mail: adam. to understanding non-pollinator agents of selection and, by [email protected] extension, floral evolution. ©2016JohnWiley&SonsLtd 2123 2124 A. B. Roddy et al. Even basic information about flower water relations, such as regulate water loss rates, such as stomatal density and size the mechanisms used to deliver water, are not clear. Basal and epidermal conductance to water vapour. Positive correla- angiosperm flowers from the ANA grade (Amborella, tions between Kflower and other hydraulic traits would suggest Nymphaeales, Austrobaileyales) and magnoliids seem to rely that these traits are important in determining the efficiency of predominantly on continuous delivery of water by the xylem water flow through flowers. We further predicted that positive (Feild et al. 2009a, 2009b). Such a strategy could be costly correlations between Kflower and hydraulic traits, particularly because it would require a vascular system capable of meeting xylem traits, should exist for magnoliid flowers because flowers potentially large transpiration demands and risky because it of two of these genera, Magnolia and Calycanthus, have been would require flower water potential to decline diurnally with shown to maintain a hydraulic connection to the stem xylem stem water potential. In contrast, some eudicot flowers and throughout anthesis (Feild et al. 2009b; Roddy et al.inprep.). petals tend to have higher, less negative water potentials than If flowers from the eudicots, and possibly also the monocots, subtending bracts and leaves (Chapotin et al. 2003; Trolinder rely more heavily on a different source of water during anthesis et al. 1993). These ‘reverse’ water potential gradients imply that (e.g. either delivery by the phloem or depletion of stored for these flowers to remain hydrated, water must be imported water), then they should not exhibit positive correlations against an apoplastic water potential gradient in the xylem. between xylem traits and Kflower. Given that much of the These authors have concluded that flowers are hydrated pre- variation in traits may be due to differences among clades, dominantly by the phloem, which is primarily responsible for we used correlations of phylogenetic independent contrasts the transport of photosynthates throughout the plant. In con- (PICs) to test whether pairs of traits co-evolved (Felsenstein trast to the xylem, the phloem has much higher hydraulic resis- 1985). Co-evolution of traits would further imply a functional tance and lower water flux rates because phloem transport link. occurs symplastically (Münch 1930; Nobel 1983; Savage et al. 2016; Windt et al. 2009). However, implicating the phloem as the sole source of floral water is not necessary to explain the re- MATERIALS AND METHODS sults from these studies; flowers can be xylem-hydrated and still Plant material maintain higher water potentials than leaves so long as they are fl more negative than the stem xylem (e.g. Feild et al. 2009a, We collected owering shoots from around the University of 2009b), or they could rely on large amounts of water imported California, Berkeley campus and from the University of early in development (e.g. during bud expansion) and stored California Botanic Garden during the springs of 2013 and locally (i.e. hydraulic capacitance; Chapotin et al. 2003) that 2014, and from the Marsh Botanical Garden, New Haven, could be depleted throughout anthesis. Regardless of whether CT, in the spring of 2015. All plants had been kept well- flowers have high hydraulic capacitance or rely on water deliv- watered. We chose a phylogenetically diverse set of species that fl ered by the phloem, maintaining a higher water status in varied by almost two orders of magnitude in oral display size flowers could result in water being drawn back into the stem (Table 1). These species also varied morphologically, from fl during the day, as has been shown to occur in fleshy fruits owers with undifferentiated perianths to those with a fully (Higuchi and Sakuratani 2006). The supposed dichotomy differentiated calyx and corolla and from those with free petals between xylem-hydration and phloem-hydration in flowers is to those with sympetalous connation. Additionally, we fl fl likely not a dichotomy after all but rather a spectrum between included in orescences of Cornus orida (Cornaceae), which fl more or less contributions from the phloem. In fruits, there has have small, inconspicuous owers but large, white bracts as fl been a similar debate about the contribution of the phloem their showy organs. Although the showy oral structures of