Abstract
Extreme drought conditions across the globe are impacting biodiversity with serious implications for the persistence of native species. However, quantitative data on drought tolerance is not available for diverse flora to inform conservation management. We quantified physiological drought tolerance in the diverse Hakea genus (Proteaceae) to test predictions based on climatic-origin, life history and functional traits. We sampled terminal branches of replicate plants of 16 species in a common garden. Xylem cavitation was induced in branches under varying water potential (tension) in a centrifuge and the tension generating 50% loss of conductivity (stem P50) was characterized as a metric for drought tolerance. The same branches were used to estimate plant functional traits, including wood density, specific leaf area, and Huber value (sap flow area to leaf area ratio). There was significant variation in stem P50 among species, which was negatively associated with the species climate-origin (rainfall and aridity). Drought tolerance did not differ among life histories; however, a drought avoidance strategy with terete leaf form and greater Huber value may be important for species to colonize and persist in the arid biome. Our findings will contribute to future prediction of species vulnerability to drought and adaptive management under climate change.
Keywords: Aridity, climate change, drought tolerance, life-history, functional traits, Proteaceae.
Acknowledgement : We appreciate the support of Benedict Lyte from The Royal Botanic Garden, Sydney for granting us excess to their living collections. Nzie Peter for support in data collection and Rosana Lopez for technical insights into the hydraulic techniques. Australian Postgraduate Award (Western Sydney University), Ecological Society of Australia-Holsworth Wildlife Research Endowment grant to O.O.O. This project has been supported by the New South Wales Government’s Department for Planning, Industry and Environment, Saving Our Species grant to P.D.R and D.T.

Introduction

The impacts of drought on diverse biomes across the globe are substantial, with prolonged drought resulting in forest dieback and plant mortality, changes in species distribution, local extinction and decline in ecosystem function and resilience (Allen et al., 2010; Goulden and Bales, 2019; Powers et al., 2020). Predicting the impacts of drought on biomes and plant lineages remains a challenging task for scientists, as most predictions relying on species distribution models (SDM) and climatic niche data lack the species physiological tolerance (Fitzpatrick et al., 2008; McDowell et al., 2008; Razgour et al., 2019; Urban, 2015). Hence, quantifying species physiological thresholds is key to understanding how plants will cope with extreme climatic-induced events such as drought in the future (Allen et al., 2010).
One promising strategy to quantify physiological tolerance to drought is by characterizing hydraulic traits in relation to water limitation (Choat et al., 2012; Martin-StPaul et al., 2017). This is particularly important, as studies have shown that most flowering plants (angiosperms) function close to their hydraulic safety margin (minimum xylem pressure experienced in the field - water potential causing 50% loss of conductivity (P50)), and are vulnerable to climate change (Choat et al., 2012). Under prolonged drought conditions, stomatal closure is unable to prevent the continuous decline of the xylem pressure, leading to cavitation, a phase change from liquid water to gas, and the formation of gas emboli (Choat et al., 2012). This results in loss of xylem hydraulic conductivity, and in severe cases, hydraulic failure and subsequent mortality (Anderegg et al., 2016; Brodribb and Cochard, 2009; Cochard, 2014; McDowell et al., 2008; Pockman et al., 1995; Urli et al., 2013). A large body of evidence has shown species drought tolerance is quantitatively linked with resistance to cavitation in woody species (Adams et al., 2017; Brodribb and Cochard, 2009; Choat et al., 2012; Kursar et al., 2009; McDowell et al., 2008; Pockman et al., 1995). The xylem tensions associated with irreversible damage (hydraulic failure) are approximated by P50 in gymnosperms and by P88 (i.e. water potential at 88% loss of conductivity) in angiosperms (Anderegg et al., 2016; Urli et al., 2013), possibly reflecting structural and functional differences in water transport systems (Choat et al. 2018).
Plants have adapted to water deficit through a wide range of life history and functional traits, with underlying anatomical and physiological mechanisms enabling them to colonise and persist in variable climate. P50 is known to be correlated with life history (Pratt et al., 2007), structure and function (Brodribb and Holbrook, 2004; Jacobsen et al., 2007), and species climate range (Bourne et al., 2017). For instance, in drier climates species tend to have higher wood density, which provides greater resistance to xylem conduit implosion under high xylem tensions and is strongly correlated with P50 (Barotto et al., 2018; Hacke et al., 2001; Jacobsen et al., 2005). Huber value (HV: ratio of sapwood area to leaf area) is observed to be negatively related to site water availability and P50, such that species in drier climates have higher HV and more negative P50 than species in wetter climates (Gotsch et al., 2010; Markesteijn et al., 2011). Leaf size has been observed to be negatively related to drought tolerance (P50) such that species with small leaves tend to be more cavitation resistant (Markesteijn et al., 2011; Schreiber et al., 2016). Studies have quantified and explored species vulnerability to climate-induced drought (P50) in relation to functional traits across biomes (Blackman et al., 2017, 2014; Bourne et al., 2017; Larter et al., 2017; Li et al., 2019, 2018; Lucani et al., 2019; Martorell et al., 2014; Nardini and Luglio, 2014; Pita et al., 2003). However, our knowledge on the drought tolerance (P50) of diverse related species in relation to the interactive effects of functional and life-history traits on species survival across biomes remain limited.
The Hakea genus is an ideal candidate for exploring variation in physiological drought tolerance across contrasting biomes. This is because Hakea is one of the two genera (the other beingGrevillea ) within the ancient Gondwana plant family Proteaceae, that have successfully transitioned into the arid biome in the Australian continent. Hakea also display a wide variation in functional and life history traits within and among biomes. For instance, some species re-sprout either from root suckers, epicormic or lignotuber buds after disturbance such as fire and drought (e.g. H. purpurea, H. drupacea and H. bakeriana ), while other Hakea species must rely on seed production (e.g. H. sericea ) (Clarke et al., 2013; Groom and Lamont, 1996; Weston, 1995). Leaf morphology varies greatly among species, with broad-leaved (e.g. H. dactyloides ,H. cristata, H. bucculenta ) and terete leaved species (e.g.H. leucoptera, H. tephrospermum, H. sericea ) (Groom and Lamont, 1996). The differences in functional traits and life history forms among the genus could influence species response to stress conditions (Groom and Lamont, 1996; Zeppel et al., 2015). Studies have shown that resprouting species tend to allocate more biomass to roots than shoots, as well as exhibiting lower rates of photosynthesis, hydraulic conductivity, and transpiration (Groom and Lamont, 1996; Hacke et al., 2001; Hernández et al., 2011; Vilagrosa et al., 2014; Zeppel et al., 2015). Leaf shape influences species exposure to drought, as such within warmer and drier sites, species tend to be needle-leaved (Groom and Lamont, 1996; Wright et al., 2017). Resprouting capacity and needle-leaves support a drought avoidance strategy, and as such may have variable drought tolerance (Groom and Lamont, 1996; Vilagrosa et al., 2014; Zeppel et al., 2015).
In this study we aimed to determine the drought tolerance ofHakea species that differ in life history and climatic niches to investigate what attributes are predictive of aridity. Firstly, we hypothesized that there will be significant variation in drought tolerance between species, and that this would be predicted by different life-histories. Specifically, resprouting species will have higher P50 than non-resprouting species, and that needle-leaved species will have higher P50 than broad-leaved species (Groom and Lamont, 1996; Hernández et al., 2011; Zeppel et al., 2015). Secondly, we hypothesized that drought tolerance (P50) will be predicted by species climate such that species in drier climates will have higher P50 than species in wetter climates (Bourne et al., 2017; Larter et al., 2017; Trueba et al., 2017). Thirdly, drought tolerance (P50) will be positively correlated to Huber Value (HV) and wood density (WD), and negatively correlated to specific leaf area (SLA). This study will therefore provide empirical evidence on species drought tolerance (P50) to inform conservation management of diverse native flora.

Materials and Methods

Experimental design and Species selection

All samples were collected from the same site, the Australian Botanic Garden (ABG), Mount Annan, NSW, Australia (GPS location: Lat. -34.0703, Log. 150.7668, average annual rainfall of 759 mm (2007-2016)) and were well-watered via irrigation systems (simulating a common garden design). Comparing multiple species from the same Hakea genus in a common garden minimizes environmental effects and allows quantification of genetically determined trait variation. Using this approach, we examined the variation in functional and hydraulic traits of a diverse array of species sampled from across the Hakea phylogeny (Cardillo et al. 2017). A total of 16 species were selected to represent a wide range of vegetation type, biome, climate, and life histories (Table 1). Species occurrence records were downloaded from the Australian Living Atlas (ALA) (https://www.ala.org.au/, 2019). Vegetation type was defined according to the World Wildlife Fund (WWF) as abbreviated by Cardillo et al. (2017); Arid (Deserts and Xeric Shrublands), Mediterranean (Mediterranean Forests, Woodlands and Scrub), Forest (Temperate Broadleaf and Mixed Forests), Grasslands (Temperate Grasslands, Savannas, and Shrublands). The vegetation harboring greater than 50% of the species occurrence records was assigned as its vegetation type. Biome was defined based on the aridity index (UNEP, 1997) as broadly humid and arid (aridity index > 0.5, < 0.5, respectively). The climate summary details for each species distribution was obtained from The Atlas of Living Australia using R v3.6.3 (RCoreTeam, 2020).

Sampling of Plant material

Three individuals for each species were sampled from the AGB. Terminal full sunlight, north-facing branch that were ca. 90 cm long were sampled and placed into a black plastic bag with wet tissue paper and transported immediately to the laboratory (<90 minutes). Samples were stored in a cold room at 4oC until they were processed (within 10 days). A standardized 50 cm branch was cut under water from the terminal end of the collected samples, from which the bottom 10 cm was excised, barked removed to estimate the sap flow area and then oven dried to obtain the wood density (WD: oven dry mass/volume). All leaves were removed from the remaining 40 cm branch and leaf area measured using the Li Cor 3100 leaf area meter. Leaf material was oven dried at 70 C for 48 h prior to obtaining the dry mass. Specific leaf area (SLA, mm2mg-1) was obtained by dividing the total leaf area by the leaf oven dry mass. The ratio of the sapwood area to leaf area was described as the Huber value (HV).