RESEARCH ARTICLE Assessing vulnerability to embolism and hydraulic safety margins in reed-like Restionaceae A. G. West1 , K. Atkins1, J. J. van Blerk1 & R. P. Skelton2,3 1 Department of Biological Sciences, University of Cape Town, Rondebosch, South Africa 2 Fynbos Node, South African Environmental Observation Network, Newlands, South Africa 3 Department of Animal, Plant and Environmental Sciences, University of the Witwatersrand, Johannesburg, South Africa Keywords Embolism; graminoids; optical method; P50; pneumatron; Restionaceae; turgor loss point. Correspondence A. G. West, Department of Biological Sciences, University of Cape Town, Rondebosch, South Africa. E-mail: adam.west@uct.ac.za Editor C. Werner Received: 30 November 2023; Accepted: 12 March 2024 doi:10.1111/plb.13644 ABSTRACT • The African Restionaceae (Poales), the dominant graminoid layer in the megadiverse Cape Floristic Region of South Africa, are distributed across a wide range of moisture availability, yet currently there is very little known about the underlying hydraulics of this group. • We tested two methods for measuring culm vulnerability to embolism, the optical and pneumatic methods, in three species of Cannomois ranging in habitat from semi-riparian (Cannomois virgata) to dryland (Cannomois parviflora and C. congesta). Estimates of culm xylem vulnerability were coupled with measures of turgor loss point (ΨTLP) and minimum field water potential (ΨMD) to assess hydraulic safety margins. • The optical and pneumatic methods produced similar estimates of P50, but differed for P12 and P88. All three species were quite vulnerable to embolism, with P50 of – 1.9 MPa (C. virgata), �2.3 MPa (C. congesta), and �2.4 MPa (C. parviflora). Estimates of P50, ΨTLP and ΨMD aligned with habitat moisture stress, with highest values found in the semi-riparian C. virgata. Consistent differences in P50, ΨMD and ΨTLP between species resulted in consistent hydraulic safety margins across species of 0.96� 0.1 MPa between ΨMD and P50, with onset of embolism occurring 0.43� 0.04MPa after ΨTLP for all three species. • Our study demonstrates that restio occupancy of dry environments involves more than the evolution of highly resistant xylem, suggesting that other aspects of water relations are key to understanding trait–environment relationships in this group. INTRODUCTION Graminoids are a highly diverse and ecologically important group of plants, influencing ecosystem productivity structure (Bond & Midgley 2012; February et al. 2013), distribution (Sta- ver et al. 2011), drought sensitivity (Knapp et al. 2002; Craine et al. 2013; Van Sundert et al. 2021), carbon storage (Goud et al. 2022; Zhou et al. 2022), herbivore abundance (Staver et al. 2019) and fire regimes (Karp et al. 2021) globally. The tol- erance of graminoids to environmental stress is an important factor influencing their productivity (Fay et al. 2003) and dis- tribution (Jurjavcic et al. 2002). A critical environmental stress is water availability, with both a surfeit and deficit of water pre- senting stressful conditions that can differentially affect co-occurring species (Swemmer et al. 2006), and structure communities (Silvertown et al. 1999; Araya et al. 2011; Silver- town et al. 2015). Climate models predict an intensification of the water cycle in future, increasing the frequency and duration of drought and flooding events globally (IPCC 2022), making understanding the tolerance of moisture extremes in grami- noids increasingly important. A key trait for drought tolerance in plants is the resistance of the xylem to embolism and hydraulic failure (Tyree & Sperry 1989). Under severe drought stress, plant water potentials can fall low enough to induce air to enter the xylem of the plant, causing emboli that disrupt water transport and can result in desiccation and death of the plant through hydraulic failure (McDowell et al. 2008; Brodribb et al. 2021). Hydraulic failure is thought to be the main cause of drought-related woody plant mortality globally (Anderegg et al. 2012; Adams et al. 2017; Choat et al. 2018) and has received considerable attention in the literature (Choat et al. 2018; Brodersen et al. 2019; Brodribb et al. 2020). However, compared to woody plants, relatively little is known about the importance of hydraulic failure in grami- noids. There are reasons to assume that hydraulic failure might be a less important trait for graminoids than woody plants because of the lower investment in biomass supporting the pho- tosynthetic tissue, as well as the ability of some graminoids to escape drought phenologically, to resprout or to potentially refill embolized vessels by positive xylem pressure (Holloway-Phillips & Brodribb 2011; Cao et al. 2012). However, recent work sug- gests that there is considerable variation in graminoid physiologi- cal drought tolerance (Craine et al. 2013) and resistance to embolism (Jacob et al. 2022) that is comparable to that of woody plants (Lens et al. 2016). As such, hydraulic failure, or the avoid- ance thereof, may be an important factor influencing crop resil- ience (Corso et al. 2020), graminoid productivity (Jacob et al. 2022) and structuring communities (Ocheltree et al. 2020), warranting further research into the hydraulic traits of grami- noids globally. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 633 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Plant Biology ISSN 1435-8603 https://orcid.org/0000-0002-9352-9282 https://orcid.org/0000-0002-9352-9282 https://orcid.org/0000-0003-1626-3270 https://orcid.org/0000-0003-1626-3270 https://orcid.org/0000-0003-2768-6420 https://orcid.org/0000-0003-2768-6420 mailto:adam.west@uct.ac.za http://creativecommons.org/licenses/by-nc-nd/4.0/ http://creativecommons.org/licenses/by-nc-nd/4.0/ http://creativecommons.org/licenses/by-nc-nd/4.0/ The Restionaceae (hereafter “restios”) are reed-like mono- cots that constitute the bulk of the graminoid layer in the mega-diverse sclerophyllous Fynbos Biome of the Cape Floris- tic Region (CFR) in southwestern Africa (Rebelo et al. 2006). The restios differ from grasses in that they have photosynthetic culms rather than leaves, and cluster roots rather than mycor- rhizal symbionts (Bell et al. 2000), and appear to dominate the more nutrient-poor soils of the region (Goldblatt 1978; Camp- bell 1983; Bond & Goldblatt 1984; Hoffman et al. 1987; Bergh et al. 2014). The restios represent a fascinating graminoid group to explore hydraulically. The 350 species of African restios are well resolved phylogenetically (Briggs & Linder 2009) and occur across the full hydrological range in the fynbos, from the lowest rainfall on well-drained sand to the highest rainfall or perenni- ally waterlogged habitats (Born & Linder 2018). Additionally, they have been shown to be strongly segregated by moisture availability patterns across small scales, with the proportion of time spent under flooding-induced or drought-induced stress correlating with species occurrence (Araya et al. 2011; Araya et al. 2012) and forming part of the fundamental niche (Born & Linder 2018). The ability to tolerate flooding stress has been well linked to the presence of aerenchymatous roots in the res- tios (Huber & Linder 2012), however less attention has been paid to the adaptations and trade-offs associated with coping with drought-related stress in this group, and the extent to which this relates to fundamental and realized niches. Interest- ingly, restios in mesic mountain fynbos have been shown to be remarkably drought resistant (West et al. 2012), maintaining high water potentials and low rates of transpiration over the dry summer (Skelton et al. 2023) despite being shallow rooted (West et al. 2012; Skelton et al. 2023); a phenomenon that has been attributed to their ability to utilize cloud (Marloth 1903, 1905; Nagel 1956) and dew moisture (Skelton et al. 2023) dur- ing rain-free periods. However, the broad environmental range, and considerable lability in root architecture (Ehmig & Lin- der 2020) of the restios, demand that traits relating to drought stress are more widely investigated in this group. An important step towards this, is the development of a reliable method of measuring drought-related vulnerability to embolism in this group. Here we test two methods of obtaining estimates of xylem vulnerability to drought-induced embolism for restio culms. The morphology of these culms, with sparsely distributed vas- cular bundles embedded in ground tissue, often with a central cavity (Fig. 1), presents a potential challenge for obtaining reli- able estimates of vulnerability to embolism. We derived esti- mates of vulnerability to embolism using the optical vulnerability method (Brodribb et al. 2016; Brodribb et al. 2017) and the automated pneumatic method, or ‘pneu- matron’ (Pereira et al. 2016, 2020). The optical method has been used successfully on graminoid leaves previously (e.g., Johnson et al. 2018; Corso et al. 2020; Jacob et al. 2022), but is yet to be tested on graminoid shoots or culms. An advantage of the optical method is that it is able to directly assess xylem embolism through image analysis (Jacob et al. 2022). In con- trast, the pneumatic method relies on the assumption that air discharged under a mild vacuum from an attached plant seg- ment is proportional to the amount of xylem embolism that segment has experienced (Pereira et al. 2016; Yang et al. 2022; Paligi et al. 2023). The pneumatic method has been used successfully to generate xylem vulnerability curves in a wide range of woody angiosperm species, including leaves (Pereira et al. 2020), but it has yet to be tested on graminoid shoots or culms. It is possible that gas diffusion from non-xylary tissues, or changes in volume during dehydration, may obscure gas dif- fusion from embolized xylem in graminoid culms. As such, this method may not be well suited to measuring xylem embolism in restioid culms that may distort or collapse during dehydra- tion. Yet, the relative ease and simplicity of the technique demands that it be tested. We chose three restio species of the genus Cannomois P.Beauv. ex Desv. from a range of habitats, from riparian to dryland, to test our methods. We measured vulnerability to embolism using the optical and pneumatic methods, minimum seasonal water potentials in the field (Ψmin) and pressure–volume curves. We hypothesized that culm xylem would embolise at lower water potentials than those associated with turgor loss point and regular seasonal in situ moisture stress in all species, and that vulnerability to embolism would be correlated with environmental conditions (more vulnerable in wetter habitats). We inferred that the creation of realistic vulnerability curves, with agreement across the methods and traits, would indicate that hydraulic trait measurements can be successfully obtained in restio culms and can be applied more broadly within the Restionaceae and across other morphologi- cally similar graminoid species. MATERIAL AND METHODS Sample selection and collection The study was conducted on three common species of Restio- naceae from the Cannomois genus representing a range of habi- tats and environmental conditions across the Cape Floristic Region. Species descriptions below were taken from Lin- der (2018). Cannomois virgata (Rottb.) Steud. is mostly found on south-facing slopes between 30 and 1800 m, in mesic areas with relatively high rainfall or in valley bottom seeps. Plants are 1.0–1.5-m tall with a well-developed spreading rhizome and large culms between 7.5 and 10.0 mm in diameter. Culms are robust, with five to eight thick-walled sclerenchyma layers, and a central ground tissue containing a central cavity. For this study, female plants were sampled from a steep south-facing slope adjacent to a river, in Jonkershoek Nature Reserve (33°58.40 S, 18°56.50 E, 270 m a.s.l., 1600 mm MAP). Canno- mois congesta Mast. is mostly found on well-drained pebbly or sandy soils between 1100 and 2000 m. Plants are 0.5–1.0-m tall, compact, without spreading rhizomes, and culms approxi- mately 1–2 mm in diameter. Culms have three to six thin-walled sclerenchyma layers, and a central cavity. For this study, female plants were sampled from a flat, sandy plain at Jonaskop (33°56.50 S, 19°31.50 E, 990 m a.s.l., 500 mm MAP). Cannomois parviflora (Thunb.) Pillans incorporates numerous synonyms, including Cannomois acuminata (Kunth.) Pillans. It is mostly found on well-drained pebbly or sandy soils between 150 and 1500 m. Plants are 0.5–1.0-m tall, with spreading rhi- zomes, and culms approximately 1–2 mm in diameter. Culms have four to six thick-walled sclerenchyma layers, and central ground tissue with scattered cavities or a central cavity. For this study, female plants were sampled from a NE-facing pebbly slope at Algeria in the Cederberg (32°22.40 S, 19°3.20 E, Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 634 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 550 m a.s.l., 675 mm MAP). A plant water relations study in this location in the 1980s (Miller et al. 1984) identified this spe- cies as Cannomois accuminata (Kunth) Pillans, which is now a synonym for C. parviflora (Linder 2018). We collected entire rhizomes, with culms attached, from at least three healthy female individuals of each species. Care was taken to keep the plant intact when sampling so as not to induce any embolism into the culms. Rhizomes were placed in large plastic bags with damp paper towels, which were then sealed to prevent further water loss while being transported to the laboratory for processing. Two sampling periods were con- ducted. The first, in October 2021, sampled plant material for optical vulnerability curves. The second sampling period, in June 2022, sampled material for pressure–volume curves and pneumatic vulnerability curves. Identical methods were followed in both sampling campaigns, and the same popula- tions were sampled. Field xylem water potential Predawn and midday xylem water potentials were measured on at least four individuals of each species in the field dur- ing February and March 2021 using a Scholander-type pres- sure chamber (PMS Instruments, Corvallis, Oregon, USA). Midday measurements were made between 12:00 and 14:00 h and predawn measurements were made approximately 1 h before sunrise. For each measurement, a single, fully devel- oped culm was excised and then wrapped in moist paper towel and placed in a plastic bag to prevent further water loss while being measured. Fig. 1. Cross- and transverse sections of culms showing vascular bundles for Cannomois parviflora (a), C. congesta (b) and C. virgata (c). Vascular bundles were stained by perfusing safranin dye through the culm. Note the presence of large central cavity. Transverse sections show the “window” cut for the optical vul- nerability technique, exposing the xylem embedded within the collenchyma, by removing the epidermal layers. Scale bar = 0.5 mm. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 635 West, Atkins, van Blerk & Skelton Assessing vulnerability to embolism in Restionaceae 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense Culm anatomy and vessel length To examine the xylem anatomy in our species, the xylem of a representative culm of each species was stained with 0.1% aqueous safranin dye solution in distilled water. Culm sections spanning two nodes (i.e., consisting of one full intact inter- node) were cut underwater using a single-edged carbon steel blade. The basal end of the section was attached to a 5 ml syringe, containing dye, and pressure was applied until a steady flow of dye came out the distal end. Thereafter, the segments were flushed with water until the water ran clear of excess dye. A window of epidermis and chlorenchyma tissue was carefully removed in the same way a culm would be prepped for optical scanning (see “Optical vulnerability measurements” below) to expose stained xylem tissue. Cross-sections were taken by hand sectioning. Lateral and cross-sectional images were captured using a dissecting microscope (Lecia EZ4 HD). To test whether vessels and the hollow central cavity spanned nodes, we per- fused pressurized air (100 kPa) into the proximal end of a culm while placing the distal end into a bucket of water. The distal end was then progressively cut back until bubbles could be seen streaming from the cut surface. For all species, vessels and the central cavity did not span the nodes. Thus, for pneumatic measurements, culms were excised just below the most basal node. This eliminated the concern of open vessels spanning the culm and minimized the air volume of the open central cavity. Optical vulnerability measurements We used the optical method described in Brodribb et al. (2016), Brodribb et al. (2017), and Skelton et al. (2018) to generate xylem vulnerability curves for restioid culms. Full details of the method, including an overview of the technique, image processing, as well as scripts to guide image capture and analysis, are also available at http://www.opensourceov.org. We attached culms, still attached to the rhizome, to flatbed scanners (Epson perfection V800 or V850 Scanner, Epson USA) to generate a time series of images of an exposed section of the xylem within the culm as the entire plant desiccated over a few days. We measured a single culm per individual plant (n= 3 individuals) for both C. virgata and C. parviflora. For C. congesta, we measured three culms per individual, for three individuals. The three culms per individual were subsequently averaged per individual (resulting in n= 3 individuals). The xylem was exposed by carefully removing the epidermal and chlorenchymal layer from a small section of the culm, using a dissecting microscope and a single-edged carbon-steel blade (Fig. 1). During this procedure, considerable care was taken to not cut any xylem tissue and to ensure that the exposed vascu- lar tissue remained moist by periodically dousing it in water to prevent any air entering the xylem before measurement. The culm was then attached to the scanner with the window of exposed xylem tissue facing down. Each window was sur- rounded by a rectangular putty (Prestik, Bostik, SA) frame to create a containment well for ultrasound gel (Tensive ultra- sound gel; Parker Labs, America). The ultrasound gel ensured that the window did not desiccate too rapidly without imped- ing the transmission of light to the xylem surface. Each well was then covered by a glass microscope slide and was secured to the scanner bed with duct tape to prevent any movement which may interfere with measurements during the dry down process. Culms were then scanned in reflective mode at 4800 dpi every 5 min over a period of a few days, allowing us to track embolism events over time within the xylem in each sample. Concurrently to optical scanning, we measured the decline in xylem water potential of culms attached to the same rhizome as the scanned culms using a Scholander-type pressure cham- ber (PMS Instruments, Corvallis, OR, USA). Culm water potential was periodically measured until the water potential fell too low for reliable pressure chamber measurements for these species (between �2 and �3MPa). This related to a min- imum water potential measurement equal to or less than P50 for all species. The optical measurements where then continued for at least 36 h after this point until the culms were visibly des- iccated and discoloured. Upon completion, image sequences were analysed to identify embolism events, seen as changes in the reflection of the stem xylem. Image subtraction of subsequent images conducted in ImageJ (National Institutes of Health, Bethesda, MD, USA) was used to reveal rapid changes in light transmission or con- trast produced by each embolism event. Slow movements of the culms caused by drying could easily be distinguished from embolism events and were filtered from the analysis. Embolism events were thresholded, allowing automated counting of each event using the analyse-stack function in ImageJ. From the thresholded stack of embolism events we could extract a time-resolved count of embolism events (using the timestamp of each image). We then converted the raw embolism counts to a percentage of total pixels embolized, producing a dataset of time-resolved percentage embolism. The time-resolved per- centage embolism data were combined with the xylem water potential timeline to estimate the culm xylem water potential associated with each embolism event. The water potential data were modelled using segmented regressions that described the rapid initial phase of water potential decline to turgor loss point, followed by a more gradual secondary decline. This sec- ondary decline was then extrapolated to match the end of the optical data time series. Vulnerability to embolism was recorded as the relationship between percentage embolism and water potential (Ψ), and modelled using a sigmoid function (Pammenter & van Wiligen 1998): Percentage embolism ¼ 100 1þ ea Ψ�bð Þð Þ (1) where a corresponds to the sensitivity to decreasing water potential (proportional to the slope of the equation) and b is the water potential associated with 50% embolism (P50, MPa). P88 and P12 were calculated using the equations of Domec & Gartner (2001). 95% confidence limits of the vulnerability curves were calculated through bootstrapping. Pneumatic vulnerability measurements We constructed automated ‘pneumatrons’, with ports for two culms, programmed to measure the air discharge from the attached culms every 15 min following the protocol of Pereira et al. (2020) and Trabi et al. (2021). Individual culms were excised from field-collected plants that were being kept in large, moistened plastic bags in the laboratory. Every effort Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 636 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense http://www.opensourceov.org was made to prevent culm desiccation occurring prior to mea- surements commencing on the pneumatrons. Prior to excision from the plant, individual culms were placed in smaller plastic bags with moist paper towel and sealed basally. This was done inside the larger plastic bag containing the entire plant. The culm was then excised at the base and connected directly to a port on a pneumatron using tight-fitting silicon tubing. Experimentation showed that leaks were best prevented using contact adhesive (Genkem) around the outside of the culm/- tube interface. The flexibility of the contact adhesive was suffi- cient to prevent leaks even as the culm shrunk during dry down. After all culms were attached, and the pneumatrons had gone through a few cycles to verify there were no leaks, the bags were removed from the culms, and the benchtop dry-down commenced. Each measurement cycle commenced with pulling a vacuum on the culm that reduced the pressure in the manifold to approximately 30 kPa. The pressure in the discharge tube was then measured for 30 s, following which the tube was vented and returned to atmospheric pressure. Air discharge (AD, μl) was calculated, following the methods of Pereira et al. (2016, 2020), as: AD ¼ Pf�Pið Þ � Vr Patm � � � 106 (2) where Patm is atmospheric pressure (kPa), Vr is the discharge tube volume of the pneumatron (�0.001 l), Pi is the initial pressure (kPa), measured immediately after imposing the par- tial vacuum on the culm, and Pf is the final pressure (kPa), measured after 5 s. The 5-s timeperiod was empirically deter- mined as yielding the best results, as longer times tended to result in greater noise and less differentiation in AD across the entire dry-down, possibly due to diffusion of air through the non-xylary components of the culm. We ran two rounds of pneumatic measurements for each species, attempting to obtain sufficient culm replication. We aimed to measure three individual plants per species, with each individual plant represented by the average of two culms. How- ever, several culms provided no coherent air discharge trend per round. For C. congesta and C. parviflora we obtained curves for three individual plants, but for C. parviflora we only recov- ered curves for two individuals. This would need to be improved upon for ecological studies seeking sufficient statisti- cal power to detect small differences between species and popu- lations. During each round of measurements, culm water potential was measured periodically on additional excised culms that were handled in as similar a manner to pneumatron culms as possible. These culms were excised at the same time as the pneumatron culms and were placed alongside the pneuma- tron culms in the laboratory to dry down together. Measure- ments were made and modelled in the same manner as described for the optical method above. Each water potential culm was only measured once, to prevent the possibility of cumulative damage skewing the water potential measurements over the course of the dry down. The pneumatron measurements were terminated when the culms were visibly desiccated and discoloured and the air dis- charge curve had plateaued. At the conclusion of the dry-down, percent air discharge (PAD, %) was calculated as: PAD ¼ 100� AD�ADminð Þ ADmax�ADminð Þ (3) where ADmin is the minimum and ADmax is the maximum amount of air discharge (μl) measured over the dry-down period. These time-resolved percentage air discharge data were combined with the xylem water potential (Ψ) timeline and modelled using a sigmoid function (Pammenter & van Wiligen 1998) as for the optical curves. Pressure–volume curves Pressure–volume curves were created for three culms per spe- cies by repeatedly measuring the xylem water potential (follow- ing the methods described above), together with the culm mass, as the culms desiccated from a fully hydrated state in the laboratory. Once the measurements were completed, the culms were dried at 60 °C for 48 h and then weighed to obtain the dry mass. The relative water content (RWC; %) was then calculated as: RWC ¼ mw�md ms�md � 100 (4) where mw is the wet mass at any given time, md is the dry mass, and ms is the saturated mass which was the first mass measure- ment taken on the culms when they were fully hydrated. For each culm a pressure–volume curve was plotted as �1/Ψ (MPa�1) versus 100 – RWC (%) and the turgor loss point (ΨTLP; MPa), osmotic potential (Ψ0), and relative water con- tent at turgor loss point (RWCTLP) were calculated using stan- dard procedures (Nicotra et al. 2010; Bartlett et al. 2012). Hydraulic safety margins Hydraulic safety margins were calculated in two ways: HSMΨMD was calculated as the difference between minimum field water potential (ΨMD) and the P50 obtained from averag- ing the optical and pneumatic estimates. HSMΨTLP was calcu- lated as the difference between the turgor loss point (ΨTLP) and the P50 obtained from averaging the optical and pneumatic estimates. Error was propagated as: ε A�Bð Þ ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ε2A þ ε2Bð Þ q (5) where A and B are the two means being subtracted and ε is the error associated with that mean. Statistical analyses All analyses were performed using R Statistical Software (ver- sion 4.2.1; R Core Team 2022) and the “tidyverse” suite (Wick- ham et al. 2019). Segmented regressions were fit to optical and pneumatron dry-down water potential time series using the “segmented” package in R (version 1.6-0; Muggeo 2008). Sig- moid curves were fit to optical and pneumatron data using “nls” (R Core Team 2022). 95% confidence intervals of species mean curves were calculated by bootstrapping. Plots were pro- duced using “ggplot2” (Wickham 2016) and “ggpubr” (Kas- sambara 2020). Differences in model parameters between Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 637 West, Atkins, van Blerk & Skelton Assessing vulnerability to embolism in Restionaceae 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense species, method and their interaction were tested for using two-way ANOVA and Tukey HSD post-hoc comparisons. Where a factor not significant in the full model, a reduced model, only containing the significant factors, was used following the pro- cedures of Crawley (2012). RESULTS Culm anatomy with regard to optical and pneumatic methods The three Cannomois species examined in this study all had culms with a similar anatomical structure, with distributed vas- cular bundles surrounded by parenchymatous tissue, and a large hollow cavity in center of the culm. Our staining showed that there was considerable xylem visible from the cleared epi- dermal layer (Fig. 1), suitable for a robust optical measure- ment. To date the pneumatic method has not been successfully applied to herbaceous monocots. However, as the pneumatic method normalizes air discharge through the course of a dry-down to the initial air discharge from the hydrated culm, the presence of a large fraction of parenchymatous tissue and a large central cavity are not necessarily problematic for this technique. Potential challenges are the sparse vascular bundles resulting in a small air discharge volume requiring a low vol- ume discharge tube to obtain sufficient resolution (see Pereira et al. 2020), the potential for early gas discharge from residual water on the exposed cut surface (e.g., Miranda et al. 2023) and potential changes in the volume of airspace in the culm during the dry-down which could potentially obscure patterns of air discharge from the xylem. Optical vulnerability curves Xylem embolism events were readily discernable in the imagery as sharply defined lines (Fig. 2). In contrast, movement or col- lapse of non-xylary tissues resulted in more diffuse changes that were easy to differentiate from the xylem and thus exclude from the analysis. As the culms were still attached to intact rhi- zomes, the plants dried-down slowly, taking up to 6 days to complete the dry-down (Fig. 3). The culms for C. congesta and C. virgata dried-down at similar rates, evidenced by the similar progression of percentage embolism and water potential with time for each culm (Fig. 3b, c, e, f). As such, a single relation- ship between water potential and time was used to create the vulnerability curves for these species (Fig. 3h, i). For C. parvi- flora, it was apparent from both the optical and the water potential data that the progression of the dry-down was more rapid for one of the individuals than for the other two (Fig. 3a, d). As such, we fitted a separate water potential versus time relationship for this one individual. Doing so reduced the vari- ation in vulnerability curves between the individuals (Fig. 3g), which we regarded as a more accurate reflection of variation between the individuals. Pneumatic vulnerability curves We obtained reasonable air discharge curves from the three spe- cies of Cannomois (Fig. 4). The larger, more rigid culms of C. virgata produced a greater gas discharge volume than the smal- ler, more flexible culms of C. parviflora and C. congesta. For the latter two species, pressure differences through the dry-down were small, despite minimizing the discharge tube volume to �1ml. This resulted in less resolution for the C. parviflora and C. congesta curves than for C. virgata. This could be improved in future by further minimizing discharge tube volume. For the pneumatic method, the dry-down progressed much faster than for the optical method, as the culms were excised from the plant, rather than attached to the rhizomes. For both C. parviflora and C. congesta, the culms dried down at similar rates, as such a single relationship between water potential and time (Fig. 4d, e) was used to create the vulnerability curves for each of these species (Fig. 4g, h). For C. virgata, the culms were bagged overnight during the first measurement round (Fig. 4c), but not for the second round of measurements. As such, we used separate water potential versus time relationships Fig. 2. Dehydration sequence of exposed culm xylem sections, for the opti- cal technique, for three Cannomois species. In each row the first image depicts the culm with the exposed “window” showing xylem vessels. Subse- quent images in each row depict cumulative embolism events in each culm, from fully hydrated to the given water potential. Emboli are coloured from earlier (colder/darker colours) to later events (warmer/lighter colours). Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 638 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense for each measurement round for this species (Fig. 4g) to create the vulnerability curves (Fig. 4i). Comparison of vulnerability curve methods The optical and pneumatic methods produced similar esti- mates of P50 for all three species but differed for P12 and P88 (Tables 1 and 2, Fig. 5), with the pneumatron producing higher estimates of P12 and lower estimates of P88 than the optical method (Figs 5 and 6). There was a strong species effect for both methods for P50, P12 and P88 (Table 2, Fig. 5). For P50, both methods produced a significantly higher P50 for C. virgata compared to C. parvi- flora and C. congesta, but no significant difference between C. parviflora and C. congesta (Fig. 5a). For P12, the optical method produced no difference between the species, whereas the pneu- matic method resulted in higher estimates of P12 for C. con- gesta and C. virgata, but not for C. parviflora, resulting in a significant species:method interaction (Fig. 5b). For P88, both methods produced similar relative patterns for the species, Fig. 3. Optical vulnerability curves for three Cannomois species. (a–c) Timeline of percentage embolism for each species, where each solid line represents an individual plant. For C. congesta, three culms (dashed lines) were averaged per plant (solid lines). (d–f) Timeline of water potential for each species. Points indi- cate observed values recorded using the Scholander-type pressure chamber, lines indicate interpolated xylem water potential (see Material and Methods for details). (h–j) Xylem vulnerability curves for each species. Coloured lines correspond to lines in (a–c), solid black lines are mean xylem vulnerability curves, filled black circles indicate the P50 values, and the grey shaded areas indicate 95% confidence intervals for each species. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 639 West, Atkins, van Blerk & Skelton Assessing vulnerability to embolism in Restionaceae 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense with significant differences found between C. congesta and C. virgata, but not for C. parviflora (Fig. 5c). Vulnerability to embolism, pressure–volume curves and hydraulic safety margins All three of the Cannomois species were quite vulnerable to embolism, with values for P50 ranging between �1.9 MPa (C. virgata), �2.3 MPa (C. congesta) and �2.4 MPa (C. parviflora) (Table 1, Fig. 5). This relative ranking of P50 between species corresponded with the values obtained for midday water potential (ΨMD), turgor loss point (ΨTLP) and osmotic poten- tial at full turgot (ΨO) (Table 1, Fig. 7). ΨTLP varied signifi- cantly between the species, with C. virgata having a significantly higher turgor loss point than C. congesta and C. parviflora (Table 1, Figure S1). A similar result was found for ΨO, with significant differences being found between C. parvi- flora and C. virgata, whereas RWCTLP was significantly higher for C. parviflora than for C. congesta and C. virgata (Table 1, Figure S1). Fig. 4. Pneumatic vulnerability curves for three Cannomois species. (a–c) Timeline of percentage embolism, based on air discharge, for each species. Colours represent individual plants, with symbols reflecting measurement round. Open symbols (for C. congesta and C. virgata) represent individual culms that were averaged per plant (filled symbols). (d–f) Timeline of water potential for each species. Points indicate observed values recorded using the Scholander-type pres- sure chamber, symbols reflect measurement round, lines indicate interpolated xylem water potential (see Material and Methods for details). (h, i) Xylem vulner- ability curves for each species. Coloured symbols as per panel (a–c), solid black lines are mean xylem vulnerability curves, filled black circles indicate the P50 values, and the grey shaded areas indicate 95% confidence intervals for each species. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 640 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense The consistent differences in P50, ΨMD and ΨTLP between species resulted in consistent hydraulic safety margins, with the average HSMΨMD across species being 0.95� 0.1 MPa, and the average HSMΨTLP across species being 0.69� 0.05MPa (Table 1, Fig. 7). The P12 estimates obtained from the optical method indicated that the onset of embolism occurred approx- imately 0.43� 0.04 MPa after ΨTLP for all three species (Fig. 6). P12 estimates from the pneumatic method were more variable, with P12<ΨTLP for C. parviflora, but P12>ΨTLP for C. congesta and C. virgata (Fig. 6). DISCUSSION The extent to which aboveground hydraulic traits influence species distributions of graminoids is relatively understudied globally. This is certainly the case for the restios of the Cape Floristic Region (CFR), where strong hydrological niche segregation is seen (Araya et al. 2011), as well as a high degree of root lability (Ehmig & Linder 2020), yet little is known about the aboveground hydraulic traits of these species and the extent to which these influence this niche segregation. Here we demonstrate that the optical and pneumatic methods can suc- cessfully generate robust estimates of vulnerability to embolism for restios, that will enable future examination of trait–envi- ronment relationships in this and other morphologically simi- lar graminoid groups. Doing so allows us to demonstrate that restio occupancy of dry environments involves more than the evolution of highly resistant xylem, suggesting that other aspects of water relations are key to understanding trait–envi- ronment relationships in this diverse and ecologically impor- tant group (West et al. 2012; Skelton et al. 2023). The optical vulnerability method appears to be a reliable method for detecting the dynamics of embolism formation in restio culms. The optical method produced steep sigmoidal curves, with all estimates of vulnerability to embolism (P50, P12, P88) occurring at more negative water potentials than Ψmin and ΨTLP for all species (Fig. 6a), resulting in positive hydraulic safety margins (Table 2), thereby providing support for the embolism avoidance hypothesis (e.g., Martin-StPaul et al. 2017; Skelton et al. 2021). The relative rankings of the estimates of vulnerability to embolism (P50, P12, P88) were cor- related with the relative values of Ψmin and ΨTLP, with the more mesic C. virgata appearing more vulnerable to embolism, and with higher Ψmin and ΨTLP than the dryland C. congesta and C. parviflora (Fig. 7a). We could not distinguish between the two dryland species in this study, and increased sample size will be necessary if attempting to detect differences between such similar species in future. The P50 estimated by the pneumatic method was very consis- tent with that from the optical method (Fig. 6). However, there Table 1. Parameters extracted from vulnerability and pressure volume curves as well as measured field water potentials for Cannomois parviflora, C. congesta and C. virgata. C. parviflora C. congesta C. virgata F (df) P P50_opt (MPa) �2.32� 0.14 �2.29� 0.06 �1.93� 0.05 P12_opt (MPa) �2.12� 0.14 �1.89� 0.04 �1.79� 0.03 P88_opt (MPa) �2.52� 0.15 �2.69� 0.09 �2.07� 0.09 P50_pneu (MPa) �2.53� 0.03 �2.25� 0.02 �1.82� 0.08 P12_pneu (MPa) �2.26� 0.04 �1.38� 0.03 �0.95� 0.18 P88_pneu (MPa) �2.80� 0.02 �3.11� 0.06 �2.69� 0.13 ΨTLP (MPa) �1.70� 0.09b �1.52� 0.02b �1.28� 0.02a 15.9 (2, 6) 0.004 Ψo (MPa) �1.47� 0.15b �1.11� 0.02ab �0.99� 0.04a 8.09 (2, 6) 0.020 RWCTLP (%) 96.2� 0.73b 90.8� 1.0a 89.9� 0.96a 16.0 (2, 6) 0.004 ΨPD (MPa) �0.49� 0.14 �0.55� 0.08 �0.46� 0.05 0.12 (2, 35) 0.888 ΨMD (MPa) �1.34� 0.12 �1.26� 0.09 �1.11� 0.14 0.31 (2, 35) 0.734 HSMΨMD 1.09� 0.15 1.01� 0.10 0.77� 0.15 HSMΨTLP 0.73� 0.12 0.75� 0.04 0.60� 0.07 P50_opt, P12_opt, P88_opt reflect the water potentials at 50%, 12% and 88% embolism measured with the Optical technique. P50_pneu, P12_pneu, P88_pneu reflect the water potentials at 50%, 12% and 88% of maximum air discharge measured with the pneumatic technique. ΨTLP, turgor loss point; ΨO, osmotic potential at full turgor; RWCTLP, relative water content at the turgor loss point; ΨPD, pre-dawn water potential; ΨMD, midday water potential; HSMΨMD, hydraulic safety margin (defined as ΨMD – P50); HSMΨTLP, hydraulic safety margin (defined as ΨTLP – P50). Results from a one-way ANOVA are shown for the pressure volume curve parameters and the field water potential data. Significant differences are shown in bold. Where the ANOVA was significant, letters reflect dif- ferences as detected by Tukey HSD post-hoc tests. Results from a two-way ANOVA for the vulnera- bility curve parameters are shown in Table 2. Table 2. Results from two-way ANOVA testing for differences between P50, P12 and P88 by species and method (optical versus pneumatic) with interac- tion (model: y� species ×method). Parameter Factor df F value P P50 Species 2, 14 21.6 5.22E-05 P12 Species 2, 11 29.1 4.1E-05 Method 1, 11 26.4 3.2E-04 Species:Method 2, 12 10.3 3.0E-03 P88 Species 2, 13 12.5 9.4E-04 Method 1, 13 26.4 1.9E-04 Factors and the interaction were dropped from models when not significant following the procedures of Crawley (2012). The simplest model is reported below. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 641 West, Atkins, van Blerk & Skelton Assessing vulnerability to embolism in Restionaceae 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense was disagreement between the methods regarding P12 and P88 (Fig. 6). Recent work indicates that sap residue in cut xylem conduits can result in early gas discharge, influencing the slope of vulnerability curves produced by the pneumatic method for some woody species (Miranda et al. 2023). As demonstrated by Miranda et al. (2023), this early gas discharge only occurred at Fig. 5. Comparison of critical water potential thresholds, P50 (a), P12 (b) and P88 (c), for three species of Cannomois measured by the optical (solid boxes) and pneumatic (hatched boxes) vulnerability methods. Horizontal bars represent the minimum and maximum (outer bars connected by dashed lines), the median (inner, thick lines), and the first quartile and third quartile in the dataset of each species. Open circles represent outliers. Significant factors (two-way ANOVA) are shown in each panel, with letters reflecting significant differences between treatments as determined by Tukey HSD post-hoc tests. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 642 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense water potentials greater than turgor loss point and was not related to embolism formation. Correcting for this early gas dis- charge resulted in strong agreement between the pneumatic and optical methods. A similar artefact may have been present in our measurements on restios. All our pneumatic curves displayed early gas discharge at odds with the onset of embolism shown by the optical curves (Fig. 6). It is possible that this early gas dis- charge resulted from sap residue in cut xylem conduits, as well as from the extensive parenchymatous tissue surrounding the vascular bundles, prior to the onset of the turgor loss point. Additionally, potential changes in discharge tube volume during culm desiccation might influence the calculation of air discharge if not properly quantified. These factors were not directly tested in this study, and remain future research directions to explore. Nevertheless, despite possible contributions of non-embolism related gas discharge, pneumatic estimates of P50 can be robust (Pereira et al. 2016), as appears to be the case here, given the similarity between the two methods (Fig. 6). As such, we con- sider the pneumatic method to be a useful way to estimate P50 for these species. Additionally, future methods developments will no doubt improve pneumatic estimates of P50, P12 and P88. Future studies might also examine hypotheses linking inter- culm (or intra-individual) variability in vulnerability to indi- vidual survival in the Restionaceae. High inter-leaf variability in xylem resistance to embolism in Persea has been linked with heterogeneity in drought-induced leaf mortality across the can- opy, which may act as a buffer against complete canopy death during prolonged drought (Cardoso et al. 2020). We have observed patchy dieback across Cannomois individuals during Fig. 6. Methods comparison of the optical and pneumatic vulnerability curves for three species of Cannomois. Thick solid coloured lines represent the mean optical vulnerability curve, with 95% confidence intervals shown by coloured shading and individual culms shown by thin solid lines. Mean pneumatic vulnerability curves are represented by dashed, black lines, with 95% confidence intervals shown by grey shading, and individual culms shown by grey symbols. Critical water potential thresholds are shown on the mean vulnerability curves as filled circles (P50), down-facing triangles (P12) and upward-facing triangles (P88). Also shown is the water potential associ- ated with turgor loss of culms (vertical coloured dashed lines) and minimum field xylem water potential (coloured horizontal bars reflecting range, with vertical line indicating the mean). Fig. 7. Xylem vulnerability curves for the three species generated using the optical (a) and the pneumatic (b) vulnerability techniques. In each panel the solid black lines are mean xylem vulnerability curves for each species and the shading indicates the 95% confidence intervals. Critical water potential thresholds are shown on the mean vulnerability curves as filled circles (P50), down-facing triangles (P12) and upward-facing triangles (P88). Also shown is the water potential associated with turgor loss of culms (vertical coloured dashed lines) and minimum field xylem water potential (coloured horizontal bars reflecting range, with vertical line indicating the mean). Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 643 West, Atkins, van Blerk & Skelton Assessing vulnerability to embolism in Restionaceae 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense prolonged dry periods (see Plate 3.4, Skelton 2014) and future studies might examine if the inter-culm variation in vulnerabil- ity, seen most noticeably with the optical method (Fig. 3), cor- relates with these phenomena in the field. Our estimates of P50 for our three Cannomois species were between �1.9 and �2.4 MPa (Table 2), which is suggestive of a relatively high vulnerability to embolism for these perennial graminoids, growing in seasonally dry environments. Previous estimates of P50 for graminoids range widely between �0.5 and �7.6 MPa (e.g., Holloway-Phillips & Brodribb 2011; Skelton et al. 2015; Lens et al. 2016; Ocheltree et al. 2016; Johnson et al. 2018). However, these studies used a wide variety of methods, which renders direct comparison potentially chal- lenging. Directly comparable studies for graminoids, using the optical technique, have measured P50 of between �2.21 and �2.87MPa for wheat (Johnson et al. 2018; Corso et al. 2020), and between �4.4 and �6.1 MPa for five different pasture grasses (Jacob et al. 2022). Our finding of greater vulnerability to embolism in three perennial restio species than that found in an annual crop spe- cies, such as wheat, demands explanation. One possible expla- nation could be that individual culms are not built to endure negative water potentials through the dry season, having nega- tive safety margins and being effectively drought-deciduous. Several lines of evidence indicate that this is not the case in most Restionaceae species, including Cannomois. While a pre- vious study found negative safety margins for two species of restios (H. aristata and C. congesta) (Skelton et al. 2015), the estimates of P50 for these species were produced using air- injection and hydraulic methods that resulted in exponential curves that most likely overestimated vulnerability (e.g., Cochard et al. 2013). Our study supports this by demonstrating sigmoidal vulnerability curves, with positive safety margins for these species, which we consider a considerable advance in our ability to accurately measure vulnerability curves for these spe- cies. Furthermore, Cannomois are evergreen perennials, retain- ing functional culms all year round (Linder 2018; Skelton et al. 2023). They do not have sterile culms, which may follow more acquisitive strategies, retaining only fertile culms which are more conservative (Ehmig et al. 2019). Individual culms live for several years, commencing growth before the start of the wet winter (van Blerk et al. 2022), with the female plants of C. congesta holding a single nut for 2 years at the distal tip of the culm before it is mature and released (van Blerk et al. 2017). Thus, inability to survive at least one dry season would result in considerable fitness costs for these plants. Drought-deciduousness is also extremely rare in the fynbos, presumably due to the paucity of nutrients, which make an evergreen, sclerophyllous strategy most adaptive. Another possible explanation for the relatively high vulnera- bility to embolism is that these restios are able to avoid nega- tive water potentials in their culms throughout the dry season, thereby maintaining a positive hydraulic safety margin and avoiding embolism in the culms. This study, as well as previous field observations on restios, support this explanation. In this study, the onset of embolism occurred after both the ΨTLP and Ψmin for all species (Fig. 6). While there are no comparable measurements of ΨTLP in the literature, our measurements of Ψmin are comparable to previous studies. Minimum water potentials for Cannomois acuminata (now C. parviflora) were measured as >�1.9 MPa between October to May at the same field site used for this study (Miller et al. 1984; von Willert et al. 1989). Similarly, a multi-year field ecophysiological inves- tigation of the shallow-rooted C. congesta found that Ψmin never fell below �2MPa, even in a particularly dry year, typi- cally remaining >�1.2 MPa throughout the year (Skelton et al. 2023). Gas exchange and sap flow were also maintained in these culms throughout the year, at relatively low levels (Skelton et al. 2013; Skelton et al. 2023), most likely sustained by the uptake of the regular dew and cloud moisture events during rain-free periods (Skelton et al. 2023). This combina- tion of low rates of gas exchange with regular moisture uptake may enable the success of these leafless culms through the rela- tively mild summers typical of the Cape mountain fynbos. The ability of restios to intercept and trap cloud moisture (Marloth 1903, 1905; Nagel 1956), has the potential to alleviate summer drought stress in areas receiving frequent bouts of summer cloud. This was demonstrated by an in-situ drought experiment in mountain fynbos, where all summer rainfall, but not cloud moisture, was removed from natural plots contain- ing proteas, ericas and restios (West et al. 2012). Despite hav- ing no summer rainfall, water potentials in the restios (Hypodiscus aristatus and Staberoha cernua) never fell below �2MPa, presumably because of the regular uptake of cloud moisture that would frequently wet the foliage over the sum- mer, whereas co-occurring shallow-rooted Erica species experi- enced water potentials down to �10MPa. As a result, there was no drought impact on the restios, whereas Erica species suffered dieback and mortality (West et al. 2012). The ability to obtain robust estimates of vulnerability to embo- lism for reed-like graminoids should enable exploration of the importance of water relations parameters in the restios and other morphologically similar groups. For the restios, there are several pressing lines for investigation. The 350 species of restios occur across the full hydrological range in the CFR. Yet to date, only a handful of species have received close attention, mostly in rela- tively mesic mountain fynbos (Miller et al. 1984; von Willert et al. 1989; West et al. 2012; Skelton et al. 2023). It is of particular interest to determine the hydraulic safety margins in restios across their full hydrologic range, and the extent to which future drought may impact this ecologically important component of the CFR. This is a pressing research agenda, given the strong like- lihood of increasingly severe drought in the region (IPCC 2022). There is also interesting work to be done in uncovering the hydraulic basis for the strong hydrologic niche segregation seen in this group (Araya et al. 2011), particularly in the context of increased groundwater abstraction from the CFR to meet human water needs (February et al. 2004). Lastly, there is considerable interest in the role that hydrologic refugia (Mclaughlin et al. 2017) may have played in the evolution of the Cape Flora. The ability to add robust hydraulic traits to restio distribution and phylogeny should considerably aid this research. ACKNOWLEDGEMENTS The authors wish to acknowledge the Du Plessis family for access to Jonaskop, and Cape Nature (Permit number 0052- AAA008-00008 and CN32-28-14428) for allowing access to the Jonaskop, Cederberg and Jonkershoek study sites. Financial support was received from a FLAIR Fellowship from the Royal Society and the African Academy of Sciences awarded to RPS (Award number FLR\R1\191609). The FLAIR Fellowship Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 644 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense Programme is a partnership between the African Academy of Sciences and the Royal Society funded by the UK Govern- ment’s Global Challenges Research Fund. The authors acknowledge Brad Ripley and Liam Reynolds for assistance in developing the pneumatron method. AUTHOR CONTRIBUTIONS AGW and RPS designed the study. AGW and RPS performed the experiments and collected all the data. KA assisted in devel- oping the optical method for use on restios. JvB provided anatomical sections. Data were analysed by AGW, RPS and JvB. AGW wrote the manuscript with revisions from RPS, JvB and KA. SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of the article. Figure S1. (a) Pressure-volume curves for three species of Cannomois. REFERENCES Adams H.D., Zeppel M.J.B., Anderegg W.R.L., Hart- mann H., Landhäusser S.M., Tissue D.T., Huxman T.E., Hudson P.J., Franz T.E., Allen C.D., Anderegg L.D., Barron-Gafford G.A., Beerling D.J., Breshears D.D., Brodribb T.J., Bugmann H., Cobb R.C., Collins A.D., Dickman L.T., Duan H., Ewers B.E., Galiano L., Galvez D.A., Garcia-Forner N., Gaylord M.L., Germino M.J., Gessler A., Hacke U.G., Hakamada R., Hector A., Jenkins M.W., Kane J.M., Kolb T.E., Law D.J., Lewis J.D., Limousin J.-M., Love D.M., Maca- lady A.K., Martı́nez-Vilalta J., Mencuccini M., Mitch- ell P.J., Muss J.D., O’Brien M.J., O’Grady A.P., Pangle R.E., Pinkard E.A., Piper F.I., Plaut J.A., Pock- man W.T., Quirk J., Reinhardt K., Ripullone F., Ryan M.G., Sala A., Sevanto S., Sperry J.S., Vargas R., Ven- netier M., Way D.A., Xu C., Yepez E.A., McDowell N.G. (2017) A multi-species synthesis of physiologi- cal mechanisms in drought-induced tree mortality. Nature Ecology & Evolution, 1, 1285–1291. Anderegg W.R.L., Berry J.A., Smith D.D., Sperry J.S., Anderegg L.D.L., Field C.B. (2012) The roles of hydraulic and carbon stress in a widespread climate- induced forest die-off. roceedings of the National Academy of Sciences of the United States of America, 109, 233–237. Araya Y.N., Silvertown J., Gowing D.J. (2012) Do niche-structured plant communities exhibit phylo- genetic conservatism? A test case in an endemic clade. Journal of Ecology, 100, 1434–1439. Araya Y.N., Silvertown J., Gowing D.J., McConway K.J., Peter L.H., Midgley G. (2011) A fundamental, eco-hydrological basis for niche segregation in plant communities. New Phytologist, 189, 253–258. Bartlett M.K., Scoffoni C., Sack L. (2012) The determi- nants of leaf turgor loss point and prediction of drought tolerance of species and biomes: a global meta-analysis. Ecology Letters, 15, 393–405. Bell T.L., Stock W.D., Linder H.P. (2000) Ecophysio- logical investigations of the distribution of Poaceae and Restionaceae in the cape floristic region, South Africa, Grasses: Systematics and evolution. CSIRO, Melbourne, Australia, pp 267–278. Bergh N.G., Verboom G., Rouget M., Cowling R.M. (2014) Vegetation types of the greater cape floristic region, Fynbos: Ecology, evolution, and conservation of a megadiverse region. Oxford University Press, Oxford, UK, pp 1–25. Bond P., Goldblatt P. (1984) Plants of the Cape Flora: a descriptive catalogue. Journal of South African Bot- any, 7, 304–306. Bond W.J., Midgley G.F. (2012) Carbon dioxide and the uneasy interactions of trees and savannah grasses. Philosophical Transactions of the Royal Soci- ety, B: Biological Sciences, 367, 601–612. Born J., Linder H.P. (2018) Water availability, funda- mental niches and realized niches: a case study from the Cape flora. Austral Ecology, 6, 696–705. Briggs B.G., Linder H.P. (2009) A new subfamilial and tribal classification of Restionaceae (Poales). Telopea, 12, 333–345. Brodersen C.R., Roddy A.B., Wason J.W., McElrone A.J. (2019) Functional status of xylem through time. Annual Review of Plant Biology, 70, 407–433. Brodribb T., Brodersen C.R., Carriqui M., Tonet V., Rodriguez D.C., McAdam S. (2021) Linking xylem network failure with leaf tissue death. New Phytolo- gist, 232, 68–79. Brodribb T.J., Carriqui M., Delzon S., Lucani C. (2017) Optical measurement of stem xylem vulnerability. Plant Physiology, 174, 2054–2061. Brodribb T.J., Powers J., Cochard H., Choat B. (2020) Hanging by a thread? Forests and drought. Science, 368, 261–266. Brodribb T.J., Skelton R.P., MCAdam S.A.M., Bienaimé D., Lucani C.J., Marmottant P. (2016) Visual quanti- fication of embolism reveals leaf vulnerability to hydraulic failure. New Phytologist, 209, 1403–1409. Campbell B.M. (1983) Montane plant environments in the fynbos biome. Bothalia, 14, 283–298. Cao K.-F., Yang S.-J., Zhang Y.-J., Brodribb T.J. (2012) The maximum height of grasses is determined by roots: root pressure determines height in bamboo. Ecology Letters, 15, 666–672. Cardoso A.A., Batz T.A., McAdam S.A.M. (2020) Xylem embolism resistance determines leaf mortality during drought in Persea americana. Plant Physiol- ogy, 182, 547–554. Choat B., Brodribb T.J., Brodersen C.R., Duursma R.A., López R., Medlyn B.E. (2018) Triggers of tree mortality under drought. Nature, 558, 531–539. Cochard H., Badel E., Herbette S., Delzon S., Choat B., Jansen S. (2013) Methods for measuring plant vul- nerability to cavitation: a critical review. Journal of Experimental Botany, 64, 4779–4791. Corso D., Delzon S., Lamarque L.J., Cochard H., Torres-Ruiz J.M., King A., Brodribb T. (2020) Nei- ther xylem collapse, cavitation, or changing leaf con- ductance drive stomatal closure in wheat. Plant, Cell & Environment, 43, 854–865. Craine J.M., Ocheltree T.W., Nippert J.B., Towne E.G., Skibbe A.M., Kembel S.W., Fargione J.E. (2013) Global diversity of drought tolerance and grassland climate- change resilience. Nature Climate Change, 3, 63–67. Crawley M. (2012) Analysis of variance, The R book. John Wiley & Sons, Chichester, UK, pp 498–536. Domec J.-C., Gartner B.L. (2001) Cavitation and water storage capacity in bole xylem segments of mature and young Douglas-fir trees. Trees, 15, 204–214. Ehmig M., Coiro M., Linder H.P. (2019) Ecophysiolog- ical strategy switch through development in heteroblastic species of mediterranean ecosystems – An example in the African Restionaceae. Annals of Botany, 123, 611–623. Ehmig M., Linder H.P. (2020) Unexpected diversity and evolutionary lability in root architectural eco- morphs in the rushes of the hyperdiverse Cape flora. New Phytologist, 227, 216–231. Fay P.A., Carlisle J.D., Knapp A.K., Blair J.M., Collins S.L. (2003) Productivity responses to altered rainfall patterns in a C4 -dominated grassland. Oecologia, 137, 245–251. February E.C., Bond W.J., Taylor R., Newton R. (2004) Will water abstraction from the Table Mountain aquifer threaten endemic species? A case study at Cape Point, Cape Town. South African Journal of Sci- ence, 100, 253. February E.C., Higgins S.I., Bond W.J., Swemmer L. (2013) Influence of competition and rainfall manip- ulation on the growth responses of savanna trees and grasses. Ecology, 94, 1155–1164. Goldblatt P. (1978) An analysis of the flora of southern Africa: its characteristics, relationships, and orgins. Annals of the Missouri Botanical Garden, 65, 369–436. Goud E.M., Touchette S., Strachan I.B., Strack M. (2022) Graminoids vary in functional traits, carbon dioxide and methane fluxes in a restored peatland: implications for modelling carbon storage. Journal of Ecology, 110, 2105–2117. Hoffman M.T., Moll E.J., Boucher C. (1987) Post-fire succession at Pella, a south African lowland fynbos site. South African Journal of Botany, 53, 370–374. Holloway-Phillips M.-M., Brodribb T.J. (2011) Mini- mum hydraulic safety leads to maximum water-use efficiency in a forage grass: minimum hydraulic safety, maximum water-use efficiency. Plant, Cell & Environment, 34, 302–313. Huber M., Linder H.P. (2012) The evolutionary loss of aerenchyma limits both realized and fundamental ecohydrological niches in the Cape reeds (Restiona- ceae). Journal of Ecology, 100, 1338–1348. IPCC (2022) Climate change 2021: the physical science basis. Contribution of working group I to the sixth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cam- bridge, UK. Jacob V., Choat B., Churchill A.C., Zhang H., Barton C.V.M., Krishnananthaselvan A., Post A.K., Power S.A., Medlyn B.E., Tissue D.T. (2022) High safety margins to drought-induced hydraulic failure found in five pasture grasses. Plant, Cell & Environment, 45, 1631–1646. Johnson K.M., Jordan G.J., Brodribb T.J. (2018) Wheat leaves embolized by water stress do not recover func- tion upon rewatering. Plant, Cell & Environment, 41, 2704–2714. Jurjavcic N., Harrison S., Wolf A. (2002) Abiotic stress, competition, and the distribution of the native Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 645 West, Atkins, van Blerk & Skelton Assessing vulnerability to embolism in Restionaceae 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense annual grass Vulpia microstachys in a mosaic envi- ronment. Oecologia, 130, 555–562. Karp A.T., Faith J.T., Marlon J.R., Staver A.C. (2021) Global response of fire activity to late quaternary grazer extinctions. Science, 374, 1145–1148. Kassambara A. (2020) ggpubr: “ggplot2” based publi- cation ready plots. Knapp A.K., Fay P.A., Blair J.M., Collins S.L., Smith M.D., Carlisle J.D., Harper C.W., Danner B.T., Lett M.S., McCarron J.K. (2002) Rainfall variability, car- bon cycling, and plant species diversity in a Mesic grassland. Science, 298, 2202–2205. Lens F., Picon-Cochard C., Delmas C.E.L., Signarbieux C., Buttler A., Cochard H., Jansen S., Chauvin T., Doria L.C., Del Arco M., Delzon S. (2016) Herba- ceous angiosperms are not more vulnerable to drought-induced embolism than angiosperm trees. Plant Physiology, 172, 661–667. Linder H.P. (2018) African Restionaceae, electronic dataset. Available from https://www.systbot.uzh. ch/en/Bestimmungsschluessel/Restionaceae.html (accessed July 2022). Marloth R. (1903) Results of experiments on Table Mountain for ascertaining the amount of moisture deposited from the south-east clouds. Transactions of the South African Philosophical Society, 14, 403–408. Marloth R. (1905) Results of further experiments on Table Mountain for ascertaining the amount of moisture deposited from the south-east clouds. Transactions of the South African Philosophical Soci- ety, 16, 97–105. Martin-StPaul N., Delzon S., Cochard H. (2017) Plant resistance to drought depends on timely stomatal closure. Ecology Letters, 20, 1437–1447. McDowell N., Pockman W.T., Allen C.D., Breshears D.D., Cobb N., Kolb T., Plaut J., Sperry J., West A., Williams D.G., Yepez E.A. (2008) Mechanisms of plant survival and mortality during drought: why do some plants survive while others succumb to drought? New Phytologist, 178, 719–739. Mclaughlin B.C., Ackerly D.D., Klos P.Z., Natali J., Dawson T.E., Thompson S.E. (2017) Hydrologic refugia, plants, and climate change. Global Change Biology, 23, 2941–2961. Miller J.M., Miller P.C., Miller P.M. (1984) Leaf conduc- tances and xylem pressure potentials in fynbos plant species. South African Journal of Science, 80, 381–385. Miranda M.T., Pereira L., Pires G.S., Guan X., Silva L.M., Kreinert S., Machado E.C., Jansen S., Ribeiro R.V. (2023) Xylem sap residue in cut-open conduits can affect gas discharge in pneumatic experiments. Muggeo V.M.R. (2008) Segmented: an R package to fit regression models with broken-line relationships. R News, 8, 20–25. Nagel J.F. (1956) Fog precipitation on Table Mountain. Quarterly Journal of the Royal Meteorological Society, 82, 452–460. Nicotra A., Sack L., Santiago L., Cornwell W. (2010) PrometheusWiki: protocols, methods, explanations and updated standards for ecological and environ- mental plant physiology. Ocheltree T.W., Mueller K.M., Chesus K., LeCain D.R., Kray J.A., Blumenthal D.M. (2020) Identification of suites of traits that explains drought resistance and phenological patterns of plants in a semi-arid grass- land community. Oecologia, 192, 55–66. Ocheltree T.W., Nippert J.B., Prasad P.V.V. (2016) A safety vs efficiency trade-off identified in the hydrau- lic pathway of grass leaves is decoupled from photosynthesis, stomatal conductance and precipita- tion. New Phytologist, 210, 97–107. Paligi S.S., Link R.M., Isasa E., Bittencourt P., Cabral J.S., Jansen S., Oliveira R.S., Pereira L., Schuldt B. (2023) Assessing the agreement between the pneu- matic and the flow-centrifuge method for estimating xylem safety in temperate diffuse-porous tree spe- cies. Plant Biology, 25, 1171–1185. Pammenter N.W., van der Willigen C. (1998) A mathe- matical and statistical analysis of the curves illustrat- ing vulnerability of xylem to cavitation. Tree Physiology, 18, 589–593. Pereira L., Bittencourt P.R.L., Oliveira R.S., Junior M.B.M., Barros F.V., Ribeiro R.V., Mazzafera P. (2016) Plant pneumatics: stem air flow is related to embolism – New perspectives on methods in plant hydraulics. New Phytologist, 211, 357–370. Pereira L., Bittencourt P.R.L., Pacheco V.S., Miranda M.T., Zhang Y., Oliveira R.S., Groenendijk P., Machado E.C., Tyree M.T., Jansen S., Rowland L., Ribeiro R.V. (2020) The Pneumatron: an automated pneumatic apparatus for estimating xylem vulnera- bility to embolism at high temporal resolution. Plant, Cell & Environment, 43, 131–142. R Core Team (2022) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Rebelo A.G., Boucher C., Helme N., Mucina L., Ruth- erford M.C. (2006) Fynbos Biome, The vegetation of South Africa, Lesotho and Swaziland. South African National Biodversity Institute, Pretoria, South Africa, pp 53–218. Silvertown J., Araya Y., Gowing D. (2015) Hydrological niches in terrestrial plant communities: a review. Journal of Ecology, 103, 93–108. Silvertown J., Dodd M.E., Gowing D.J.G., Mountford J.O. (1999) Hydrologically defined niches reveal a basis for species richness in plant communities. Nature, 400, 61–63. Skelton R.P. (2014) The role of hydraulic strategies in understanding the response of fynbos to drought PhD Thesis. University of Cape Town, South Africa. Skelton R.P., Anderegg L.D.L., Diaz J., Kling M.M., Papper P., Lamarque L.J., Delzon S., Dawson T.E., Ackerly D.D. (2021) Evolutionary relationships between drought-related traits and climate shape large hydraulic safety margins in western north American oaks. Proceedings of the National Academy of Sciences of the United States of America, 118, e2008987118. Skelton R.P., Dawson T.E., Thompson S.E., Shen Y., Weitz A.P., Ackerly D. (2018) Low vulnerability to xylem embolism in leaves and stems of North Amer- ican oaks. Plant Physiology, 177(3), 1066–1077. https://doi.org/10.1104/pp.18.00103 Skelton R.P., West A.G., Buttner D., Dawson T.E. (2023) Consistent responses to moisture stress despite diverse growth forms within mountain fyn- bos communities. Oecologia, 201, 323–339. Skelton R.P., West A.G., Dawson T.E. (2015) Predict- ing plant vulnerability to drought in biodiverse regions using functional traits. Proceedings of the National Academy of Sciences of the United States of America, 112, 5744–5749. Skelton R.P., West A.G., Dawson T.E., Leonard J.M. (2013) External heat-pulse method allows compara- tive sapflow measurements in diverse functional types in a Mediterranean-type shrubland in South Africa. Functional Plant Biology, 40, 1076–1087. Staver A.C., Archibald S., Levin S. (2011) Tree cover in sub-Saharan Africa: rainfall and fire constrain forest and savanna as alternative stable states. Ecology, 92, 1063–1072. Staver A.C., Wigley-Coetsee C., Botha J. (2019) Grazer movements exacerbate grass declines during drought in an African savanna. Journal of Ecology, 107, 1482–1491. Swemmer A.M., Knapp A.K., Smith M.D. (2006) Growth responses of two dominant C4 grass species to altered water availability. International Journal of Plant Sciences, 167, 1001–1010. Trabi C.L., Pereira L., Guan X., Miranda M.T., Bitten- court P.R.L., Oliveira R.S., Ribeiro R.V., Jansen S. (2021) A user manual to measure gas diffusion kinetics in plants: pneumatron construction, opera- tion, and data analysis. Frontiers in Plant Science, 12, 633595. Tyree M.T., Sperry J.S. (1989) Vulnerability of xylem to cavitation and embolism. Annual Review of Plant Physiology and Plant Molecular Biology, 40, 19–36. van Blerk J.J., West A.G., Midgley J.J. (2017) Restio culm felling is a consequence of pre-dispersal seed predation by the rodent Rhabdomys pumilio in the fynbos. South African Journal of Botany, 112, 123– 127. van Blerk J.J., West A.G., Skelton R.P., Midgley J.J. (2022) Phenological asynchrony between sexes of Restionaceae can explain culm δ13C differences. Aus- tral Ecology, 47, 1200–1207. Van Sundert K., Arfin Khan M.A.S., Bharath S., Buck- ley Y.M., Caldeira M.C., Donohue I., Dubbert M., Ebeling A., Eisenhauer N., Eskelinen A., Finn A., Gebauer T., Haider S., Hansart A., Jentsch A., Kübert A., Nijs I., Nock C.A., Nogueira C., Porath- Krause A.J., Radujković D., Raynaud X., Risch A.C., Roscher C., Scherer-Lorenzen M., Schuchardt M.A., Schütz M., Siebert J., Sitters J., Spohn M., Virtanen R., Werner C., Wilfahrt P., Vicca S. (2021) Fertilized graminoids intensify negative drought effects on grassland productivity. Global Change Biology, 27, 2441–2457. von Willert D.J., Herppich M., Miller J.M. (1989) Pho- tosynthetic characteristics and leaf water relations of mountain fynbos vegetation in the Cedarberg area (South Africa). South African Journal of Botany, 55, 288–298. West A.G., Dawson T.E., February E.C., Midgley G.F., Bond W.J., Aston T.L. (2012) Diverse functional responses to drought in a Mediterranean-type shrub- land in South Africa. New Phytologist, 195, 396–407. Wickham H. (2016) ggplot2: elegant graphics for data analysis. Springer, New York, USA. Wickham H., Averick M., Bryan J., Chang W., McGo- wan L.D., François R., Grolemund G., Hayes A., Henry L., Hester J., Kuhn M., Pedersen T., Miller E., Bache S., Müller K., Ooms J., Robinson D., Seidel D., Spinu V., Takahashi K., Vaughan D., Wilke C., Woo K., Yutani H. (2019) Welcome to the tidyverse. Journal of Open Source Software, 4, 1686. Yang D., Pereira L., Peng G., Ribeiro R.V., Kaack L., Jansen S., Tyree M.T. (2022) A unit pipe pneumatic model to simulate gas kinetics during measurements of embolism in excised angiosperm xylem. Tree Physiology, 43, 88–101. Zhou Y., Singh J., Butnor J.R., Coetsee C., Boucher P.B., Case M.F., Hockridge E.G., Davies A.B., Staver A.C. (2022) Limited increases in savanna carbon stocks over decades of fire suppression. Nature, 603, 445–449. Plant Biology 26 (2024) 633–646 © 2024 The Authors. Plant Biology published by John Wiley & Sons Ltd on behalf of German Society for Plant Sciences, Royal Botanical Society of the Netherlands. 646 Assessing vulnerability to embolism in Restionaceae West, Atkins, van Blerk & Skelton 14388677, 2024, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/plb.13644 by U niversity O f W itw atersrand, W iley O nline L ibrary on [22/11/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://www.systbot.uzh.ch/en/Bestimmungsschluessel/Restionaceae.html https://www.systbot.uzh.ch/en/Bestimmungsschluessel/Restionaceae.html https://doi.org/10.1104/pp.18.00103 Abstract Sample selection and collection Field xylem water potential plb13644-fig-0001 Culm anatomy and vessel length Optical vulnerability measurements Pneumatic vulnerability measurements Pressure-volume curves Hydraulic safety margins Statistical analyses Culm anatomy with regard to optical and pneumatic methods Optical vulnerability curves Pneumatic vulnerability curves plb13644-fig-0002 Comparison of vulnerability curve methods plb13644-fig-0003 Vulnerability to embolism, pressure-volume curves and hydraulic safety margins plb13644-fig-0004 plb13644-fig-0005 plb13644-fig-0006 plb13644-fig-0007 plb13644-supitem References plb13644-bib-0001 plb13644-bib-0002 plb13644-bib-0003 plb13644-bib-0004 plb13644-bib-0005 plb13644-bib-0006 plb13644-bib-0007 plb13644-bib-0008 plb13644-bib-0009 plb13644-bib-0010 plb13644-bib-0011 plb13644-bib-0012 plb13644-bib-0013 plb13644-bib-0014 plb13644-bib-0015 plb13644-bib-0016 plb13644-bib-0017 plb13644-bib-0018 plb13644-bib-0019 plb13644-bib-0020 plb13644-bib-0021 plb13644-bib-0022 plb13644-bib-0023 plb13644-bib-0024 plb13644-bib-0025 plb13644-bib-0026 plb13644-bib-0027 plb13644-bib-0028 plb13644-bib-0029 plb13644-bib-0030 plb13644-bib-0031 plb13644-bib-0032 plb13644-bib-0033 plb13644-bib-0034 plb13644-bib-0035 plb13644-bib-0036 plb13644-bib-0037 plb13644-bib-0038 plb13644-bib-0039 plb13644-bib-0040 plb13644-bib-0041 plb13644-bib-0042 plb13644-bib-0043 plb13644-bib-0044 plb13644-bib-0045 plb13644-bib-0046 plb13644-bib-0047 plb13644-bib-0048 plb13644-bib-0049 plb13644-bib-0050 plb13644-bib-0051 plb13644-bib-0052 plb13644-bib-0053 plb13644-bib-0054 plb13644-bib-0055 plb13644-bib-0056 plb13644-bib-0057 plb13644-bib-0058 plb13644-bib-0059 plb13644-bib-0060 plb13644-bib-0061 plb13644-bib-0062 plb13644-bib-0063 plb13644-bib-0064 plb13644-bib-0065 plb13644-bib-0066 plb13644-bib-0067 plb13644-bib-0068 plb13644-bib-0069 plb13644-bib-0070 plb13644-bib-0071 plb13644-bib-0072 plb13644-bib-0073 plb13644-bib-0074 plb13644-bib-0075 plb13644-bib-0076 plb13644-bib-0077 plb13644-bib-0078 plb13644-bib-0079 plb13644-bib-0080 plb13644-bib-0081 plb13644-bib-0082 plb13644-bib-0083 plb13644-bib-0084