Xylem adjusts to maintain efficiency across a steep precipitation gradient in two coexisting generalist species

Xylem adjusts to maintain efficiency across a steep precipitation gradient in two coexisting... Abstract Background and Aims Trees adjust the configuration of their conductive system in response to changes in water availability, maximizing efficiency in wet environments and increasing safety in dry habitats. However, evidence of this general trend is not conclusive. Generalist species growing across broad climatic gradients provide an ideal framework to assess intra-specific xylem adjustments under contrasting environmental conditions. Our aims were to compare the response of xylem traits to variations in precipitation of two co-occurring generalist tree species, and to assess climate control on xylem trait variability and co-ordination. Methods We evaluated xylem traits of Embothrium coccineum (Proteaceae, evergreen) and Nothofagus antarctica (Nothofagaceae, deciduous) in three areas across an abrupt precipitation gradient, from 500 to 2500 mm, in southern Chile. We measured wood density, vessel lumen area and density, percentage of conductive area and vessel grouping, and estimated the hydraulic function from anatomical measurements in 60 individuals per species. Key Results Both species shared a common pattern of response along the precipitation gradient, with an increase in vessel density with dryness, but without changes in estimated hydraulic conductivity. Xylem traits in E. coccineum were more variable and more responsive to the climate gradient, decreasing vessel lumen area and increasing wood density, whereas vessel grouping showed contrasting patterns between species. Additionally, the analysis of trait co-ordination at the individual level revealed a tighter co-ordination among xylem traits in E. coccineum. Conclusions Estimated xylem efficiency was maintained in combination with different levels of expected xylem safety within species. Reduction in vessel lumen area was compensated through large increases in vessel density, thus breaking the trade-off between xylem efficiency and safety. Otherwise, the existence of alternative internal adjustments in coexisting species to face similar climatic constraints might increase resilience of temperate forests against unpredictable changes in climatic conditions. Embothrium coccineum, Nothofagus antarctica, Patagonia, precipitation gradient, temperate rainforest, vessel grouping, vessel traits, wood density, xylem trait co-ordination INTRODUCTION Increasing temperature and a higher frequency and intensity of drought events are triggering worldwide tree dieback (e.g. Allen et al., 2010; Sánchez-Salguero et al., 2012; Greenwood et al., 2017), clearly evidencing that climate change is already impacting forest ecosystems. Understanding the response of forests to this novel scenario requires a better comprehension of plants’ adjustments to climate variability. Natural gradients provide an ideal framework to assess this question (De Frenne et al., 2013), since they allow the evaluation of species’ adjustments across contrasting environmental conditions. The adaptive potential of widespread species has been attributed to intra-specific variation in functional traits (Burns, 2004; Richards et al., 2005; Violle and Jiang, 2009); thus, linking patterns of trait variation with species’ ability to inhabit environmental gradients may contribute to prediction of community responses to local or global environmental changes and hence provide valuable insights into the mechanisms that control local functional diversity. Water availability is considered a key factor underlying trait variability among populations and communities (Maherali et al., 2004). The relationship between plants and water flow is mediated by the configuration of the hydraulic system (Tyree and Zimmermann, 2002). Hydraulic efficiency of the stem xylem conductive system is determined by hydraulic conductivity, i.e. the volume of sap that can be moved through the xylem normalized by time, cross-sectional area, length and pressure gradient. Hydraulic safety refers to the ability to maintain xylem conductivity at decreasing water potential, i.e. without occurrence of embolism (Hacke and Sperry, 2001; Tyree and Zimmermann, 2002). Efficiency and safety have been considered to represent a trade-off in hydraulic system functionality (Sperry, 2003). As a result, xylem traits show conspicuous anatomical responses to environmental conditions. The vessel lumen area of angiosperms diminishes in plants inhabiting dry or cold areas, and increases in moist and warm environments (Sperry et al., 2008; Hacke et al., 2017), because wide vessels prioritize water conductivity at the expense of higher embolism risk by either drought or frost (Tyree and Sperry, 1989). However, evidence of this safety–efficiency trade-off is not conclusive, and some species may show simultaneously reduced xylem safety and low hydraulic efficiency (Maherali et al., 2004; Gleason et al., 2016a). Several potential factors may modulate the relationship between the conductive system of plants and the environment. First, wood traits are co-ordinated with leaf-level adjustments (Castro-Díez et al., 1998; Martínez-Vilalta et al., 2009), especially under harsh climatic conditions (Zeballos et al., 2017). For instance, leaves may close their stomata to avoid water losses by evapotranspiration when water is scarce, thus reducing embolism risk (Martínez-Vilalta and García-Forner, 2016). Secondly, leaf habit may alter the relationship between xylem and climate (Castro-Díez et al., 1998; Pérez-de-Lis et al., 2018). Deciduous trees shed leaves during the unfavourable season and thus reduce plant hydraulic demands, enabling them to tolerate large hydraulic conductance losses (Sperry et al., 1994). As a result, deciduous species may have larger vessels compared with evergreen congenerics, since vessel lumen area is less constrained by unfavourable conditions (Castro-Díez et al., 1998; Gorsuch et al., 2001). The relationship of xylem traits to climate may also shift between sexes in dioecious species (Olano et al., 2017), and in other cases xylem traits may even be unresponsive to environmental gradients (Martínez-Vilalta et al., 2009). All these factors make xylem adjustments across environmental gradients not easily predictable for coexisting species differing in leaf traits, leaf habit or reproductive conditions. In this study, we focused on two widespread, small-statured tree species with contrasting foliar habit that grow in southern Chile, the evergreen Embothrium coccineum and the deciduous Nothofagus antarctica. We evaluated xylem adjustments of both species across an abrupt gradient of annual precipitation from inland (500 mm) to coastal (2500 mm) forests in Patagonia, Chile, also considering variations in temperature. We measured wood density, vessel lumen area and density, percentage of conductive area and vessel grouping, and estimated the hydraulic function from anatomical measurements (hydraulically weighted mean diameter and xylem-specific hydraulic conductivity) (1) to assess whether xylem traits show similar responses to the environmental gradient in both species; and (2) to estimate to what extent xylem trait variability was controlled by climate. We also evaluated co-ordination among xylem traits in both species, since inter-relations among xylem traits commonly observed in inter-specific comparisons, such as the positive correlations between conduit dimensions and hydraulic conductivity, and between wood density and vulnerability to embolism (Hacke et al., 2001; Chave et al., 2009; Lachenbruch and McCulloh, 2014; Rosner, 2017), might be weak at the intra-specific level (Martínez-Vilalta et al., 2009; Schreiber et al., 2011). We hypothesized that patterns of xylem trait variation along the precipitation gradient would be similar in both species according to the safety–efficiency trade-off. Despite the phylogenetic distance between both species, their xylem anatomies were expected to differ. We also hypothesized that xylem trait variation in the deciduous N. antarctica would show weaker correlation with climate than in the evergreen E. coccineum. MATERIALS AND METHODS Species and research site descriptions Embothrium coccineum J.R. & G. Forster (Proteaceae) and Nothofagus antarctica (Foster) Oerst. (Nothofagaceae) are two small-statured tree species (5–8 m mean height) that are evergreen and deciduous, respectively. Both species have a broad latitudinal distribution throughout Chile and Argentina, from central Chile (34°S) to Tierra del Fuego (55°S), ranging from superhumid to dry sites, and from sea level to, in many cases, the treeline limit (Fajardo and Piper, 2015). In the Aysén region (Patagonia, Chile), these two species prevail across an abrupt precipitation gradient (2500 to 600 mm year–1, 9.2–11 °C temperature difference between mean warmest and mean coldest months; Fig. 1; Table 1). In this study, we worked in three areas across this precipitation gradient that differ considerably in climatic conditions and hence in vegetation. Fig. 1. View largeDownload slide Location of the study areas (wet, mesic and dry), sites (four per area) and climatic diagrams for each study area. Ak, Aiken Park; At, Atravesado Lake; Ci, Baguales; Cl, Claro River; Cu, Cuervo River; Ib, Puerto Ibáñez; Lv, Levicán Peninsula; Pa, Pangal Valley; Pf, head of Pangal Valley; Rc, Reserva Coyhaique; Ro, Rosado Mountain; Sa, Ibáñez River waterfall. Note that central months in climatic diagrams correspond to summer in the southern hemisphere. Climatic data were obtained from the Climate Forecast System Reanalysis (CFSR) data set, compiled by the National Center for Environmental Prediction (NCEP), USA (Globalweather 2016). The scale bar in the region map represents 50 km. Fig. 1. View largeDownload slide Location of the study areas (wet, mesic and dry), sites (four per area) and climatic diagrams for each study area. Ak, Aiken Park; At, Atravesado Lake; Ci, Baguales; Cl, Claro River; Cu, Cuervo River; Ib, Puerto Ibáñez; Lv, Levicán Peninsula; Pa, Pangal Valley; Pf, head of Pangal Valley; Rc, Reserva Coyhaique; Ro, Rosado Mountain; Sa, Ibáñez River waterfall. Note that central months in climatic diagrams correspond to summer in the southern hemisphere. Climatic data were obtained from the Climate Forecast System Reanalysis (CFSR) data set, compiled by the National Center for Environmental Prediction (NCEP), USA (Globalweather 2016). The scale bar in the region map represents 50 km. Table 1. Climatic characteristics of the study sites (based on the data set obtained from ClimateSA v1.0 for the period 1981–2010; see text) Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  MAT, mean annual temperature; MAP, mean annual precipitation; ACMT, average coldest month temperature; NFFD, number of frost-free days; CMD, Hargreaves climatic moisture deficit. View Large Table 1. Climatic characteristics of the study sites (based on the data set obtained from ClimateSA v1.0 for the period 1981–2010; see text) Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  MAT, mean annual temperature; MAP, mean annual precipitation; ACMT, average coldest month temperature; NFFD, number of frost-free days; CMD, Hargreaves climatic moisture deficit. View Large The wettest area was located around the city of Puerto Aysén (45°24′27″S, 72°41′39″W, approx. 30 m a.s.l.), at the westernmost part of the gradient. Here, annual precipitation is 2034 mm year–1 (Puerto Aysén weather station, 32 m a.s.l.; Dirección General de Aguas, 2007–2017). Precipitation is evenly distributed throughout the year. Mean summer temperature (December–February) is 12.7 °C, whereas mean winter temperature (June–August) is 4.4 °C (data for the period 1981–2010 generated with the ClimateSA v1.0 software package, available at http://tinyurl.com/ClimateSA;Hamann et al., 2013). In this area, both target species occur within a species-rich temperate rainforest that prevails under a hyper-humid climatic regime (Luebert and Pliscoff, 2006). Dominant species include Drimys winteri (Winteraceae), Nothofagus betuloides and N. nitida (Nothofagaceae), Amomyrtus luma (Myrtaceae), Laureliopsis philippiana (Atherospermataceae), Lomatia ferruginea (Proteaceae), Caldcluvia paniculata and Weinmannia trichosperma (Cunnoniaceae), Podocarpus nubigenus (Podocarpaceae) and Pilgerodendron uviferum (Cupressaceae), among others. The mesic area was located around the city of Coyhaique (45°34′47″S, 72°03′40″W, approx. 280 m a.s.l.). Mean annual precipitation is 910 mm year–1 (Coyhaique National Reserve weather station, 400 m a.s.l.; Dirección General de Aguas, 2002–2015), with a slight drop in precipitation in February. This is the coldest area, since mean summer temperature (December–February) is 11.2 °C, whereas mean winter temperature (June–August) is 1.7 °C. The forest here has lower species richness, and is dominated by Nothofagus pumilio, N. antarctica, N. dombeyi, E. coccineum, Ribes magellanicum (Saxifragaceae) and Discaria chacaye (Rhamnaceae). The dry area was located around the village of Puerto Ibáñez (46°17′35″S, 71°56′19″W, approx. 320 m a.s.l.). Mean annual precipitation is 580 mm, with wet winters (around 80–90 mm month–1) and relatively dry summers (around 40–50 mm month–1). Mean summer temperature (December–February) is 12.4 °C, whereas mean winter temperature (June–August) is 2.3 °C. Vegetation here is a sclerophyll, short-stature forest dominated by D. chacaye, Schinus marchandii (Anacardiaceae), Maytenus boaria (Celastraceae), Colletia spinosissima (Rhamnaceae), E. coccineum and N. antarctica. Sampling, tissue collection and processing Sampling was conducted in November–December 2013. Four sites separated by a minimum distance of 5 km were selected in each of the three contrasting precipitation areas to enhance replication and include variability within areas (Fig. 1; Table 1). At each site, five individuals per species were sampled. We restricted the sampling to adult, unshaded trees without browsing or other apparent damage. For each individual tree, we cut at approx. 2 m height one 2 m long terminal, sun-exposed branch with fully expanded leaves using a telescoping pole (ARS Corporation, Sakai, Japan). This standardization of branch length minimizes xylem anatomical and hydraulic differences related to distance to the branch tip (Carrer et al., 2015). For each branch, we selected and cut two 2 cm wide pieces of wood of similar diameter (1.5–2 cm). Both wood pieces were labelled and placed in a cooler for transportation. In the laboratory, the fresh volume for one of the two pieces of branch per individual was determined by submerging the wood (without bark) into a glass beaker on a scale. The mass difference caused by the sample, which equals the volume of water displaced, was recorded and converted to volume based on the density of water as 1.0 g cm–3 at standard temperature and pressure. Samples were then dried in a forced-air oven (Memmert GmbH, Schwabach, Germany) at 70 °C for 72 h and the dry mass was subsequently measured. Wood density (WD) was then calculated as the oven-dry mass per green volume (Williamson and Wiemann, 2010). Quantitative wood anatomy The second piece of branch was used to measure xylem anatomy. Xylem anatomical analyses followed the protocol proposed by von Arx et al. (2016). Anatomical cross-sections of 10 μm thickness from the remaining pieces of wood were produced with a sledge microtome (Gärtner et al., 2015). These cross-sections were then placed on a slide and stained with Alcian blue (1 % solution in acetic acid) and safranin (1 % solution in ethanol). Afterwards, the cross-sections were dehydrated using a series of ethanol solutions of increasing concentration, washed with xylol and permanently preserved by embedding them in Eukitt glue. Overlapping images covering a complete radius from pith to bark were captured with a Nikon D90 digital camera mounted on a Nikon Eclipse 50i optical microscope with ×40 magnification (corresponding to a resolution of 0.6134 pixels μm–1) and merged to a single image using PTGUI v8.3.10 Pro (New House Internet Services B.V., Rotterdam, The Netherlands). In each image, we measured seven anatomical traits associated with hydraulic function in the last complete annual ring (i.e. the ring formed in the growing season 2012–2013). Five of these traits were directly measured from images, and the others were derived from these measurements. Directly measured traits were: (1) vessel density (VD; no. mm–2), which is the number of vessels per mm2; (2) mean vessel lumen area (MVA; μm2), which is the average lumen area of vessels; (3) percentage of conductive area per ring (CA), obtained as the cumulative lumen area of all counted vessels divided by the ring area; (4) vessel grouping index per annual ring (VGI), which is the mean number of vessels with contiguous cell walls (von Arx et al., 2013); and (5) vessel solitary fraction per annual ring (VSF), i.e. the percentage of solitary vessels with respect to all vessels in the ring. From these data, we calculated two derived traits: (6) hydraulically weighted mean diameter (Dh; μm), based on the vessels’ contribution to hydraulic conductance according to the Hagen–Poiseuille law, obtained as Dh = Σd5/Σd4, where d is the lumen diameter of each vessel (see Kolb and Sperry, 1999); and (7) an anatomy-based estimate of specific hydraulic conductivity (Ks; m2 s–1 MPa–1), which is the hydraulic conductivity per unit area, where the hydraulic conductance of each cell is calculated following Nonweiler (1975) and considering the ovality of cells (see equations in Olano et al., 2017). Vessel anatomical features were measured using ROXAS v3.0 (von Arx and Dietz, 2005; von Arx and Carrer, 2014), a specific image-analysis tool based on Image-Pro Plus (Media Cybernetics, Silver Spring, MD, USA). We first adjusted ROXAS settings to create different configurations of parameters used for automatic identification of vessel lumina. For each species, we performed a preliminary visual exploration to adjust the maximum and minimum vessel lumen area, vessel ovality and several colour parameters. With this information, we created specific configurations for the automatic analysis. In some cases, several configurations differing in parameters related to colour and vessel shape were needed to account for the variability in image quality. We used these configurations to analyse all samples automatically. The automatic output was then manually edited by drawing the last ring boundaries, deleting erroneously detected vessels (e.g. in parenchyma rays) and rectifying additional misidentifications. Then, exclusion areas were used to leave out parts of the image with lower quality (e.g. broken vessels introduced during sample preparation). In the last step of the manual editing, the 100 largest vessels per image were checked using the ‘Outlier search’ tool in ROXAS to find vessels that remained erroneously merged during the automatic analysis due to the presence of low contrast cell walls in some parts of the image. In high-quality images, this outlier search was only performed until no correction was needed after checking 20 consecutive vessels. Climatic data We used the ClimateSA v1.0 software package to generate data for the period 1981–2010 for each of the 12 sampling sites. This software is available at http://tinyurl.com/ClimateSA (Hamann et al., 2013). Climatic data have been developed with the parameter-elevation regressions on an independent slopes model (PRISM), an expert interpolation approach described by Daly et al. (2008) which uses physiographic information to better predict climate patterns in mountainous terrain. Although sites within the same area were closer to each other than to sites within the other areas, they were far enough from one another to show differences in climate. We obtained data of mean annual temperature, mean coldest month temperature, number of frost-free days, mean annual precipitation and Hargreaves climatic moisture deficit (Table 1). Climatic moisture deficit was obtained as the sum of the monthly difference between a reference evaporation calculated with the Hargreaves equation with a latitude correction applied (see Hamann et al., 2013) and precipitation. Data analyses Vessel lumen area and their contribution to potential hydraulic conductivity. For each species and site, we calculated the frequency distribution of vessels according to their lumen area. We grouped vessels by lumen area into 200 μm2 and 100 μm2 breaks (for E. coccineum and N. antarctica, respectively), i.e. we counted the number of vessels whose lumen area fell under each of these breaks or size classes. We also obtained the relative contribution of each of these classes to total water flow based on theoretical conductivity of each vessel in the class, according to Poiseuille’s law. Frequency distributions of vessels in each class and theoretical hydraulic conductivity were compared at the species level between sites using two-sample Kolmogorov–Smirnov (K–S) tests (von Arx et al., 2012). The vessel lumen area class at which 50 % of total water flow is achieved was also obtained per species and site. Variation in xylem traits along the gradient. In a preliminary step, we evaluated redundancy among xylem traits (i.e. the seven anatomical variables and WD) by examining the correlation matrix among them. Vessel solitary fraction and MVA were excluded from subsequent statistical analyses due to their high correlations with VGI and Dh, respectively (r > 0.9 in both cases). We previously obtained coefficients of variation (CVs) of all xylem traits per species within sites (n = 5 individuals per site of each species), within precipitation areas (n = 20 individuals per area of each species) and at the species level (n = 60 individuals of each species). To evaluate intra-specific variations in xylem traits along the precipitation gradient, we fitted separate linear mixed models (LMMs) per species and trait. We considered the study area of precipitation as a fixed factor with three levels (wet, mesic and dry). Site was included as a random factor nested in area to incorporate the grouping structure of our data, since individuals within a site can show trait values more similar to each other than to individuals from other sites within a given area. Models were fitted using restricted maximum likelihood (REML), and normalized residuals were extracted. Residuals were checked for normality and homoscedasticity, and models were re-fitted allowing them to have different variances per area in cases where these assumptions were not met. We fitted revised models with maximum likelihood, compared them using the Akaike information criterion (AIC), selected models that minimized the AIC value and checked again for normality and homoscedasticity of residuals. Finally, the optimal models were re-fitted with REML (Zuur et al., 2009). We used the ‘nlme’ package (Pinheiro et al., 2016) in R version 3.4.1 environment (R Core Team, 2016) to fit LMM. We also compared global patterns of intra-specific variation in xylem traits along the precipitation gradient for the two species. To do so, we performed a principal component analysis (PCA) on an individual × xylem traits matrix to determine their variation structure. Then, we used canonical ordination techniques to evaluate the effect of species and area (and their interaction) on xylem traits. We performed a redundancy analysis (RDA; ter Braak, 1986), which combines multiple regressions with PCA, to relate the dependent matrix (xylem traits) to an explanatory matrix (species and area). We also performed partial RDA analyses to discriminate the independent effects of species and area on xylem traits. In all cases, the significance of explanatory variables was evaluated using a Monte Carlo test with 9999 permutations. These analyses were performed with the ‘vegan’ package (Oksanen et al., 2016). Effects of climate on xylem trait variation. We used Pearson correlations to relate mean coldest month temperature and climatic moisture deficit with xylem traits. In addition, we used structural equation models (SEMs) to evaluate direct and indirect effects of these climatic variables on VD, Dh, Ks and WD, as well as the inter-relationships among these xylem traits and to CA, following the suggestion by Gleason et al. (2016b). These models were based on the hypothesized xylem a djustments to differences in water availability and continentality (see the Introduction). We adjusted two models, one per species. Model parameters were estimated with maximum likelihood, and global model fit was assessed by using the goodness of fit index (GFI). The magnitude and significance of the direct and indirect effects were estimated from the standardized path coefficients by a multivariate Wald test. Structural equation models were performed with AMOS 18.0 software (AMOS Development Corp., Mount Pleasant, SC, USA). RESULTS Xylem anatomy and vessel lumen area distributions Nearly one hundred thousand (97 711) vessels were measured in the last complete growth ring of the 120 sampled trees. In N. antarctica, the arrangement of the vessels was diffuse-porous with a slight tendency to semi-ring porosity (Detmann et al., 2013), whereas the xylem of E. coccineum was semi-ring porous with vessels arranged in tangential bands (Fig. 2). Vessel lumen area showed large variability, ranging from 38 to 4389 μm2 in E. coccineum, and from 27 to 2058 μm2 in N. antarctica. Small vessels were by far the most abundant in both species (Fig. 3A, B). However, the relative proportion of large vessels was higher in wet sites than in mesic and dry sites (P < 0.001 for all paired K–S tests comparing frequency distributions of vessel lumen area). This shift towards larger vessels in wet sites had a strong impact on their relative contribution to potential hydraulic conductivity (Fig. 3C, D). The vessel lumen area class at which 50 % of total water flow occurred was larger at the wet site, especially in E. coccineum (P < 0.001 for all paired K–S tests comparing curves of estimated hydraulic conductivity). In other words, small changes in vessel lumen size distribution from wet to dry areas led to strong changes in vessel class contribution to water flow. Fig. 2. View largeDownload slide Cross-sections of branch segments of the study species. Measured vessel lumina are filled in light blue, ring borders in yellow. Scale bars represent 100 μm. Fig. 2. View largeDownload slide Cross-sections of branch segments of the study species. Measured vessel lumina are filled in light blue, ring borders in yellow. Scale bars represent 100 μm. Fig. 3. View largeDownload slide Frequency distributions of vessel number and potential vessel hydraulic conductivity per vessel lumen area classes. Vessel lumen areas were divided in 200 μm2 size and 100 μm2 size classes for E. coccineum and N. antarctica, respectively. (A and B) The proportion of vessels in each size class. (C and D) The contribution of estimated hydraulic conductivity of vessels in each size class to total conductivity per ring. Distributions are presented separately per species and area. Triangles represent the size class at which 50 % of conductivity is reached. Fig. 3. View largeDownload slide Frequency distributions of vessel number and potential vessel hydraulic conductivity per vessel lumen area classes. Vessel lumen areas were divided in 200 μm2 size and 100 μm2 size classes for E. coccineum and N. antarctica, respectively. (A and B) The proportion of vessels in each size class. (C and D) The contribution of estimated hydraulic conductivity of vessels in each size class to total conductivity per ring. Distributions are presented separately per species and area. Triangles represent the size class at which 50 % of conductivity is reached. Variation in xylem traits along the gradient Xylem traits were in general more variable in E. coccineum than in N. antarctica (Table 2). Wood density was the least variable trait (<10 % of variation in all cases), whereas VD reached >50 % of variation in E. coccineum. The highest variation was found at the species level (i.e. considering all sampled individuals), but variability within areas was very close to the global variation in all traits of both species. Variation within sites was clearly lower than within areas. Table 2. Coefficients of variation of xylem traits observed within sites (CVsite, n = 5), within areas (CVarea, n = 20) and at the species level (CVtotal, n = 60)   Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99    Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99  Average values of CVsite and CVarea are shown. WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Table 2. Coefficients of variation of xylem traits observed within sites (CVsite, n = 5), within areas (CVarea, n = 20) and at the species level (CVtotal, n = 60)   Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99    Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99  Average values of CVsite and CVarea are shown. WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Three xylem traits varied along the precipitation gradient in at least one species. Wood density increased significantly from wet to mesic and dry areas in E. coccineum (0.557 ± 0.006 g cm–3; mean ± s.e. at the species level), whereas it did not differ among areas in N. antarctica (0.562 ± 0.004 g cm–3, Fig. 4A). The hydraulically weighted mean diameter in E. coccineum (35.7 ± 0.7 μm) was significantly higher in the wet area than in the mesic and dry areas (Fig. 4B), but this trait remained constant in N. antarctica (28.0 ± 0.4 μm). Vessel density (187 ± 13 vessels mm–2 in E. coccineum; 434 ± 16 vessels mm–2 in N. antarctica) varied along the gradient in both species, increasing significantly from wet to dry sites (Fig. 4C). The percentage of the conductive area (10.68 ± 0.60 % in E. coccineum; 17.85 ± 0.64 % in N. antarctica) remained constant along the gradient (Fig. 4D), and so did Ks (2.33 × 10–7 ± 1.44 10–8 kg m–2 MPa–1 s–1 in E. coccineum; 3.39 × 10–7±137 10–8 kg m–2 MPa–1 s–1 in N. antarctica; Fig. 4E). Vessel grouping revealed a non- significant trend towards reduced VGI and increased VSF from wet to dry areas in E. coccineum (Fig. 4F). Nothofagus antarctica showed the inverse trend, but this was still not significant. Fig. 4. View largeDownload slide Mean values and confidence intervals at P = 0.95 of xylem traits per area for Nothofagus antarctica and Embothrium coccineum. Letters indicate significant differences obtained with linear mixed models considering area (wet, mesic and dry) as a fixed factor and site nested in area as a random factor. Fig. 4. View largeDownload slide Mean values and confidence intervals at P = 0.95 of xylem traits per area for Nothofagus antarctica and Embothrium coccineum. Letters indicate significant differences obtained with linear mixed models considering area (wet, mesic and dry) as a fixed factor and site nested in area as a random factor. Global variation patterns in xylem traits The first two axes of the PCA explained 81.05 % of the variance in xylem traits (Fig. 5). The first axis was related to CA, VD and Ks, and sorted individuals according to species identity. Individuals of N. antarctica showed in general higher VD, CA and Ks than individuals of E. coccineum. These three traits were orthogonal to WD and Dh, which were negatively correlated (r = –0.46; P < 0.001) and drove variation along the second axis. This axis was related to the precipitation gradient, with individuals showing higher WD and lower Dh at the dry area, and the opposite pattern at the wet area. The RDA confirmed the impact of species and moisture gradient on xylem configuration, as both factors explained 61.87 % of the variability in xylem traits (F = 36.994, P = 0.001). Partial RDAs determined the overwhelming effect of species (52.18 % of explained variance; F = 147.09, P < 0.001), but also the robustness of the effect of the climatic gradient (6.89 % of explained variance associated with area, F = 9.72, P < 0.001). Species per area interaction was not significant (P = 0.769), indicating a common response of both species in different areas. Fig. 5. View largeDownload slide Principal component analysis (PCA) biplot showing the position of individual branches along the two first principal component axes. Colours indicate areas along the gradient (black = wet, dark grey = mesic, light grey = dry). Large open symbols and error bars represent average position and standard errors per species and area. Abbreviations: VD, vessel density; Dh, hydraulically weighted mean diameter; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity; CA, percentage of conductive area; VGI, vessel grouping index; WD, wood density. Fig. 5. View largeDownload slide Principal component analysis (PCA) biplot showing the position of individual branches along the two first principal component axes. Colours indicate areas along the gradient (black = wet, dark grey = mesic, light grey = dry). Large open symbols and error bars represent average position and standard errors per species and area. Abbreviations: VD, vessel density; Dh, hydraulically weighted mean diameter; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity; CA, percentage of conductive area; VGI, vessel grouping index; WD, wood density. Climatic control on xylem trait variation Correlations and SEM showed a stronger climatic control on xylem traits in E. coccineum than in N. antarctica. The average coldest month temperature correlated negatively with WD and VSF, and positively with Dh and MVA in E. coccineum (Table 3). In contrast, temperature was only marginally correlated with VD in N. antarctica. The effect of climatic moisture deficit was more similar between species. Climatic moisture deficit correlated positively with VD and negatively with Dh and MVA in both species, although negative correlations were only marginally significant in N. antarctica. An additional positive correlation between climatic moisture deficit and WD was observed in E. coccineum. Table 3. Pearson correlation coefficients between climatic variables [mean coldest month temperature, climatic moisture deficit (CMD)] and xylem traits   E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201    E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201  Significant correlations are highlighted in bold (P < 0.05) and in bold italics (P < 0.1). WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Table 3. Pearson correlation coefficients between climatic variables [mean coldest month temperature, climatic moisture deficit (CMD)] and xylem traits   E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201    E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201  Significant correlations are highlighted in bold (P < 0.05) and in bold italics (P < 0.1). WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Structural equation models showed that neither temperature nor climatic moisture deficit affected xylem conductivity (Ks) directly, but rather via their influence on xylem traits and CA (especially in E. coccineum; Fig. 6). Covariation between temperature and climatic moisture deficit resulted in the effects of temperature on xylem traits being closely aligned with the effects of moisture deficit. In N. antarctica, the only xylem trait affected by climate was Dh, which decreased as climatic moisture deficit increased (Fig. 6A). Increases in Dh led to increases in CA, which was in turn the main determinant of Ks. However, CA was also strongly affected by VD, which was independent of climatic moisture deficit at the individual level in this species. In E. coccineum, both Dh and VD were controlled by climatic moisture deficit and co-varied negatively, indicating that in drier sites, trees had smaller vessels and higher vessel density (Fig. 6B). In addition, trees in drier sites had denser wood, as WD was directly controlled by climatic moisture deficit but also indirectly by Dh and by its covariation with VD. In the same way as for N. antarctica, the main determinant of Ks was CA, which in turn was strongly controlled by VD but also by Dh. As a result, both Dh and VD determined Ks in E. coccineum through their effect on CA, but also through direct effects. Interestingly, CA did not affect WD in any species. Both SEMs showed good fit to the data (GFI = 0.999 for E. coccineum, GFI = 0.992 for N. antarctica; values above 0.90 indicate a good fit). Fig. 6. View largeDownload slide Adjusted structural equation models (SEMs) for the effects of climate (mean coldest month temperature and climatic moisture deficit) on xylem traits, and inter-relations among xylem traits. Arrows indicate the direction of evaluated relationships (paths). Arrow width is proportional to path coefficients. Significant path coefficients (P < 0.05) are shown, while non-significant paths appear as light grey arrows. Ks, estimated xylem-specific hydraulic conductivity. Goodness of fit index (GFI) = 0.999 for E. coccineum and 0.992 for N. antarctica. Fig. 6. View largeDownload slide Adjusted structural equation models (SEMs) for the effects of climate (mean coldest month temperature and climatic moisture deficit) on xylem traits, and inter-relations among xylem traits. Arrows indicate the direction of evaluated relationships (paths). Arrow width is proportional to path coefficients. Significant path coefficients (P < 0.05) are shown, while non-significant paths appear as light grey arrows. Ks, estimated xylem-specific hydraulic conductivity. Goodness of fit index (GFI) = 0.999 for E. coccineum and 0.992 for N. antarctica. DISCUSSION Both species shared most anatomical adjustments to the precipitation gradient, despite their different xylem anatomies and phylogenetic distance, supporting our first hypothesis. Only vessel grouping showed contrasting patterns in both species. Otherwise, the strength of climatic control on xylem traits varied between species: climatic control on xylem traits was tighter in the evergreen E. coccineum than in the deciduous N. antarctica, also supporting our second hypothesis. In our assessment, some of the studied traits remained constant across the gradient when they were compared among areas; however, when the focus was established on the analysis of trait co-ordination at the individual level, additional xylem adjustments were revealed. Variation in xylem traits along the gradient Vessel density increased at the drier edge of the gradient in both species, and the evergreen E. coccineum also experienced a decrease in Dh. Small vessels are indisputably more resistant to embolism, thus a reduction in the lumen area of the largest vessels under dry conditions could be interpreted as an increase in xylem safety (Tyree and Zimmermann, 2002). Higher xylem safety as a response to dryness has been previously reported within and across species, although some discrepancies emerge among studies. Across species, Maherali et al. (2004) observed increased xylem safety with dryness in evergreen angiosperms but detected a lack of correlation between climate and hydraulic safety in deciduous angiosperms. A similar pattern was observed among congeneric species with contrasting leaf habit (Villar-Salvador et al., 1997). Our results partially agree with both studies, since the deciduous N. antarctica experienced a similar pattern of Dh reduction with dryness, although it was only marginally significant. In this sense, our observations do agree with observed within-species reduction in MVA (strongly correlated with Dh) along aridity gradients for both evergreen (Fisher et al., 2007; von Arx et al., 2012) and deciduous species (Schreiber et al., 2015). A striking result of our study is that the reduction in Dh was not linked to a drop of Ks, probably because of the simultaneous increase in VD. Thus, estimated xylem efficiency was maintained in combination with different levels of expected xylem safety, departing from the safety–efficiency trade-off hypothesis (Sperry, 2003; Zanne et al., 2010). In addition, this departure might be related to the existence of different anatomical traits that operate simultaneously to determine total xylem efficiency (such as vessel diameter and length, inter-vessel pit membrane thickness and/or vessel connectivity), and to the fact that each of these features may have separate and different effects on safety (Gleason et al. 2016a). Changes in anatomical features did not occur linearly across the gradient. In fact, vessel traits in the dry and mesic areas showed less difference from each other than with respect to the wet area. Specifically, WD and Dh in E. coccineum showed similar average values in the dry and mesic areas. The existence of significant correlations of these traits with both climatic moisture deficit and temperature of the coldest month may give some clues to explain this similarity between sites. Temperature of the coldest month was lower on average in the mesic than in the dry area, suggesting a higher risk of freezing-induced embolism in the mesic area. As a response, plants may reduce vessel lumen area and increase the density of narrower vessels to prevent freeze–thaw embolism (Castro-Díez et al., 1998; Medeiros and Pockmann, 2014), similarly as for facing moisture deficit. However, according to the precipitation gradient, the dry area had a higher risk of drought-induced embolism, and therefore smaller vessels might help here to prevent embolism. The combination of both milder temperatures and higher moisture deficit may therefore have favoured adjustments at the xylem level similar to a combination of lower temperature with lower climatic moisture deficit. Interestingly, vessel grouping showed contrasting patterns in both species. In the evergreen E. coccineum, vessels tended to be more grouped in warmer and moister areas. This is in line with a positive effect of winter temperature on vessel grouping observed in the deciduous shrub Betula nana on Greenland (Nielsen et al., 2017). According to the air-seeding hypothesis (Alder et al., 1997; Wheeler et al., 2005; Loepfe et al., 2007), the larger the contact surface among vessels, the higher the risk of embolism spreading from one vessel to the next by the aspiration of air through the pit pores (Sperry and Tyree, 1988; Brodersen, 2013; Brodersen et al., 2013). The lower proportion of grouped vessels in dry and cold areas might thus be interpreted as an adjustment to lower the risk of embolism of connected vessels under frost and water limitation. In contrast, in the deciduous N. antarctica, the opposite trend was observed: vessels were more grouped in cold and dry areas. A high degree of vessel grouping might provide alternative pathways for water transport when a vessel is blocked by an embolism, allowing the water flow to continue through one or more adjacent vessels and thus improving the hydraulic redundancy (Baas et al., 1983; Tyree et al., 1994). However, our contrasting results in two coexisting widespread species support the controversy about the functional role of vessel grouping under water limitation (von Arx et al., 2013), pointing to a high degree of species specificity in this trait irrespective of the study system and a relative independence of vessel grouping from the other xylem traits. Co-ordination among xylem traits The analysis of the co-ordination between xylem traits at the plant level provided clues to interpret xylem adjustments across the climatic gradient. A remarkable result was that the percentage of conductive area (CA) was the main determinant of Ks in both species. CA is defined as VD times MVA (Carlquist, 2001). While the influence of VD and MVA on CA is thus linear, similar CA values might have led to very different Ks values depending on vessel lumen area distribution. The reason for this is that Ks scales to the fourth power of diameter (Tyree and Ewers, 1991), whereas Ks scales proportionately with vessel density, i.e. many small vessels within a given xylem area would form a less efficient xylem than fewer larger vessels. Yet, our results indicate that changes in vessel density and vessel lumen area distribution were co-ordinated to keep Ks constant. However, co-ordination among vessel traits at the plant level has been poorly explored so far, and thus more research is needed to evaluate the role played by trait integration on the hydraulic function (Gleason et al., 2016b). An additional result was that xylem traits in the evergreen E. coccineum showed a stronger climate control than in the deciduous N. antarctica. This tighter climatic control was indicated by the direct effect of climatic moisture deficit on Dh, VD and WD, supporting the higher xylem trait dependence from climate in evergreen species (Sperry et al., 1994; Villar-Salvador et al., 1997; Castro-Díez et al., 1998; Maherali et al., 2004; Pérez-de-Lis et al., 2018). Similarly, inter-relations among xylem traits were stronger in E. coccineum than in N. antarctica, including a negative covariation between VD and Dh that was not detected in the deciduous species. Wood density and climate Wood density was the trait with the lowest variability. Among several functional traits, WD has continuously shown a very low intra-specific variation, commonly lower than 10 % (Fajardo and Piper, 2011; Siefert et al., 2015; Fajardo, 2016, 2018). Nevertheless, WD variability was strongly correlated with climate in the evergreen E. coccineum, with denser wood occurring under dry and cold conditions as expected according to its positive correlation with resistance to hydraulic embolism (e.g. Hacke et al., 2001; Detmann et al., 2013; Rosner, 2017). In general, WD showed a negative correlation with Dh and was orthogonal to the other xylem traits. However, SEMs showed that at the individual level, WD in E. coccineum was determined by climatic moisture deficit and Dh, while it was completely independent of climate and of the other xylem traits in N. antarctica. The climatic independence of WD in N. antarctica is similar to that observed in the congeneric N. pumilio across elevational gradients (Fajardo, 2018), and to trends observed across species (Preston et al., 2006). A possible explanation for the independence between WD and the other traits is that WD was measured in the whole branch section, whereas anatomical traits were measured only in the last growth ring. In addition, WD may be more linked to the proportion of fibres or to wood elements other than vessels (Zanne et al., 2010; Zieminska et al., 2013). Conclusions We demonstrate that vessel size-related reduction in hydraulic efficiency within a species can be compensated through increases in vessel density. This implies a departure from the safety–efficiency trade-off, since estimated xylem efficiency can in fact be maintained in combination with different levels of expected xylem safety. Otherwise, the existence of contrasting relationships of xylem traits with climate in both study species indicates that predictions of community responses to local or environmental changes should consider the potential existence of differences in trait sensitivity to climatic variations among coexisting species. The fact that coexisting species may achieve similar responses to environmental gradients through alternative internal adjustments might increase resilience of temperate forests against unpredictable changes in climatic conditions. ACKNOWLEDGEMENTS We thank Laura Sánchez-Jardón for help with the fieldwork, and Miguel García Hidalgo and Gonzalo Juste for their invaluable assistance with laboratory tasks. We greatly acknowledge the enthusiastic review by Sean M. Gleason, as well as the constructive comments from another anonymous referee. This work was supported by the Chilean Fondo Nacional de Desarrollo Científico y Tecnológico [project FONDECYT 1160329 to A.F.]; the Spanish Ministry of Science and Innovation [FPI grant BES-2010-036041 to A.I.G.C]; the Spanish Ministry of Economy and Competitiveness [projects SPRING-CGL2017-87309-P and CGL2012-34209, excellence networks CGL2014-53840-REDT and CGL2016-81706-REDT, Juan de la Cierva-Formación grant FJCI-2015-24770 to A.I.G.C.]; COST Action FP1106 [regular STSM to A.I.G.C.]; and the Swiss State Secretariat for Education, Research and Innovation SERI [SBFI C12.0100 to G.v.A.]. LITERATURE CITED Alder NN, Pockman WT, Sperry JS, Nuismer S. 1997. Use of centrifugal force in the study of xylem cavitation. Journal of Experimental Botany  48: 665– 674. Google Scholar CrossRef Search ADS   Allen CD, Macalady AK, Chenchouni Het al.   2010. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management  259: 660– 684. Google Scholar CrossRef Search ADS   von Arx G, Dietz H. 2005. Automated image analysis of annual rings in the roots of perennial forbs. International Journal of Plant Sciences  166: 723– 732. Google Scholar CrossRef Search ADS   von Arx G, Carrer M. 2014. ROXAS – a new tool to build centuries-long tracheid lumen chronologies in conifers. Dendrochronologia  32: 290– 293. Google Scholar CrossRef Search ADS   von Arx G, Archer SR, Hughes MK. 2012. Long-term functional plasticity in plant hydraulic architecture in response to supplemental moisture. Annals of Botany  109: 1091– 1100. Google Scholar CrossRef Search ADS   von Arx G, Kueffer C, Fonti P. 2013. Quantifying plasticity in vessel grouping – added value from the image analysis tool ROXAS. IAWA Journal  34: 433– 445. Google Scholar CrossRef Search ADS   von Arx G, Crivellaro A, Prendin AL, Cufar K, Carrer M. 2016. Quantitative wood anatomy – practical guidelines. Frontiers in Plant Sciences  7: 781. Baas P, Werker E, Fahn A. 1983. Some ecological trends in vessel characters. IAWA Journal  4: 141– 159. Google Scholar CrossRef Search ADS   ter Braak CJF. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology  67: 1167– 1179. Google Scholar CrossRef Search ADS   Brodersen CR. 2013. Visualizing wood anatomy in three dimensions with high-resolution X-ray micro-tomography (μCT) – a review. IAWA Journal  34: 408– 424. Google Scholar CrossRef Search ADS   Brodersen CR, McElrone AJ, Choat B, Lee EF, Shackel KA, Matthews MA. 2013. In vivo visualizations of drought-induced embolism spread in Vitis vinifera. Plant Physiology  161: 1820– 1829. Google Scholar CrossRef Search ADS   Burns KC. 2004. Patterns in specific leaf area and the structure of a temperate heath community. Diversity and Distribution  10: 105– 112. Google Scholar CrossRef Search ADS   Carlquist S. 2001. Comparative wood anatomy: systematic, ecological, and evolutionary aspects of dicotyledon wood , 2nd edn. Berlin: Springer. Carrer M, von Arx G, Castagnieri D, Petit G. 2015. Distilling environmental information from time series of conduit size: the standardization issue and its relation to allometric and hydraulic constraints. Tree Physiology  35: 27– 33. Google Scholar CrossRef Search ADS   Castro-Díez P, Villar-Salvador P, Pérez-Rontomé C, Maestro-Martínez M, Montserrat-Martí G. 1998. Leaf morphology, leaf chemical composition and stem xylem characteristics in two Pistacia (Anacardiaceae) species along a climatic gradient. Flora  193: 195– 202. Google Scholar CrossRef Search ADS   Chave J, Coomes D, Jansen S, Lewis SL, Swenson NG, Zanne AE. 2009. Towards a worldwide wood economics spectrum. Ecology Letters  12: 351– 366. Google Scholar CrossRef Search ADS   Daly C, Halbleib M, Smith JIet al.   2008. Physiographically-sensitive mapping of temperature and precipitation across the conterminous United States. International Journal of Climatology  28: 2031– 2064. Google Scholar CrossRef Search ADS   De Frenne P, Graae BJ, Rodríguez-Sánchez Fet al.   2013. Latitudinal gradients as natural laboratories to infer species’ responses to temperature. Journal of Ecology  101: 784– 795. Google Scholar CrossRef Search ADS   Dettmann S, Pérez CA, Thomas FM. 2013. Xylem anatomy and calculated hydraulic conductance of four Nothofagus species with contrasting distribution in South-Central Chile. Trees – Structure and Function  27: 685– 696. Google Scholar CrossRef Search ADS   Fajardo A. 2016. Wood density is a poor predictor of competitive ability among individuals of the same species. Forest Ecology and Management  372: 217– 225. Google Scholar CrossRef Search ADS   Fajardo A. 2018. Insights into intraspecific wood density variation and its relationship with growth, height and elevation in a treeline species. Plant Biology  20: 456– 464. Google Scholar CrossRef Search ADS   Fajardo A, Piper FI. 2001. Intraspecific trait variation and covariation in a widespread tree species (Nothofagus pumilio) in southern Chile. New Phytologist  189: 259– 271. Google Scholar CrossRef Search ADS   Fajardo A, Piper FI. 2015. High foliar nutrient concentrations and resorption efficiency in Embothrium coccineum (Proteaceae) in southern Chile. American Journal of Botany  102: 208– 216. Google Scholar CrossRef Search ADS   Fisher JB, Goldstein G, Jones TJ, Cordell S. 2007. Wood vessel diameter is related to elevation and genotype in the Hawaiian tree Metrosideros polymorpha (Myrtaceae). American Journal of Botany  94: 709– 715. Google Scholar CrossRef Search ADS   Gärtner H, Lucchinetti S, Schweingruber FH. 2015. A new sledge microtome to combine wood anatomy and tree-ring ecology. IAWA Journal  36: 452– 459. Google Scholar CrossRef Search ADS   Gleason SM, Westoby M, Jansen Set al.   2016a. Weak tradeoff between xylem safety and xylem-specific hydraulic efficiency across the world’s woody plant species. New Phytologist  209: 123– 136. Google Scholar CrossRef Search ADS   Gleason SM, Westoby M, Jansen Set al.   2016b. On research priorities to advance understanding of the safety–efficiency tradeoff in xylem. New Phytologist  211: 1156– 1158. Google Scholar CrossRef Search ADS   Gorsuch DM, Oberbauer SF, Fisher JB. 2001. Comparative vessel anatomy of arctic deciduous and evergreen dicots. American Journal of Botany  88: 1643– 1649. Google Scholar CrossRef Search ADS   Greenwood S, Ruiz-Benito P, Martínez-Vilalta Jet al.   2017. Tree mortality across biomes is promoted by drought intensity, lower wood density and higher specific leaf area. Ecology Letters  20: 539– 553. Google Scholar CrossRef Search ADS   Hacke UG, Sperry JS. 2001. Functional and ecological xylem anatomy. Perspectives in Plant Ecology, Evolution and Systematics  4: 97– 115. Google Scholar CrossRef Search ADS   Hacke UG, Sperry JS, Pockman WT, Davis SD, McCulloh KA. 2001. Trends in wood density and structure are linked to prevention of xylem implosion by negative pressure. Oecologia  126: 457– 461. Google Scholar CrossRef Search ADS   Hacke UG, Spicer R, Schreiber SG, Plavcová L. 2017. An ecophysiological and developmental perspective on variation in vessel diameter. Plant, Cell and Environment  40: 831– 845. Google Scholar CrossRef Search ADS   Hamann A, Wang T, Spittlehouse DL, Murdock TQ. 2013. A comprehensive, high-resolution database of historical and projected climate surfaces for western North America. Bulletin of the American Meteorological Society  94: 1307– 1309. Google Scholar CrossRef Search ADS   Kolb K, Sperry JS. 1999. Differences in drought adaptation between subspecies of sagebrush (Artemisia tridentata). Ecology  80: 2373– 2384. Google Scholar CrossRef Search ADS   Lachenbruch B, McCulloh KA. 2014. Traits, properties, and performance: how woody plants combine hydraulic and mechanical functions in a cell, tissue, or whole plant. New Phytologist  204: 747– 764. Google Scholar CrossRef Search ADS   Loepfe L, Martínez-Vilalta J, Pinol J, Mencuccini M. 2007. The relevance of xylem network structure for plant hydraulic efficiency and safety. Journal of Theoretical Biology  247: 788– 803. Google Scholar CrossRef Search ADS   Luebert F, Pliscoff P. 2006. Sinopsis bioclimática y vegetacional de Chile . Santiago: Editorial Universitaria. Maherali H, Pockman WT, Jackson RB. 2004. Adaptive variation in the vulnerability of woody plants to xylem cavitation. Ecology  85: 2184– 2199. Google Scholar CrossRef Search ADS   Martínez-Vilalta J, García-Forner N. 2016. Water potential regulation, stomatal behavior and hydraulic transport under drought: deconstructing the iso/anisohydric concept. Plant, Cell and Environment  40: 962– 976. Google Scholar CrossRef Search ADS   Martínez-Vilalta J, Cochard H, Mencuccini Met al.   2009. Hydraulic adjustments of Scots pine across Europe. New Phytologist  184: 353– 364. Google Scholar CrossRef Search ADS   Medeiros JS, Pockman WT. 2014. Freezing regime and trade-offs with water transport efficiency generate variation in xylem structure across diploid populations of Larrea sp. (Zygophyllaceae). American Journal of Botany  101: 598– 607. Google Scholar CrossRef Search ADS   Nielsen SS, von Arx G, Damgaard CFet al.   2017. Xylem anatomical trait variability provides insight on the climate–growth relationship of Betula nana in Western Greenland. Arctic, Antarctic, and Alpine Research  49: 359– 371. Google Scholar CrossRef Search ADS   Nonweiler TRF. 1975. Flow of biological fluids through non-ideal capillaries. In: Zimmermann MH, Milburn JA, eds. Encyclopaedia of plant physiology, New Series, Vol 1. Transport in plants. I. Phloem transport . Berlin: Springer, 474– 477. Oksanen J, Blanchet FG, Friendly Met al.   2016. vegan: community ecology package. R package version 2.4-1 . https://CRAN.R-project.org/package=vegan Olano JM, González-Muñoz N, Arzac Aet al.   2017. Sex determines xylem anatomy in a dioecious conifer: hydraulic consequences in a drier world. Tree Physiology  37: 1493– 1502. Google Scholar CrossRef Search ADS   Pérez-de-Lis G, Rozas V, Vázquez-Ruiz RA, García-González I. 2018. Do ring-porous oaks prioritize earlywood vessel efficiency over safety? Environmental effects on vessel diameter and tyloses formation. Agricultural and Forest Meteorology  248: 205– 214. Google Scholar CrossRef Search ADS   Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core Team. 2016. nlme: linear and nonlinear mixed effects models. R package version 3.1-128 , URL: http://CRAN.R-project.org/package=nlme>. Preston KA, Cornwell WD, DeNoyer JL. 2006. Wood density and vessel traits as distinct correlates of ecological strategy in 51 California coast range angiosperms. New Phytologist  170: 807– 818. Google Scholar CrossRef Search ADS   R Core Team. 2016. R: a language and environment for statistical computing . Vienna: R Foundation for Statistical Computing. URL https://www.R-project.org/. Richards CL, Pennings SC, Donovan LA. 2005. Habitat range and phenotypic variation in salt marsh plants. Plant Ecology  176: 263– 273. Google Scholar CrossRef Search ADS   Rosner S. 2017. Wood density as a proxy for vulnerability to cavitation: size matters. Journal of Plant Hydraulics  4: e-001. doi: 10.20870/jph.2017.e001. Google Scholar CrossRef Search ADS   Sánchez-Salguero R, Navarro-Cerrillo R, Camarero JJ, Fernández-Cancio Á. 2012. Selective drought-induced decline of pine species in southeastern Spain. Climate Change  113: 767– 785. Google Scholar CrossRef Search ADS   Schreiber SG, Hacke UG, Hamann A, Thomas BR. 2011. Genetic variation of hydraulic and wood anatomical traits in hybrid poplar and trembling aspen. New Phytologist  190: 150– 160. Google Scholar CrossRef Search ADS   Schreiber SG, Hacke UG, Hamann A. 2015. Variation of xylem vessel diameters across a climate gradient: insight from a reciprocal transplant experiment with a widespread boreal tree. Functional Ecology  29: 1392– 1401. Google Scholar CrossRef Search ADS   Siefert A, Violle C, Chalmandrier Let al.   2015. A global meta-analysis of the relative extent of intraspecific trait variation in plant communities. Ecology Letters  18: 1406– 1419. Google Scholar CrossRef Search ADS   Sperry JS. 2003. Evolution of water transport and xylem structure. International Journal of Plant Sciences  164: S115– S127. Google Scholar CrossRef Search ADS   Sperry JS, Tyree MT. 1988. Mechanism of water stress-induced xylem embolism. Plant Physiology  88: 581– 587. Google Scholar CrossRef Search ADS   Sperry JS, Nichols KL, Sullivan JEM, Eastlack SE. 1994. Xylem embolism in ring-porous, diffuse-porous and coniferous trees of Northern Utah and interior Alaska. Ecology  75: 1736– 1752. Google Scholar CrossRef Search ADS   Sperry JS, Meinzer FC, McCulloh KA. 2008. Safety and efficiency conflicts in hydraulic architecture: scaling from tissues to trees. Plant, Cell and Environment  31: 632– 645. Google Scholar CrossRef Search ADS   Tyree MT, Ewers FW. 1991. The hydraulic architecture of trees and other woody plants. New Phytologist  119: 345– 360. Google Scholar CrossRef Search ADS   Tyree MT, Sperry JS. 1989. Vulnerability of xylem to cavitation and embolism. Annual Review of Plant Physiology and Plant Molecular Biology  40: 19– 38. Google Scholar CrossRef Search ADS   Tyree MT, Zimmermann MH. 2002. Xylem structure and the ascent of sap . Berlin: Springer-Verlag. Google Scholar CrossRef Search ADS   Tyree MT, Davis SD, Cochard H. 1994. Biophysical perspectives of xylem evolution: is there a tradeoff of hydraulic efficiency for vulnerability to dysfunction? IAWA Journal  15: 335– 360. Google Scholar CrossRef Search ADS   Villar-Salvador P, Castro-Díez P, Pérez-Rontomé C, Montserrat-Martí G. 1997. Stem xylem features in three Quercus (Fagaceae) species along a climatic gradient in NE Spain. Trees – Structure and Function  12: 90– 96. Violle C, Jiang L. 2009. Towards a trait-based quantification of species niche. Journal of Plant Ecology  2: 87– 93. Google Scholar CrossRef Search ADS   Wheeler JK, Sperry JS, Hacke UG, Hoang N. 2005. Inter-vessel pitting and cavitation in woody Rosaceae and other vesselled plants: a basis for a safety versus efficiency trade-off in xylem transport. Plant, Cell and Environment  28: 800– 812. Google Scholar CrossRef Search ADS   Williamson GB, Wiemann MC. 2010. Measuring wood specific gravity… Correctly. American Journal of Botany  97: 519– 524. Google Scholar CrossRef Search ADS   Zanne AE, Westoby M, Falster DSet al.   2010. Angiosperm wood structure: global patterns in vessel anatomy and their relation to wood density and potential conductivity. American Journal of Botany  97: 207– 215. Google Scholar CrossRef Search ADS   Zeballos SR, Giorgis MA, Cabido M, Gurvich DE. 2017. Unravelling the coordination between leaf and stem economics spectra through local and global scale approaches. Austral Ecology  42: 394– 403. Google Scholar CrossRef Search ADS   Ziemińska K, Butler DW, Gleason SM, Wright IJ, Westoby M. 2013. Fibre walls and lumen fractions drive wood density variation across 24 Australian angiosperms. AoB Plants  5: plt046. doi: 10.1093/aobpla/plt046. Google Scholar CrossRef Search ADS   Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. 2009. Mixed effects models and extensions in ecology with R . New York: Springer. Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Botany Oxford University Press

Xylem adjusts to maintain efficiency across a steep precipitation gradient in two coexisting generalist species

Loading next page...
 
/lp/ou_press/xylem-adjusts-to-maintain-efficiency-across-a-steep-precipitation-azNiX3Z96F
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0305-7364
eISSN
1095-8290
D.O.I.
10.1093/aob/mcy088
Publisher site
See Article on Publisher Site

Abstract

Abstract Background and Aims Trees adjust the configuration of their conductive system in response to changes in water availability, maximizing efficiency in wet environments and increasing safety in dry habitats. However, evidence of this general trend is not conclusive. Generalist species growing across broad climatic gradients provide an ideal framework to assess intra-specific xylem adjustments under contrasting environmental conditions. Our aims were to compare the response of xylem traits to variations in precipitation of two co-occurring generalist tree species, and to assess climate control on xylem trait variability and co-ordination. Methods We evaluated xylem traits of Embothrium coccineum (Proteaceae, evergreen) and Nothofagus antarctica (Nothofagaceae, deciduous) in three areas across an abrupt precipitation gradient, from 500 to 2500 mm, in southern Chile. We measured wood density, vessel lumen area and density, percentage of conductive area and vessel grouping, and estimated the hydraulic function from anatomical measurements in 60 individuals per species. Key Results Both species shared a common pattern of response along the precipitation gradient, with an increase in vessel density with dryness, but without changes in estimated hydraulic conductivity. Xylem traits in E. coccineum were more variable and more responsive to the climate gradient, decreasing vessel lumen area and increasing wood density, whereas vessel grouping showed contrasting patterns between species. Additionally, the analysis of trait co-ordination at the individual level revealed a tighter co-ordination among xylem traits in E. coccineum. Conclusions Estimated xylem efficiency was maintained in combination with different levels of expected xylem safety within species. Reduction in vessel lumen area was compensated through large increases in vessel density, thus breaking the trade-off between xylem efficiency and safety. Otherwise, the existence of alternative internal adjustments in coexisting species to face similar climatic constraints might increase resilience of temperate forests against unpredictable changes in climatic conditions. Embothrium coccineum, Nothofagus antarctica, Patagonia, precipitation gradient, temperate rainforest, vessel grouping, vessel traits, wood density, xylem trait co-ordination INTRODUCTION Increasing temperature and a higher frequency and intensity of drought events are triggering worldwide tree dieback (e.g. Allen et al., 2010; Sánchez-Salguero et al., 2012; Greenwood et al., 2017), clearly evidencing that climate change is already impacting forest ecosystems. Understanding the response of forests to this novel scenario requires a better comprehension of plants’ adjustments to climate variability. Natural gradients provide an ideal framework to assess this question (De Frenne et al., 2013), since they allow the evaluation of species’ adjustments across contrasting environmental conditions. The adaptive potential of widespread species has been attributed to intra-specific variation in functional traits (Burns, 2004; Richards et al., 2005; Violle and Jiang, 2009); thus, linking patterns of trait variation with species’ ability to inhabit environmental gradients may contribute to prediction of community responses to local or global environmental changes and hence provide valuable insights into the mechanisms that control local functional diversity. Water availability is considered a key factor underlying trait variability among populations and communities (Maherali et al., 2004). The relationship between plants and water flow is mediated by the configuration of the hydraulic system (Tyree and Zimmermann, 2002). Hydraulic efficiency of the stem xylem conductive system is determined by hydraulic conductivity, i.e. the volume of sap that can be moved through the xylem normalized by time, cross-sectional area, length and pressure gradient. Hydraulic safety refers to the ability to maintain xylem conductivity at decreasing water potential, i.e. without occurrence of embolism (Hacke and Sperry, 2001; Tyree and Zimmermann, 2002). Efficiency and safety have been considered to represent a trade-off in hydraulic system functionality (Sperry, 2003). As a result, xylem traits show conspicuous anatomical responses to environmental conditions. The vessel lumen area of angiosperms diminishes in plants inhabiting dry or cold areas, and increases in moist and warm environments (Sperry et al., 2008; Hacke et al., 2017), because wide vessels prioritize water conductivity at the expense of higher embolism risk by either drought or frost (Tyree and Sperry, 1989). However, evidence of this safety–efficiency trade-off is not conclusive, and some species may show simultaneously reduced xylem safety and low hydraulic efficiency (Maherali et al., 2004; Gleason et al., 2016a). Several potential factors may modulate the relationship between the conductive system of plants and the environment. First, wood traits are co-ordinated with leaf-level adjustments (Castro-Díez et al., 1998; Martínez-Vilalta et al., 2009), especially under harsh climatic conditions (Zeballos et al., 2017). For instance, leaves may close their stomata to avoid water losses by evapotranspiration when water is scarce, thus reducing embolism risk (Martínez-Vilalta and García-Forner, 2016). Secondly, leaf habit may alter the relationship between xylem and climate (Castro-Díez et al., 1998; Pérez-de-Lis et al., 2018). Deciduous trees shed leaves during the unfavourable season and thus reduce plant hydraulic demands, enabling them to tolerate large hydraulic conductance losses (Sperry et al., 1994). As a result, deciduous species may have larger vessels compared with evergreen congenerics, since vessel lumen area is less constrained by unfavourable conditions (Castro-Díez et al., 1998; Gorsuch et al., 2001). The relationship of xylem traits to climate may also shift between sexes in dioecious species (Olano et al., 2017), and in other cases xylem traits may even be unresponsive to environmental gradients (Martínez-Vilalta et al., 2009). All these factors make xylem adjustments across environmental gradients not easily predictable for coexisting species differing in leaf traits, leaf habit or reproductive conditions. In this study, we focused on two widespread, small-statured tree species with contrasting foliar habit that grow in southern Chile, the evergreen Embothrium coccineum and the deciduous Nothofagus antarctica. We evaluated xylem adjustments of both species across an abrupt gradient of annual precipitation from inland (500 mm) to coastal (2500 mm) forests in Patagonia, Chile, also considering variations in temperature. We measured wood density, vessel lumen area and density, percentage of conductive area and vessel grouping, and estimated the hydraulic function from anatomical measurements (hydraulically weighted mean diameter and xylem-specific hydraulic conductivity) (1) to assess whether xylem traits show similar responses to the environmental gradient in both species; and (2) to estimate to what extent xylem trait variability was controlled by climate. We also evaluated co-ordination among xylem traits in both species, since inter-relations among xylem traits commonly observed in inter-specific comparisons, such as the positive correlations between conduit dimensions and hydraulic conductivity, and between wood density and vulnerability to embolism (Hacke et al., 2001; Chave et al., 2009; Lachenbruch and McCulloh, 2014; Rosner, 2017), might be weak at the intra-specific level (Martínez-Vilalta et al., 2009; Schreiber et al., 2011). We hypothesized that patterns of xylem trait variation along the precipitation gradient would be similar in both species according to the safety–efficiency trade-off. Despite the phylogenetic distance between both species, their xylem anatomies were expected to differ. We also hypothesized that xylem trait variation in the deciduous N. antarctica would show weaker correlation with climate than in the evergreen E. coccineum. MATERIALS AND METHODS Species and research site descriptions Embothrium coccineum J.R. & G. Forster (Proteaceae) and Nothofagus antarctica (Foster) Oerst. (Nothofagaceae) are two small-statured tree species (5–8 m mean height) that are evergreen and deciduous, respectively. Both species have a broad latitudinal distribution throughout Chile and Argentina, from central Chile (34°S) to Tierra del Fuego (55°S), ranging from superhumid to dry sites, and from sea level to, in many cases, the treeline limit (Fajardo and Piper, 2015). In the Aysén region (Patagonia, Chile), these two species prevail across an abrupt precipitation gradient (2500 to 600 mm year–1, 9.2–11 °C temperature difference between mean warmest and mean coldest months; Fig. 1; Table 1). In this study, we worked in three areas across this precipitation gradient that differ considerably in climatic conditions and hence in vegetation. Fig. 1. View largeDownload slide Location of the study areas (wet, mesic and dry), sites (four per area) and climatic diagrams for each study area. Ak, Aiken Park; At, Atravesado Lake; Ci, Baguales; Cl, Claro River; Cu, Cuervo River; Ib, Puerto Ibáñez; Lv, Levicán Peninsula; Pa, Pangal Valley; Pf, head of Pangal Valley; Rc, Reserva Coyhaique; Ro, Rosado Mountain; Sa, Ibáñez River waterfall. Note that central months in climatic diagrams correspond to summer in the southern hemisphere. Climatic data were obtained from the Climate Forecast System Reanalysis (CFSR) data set, compiled by the National Center for Environmental Prediction (NCEP), USA (Globalweather 2016). The scale bar in the region map represents 50 km. Fig. 1. View largeDownload slide Location of the study areas (wet, mesic and dry), sites (four per area) and climatic diagrams for each study area. Ak, Aiken Park; At, Atravesado Lake; Ci, Baguales; Cl, Claro River; Cu, Cuervo River; Ib, Puerto Ibáñez; Lv, Levicán Peninsula; Pa, Pangal Valley; Pf, head of Pangal Valley; Rc, Reserva Coyhaique; Ro, Rosado Mountain; Sa, Ibáñez River waterfall. Note that central months in climatic diagrams correspond to summer in the southern hemisphere. Climatic data were obtained from the Climate Forecast System Reanalysis (CFSR) data set, compiled by the National Center for Environmental Prediction (NCEP), USA (Globalweather 2016). The scale bar in the region map represents 50 km. Table 1. Climatic characteristics of the study sites (based on the data set obtained from ClimateSA v1.0 for the period 1981–2010; see text) Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  MAT, mean annual temperature; MAP, mean annual precipitation; ACMT, average coldest month temperature; NFFD, number of frost-free days; CMD, Hargreaves climatic moisture deficit. View Large Table 1. Climatic characteristics of the study sites (based on the data set obtained from ClimateSA v1.0 for the period 1981–2010; see text) Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  Site  Area  Latitude (°)  Longitude (°)  Elevation (m)  MAT (°C)  MAP (mm)  ACMT (°C)  NFFD  CMD (mm)  Pangal Valley  Wet  45.304  72.626  154  8.7  1837  4.0  306  138  Head of Pangal Valley  Wet  45.212  72.620  215  8.4  1839  3.8  301  141  Aikén Park  Wet  45.464  72.755  230  8.1  2135  3.5  297  53  Cuervo River  Wet  45.337  73.030  50  9.1  2235  4.9  321  28  Coyhaique Reserve  Mesic  45.533  72.010  784  5.4  938  0.0  209  502  CIEP  Mesic  45.532  72.071  335  7.6  1130  2.2  271  477  Rosado Mountain  Mesic  45.448  72.073  756  5.6  1056  0.2  215  460  Atravesado Lake  Mesic  45.673  72.194  294  7.7  1234  2.4  275  430  Puerto Ibáñez  Dry  46.268  71.932  423  7.0  551  1.2  248  731  Ibáñez River waterfall  Dry  46.262  71.995  221  8.0  604  2.3  274  715  Claro River  Dry  46.266  72.005  247  7.9  608  2.3  272  699  Levicán Peninsula  Dry  46.309  71.973  387  7.2  557  1.4  253  727  MAT, mean annual temperature; MAP, mean annual precipitation; ACMT, average coldest month temperature; NFFD, number of frost-free days; CMD, Hargreaves climatic moisture deficit. View Large The wettest area was located around the city of Puerto Aysén (45°24′27″S, 72°41′39″W, approx. 30 m a.s.l.), at the westernmost part of the gradient. Here, annual precipitation is 2034 mm year–1 (Puerto Aysén weather station, 32 m a.s.l.; Dirección General de Aguas, 2007–2017). Precipitation is evenly distributed throughout the year. Mean summer temperature (December–February) is 12.7 °C, whereas mean winter temperature (June–August) is 4.4 °C (data for the period 1981–2010 generated with the ClimateSA v1.0 software package, available at http://tinyurl.com/ClimateSA;Hamann et al., 2013). In this area, both target species occur within a species-rich temperate rainforest that prevails under a hyper-humid climatic regime (Luebert and Pliscoff, 2006). Dominant species include Drimys winteri (Winteraceae), Nothofagus betuloides and N. nitida (Nothofagaceae), Amomyrtus luma (Myrtaceae), Laureliopsis philippiana (Atherospermataceae), Lomatia ferruginea (Proteaceae), Caldcluvia paniculata and Weinmannia trichosperma (Cunnoniaceae), Podocarpus nubigenus (Podocarpaceae) and Pilgerodendron uviferum (Cupressaceae), among others. The mesic area was located around the city of Coyhaique (45°34′47″S, 72°03′40″W, approx. 280 m a.s.l.). Mean annual precipitation is 910 mm year–1 (Coyhaique National Reserve weather station, 400 m a.s.l.; Dirección General de Aguas, 2002–2015), with a slight drop in precipitation in February. This is the coldest area, since mean summer temperature (December–February) is 11.2 °C, whereas mean winter temperature (June–August) is 1.7 °C. The forest here has lower species richness, and is dominated by Nothofagus pumilio, N. antarctica, N. dombeyi, E. coccineum, Ribes magellanicum (Saxifragaceae) and Discaria chacaye (Rhamnaceae). The dry area was located around the village of Puerto Ibáñez (46°17′35″S, 71°56′19″W, approx. 320 m a.s.l.). Mean annual precipitation is 580 mm, with wet winters (around 80–90 mm month–1) and relatively dry summers (around 40–50 mm month–1). Mean summer temperature (December–February) is 12.4 °C, whereas mean winter temperature (June–August) is 2.3 °C. Vegetation here is a sclerophyll, short-stature forest dominated by D. chacaye, Schinus marchandii (Anacardiaceae), Maytenus boaria (Celastraceae), Colletia spinosissima (Rhamnaceae), E. coccineum and N. antarctica. Sampling, tissue collection and processing Sampling was conducted in November–December 2013. Four sites separated by a minimum distance of 5 km were selected in each of the three contrasting precipitation areas to enhance replication and include variability within areas (Fig. 1; Table 1). At each site, five individuals per species were sampled. We restricted the sampling to adult, unshaded trees without browsing or other apparent damage. For each individual tree, we cut at approx. 2 m height one 2 m long terminal, sun-exposed branch with fully expanded leaves using a telescoping pole (ARS Corporation, Sakai, Japan). This standardization of branch length minimizes xylem anatomical and hydraulic differences related to distance to the branch tip (Carrer et al., 2015). For each branch, we selected and cut two 2 cm wide pieces of wood of similar diameter (1.5–2 cm). Both wood pieces were labelled and placed in a cooler for transportation. In the laboratory, the fresh volume for one of the two pieces of branch per individual was determined by submerging the wood (without bark) into a glass beaker on a scale. The mass difference caused by the sample, which equals the volume of water displaced, was recorded and converted to volume based on the density of water as 1.0 g cm–3 at standard temperature and pressure. Samples were then dried in a forced-air oven (Memmert GmbH, Schwabach, Germany) at 70 °C for 72 h and the dry mass was subsequently measured. Wood density (WD) was then calculated as the oven-dry mass per green volume (Williamson and Wiemann, 2010). Quantitative wood anatomy The second piece of branch was used to measure xylem anatomy. Xylem anatomical analyses followed the protocol proposed by von Arx et al. (2016). Anatomical cross-sections of 10 μm thickness from the remaining pieces of wood were produced with a sledge microtome (Gärtner et al., 2015). These cross-sections were then placed on a slide and stained with Alcian blue (1 % solution in acetic acid) and safranin (1 % solution in ethanol). Afterwards, the cross-sections were dehydrated using a series of ethanol solutions of increasing concentration, washed with xylol and permanently preserved by embedding them in Eukitt glue. Overlapping images covering a complete radius from pith to bark were captured with a Nikon D90 digital camera mounted on a Nikon Eclipse 50i optical microscope with ×40 magnification (corresponding to a resolution of 0.6134 pixels μm–1) and merged to a single image using PTGUI v8.3.10 Pro (New House Internet Services B.V., Rotterdam, The Netherlands). In each image, we measured seven anatomical traits associated with hydraulic function in the last complete annual ring (i.e. the ring formed in the growing season 2012–2013). Five of these traits were directly measured from images, and the others were derived from these measurements. Directly measured traits were: (1) vessel density (VD; no. mm–2), which is the number of vessels per mm2; (2) mean vessel lumen area (MVA; μm2), which is the average lumen area of vessels; (3) percentage of conductive area per ring (CA), obtained as the cumulative lumen area of all counted vessels divided by the ring area; (4) vessel grouping index per annual ring (VGI), which is the mean number of vessels with contiguous cell walls (von Arx et al., 2013); and (5) vessel solitary fraction per annual ring (VSF), i.e. the percentage of solitary vessels with respect to all vessels in the ring. From these data, we calculated two derived traits: (6) hydraulically weighted mean diameter (Dh; μm), based on the vessels’ contribution to hydraulic conductance according to the Hagen–Poiseuille law, obtained as Dh = Σd5/Σd4, where d is the lumen diameter of each vessel (see Kolb and Sperry, 1999); and (7) an anatomy-based estimate of specific hydraulic conductivity (Ks; m2 s–1 MPa–1), which is the hydraulic conductivity per unit area, where the hydraulic conductance of each cell is calculated following Nonweiler (1975) and considering the ovality of cells (see equations in Olano et al., 2017). Vessel anatomical features were measured using ROXAS v3.0 (von Arx and Dietz, 2005; von Arx and Carrer, 2014), a specific image-analysis tool based on Image-Pro Plus (Media Cybernetics, Silver Spring, MD, USA). We first adjusted ROXAS settings to create different configurations of parameters used for automatic identification of vessel lumina. For each species, we performed a preliminary visual exploration to adjust the maximum and minimum vessel lumen area, vessel ovality and several colour parameters. With this information, we created specific configurations for the automatic analysis. In some cases, several configurations differing in parameters related to colour and vessel shape were needed to account for the variability in image quality. We used these configurations to analyse all samples automatically. The automatic output was then manually edited by drawing the last ring boundaries, deleting erroneously detected vessels (e.g. in parenchyma rays) and rectifying additional misidentifications. Then, exclusion areas were used to leave out parts of the image with lower quality (e.g. broken vessels introduced during sample preparation). In the last step of the manual editing, the 100 largest vessels per image were checked using the ‘Outlier search’ tool in ROXAS to find vessels that remained erroneously merged during the automatic analysis due to the presence of low contrast cell walls in some parts of the image. In high-quality images, this outlier search was only performed until no correction was needed after checking 20 consecutive vessels. Climatic data We used the ClimateSA v1.0 software package to generate data for the period 1981–2010 for each of the 12 sampling sites. This software is available at http://tinyurl.com/ClimateSA (Hamann et al., 2013). Climatic data have been developed with the parameter-elevation regressions on an independent slopes model (PRISM), an expert interpolation approach described by Daly et al. (2008) which uses physiographic information to better predict climate patterns in mountainous terrain. Although sites within the same area were closer to each other than to sites within the other areas, they were far enough from one another to show differences in climate. We obtained data of mean annual temperature, mean coldest month temperature, number of frost-free days, mean annual precipitation and Hargreaves climatic moisture deficit (Table 1). Climatic moisture deficit was obtained as the sum of the monthly difference between a reference evaporation calculated with the Hargreaves equation with a latitude correction applied (see Hamann et al., 2013) and precipitation. Data analyses Vessel lumen area and their contribution to potential hydraulic conductivity. For each species and site, we calculated the frequency distribution of vessels according to their lumen area. We grouped vessels by lumen area into 200 μm2 and 100 μm2 breaks (for E. coccineum and N. antarctica, respectively), i.e. we counted the number of vessels whose lumen area fell under each of these breaks or size classes. We also obtained the relative contribution of each of these classes to total water flow based on theoretical conductivity of each vessel in the class, according to Poiseuille’s law. Frequency distributions of vessels in each class and theoretical hydraulic conductivity were compared at the species level between sites using two-sample Kolmogorov–Smirnov (K–S) tests (von Arx et al., 2012). The vessel lumen area class at which 50 % of total water flow is achieved was also obtained per species and site. Variation in xylem traits along the gradient. In a preliminary step, we evaluated redundancy among xylem traits (i.e. the seven anatomical variables and WD) by examining the correlation matrix among them. Vessel solitary fraction and MVA were excluded from subsequent statistical analyses due to their high correlations with VGI and Dh, respectively (r > 0.9 in both cases). We previously obtained coefficients of variation (CVs) of all xylem traits per species within sites (n = 5 individuals per site of each species), within precipitation areas (n = 20 individuals per area of each species) and at the species level (n = 60 individuals of each species). To evaluate intra-specific variations in xylem traits along the precipitation gradient, we fitted separate linear mixed models (LMMs) per species and trait. We considered the study area of precipitation as a fixed factor with three levels (wet, mesic and dry). Site was included as a random factor nested in area to incorporate the grouping structure of our data, since individuals within a site can show trait values more similar to each other than to individuals from other sites within a given area. Models were fitted using restricted maximum likelihood (REML), and normalized residuals were extracted. Residuals were checked for normality and homoscedasticity, and models were re-fitted allowing them to have different variances per area in cases where these assumptions were not met. We fitted revised models with maximum likelihood, compared them using the Akaike information criterion (AIC), selected models that minimized the AIC value and checked again for normality and homoscedasticity of residuals. Finally, the optimal models were re-fitted with REML (Zuur et al., 2009). We used the ‘nlme’ package (Pinheiro et al., 2016) in R version 3.4.1 environment (R Core Team, 2016) to fit LMM. We also compared global patterns of intra-specific variation in xylem traits along the precipitation gradient for the two species. To do so, we performed a principal component analysis (PCA) on an individual × xylem traits matrix to determine their variation structure. Then, we used canonical ordination techniques to evaluate the effect of species and area (and their interaction) on xylem traits. We performed a redundancy analysis (RDA; ter Braak, 1986), which combines multiple regressions with PCA, to relate the dependent matrix (xylem traits) to an explanatory matrix (species and area). We also performed partial RDA analyses to discriminate the independent effects of species and area on xylem traits. In all cases, the significance of explanatory variables was evaluated using a Monte Carlo test with 9999 permutations. These analyses were performed with the ‘vegan’ package (Oksanen et al., 2016). Effects of climate on xylem trait variation. We used Pearson correlations to relate mean coldest month temperature and climatic moisture deficit with xylem traits. In addition, we used structural equation models (SEMs) to evaluate direct and indirect effects of these climatic variables on VD, Dh, Ks and WD, as well as the inter-relationships among these xylem traits and to CA, following the suggestion by Gleason et al. (2016b). These models were based on the hypothesized xylem a djustments to differences in water availability and continentality (see the Introduction). We adjusted two models, one per species. Model parameters were estimated with maximum likelihood, and global model fit was assessed by using the goodness of fit index (GFI). The magnitude and significance of the direct and indirect effects were estimated from the standardized path coefficients by a multivariate Wald test. Structural equation models were performed with AMOS 18.0 software (AMOS Development Corp., Mount Pleasant, SC, USA). RESULTS Xylem anatomy and vessel lumen area distributions Nearly one hundred thousand (97 711) vessels were measured in the last complete growth ring of the 120 sampled trees. In N. antarctica, the arrangement of the vessels was diffuse-porous with a slight tendency to semi-ring porosity (Detmann et al., 2013), whereas the xylem of E. coccineum was semi-ring porous with vessels arranged in tangential bands (Fig. 2). Vessel lumen area showed large variability, ranging from 38 to 4389 μm2 in E. coccineum, and from 27 to 2058 μm2 in N. antarctica. Small vessels were by far the most abundant in both species (Fig. 3A, B). However, the relative proportion of large vessels was higher in wet sites than in mesic and dry sites (P < 0.001 for all paired K–S tests comparing frequency distributions of vessel lumen area). This shift towards larger vessels in wet sites had a strong impact on their relative contribution to potential hydraulic conductivity (Fig. 3C, D). The vessel lumen area class at which 50 % of total water flow occurred was larger at the wet site, especially in E. coccineum (P < 0.001 for all paired K–S tests comparing curves of estimated hydraulic conductivity). In other words, small changes in vessel lumen size distribution from wet to dry areas led to strong changes in vessel class contribution to water flow. Fig. 2. View largeDownload slide Cross-sections of branch segments of the study species. Measured vessel lumina are filled in light blue, ring borders in yellow. Scale bars represent 100 μm. Fig. 2. View largeDownload slide Cross-sections of branch segments of the study species. Measured vessel lumina are filled in light blue, ring borders in yellow. Scale bars represent 100 μm. Fig. 3. View largeDownload slide Frequency distributions of vessel number and potential vessel hydraulic conductivity per vessel lumen area classes. Vessel lumen areas were divided in 200 μm2 size and 100 μm2 size classes for E. coccineum and N. antarctica, respectively. (A and B) The proportion of vessels in each size class. (C and D) The contribution of estimated hydraulic conductivity of vessels in each size class to total conductivity per ring. Distributions are presented separately per species and area. Triangles represent the size class at which 50 % of conductivity is reached. Fig. 3. View largeDownload slide Frequency distributions of vessel number and potential vessel hydraulic conductivity per vessel lumen area classes. Vessel lumen areas were divided in 200 μm2 size and 100 μm2 size classes for E. coccineum and N. antarctica, respectively. (A and B) The proportion of vessels in each size class. (C and D) The contribution of estimated hydraulic conductivity of vessels in each size class to total conductivity per ring. Distributions are presented separately per species and area. Triangles represent the size class at which 50 % of conductivity is reached. Variation in xylem traits along the gradient Xylem traits were in general more variable in E. coccineum than in N. antarctica (Table 2). Wood density was the least variable trait (<10 % of variation in all cases), whereas VD reached >50 % of variation in E. coccineum. The highest variation was found at the species level (i.e. considering all sampled individuals), but variability within areas was very close to the global variation in all traits of both species. Variation within sites was clearly lower than within areas. Table 2. Coefficients of variation of xylem traits observed within sites (CVsite, n = 5), within areas (CVarea, n = 20) and at the species level (CVtotal, n = 60)   Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99    Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99  Average values of CVsite and CVarea are shown. WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Table 2. Coefficients of variation of xylem traits observed within sites (CVsite, n = 5), within areas (CVarea, n = 20) and at the species level (CVtotal, n = 60)   Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99    Embothrium coccineum  Nothofagus antarctica  Trait  CVtotal  CVarea  CVsite  CVtotal  CVarea  CVsite  WD  7.84  6.10  5.31  5.76  5.61  4.73  VD  52.92  50.78  41.19  28.47  24.82  20.13  CA  43.89  44.17  37.76  27.79  26.81  22.88  MVA  29.45  26.35  24.58  18.77  18.27  16.72  Dh  15.37  13.35  12.65  10.26  10.02  9.17  VGI  23.67  23.86  19.87  14.93  14.62  12.19  VSF  36.88  36.14  32.21  25.97  24.68  21.99  Ks  47.84  47.37  40.92  31.38  31.19  27.99  Average values of CVsite and CVarea are shown. WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Three xylem traits varied along the precipitation gradient in at least one species. Wood density increased significantly from wet to mesic and dry areas in E. coccineum (0.557 ± 0.006 g cm–3; mean ± s.e. at the species level), whereas it did not differ among areas in N. antarctica (0.562 ± 0.004 g cm–3, Fig. 4A). The hydraulically weighted mean diameter in E. coccineum (35.7 ± 0.7 μm) was significantly higher in the wet area than in the mesic and dry areas (Fig. 4B), but this trait remained constant in N. antarctica (28.0 ± 0.4 μm). Vessel density (187 ± 13 vessels mm–2 in E. coccineum; 434 ± 16 vessels mm–2 in N. antarctica) varied along the gradient in both species, increasing significantly from wet to dry sites (Fig. 4C). The percentage of the conductive area (10.68 ± 0.60 % in E. coccineum; 17.85 ± 0.64 % in N. antarctica) remained constant along the gradient (Fig. 4D), and so did Ks (2.33 × 10–7 ± 1.44 10–8 kg m–2 MPa–1 s–1 in E. coccineum; 3.39 × 10–7±137 10–8 kg m–2 MPa–1 s–1 in N. antarctica; Fig. 4E). Vessel grouping revealed a non- significant trend towards reduced VGI and increased VSF from wet to dry areas in E. coccineum (Fig. 4F). Nothofagus antarctica showed the inverse trend, but this was still not significant. Fig. 4. View largeDownload slide Mean values and confidence intervals at P = 0.95 of xylem traits per area for Nothofagus antarctica and Embothrium coccineum. Letters indicate significant differences obtained with linear mixed models considering area (wet, mesic and dry) as a fixed factor and site nested in area as a random factor. Fig. 4. View largeDownload slide Mean values and confidence intervals at P = 0.95 of xylem traits per area for Nothofagus antarctica and Embothrium coccineum. Letters indicate significant differences obtained with linear mixed models considering area (wet, mesic and dry) as a fixed factor and site nested in area as a random factor. Global variation patterns in xylem traits The first two axes of the PCA explained 81.05 % of the variance in xylem traits (Fig. 5). The first axis was related to CA, VD and Ks, and sorted individuals according to species identity. Individuals of N. antarctica showed in general higher VD, CA and Ks than individuals of E. coccineum. These three traits were orthogonal to WD and Dh, which were negatively correlated (r = –0.46; P < 0.001) and drove variation along the second axis. This axis was related to the precipitation gradient, with individuals showing higher WD and lower Dh at the dry area, and the opposite pattern at the wet area. The RDA confirmed the impact of species and moisture gradient on xylem configuration, as both factors explained 61.87 % of the variability in xylem traits (F = 36.994, P = 0.001). Partial RDAs determined the overwhelming effect of species (52.18 % of explained variance; F = 147.09, P < 0.001), but also the robustness of the effect of the climatic gradient (6.89 % of explained variance associated with area, F = 9.72, P < 0.001). Species per area interaction was not significant (P = 0.769), indicating a common response of both species in different areas. Fig. 5. View largeDownload slide Principal component analysis (PCA) biplot showing the position of individual branches along the two first principal component axes. Colours indicate areas along the gradient (black = wet, dark grey = mesic, light grey = dry). Large open symbols and error bars represent average position and standard errors per species and area. Abbreviations: VD, vessel density; Dh, hydraulically weighted mean diameter; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity; CA, percentage of conductive area; VGI, vessel grouping index; WD, wood density. Fig. 5. View largeDownload slide Principal component analysis (PCA) biplot showing the position of individual branches along the two first principal component axes. Colours indicate areas along the gradient (black = wet, dark grey = mesic, light grey = dry). Large open symbols and error bars represent average position and standard errors per species and area. Abbreviations: VD, vessel density; Dh, hydraulically weighted mean diameter; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity; CA, percentage of conductive area; VGI, vessel grouping index; WD, wood density. Climatic control on xylem trait variation Correlations and SEM showed a stronger climatic control on xylem traits in E. coccineum than in N. antarctica. The average coldest month temperature correlated negatively with WD and VSF, and positively with Dh and MVA in E. coccineum (Table 3). In contrast, temperature was only marginally correlated with VD in N. antarctica. The effect of climatic moisture deficit was more similar between species. Climatic moisture deficit correlated positively with VD and negatively with Dh and MVA in both species, although negative correlations were only marginally significant in N. antarctica. An additional positive correlation between climatic moisture deficit and WD was observed in E. coccineum. Table 3. Pearson correlation coefficients between climatic variables [mean coldest month temperature, climatic moisture deficit (CMD)] and xylem traits   E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201    E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201  Significant correlations are highlighted in bold (P < 0.05) and in bold italics (P < 0.1). WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Table 3. Pearson correlation coefficients between climatic variables [mean coldest month temperature, climatic moisture deficit (CMD)] and xylem traits   E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201    E. coccineum  N. antarctica  Trait  Coldest temperature  CMD  Coldest temperature  CMD  WD  –0.621  0.696  –0.318  0.393  VD  –0.225  0.642  –0.573  0.594  CA  0.173  0.230  –0.488  0.367  MVA  0.714  –0.824  0.261  –0.519  Dh  0.694  –0.795  0.237  –0.537  VGI  0.428  –0.249  –0.486  0.343  VSF  –0.610  0.417  0.480  –0.396  Ks  0.468  –0.183  –0.404  0.201  Significant correlations are highlighted in bold (P < 0.05) and in bold italics (P < 0.1). WD, wood density; VD, vessel density; CA, percentage of conductive area; MVA, mean vessel lumen area; Dh, hydraulically weighted mean diameter; VGI, vessel grouping index; VSF, vessel solitary fraction; Ks, anatomy-based estimate of xylem-specific hydraulic conductivity. View Large Structural equation models showed that neither temperature nor climatic moisture deficit affected xylem conductivity (Ks) directly, but rather via their influence on xylem traits and CA (especially in E. coccineum; Fig. 6). Covariation between temperature and climatic moisture deficit resulted in the effects of temperature on xylem traits being closely aligned with the effects of moisture deficit. In N. antarctica, the only xylem trait affected by climate was Dh, which decreased as climatic moisture deficit increased (Fig. 6A). Increases in Dh led to increases in CA, which was in turn the main determinant of Ks. However, CA was also strongly affected by VD, which was independent of climatic moisture deficit at the individual level in this species. In E. coccineum, both Dh and VD were controlled by climatic moisture deficit and co-varied negatively, indicating that in drier sites, trees had smaller vessels and higher vessel density (Fig. 6B). In addition, trees in drier sites had denser wood, as WD was directly controlled by climatic moisture deficit but also indirectly by Dh and by its covariation with VD. In the same way as for N. antarctica, the main determinant of Ks was CA, which in turn was strongly controlled by VD but also by Dh. As a result, both Dh and VD determined Ks in E. coccineum through their effect on CA, but also through direct effects. Interestingly, CA did not affect WD in any species. Both SEMs showed good fit to the data (GFI = 0.999 for E. coccineum, GFI = 0.992 for N. antarctica; values above 0.90 indicate a good fit). Fig. 6. View largeDownload slide Adjusted structural equation models (SEMs) for the effects of climate (mean coldest month temperature and climatic moisture deficit) on xylem traits, and inter-relations among xylem traits. Arrows indicate the direction of evaluated relationships (paths). Arrow width is proportional to path coefficients. Significant path coefficients (P < 0.05) are shown, while non-significant paths appear as light grey arrows. Ks, estimated xylem-specific hydraulic conductivity. Goodness of fit index (GFI) = 0.999 for E. coccineum and 0.992 for N. antarctica. Fig. 6. View largeDownload slide Adjusted structural equation models (SEMs) for the effects of climate (mean coldest month temperature and climatic moisture deficit) on xylem traits, and inter-relations among xylem traits. Arrows indicate the direction of evaluated relationships (paths). Arrow width is proportional to path coefficients. Significant path coefficients (P < 0.05) are shown, while non-significant paths appear as light grey arrows. Ks, estimated xylem-specific hydraulic conductivity. Goodness of fit index (GFI) = 0.999 for E. coccineum and 0.992 for N. antarctica. DISCUSSION Both species shared most anatomical adjustments to the precipitation gradient, despite their different xylem anatomies and phylogenetic distance, supporting our first hypothesis. Only vessel grouping showed contrasting patterns in both species. Otherwise, the strength of climatic control on xylem traits varied between species: climatic control on xylem traits was tighter in the evergreen E. coccineum than in the deciduous N. antarctica, also supporting our second hypothesis. In our assessment, some of the studied traits remained constant across the gradient when they were compared among areas; however, when the focus was established on the analysis of trait co-ordination at the individual level, additional xylem adjustments were revealed. Variation in xylem traits along the gradient Vessel density increased at the drier edge of the gradient in both species, and the evergreen E. coccineum also experienced a decrease in Dh. Small vessels are indisputably more resistant to embolism, thus a reduction in the lumen area of the largest vessels under dry conditions could be interpreted as an increase in xylem safety (Tyree and Zimmermann, 2002). Higher xylem safety as a response to dryness has been previously reported within and across species, although some discrepancies emerge among studies. Across species, Maherali et al. (2004) observed increased xylem safety with dryness in evergreen angiosperms but detected a lack of correlation between climate and hydraulic safety in deciduous angiosperms. A similar pattern was observed among congeneric species with contrasting leaf habit (Villar-Salvador et al., 1997). Our results partially agree with both studies, since the deciduous N. antarctica experienced a similar pattern of Dh reduction with dryness, although it was only marginally significant. In this sense, our observations do agree with observed within-species reduction in MVA (strongly correlated with Dh) along aridity gradients for both evergreen (Fisher et al., 2007; von Arx et al., 2012) and deciduous species (Schreiber et al., 2015). A striking result of our study is that the reduction in Dh was not linked to a drop of Ks, probably because of the simultaneous increase in VD. Thus, estimated xylem efficiency was maintained in combination with different levels of expected xylem safety, departing from the safety–efficiency trade-off hypothesis (Sperry, 2003; Zanne et al., 2010). In addition, this departure might be related to the existence of different anatomical traits that operate simultaneously to determine total xylem efficiency (such as vessel diameter and length, inter-vessel pit membrane thickness and/or vessel connectivity), and to the fact that each of these features may have separate and different effects on safety (Gleason et al. 2016a). Changes in anatomical features did not occur linearly across the gradient. In fact, vessel traits in the dry and mesic areas showed less difference from each other than with respect to the wet area. Specifically, WD and Dh in E. coccineum showed similar average values in the dry and mesic areas. The existence of significant correlations of these traits with both climatic moisture deficit and temperature of the coldest month may give some clues to explain this similarity between sites. Temperature of the coldest month was lower on average in the mesic than in the dry area, suggesting a higher risk of freezing-induced embolism in the mesic area. As a response, plants may reduce vessel lumen area and increase the density of narrower vessels to prevent freeze–thaw embolism (Castro-Díez et al., 1998; Medeiros and Pockmann, 2014), similarly as for facing moisture deficit. However, according to the precipitation gradient, the dry area had a higher risk of drought-induced embolism, and therefore smaller vessels might help here to prevent embolism. The combination of both milder temperatures and higher moisture deficit may therefore have favoured adjustments at the xylem level similar to a combination of lower temperature with lower climatic moisture deficit. Interestingly, vessel grouping showed contrasting patterns in both species. In the evergreen E. coccineum, vessels tended to be more grouped in warmer and moister areas. This is in line with a positive effect of winter temperature on vessel grouping observed in the deciduous shrub Betula nana on Greenland (Nielsen et al., 2017). According to the air-seeding hypothesis (Alder et al., 1997; Wheeler et al., 2005; Loepfe et al., 2007), the larger the contact surface among vessels, the higher the risk of embolism spreading from one vessel to the next by the aspiration of air through the pit pores (Sperry and Tyree, 1988; Brodersen, 2013; Brodersen et al., 2013). The lower proportion of grouped vessels in dry and cold areas might thus be interpreted as an adjustment to lower the risk of embolism of connected vessels under frost and water limitation. In contrast, in the deciduous N. antarctica, the opposite trend was observed: vessels were more grouped in cold and dry areas. A high degree of vessel grouping might provide alternative pathways for water transport when a vessel is blocked by an embolism, allowing the water flow to continue through one or more adjacent vessels and thus improving the hydraulic redundancy (Baas et al., 1983; Tyree et al., 1994). However, our contrasting results in two coexisting widespread species support the controversy about the functional role of vessel grouping under water limitation (von Arx et al., 2013), pointing to a high degree of species specificity in this trait irrespective of the study system and a relative independence of vessel grouping from the other xylem traits. Co-ordination among xylem traits The analysis of the co-ordination between xylem traits at the plant level provided clues to interpret xylem adjustments across the climatic gradient. A remarkable result was that the percentage of conductive area (CA) was the main determinant of Ks in both species. CA is defined as VD times MVA (Carlquist, 2001). While the influence of VD and MVA on CA is thus linear, similar CA values might have led to very different Ks values depending on vessel lumen area distribution. The reason for this is that Ks scales to the fourth power of diameter (Tyree and Ewers, 1991), whereas Ks scales proportionately with vessel density, i.e. many small vessels within a given xylem area would form a less efficient xylem than fewer larger vessels. Yet, our results indicate that changes in vessel density and vessel lumen area distribution were co-ordinated to keep Ks constant. However, co-ordination among vessel traits at the plant level has been poorly explored so far, and thus more research is needed to evaluate the role played by trait integration on the hydraulic function (Gleason et al., 2016b). An additional result was that xylem traits in the evergreen E. coccineum showed a stronger climate control than in the deciduous N. antarctica. This tighter climatic control was indicated by the direct effect of climatic moisture deficit on Dh, VD and WD, supporting the higher xylem trait dependence from climate in evergreen species (Sperry et al., 1994; Villar-Salvador et al., 1997; Castro-Díez et al., 1998; Maherali et al., 2004; Pérez-de-Lis et al., 2018). Similarly, inter-relations among xylem traits were stronger in E. coccineum than in N. antarctica, including a negative covariation between VD and Dh that was not detected in the deciduous species. Wood density and climate Wood density was the trait with the lowest variability. Among several functional traits, WD has continuously shown a very low intra-specific variation, commonly lower than 10 % (Fajardo and Piper, 2011; Siefert et al., 2015; Fajardo, 2016, 2018). Nevertheless, WD variability was strongly correlated with climate in the evergreen E. coccineum, with denser wood occurring under dry and cold conditions as expected according to its positive correlation with resistance to hydraulic embolism (e.g. Hacke et al., 2001; Detmann et al., 2013; Rosner, 2017). In general, WD showed a negative correlation with Dh and was orthogonal to the other xylem traits. However, SEMs showed that at the individual level, WD in E. coccineum was determined by climatic moisture deficit and Dh, while it was completely independent of climate and of the other xylem traits in N. antarctica. The climatic independence of WD in N. antarctica is similar to that observed in the congeneric N. pumilio across elevational gradients (Fajardo, 2018), and to trends observed across species (Preston et al., 2006). A possible explanation for the independence between WD and the other traits is that WD was measured in the whole branch section, whereas anatomical traits were measured only in the last growth ring. In addition, WD may be more linked to the proportion of fibres or to wood elements other than vessels (Zanne et al., 2010; Zieminska et al., 2013). Conclusions We demonstrate that vessel size-related reduction in hydraulic efficiency within a species can be compensated through increases in vessel density. This implies a departure from the safety–efficiency trade-off, since estimated xylem efficiency can in fact be maintained in combination with different levels of expected xylem safety. Otherwise, the existence of contrasting relationships of xylem traits with climate in both study species indicates that predictions of community responses to local or environmental changes should consider the potential existence of differences in trait sensitivity to climatic variations among coexisting species. The fact that coexisting species may achieve similar responses to environmental gradients through alternative internal adjustments might increase resilience of temperate forests against unpredictable changes in climatic conditions. ACKNOWLEDGEMENTS We thank Laura Sánchez-Jardón for help with the fieldwork, and Miguel García Hidalgo and Gonzalo Juste for their invaluable assistance with laboratory tasks. We greatly acknowledge the enthusiastic review by Sean M. Gleason, as well as the constructive comments from another anonymous referee. This work was supported by the Chilean Fondo Nacional de Desarrollo Científico y Tecnológico [project FONDECYT 1160329 to A.F.]; the Spanish Ministry of Science and Innovation [FPI grant BES-2010-036041 to A.I.G.C]; the Spanish Ministry of Economy and Competitiveness [projects SPRING-CGL2017-87309-P and CGL2012-34209, excellence networks CGL2014-53840-REDT and CGL2016-81706-REDT, Juan de la Cierva-Formación grant FJCI-2015-24770 to A.I.G.C.]; COST Action FP1106 [regular STSM to A.I.G.C.]; and the Swiss State Secretariat for Education, Research and Innovation SERI [SBFI C12.0100 to G.v.A.]. LITERATURE CITED Alder NN, Pockman WT, Sperry JS, Nuismer S. 1997. Use of centrifugal force in the study of xylem cavitation. Journal of Experimental Botany  48: 665– 674. Google Scholar CrossRef Search ADS   Allen CD, Macalady AK, Chenchouni Het al.   2010. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management  259: 660– 684. Google Scholar CrossRef Search ADS   von Arx G, Dietz H. 2005. Automated image analysis of annual rings in the roots of perennial forbs. International Journal of Plant Sciences  166: 723– 732. Google Scholar CrossRef Search ADS   von Arx G, Carrer M. 2014. ROXAS – a new tool to build centuries-long tracheid lumen chronologies in conifers. Dendrochronologia  32: 290– 293. Google Scholar CrossRef Search ADS   von Arx G, Archer SR, Hughes MK. 2012. Long-term functional plasticity in plant hydraulic architecture in response to supplemental moisture. Annals of Botany  109: 1091– 1100. Google Scholar CrossRef Search ADS   von Arx G, Kueffer C, Fonti P. 2013. Quantifying plasticity in vessel grouping – added value from the image analysis tool ROXAS. IAWA Journal  34: 433– 445. Google Scholar CrossRef Search ADS   von Arx G, Crivellaro A, Prendin AL, Cufar K, Carrer M. 2016. Quantitative wood anatomy – practical guidelines. Frontiers in Plant Sciences  7: 781. Baas P, Werker E, Fahn A. 1983. Some ecological trends in vessel characters. IAWA Journal  4: 141– 159. Google Scholar CrossRef Search ADS   ter Braak CJF. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology  67: 1167– 1179. Google Scholar CrossRef Search ADS   Brodersen CR. 2013. Visualizing wood anatomy in three dimensions with high-resolution X-ray micro-tomography (μCT) – a review. IAWA Journal  34: 408– 424. Google Scholar CrossRef Search ADS   Brodersen CR, McElrone AJ, Choat B, Lee EF, Shackel KA, Matthews MA. 2013. In vivo visualizations of drought-induced embolism spread in Vitis vinifera. Plant Physiology  161: 1820– 1829. Google Scholar CrossRef Search ADS   Burns KC. 2004. Patterns in specific leaf area and the structure of a temperate heath community. Diversity and Distribution  10: 105– 112. Google Scholar CrossRef Search ADS   Carlquist S. 2001. Comparative wood anatomy: systematic, ecological, and evolutionary aspects of dicotyledon wood , 2nd edn. Berlin: Springer. Carrer M, von Arx G, Castagnieri D, Petit G. 2015. Distilling environmental information from time series of conduit size: the standardization issue and its relation to allometric and hydraulic constraints. Tree Physiology  35: 27– 33. Google Scholar CrossRef Search ADS   Castro-Díez P, Villar-Salvador P, Pérez-Rontomé C, Maestro-Martínez M, Montserrat-Martí G. 1998. Leaf morphology, leaf chemical composition and stem xylem characteristics in two Pistacia (Anacardiaceae) species along a climatic gradient. Flora  193: 195– 202. Google Scholar CrossRef Search ADS   Chave J, Coomes D, Jansen S, Lewis SL, Swenson NG, Zanne AE. 2009. Towards a worldwide wood economics spectrum. Ecology Letters  12: 351– 366. Google Scholar CrossRef Search ADS   Daly C, Halbleib M, Smith JIet al.   2008. Physiographically-sensitive mapping of temperature and precipitation across the conterminous United States. International Journal of Climatology  28: 2031– 2064. Google Scholar CrossRef Search ADS   De Frenne P, Graae BJ, Rodríguez-Sánchez Fet al.   2013. Latitudinal gradients as natural laboratories to infer species’ responses to temperature. Journal of Ecology  101: 784– 795. Google Scholar CrossRef Search ADS   Dettmann S, Pérez CA, Thomas FM. 2013. Xylem anatomy and calculated hydraulic conductance of four Nothofagus species with contrasting distribution in South-Central Chile. Trees – Structure and Function  27: 685– 696. Google Scholar CrossRef Search ADS   Fajardo A. 2016. Wood density is a poor predictor of competitive ability among individuals of the same species. Forest Ecology and Management  372: 217– 225. Google Scholar CrossRef Search ADS   Fajardo A. 2018. Insights into intraspecific wood density variation and its relationship with growth, height and elevation in a treeline species. Plant Biology  20: 456– 464. Google Scholar CrossRef Search ADS   Fajardo A, Piper FI. 2001. Intraspecific trait variation and covariation in a widespread tree species (Nothofagus pumilio) in southern Chile. New Phytologist  189: 259– 271. Google Scholar CrossRef Search ADS   Fajardo A, Piper FI. 2015. High foliar nutrient concentrations and resorption efficiency in Embothrium coccineum (Proteaceae) in southern Chile. American Journal of Botany  102: 208– 216. Google Scholar CrossRef Search ADS   Fisher JB, Goldstein G, Jones TJ, Cordell S. 2007. Wood vessel diameter is related to elevation and genotype in the Hawaiian tree Metrosideros polymorpha (Myrtaceae). American Journal of Botany  94: 709– 715. Google Scholar CrossRef Search ADS   Gärtner H, Lucchinetti S, Schweingruber FH. 2015. A new sledge microtome to combine wood anatomy and tree-ring ecology. IAWA Journal  36: 452– 459. Google Scholar CrossRef Search ADS   Gleason SM, Westoby M, Jansen Set al.   2016a. Weak tradeoff between xylem safety and xylem-specific hydraulic efficiency across the world’s woody plant species. New Phytologist  209: 123– 136. Google Scholar CrossRef Search ADS   Gleason SM, Westoby M, Jansen Set al.   2016b. On research priorities to advance understanding of the safety–efficiency tradeoff in xylem. New Phytologist  211: 1156– 1158. Google Scholar CrossRef Search ADS   Gorsuch DM, Oberbauer SF, Fisher JB. 2001. Comparative vessel anatomy of arctic deciduous and evergreen dicots. American Journal of Botany  88: 1643– 1649. Google Scholar CrossRef Search ADS   Greenwood S, Ruiz-Benito P, Martínez-Vilalta Jet al.   2017. Tree mortality across biomes is promoted by drought intensity, lower wood density and higher specific leaf area. Ecology Letters  20: 539– 553. Google Scholar CrossRef Search ADS   Hacke UG, Sperry JS. 2001. Functional and ecological xylem anatomy. Perspectives in Plant Ecology, Evolution and Systematics  4: 97– 115. Google Scholar CrossRef Search ADS   Hacke UG, Sperry JS, Pockman WT, Davis SD, McCulloh KA. 2001. Trends in wood density and structure are linked to prevention of xylem implosion by negative pressure. Oecologia  126: 457– 461. Google Scholar CrossRef Search ADS   Hacke UG, Spicer R, Schreiber SG, Plavcová L. 2017. An ecophysiological and developmental perspective on variation in vessel diameter. Plant, Cell and Environment  40: 831– 845. Google Scholar CrossRef Search ADS   Hamann A, Wang T, Spittlehouse DL, Murdock TQ. 2013. A comprehensive, high-resolution database of historical and projected climate surfaces for western North America. Bulletin of the American Meteorological Society  94: 1307– 1309. Google Scholar CrossRef Search ADS   Kolb K, Sperry JS. 1999. Differences in drought adaptation between subspecies of sagebrush (Artemisia tridentata). Ecology  80: 2373– 2384. Google Scholar CrossRef Search ADS   Lachenbruch B, McCulloh KA. 2014. Traits, properties, and performance: how woody plants combine hydraulic and mechanical functions in a cell, tissue, or whole plant. New Phytologist  204: 747– 764. Google Scholar CrossRef Search ADS   Loepfe L, Martínez-Vilalta J, Pinol J, Mencuccini M. 2007. The relevance of xylem network structure for plant hydraulic efficiency and safety. Journal of Theoretical Biology  247: 788– 803. Google Scholar CrossRef Search ADS   Luebert F, Pliscoff P. 2006. Sinopsis bioclimática y vegetacional de Chile . Santiago: Editorial Universitaria. Maherali H, Pockman WT, Jackson RB. 2004. Adaptive variation in the vulnerability of woody plants to xylem cavitation. Ecology  85: 2184– 2199. Google Scholar CrossRef Search ADS   Martínez-Vilalta J, García-Forner N. 2016. Water potential regulation, stomatal behavior and hydraulic transport under drought: deconstructing the iso/anisohydric concept. Plant, Cell and Environment  40: 962– 976. Google Scholar CrossRef Search ADS   Martínez-Vilalta J, Cochard H, Mencuccini Met al.   2009. Hydraulic adjustments of Scots pine across Europe. New Phytologist  184: 353– 364. Google Scholar CrossRef Search ADS   Medeiros JS, Pockman WT. 2014. Freezing regime and trade-offs with water transport efficiency generate variation in xylem structure across diploid populations of Larrea sp. (Zygophyllaceae). American Journal of Botany  101: 598– 607. Google Scholar CrossRef Search ADS   Nielsen SS, von Arx G, Damgaard CFet al.   2017. Xylem anatomical trait variability provides insight on the climate–growth relationship of Betula nana in Western Greenland. Arctic, Antarctic, and Alpine Research  49: 359– 371. Google Scholar CrossRef Search ADS   Nonweiler TRF. 1975. Flow of biological fluids through non-ideal capillaries. In: Zimmermann MH, Milburn JA, eds. Encyclopaedia of plant physiology, New Series, Vol 1. Transport in plants. I. Phloem transport . Berlin: Springer, 474– 477. Oksanen J, Blanchet FG, Friendly Met al.   2016. vegan: community ecology package. R package version 2.4-1 . https://CRAN.R-project.org/package=vegan Olano JM, González-Muñoz N, Arzac Aet al.   2017. Sex determines xylem anatomy in a dioecious conifer: hydraulic consequences in a drier world. Tree Physiology  37: 1493– 1502. Google Scholar CrossRef Search ADS   Pérez-de-Lis G, Rozas V, Vázquez-Ruiz RA, García-González I. 2018. Do ring-porous oaks prioritize earlywood vessel efficiency over safety? Environmental effects on vessel diameter and tyloses formation. Agricultural and Forest Meteorology  248: 205– 214. Google Scholar CrossRef Search ADS   Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core Team. 2016. nlme: linear and nonlinear mixed effects models. R package version 3.1-128 , URL: http://CRAN.R-project.org/package=nlme>. Preston KA, Cornwell WD, DeNoyer JL. 2006. Wood density and vessel traits as distinct correlates of ecological strategy in 51 California coast range angiosperms. New Phytologist  170: 807– 818. Google Scholar CrossRef Search ADS   R Core Team. 2016. R: a language and environment for statistical computing . Vienna: R Foundation for Statistical Computing. URL https://www.R-project.org/. Richards CL, Pennings SC, Donovan LA. 2005. Habitat range and phenotypic variation in salt marsh plants. Plant Ecology  176: 263– 273. Google Scholar CrossRef Search ADS   Rosner S. 2017. Wood density as a proxy for vulnerability to cavitation: size matters. Journal of Plant Hydraulics  4: e-001. doi: 10.20870/jph.2017.e001. Google Scholar CrossRef Search ADS   Sánchez-Salguero R, Navarro-Cerrillo R, Camarero JJ, Fernández-Cancio Á. 2012. Selective drought-induced decline of pine species in southeastern Spain. Climate Change  113: 767– 785. Google Scholar CrossRef Search ADS   Schreiber SG, Hacke UG, Hamann A, Thomas BR. 2011. Genetic variation of hydraulic and wood anatomical traits in hybrid poplar and trembling aspen. New Phytologist  190: 150– 160. Google Scholar CrossRef Search ADS   Schreiber SG, Hacke UG, Hamann A. 2015. Variation of xylem vessel diameters across a climate gradient: insight from a reciprocal transplant experiment with a widespread boreal tree. Functional Ecology  29: 1392– 1401. Google Scholar CrossRef Search ADS   Siefert A, Violle C, Chalmandrier Let al.   2015. A global meta-analysis of the relative extent of intraspecific trait variation in plant communities. Ecology Letters  18: 1406– 1419. Google Scholar CrossRef Search ADS   Sperry JS. 2003. Evolution of water transport and xylem structure. International Journal of Plant Sciences  164: S115– S127. Google Scholar CrossRef Search ADS   Sperry JS, Tyree MT. 1988. Mechanism of water stress-induced xylem embolism. Plant Physiology  88: 581– 587. Google Scholar CrossRef Search ADS   Sperry JS, Nichols KL, Sullivan JEM, Eastlack SE. 1994. Xylem embolism in ring-porous, diffuse-porous and coniferous trees of Northern Utah and interior Alaska. Ecology  75: 1736– 1752. Google Scholar CrossRef Search ADS   Sperry JS, Meinzer FC, McCulloh KA. 2008. Safety and efficiency conflicts in hydraulic architecture: scaling from tissues to trees. Plant, Cell and Environment  31: 632– 645. Google Scholar CrossRef Search ADS   Tyree MT, Ewers FW. 1991. The hydraulic architecture of trees and other woody plants. New Phytologist  119: 345– 360. Google Scholar CrossRef Search ADS   Tyree MT, Sperry JS. 1989. Vulnerability of xylem to cavitation and embolism. Annual Review of Plant Physiology and Plant Molecular Biology  40: 19– 38. Google Scholar CrossRef Search ADS   Tyree MT, Zimmermann MH. 2002. Xylem structure and the ascent of sap . Berlin: Springer-Verlag. Google Scholar CrossRef Search ADS   Tyree MT, Davis SD, Cochard H. 1994. Biophysical perspectives of xylem evolution: is there a tradeoff of hydraulic efficiency for vulnerability to dysfunction? IAWA Journal  15: 335– 360. Google Scholar CrossRef Search ADS   Villar-Salvador P, Castro-Díez P, Pérez-Rontomé C, Montserrat-Martí G. 1997. Stem xylem features in three Quercus (Fagaceae) species along a climatic gradient in NE Spain. Trees – Structure and Function  12: 90– 96. Violle C, Jiang L. 2009. Towards a trait-based quantification of species niche. Journal of Plant Ecology  2: 87– 93. Google Scholar CrossRef Search ADS   Wheeler JK, Sperry JS, Hacke UG, Hoang N. 2005. Inter-vessel pitting and cavitation in woody Rosaceae and other vesselled plants: a basis for a safety versus efficiency trade-off in xylem transport. Plant, Cell and Environment  28: 800– 812. Google Scholar CrossRef Search ADS   Williamson GB, Wiemann MC. 2010. Measuring wood specific gravity… Correctly. American Journal of Botany  97: 519– 524. Google Scholar CrossRef Search ADS   Zanne AE, Westoby M, Falster DSet al.   2010. Angiosperm wood structure: global patterns in vessel anatomy and their relation to wood density and potential conductivity. American Journal of Botany  97: 207– 215. Google Scholar CrossRef Search ADS   Zeballos SR, Giorgis MA, Cabido M, Gurvich DE. 2017. Unravelling the coordination between leaf and stem economics spectra through local and global scale approaches. Austral Ecology  42: 394– 403. Google Scholar CrossRef Search ADS   Ziemińska K, Butler DW, Gleason SM, Wright IJ, Westoby M. 2013. Fibre walls and lumen fractions drive wood density variation across 24 Australian angiosperms. AoB Plants  5: plt046. doi: 10.1093/aobpla/plt046. Google Scholar CrossRef Search ADS   Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. 2009. Mixed effects models and extensions in ecology with R . New York: Springer. Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Annals of BotanyOxford University Press

Published: May 24, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off