SEARCH

SEARCH BY CITATION

Keywords:

  • anatomy;
  • capacitance;
  • cavitation;
  • Ceanothus;
  • evolution;
  • Rhamnus;
  • trade-offs;
  • xylem

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • • 
    Here, hypotheses about stem and root xylem structure and function were assessed by analyzing xylem in nine chaparral Rhamnaceae species.
  • • 
    Traits characterizing xylem transport efficiency and safety, mechanical strength and storage were analyzed using linear regression, principal components analysis and phylogenetic independent contrasts (PICs).
  • • 
    Stems showed a strong, positive correlation between xylem mechanical strength (xylem density and modulus of rupture) and xylem transport safety (resistance to cavitation and estimated vessel implosion resistance), and this was supported by PICs. Like stems, greater root cavitation resistance was correlated with greater vessel implosion resistance; however, unlike stems, root cavitation resistance was not correlated with xylem density and modulus of rupture. Also different from stems, roots displayed a trade-off between xylem transport safety from cavitation and xylem transport efficiency. Both stems and roots showed a trade-off between xylem transport safety and xylem storage of water and nutrients, respectively.
  • • 
    Stems and roots differ in xylem structural and functional relationships, associated with differences in their local environment (air vs soil) and their primary functions.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The xylem of angiosperms has three basic functions, namely long-distance water transport, mechanical support of the plant body, and storage of water and nutrients. Understanding how these xylem functions are integrated to maximize fitness is critical for testing ecological and evolutionary hypotheses concerning the diversity of structural, physiological and life-history types found among woody plant species.

Analyses of stem xylem transport show that a greater ability to conduct water safely (i.e. greater resistance to cavitation) is correlated with less efficient water transport (Hacke et al., 2006). For freezing-induced cavitation, this relationship occurs because vessels with larger volumes (and diameters) are simultaneously more efficient at transport and more prone to freeze–thaw-induced cavitation (Davis et al., 1999a; Pittermann & Sperry, 2003; Pratt et al., 2005). For water stress-induced cavitation, the trade-off between safety and efficiency is explained by the pit area hypothesis. According to this model, taxa with more efficient transport are more susceptible to cavitation because they have a greater pitted area per vessel. More pit area leads to a greater probability of having a large pit membrane pore, and larger pores are more prone to cavitation by air seeding (Hargrave et al., 1993; Choat et al., 2003; Wheeler et al., 2005; Hacke et al., 2006). Alternatively, more pit area may lead to a greater probability that a larger pit pore is adjacent to a gas-filled space.

The mechanical and transport functions of xylem may be interdependent, even though the cells in angiosperms that provide most of the mechanical support (fibers) are not the same cells that provide most of the water transport (vessels). At the cellular level, vessel pairs are exposed to large hoop stresses that could lead to cell collapse if the adjoining vessel wall thickness (t) to lumen (b) ratio ((t/b)2) is not high enough (Hacke et al., 2001; Cochard et al., 2004; Brodribb & Holbrook, 2005). Thus, in addition to avoiding cavitation by air-seeding, safe xylem transport is also characterized by resistance to vessel implosion. At the tissue level, the fiber matrix surrounding the vessels also appears to be important in resisting vessel wall collapse under negative pressures (Hacke et al., 2001; Jacobsen et al., 2005, in press). For stems, these patterns lead to a positive relationship between xylem safety from cavitation and mechanical strength (xylem density, modulus of rupture (MOR) and modulus of elasticity (MOE)).

Storage of both water and nutrients in stem xylem are important factors affecting plant fitness and life history. For example, stored water helps to buffer against severe seasonal water deficits (Waring & Running, 1978; Stratton et al., 2000; Pratt & Black, 2006). Xylem is also an important storage tissue for nutrients, such as nitrogen and starch (Sauter & van Cleve, 1994; Kozlowski & Pallardy, 1996).

Although storage is important, much research on xylem function has focused on xylem transport and mechanics, and thus we currently lack information on how xylem storage impacts the water transport and mechanical support functions of xylem. Moreover, relatively little is known about how these relationships vary between root and stem xylem. It is possible that roots have different transport and mechanical constraints compared with stems because their mechanical environment is a soil matrix and roots generally experience less negative hydrostatic pressures than stems. Furthermore, roots uniquely function in below-ground anchorage, radial transport and storage.

The purpose of the study presented here was to examine relationships between xylem transport and the mechanical and storage traits of stems and roots of nine Rhamnaceae species. Xylem transport was estimated in the two categories of safety and efficiency. Xylem safety was quantified as resistance to water stress-induced xylem cavitation (Ψ50; i.e. the water potential at 50% loss of hydraulic conductivity), estimated vessel pair implosion resistance (inline image) and vessel redundancy. Xylem efficiency was quantified as xylem-specific hydraulic conductivity (Ks) and hydraulic mean vessel diameter (dh). Storage of water was quantified, for stems only, as capacitance, and potential nutrient storage was estimated as the percentage parenchyma in cross-section. Stem and root biomechanics were quantified as xylem density, MOR, MOE and cross-sectional fiber wall area. Relationships between traits were assessed using linear regressions and principal component (PC) analysis.

The taxa sampled in the present study are all in the family Rhamnaceae and form three clades (Fig. 1). Of the six Ceanothus species sampled, three fall in each of two well-supported subgenera that diverged between 18 and 39 million years ago and subsequently spread throughout the California Floristic Province and beyond (Mason, 1942; Hardig et al., 2000). Ceanothus spp., in the subgenus Cerastes, generally lack the ability to resprout following fire, have small thick leaves with stomatal crypts, have relatively shallow roots (Hellmers et al., 1955) and occupy relatively arid microsites (McMinn, 1942; Davis et al., 1999b; Ackerly, 2003). By contrast, Ceanothus spp., of the subgenus Ceanothus, often sprout following fire, have larger and thinner leaves lacking stomatal crypts and occupy more mesic microsites (Meentemeyer & Moody, 2002). The divergence of the two subgenera resulted in basal divergence in a suite of adaptive traits related to stress tolerance and life-history type, many of which have been conserved across extant taxa. Sampled species within the Rhamnus clade are all resprouters following fire, but other than that they tend to be more diverse in their range of leaf traits and apparent microsite aridity preference than species within either the Cerastes or Ceanothus clades. Because of the known phylogenetic structure of numerous traits related to life history and stress tolerance among the sampled taxa, we calculated phylogenetic independent contrasts (PICs) for sampled traits. This approach assesses the influence of phylogeny on trait correlations and accounts for the statistical nonindependence of species (Felsenstein, 1985). However, because a fully resolved phylogeny is unavailable for the species sampled here, as well as other uncertainties when applying PICs in the present study (see the Materials and Methods), we report correlations with and without phylogenetic correction. Reporting both also has the benefit of allowing examination of ecological as well as evolutionary relationships in xylem traits.

image

Figure 1. Phylogeny used to determine phylogenetic independent contrasts among nine species, followed by abbreviations in parentheses (a). (b and c) The three alternative resolutions for each of the two polytomies existing in the phylogeny (Jacobsen et al., in press). All of these alternative trees and combinations were considered in analyses, resulting in nine possible trees. The phylogeny was constructed using published results referenced in the text, and all branch lengths are assumed to be equal. Although we place Rhamnus californica in the genus Rhamnus here, recent molecular analyses have placed it in the genus Frangula (Bolmgren & Oxelman, 2004).

Download figure to PowerPoint

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Naturally occurring populations of nine Rhamnaceae shrub species were sampled in the Santa Monica Mountains in Los Angeles County, California (see Pratt et al., 2007 for a description of study sites). These nine species fall into three well-supported monophyletic clades (Fig. 1) (McMinn, 1942; Hardig et al., 2000). Nomenclature for species follows Hickman (1993).

Seasonal water status, hydraulic conductivity and vulnerability of the xylem to cavitation were all measured as described in Pratt et al. (2007). Xylem transverse areas of stems and roots were measured to calculate Ks, which is one measure of xylem hydraulic efficiency. Stems were thin-sectioned, mounted in glycerol, digitally photographed and the xylem area was measured with scionimage (scionimage version Beta 3b; Scion Corporation, Frederick, MD, USA). The pith was measured and subtracted from the xylem area.

Xylem anatomical traits were measured on the same stems and roots in which vulnerability to cavitation was measured. For each organ, images were taken of several wedge-shaped cross-sectional sectors, to sample for vessel and fiber features, using a digital camera (Spot RT color camera; Diagnostic Instruments, Sterling Heights, MI, USA) attached to a light microscope (Nikon Microphot-FX; Nikon, Garden City, NY, USA). ScionImage was used for all image analyses and was performed at California State University, Bakersfield (ScionImage version 4.03.2; Scion Corporation). Vessel lumen diameter (d) and wall thickness, fiber lumen diameter and wall thickness, and percentage transverse fiber, vessel and parenchyma area, were measured. All of the vessels and fibers in sectors were measured until 100 vessels and 100 fibers had been measured in a given organ. The dh value was calculated from the formula dh = (∑d5)/(∑d4), based upon all the sampled vessels in an organ. This parameter weights the vessel diameters by their hydraulic contribution, which is a function of the diameter to the fourth power. The inline image (Hacke et al., 2001) was determined for those vessels, within the 100 vessels sampled per organ that formed pairs in which one or both vessels had diameters ± 5 µm of the calculated dh, with t as the combined wall thickness of adjoining vessels and b as the lumen diameter of the selected vessel. The percentage transverse area of cell types was estimated through measurements of vessel, fiber and parenchyma area in four randomly chosen cross-sectional sectors per organ at ×200 magnification. Fiber wall area in a cross-section was calculated using the average (n = 100) fiber wall to lumen ratio for each stem or root and the cross-sectional area of fibers for that same organ.

Xylem density was measured on stems used for vulnerability to cavitation measurements, following the methods of Hacke et al. (2000). Segments 3 cm long were soaked in water under vacuum until gas bubbles no longer appeared (c. 24 h). Bark and pith were removed from segments and the density was measured by water displacement on an analytical balance (AE163; Mettler-Toledo, Columbus, OH, USA). Samples were dried in an oven at 70°C, for at least 48 h, and the dry weight of the xylem was measured for calculation of densities: dry weight (kg)/fresh volume (m3).

The mechanical properties of stems and roots were sampled on organs similar in size to those used for vulnerability curves (Pratt et al., 2007). Stems were sampled in October 2005 and roots were sampled in July 2002. Samples were shipped overnight on ice to Michigan State University where they were measured using an Instron Universal Machine (Model 4202; Instron Corporation, Canton, MA, USA). Stem and root segments were kept at approx. 10°C until MOE and MOR were measured. A four-point bending test, with a compression load cell of 500 N, was conducted as described by Woodrum et al. (2003) and Jacobsen et al. (2005) on stem segments cut to 0.275 m long.

Xylem capacitance was measured to estimate the water storage capacity of the xylem. This was performed by generating pressure–volume curves for xylem disks, with pith and bark removed, that measured c. 8 mm in diameter and 4.5 mm long. These dimensions were used because they were the largest that would fit in our C-52 psychrometer chambers (Wescor Inc., Logan, UT, USA). Plant stems for this procedure were removed in the early morning from plants growing at our field sites during the rainy winter months (28 October 2004 to 7 March 2005). The ends of cut stems (c. 0.5 m long) were immediately submerged in water to hydrate them and the leaves were sealed in a large plastic bag to minimize water loss. In the laboratory, stems were hydrated for c. 2 h. Branchlet water potential was measured with a pressure chamber before removing xylem disks, and ranged from −0.16 to −1.01 MPa. These values were similar to the predawn branchlet water potential measured in the field during the winter rainy season (Pratt et al., 2007).

To construct pressure–volume curves, the xylem was allowed to dehydrate over time and the xylem water potential (Ψx) and mass were measured during drying. To prepare the xylem disks, xylem disks were cut in half under water, the pith was removed with forceps under a dissecting microscope and the bark was removed. Following this, the disks were blotted, weighed on an analytical balance (AE163; Mettler-Toledo), and then sampled for Ψx in psychrometers connected to a datalogger (Model CR-7; Campbell Scientific, Logan, UT, USA). This first mass was taken as the saturated mass (i.e. the mass near saturated water content). Psychrometers were calibrated with salt solutions of known water potential. Data were generated by taking measurements of tissue mass and Ψx over time (c. 7 d) as the disks dried down. Xylem mass was weighed before and after insertion into psychrometric chambers. Xylem dried rapidly when left in open air, but maintained a relatively constant mass while in chambers. Relative water content (RWC) was calculated as ((fresh mass – dry mass)/(saturated mass – dry mass) × 100). Capacitance was calculated as the change in percentage RWC per unit change in Ψx. The range in Ψx, used in the calculation for each species, was the difference in branchlet Ψx measured between the dry and wet seasons for plants growing at each field site (Pratt et al., 2007).

Relationships between physiological and anatomical traits were analyzed as linear regressions (Minitab 14.12; State College, State College, PA, USA). The a priori goal was to determine associations between traits that characterize xylem transport, mechanics and storage. We quantified 12 traits associated with these three xylem functions. Because of the relatively small sample size, α was set at 0.10 to reduce the chance of committing type II errors. A sequential Bonferroni correction was used to hold the experiment-wise type I error rate at 0.10 (Sokal & Rohlf, 1995). Differences between stem and root traits were analyzed using analysis of variance (anova), with species and organ (stem and root) as independent variables (α = 0.05). Data were transformed to satisfy assumptions of parametric models, and percentage data were arcsine transformed.

Because closely related species tend to have traits that are more similar than is the case for distantly related species, we also analyzed the data taking into account the evolutionary relationships of the sampled species (Preston & Ackerly, 2003; Pratt et al., 2007). To do this we used PICs that allowed us to test for correlated evolutionary change between two traits (Felsenstein, 1985). Contrasts were calculated using compare software (Martins, 2004).

PIC analyses should be interpreted in light of several limitations in relation to the analyses of the present study (see also Ackerly, 2004). First, the phylogeny of the sampled taxa is not fully resolved, resulting in nine possible phylogenies (Fig. 1). Contrasts were calculated for all nine possible phylogenies (Fig. 1). Correcting for multiple comparisons is customary in such a case, but this would have severely limited our power and inflated our type II error, especially with the relatively low sample size in the present study. Thus, for each regression between two traits, the average P-value for all nine phylogenies was calculated. We felt justified in taking the average because when the average correlation between two traits was significant, it was generally significant across all nine phylogenies. The few occurrences when this was not the case are highlighted in the Discussion. Second, information was not available on the phylogenetic branch lengths, which is proportional to the degree of trait change between an ancestor and a descendent under a Brownian motion model of evolution (Felsenstein, 1985). Contrast calculations were based on equal branch lengths, a reasonable assumption in the absence of branch length information, and one that minimizes type I error rates (Ackerly, 2000). Finally, the objective of the present study was to sample nine species intensively, and thus the sample size in the present study is relatively small. On account of this, we set α at 0.10 to reduce the chance of committing type II errors. Because of the small sample size, poor phylogeny resolution and lack of branch length information, the PIC analyses should be viewed with caution. However, it should be noted that the raw data analyses should also be viewed with caution because testing the hypothesis that two traits are functionally related using phylogenetically uncorrected data leads to an inflation of type I error (Martins & Garland, 1991). Because of the different limitations of raw data and PICs we present both.

Principal components analysis (PCA) was used to examine relationships among traits simultaneously. Trait values were standardized by calculating z-scores before analysis (McGarigal et al., 2000).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The degree of water stress experienced by plants growing in the field during the dry season was correlated to cavitation resistance of stems and roots (Fig. 2). The slope and standard error of the regression between Ψ50 and minimum predawn dry-season water potential (Ψmin) was 1.29 ± 0.25 for stems and 1.14 ± 0.37 for roots, with neither of the slopes being different from 1 (Fig. 2). The intercept of the stems and roots were significantly different, indicating that stems were more resistant to cavitation than roots (P < 0.01; Fig. 2; Table 1). PICs for these values were significantly correlated for stems, but not for roots (Fig. 2).

image

Figure 2. The water potential at 50% loss of hydraulic conductivity (Ψ50) for stems (black text, closed symbols and solid line) and roots (gray text, open symbols and dotted line), plotted as a function of dry season predawn water potential (Ψmin). Phylogenetic independent contrasts for one of nine phylogenies (see the Materials and Methods) for Ψ50 are plotted as a function of predawn Ψmin contrasts (inset). Abbreviations for species are given in Fig. 1. Noncontrast data points are means ± 1 standard error (n = 6), and lines are linear regressions. Regression lines are only shown that are statistically significant (P = 0.001 for stems and 0.02 for roots, and average P = 0.03 for stems in inset). The dashed line indicates a 1 : 1 line.

Download figure to PowerPoint

Table 1.  Stem and root xylem traits compared via analysis of variance (anova)
 Ks (kg s−1 MPa−1 m−1)dh (µm)inline imageΨ50 (MPa)Density (kg m3)MOR (N mm2)MOE (N mm2)FWA (%)Parenchyma (%)
  1. Data are means (n = 9) and 1 standard error in parentheses. Capacitance was not included because it was only measured on stems.

  2. See Fig. 1 for trait abbreviations and see the Materials and Methods for their corresponding xylem functions.

  3. Density, refers to xylem density; dh, hydraulic mean vessel diameter; FWA, fiber wall area; Ks, xylem-specific hydraulic conductivity; MOE, modulus of elasticity; MOR, modulus of rupture; inline image, vessel pair implosion resistance; Ψ50, water potential at 50% loss of hydraulic conductivity.

Stems1.10 (0.07)32.9 (1.1)0.029 (0.001)−6.03 (0.77)616 (23)207 (9)7018 (531)58.5 (1.5)13.0 (1.0)
Roots6.19 (1.60)61.5 (4.4)0.033 (0.003)−3.77 (0.79)582 (16)202 (20)8334 (897)49.3 (1.2)17.5 (1.2)
P< 0.001< 0.0010.01< 0.0010.0020.950.19< 0.001< 0.001

Greater resistance to water stress-induced cavitation (Ψ50) was correlated with lower values of Ks for roots (Fig. 3), and this was true for raw trait values (Fig. 3) as well as for PIC values (Fig. 3). By contrast, stem Ψ50 was not correlated with Ks for raw data (Fig. 3) or PICs (Fig. 3). Root Ks was greater than stem Ks (Table 1).

image

Figure 3. The water potential at 50% loss of hydraulic conductivity (Ψ50) for stems (black text, closed symbols and solid line) and roots (gray text, open symbols and dotted line), plotted as a function of xylem-specific hydraulic conductivity (Ks). Phylogenetic independent contrasts for one of nine phylogenies (see the Materials and Methods) for Ψ50 are plotted as a function of Ks contrasts (inset). Abbreviations for species are given in Fig. 1. Data points are means ± 1 standard error (n = 6), and lines are linear regressions. Regression lines are only shown when they are statistically significant (P = 0.03 for root data (P-value and r2 are for semilog transformed root data), and average P = 0.07 in inset); ns, nonsignificant regression.

Download figure to PowerPoint

Resistance to Ψ50 was correlated with several mechanical traits in stems but not in roots (Fig. 4). In stems, increasing xylem density, MOR and inline image were correlated with increased cavitation resistance (Fig. 4a–c). Like stems, more negative Ψ50 values in roots were correlated with increasing inline image (Fig. 4c); however, root xylem density (Fig. 4a) and MOR were not correlated with Ψ50 (Fig. 4b). Root inline image was generally greater than or equal to stem inline image (Fig. 4c) and, when analyzed across all species, inline image was significantly greater for roots compared with stems (Table 1).

image

Figure 4. Stem (black text, closed symbols and solid line) and root (gray text, open symbols and dotted line) xylem density (a), modulus of rupture (MOR) (b) and vessel thickness to span ratio inline image (c), plotted as a function of the water potential at 50% loss in hydraulic conductivity (Ψ50). The dashed line in (c) represents the line of theoretical implosion for vessels, below which vessels are in danger of collapsing (Hacke et al., 2001). MOR and modulus of elasticity (MOE) were correlated for stems and roots (r2 = 0.55, P = 0.02), but only MOR is shown, for brevity. Phylogenetic independent contrasts for xylem density (inset in a), MOR (inset in b) and inline image × 100 (inset in c), are plotted as a function of Ψ50 contrasts. Abbreviations for species are in Fig. 1. Noncontrast data points are means ± 1 standard error (n = 6), and lines are linear regressions. Regression lines are only shown that are statistically significant and others are noted as ns, for not significant.

Download figure to PowerPoint

Correlations among stem PICs for Ψ50 with xylem density and MOR were significant (Fig. 4a,b), whereas Ψ50 and inline image were not correlated (Fig. 4c). In contrast to stems, root PICs for Ψ50 were not correlated with xylem density or MOR (Fig. 4a,b), but were correlated with inline image (Fig. 4c).

In transverse xylem sections of stems and roots, stems and roots differed in their percentages of vessel, parenchyma and fiber area (Fig. 5; Table 1). The percentage fiber area was 13% greater in stems than in roots. In roots, the percentage area of parenchyma was c. 4% greater than in stems, and the percentage area of vessels was 9% greater than stems. The percentage of fiber wall area was also greater in transverse section for stems compared with roots (Table 1).

image

Figure 5. Stem and root percentage of cell types in xylem cross-sections. Data are averages ± 1 standard error (n = 9). Vessel and fiber areas refer to wall plus lumen area.

Download figure to PowerPoint

The percentages of vessel and parenchyma area were correlated with Ψ50 for roots, but not for stems, and only for raw values, not PICs (Fig. 6). Root percentage vessel and parenchyma areas were both correlated with Ψ50, such that greater vessel and lower parenchyma areas were correlated with more negative Ψ50 (Fig. 6a,b). Thus, the roots with the most negative Ψ50 values tended to have a larger percentage of vessel area and a smaller percentage of parenchyma area in the xylem transverse section. This stands in contrast to stems where neither vessel nor parenchyma were correlated with Ψ50 (Fig. 6a,b).

image

Figure 6. Stem (black text, closed symbols and solid line) and root (gray text, open symbols and dotted line) cross-sectional area percentage of vessel (a) and parenchyma (b) areas plotted as a function of the water potential at 50% loss in hydraulic conductivity (Ψ50). Phylogenetic independent contrasts for one of nine phylogenies for vessel (a) and parenchyma (b) areas are plotted as a function of Ψ50 contrasts in the insets. Abbreviations for species are in Fig. 1. Noncontrast data points are means ± 1 standard error (N = 6), and lines are linear regressions. Regression lines are only shown that are statistically significant and others are noted as ns, for not significant.

Download figure to PowerPoint

Most xylem mechanical traits, such as xylem density, MOR and MOE, did not correlate with Ks for both stems and roots (data not shown). However, vessel inline image was inversely correlated with Ks for roots, but not for stems (Fig. 7), and this was true for raw data and PICs (Fig. 7).

image

Figure 7. Stem (black text, closed symbols and solid line) and root (gray text, open symbols and dotted line) vessel thickness to span ratio, inline image, plotted as a function of the xylem-specific hydraulic conductivity (Ks). Phylogenetic independent contrasts for one of nine phylogenies for inline image × 100 are plotted as a function of Ks contrasts (inset). Abbreviations for species are in Fig. 1. Noncontrast data points are means ± 1 standard error (n = 6), and lines are linear regressions. P-values for regressions are 0.01 for roots for both raw and contrast data, and stems are not significant (ns).

Download figure to PowerPoint

Xylem capacitance, measured only for stems, was strongly correlated with xylem density, Ψ50 and vessel inline image (Fig. 8). Greater xylem capacitance was correlated with lower xylem density (Fig. 8a). Likewise, greater capacitance was correlated with a lower degree of cavitation resistance (less negative Ψ50; Fig. 8b). Greater capacitance was also correlated with reduced inline image (Fig. 8c). Lastly, greater capacitance was correlated with greater fiber lumen diameter (r2 = 0.60; P = 0.01), and greater total fiber and vessel lumen in cross-section (r2 = 0.42; P = 0.05; data not shown). PICs for stems were also correlated for capacitance and xylem density (Fig. 8a), and capacitance and Ψ50 (Fig. 8b), but PICs were not correlated for capacitance and inline image (Fig. 8c).

image

Figure 8. Stem xylem density (a), water potential at 50% loss in hydraulic conductivity (Ψ50; (b) and vessel resistance to implosion (c); inline image), plotted as a function of xylem water storage (capacitance). Phylogenetic independent contrasts for one of nine phylogenies for xylem density (a), Ψ50 (b) and inline image (c), plotted as a function of capacitance contrasts, are shown as insets. Abbreviations for species are as detailed in Fig. 1. Noncontrast data points are means ± 1 standard error (n = 6), and lines are linear regressions. Regression lines are only shown that are statistically significant (ns, not significant).

Download figure to PowerPoint

PCs analyses summarize and confirm the linear regression analyses (Fig. 9). For stems, traits associated with mechanical strength (MOR, MOE, fiber wall area and xylem density) clustered with the xylem transport safety measures of Ψ50, inline image and Ψmin, and loaded strongly along PC 1 (Fig. 9a). Traits associated with xylem efficiency (Ks and dh) and water storage (capacitance) were negatively associated with the mechanical and transport safety traits; however, the efficiency traits had relatively low loading values (< 0.3), indicating that the negative association was weak. The traits most strongly associated with PC 2 were percentage cell areas (vessel and parenchyma) and the biomechanical traits of inline image and MOE. The clearest pattern among these traits is that high percentages of parenchyma were associated with lower values of MOE, indicating the importance of parenchyma in stem flexibility.

image

Figure 9. Principal components analyses for stems (a) and roots (b). Trait abbreviations follow the previous figures, but also include MOE (modulus of elasticity), dh (hydraulic mean vessel diameter) and density (xylem density). Solid circles surround traits that loaded most strongly on principal component (PC) 1, and traits circled with dashed lines loaded most strongly on PC 2. Overlap between circles denotes traits that loaded most strongly on PC 2, but also showed relatively strong loading on PC 1. Absolute values for water potential at 50% loss of hydraulic conductivity (Ψ50) and dry season predawn water potential (Ψmin) were used.

Download figure to PowerPoint

Like stems, root mechanical (MOE and MOR) and transport safety (Ψ50, Ψmin and inline image) traits clustered at one end of PC 1; however, unlike stems, the mechanical traits of xylem density and fiber wall area did not cluster with the other traits (Fig. 9b). Also, as in stems, transport efficiency traits (Ks and dh) clustered and loaded most heavily along PC 1, opposite the transport safety and mechanical traits (Fig. 9b). In the case of roots, the negative loading scores for the efficiency traits were relatively strong (> 0.3), showing the trade-off between transport safety and efficiency present in roots. The strongest loading scores along PC 2 were for percentage cell areas (vessel and parenchyma) and the mechanical traits of xylem density and fiber wall area (Fig. 9b).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The species most resistant to cavitation (more negative Ψ50) experienced the most negative predawn water potentials during the dry season (Ψmin). The slope of this relationship was not different from 1.0 for stems and roots, which is similar to that reported in a study across 26 chaparral species sampled in the Santa Monica Mountains (Jacobsen et al., in press) and among shrubs in the great basin (Hacke et al., 2000). For stems of all species, data points were not different from, or fell below, the 1 : 1 line, indicating that they were predicted to have ≤ 50% loss in hydraulic conductivity during the dry season, assuming that the xylem water potential at midday was not significantly different from Ψmin. This agrees with the recent stem measurements of dry season native embolism of 10 chaparral species (including four Ceanothus spp.) in the Santa Monica Mountains, which found a mean and standard error percentage loss of conductivity of 42 ± 8%, a value not different from 50 (A. Jacobsen, unpublished). This supports our assumption that Ψmin is a robust estimate of minimum xylem water potential. The fact that Ψ50 strongly correlates to Ψmin shows that Ψ50 is a precise index of the maximum seasonal water stress experienced by the xylem of chaparral shrubs under natural field conditions. For roots, many species fell above the 1 : 1 line, indicating that they were predicted to have > 50% loss in hydraulic conductivity, which is caused by the greater vulnerability of roots to cavitation than stems. Differences between stem and root cavitation resistance is fully discussed in a separate study (Pratt et al., 2007).

A greater degree of mechanical strength was correlated with greater resistance to cavitation for stems. This was true at the cellular level (inline image), tissue level (xylem density) and organ level (modulus of rupture). This probably reflects the need for xylem cells to resist the stresses created by negative hydrostatic pressures in the xylem (Hacke et al., 2001; Jacobsen et al., 2005). The correlations of PICs held for xylem density and MOR, indicating that the relationships were not significantly influenced by sampled taxa phylogeny. Estimated vessel implosion resistance (inline image) and cavitation resistance for stems were not correlated using PICs. This finding is similar to that of Jacobsen et al. (in press), who found that vessel implosion resistance and cavitation resistance were not evolutionarily correlated among 26 species of chaparral shrubs. It appears that the evolution of increased cavitation resistance was not concurrent with the evolution of increased inline image in these species. This is surprising because increased cavitation resistance is expected to lead to greater negative hydrostatic pressures and hoop stresses on vessel walls, necessitating higher inline image to resist collapse. On the other hand, developing excessively high inline image is presumably selected against as inefficient resource allocation, and it may reduce transport efficiency in roots (Fig. 7). Vessel implosion is a threshold-type response, and as long as inline image is above the implosion threshold (dashed line in Fig. 4c), vessels should be safe from implosion. Thus, the evolution of increased cavitation resistance need not necessitate an increase in inline image as long as the vessel pairs remain above the implosion threshold (Jacobsen et al., in press). The role of fibers in supporting vessels against implosion further complicates the inline image implosion resistance model, and more work is needed to more fully understand this relationship, a fact highlighted by the lack of correlation with PICs in both the present study and that of Jacobsen et al. (in press).

For roots, the cavitation resistance and mechanical trends were different from stems. There was no correlation between Ψ50 and xylem density among roots. This correlation is nearly universal among woody stems, and may represent a fundamental difference between root and stem xylem (Hacke & Sperry, 2001; Jacobsen et al., in press). Moreover, root resistance to cavitation and resistance to mechanical breakage were not correlated, as has been shown for stems (Jacobsen et al., 2005, in press). On the other hand, greater cavitation resistance was correlated with greater estimated vessel implosion resistance for raw data and PICs for roots. We also note that the measured estimates of vessel implosion resistance were generally greater or equal for roots compared with stems. This is unexpected because stems will typically experience more severe negative hydrostatic pressures than roots. These data suggest that in roots, vessel structure may be more important to resisting implosion than the surrounding fiber matrix as compared with stems. Roots are surrounded by the soil matrix and thus, for mechanical support of the plant body, may not require as robust a fiber matrix as found in stems. In stems, the fiber matrix performs the dual function of supporting the plant body against gravity and supporting vessels from imploding under stresses caused by negative hydrostatic pressures (however, see Jacobsen et al., 2007).

The MOR and MOE values were not significantly greater for stems than for roots, although stems have more fibers than roots. In addition, stems have more cell wall area (fiber and vessel) in cross-section than roots (R. B. Pratt, unpublished). The reason for comparable mechanical strength in roots may relate to bark differences in stems and roots. Roots generally have a thicker bark than stems (Esau, 1965), which may have contributed significantly to root mechanics, as has been shown in stems of other species (Niklas, 1999). In addition, parenchyma can significantly contribute to organ mechanical strength (Beery et al., 1983; Woodrum et al., 2003). Further work is needed to understand the structural basis for the relatively high root mechanical strength against bending stress.

In roots there was a clear trade-off between xylem transport efficiency and transport safety, whereas stems exhibited no such trade-off (i.e. stem Ψ50 and Ks were not correlated, and nor were inline image and Ks). The reason for this trade-off may relate to vessel resistance to implosion. Among roots, the fiber matrix does not provide as much support as in stems, and greater inline image may be necessary to avoid vessel collapse under negative hydrostatic pressures. Increasing vessel wall thickness relative to the lumen diameter (i.e. inline image) may limit lumen diameter size. As hydraulic efficiency is proportional to the lumen radius to the fourth power, as described by the Hagen-Poiseuille relationship, even small decreases in lumen diameter could result in considerable efficiency declines. An interesting independent test of this pattern could be to assess the safety vs efficiency trade-off in mechanical parasites, such as lianas, which presumably do not have as robust a fiber matrix as free-standing plants. We also do not rule out the possibility that the pit area hypothesis could explain the observed trade-off (see the Introduction; Wheeler et al., 2005; Hacke et al., 2006), but a test of this will require further study.

In contrast to our findings in roots, a recent study found no evidence for correlated evolution between Ψ50 and Ks in roots sampled across a broad range of 16 temperate forest tree taxa, including three conifers (Maherali et al., 2006). At the cellular level, Ks is influenced by conduit diameter and length, the pitted area of a conduit and the conductivity of pits. Cavitation resistance is determined by the largest diameter pore in a pit membrane adjacent to an air-filled area, the vessel pitted area and vessel implosion resistance. Large differences in these traits across sampled taxa may confound attempts to analyze trade-offs; thus, there is great value in comparing taxa with similar xylem anatomy (Maherali et al., 2006). The taxa sampled here are from a common lineage and have similar xylem anatomy, and our results suggest a safety vs efficiency trade-off in this lineage. Because of the numerous anatomical and physiological traits influencing Ψ50 and Ks, a range of relationships between Ψ50 and Ks might be expected in different lineages and across broader taxonomic scales (Maherali et al., 2004; Edwards, 2006).

Stems did not exhibit a correlation between percentage fiber wall area and Ks, suggesting no trade-off between efficiency and safety from mechanical stresses. In addition, there was no association between Ks and MOR here or in a previous study (Woodrum et al., 2003). Thus, in stems we find no evidence for a trade-off between transport efficiency and mechanical support.

The correlations between the percentages of cell types in cross-section and resistance to cavitation yield further insight into differences between stems and roots. Previous studies of stems have found a correlation between fiber mechanical strength (cross-sectional fiber wall area) and resistance to water stress-induced cavitation (Jacobsen et al., 2005, in press), suggesting a link between fiber properties and vessel resistance to negative hydrostatic pressures. We found that an increased percentage of fiber wall area in cross-section was not significantly correlated (after Bonferroni correction) with greater cavitation resistance (P = 0.08; data not shown). Roots also had no such correlation, suggesting a minimal role of fibers in vessel support. Therefore, in the present study there is no support for the hypothesis that increased fiber wall area is important in resisting negative hydrostatic pressures. It may not be necessary to increase the proportion of fiber wall area to resist negative pressure stress, and an alternative is that fibers may be differentially arranged to achieve adequate mechanical support.

The correlation between increased percentage vessel area and increased cavitation resistance in roots may suggest a functional relationship between these two traits; however, this is unlikely because of a lack of correlation with PICs. All things being equal, xylem transport efficiency increases with an increase in the area of vessels in cross-section. However, the chief advantage of increased vessel area for the species sampled here may be its correlation with more vessels per unit area (i.e. greater vessel redundancy) (R. B. Pratt, unpublished). Greater vessel redundancy helps to ensure sustained transport if transport fails in some vessels because of cavitation or other damage (Carlquist, 1977; Zimmermann, 1983). Thus, root xylem that is more cavitation resistant appears to be more redundant, making it simultaneously more resistant and tolerant to hydraulic failure. The lack of correlation with PICs suggests that these two traits evolved independently of one another. It would be useful to evaluate samples from a broader range of taxa to establish if these traits are evolutionarily correlated across broader taxonomic scales.

Reduced percentage parenchyma area was correlated with greater resistance to cavitation in roots, but not in stems. This relationship suggests that root transport safety comes at the cost of reduced parenchyma area. Living parenchyma cells have the disadvantage of respiring and thus adding to maintenance respiration; however, root parenchyma cells are important for storing starch and nutrients, as well as for differentiating into lateral roots. Thus, a higher degree of water stress tolerance may reduce the carbohydrate and nutrient storage capacity of roots, and reduce the ability of roots to develop lateral roots to exploit resource pulses opportunistically or grow following injury. Consistent with this, species in the present study that are nonsprouters after fire have the lowest proportion of root parenchyma, and presumably rely less on root storage than postfire sprouters (Fig. 6b; lowest percentage parenchyma areas for roots are all nonsprouters = Ccr, Ccu, Cm, Co).

None of the percentages of cells in cross-section were correlated with cavitation resistance when analyzed using PICs. From the existing literature, it is unclear how evolutionarily labile are cell-type percentages (Preston et al., 2006). A previous study of five maples also shows that the percentage of xylem cell types in stems is phylogenetically conserved (Woodrum et al., 2003). It is worth noting, in the present study, that for correlations among cell percentages that were significant with raw data also showed several significant correlations among the nine phylogenies using PICs, even though the average P-value was not significant. This indicates that the percentage cell area analyses were sensitive to phylogeny, and a fully resolved phylogeny is needed to assess the relationships between cell type percentages and cavitation resistance in greater detail.

A greater ability of stem xylem to store water is strongly linked with reduced mechanical support and transport safety. For nonsucculent stems, greater water storage capacity is increased primarily by increasing fiber and vessel lumen diameters (Desch, 1973). For fibers, greater lumen volume should lead to reduced xylem density and MOR. For vessels, greater lumen diameter may lead to increased susceptibility to vessel implosion unless there is a concomitant increase in vessel wall thickness. The importance of xylem water storage in woody plant stems for most species is unclear. However, it has been estimated that stored water can contribute from 10 to 50% of the daily transpiration in certain tree species (Gartner & Meinzer, 2005). Furthermore, the strong inverse correlations, shown here, between transport safety and capacitance, and the correlations between water storage and seasonal changes in water status, are suggestive of an important role for stem water storage (Stratton et al., 2000; Pratt & Black, 2006).

In summary, stems and roots have different xylem structural and functional relationships (summarized in Fig. 9). Stems are most strongly characterized by positive associations between mechanical strength and transport safety (i.e. resistance to water stress-induced cavitation). Stem mechanical strength vs safety relationships are found at the tissue (xylem density) and organ (MOE and MOR) levels, and are evolutionarily correlated. Stem mechanical strength appears to be important in maintaining xylem transport under negative hydrostatic pressure, probably as an aid to resist vessel implosion as well as resisting mechanical stresses caused by gravity. Roots show some important differences when compared with stems. Unlike stems, mechanical strength is not correlated with transport safety in roots. This difference between stems and roots probably relates to the contrasting mechanical environment of these two organs. This mechanical difference may also help to explain the relatively strong root transport safety vs efficiency trade-off because root vessels are not as well supported by a robust fiber matrix as are stems, and thus as root vessels become wider (more efficient) they become more susceptible to implosion (less safe).

Both stems and roots show a trade-off between transport safety and storage. For stems, this trade-off is solely related to water storage (capacitance) and not potential nutrient storage (percentage parenchyma). Roots show a trade-off between potential nutrient storage and transport safety, although this trade-off was not supported by PICs. Capacitance was not measured for roots. It would be valuable to study nutrient storage vs transport safety on a broader range of taxa to evaluate this potential trade-off in roots in greater detail.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

RBP and ALJ thank the California State University, Bakersfield Biology Department, for making available laboratory resources during the collection of significant data and writing/analysis for this project. Special thanks are owed to David Ackerly for valuable discussions. Three anonymous reviewers greatly improved the manuscript. The authors acknowledge grants NSF-RUI IBN-0130870 and NSF-REU DBI-0452587 to S. D. Davis and F. W. Ewers and NSF IBN-0131247 to F. W. Ewers. Marcia Murry, Madeleine Ewers, Ruchika Mohla, Dustin Long and Krystle Beglari are heartily thanked for help in root excavations.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Ackerly DD. 2000. Taxon sampling, correlated evolution, and independent contrasts. Evolution 54: 14801492.
  • Ackerly DD. 2003. Community assembly, niche conservatism, and adaptive evolution in changing environments. International Journal of Plant Science 164: 165184.
  • Ackerly DD. 2004. Adaptation, niche conservatism, and convergence: comparative studies of leaf evolution in the California chaparral. The American Naturalist 163: 654671.
  • Beery WH, Ifju G, McLain TE. 1983. Quantitative wood anatomy-relating anatomy to transverse tensile strength. Wood and Fiber Science 15: 395407.
  • Bolmgren K, Oxelman B. 2004. Generic limits in Rhamnus L. s.1. (Rhamnaceae) inferred from nuclear and chloroplast DNA sequence phylogenies. Taxon 53: 383390.
  • Brodribb TJ, Holbrook NM. 2005. Water stress deforms tracheids peripheral to the leaf vein of a tropical conifer. Plant Physiology 137: 11391146.
  • Carlquist S. 1977. Ecological factors in wood evolution: a floristic approach. American Journal of Botany 64: 887896.
  • Choat B, Ball M, Luly J, Holtum J. 2003. Pit membrane porosity and water stress-induced cavitation in four co-existing dry rainforest tree species. Plant Physiology 141: 4148.
  • Cochard H, Froux F, Mayr S, Coutand C. 2004. Xylem wall collapse in water-stressed pine needles. Plant Physiology 134: 401408.
  • Davis SD, Sperry JS, Hacke UG. 1999a. The relationship between xylem conduit diameter and cavitation caused by freezing. American Journal of Botany 86: 13671372.
  • Davis SD, Ewers FW, Wood J, Reeves JJ, Kolb KJ. 1999b. Differential susceptibility to xylem cavitation among three pairs of Ceanothus species in the Transverse Mountain Ranges of southern California. Ecoscience 6: 180186.
  • Desch HE. 1973. Timber: its structure and properties. London, UK: The Macmillan Press.
  • Edwards EJ. 2006. Correlated evolution of stem and leaf hydraulic traits in Pereskia (Cactaceae). New Phytologist 172: 479489.
  • Esau K. 1965. Plant Anatomy. New York, NY, USA: John Wiley and Sons, Inc.
  • Felsenstein J. 1985. Phylogenies and the comparative method. American Naturalist 125: 115.
  • Gartner BL, Meinzer FC. 2005. Structure–function relationships in sapwood water transport and storage. In: HolbrookNM, ZwienieckiMA, eds. Vascular transport in plants. San Diego, CA, USA: Elsevier, 564.
  • Hacke UG, Sperry JS. 2001. Functional and ecological xylem anatomy. Perspectives in Plant Ecology, Evolution and Systematics 4/2: 97115.
  • Hacke UG, Sperry JS, Pittermann J. 2000. Drought experience and cavitation resistance in six shrubs from the Great Basin, Utah. Basic and Applied Ecology 1: 3141.
  • 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: 457461.
  • Hacke UG, Sperry JS, Wheeler JK, Castro L. 2006. Scaling of angiosperm xylem structure with safety and efficiency. Tree Physiology 26: 689701.
  • Hardig TM, Soltis PS, Soltis DE. 2000. Diversification of the North American shrub genus Ceanothus (Rhamnaceae): conflicting phylogenies from nuclear ribosomal DNA and chloroplast DNA. American Journal of Botany 87: 108123.
  • Hargrave KR, Kolb KJ, Ewers FW, Davis SD. 1993. Conduit diameter and drought-induced embolism in Salvia mellifera Greene (Labiatae). New Phytologist 126: 695705.
  • Hellmers H, Horton JS, Juhren G, O’Keefe J. 1955. Root systems of some chaparral plants in southern California. Ecology 36: 667678.
  • Hickman JC. 1993. The Jepson manual: higher plants of California. Berkeley, CA, USA: University of California Press.
  • Jacobsen AL, Ewers FW, Pratt RB, Paddock WA, Davis SD. 2005. Do xylem fibers affect vessel cavitation resistance? Plant Physiology 139: 546556.
  • Jacobsen AL, Pratt RB, Ewers FW, Davis SD. (in press). Cavitation resistance among twenty-six chaparral species of southern California. Ecological Monographs.
  • Jacobsen AL, Agenbag L, Esler KJ, Pratt RB, Ewers FW, Davis SD. 2007. Xylem density, biomechanics and anatomical traits correlate with water stress in 17 evergreen shrub species of the Mediterranean-type climate region of South Africa. Journal of Ecology 95: 171183.
  • Kozlowski TT, Pallardy SG. 1996. Physiology of woody plants. San Diego, CA, USA: Academic Press.
  • Maherali H, Pockman WT, Jackson RB. 2004. Adaptive variation in the vulnerability of woody plants to xylem cavitation. Ecology 85: 21842199.
  • Maherali H, Moura CF, Caldeira MC, Willson CJ, Jackson RB. 2006. Functional coordination between leaf gas exchange and vulnerability to xylem cavitation in temperate forest trees. Plant, Cell & Environment 29: 571583.
  • Martins EP. 2004. COMPARE, v, 4.6b. Computer programs for the statistical analysis of comparative data. Bloomington, IN, USA: Indiana University (distributed by the author at http://Compare.Bio.Indiana.Edu/.Departmentofbiology).
  • Martins EP, Garland T. 1991. Phylogenetic analyses of the correlated evolution of continuous characters – a simulation study. Evolution 45: 534557.
  • Mason HL. 1942. [Ceanothus.] The distributional history and fossil record. In: Van RenssalaerM, McMinnHE, eds. Ceanothus. Santa Barbara, CA, USA: Santa Barbara Botanic Garden, 281303.
  • McGarigal K, Cushman S, Stafford S. 2000. Multivariate statistics for wildlife and ecology research. New York, NY, USA: Springer.
  • McMinn HE. 1942. Ceanothus: part II; a systematic study of the genus Ceanothus. Santa Barbara, CA, USA: Santa Barbara Botanic Garden.
  • Meentemeyer RK, Moody A. 2002. Distribution of plant life history types in California chaparral: the role of topographically-determined drought severity. Journal of Vegetation Science 13: 6778.
  • Niklas KJ. 1999. The mechanical role of bark. American Journal of Botany 86: 465469.
  • Pittermann J, Sperry JS. 2003. Tracheid diameter is the key trait determining the extent of freezing-induced embolism in conifers. Tree Physiology 23: 907914.
  • Pratt RB, Black RA. 2006. Do invasive trees have a hydraulic advantage over native trees? Biological Invasions 8: 13311341.
  • Pratt RB, Ewers FW, Lawson MC, Jacobsen AL, Brediger MM, Davis SD. 2005. Mechanisms for tolerating freeze-thaw stress of two evergreen chaparral species: Rhus ovata and Malosma laurina (Anacardiaceae). American Journal of Botany 92: 11021113.
  • Pratt RB, Jacobsen AL, Golgotiu KA, Sperry JS, Ewers FW, Davis SD. (in press). Life history type and water stress tolerance in nine California chaparral species (Rhamnaceae). Ecological Monographs.
  • Preston KA, Ackerly DD. 2003. Hydraulic architecture and evolution of shoot allometry in contrasting climates. American Journal of Botany 90: 15021512.
  • Preston KA, Cornwell WK, DeNoyer JL. 2006. Wood density and vessel traits as distinct correlates of ecological strategy in 51 California coast range angiosperms. New Phytologist 170: 807818.
  • Sauter JJ, Van Cleve B. 1994. Storage, mobilization and interrelationships of starch, sugars, protein and fat in the ray storage tissue of poplar trees. Trees-Structure and Function 8: 297304.
  • Sokal RR, Rohlf FJ. 1995. Biometery. New York, NY, USA: W.H. Freeman.
  • Stratton L, Goldstein G, Meinzer FC. 2000. Stem water storage capacity and efficiency of water transport: their functional significance in a Hawaiian dry forest. Plant, Cell & Environment 23: 99106.
  • Waring RH, Running SW. 1978. Sapwood water storage: its contribution to transpiration and effect upon water conductance through the stems of old-growth Douglas-fir. Plant, Cell & Environment 1: 131140.
  • 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 & Environment 28: 800812.
  • Woodrum CL, Ewers FW, Telewski FW. 2003. Hydraulic, biomechanical, and anatomical interactions of xylem from five species of Acer (Aceraceae). American Journal of Botany 90: 693699.
  • Zimmermann MH. 1983. Xylem structure and the ascent of sap. New York, NY, USA: Springer-Verlag.