Geographic And Seasonal Variation In Chaparral Vulnerability To Cavitation

2014·12.6 MB·English

by  Anna L Jacobsen

'g^THSOW^
DEC 1 7 2014
Madrono, VoL 61, No. 4, pp. 317-327, 2014
GEOGRAPHIC AND SEASONAL VARIATION IN CH?rPARgAL
VULNERABILITY TO CAVITATION
Anna L. Jacobsen* and R. Brandon Pratt
Department of Biology, California State University, Bakersfield, 9001 Stockdale Hwy,
CA
Bakersfield, 93311
^[email protected]
Stephen D. Davis
Natural Science Division, Pepperdine University, 24255 Pacific Coast Hwy,
CA
Malibu, 90263
Michael F. Tobin
Department of Natural Sciences, University of Houston Downtown, 1 Main St,
TX
Houston, 77002
Abstract
Resistance ofstem xylem to water stress-induced cavitation and embolism among chaparral shrub
species in California has been extensively studied, providing the opportunity to examine broad
patternsincavitation resistance. Weusedpreviouslypublished aswell as unpublished vulnerabilityto
cavitationcurvedatafrom 16chaparralshrubspeciesofsouthernCaliforniatoexaminethevariability
ofcavitation resistance across sites, regions, and seasons. Additionally, these data provided a unique
opportunityto address a recentmethodological debatewithin the field ofplanthydraulics. We found
that different methods, specifically a centrifugemethod and a dehydration method, produced similar
results (P = 0.184). Vulnerability to cavitation varied seasonally, with species exhibiting greater
susceptibility to water-stress induced cavitation during the wet season (P = 0.003). Cavitation
resistancedidnotdifferamongsitesthatwerelessthan 10kmaparteventhoughthesesitesdifferedin
theircoastalexposure,precipitation,andtemperatures(P = 0.476). However,acrosslargergeographic
distancesandwithincreasedclimaticdivergence, cavitationresistancesignificantlyvaried (P = 0.005),
withpopulations from a higherrainfall mountain rangeexhibitinggreater susceptibilitytocavitation.
These data suggest that species may be particularly susceptible to the onset ofearly summer drought
before xylem has hardened. Variation in cavitation resistance may be limited locally, but broadly
dispersedspeciesmaydivergeincavitationresistanceacrosstheirrange. Maintainingpopulationsthat
vary in cavitation resistance may be an important component ofspecies conservation planning in an
era ofincreased climatic variability.
Key Words: Cavitation resistance, chaparral, drought, embolism, water potential, water relations,
water stress, xylem.
Cavitation and subsequent embolism of xy- sites and during dry years may lead to intraspe-
lem conduits can occur when plants experience cificvariation amongpopulations in some species
water stress (Davis et al. 2002; Tyree and (Mencuccini and Comstock 1997; Kavanagh et
Zimmermann 2002) and results in reduced al. 1999; Kolb and Sperry 1999b; Pratt et al.
hydraulic transport efficiency. This may be 2012), although previous studies have also found
particularly harmful during periods of extreme that some climatically divergent populations do
or protracted drought when catastrophic hy- not vary greatly in cavitation resistance (Men-
draulic failure may lead to plant dieback and cuccini and Comstock 1997; Matzner et al. 2001;
death (McDowell et al. 2008). Cavitation- Stout and Sala 2003; Lamy et al. 2011). Species
induced hydraulic failure has been linked to may also be phenotypically plastic in their xylem
plant dieback and mortality during both long- vulnerability to cavitation, although this has
and short-term droughts (Rice et al. 2004; been relatively little studied (Holste et al. 2006;
Anderegg et al. 2013; Paddock et al. 2013; Pratt Beikircher and Mayr 2009; Mayr et al. 2010;
et al. 2014) and when plants were experimentally Fichot et al. 2010; Awad et al. 2010; Plavcova
exposed to water stress (Pratt et al. 2008). and Hacke 2012).
Xylem cavitation resistance is an important Co-occurring species of chaparral shrubs in
plant functional trait that varies among ecosys- southern California exhibit highly divergent
tems (Maherali et al. 2004; Choat et al. 2012) levels of cavitation resistance (Davis et al. 1999;
and, at smaller scales, among plant communities Jacobsen et al. 2007a; Pratt et al. 2007b) and
(Jacobsen et al. 2007b; Hacke et al. 2009). For include some of the most cavitation resistant
species with broad distributions, selection at drier angiosperms ever measured (Maherali et al.
MADRONO
318 [Vol. 61
Fig. 1. Vulnerability to cavitation curves were analyzed from species occurring in southern California that had
beenmeasuredatmultiplelocations. Dataweredividedintotworegions,eithertheSantaMonicaMountainsorthe
San Gabriel Mountains. Within the Santa Monica Mountains, vulnerability curves were further divided into
subpopulations sampled at sites located nearthecoastand those sampled frominland sites. Selectedhighways and
cities from the study area are indicated on the map and themean annual precipitation forthe sites and regions (as
reported in the previously published studies included in Table 1) are shown.
2004). However, large scale mortality ofchapar- and Davis 1996; Davis et al. 1999a, 2002;
ral shrubs has been reported during extreme Jacobsen et al. 2005, 2007a, b; Pratt et al.
drought years, particularly at a desert-chaparral 2007b), which enabled us to evaluate the impact
ecotone (Paddock et al. 2013). of differences between sites and regions on
Chaparral shrubs exhibit multiple life history vulnerability to cavitation within repeatedly
types, largely defined by their post-fire response. measured species ofthis plant community. Using
This includes species that resprout from under- previously published and some previously un-
ground storage structures following fire, species published data we examined if cavitation resis-
that recruit post-fire through the germination of tance ofchaparral shrubs 1) varied locally among
fire-cued seeds, and species that employ both of shrub subpopulations at sites that were located in
these strategies. These differential life history close proximity but differed in their ocean
types are related to large functional differences exposure (coast vs. inland) and 2) varied region-
between some chaparral shrub species at both the ally between shrub populations from two moun-
adult (Jacobsen et al. 2007a; Pratt et al. 2007a, b) tain ranges occurring in southern California
and seedling (Pratt et al. 2008, 2010, 2012) stages. (Santa Monica Mts. vs. San Gabriel Mts.)
Furthermore, drought-induced mortality ofseed- (Fig. 1). We predicted that populations from
lings may be common, particularly among post- coastally exposed sites would be less cavitation
fire seeding species whose seedlings must survive resistant than inland sites, due to the stronger
the protracted summer dry period that occurs in maritime climatic influence along the coast
the Mediterranean-type climate region of south- (Vasey et al. 2012). We predicted that popula-
ern California (Frazer and Davis 1988; Thomas tions from the Santa Monica Mountains would
and Davis 1989). Selection for higher cavitation be more cavitation resistant than the San Gabriel
resistance may therefore be expected at drier and Mountains due to higher mean annual precipita-
more interior sites, where mortality has been tion occurring in the San Gabriel Mountains,
observed, compared to species occurring in more consistent with previous studies examining the
mesic or coastal sites that experience less water impact ofprecipitation orwatering treatments on
stress. These patterns may become more pro- cavitation resistance (Mencuccini and Comstock
nounced across broader, regional scales. 1997; Kolb and Sperry 1999b; Helms 2009; Awad
The cavitation resistance of chaparral shrubs et al. 2010).
has been extensively studied, particularly in This analysis was complicated by the recent
southern California (Jarbeau et al. 1995; Redfeldt suggestion that some methods used to construct
2014 JACOBSEN ET AL.: CHAPARRAL CAVITATION RESISTANCE 319
]
vulnerability to cavitation curves may not pro- then branches were cut underwater, alternately at
m
duce reliable data (Choat et al. 2010; Cochard et each end, to excise a 0.10-0.14 central stem
al. 2010). In particular, it has been suggested that segment for measurement. This sampling proto-
centrifuge-based data may produce a measure- col reduces the negative pressure in the xylem
ment artifact, especially in species with long prior to the extraction of the measured sample,
vessels. The dataset compiled for the present which maybeimportant in orderto avoid artifoct
study provided an ideal test of this suggestion, when sampling some species (Wheeler et al.
because both dehydration and standard centri- 2013). Hydraulic conductivity (Kh) of stem
fuge curves have been used to construct vulner- segments was measured both immediately fol-
abilitycurves on the samechaparral species atthe lowing excision ofthe stem segment and follow-
same sites and during the same season. Addi- ing a one hour flush at 100 kPa using a degassed
tionally, many chaparral shrub species have long and ultra-filtered solution (Jacobsen et al. 2007a).
maximum vessels, including several species with The percentage loss in hydraulic conductivity
maximum vessel lengths greater than one meter (PLC) was determined for each stem segment
We
in length (Jacobsen et al. 2012). also (Sperry et al. 1988).
evaluated the impact of the season (wet vs. dry) Centrifuge vulnerability curves were construct-
during which curves were constructed, because ed using the methods described in Pratt et al.
m m
this has previously been reported to impact (2007a). Stem segments 0.14 or 0.27 in
vulnerability to cavitation (Kolb and Sperry length were excised, underwater, from larger
1999a; Jacobsen et al. 2007b). branches that had been harvested in the field.
Segments were then flushed as described above
Methods before being subjected to increasing negative
water potentials by being spun in a custom
Previously published vulnerability curves con- centrifuge rotor (Alder et al. 1997; Tobin et al.
ducted on chaparral shrub species from the Santa 2013). Hydraulic conductivity was measured
Monica Mountains and the San Gabriel Moun- following each spin and these were used to
tains of southern California were compiled calculate the PLC at each imposed water
(Jarbeau et al. 1995; Redfeldt and Davis 1996; potential relative to the initial flushed value.
Davis et al. 1999a; Davis et al. 2002; Jacobsen Vulnerability curves, whether generated using
et al. 2005, 2007a, b, Pratt et al. 2007b). Pre- dehydration or centrifuge based data, were
viously unpublished vulnerability curves mea- plotted asthePLCwithdecliningwaterpotential.
sured by the authors from 2004-2012 were also All data fromprevious studies were replotted and
included in analyses (Table 1). This included data reanalyzed for the current study. Water potential
for 16 species forwhichcurves were reported that and PLC data from each curve and study were fit
had been measured on the same species across using a Weibull curve (Microsoft Excel 2010,
multiple sites (coastal vs. inland within the Santa Microsoft, Redmond, WA). This curve was then
Monica Mts.), regions (Santa Monica Mts. vs. used to calculate the water potential at 50% loss
San Gabriel Mts.), and/or seasons (wet vs. dry) in hydraulic conductivity (P50) for each curve.
and using two different methods (dehydration vs. The P50 value was used as an estimate ofspecies
centrifuge) (Fig. 1). Wet season vulnerability cavitation resistance, as this is the most com-
curves were defined as those that were measured monly compared and reported value in species
between December and May for any sample year comparisons (Choat et al. 2012). Additionally,
across all ofthe studies listed above, when plants the alpha (shape parameter), which determines if
were hydrated in the field (Jacobsen et al. 2008). a curve is concave or convex, and the beta (scale
Rainfall typically occurs during November to parameter), whichdetermines the “stretch” ofthe
April at these sites and regions, with plant water curve along the x-axis, for the Weibull fit were
status declining in July following the use of soil recorded. These fit parameters were compared
moisture reserves (Jacobsen et al. 2008). Dry later, ifwe found that there was a significant shift
season vulnerability curves were defined as those in the P50 between our comparison groups, to
that were measured during July to October when determine if the change in cavitation resistance
plants are dehydrated in the field and were no was due to a shift in the shape or scale of the
longer growing (Jacobsen et al. 2007b). vulnerability to cavitation curve.
Dehydration vulnerability curves were con- Differencesin the P50 ofspecies as impacted by
structed using the methods described in Jacobsen site, region, season, or method were analyzed
et al. (2007a). Briefly, large branches, longer than using paired t-tests with data paired by species
the longest vessels as determined via air injection, (Release 16.1.0, Minitab, State College, PA). For
were collected in the field. Branches were allowed these comparisons, data were matched by the
to dehydrate for varying periods of time in the parameters that were not being evaluated in the
laboratory before being tightly bagged overnight current comparison and a single pair ofmean P50
to allow them to equilibrate. Water potentials of was generated for each species (e.g., the site
branches were measured the next morning and comparisons were conducted on species data
M " s
MADRONO
320 [Vol. 61
cIwO(z^[iaUoTlV ^’1«uOcS^uc3/Hl A^^W+•<<Vua£^<'1-u<uk' I5g(<j>DSD_U'^^4WsOmd; d.dS^OOQddd0>2!nn3 -<O^Go:«0COGGN,!3O-. £^t3g|wd31.2^.\oO^«dGd>.^2ov2._-^Oo[OaoG0S0dGdgi33--. a(sO^0Cd^N3'--^'_'adGaSd_dl-d^OaSH0ddH^N^Gi3r, a a •-QG^dGZ* d3o+3dG-^<^'.5aQiTGdoad.0-daOj.3iS(TN) ^02( t20G^a—vI 4o2dD1-»aa-Hv ooI[2TG(). «accCddO Xr2-)(. ,drb •2aoo<aemddp9aG.aNt". dOO,^^3^0OrGd0d2Nn3b3 XSdoo^meaZd't •2X,’a^GPc&G5agd XXOrG«G0d^d0a'3 XXX^ddG&;’=-C(aoSdd:0d8N:3 ^a"X;dd«Ga'_5Sdoa^fO^-:aSddG—daGd8Ndgi^^^aadadd0MSg3'!3.SaOO^O^oSod-8Gd<'d'.a-ddcGaud2G
O *— -S
0^ZOS_^Z^0) ^.‘crscoc3c/«T-oO(-^g^SN3 aO0p0nIu. +1 +1 ira0+n^01 0qda!c+>-ni; 0aiddd+-0-n_1 \T0q"01 a0O(rd+NNn-1 mdrd+--1 d(Od(a+-NNN1 rm0drd+--1 000dad+000-1 0(0drd+-N1-
w
S
^^. •aocC/i*g''-«o «a .2^5 -H r-H (M -H
« -o
£Q <U fi C^
^I a 5 s ^
<J ’"<osu,fwO. fS^ol'i, ^r—O^ m m mroT-HfNlin CNjinrsd-
?3 S G 0) fSi (N ro (N
^< aS-2^ ba2D2a^
u o^'Sa^o0i0.2c3G33 ro fNl m ro ro m mmcNrom —ncNd'cn
2 'o
wQ (^gU p ZCO -H O O O ro (N d- O <N
aI oS! "^d '<5u
c« a3= 8a a
0u0 „ ’E^§•c§^ <N O (N O O O O ’— o
D^ 3cG^ 2 a(SD
'd
Z >c 3
2H< (uM-U P«^ d- ^
H
>
u< a.M0 d «^ LL
o XG3 .a 3
H 0c -0G03 013 a
-•0gd3 .00a33 ’CdVO cfa0c <bO uo u ua d G ad Xa ^ q
<w "o d X-0o3
z ]> 'd
J S u a
>ZU0). |a0^31a ’GO 003 I >2 zG EGocn8d0^:3 waS ^d d .0d0233 Sbbffll
z-E1g2z<C--QJ^232^^-GaG<>13 d2GC-^Ga3^"G3D3 i^S-sGn--IIGudd -0^Z^ 1sa§ ^a8-a aGSSs: qJ0d0dC33 XI“^^<sG^GGGi3:ws.w3ddG0ydy3.,~o1C22G0§O5E00dGadPc-33d "<GGK5G^sGj 0d0Odaad33 Xdd0Cd0333 <G^S^aIJfd0d00e333 "~a§G§ssSG Ea^addd0033 |soId tG ^S-qad.tZgiiQMi-<SdnadT 13 '<I^S§fg<To0OO3
2014] JACOBSEN ET AL.: CHAPARRAL CAVITATION RESISTANCE 321
Fig. 2. Vulnerability to cavitation curves for 16 chaparral species, showingthe increase in the percentage loss in
hydraulic conductivity with decreasing pressure. Curves represent pooled data from all reported vulnerability
curvesforthesespeciesfromtheSantaMonicaMountainsand SanGabrielMountainsofsouthernCalifornia. The
shown curve is a Weibull curve fit to pooled data for each species. See Table 1 for species abbreviations.
matched for season and method, the season Results
comparison was conducted on data matched
for site and method, and the method comparison Sixteen chaparral shrub species varied greatly
was conducted on data matched for site and in their resistance to cavitation (Fig. 2). When
season). It was not possible to use a more pooled across studies, species cavitation resis-
complicated model to analyze these data due to tance (as estimated by the water potential at 50%
a lack of replication for many species and loss in hydraulic conductivity; R50) ranged from
comparisons (Table 1). In most cases, this lack —0.7 MPa to —8.32 MPa (Table 1), with values
ofreplication was most commonly due to limited as high as —0.04 MPa and as low as —10.48 MPa
species range, for instance some species occur for mean species values from individual studies.
only along the coast or only within the Santa The use of different methods (i.e., centrifuge-
Monica Mountains. based curves vs. dehydration-based curves) did
To further examine the influence of site, not result in a significant difference in P50
region, season, and method on cavitation resis- (Fig. 3A, B; P = 0.184, T = -1.44, n = 10
tance, we plotted species pairs ofP50 as theywere matched species pairs).
calculated above. These datawere analyzed using The season in which species were sampled
SMA regression (R Statistical Software, SMATR significantly impacted P50 (Fig. 3C, D; P =
package, Warton et ah 2006). We compared 0.003, T = —3.70, n = 12 matched species pairs).
whether the slopes of these regressions were This difference was due primarily to a shift in the
different than one and whether the intercept scale ofthe vulnerabilitycurve (betaparameter; P
was different from zero as an additional way to = 0.002, T = 3.93) rather than a change in the
examine if site, season, region, or method shape ofthe curve (alpha parameter; P = 0.053,
influenced chaparral shrub cavitation resistance. T = 2.17). Vulnerability curves measured during
MADRONO
322 [Vol. 61
Fig. 3. Vulnerability to cavitation curves pooled across species ofchaparral shrub species as they vary with the
method used to construct curves (A, B), the season of measurement (C, D), the site from which species were
sampled(E, F), ortheregionthatwassampled(G, H). VulnerabilitytocavitationcurvesareshowninpanelsA, C,
E, and Gand themean ±1 SEpressureat 50% lossinhydraulicconductivity(Pso) is showninpanels B, D, F, and
H. Then indicates the numberofspeciesforwhichvulnerabilitycurveswere sampledusingdifferentmethodsorin
different seasons, sites, or regions and different lowercase letters indicate significant differences.
2014] JACOBSEN ET AL.: CHAPARRAL CAVITATION RESISTANCE 323
Wet Season (MPa)
Fig. 4. Vulnerability to cavitation estimated as the pressure at 50% loss in hydraulic conductivity (P50) for
chaparralshrubspeciesasitvariedbyvulnerabilitycurvemethod(A),measurementseason(B), site(C), andregion
(D). For each regression, species data were matched for all ofthe parameters not included in a given comparison
(e.g., species data for vulnerability curve methods were matched for season, site, and region) and a species mean
was calculated from the matched data. Data were fit using SMA regression and the slope and intercept for
regressions are shown in each panel along with the 95% confidence intervals for these parameters in parentheses.
The dashed line in each panel is the 1:1 line. Species abbreviations are listed in Table 1.
the wet season had ^50 that were approximately ly 3.1 MPa more vulnerable to cavitation
1.4 MPa more vulnerable to cavitation than P50 compared to populations of the same species
measured from the same subpopulation during occurring in the Santa Monica Mountains
the dry season (Fig. 3D). (Fig. 3H).
Within a single mountain range, the Santa Across all comparisons (method, season, site,
Monica Mountains of southern California, P50 and region), the slopes of regressions were not
did not vary among subpopulations of species different from one in any case (P > 0.05 for all;
occurring along the coast versus those occurring see slope and intercepts with 95% confidence
inland (Fig. 3E, F; P = 0.476, T = -0.76, n = 7 intervals in Fig. 4). Similarly, the intercepts of
matched species pairs). regressions were not different from zero in any
Across larger geographic distances, the P50 of case (Fig. 4; P > 0.05 for all).
populations occurring in the Santa Monica
Mountains differed significantly from those Discussion
occurring in the San Gabriel Mountains
(Fig. 3G, H; P - 0.005, T = -4.04, n = 8 Methodological Considerations
matched species pairs). This differencewasdue to
a change in the shape of the vulnerability curve We found that both centrifuge and dehydra-
(alpha parameter; P = 0.004, T = 4.16), rather tion based vulnerability curves produced similar
than a shift in the scale ofthe vulnerability curve estimates of cavitation resistance. This is consis-
(beta parameter; P = 0.553, T = 0.62). Vulner- tent with several recent studies that have shown
ability curves measured on populations occurring that these methods produce similar results
in the San Gabriel Mountains were approximate- (Jacobsen and Pratt 2012; Sperry et al. 2012;
MADRONO
324 [Vol. 61
Tobin et al. 2013). Some prior reports of 1.4 m (Ro), 1.1m (Fc), and 1.2 m (Ri) (Jacobsen
disagreement between methods may be due to et al. 2012; See Table 1 for species codes), which
the use of a different centrifuge technique would clearly have vessels open through the
(Cochard et al. 2010) or the confounding of measured samples. There was also methodolog-
technique comparisons with the pooling of data ical agreement between species with shorter
generated from both flushed and non-flushed maximum vessel lengths, including 0.3 m (Af),
m m
curves (Cochard and Delzon 2013). Dehydration 0.5 (Cm), and 0.3 (Cs) (Jacobsen et al. 2012;
curves are most often based on percentage loss in See Table 1 for species codes). Additionally,
conductivity (PLC) obtained by taking initial, or when matched for region and season, we found
native, measurement of conductivity (K^) fol- good agreement between dehydration and centri-
lowed by the flushing of samples to measure the fuge curves, including for species with r-shaped
maximum potential conductivity (Ksmax) (howev- curves. Consistent with the findings of previous
er, see Jacobsen and Pratt 2012). These curves studies (Sperry et al. 2012; Christman et al. 2012;
would not be expected to be comparable to Tobin et al. 2013), this suggests that these curves
centrifuge curves generated using non-flushed describe a genuine species strategy. This finding
stems, because non-flushed curves begin at the highlights the diversity of hydraulic strategies
native rather than the .^smax (see Sperry et al. employed by plants, including many chaparral
2012). This results in non-flushed centrifuge shrub species.
curves exhibiting a flat initial portion of their
curve, until the point that the centrifuged Seasonal Effects
pressure exceeds the native pressure of the
samples, which often results in a sigmoidal in Cavitation resistance differed depending on the
shape. Non-flushed centrifuge curves may dras- season during which measurements were con-
tically differ in shape compared to flushed curves ducted. Across species there was a general shift
(Sperry et al. 2012). Thus, the agreement of toward increased resistance from the wet season
methods in the present study may be at least to the dry season. This isconsistentwith ageneral
partially due to the careful matching of samples “hardening” of the xylem as plants reduced
and the inclusion of only flushed-sample centri- growth, produced latewood, and transitioned
fuge curves. into the dry season during which they experience
Another issue ofrepeated concern with centri- lower water potentials. This same pattern has
fuge-based curves has been the influence ofopen been described previously in arid and semi-arid
vessels (i.e., vessels that are longer than the plant communities (Kolb and Sperry 1999a;
measured sample length and therefore do not Jacobsen et al. 2007b; Helms 2009). Changes in
contain a terminal vessel element within the xylem function seasonally have also been de-
sample) (Choat et al. 2010; Cochard et al. scribed for other hydraulic traits and in other
2010). This issue has been experimentally exam- plant communities, including large seasonal
ined using both alteration ofthe number ofopen changes in hydraulic conductivity (Sperry et al.
vessels (Jacobsen and Pratt 2012) and short and 1987; Tibbetts and Ewers 2000; Jacobsen et al.
long vesselled comparisons (Sperry et al. 2012; 2007b; Choatet al. 2010). This suggeststhatplant
Tobin et al. 2013) and open vessels have not been hydraulics may be quite variable intra-annually
found to impact vulnerability curves when the and that measurements of plant hydraulics
standard Alder et al. (1997) centrifuge technique conducted at a single sampling time may not
is used. Additionally, this issue has been tied to present a complete picture of the hydraulic
the shape ofvulnerability curves, with particular strategy of a species. Additionally, measurement
concern raised about “r” shaped (or exponential of PLC or conductivity across seasons may not
shaped) vulnerability curves (Cochard and Del- be comparable to measurements, such as vulner-
zon 2013), although careful study of r-shaped ability curves, that are measured at a single time
curve species, including single vessel air injection point. Finally, increased resistance later in the
measures, haveconfirmed thevalidity ofr-shaped season suggests that plants may be particularly
curves (Sperry et al. 2012; Christman et al. 2012; sensitive to early season drought and this may be
Tobin et al. 2013). an important area ofresearch in predicting plant
Chaparral species exhibit a wide range in the responses to more variable climate patterns.
shape and scale of their vulnerability curves and
are also very divergent in the length of xylem Geographic Variability
vessels found among species, providing the
opportunity for us to evaluate both the impact Across shorter distances and between subpop-
oflong vessels on curves as well as the validity of ulations within a single mountain range (i.e.,
vulnerability curves of differing shapes. Centri- sites), we did not find that cavitation resistance
fuge and dehydration based vulnerability curves varied even though both the mean annual
were not different across species. This included precipitation and temperature differed among
species with maximum vessel lengths as long as coastal and inland sites. This is consistent with
2014] JACOBSEN ET AL.; CHAPARRAL CAVITATION RESISTANCE 325
recent research on a single chaparral lineage, most climate change scenarios and patterns of
ArctostaphylosAdans., that found thatcavitation precipitation are expected to change, although
resistance did not differ between foggy coastal California topography makes fine-scale predic-
sites and non-foggy inland sites along the central tions difficult (Cayan et al. 2008). Thus, preser-
coast of California (Jacobsen and Pratt 2013), vation of hydraulic trait diversity among chap-
even though the water potentials of plants at arral species and populations may be key to long-
these sites differed (Vasey et al. 2012). It may be term resilience of chaparral shrub species and
that across relatively short distances (<10 km), communities.
enough genetic material is exchanged to prevent
subpopulations from diverging in response to Conclusions
local climatic conditions. This may also partially
explain the high mortality observed among some Chaparral species vary in their cavitation
species during drought where they occur at the resistance seasonally and among populations that
drier end oftheir range (Paddock et al. 2013). occurred in regions with varying mean annual
Across larger distances and among disjunct precipitation, suggesting that the timing ofwater
populations occurring within two mountain stress and the traits of different populations are
ranges, cavitationresistance significantlydiffered. both important factors likely to influence the
Populations occurring in the San Gabriel Moun- long-term resilience of chaparral communities.
tains, which receive considerably more mean Additionally, chaparral species are highly vari-
annual precipitation (855 mm) than the Santa able in their vulnerability to cavitation, including
Monica Mountains (460 mm), were more vulner- the presence of some highly vulnerable species
able to cavitation. This was consistent with our that exhibit r-shaped '(or exponential shaped)
prediction and with a previous study that vulnerability curves as well as more resistant
compared two chaparral shrub species between species. More research is needed to more fully
these two regions (Helms 2009). This is also understand these very different, yet co-occurring,
consistent with previous studies that have de- hydraulic strategies.
scribed populations from more mesic areas as
being more vulnerable to cavitation in some Acknowledgments
species (Mencuccini and Comstock 1997; Kava- RBPthanksNSF (IOS-0845125) and theAndrewW.
nagh et al. 1999; Kolb and Sperry 1999b) as well Mellow Foundation for support of this project. ALJ
as findings from shrub species occurring in other thanks NSF (lOS-1252232).
Mediterranean-type climate regions (Pratt et al.
2012). Additionally, the San Gabriel Mountains Literature Cited
sites experience winter freezing and freeze-thaw Alder, N. N., W. T. Pockman, J. S. Sperry, and S.
events that may impact the hydraulics of plants Nuismer. 1997. Use of centrifugal force in the
from the San Gabriel Mountains (Davis et al. study ofxylemcavitation. Journal ofExperimental
1999b; Cordero and Nilsen 2002), although it is Botany 48:665-674.
not clear the impact this may have on the present Anderegg, W. R. L., L. PlavcovA, L. D. L. Ander-
comparison because freezing also occurs in some EGG, U. G. Hacke,J. A. Berry,andC. B. Field.
sites within the Santa Monica Mountains (Pratt 2013. Drought’s legacy: multiyear hydraulic dete-
rioration underlies widespread aspen forest die-off
esthaol.wn200t5o). iFimnpaalclyt, svouillnperroapbeirltiiteys htaovecaavlsiotabteieonn and portends increased risk. Global Change
Biology 19:1188-1196.
(Sperry and Hacke 2002) and it should be noted Awad, H., T. Barigah, E. Badel, H. Cochard, and
that soil composition was different between these S. Herbette. 2010. Poplar vulnerability to xylem
regions, with the Santa Monica Mountains cavitation acclimates to drier soil conditions.
composed predominantly ofsedimentary derived Physiologia Plantarum 139:280-288.
soils and the San Gabriel Mountains composed Baldwin, B. G., D. H. Goldman, D. J. Keil, R.
predominantly ofplutonic derived soils (Schoen- Patterson, T. J. Rosatti, and D. H. Wilken,
herr 1992). (eds.). 2012. TheJepson Manual: vascularplantsof
California, 2nd ed. University ofCalifornia Press,
Berkeley, CA.
Hydraulic Trait Diversity Beikircher, B. and S. Mayr. 2009. Intraspecific
differences in drought tolerance and acclimation in
It is not clear whether these differences hydraulics of Ligustrum vulgare and Viburnum
represent phenotypic plasticity in xylem function lantana. Tree Physiology 29:765-775.
worouilfdtrheeqyuirreepersetsaebnltisghemneentticofdiafcfeoremnmceosn, gaanrddeint CayaTny,reDe.,Ra.,ndE.KP.. MHaauyrheore,.M2.00D8.. DCelitmtaitnegecrh,anMg.e
scenarios for the California Region. Climate
with plants from both regions to determine.
Change 87:S21-S42.
However, these data suggest that there is Choat, B., W. M. Drayton, C. Brodersen, M. A.
intraspecific variability in cavitation resistance Matthews, K. A. Shackel, H. Wada, andA. J.
among these chaparral shrub species. California McElrone. 2010. Measurementofvulnerabilityto
climate is expected to warm significantly under water stress-induced cavitation in grapevine: a
MADRONO
326 [Vol. 61
comparison of four techniques applied to long- response to differenceenvironmental effects. Plant,
vesseled species. Plant, Cell & Environment Cell & Environment 29:836-843.
33:1502-1512. Jacobsen, A. L. 2008. Comparative community
Del,zS.onJ,anRs.enB,haT.skJa.rB,roSd.rJi.bBbu,ccHi.,CTo.chS.arFdie,ldS,. pahmyosnioglotghyr:eensoenmcio-navriedrgsehnrcueb icnomwmautneirtierse.latNieonws
S. M. Gleason, U. G. Hacke, A. L. Jacobsen, F. Phytologist 180:100-113.
Lens, H. Maherali, J. Martinez-Vilalta, S. . 2013. Vulnerability to cavitation of central
Mayr, M. Mencuccini, P. J. Mitchell, A. California Arctostaphylos (Ericaceae): a new anal-
Nardini, j. Pittermann, R. B. Pratt, J. S. ysis. Oecologia 171:329-334.
Sperry, M. Westoby, I. J. Wright, and A. E. F. W. Ewers, R. B. Pratt, W. A. Paddock
Zanne. 2012. Global convergence in the vulnera- III,, AND S. D. Davis. 2005. Do xylem fibers affect
bility offorest to drought. Nature 491:752-756. vessel cavitation resistance? Plant Physiology
Christman, M. A., J. S. Sperry, and D. D. Smith. 139:546-556.
2012. Rare pits, large vessels and extreme vulner- AND R. B. Pratt. 2012. No evidence for an
ability to cavitation in a ring-porous species. New open vessel effect in centrifuge-based vulnerability
Phytologist 193:713-720. curvesofalong-vesselledliana{Vitis vinifera). New
CoCHARD, H. AND S. Delzon. 2013. Hydraulic failure Phytologist 194:982-990.
andrepairarenotroutineintrees.AnnalsofForest , , S. D. Davis, and F. W. Ewers.
Science 70:659-661. 2007b. Cavitation resistance and seasonal hydrau-
S. Herbette, T. Barigah, E. Badel, M. lics differ among three arid Californian plant
Enn,ajeh, AND A. ViLAGROSA. 2010. Does sample communities. Plant, Cell & Environment 30:
length influence the shape of xylem embolism 1599-1609.
vsuplinnenrianbgilitteychnciuqruvee.s?PlAantt,estCelwlit&h tEhneviCraovnimternotn 200,7a. Cavit,atF.ionW.resEiwstearnsce, aamnodngS.26D.chaDpaavrirsa.l
33:1543-1552. species of southern California. Ecological Mono-
Cordero, A. R. and E. T. Nilsen. 2002. Effects of graphs 77:99-115.
hsyudmrmauelricdrcoonudguhcttivaintdy woifntRehrodfordeeeznidnrgononspesctieems Ewe,rs. 201,2M..AF.Tgolobbianl,Uan.alGy.siHsacokfex,yalenmdvFe.sWse.l
from contrasting climates. Tree Physiology 22: length in woody plants. American Journal of
919-928. Botany 99:1583-1591.
Davis, S. D., F. W. Ewers, J. S. Sperry, K. A. Jarbeau, j. a., F. W. Ewers, and S. D. Davis. 1995.
PoRTwooD, M. C. Crocker, and G. C. Adams. The mechanism of water-stress-induced embolism
2002. Shoot dieback during prolonged drought in in two species of chaparral shrubs. Plant, Cell &
Ceanothus chaparral of California: a possible case Environment 18:189-196.
of hydraulic failure. American Journal of Botany Kavanagh, K. L., B. j. Bonds, S. N. Aitken, B. L.
89:820-838. Gartner, and S. Knowe. 1999. Shoot and root
, , J. Wood, J. J. Reeves, and K. J. vulnerability to xylem cavitation in four popula-
Kolb. 1999a. Differential susceptibility to xylem tions of Douglas-fir seedlings. Tree Physiology
cavitation among three pairs of Ceanothus species 19:31-37.
in the Transverse Mountain Ranges of southern Kolb, K. J. and J. S. Sperry. 1999a. Transport
California. Ecoscience 6:180-186. constraints on water use by the Great Basin shrub,
, J. S. Sperry, and U. G. Hacke. 1999b. The Artemisia tridentata. Plant, Cell & Environment
relationship between xylem conduit diameter and 22:925-935.
cavitation caused by freezing. American Journal of . 1999b. Differences in drought adaptation
Botany 86:1367-1372. between subspecies ofsagebrush {Artemisia triden-
Fichot, R., T. S. Barigah, S. Chamaillard, D. Le tata). Ecology 80:2373-2384.
Thiec, F. Laurans, H. Cochard, and F. Lamy, J.-B., L. Bouffier, R. Burlett, C. Plomion,
Brignolas. 2010. Common trade-offs between H. CocHARD, AND S. Delzon. 2011. Uniform
xylem resistance to cavitation and otherphysiolog- selection as a primary force reducing population
ical traits do not hold among unrelated Populus genetic differentiation of cavitation resistance
deltoides X Populus nigra hybrids. Plant, Cell & across a species range. PLoS ONE 6:e23476,
Environment 33:1553-1568. doi:10.1371/journal.pone.0023476
Frazer, J. M. and S. D. Davis. 1988. Differential Maherali, H., W. T. Pockman, and R. B. Jackson.
survival of chaparral seedlings during the first 2004. Adaptive variation in the vulnerability of
summer drought after wildfire. Oecologia 76:215- woody plants to xylem cavitation. Ecology
221 85:2184-2199.
.
Hacke, U. G., A. L. Jacobsen, and R. B. Pratt. Matzner, S. L., K. J. Rice, and J. H. Richards.
2009. Xylem function of arid-land shrubs from 2001. Intra-specific variation in xylem cavitation
California, USA: an ecological and evolutionary in interior live oak {Quercus wislizenii A. DC.).
analysis. Plant, Cell & Environment 32:1324—1333. Journal ofExperimental Botany 52:783-789.
Helms, A. M. 2009. The effect of severe drought on Mayr, S., B. Beikircher, M. A. Obkircher, and P.
xylem water transport in two chaparral species in Schmid. 2010. Hydraulic plasticity and limitations
the San Gabriel and Santa Monica Mountains. of alpine Rhododendron species. Oecologia 164:
Honorsthesis, Pepperdine University, Malibu, CA. 321-330.
Holste, E. K., M. j. Jerke, and S. L. Matzner. McDowell, N., W. T. Pockman, C. D. Allen, D. D.
2006. Long-term acclimatization of hydraulic Breshears, N. Cobb, T. Kolb, J. Plaut, J.
properties, xylem conduit size, wall strength and Sperry, A. West, D. G. Williams, and E. A.
cavitation resistance in Phaseolus vulgaris in Yepez. 2008. Mechanisms of plant survival and