AJB Advance Article published on September 16, 2016, as 10.3732/ajb.1600167.
The latest version is at http://www.amjbot.org/cgi/doi/10.3732/ajb.1600167
RESEARCH ARTICLE
A M E R I C A N J O U R N A L O F B O TA N Y
Seasonal changes in tissue-water relations for eight
species of ferns during historic drought in California1
Helen I. Holmlund2,6, Victoria M. Lekson2, Breahna M. Gillespie3, Nicole A. Nakamatsu2, Amanda M. Burns4, Kaitlyn E. Sauer2,
Jarmila Pittermann5, and Stephen D. Davis2
PREMISE OF THE STUDY: California experienced severe drought between 2012 and 2016. During this period, we compared seasonal changes in tissue-water
relations among eight fern species in the Santa Monica Mountains of southern California to elucidate differential mechanisms of drought survival and
physiological performance during extreme water deficits.
METHODS: We monitored seasonal changes in water potential (Ψmd) and dark-adapted chlorophyll fluorescence (Fv/Fm), assessed tissue-water relations
including osmotic potential at saturation and the turgor loss point (Ψπ, sat and Ψπ, tlp), and measured, for two evergreen species, xylem-specific and leafspecific hydraulic conductivity (Ks and Kl) and vulnerability of stem xylem to water stress-induced embolism (water potential at 50% loss hydraulic conductivity, Ψ50).
KEY RESULTS: Species grew in either riparian or chaparral understory. The five chaparral species had a wider range of seasonal water potentials, root
depths, and frond phenological traits, including one evergreen, two summer-deciduous, and two desiccation-tolerant (resurrection) species. Evergreen
species were especially diverse, with an evergreen riparian species maintaining seasonal water potentials above −1.3 MPa, while an evergreen chaparral
species had seasonal water potentials below −8 MPa. In those two species the Ψ50 values were −2.5 MPa and −4.3 MPa, respectively.
CONCLUSIONS: Observed differences in physiological performance among eight fern species reflected niche partitioning in water utilization and habitat
preference associated with distinct phenological traits. We predict differential survival among fern species as future drought events in California intensify,
with desiccation-tolerant resurrection ferns being the most resistant.
KEY WORDS chlorophyll fluorescence; ferns; niche segregation; osmotic adjustment; pteridophyte; water utilization; xylem embolism
The Santa Monica Mountain range (SMM) of southern California
is home to some of the most dehydration-tolerant angiosperms
based on susceptibility to water stress-induced embolism of stem
xylem (Jacobsen et al., 2008; Pratt et al., 2008). Several species of
evergreen chaparral shrubs in the SMM survive xylem water potentials below −11 MPa (Davis et al., 2002; Jacobsen et al., 2007; Pratt
et al., 2007). The climate is mediterranean, characterized by long,
dry summers and cool, wet winters (Aschmann, 1973; Cowling
et al., 2005). Chaparral shrubs growing in the SMM typically survive
6 to 9 mo with little to no precipitation, but in 2013 they experienced
1
Manuscript received 19 April 2016; revision accepted 25 August 2016.
Pepperdine University 24255 Pacific Coast Highway, Malibu, California 90263 USA;
3
San Diego State University, 5500 Campanile Drive, San Diego, California 92182 USA;
4
Berea College, 101 Chestnut Street, Berea, Kentucky 40403 USA; and
5
University of California, Santa Cruz, 1156 High St, Santa Cruz, California 95064 USA
6
Author for correspondence (e-mail: [email protected])
doi:10.3732/ajb.1600167
2
only 28 mm of rain over 10 mo, that led to excessive defoliation and
whole plant mortality (Coates et al., 2015). A recent study based on
tree ring analysis suggests that the 2012–2014 drought in California
was the worst in the last 1200 yr (Griffin and Anchukaitis, 2014). In
addition, recent climate models forecast 80% probability of intensified multidecadal megadroughts (>35 yr) in California the later
part of this century (Cook et al., 2015).
In the SMM, ferns not only occur in riparian habitats but also
extend their range into adjacent chaparral understories (Fig. 1A).
Remarkably, varied evergreen species occupy perennial waterfalls
and streams as well as extremely dry soils beneath drought-hardy
Ceanothus chaparral (Fig. 1A, Cm). For example, several shallowrooted Ceanothus species in the SMM typically achieve minimum
seasonal water potentials of −8 MPa (Frazer and Davis, 1988;
Thomas and Davis, 1989; Kolb and Davis, 1994; Jacobsen et al.,
2007). Theoretically, the tissues of understory evergreen ferns
should achieve water potential equilibrium with the bulk soil in the
A M E R I C A N J O U R N A L O F B OTA N Y 103(9): 1–11, 2016; http://www.amjbot.org/ © 2016 Botanical Society of America • 1
Copyright 2016 by the Botanical Society of America
2
•
A M E R I C A N J O U R N A L O F B OTA N Y
FIGURE 1 Cold Creek Canyon watershed in the Santa Monica Mountains of southern California. (A) Overview of riparian vegetation (R) adjacent upland
chaparral dominated by the shrub species Ceanothus megacarpus (Cm) undergoing drought-induced dieback in December 2014. (B) Chaparral understory showing four co-occurring species of ferns in April 2016, dehydration-tolerant Dryopteris arguta (Da), dehydration-sensitive Adiantum jordanii
(Aj), and desiccation-tolerant Pellaea andromedifolia (Pan) and Pentagramma triangularis subsp. triangularis (Ptt). Bar = 10 cm. These four species partition rooting depths that range in mean values between 8 cm and 36 cm ( Table 1) as well as differential minimum seasonal water potentials that unambiguously define the water potential gradient between plant tissues and bulk soil that drives soil moisture extraction, ranging between −2.6 MPa
and −8.4 MPa ( Table 1). (C) Same view as (B), but taken in June 2016 after Aj, Pan, and Ptt have initiated seasonal dormancy in response to drying soil
( Table 1), while Da remains evergreen. Dryopteris arguta in (D) May 2013, (E) February 2014, (F) March 2014, and (G) December 2014. Photographs:
Stephen Davis.
same rooting zone occupied by evergreen chaparral and thus
should also experience seasonal water potentials approaching −8
MPa (Fig. 1E). Additional species of non-evergreen ferns co-occurring in dry-land chaparral were either summer-deciduous or resurrection ferns, and remarkably these species grow adjacent to each
other (Fig. 1B, C). Altogether we observed four frond phenological
traits in our eight fern species: evergreen, fall-deciduous, summerdeciduous, and resurrection (Table 1). We determined how such an
array of co-occurring fern species, displaying contrasting frond
phenological traits, would differentially respond to California’s unprecedented drought of 2012–2016. We hypothesized that the eight
species examined would have a spectrum of water utilization patterns reflecting niche segregation and large differences in dehydration tolerance of both their symplastic tissues (minimum seasonal
water potential, osmotic adjustment in the turgor loss point, and
chlorophyll fluorescence) and their apoplastic tissues (dysfunction
of their xylem water transport system, i.e., water potential causing
50% loss in stipe hydraulic conductivity = Ψ50). Furthermore, we
hypothesized that evergreen species would display the greatest differences in dehydration tolerance because they grow either in perennial water or under dry chaparral.
MATERIALS AND METHODS
Plant material—Fern samples were collected from study sites in the
SMM (Fig. 1A). All eight fern species grow in the same Cold Creek
watershed. Three of the species sampled grow in or adjacent the
perennial headwaters of the Cold Creek Preserve off Stunt Road
[34°05′18″N, 118°38′44″W; Adiantum capillus-veneris L., Pteridium
S E P T E M B E R 2016 , V O LU M E 103
aquilinum (L.) Kuhn var. pubescens Underw., Woodwardia fimbriata Sm.]. Five additional species occur at a drier site along the Backbone Trail off Piuma Road (34°04′34 ″N, 118°41′10 ″W; Adiantum
jordanii Müll. Hal., Dryopteris arguta Kaulf., Pellaea andromedifolia
Kaulf., Pentagramma triangularis Kaulf. subsp. triangularis, Polypodium californicum Kaulf., Fig. 1A, Table 1). The two sites are separated by less than 4 km. Nomenclature follows Baldwin et al. (2012).
Plant root depth—Root depths were estimated after careful excava-
tion of rhizome and root systems for eight fern species at our riparian and chaparral study sites. Measurements were made in May and
June of 2016 by using a ruler or meter stick (n = 6–12, Table 1).
• H O L M LU N D E T A L. — S E A S O N A L WAT E R R E L AT I O N S I N F E R N S
• 3
sealed in tin sample canisters and returned to the laboratory
within 1 h for determination of soil water potential using dewpoint hygrometry (model WP4C, Decagon Devices). We estimated percentage shade for each fern species at riparian and
chaparral study sites using a sunfleck ceptometer (model LP-80,
Decagon Devices). Photosynthetically active radiation (PAR), in
the 400–700 nm spectrum, was measured using a probe 86 cm
long, housing 80 quantum sensors. The probe was held level, and
readings were averaged at four compass directions, north, south,
east, and west. Measurements were taken at the center of each
fern, at the topmost frond of each of eight species (n = 12–14).
Percentage shade was calculated as the percentage reduction in
PAR between open canopy and topmost fronds.
Midday water potential—At midday (between 1200 and 1500 h),
frond tips were cut, immediately sealed in plastic bags, placed on
ice, and transported back to the laboratory in less than 45 min. Water potentials were measured using a Scholander-Hammel pressure
chamber (model 1001, PMS Instrument Co., Corvallis, Oregon,
USA). Low water potentials in D. arguta measured by pressure
chamber on 17 November 2014 (−7.49 MPa ± 0.23, SE, n = 7) were
not significantly different (P > 0.9, t = 0.10, df = 12) from water
potentials measured by dew-point hygrometry on the same frond
tissues (−7.53 MPa ± 0.29) (model WP4C, Decagon Devices, Pullman, Washington, USA), giving us confidence in the accuracy of
the pressure chamber method. Seasonal water potentials were measured every 4–6 wk and more frequently after rain events to track
the rehydration process in fern tissues. We attempted to capture
these seasonal water stress dynamics on frond tissue in situ by
tracking chlorophyll fluorescence, described below.
Chlorophyll fluorescence—Dark-adapted chlorophyll fluorescence
(Fv/Fm) of fronds was measured at midday just before measurements of water potential using a pulse-modulated fluorometer
(model PS1-FL, Opti Sciences, Hudson, New Hampshire, USA). At
midday, representative central regions of mature fronds chosen for
measurement were covered with dark-adaption clips for 20 min before determination of Fv/Fm. Additional daily measurements were
collected more frequently following rain events to track the dynamic return of chloroplast function in resurrection ferns in comparison to evergreen ferns.
Pressure–volume curves—We used a modified pressure–volume
technique of Tyree and Hammel (1972) to estimate the following
tissue-water relations: osmotic potential at full tissue hydration (Ψπ, sat),
osmotic potential at the turgor loss point of tissues (Ψπ, tlp), bulk
modulus of tissue elasticity (ε), and tissue-water capacitance (C).
The method used has been fully described by Davis and Mooney
(1986). Frond samples were gathered early in the morning, between
the hours of 06:30 and 08:30, following natural overnight hydration. The tips of fronds (10–20 cm long) were cut under water in the
field, placed in 50-mL tubes of water, and covered individually with
plastic bags. All tubes were placed upright in a cooler, which was
also covered with a large plastic bag, and transported in less than
45 min back to the laboratory. Frond samples were rehydrated for
1–12 h in spring and 1–2 h in summer and fall, when the tissues
were relatively dry. All leaves were carefully examined with a magnifying glass for signs of over-hydration, including water-filled
spaces between mesophyll cells, to prevent abnormal shifts in
cell wall elasticity and osmotic potential (cf. Meinzer et al., 1986;
Saruwatari and Davis, 1989).
We generated pressure–volume curves by taking sequential water potential measurements (Ψ) as fronds were allowed to air dry on
a bench top. The mass before and after Ψ determination was obtained using an analytical balance. After the curve was complete,
leaf samples were placed in a drying oven at 65°C for at least 48 h
before recording the dry mass. For each data point, relative water
content (RWC) was obtained using the equation:
Microclimate data—Rainfall data were collected less than 1 km
from our Piuma Road study site at the Monte Nido Weather Station (34°04′41″N, 118°41′38″W), operated by the Los Angeles
Department of Public Works (Fig. 2C). Percentage soil moisture,
from soil surface to a depth of 12 cm, was estimated in June 2016
on 12 individuals from each of the eight fern species using a time
domain reflectometer, equipped with 12-cm long probes (HydroSense Soil Water Measurement System, Campbell Scientific,
Logan, Utah, USA). Although soil moisture at each site likely
varies seasonally, we chose to measure the soil moisture of the
top 12 cm in June when the shallow-rooted species go dormant.
We also measured soil water potentials at 10 and 20 cm depths at
the Piuma Road site in June 2014 and October 2014 to track the
decline in soil moisture resources during the summer and fall
rainless period characteristic of the mediterranean-type climate
in the SMM. About 20 g of soil were removed at each depth using a soil sampler (model KB, Oakfield Apparatus, Oakfield,
Wisconsin, USA). Soil was extracted approximately 3 m apart
along a transect 80 m in length (n = 23–28). Soil was immediately
RWC =
FW − DW
SW − DW
where FW = fresh mass, DW = dry mass, and SW = mass at saturation. As done by Saruwatari and Davis (1989), we partially rehydrated fronds during summer and fall dry periods to avoid
over-hydration of tissues that might impact cell wall elasticity. Because we only partially rehydrated the fronds, it was necessary to
calculate the true saturated mass (SW), which we estimated using a
linear regression of the initial FW vs. Ψ values, calculating SW at
the intercept where Ψ = 0 MPa.
To determine Ψπ, tlp, a plot of 1/Ψ vs. RWC was generated, and the
point at which the graph linearized was taken as the Ψπ, tlp (Davis and
Mooney, 1986). For D. arguta, Ψπ, tlp was measured in June, July, September, November, and December of 2014. For W. fimbriata, Ψπ, tlp
was measured in June 2014, July 2013, and September 2015. All eight
species were compared in June 2014 (Table 2). Ψπ, sat was determined
Notes: Wf = Woodwardia fimbriata, Acv = Adiantum capillus-veneris, Paq = Pteridium aquilinum var. pubescens, Da = Dryopteris arguta, Pc = Polypodium californicum, Aj = Adiantum jordanii, Pan = Pellaea andromedifolia, Ptt = Pentagramma
triangularis subsp. triangularis, E = evergreen, D = deciduous, R = resurrection, S = sensitive, T = tolerant. Different uppercase letters A–E denote significant difference in mean values by a Tukey’s test (P < 0.05) that was performed post hoc
after a one-way ANOVA failed to support the null hypothesis (P < 0.005). Frond seasonal Ψmin: F7, 41 = 216.9, n = 6–7; soil moisture: F7, 89 = 36.93, n = 12; maximum root depth: F7, 75 = 120.7, n = 6–12, percentage shade: F7, 92 = 3.16, n = 12–14.
a
Weeks during which species had no photosynthetically active tissue from June 2013 to May 2014.
b
The seasonal minimum water potentials reflect values that could be measured with the pressure chamber. For the two resurrection species, the true seasonal minimum water potentials were much lower, since the frond tissues equilibrate
with dry air during desiccation.
74 ± 8 AB
91 ± 6 A
68 ± 10 AB
76 ± 7 AB
97 ± 1 A
73 ± 5 AB
57 ± 10 B
87 ± 4 A
0.110 ± 0.008 A
0.024 ± 0.004 B
0.238 ± 0.012 C
0.363 ± 0.019 D
0.030 ± 0.003 B
0.201 ± 0.015 C
0.121 ± 0.009 A
0.078 ± 0.005 A
52 ± 9 A
38 ± 4 AB
29 ± 8 B
6 ± 1 CD
4±0D
7 ± 1 CD
6 ± 1 CD
10 ± 1 C
−1.24 ± 0.07 A
−1.81 ± 0.05 AB
−2.13 ± 0.02 BC
−8.43 ± 0.28 D
−2.81 ± 0.10 E
−3.24 ± 0.16 E
−2.61 ± 0.19 CE
−2.84 ± 0.16 E
Dehydration S
Dehydration S
Dehydration S
Dehydration T
Dehydration S
Dehydration S
Desiccation T
Desiccation T
E
E
D (fall)
E
D (summer)
D (summer)
R
R
Wf
Acv
Paq
Da
Pc
Aj
Pan
Ptt
Riparian
Riparian
Riparian
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
0
0
13
0
37
43
24
24
Mean % soil moisture
± SE June 2016, 0.12 m
Mean frond seasonal
Ψmin ± SE (MPa) b
Frond
phenological trait
Habitat
Frond seasonal
dormancy (wk/yr) a
Frond tissue-water
relations
Mean maximum
root depth ± SE (m)
Mean %
shade ± SE
A M E R I C A N J O U R N A L O F B OTA N Y
Species
•
TABLE 1. Summary of traits and abiotic factors for eight fern species growing in a common watershed in the Santa Monica Mountains of southern California.
4
by extending the regression line to the point where x = 0 (RWC =
100). ε was determined from the slope m of the line of Ψp vs. RWC
and the RWC at the turgor loss point (RWC°), such that ε= mRWC°,
and C was determined as the slope of RWC vs. Ψ (Table 2).
Hydraulic conductivity and cavitation resistance—We chose to examine hydraulic conductivity and cavitation resistance in D. arguta
and W. fimbriata because of their frond persistence and larger stipe
diameter, which facilitates measurable flow rates. Evergreen A. capillus-veneris had tiny, fragile fronds and flow rates that were difficult to measure. Hydraulic conductivity was measured using the
methods of Sperry et al. (1988). Fronds were cut under water at
predawn, between the hours of 06:00 and 06:30, and double-bagged
to reduce water loss. The double-bagged samples were additionally
covered with a black plastic bag to promote further stomatal closure
and reduce water loss during immediate transport back to the
laboratory. In the laboratory, fronds were submerged in water, and
alternate ends of each stipe were cut under water to prevent the introduction of air into xylem conduits. Then both ends of each stipe
were trimmed with a new razor blade to generate a 140-mm stipe
segment with minimal damage to vascular tissue.
To assess flow rate, we connected each stem to a miniaturized
hydraulic apparatus. For hydraulic conductivity measurements,
we used a 20 mM KCl solution, degassed with a vacuum pump for
20 min, and passed through a 0.1 μm filter. Hydrostatic pressures of
5–7 kPa were generated by an elevated reservoir that caused the
KCl solution to flow through each stipe segment onto an analytical
balance to estimate flow rate gravimetrically. The balance was connected to a computer system with custom-designed software
(Sperry, 1996) to calculate conductivity based on the flow rate,
room temperature, hydrostatic pressure, and stipe length. Hydraulic conductivity (K, kg·m·MPa−1·s−1) was calculated as the volume
flow rate of KCl solution (q, kg·s−1) through a given stipe segment
(dx, m) divided by the pressure gradient (dP/dx, MPa·m−1).
K=
q
dP / dx
We also estimated xylem area-specific conductivity Ks, by dividing
K by the transverse xylem area of the fern stipe as well as leaf areaspecific conductivity Kl, by dividing the pinnae area supplied by the
stipe. Ks represents the transport efficiency of xylem water by the
stipe, whereas Kl represents the supply efficiency of xylem water to
pinnae. We measured transverse xylem area with a compound microscope (model SZH, Olympus, Lake Success, New York, USA)
and pinnae area with a leaf area meter (model LI-3100, Li-Cor, Lincoln, Nebraska, USA).
Percentage loss in hydraulic conductivity due to air embolism
(PLC) under natural field conditions, at predawn, is often referred
to as native embolism (PLCnative) and was calculated by the following equation (Hacke et al., 2015):

K ¬
PLC native = žžž1 int ­­­ ×1 00%
žŸ K max ­®
where Kint is the initial (native) conductivity before removal of air
emboli and Kmax is the conductivity after removal of air emboli by
degassing in 20 mM KCl solution for 60 min.
S E P T E M B E R 2016 , V O LU M E 103
• H O L M LU N D E T A L. — S E A S O N A L WAT E R R E L AT I O N S I N F E R N S
• 5
FIGURE 2 Seasonal trends in (A) midday water potential (n = 3–7), (B) chlorophyll fluorescence (n = 7–39), and (C) 5-d precipitation from July 2013 to
February 2016, obtained from the Monte Nido Weather Station, Los Angeles Department of Public Works. Breaks in the line graphs indicate dormancy.
Color indicates frond phenological trait in eight species: green = evergreen (Dryopteris arguta, Woodwardia fimbriata, Adiantum capillus-veneris), orange = deciduous (Polypodium californicum, Adiantum jordanii, Pteridium aquilinum var. pubescens), purple = resurrection (Pentagramma triangularis
subsp. triangularis, Pellaea andromedifolia). Asterisks indicate time that chaparral soil water potential was measured in *June 2014 (−30.6 ± 3.6 MPa at
10 cm, −4.1 ± 0.6 MPa at 20 cm, n = 28) and **October 2014 (−59.1 ± 4.2 MPa at 10 cm, −25.8 ± 1.8 MPa at 20 cm, n = 23). Bars = ±1 SE.
6
•
A M E R I C A N J O U R N A L O F B OTA N Y
TABLE 2. Frond tissue-water relations calculated from pressure–volume curves on eight fern species growing in a common watershed in the Santa Monica
Mountains.
Species
Wf
Acv
Paq
Da
Pc
Aj
Pan
Ptt
Mean Ψπ, TLP ± SE
(MPa) (June) a
Mean Ψπ, TLP ± SE (MPa)
(September) b
Mean Ψπ,, Sat ± SE
(MPa) (June) a
Mean ε ± SE
(MPa) (June) a
Mean C ± SE (MPa−1)
(June) a
−1.34 ± 0.10 A
−1.21 ± 0.05 A
−2.06 ± 0.03 B
−1.91 ± 0.15 B
−1.85 ± 0.04 B
−1.83 ± 0.05 B
−1.46 ± 0.05 A
−1.29 ± 0.06 A
−1.47 ± 0.02 A
—
—
−4.68 ± 0.29 B
—
—
—
—
−1.15 ± 0.09 AB
−1.02 ± 0.04 A
−1.80 ± 0.06 C
−1.44 ± 0.13 BD
−1.55 ± 0.05 CD
−1.28 ± 0.09 ABD
−0.97 ± 0.05 A
−1.12 ± 0.06 AB
24.7 ± 3.7AB
18.9 ± 3.4 AC
34.1 ± 1.8 B
12.8 ± 3.5 C
25.1 ± 1.9 AB
9.5 ± 1.9 C
10.1 ± 1.4 C
7.9 ± 0.4 C
3.8 ± 0.6 AB
5.4 ± 1.0 ABC
2.4 ± 0.1 A
7.4 ± 1.5 BC
3.4 ± 0.4 AB
7.0 ± 1.1 BC
5.2 ± 0.5 ABC
8.0 ± 0.6 C
Notes: Species abbreviations are the same as in Table 1.
Different uppercase letters denote significant difference (P < 0.05) by one-way ANOVA followed by a Tukey’s HSD post hoc test, F7, 51 = 20.26 (June Ψπ, tlp), 12.73 (June Ψπ, sat), 12.84 (June ε), 5.11
(June C), P < 0.001, n = 6–7.
b
Different uppercase letters denote significant difference in mean values by unpaired Student’s t test, P < 0.0001, t = 12.93, df = 10, n = 6-7.
a
We constructed vulnerability to embolism curves using the centrifuge method of Alder et al. (1997). Spinning the stem at a given
speed generates a known centrifugal force on the water in the middle of the stem, which can be used to replicate biologically relevant
tensions. These tensions have been shown to match leaf water potentials (Holbrook et al., 1995). After initial removal of air emboli
by vacuum treatment, Kmax was measured, and then stipe segments
were repeatedly measured after increasing rates of centrifugation
that generated a series of increasing tensions of −0.5 MPa, −2.0
MPa, −4.0 MPa, −6.0 MPa, and −8.0 MPa for D. arguta and tensions
of −0.5 MPa, −2.0 MPa, −4.0 MPa, −5.0 MPa, and −6.0 MPa for W.
fimbriata. All stems were spun for 5 min at each tension. Percentage
loss in hydraulic conductivity of stipe segments due to air embolism
at each tension (PLCt) was calculated using the equation
ž
PLC = žžžžž1K Kt
t
Ÿ
¬
­­­
­­
­
max ®­
× 100%
where Kt is the conductivity after centrifugation at each tension
listed above. We generated vulnerability to embolism curves by
graphing PLCt vs. Ψ (tension in MPa). Data were fitted with a polynomial curve (first-degree for D. arguta and second-degree for
W. fimbriata). Curve fits were selected based on visual inspection of
fit and r2 = 0.999 for corrected mean loss of conductivity, %, vs. tension, MPa, in the linear fit for D. arguta and the quadratic fit for
W. fimbriata. To compare species, we calculated the mean tension at
50% loss of conductivity (Ψ50) and at 80% loss of conductivity (Ψ80)
from each curve (n = 5–6).
We corrected PLCnative for xylem fatigue by using the PLC measured
by centrifugation at −0.5 MPa tension as the zero starting point, effectively subtracting it out from all values. Since midday water potentials
were more negative than −0.5 MPa, we assumed any embolism measured at this tension by centrifugation was due to nonfunctional, fatigued conduits that did not contribute to native xylem water transport.
Essentially, it provides a Kmax that more accurately reflects the plant’s
maximum conductivity in situ. The rationale for this correction has
been fully explained previously (Hacke et al., 2000; Sperry and Hacke,
2002; Maherali et al., 2006; Jacobsen et al., 2007). Both corrected and
uncorrected values were reported for comparison.
Statistical analyses—When comparing three or more species, statistical differences were determined using a one-way ANOVA followed by a Tukey’s honestly significant difference (HSD) post hoc
test with an alpha of 0.05. When only two data sets were compared,
an unpaired Student’s t test with equal variance was employed. Linear regressions were considered significant after Pearson’s correlation analysis at P < 0.05 (SPSS Statistics, IBM, Armonk, New York,
USA). Data were log-transformed where appropriate (PLC values,
% soil moisture, % shade).
RESULTS
Seasonal changes in water potential—A comparison of midday
water potentials among eight fern species over 28 mo showed repeated annual patterns (Fig. 2A). Water potentials were generally
lowest in the late fall and highest in winter after the initiation of
rains (Fig. 2C). Seasonal water potentials were among the lowest
observed for all species in January and February of 2014, typically
the wettest months of the year (Fig. 2A, C). Seasonal water potentials were not as low in the fall of 2015 as in 2013 and 2014, possibly
reflecting the unusual summer rain event of 24 mm on 18 July 2015
(Fig. 2C).
The three evergreen species showed the greatest disparity in
midday water potentials. The perennial stream species A. capillusveneris and W. fimbriata never dropped below −2 MPa, and the
chaparral understory species D. arguta repeatedly experienced water potentials below −8 MPa (Fig. 2A). A comparison of D. arguta
midday water potentials with the dominant overstory chaparral
shrub Ceanothus megacarpus after protracted drought on 5 October 2013 indicated no significant difference between species (−8.16
MPa ± 0.34 vs. −8.89 ± 0.46 [±SE], n = 6, P > 0.2, t = −1.29, df = 10,
data not shown). This comparison between an understory evergreen fern and a dominant overstory chaparral shrub suggests approximate water potential equilibrium during summer and fall dry
seasons. The evergreen fern W. fimbriata growing in the perennial
headwaters of Cold Creek maintained the highest seasonal water
potential among the eight species examined and experienced a
minimum seasonal water potential of only −1.24 MPa (Fig. 2A,
Table 1).
Deciduous species experienced frond dieback (dormancy) in either the summer (summer-deciduous) or the fall (fall-deciduous),
as indicated by the breaks in the graph on Fig. 2A and B. The only
fall-deciduous species in this study, P. aquilinum, grew adjacent to
A. capillus-veneris and W. fimbriata in top soil that was drier than
that of W. fimbriata in June 2016, at the beginning of the summer
S E P T E M B E R 2016 , V O LU M E 103
• H O L M LU N D E T A L. — S E A S O N A L WAT E R R E L AT I O N S I N F E R N S
• 7
drought (29% vs. 52% soil moisture, Table 1). The minimum seasonal water potential of P. aquilinum was significantly different
from that of evergreen W. fimbriata (−2.13 MPa vs. −1.24 MPa, Table 1), yet not significantly different from that of evergreen waterfall
species A. capillus-veneris (−2.13 MPa vs. −1.81 MPa, Table 1).
Summer-deciduous species A. jordanii and P. californicum experienced much longer seasonal dormancy (37–43 wk, Fig. 2A,
Table 1) than the fall-deciduous species P. aquilinum (13 wk, Fig.
2A, Table 1). Resurrection ferns P. andromedifolia and P. triangularis had intermediate dormancy periods of 24 wk, primarily because they exhibited rapid revival of tissues in response to isolated
rain events (e.g., 6.8 mm on 6 January 2014 and 24 mm on 18 July
2015, Fig. 2A, C). The resurrection ferns did not break desiccationinduced dormancy during summer dry periods except following a
rare summer rain event, as was the case on 18 July 2015 (Fig. 2C).
Seasonal changes in chlorophyll fluorescence—For the three ever-
green species A. capillus-veneris, D. arguta, and W. fimbriata, seasonal changes in dark-adapted chlorophyll fluorescence at midday
(Fv/Fm) followed a pattern similar to seasonal changes in water potential (compare Fig. 2A and 2B). Fluorescence values were generally lowest in the late fall and highest in winter after the initiation of
rains (Fig. 2B, C). The three evergreen species showed the greatest
disparity in Fv/Fm. The perennial stream species A. capillus-veneris
and W. fimbriata never dropped below a Fv/Fm of 0.7, yet the Fv/Fm
of the chaparral understory species D. arguta repeatedly dropped
below 0.3 (Fig. 2B). Woodwardia fimbriata, growing in the perennial headwaters of Cold Creek, maintained the highest Fv/Fm among
the evergreen ferns examined and experienced a seasonal minimum Fv/Fm of only 0.786 (Fig. 2B). The seasonal minimum in Fv/Fm
for the third evergreen fern, A. capillus-veneris, was 0.746 (Fig. 2B).
Summer-deciduous species A. jordanii and P. californicum
avoided prolonged low Fv/Fm by foliar abscission during dry periods (Fig. 2B). The only fall-deciduous species in the comparison, P.
aquilinum, avoided low Fv/Fm by growing near a water source (Cold
Creek) and by foliar abscission during cold winter months (minimum temperatures of −8°C measured by max/min thermometer at
Cold Creek Preserve on 15 January 2013, personal observation, S.
Davis). The seasonal low Fv/Fm for P. aquilinum was only 0.822 (Fig.
2B). In contrast, resurrection ferns P. andromedifolia and P. triangularis experienced some of the lowest Fv/Fm values measured during the study, but these low values were only detected during abrupt
emergence from desiccation-induced dormancy, within 24 h after
rainfall events (Figs. 2B, 2C; 3, inset).
Correlation between seasonal Ψ and Fv/Fm—Midday water potentials plotted against dark-adapted chlorophyll fluorescence for all
eight species resulted in a Pearson’s correlation coefficient of r2 =
0.812 (n = 107, P < 0.001, df = 105) (Fig. 3). This regression analysis
excluded four data points collected during brief revival of desiccated tissues in resurrection ferns, P. andromedifolia and P. triangularis, immediately following postdrought precipitation events
(Fig. 2B). We were concerned that the abrupt revival of desiccated
tissues and chloroplast function from deep dormancy in resurrection ferns would confound seasonal correlations between gradual
seasonal decline in Ψmd and Fv/Fm (Fig. 3, inset).
Soil moisture, maximum root depth, and percentage shade—Soil
moisture at 0.12 m was significantly higher in the riparian habitat than
in the chaparral understory in June 2016 (P < 0.001, F1, 94 = 169, Table 1).
FIGURE 3 Relationship between midday water potential (Ψmd) and midday
dark-adapted chlorophyll fluorescence (Fv/Fm) in eight fern species, excluding four data points taken during the rapid rehydration phase of the resurrection ferns. Points represent mean values, bars = ±1 SE, n = 107 mean
values, r2 = 0.812, df = 105, P < 0.001. The inset graph shows the dynamic
increase in Fv/Fm for two resurrection species, P. andromedifolia (purple)
and P. triangularis (purple), in comparison to the evergreen fern D. arguta
(green), following a 24-mm rain event on 18 July 2015 (Day 0). Ψmd (MPa)
for each species is listed at Day 1 and Day 16. On Day 1, the two resurrection species have low Fv/Fm despite being relatively hydrated (−1.8 MPa)
and thus were not included in the linear regression because they would
confound the relationship between seasonal declines in Ψmd and Fv/Fm.
Soil water potentials in June 2014 and October 2014 at the Piuma Road
site were consistently lower at 10 cm (−30.6 ± 3.6 MPa and −59.1 ± 4.2
MPa, respectively) than at 20 cm (−4.1 ± 0.6 MPa and −25.8 ± 1.8 MPa,
respectively) (n = 23–28, Fig. 2A, see asterisks). Root depths were
greatest for D. arguta and least for A. capillus-veneris and P. californicum. Adiantum jordanii and P. aquilinum had intermediate root
depths, whereas P. andromedifolia, P. triangularis, and W. fimbriata
had relatively shallow root depths (P < 0.0001, F7, 75 = 121, Table 1).
Percentage shade at midday was not significantly different between
combined fern species at riparian and chaparral habitats in June 2016
(P = 0.87, F1, 94 = 0.25, Table 1). Percentage shade was significantly
higher for A. capillus-veneris, P. californicum, and P. triangularis than
for P. andromedifolia (P < 0.005, F7, 92 = 3.2, Table 1).
Comparison of frond tissue water relations by pressure–volume
curve analyses—For the three evergreen species in June 2014,
Ψπ, tlp was more negative for the chaparral understory species
D. arguta (−1.91 MPa, Table 2) than the perennial stream species
A. capillus-veneris and W. fimbriata (−1.21 and −1.34 MPa, respectively, Table 2). The fall-deciduous species, P. aquilinum, had
the most negative Ψπ, sat (−1.80 MPa, Table 2) and one of the most
negative Ψπ, tlp (−2.06 MPa, Table 2). Resurrection ferns P. andromedifolia and P. triangularis had relatively high Ψπ, sat and Ψπ, tlp and
were most similar to evergreen species growing in perennial streams
(Table 2). No clear patterns were observed for ε and C among the
eight species (Table 2).
8
•
A M E R I C A N J O U R N A L O F B OTA N Y
Repeated pressure–volume curve analyses for two evergreen
species during the summer and fall rainless periods indicated
dramatic seasonal osmotic adjustment at the turgor loss point
for the chaparral understory species D. arguta but not the perennial stream species W. fimbriata (Fig. 4). There was minimal
overlap in Ψπ, tlp between the two species with D. arguta being
consistently more negative (Fig. 4). The 3-fold lowering of Ψπ, tlp
for D. arguta during the late summer and fall (Fig. 4) was quickly
reversed with 113 mm of rain in December 2014 (Fig. 2C). The
magnitude of seasonal osmotic adjustment in D. arguta was
nearly 4 MPa in contrast to that of W. fimbriata, which was
<0.2 MPa.
Hydraulic transport properties of two evergreen ferns—The dehy-
dration-tolerant evergreen fern inhabiting chaparral understories,
D. arguta, had a 6-fold lower hydraulic conductivity (Knative), 4-fold
lower xylem-specific conductivity (Ks, native), and a 6-fold lower leafspecific conductivity (Kl, native) than the species inhabiting perennial
streams, W. fimbriata (Table 3). This pattern persisted after all air
emboli were removed (converted from Knative values to Kmax values,
Table 3). Native embolism levels (PLCnative) were only 2-fold higher
for D. arguta than W. fimbriata despite 3-fold lower water potentials at midday (Table 3).
Vulnerability to cavitation—The stipes of D. arguta were 2 to 3
MPa more resistant to water stress-induced embolism at Ψ50 and
Ψ80 than the species inhabiting perennial streams, W. fimbriata
(Fig. 5, Table 3), and this pattern held true even after Ψ50 and Ψ80
were corrected for xylem fatigue (Table 3). Ψ80 for D. arguta of
−7.57 MPa indicated that this species could experience minimum
seasonal water potentials below −8 MPa (Table 1) and yet maintain
approximately 20% of stipe xylem water transport. When field midday water potentials were used to predict native embolism levels
FIGURE 4 Seasonal changes in ΨΠ, tlp for the two evergreen species,
Woodwardia fimbriata and Dryopteris arguta. Mean values (n = 6), bars =
±1 SE. Letters denote significant difference (P < 0.05) by one-way ANOVA
followed by a Tukey’s HSD post hoc test, F7,43 = 108.27, P < 0.0001.
TABLE 3. Hydraulic conductivity of fern stipes and susceptibility of stipe
xylem to water stress-induced embolism in two evergreen species, one from
a dry chaparral habitat (Dryopteris arguta) and the other from a wet riparian
habitat (Woodwardia fimbriata), growing in the Santa Monica Mountains.
Measures a
Native K values
Knative (10−6, kg·m·s−1·MPa−1)
Ks, native (kg·m−1·s−1·MPa−1)
Kl, native (10−5, kg·m−1·s−1·MPa−1)
Ψmd (MPa)
PLCnative (%)b
PLCnative, predicted (%)b
Ψ50 (MPa)
Ψ80 (MPa)
Maximum K values
Kmax (10−6, kg·m·s−1·MPa−1)
Ks, max (kg·m−1·s−1·MPa−1)
Kl, max (10−5, kg·m−1·s−1·MPa−1)
Corrected PLC values
Ψ50, corrected (MPa)
Ψ80, corrected (MPa)
D. arguta
W. fimbriata
P value
0.86 ± 0.07
2.03 ± 0.21
1.58 ± 0.17
−3.59 ± 0.14
53.88 ± 4.60 A
43.21 ± 3.11 A
−4.32 ± 0.37
−7.57 ± 0.47
6.13 ± 1.20
9.73 ± 1.56
10.40 ± 1.59
−1.10 ± 0.07
20.64 ± 8.93 B
27.11 ± 3.54 B
−2.53 ± 0.40
−4.51 ± 0.45
<0.005
<0.001
<0.001
<0.0001
<0.05
<0.01
<0.01
<0.005
1.99 ± 0.31
4.47 ± 0.28
3.61 ± 0.56
7.53 ± 1.09
12.15 ± 1.28
12.68 ± 1.09
<0.001
<0.001
<0.0001
−5.74 ± 0.43
−8.56 ± 0.62
−3.07 ± 0.44
−5.89 ± 0.93
<0.005
<0.05
All values are means ±1 SE, n = 5–6.
Different uppercase letters denote significant difference in mean values between native
and predicted native percentage loss of conductivity (PLC), P < 0.05.
a
b
(PLCnative, predicted) by using laboratory-generated vulnerability to embolism curves, results for theoretical native embolism were not significantly different (Table 3).
DISCUSSION
To elucidate differential mechanisms of water utilization in ferns
during unprecedented drought in California, we compared the response of eight species in their natural habitats at a dry edge of California’s mediterranean-type climate, which received 185 mm of
annual rainfall from 2012–2015, less than half the annual mean of
376 mm (Los Angeles Civic Center, 1877–2015, Los Angeles Almanac, Montebello, CA, USA). Our results supported our initial hypothesis that there are significant differences in water utilization
among the eight species examined, including frond phenological
traits, water availability (root depth, soil moisture, soil water potential), and dehydration tolerance (Ψmin). Furthermore, our results
were consistent with our related hypothesis that two evergreen species in contrasting riparian and chaparral habitats would show
large differences in water utilization (Ψmin, shift in Ψπ, tlp, Ψ50, Ψ80).
As predicted, we found that fronds of two evergreen species had
both the highest and lowest seasonal water potentials among the
eight species examined, with W. fimbriata growing in perennial
streams (52 ± 9% soil moisture in June 2016, at the beginning of the
dry season) displaying the highest Ψmd and D. arguta growing under dry chaparral (6 ± 1% soil moisture in June 2016) displaying the
lowest Ψmd. Lowest values in seasonal water potential often matched
lowest values in chloroplast function, estimated by chlorophyll
fluorescence (Fv/Fm) (Fig. 3). Extremely low values of Fv/Fm were
likely driven by low rainfall that in one case was only 28 mm over a
10-mo period ending in late February 2014. The one exception to
the good match between water potential and fluorescence was low
Fv/Fm associated with abrupt revival of physiological function in
desiccated resurrection ferns immediately after rain events. Thus, it
was not surprising that when fluorescence data during this reversal
of desiccation-induced dormancy were removed from the overall
S E P T E M B E R 2016 , V O LU M E 103
FIGURE 5 Vulnerability to water stress-induced cavitation of stipe xylem
for the two evergreen species, Woodwardia fimbriata and Dryopteris arguta. Points represent mean % loss of conductivity for each centrifugegenerated tension, n = 6. Bars represent ±1 SE. Solid lines represent
standard curves, dotted lines represent curves “corrected” for xylem fatigue. Corrected curves were generated by replacing Kmax with Kh at −0.5
MPa.
regression analysis between Ψmd and Fv/Fm, midday water potential
was an improved predictor of dark-adapted fluorescence (r2 =
0.812). We think that the observed decline in Fv/Fm with Ψmd was
due primarily to the influence of low osmotic potential on chloroplast function, especially for D. arguta, which had midday water
potentials below −8 MPa and whose Fv/Fm values defined the lower
portion of the regression line for the eight species. This relationship
between Fv/Fm and Ψmd is consistent with similar results for D. arguta growing in a redwood understory during the same California
drought period, but located 530 km north in Santa Cruz (Baer et al.,
2016) with a mean annual rainfall 2-fold higher (790 mm) than the
SMM. They reported a minimum seasonal Ψmd of −3.3 MPa with an
Fv/Fm of 0.74. Our regression line applied to their Ψmd for D. arguta
predicts an Fv/Fm of 0.70, which is slightly lower than their observed
Fv/Fm. However, their D. arguta growing in a shaded redwood forest of northern California was exposed to much less radiation than
our D. arguta growing beneath a chaparral canopy that had
drought-induced leaf drop, branchlet dieback, and whole plant
mortality. Thus, higher solar radiation loads on our ferns may have
contributed to lower seasonal Fv/Fm than the ferns 530 km north in
Santa Cruz (Baer et al., 2016).
Lowest seasonal Ψmd for D. arguta was accompanied not only by
low Fv/Fm but also by low osmotic potentials of −5.7 MPa at the
turgor loss point and also high embolism levels approaching 80%
based on vulnerability to embolism curves (Ψ80 = −7.6 MPa). In
contrast, W. fimbriata would be 100% embolized at −8 MPa. This
observation is consistent with native embolism levels of 54% for D.
arguta and 21% embolism for W. fimbriata measured at Ψmd in July
2014. These results are also consistent with the observation of no
• H O L M LU N D E T A L. — S E A S O N A L WAT E R R E L AT I O N S I N F E R N S
• 9
dieback in evergreen fronds of W. fimbriata during our study but
significant dieback in fronds of D. arguta. Baer et al. (2016) calculated a Ψ50 of −3.8 MPa for D. arguta using a noncentrifuge method,
which is close to our Ψ50 of −4.3 MPa, and in the right direction if
D. arguta was adapted to a 2-fold higher rainfall at Santa Cruz than
in the SMM (cf. Jacobsen et al., 2014). We found lower levels of
xylem fatigue in W. fimbriata than in D. arguta, possibly reflecting
repeated exposure to low Ψmd in D. arguta but not W. fimbriata (cf.
Anderegg et al., 2013).
The only fall-deciduous species examined in this study, P. aquilinum, displayed a relatively high minimum seasonal Ψmd, perhaps
because it grew slightly upslope of perennial stream species A. capillus-veneris and W. fimbriata, but also because it initiated dormancy during the drought in late fall of 2013 and 2014. Frond
dieback in late November was probably a response to cold temperature, which may reach −8°C at the site during winter (measured by
max/min thermometer at Cold Creek Preserve on 15 January 2013,
S. D. Davis, personal observation). Pteridium aquilinum fronds are
known to die in response to frost, due to frond and rhizome tissue
damage (Watt, 1954; Pakeman et al., 1994). Additionally, mean vessel diameter of P. aquilinum approximates 70 μm (Brodersen et al.,
2014), which suggests that vessel conduits may be susceptible to
freezing-induced embolism. Susceptibility to freezing-induced embolism has been identified as a function of vessel and tracheid
diameters >40 μm for both angiosperms and gymnosperms (Davis
et al., 1999; Pittermann and Sperry, 2003). The fall-deciduous dormancy is probably short in the SMM because hard freezes are restricted to 3 mo in this low-elevation, coastal mountain range (cf.
Davis et al., 2007).
The fronds of summer-deciduous ferns A. jordanii and P. californicum died due to dehydration in the dry season due to tissue
dehydration of the fronds below −3 MPa. The plants remained dormant for 37–43 wk, much longer than either fall-deciduous ferns
(13 wk) or resurrection ferns (24 wk). The long dormancy for summer-deciduous ferns presumably was due to the sensitivity of frond
tissues to dehydration and the time required to regrow complete
fronds de novo during cold winter months.
In contrast, the resurrection ferns were dormant due to desiccation for only 24 wk, in part because they did not grow new tissues
de novo but rapidly rehydrated old tissues with new rain events,
even as small as 5 mm, and in several cases quickly reverted to desiccation-induced dormancy as Ψmd fell to −3 MPa (Figs. 2A; 3, inset). Dormancy and revival of frond tissues occurred multiple times
each season for desiccation-tolerant ferns, and thus total length of
dormancy was less than summer-deciduous species but longer than
the fall-deciduous species that grew streamside (13 wk dormancy).
Our comparison of eight fern species from 2013–2016 showed
striking contrasts in their response to unprecedented drought in California, at the dry edge of its mediterranean-type climate zone. Differences were partitioned not only across two habitats, riparian and
chaparral understories, but also across four frond phenological traits:
evergreen, fall deciduous, summer deciduous, and resurrection. Additionally, differences were partitioned across three types of tissuewater relations: dehydration sensitive, dehydration tolerant, and
desiccation tolerant. Dehydration-sensitive ferns in our study (1) defoliated due to dehydration below −3 MPa (A. jordanii, P. californicum), (2) defoliated before winter freeze when temperatures at the
site reached as low as −8°C (P. aquilinum), or (3) grew in a perennial
water source and remained evergreen (A. capillus-veneris, W. fimbriata). Dehydration-tolerant D. arguta persisted at low water potentials
10
•
A M E R I C A N J O U R N A L O F B OTA N Y
(−5 to −8 MPa) during the summer drought, maintaining evergreen
fronds through osmotic adjustment, cavitation resistance, and relatively deep roots. Dryopteris arguta showed fluorescence values,
Fv/Fm, approaching 0.2 just before dieback at midday water potentials
of −9 MPa (the lowest water potentials measured on individual fronds
between 2013 and 2016, Fig. 1E). Thus, we assume dieback occurred
close to this value and near a predicted xylem embolism level of ~90%
(Figs. 1E, 1F, 5). Desiccation-tolerant ferns are likely sensitive to prolonged dehydration (less than −3 MPa). However, they avoid prolonged, slow dehydration by rapidly desiccating and becoming
dormant, a process facilitated by relatively shallow roots and dry soil.
Although desiccation-tolerant ferns must pass through severe dehydration states to achieve a desiccated state, they are not equipped to
sustain frond dehydration at −5 to −8 MPa. In contrast, they are
equipped to sustain and recover from desiccation, below −30 MPa.
Shallow roots and cavitation allow them to pass through the dehydration states quickly.
This analysis for the four fern species pictured in Fig. 1B and C is
consistent with classical niche theory that co-occurring species
should partition water resource utilization to reduce competition
and allow coexistence of species in the same habitat (Hutchinson,
1957; Davis, 1991). Our data suggest both aboveground and belowground partitioning contributed to niche segregation. Aboveground, the annual duration in water utilization by fern fronds
ranged from 52 wk for the evergreen fern to 28 wk for two resurrection ferns, and 9 wk for a summer-deciduous fern. Belowground,
partitioning of water utilization was regulated by a combination of
maximum root depth characteristic of each fern species and soil
water potential in deep soil horizons during the seasonal dry period, typically June to October. Maximum root depths among species pictured in Fig. 1B and C ranged from 36.3 cm for the evergreen
fern (D. arguta) to 20.1 cm for the summer-deciduous fern (A. jordanii), to 7.8 and 12.1 cm for the two resurrection ferns. Maximum
root depth for each species not only corresponded to soil water potential at 10 cm and 20 cm in June and October, but also to frond
phenological traits, minimum seasonal water potential, and timing
of frond dehydration-sensitive and desiccation-induced dormancy.
Soil water potential along with root depths indicated that there
was inadequate soil water available to maintain frond tissues of
resurrection ferns in June 2014 (root depth ~10 cm; soil water potential at 10 cm = −30.6 MPa). Furthermore, June 2014 was coincident with their initiation of desiccation-induced dormancy.
Likewise, soil water potentials in June 2014 suggest that summerdeciduous A. jordanii was below its threshold for maintaining hydrated frond tissues (root depth ~20 cm; soil water potential at 20
cm = −4.1 MPa). Note that the soil water potential value of −4.1
MPa is below the seasonal Ψmin of A. jordanii and well below its
osmotic potential at the turgor loss point of −1.8 MPa. Furthermore, these water potential values corresponded to the observation that dehydration-induced defoliation was initiated for this
species in late June. Only the dehydration-tolerant evergreen D.
arguta pictured in Fig. 1B and C had a rooting depth that allowed
water utilization from deep soil horizons, below 20 cm, during the
peak dry season in October 2014 (root depth ~36 cm; soil water
potential at 20 cm = −25.8 MPa). To maintain fronds in an evergreen state and remain seasonally active, roots of D. arguta had to
extend below 20 cm to tap soil moisture at or above a frond Ψmin of
−8 MPa. Dryopteris arguta experienced a Ψmin approaching −8
MPa in October 2014 and an osmotic potential at the turgor loss
point approaching −6 MPa and dark-adapted chlorophyll fluores-
cence approaching 0.2, suggesting that severe dehydration stress
was hampering cellular processes and symplast function (Fig. 1E).
This severe dehydration stress in the symplast was matched by severe dehydration stress for apoplast function that approached the
water potential threshold for xylem dysfunction and loss in water
transport capacity (Ψ80 = −7.57 MPa).
Unlike the four species pictured in Fig. 1B and C, the other four
species of ferns examined in our study display less overlap in their
distribution and thus may not require niche partitioning for coexistence in the same habitat. However, they do appear to segregate along varied microsites as a form of habitat differentiation
(H. Holmlund and S. Davis, personal observations). This latter possibility awaits further investigation in ecological systems that provide a larger number of replicates than available in our Santa Monica
Mountains watershed.
The results are consistent with the suggestion of Sessa and
Givnish (2014) that the evolutionary recent diversification of ferns,
especially Dryopteris species, is more likely due to differences in water relations than differences in light regimes. We found no significant difference in percentage shade between riparian and chaparral
habitats yet significant differences in percentage soil moisture between riparian and chaparral habitats.
If we were to forecast survival and persistence of each of these
fern types with increasing intensity of drought predicted for California by the end of this century (Cook et al., 2015), we would predict that fern species with fronds that are desiccation tolerant will
be the most resilient. Under multidecadal megadrought (cf. Cook
et al., 2015), the frond dehydration-sensitive species will undergo
high mortality because perennial streams will dry up and the frond
dehydration-tolerant species will exceed their evergreen dehydration limits of −9 MPa.
ACKNOWLEDGEMENTS
The authors thank Pepperdine University and the National Science
Foundation for their support (NSF REU Site grant DBI-1062721 to
Jay Brewster at Pepperdine University, NSF grant IOS-1258186 to
J.P.). They also thank Debbie Sharpton and Jo Kitz of the Mountain
Restoration Trust for access to the reserve at Upper Cold Creek. The
authors thank the Southern California Research Learning Center
for grant support to H.I.H, V.M.L., and S.D.D. Special thanks to
Frank Ewers, Marcia Murry-Ewers, Jim Wheeler, and Frank and
Regina Merritt and for research assistance by Paul Chung, Colin
Byrne, Amir Mahmoud, Nathaniel Gruendemann, and Ala Mahmoud.
The authors thank the anonymous reviewers for their constructive
comments.
LITERATURE CITED
Alder, N. N., W. T. Pockman, J. S. Sperry, and S. Nuismer. 1997. Use of centrifugal force in the study of xylem cavitation. Journal of Experimental Botany
48: 665–674.
Anderegg, W. R., J. M. Kane, and L. D. Anderegg. 2013. Consequences of widespread tree mortality triggered by drought and temperature stress. Nature
Climate Change 3: 30–36.
Aschmann, H. 1973. Distribution and peculiarity of Mediterranean ecosystems. In F. di Castri and H. A. Mooney [eds.], Mediterranean type ecosystems, 11–19. Springer, Heidelberg, Germany.
Baer, A., J. K. Wheeler, and J. Pittermann. 2016. Not dead yet: The seasonal water relations of two perennial ferns during California’s exceptional drought.
New Phytologist 210: 122–132.
S E P T E M B E R 2016 , V O LU M E 103
Baldwin, B. G., D. H. Goldman, D. J. Keil, R. Patterson, T. J. Rosatti, and D.
H. Wilken [eds.]. 2012. The Jepson manual: Vascular plants of California.
University of California Press, Berkeley, California, USA.
Brodersen, C., S. Jansen, B. Choat, C. Rico, and J. Pittermann. 2014. Cavitation
resistance in seedless vascular plants: The structure and function of interconduit pit membranes. Plant Physiology 165: 895–904.
Coates, A. R., P. E. Dennison, D. A. Roberts, and K. L. Roth. 2015. Monitoring
the impacts of severe drought on southern California chaparral species using
hyperspectral and thermal infrared imagery. Remote Sensing 7: 14276–14291.
Cook, B. I., T. R. Ault, and J. E. Smerdon. 2015. Unprecedented 21st century drought risk in the American Southwest and Central Plains. Science
Advances 1: e1400082.
Cowling, R. M., F. Ojeda, B. B. Lamont, P. W. Rundel, and R. Lechmere-Oertel.
2005. Rainfall reliability, a neglected factor in explaining convergence and
divergence of plant traits in fire-prone Mediterranean-climate ecosystems.
Global Ecology and Biogeography 14: 509–519.
Davis, S. D. 1991. Lack of niche differentiation in adult shrubs implicates the importance of the regeneration niche. Trends in Ecology & Evolution : 272–274.
Davis, S. D., F. W. Ewers, J. S. Sperry, K. A. Portwood, M. C. Crocker, and G.
C. Adams. 2002. Shoot dieback during prolonged drought in Ceanothus
(Rhamnaceae) chaparral of California: A possible case of hydraulic failure.
American Journal of Botany 89: 820–828.
Davis, S. D., and H. A. Mooney. 1986. Tissue water relations of four co-occurring chaparral shrubs. Oecologia 70: 527–535.
Davis, S. D., R. B. Pratt, F. W. Ewers, and A. L. Jacobsen. 2007. Freezing tolerance impacts chaparral species distribution in the Santa Monica Mountains.
In D. A. Knapp [ed.], Proceedings of the 32nd Annual Southern California
Botanists Symposium: Flora and Ecology of the Santa Monica Mountains in
Fullerton, California, 2006, 159–172. California State University, Fullerton,
Fullerton, California, USA.
Davis, S. D., J. S. Sperry, and U. G. Hacke. 1999. The relationship between xylem conduit diameter and cavitation caused by freezing. American Journal
of Botany 86: 1367–1372.
Frazer, J. M., and S. D. Davis. 1988. Differential survival of chaparral seedlings
during the first summer drought after wildfire. Oecologia 76: 215–221.
Griffin, D., and K. J. Anchukaitis. 2014. How unusual is the 2012–2014
California drought? Geophysical Research Letters 41: 9017–9023.
Hacke, U. G., J. S. Sperry, and J. Pittermann. 2000. Drought experience and
cavitation resistance in six shrubs from the Great Basin, Utah. Basic and
Applied Ecology 1: 31–41.
Hacke, U. G., M. D. Venturas, E. D. MacKinnon, A. L. Jacobsen, J. S. Sperry, and
R. B. Pratt. 2015. The standard centrifuge method accurately measures vulnerability curves of long-vesselled olive stems. New Phytologist 205: 116–127.
Holbrook, N. M., M. J. Burns, and C. B. Field. 1995. Negative xylem pressures
in plants: a test of the balancing pressure technique. Science 270: 1193–1194.
Hutchinson, G. E. 1957. Concluding remarks. Cold Spring Harbor Symposium
on Quantitative Biology 22: 415–427.
Jacobsen, A. L., R. B. Pratt, S. D. Davis, and F. W. Ewers. 2008. Comparative
community physiology: Nonconvergence in water relations among three
semi-arid shrub communities. New Phytologist 180: 100–113.
• H O L M LU N D E T A L. — S E A S O N A L WAT E R R E L AT I O N S I N F E R N S
• 11
Jacobsen, A. L., R. B. Pratt, F. W. Ewers, and S. D. Davis. 2007. Cavitation
resistance among 26 chaparral species of southern California. Ecological
Monographs 77: 99–115.
Jacobsen, A. L., R. B. Pratt, S. D. Davis, and M. F. Tobin. 2014. Geographic
and seasonal variation in chaparral vulnerability to cavitation. Madroño 61:
317–327.
Kolb, K. J., and S. D. Davis. 1994. Drought tolerance and xylem embolism in
co-occurring species of coastal sage and chaparral. Ecology 75: 648–659.
Maherali, H., C. F. Moura, M. C. Caldiera, C. J. Willson, and R. B. Jackson.
2006. Functional coordination between leaf gas exchange and vulnerability
to xylem cavitation in temperate forest trees. Plant, Cell & Environment 29:
571–583.
Meinzer, F. C., P. W. Rundel, M. R. Sharifi, and E. T. Nilsen. 1986. Turgor
and osmotic relations of the desert shrub Larrea tridentata. Plant, Cell &
Environment 9: 467–475.
Pakeman, R. J., R. H. Marrs, and P. J. Jacob. 1994. A model of bracken
(Pteridium aquilinum) growth and the effects of control strategies and
changing climate. Journal of Applied Ecology 31: 145–154.
Pittermann, J., and J. Sperry. 2003. Tracheid diameter is the key trait determining the extent of freezing-induced embolism in conifers. Tree Physiology 23:
907–914.
Pratt, R. B., A. L. Jacobsen, K. A. Golgotiu, J. S. Sperry, F. W. Ewers, and S. D.
Davis. 2007. Life history type and water stress tolerance in nine California
chaparral species (Rhamnaceae). Ecological Monographs 77: 239–253.
Pratt, R. B., A. L. Jacobsen, R. Mohla, F. W. Ewers, and S. D. Davis. 2008.
Linkage between water stress tolerance and life history type in seedlings of
nine chaparral species (Rhamnaceae). Journal of Ecology 96: 1252–1265.
Saruwatari, M. W., and S. D. Davis. 1989. Tissue water relations of three chaparral shrub species after wildfire. Oecologia 80: 303–308.
Sessa, E. B., and T. J. Givnish. 2014. Leaf form and photosynthetic physiology of Dryopteris species distributed along light gradients in eastern North
America. Functional Ecology 28: 108–123.
Sperry, J. S. 1996. Measuring hydraulic conductivity: conduct.ver1.xls [online]. Department of Biology, University of Utah, Salt Lake City, Utah, USA.
Website http://biologylabs.utah.edu/sperry/methods.html#hydraulic_conductivity [accessed 24 August 2016].
Sperry, J. S., J. R. Donnelly, and M. T. Tyree. 1988. A method for measuring
hydraulic conductivity and embolism in xylem. Plant, Cell & Environment
11: 35–40.
Sperry, J. S., and U. G. Hacke. 2002. Desert shrub water relations with respect
to soil characteristics and plant functional type. Functional Ecology 16:
367–378.
Thomas, C. M., and S. D. Davis. 1989. Recovery patterns of three chaparral
shrub species after wildfire. Oecologia 80: 309–320.
Tyree, M. T., and H. T. Hammel. 1972. The measurement of the turgor pressure and the water relations of plants by the pressure-bomb technique.
Journal of Experimental Botany 23: 267–282.
Watt, A. S. 1954. Contributions to the ecology of bracken (Pteridium aquilinum). Frost and the advance and retreat of bracken. New Phytologist 53:
117–130.