Oecologia
9 Springer-Verlag1986
Oecologia (Berlin) (1986) 70:172-t 77
Water use patterns of four co-occurring chaparral shrubs
S.D. Davis* and H.A. Mooney
Department of Biological Sciences, Stanford University, Stanford, CA 94305, USA
Summary. Mixed stands of chaparral in California usually
contain several species of shrubs growing close to each other
so that aerial branches and subterranean roots overlap.
There is some evidence that roots are stratified relative to
depth. It may be that root stratification promotes sharing
of soil moisture resources. We examined this possibility by
comparing seasonal water use patterns in a mixed stand
of chaparral dominated by four species of shrubs: Quercus
durata, Heteromeles arbutifolia, Adenostoma fasciculatum,
and Rharnnus californica. We used a neutron probe and
soil phychrometers to follow seasonal depletion and recharging of soil moisture and compared these patterns to
seasonal patterns of predawn water potentials, diurnal leaf
conductances, and diurnal leaf water potentials. Our results
indicated that 1) Quercus was deeply rooted, having high
water potentials and high leaf conductances throughout the
summer drought period, 2)Heteromeles/Adenostoma were
intermediate in rooting depth, water potentials, and leaf
conductances, and 3) Rhamnus was shallow rooted, having
the lowest water potentials and leaf conductances. During
the peak of the drought, predawn water potentials for Quercus corresponded to soil water potentials at or below a
depth of 2 m, predawn water potentials of Heteromeles/
Adenostorna corresponded to a depth of 0.75 m, and predawn water potentials of Rhamnus corresponded to a depth
of 0.5 m. This study supports the concept that co-occurring
shrubs of chaparral in California utilize a different base
of soil moisture resources.
Key words: Water use - Root distribution - Water potential
predawn - Leaf conductance - Quereus durata
Heteromeles arbutifolia- Adenostorna fasciculatum - Rhamnus californica
Shrubs of the chaparral of California often form a closed,
overlapping canopy of uniform height. Roots also overlap
considerably in their horizontal distribution and extend laterally 2-3 m beyond the aerial branches of each shrub
(Kummerow et al. 1977; Kummerow and Mangan 1981).
The vertical distribution of roots, however, is evidently
* Present address: Natural Science Division, Pepperdine University, Malibu, CA 90265, USA
Offprint requests to : S.D. Davis
stratified. Direct evidence of root stratification comes from
measurements of rooting depth along road cuts and by measurements of rooting depth after hydraulic excavation
(Hellmers et al. 1955). Indirect evidence comes from measurements of drying and rewetting patterns of soil and plant
tissue during drought cycles in mixed stands of chaparral
(Rowe and Reimann 1961; Poole and Miller 1975; Burk
1978).
Differences in rooting habit and depth have been described for other plant communities and these differences
have been interpreted as means of sharing soil resources
(Cole and Holch 1941 ; Weaver and Darland 1949; Wieland
and Bazzaz 1975). The considerable depth of penetration
which has been noted for certain chaparral shrub species
(Hellmers et al. 1955) would indicate a particularly high
potential for resource sharing among species. In this study
we examine the realization of this potential.
Study site
Our study site was a mixed stand of chaparral occupying
a level portion of the northwest ridge crest of Jasper Ridge
Biological Preserve, 12 km west of Palo Alto, California.
The chaparral study site is 175 m above sea level and located on serpentine-derived soil (Page and Tabor 1967).
The chaparral is bordered by grassland with a 3 5 m transition zone (barezone) containing little vegetation.
The distribution of shrubs at our study site is illustrated
by a vegetation profile running 50 m from the transition
zone toward the interior of the chaparral stand (northwest
direction, perpendicular to the chaparral/grassland boundary, Fig. 1). Shrubs found along this transect were Quercus
durata Jeps., Adenostoma fasciculatum H. & A., Heteromeles arbutifolia M. Roem., Rhamnus californica Esch., and
Toxicodendron diversilobum (T. & G.) Green. Nomenclature follows Munz (1974). Shrub height decreased toward
the interior (away from the transition zone) of the stand
of chaparral (Table1). Adenostoma density increased
whereas Rhamnus and Toxicodendron densities decreased
toward the interior (Table 1).
A more extensive description of the study site with aerial
photographs is given elsewhere (Davis and Mooney 1985).
An extensive description of the chaparral at Jasper Ridge
is given by Cooper (1922). The climate is a mediterranean
type, with cool, wet winters and hot, dry summers.
173
5
E
4
3
,~.~
91-
~
;~. ,
:.. '
~,J ~ , ~ ~ ' ~ 1 .
f-,._
I
o
o
\
\
\ ,b\ \
\
Distonce along tronsecl' (rn)
~__To•
\ \
\
X_Rhamnus
\ \
X_Heferomeles
\ X_Quercus
X _ Adenostoma
\
\
\
\
\
\
Fig. t. Vegetation profile of shrubs
along a 5 m wide by 50 m long
transect, running perpendicular to the
barezone toward the interior of our
chaparral stand
\
Table 1. Mean height and predawn water potentials of chaparral shrubs along the 50 m transect shown in Fig. 1.
Predawn water potentials were taken on 24 August 1984
Distance along transect
(0-10 m)
Shrub
Quercus
Adenostoma
Heteromeles
Rhamnus
Toxicodendron
(10-20 m)
Plant height (m) (mean/SE/n)
3.1+0.3 (5)
2.9___0.2 (13)
2.5 • 0.2 (12)
2.4 • 0.2 (11)
2.8+0.1 (2)
2.1+0.2 (3)
1.5_+0.3 (10)
1.3•
(5)
2.7•
(3)
2.2_+0.2 (2)
(20-30 m)
(30-40 in)
(40-50 m)
1.8•
(3)
2.0_+0.1 (31)
1.0 (1)
1.6_+0.2 (6)
1.9•
(19)
1.4•
(6)
1.9_+0.1 (8)
2.0•
(16)
1.6 (1)
Distance along transect
Shrub
Quercus
Adenostoma
Heteromeles
(5-15 m)
(30-40 m)
Predawn water potential (MPa) (mean/SE/n)
- 0.32 • 0.02 (6)
- 1.84 __0.21 (6)
- 2.51 • 0.23 (6)
- 1.23 • 0.06 (6) *
- 3.10 • 0.06 (6) *
- 3.63 4- 0.21 (6) * *
Significantly different by Student t-test at: * P < 0.001, * * P < 0.01
Materials and methods
Vegetation profile
The distribution o f shrubs at the study site was analyzed
by m a k i n g a series of five transects (50 m long, 5 m wide
and spaced 10 m apart) parallel to the barezone ( N W to
SE direction) and three transects (50 m long, 5 m wide and
spaced 10 m apart) perpendicular to the barezone (NE to
SW direction). The perpendicular transects started at the
barezone and ran 50 m towards the interior o f the chaparral
stand. W e recorded the location, m a x i m u m height, and
m a x i m u m width o f each shrub growing along each transect.
The results are consolidated into the vegetation profile
shown in Fig. 1.
in the chaparral. The distance between the series of barezone and chaparral access tubes was a b o u t 8 m. W e used
a neutron p r o b e (Hydroprobe, M o d e l 503, Campbell Pacific Nuclear Corp., Campbell, California), to m o n i t o r soil
moisture at 0.2 m intervals in depth, down to 2 m, in each
o f these tubes. W e measured soil moisture two to three
times each m o n t h from 7 July 1981 to 15 N o v e m b e r 1982.
Nine soil psychrometers (Wescor, M o d e l PCT-55-30,
Logan, U t a h ) were installed at 0.25 m intervals in depth
down to 2 m, equidistant between two o f the access tubes
in the chaparral. We used a dewpoint microvoltmeter (Wescot, M o d e l HR-33T) to measure soil temperature and soil
water potential, at dawn, two to three times each month,
from 14 A u g u s t 1981 to 22 October 1982.
Soil moisture
Microclimate
W e installed access tubes 2.2-2.4 m deep, for neutron probe
determination o f soil moisture, by a m e t h o d described previously (Davis and M o o n e y 1985). Three tubes, spaced 8 m
apart, were placed in the barezone (these were used as a
control to indicate moisture content in soil devoid o f
shrubs), and three tubes, spaced 8 m apart, were placed
W e placed a shielded m a x i m u m / m i n i m u m thermometer
(Taylor, TM45-C, WeatherMeasure, Sacramento, California) 0.2 m above ground level, in the understory of shrubs,
adjacent to our access tubes. M a x i m u m / m i n i m u m temperatures were recorded one to two times a week from 5 October
1981 to 27 September 1982.
174
I00
50
g_
o2
"~
~D 0
5o
,o~
.->o~ 50
O~
i~
n50
9
.
O
--
120
:,.g,%%"%
[1::
j~d "~q "
\ . . ~
40
0 .~
O
0
N
D
d
F
M
A
M
d
J
A
S
0
Fig.2. Relative shoot elongation, precipitation, and
maximum/minimum air temperature during the
seasonal drying cycle of 1982
<~
Month of year (1981-1982)
We installed a rain gauge (Clear-Vu Rain Gage, P562,
WeatherMeasure) in the grassland adjacent to our study
site and recorded precipitation after each major storm,
Measurements on plants
Measurement of plant parameters was restricted to shrubs
growing between our soil moisture tubes in the chaparral.
This facilitated a comparison of soil water to plant water
status. Since the canopy of these same shrubs overlapped,
we assumed that their roots also overlapped in their horizontal extension (Kummerow and Mangan 198/) and hence
the roots of experimental plants experienced the same soil
environment and branches experienced the same aerial environment.
We made measurements on all representative species
of shrubs growing in our study area, except Toxicodendron
diversilobum.
Shoot elongation. Growth of our plants was monitored by
tagging 20 branches of each species and measuring shoot
elongation each month from October 1981 to October 1982.
Predawn water potentials. Leaf water potentials (Scholander
et al. /965) on Heteromeles and stem water potentials on
Quercus, Adenostoma and Rhamnus, were measured at predawn, two to three times each month, from 24 May 1981
to 9 December 1982. We made three to eight replicate measurements per species using a pressure chamber (Model 3000, Soilmoisture Inc., Santa Barbara, California) and
compressed nitrogen gas.
Diurnal leaf conductances and leaf water potentials. We
made diurnal measurements of leaf conductance to water
vapor diffusion using a steady state porometer (Model L1/600, LiCor lnc., Lincoln, Nebraska). Measurements were
made at monthly intervals, from August /981 to August
1982. We used 6-12 replicate measurements on fully expanded and illuminated leaves of each species at 2 h intervals from dawn to dusk. Paired measurements in early summer on 2 mo and 1 yr 2 mo old leaves showed no significant
difference in leaf conductance between the two age classes.
Therefore, the data for old and young leaves were pooled.
Because the weather at our study site remained similar for
several days at a time, we were able to select days with
clear skies throughout to make diurnal measurements of
leaf conductance and water potential.
On several occasions, we used tissue paper (Kimwipe,
Kimberly-Clark Corp., Neenah, Wisconsin) to remove dew
from leaves during early morning hours, before attaching
our porometer. This procedure was successful for all plants
except Adenostoma, which had dew condensed under scales
at the base of each leaf fascicle. Thus, on days when dew
was present, measurements of leaf conductance on Adenostoma were not taken until 1000 h after all dew had evaporated.
Results
Microelimate and phenology
The climate at our study site is a typical mediterranean
type with a moderate wet winter and hot, dry summer
(Aschmann /973). 1,073 mm of precipitation fell between
September and May of 1981-1982 (Fig. 2). No rain occurred during the summer months (May-September) of
1981, except for 3.3 mm on June 30 (Fig. 2). Rains ceased
in mid-April, about the same time that air temperatures
increased and shoots began their most rapid elongation
(Fig, 2).
All species of shrubs were spring growers, with Quereus
being delayed somewhat. Quercus was the only shrub whose
shoot elongation was initiated from overwintering buds.
The phenology was characteristic of sclerophyllous evergreen shrubs which have minimal growth in winter when
photoperiods are short and temperatures are cool (Fig. 2)
and maximal growth in spring when temperature, moisture,
and photoperiod are more conducive for growth (cf.
Mooney et al. 1975; Mooney/981).
Below ground environment
Soil temperature. Soil temperatures (at dawn) varied with
season and with soil depth (Fig. 3). Soil temperature pro-
175
Soil temperature (~
14
I0
6
I
0
I
I
Jan
Feb Apr
9
1982)
18
i
,
i
,
,
d
J
A
S
0
22
I
Moy Jun
--%.
,
I
dul Aug
-6
2.0
:5.0
E
.e
1
\
CL
1.2
,g
l!l
1
////
t-
"5
oo
*6
I
O9
o
I
Fig. 3. Profiles of Predawn soil temperature measured via soil psychrometers in the chaparral, during the seasonal drying cycle of
1982
34
~.// -, ~..:/ ~
2~
"~1 2.4
__
.o_
*6
A
Borezone
3o
_2.Ore
26 ~.gS.~y.,~.~:../y ~ ~a ~ , % .
Heteromeles
-120
2.8
- 80
~.2
3.6
- 40
A
1981
,
"~ 34
9
,-<j
12
Chaparral. . . . . . . . . . . . .
..~....~
.~
....;A, ,. .."~.
...-, ', .....-,, ,-
=o 30
e3
B" /
~,
t~ ,
..
14
o
_
I0
"~176
a
2.0m
~':
-
....
.
/
/
,
.
II
,
,'"
0
,e
:o
.
c
2
-120
2.8
--80
1982
3.6
:
...
2.4
M
.
~'---~5--er-~-
-%_;,.o
I.Om
IO
.
0...
~g~
.4 "~ll~Q~>0
~
2Z18
N
.4
I
i
I
M
Month of year (1982)
Fig. 5. Soil water potential in the chaparral during the seasonal
drying cycle of 1982
///
[
5.0
A
///
2.0
4.0
,]
~1
..,
~
~
. m ..
6
I~
[)
0
Month of y e a r
Fig. 6A, B. Predawn water potential of chaparral shrubs during
the drying and rewetting cycle of (A) 1981 and (B) 1982
S
"~
.I
,,I
.,I
I
I
/-.a2m
50 0 E
~
50 .-=
A S 0 N D d F M A M d d A S 0 N D
Month of year (1981- 1982)
Fig. 4A, B. Seasonal changes in soil moisture in (A) the barezone
and (B) the chaparral from January 1981 to December 1982
files were not steep; the largest gradient between 0.1 m and
2.0 m was 7.2 ~ C on J a n u a r y 1982 (Fig. 3), the mean gradient was 3.1 ~ C + 0 . 3 5 (SE, n = 2 4 ) . The most r a p i d seasonal increase in soil temperature occurred between April
and M a y , c o n c o m i t a n t with the most r a p i d increase in air
temperature, increase in shoot elongation (Fig. 2), and decrease in soil moisture (Fig. 4 B).
Soil moisture. During summer, at 0.2 m, soil moisture was
depleted more rapidly in the barezone than in the chaparral,
presumably a result o f enhanced evaporation from the exposed soil surface of the barezone (Fig. 4 A , B). A t 1 m
and 2 m there was no moisture depletion in the barezone
which indicated a lack of r o o t activity at these depths. In
contrast, in the chaparral, moisture was depleted at all levels
down to 2 m. Thus, the barezone served as a type of control
at our study site, where the activity of deep roots o f shrubs
was separated from surface evaporation.
Soil water potential. During the summer drought o f 1982
soil water potential declined in the chaparral, at all depths,
in a fashion parallel to that o f soil moisture (Fig. 5).
Plant responses
Predawn water potential. The decline in predawn water potentials of our shrubs was similar during the summer
droughts o f 1981 and 1982 (Fig. 6A, B). Both the drying
176
9
'
Quercus
9 . . . . .
/
12
'E
8
E
9 ......
o ........
Adenostome
, . . . . ' Aug' 1981
B
;
Heferomete5
Rhomnus
' 0ct'1981t
C
'Dec'1981
,,', 4
~
t
~ -o%,~'@"~,
1
,
i 9
0
i
,.~.,
i
Q..
i
.4 ..:
~. 8, -A- -4- -A" /
I
O
'
I
E '
A p r i l 1982
I
I
dune1982
,I'
E 8
E
-/\/L\
ig.. d-~r ,.-,t ..11"' l , i t
,,4
~'/~-.~lg.:~=,::~
0
--L
I
I
8
I
~4
~8
" - -,I.- -A.
~ .2. .....
" "'-.%
.
o...o...o.-.,s',
~...
~x-"~-----._.../..x.~'
"-.o.."
2
I
;..: Z,u /
/ \ 98zt
-
12
v
I
zz4
"A..A_.8
a
14
-9
~"
Ik "
o...~ ~IK
t8
zz 4
p
8
.f;.-.I'~
"0"1""0"" "0"1"~"
t4 18 2z
Hour of day (PST)
Fig. 7A-F. Diurnal changes in leaf conductance (g s) and water
potential (~/) of chaparral shrubs from August 1981 to August
1982
cycle in early summer and the rewetting cycle in early fall
indicated that rooting depths were shallow for Rhamnus,
intermediate for Heteromeles and Adenostoma and deep for
Quercus.
By the end of the summer drought of 1982 (23 October),
soil water potentials were - 2 . 8 MPa at 0.5 m, - 1 . 8 MPa
at 0.75 m, and - 0 . 6 MPa at 2 m (Fig. 5). These values corresponded to predawn water potentials of Rhamnus, Heteromeles/Adenostoma, and Quercus respectively (Fig. 6B).
The 24 m m rain in September of 1982 was detected by
our soil phychrometers at 0.1 m (Fig. 5) and by our neutron
probe at 0.2 m (Fig. 4A, B), indicating a penetration of
moisture barely to 0.2 m. Predawn water potentials of
Rhamnus and Heteromeles increased in response to this precipitation, indicating that roots were active at shallow
depths in these species. In contrast, there was no response
in predawn water potential by Adenostoma and Quercus.
Diurnal leaf conductances and plant water potentials. Diurnal
leaf conductances and plant water potentials were highest
during spring and lowest during the summer drought for
all shrubs except Quercus (Fig. 7). Leaf conductances of
Quercus were highest during the peak of the drought, when
conductances of other shrubs, especially Rhamnus, were
lowest (Fig. 7 A, F). The most negative diurnal water potential for Quercus was - 1 . 1 MPa, at a time when the water
potential of other shrubs was below - 3 . 0 MPa (Fig. 7A),
a value approaching the turgor loss point of bulk leaf or
stem tissue (Davis and Mooney 1986).
Diurnal water potentials were consistently highest for
Quercus and leaf eonductances were consistently lowest for
Rhamnus (Fig. 7). For the driest months of the year, June
through September (Figs. 4, 5, 6), Rhamnus opened its sto-
mates only during the early morning hours and dosed them
the remainder of the day (Fig. 7A, B, E, F). During these
same months, diurnal water potentials for Rhamnus reached
or exceeded the point at which turgor was lost in the bulk
tissue (Davis and Mooaey 1986), indicating that stomatal
closure was probably a result of passive water loss.
Thus, patterns of diurnal leaf conductance and plant
water potential also indicated that at our study site Quercus
had deep roots, Rhamnus shallow roots, and Heteromeles/
Adenostoma intermediate roots.
Discussion
Our results show a dramatic stratification of available soil
moisture with depth in a mixed stand of chaparral
(Fig. 4B), which does not occur in adjacent soil devoid of
shrubs (Fig. 4A). This stratification becomes more pronounced as the summer drought progresses (Fig. 5).
If shrubs differed in their rooting depth, one would predict differences in water availability and the degree of seasonal water stress, as indeed occurs (Fig. 6). Some species
(Quercus) see virtually no water stress throughout the year,
whereas others (Rhamnus) come under considerable stress
(Figs. 6, 7). Independent evidence that this indicates differential rooting depth is a) predawn water potentials of the
various species match soil water potentials at different
depths (Figs. 5, 6) and b) the pattern of change of plant
water potential with time; those which appear to be shallow
rooted by the amount of stress they undergo are also those
which respond first to a new input of soil moisture (Fig. 6 B;
cf. Wieland and Bazzaz 1975; Poole and Miller 1975; Burk
1978). This being true, one would also predict differences
in seasonal patterns of leaf conductance in response to
plants utilizing different water supplies. Again, this is what
is seen (Fig. 7). Quercus, the plant presumably getting water
from considerable depth has maximum leaf conductance
during the height of the drought. At this time, Rhamnus,
which by other evidence has the shallowest roots, has its
stomates almost closed. Heteromeles and Adenostoma fall
in between these two patterns (Fig. 7).
Burk (1978) found in a mixed stand of chaparral in
southern California (L.A. Co.) that Quercus dumosa (a
scrub oak similar to Q. durata but which does not grow
in serpentine soil) had a seasonal minimum in predawn
water potential of - 0 . 6 MPa whereas Adenostoma had a
seasonal minimum of - 1 . 3 MPa, values similar to ours
(Fig. 6). Other studies in the chaparral of California suggest
a deep rooting habit for Quercus (Cooper 1922; Hellmers
etal. 1955; Rowe and Reimann 1961). Kummerow and
Mangan (1981) present evidence that the rooting depth of
Quercus dumosa is less plastic than that of Adenostoma
because Quereus only establishes in deep soil whereas Adenostoma establishes in either deep or shallow soil.
Poole and Miller (1975) compared coastal and inland
stands of mixed chaparral of southern California (San
Diego Co.) and found that shrubs could be categorized
relative to rooting depth: three species of Rhus had deep
roots, Adenostoma and Heteromeles had roots of intermediate depth, and Arctostaphylos glauca and Ceanothus greggii
had shallow roots. In our study, Quereus durata was the
counterpart of Rhus in their study, and Rhamnus the counterpart of Arctostaphylos and Ceanothus. This indicates that
root stratification in chaparral may be common and may
provide a mechanism by which shrubs share soil resources.
177
Poole and Miller (1975) found that c h a p a r r a l shrubs
with the shallowest roots h a d leaves that were the most
xerophytic and stomates that were the least sensitive to
water stress. In contrast, in our study, the shrub with the
shallowest roots (Rhamnus) was the least xerophytic and
the most sensitive to water stress (Davis and M o o n e y 1986;
cf. Williams 1983). The differences between these two studies m a y be because Rhamnus is not a typical chaparral
shrub at our study site, but establishes as an understory
plant beneath Quereus and HeteromeIes and does poorly
once its branches emerge from the overstory o f these larger
shrubs (Fig. 1). Two observations support this conclusion.
First, we observed at the end o f summer drought in 1981,
that m a n y larger Rhamnus shrubs h a d m a j o r branches that
were dying. The first sign o f dying was an exceptionally
low water potential (below - 6.0 MPa), subsequently leaves
rolled on these branches a n d eventually abscised. We m a d e
thin sections o f stems and leaves, followed by microscopic
inspection, b u t found no evidence of fungal hyphae or necrotic tissue. W e suspect that m a j o r branches, which had
emerged from the overcanopy o f Quercus and Heteromeles,
had experienced such severe water stress that xylem vessels
cavitated, cutting off water supplied by the roots. The next
season m a n y o f these same plants were in the process o f
resprouting new branches from a r o o t crown. Second, the
distribution o f Rhamnus at our study site was restricted
to a n a r r o w belt a b o u t 20 m wide, running parallel to the
barezone (Fig. 1). All shrubs were taller and more dense
along this belt and h a d higher water potentials than their
counterparts t o w a r d the interior o f our stand o f chaparral
(Table 1). This indicated that the threshold for Rhamnus
survival was p r o b a b l y exceeded by the more open canopy
and greater water stress characteristic o f the interior o f our
stand of chaparral.
Conclusion
This study supports the concept that co-occurring shrubs
utilize a different resource base. There are a number o f
implications in these patterns. C h a p a r r a l plants are presumed to be limited in their metabolic activity by the summer drought on the one h a n d and cool temperature on
the other ( M o o n e y et al. 1975; M o o n e y 1981). Rhamnus
provides an example of this pattern. It is severly drought
limited in the summer. Quercus on the other h a n d has access
to deep water and shows relatively high conductance values
even during the extreme drought period. The summer values
are indeed higher than those occurring during the winter
when plants m a y be thermally limited. Since conductance
is positively correlated with photosynthetic capacity
( M o o n e y et al. 1975) this means that these species have
very different seasonal patterns o f carbon gain. One would
predict that Quercus fixes most o f its carbon in the summer
period in contrast to Rhamnus which m a y be comparatively
more winter active. This then brings the interesting question
o f why the stem growth periods of all species occurs during
the same time since all o f the species obviously h a v e different resource acquisition patterns. There are a number o f
possibilities. One, it m a y be that although Quercus is able
to get water at depth during the summer, nutrients are quite
limited for roots obtaining this water. It is only in the spring
when both water and nutrients (particularly nitrate) are
available and hence growth is restricted to this time in all
species. A n o t h e r possibility, which is more difficult to test,
is that having synchronous growth with other species dilutes
the potential herbivore impact on the newly expanding,
most vulnerable tissue.
Acknowledgements. We thank David Hollinger and George Koch
for helping in the installation of access tubes for our neutron probe,
Alan Grundmann for logistic support, Barbara Lilly for preparation of illustrations, and Pepperdine University for granting sabbatical leave to S.D. Davis. This study was supported by NSF
grants BSR-831-5675 to H.A. Mooney and SPI-816-5080 to S.D.
Davis.
References
Aschmann H (1973) Distribution and peculiarity of mediterranean
ecosystems. In: diCastri F, Mooney HA (eds) Mediterranean
ecosystems: origin and structure. Springer, Berlin Heidelberg
New York, pp 11-19
Burk JK (1978) Seasonal and diurnal water potentials in selected
chaparral shrubs. Am Midl Nat 99:244-248
Cole HE, Holch AE (1941) The root habits of certain weeds of
southeastern Kansas. Ecology 22:141 147
Cooper WS (1922) The broad sclerophyll vegetation of California.
Carnegie Inst Wash Pub, p 122
Davis SD, Mooney HA (1985) Comparative water relations of
adjacent California shrub and grassland communities. Oecologia (Berlin) 66 : 522-529
Davis SD, Mooney HA (1986) Tissue water relations of four cooccurring chaparral shrubs. Oecologia (Berlin) (in press)
Hellmers H, Horton JS, Juhren G, O'Keefe J (1955) Root systems
of some chaparral plants in southern California. Ecology
36 : 667-678
Kummerow J, Krause D, Jow W (1977) Root systems of chaparral
shrubs. Oecologia (Berlin) 29 : 163-177
Kummerow J, Mangan R (1981) Root systems in Quercus dumosa
Nutt. dominate chaparral in southern California. Acta Oecologia Plant 2:177-188
Mooney HA (1981) Primary production in mediterranean-climate
regions. In: diCastri F, Goodall DW, Specht RL (eds) Mediterranean-type shrublands. Elsevier, Amsterdam, pp 249 255
Mooney HA, Harrison AT, Morrow PA (1975) Environmental
limitations of photosynthesis on a California evergreen shrub.
Oecologia (Berlin) 19:293-301
Munz PA (1974) A flora of southern California. Univ Calif Press,
Berkeley, p 1086
Page BM, Tabor LL (1967) Chaotic structure and Decollement
in cenozoic rocks near Stanford University, California. Geol
Soc Am Bull 78:1-12
Poole DK, Miller PC (1975) Water relations of selected species
of chaparral and coastal sage communities. Ecology
56:1118-1128
Rowe PB, Reimann LF (1961) Water use by brush, grass, and
grass-forb vegetation. J Forestry 59:175-181
Scholander PF, Hammel HT, Bradstreet ED, Hemmingsen HA
(1965) Sap pressure in vascular plants. Science 148:33%346
Weaver JE, Darland RW (1949) Soil-root relationships of certain
native grasses in various soil types. Ecol Monogr 19:303 338
Wieland NK, Bazzaz FA (1975) Physiological ecology of three
codominant successional annuals. Ecology 56:681-688
Williams WE (1983) Optimal water-use efficiency in a California
shrub. Plant, Cell and Environ 6:145-151
Received April 18, 1986