Water supply and tree growth. Part I Water deficits


Kozlowski, T.T.

Forestry Abstracts 43(2): 57-95

1982


Water supply is the most important environmental factor determining distribution, species composition, and growth of forests. Net annual primary production of forests varies from as much as 3000 g/m2 in wet regions to negligible amounts in dry regions. The water balance of trees has been characterized by visible wilting, tissue moisture content, relative water content, saturation deficit, and water potential. Problems in use of these measures are discussed. Water deficits develop readily in forest trees, even in trees growing in wet soil, because of excess transpiration over absorption of water. Water deficits adversely affect seed germination and cause shrinkage of leaves, stems, roots, fruits, and cones. Shrinkage of tissues influences stomatal aperture, photosynthesis, movement of gases, growth measurement, dissemination of pollen and seeds, flow of latex and oleoresins, and absorption of water and ions. Some of the decrease in photosynthesis during drought is the result of increased resistance to diffusion of CO2 to chloroplasts and some to decrease in photosynthetic capacity. Water deficits inhibit shoot growth, wood production, and root growth. Yield of fruits and seeds can be inhibited at various stages of reproductive growth such as flower bud initiation, anthesis, pollination, fertilization, embryo growth, and fruit and seed enlargement. Water deficits may also induce leaf scorching and abscission, dieback of twigs and branches, and drought cracks. Severe water deficits often kill trees. Drought tolerance of trees may reflect desiccation avoidance or desiccation tolerance, with the former much more important. Desiccation avoidance is the result of one or more adaptations in leaves, stems, or roots. Among the most important of these are reduction in number and size of leaves; small, few, and sunken stomata; rapid stomatal closure; abundant leaf waxes; leaf shedding during droughts; extensive root development; capacity for twig and stem photosynthesis; living wood fibres; and strong development of palisade mesophyll. In some species osmotic adjustments result in maintenance of leaf turgor during drought periods.

CAB
COMMONWEALTH
FORESTRY
BUREAU
Forestry
Abstracts
February,
1982
Vol.
43
Review
Article
No.
2
Water
Supply
and
Tree
Growth
Part
I
Water
Deficits
T.
T.
KOZLOWSKI
Director
of
the
Biotron
and
A.J.
Riker
Professor
of
Forestry
University
of
Wisconsin
Madison,
Wisconsin
53706,
USA.
Contents
Summary
Summary
Page
I.
Introduction
57
II.
Sources
of
Water
58
III.
Measurement
of
Water
Deficits
59
IV.
Development
of
Water
Deficits
60
V.
Responses
of
Trees
to
Water.
Deficits
64
VI.
Drought
Tolerance
77
Water
supply
is
the
most
important
environmental
factor
determining
distribution,
species
composition,
and
growth
of
forests.
Net
annual
primary
production
of
forests
varies
from
as
much
as
3000
g/m
2
in
wet
regions
to
negligible
amounts
in
dry
regions.
The
water
balance
of
trees
has
been
characterized
by
visible
wilting,
tissue
moisture
content,
relative
water
content,
saturat-
ion
deficit,
and
water
potential.
Problems
in
use
of
these
measures
are
discussed.
Water
deficits
develop
readily
in
forest
trees,
even
in
trees
growing
in
wet
soil,
because
of
excess
transpiration
over
absorption
of
water.
Water
deficits
adversely
affect
seed
germination
and
cause
shrinkage
of
leaves,
stems,
roots,
fruits,
and
cones.
Shrinkage
of
tissues
influences
stomatal
aperture,
photo-
synthesis,
movement
of
gases,
growth
measurement,
dissemination
of
pollen
and
seeds,
flow
of
latex
and
oleoresins,
and
absorption
of
water
and
ions.
Some
of
the
decrease
in
photosynthesis
during
drought
is
the
result
of
increased
resistance
to
diffusion
of
CO
2
to
chloroplasts
and
some
to
decrease
in
photosynthetic
capacity.
Water
deficits
inhibit
shoot
growth,
wood
production,
and
root
growth.
Yield
of
fruits
and
seeds
can
be
inhibited
at
various
stages
of
reproductive
growth
such
as
flower
bud
initiation,
anthesis,
pollination,
fertilization,
embryo
growth,
and
fruit
and
seed
enlargement.
Water
deficits
may
also
induce
leaf
scorching
and
abscission,
dieback
of
twigs
and
branches,
and
drought
cracks.
Severe
water
deficits
often
kill
trees.
Drought
tolerance
of
trees
may
reflect
desiccation
avoidance
or
desiccation
tolerance,
with
the
former
much
more
important.
Desiccation
avoidance
is
the
result
of
one
or
more
adaptations
in
leaves,
stems,
or
roots.
Among
the
most
important
of
these
are
reduction
in
number
and
size
of
leaves;
small,
few,
and
sunken
stomata;
rapid
stomatal
closure;
abundant
leaf
waxes;
leaf
shedding
during
droughts;
extensive
root
development;
capacity
for
twig
and
stem
photosynthesis;
living
wood
fibres;
and
strong
development
of
palisade
mesophyll.
In
some
species
osmotic
adjustments
result
in
maintenance
of
leaf
turgor
during
drought
periods.
Introduction
The
distribution
of
forests,
their
species
composition,
and
productivity
are
controlled
chiefly
by
too
little
or
too
much
water
(Kramer
&
Kozlowski
1979).
In
the
equatorial
zone,
where
mean
temperatures
are
similar
throughout
the
year,
a
variety
of
vegetation
types
from
semi-desert
to
rain
forest
may
be
found.
As
the
amount
of
rainfall
increases
the
number
of
woody
species
increases
57
58
Forestry
Abstracts
1982
Vol.
43
No.
2
until
deciduous
forests
predominate,
Where
there
is
even
more
rain
and
a
shorter
dry
season,
semi-
evergreen
forests
occur.
The
wettest
portions
of
the
equatorial
zone,
with
no
distinct
dry
seasons,
support
lush
rain
forests
(Walter
1973).
In
the
corresponding
summer
dry,
winter
wet,
or
mediterr-
anean-type
climates
of
Chile
and
California,
the
dominant
growth
forms
are
similar
at
about
the
same
position
along
a
moisture
gradient
extending
from
about
1600
to
80
mm.
annual
rainfall
(Mooney
&
Dunn
1970).
Evergreen
forest
dominates
at
the
wet
end
of
the
gradient.
Towards
the
dry
end
of
the
gradient,
the
forest
is
replaced
by
dense
evergreen
shrubs,
then
by
an
open
drought-deciduous
scrub
community,
and
finally
by
an
open
community
made
up
of
drought-
deciduous
shrubs
as
well
as
many
succulents
(Mooney
et
al.
1970).
Species
diversity
is
greater
in
rain
forests
than
in
any
other
type
of
forest.
In
lowland
rain
forests
there
are
many
more
species
of
trees
and
shrubs
than
herbaceous
plants.
The
high
species
diversity
is
related
to
reproductive
isolation
associated
with
non-synchronous
flowering
(a
con-
sequence
of
the
continuously
favourable
climate)
and
low
dissemination
efficiency
(a
result
of
the
need
for
disseminules
well
supplied
with
food
to
counteract
the
severe
competition
that
seedlings
undergo)
(Daubenmire
1978).
Species
composition
of
forests
depends
not
only
on
the
total
amount
of
rainfall
but
also
on
its
annual
variability
and
seasonal
distribution.
In
the
temperate
zone
dominance
of
mesophytic
species
increases
during
a
series
of
wet
years
and
decreases
during
several
consecutive
dry
years
(Weaver
&
Albertson
1936).
The
importance
of
seasonal
distribution
of
rainfall
is
emphasized
by
failure
of
a
phenome,nally
high
annual
rainfall
(11
610
mm)
at
Cherrapungi,
India
to
support
mesophytic
forest.
Almost
all
of
the
rain
falls
in
8
consecutive
months,
with
the
other
4
months
receiving
a
total
of
less
than
100
mm.
When
more
evenly
distributed
throughout
the
year
much
less
rain
than
falls
at
Cherrapungi
can
support
rain
forest
(Daubenmire
1978).
Aridity,
which
characterizes
approximately
a
third
of
the
world's
land
area,
prevents
tree
growth
from
approaching
a
maximum.
Over
most
of
the
remaining
land
area,
tree
growth
is
reduced
in
varying
amounts
by
periodic
droughts.
Annual
net
primary
productivity
varies
from
as
much
as
3000
g/m
2
in
areas
with
abundant
and
well
distributed
rainfall
to
250
to
1000
g/m
2
in
semi-arid
regions,
and
only
25
to
400
g/m
2
in
arid
regions
(Fischer
&
Turner
1978).
The
extent
of
growth
loss
resulting
from
water
stress
is
often
not
realized
because
of
the
difficulty
of
showing
how
much
more
growth
would
occur
if
the
internal
water
balance
of
forest
trees
were
maintained
at
a
favourable
level
throughout
the
entire
growing
season
(Kozlowski
1968d).
It
cannot
be
emphasized
too
strongly
that
even
trees
in
rain
forests
recurrently
undergo
periods
of
internal
water
stress
which
reduce
their
growth.
Young
seedlings
are
particularly
sensitive
to
water
deficits
(Kozlowski
1976c).
In
Australia,
germination
of
Eucalyptus
seeds
exceeded
80%
but
seedling
survival
was
less
than
I%
because
of
desiccation
(Jacobs
1955).
In
Zimbabwe
all
young
Brachy-
stegia
spiciformis
and
Julbernardia
globillora
seedlings
that
were
established
during
the
wet
season
died
during
a
dry
period
(Strang
1966).
Even
in
the
relatively
humid
southeastern
United
States
summer
droughts
kill
many
pine
seedlings.
On
coastal
sands
in
humid
Japan
many
pine
seedlings
perish
during
midsummer
droughts
(Tazaki
et
al.
1980).
If
young
seedlings
do
survive
a
drought
their
normal
development
is
arrested
(Borger
&
Kozlowski
1972).
The
importance
of
water
deficits
in
trees
is
sometimes
underemphasized
when
growth
re-
duction
or
death
of
trees
is
attributed
to
such
factors
as
plant
competition,
disease,
or
insects.
Root
diseases
and
insect
injury
to
roots
reduce
water
absorption,
thereby
inducing
shoot
water
deficits
(Helms
et
al.
1971,
Ayres
1978).
Diversion
of
water
by
competing
herbaceous
weeds
reduced
growth
of
Pinus
resinosa
trees
by
half
(Wilde
et
al.
1968).
Gaultheria
shallon
undergrowth
in
a
Pseudotsuga
menziesii
stand
used
45
to
70%
of
the
available
water
(Tan
et
al.
1978).
Removing
understorey
broadleaved
trees
increased
growth
of
Pinus
echinata
trees
by
40%
(Roger
&
Brink-
man
1965),
further
emphasizing
the
importance
of
water
supply
to
tree
growth.
The
desiccation
of
tree
crowns
following
occlusion
of
vessels
after
infection
by
fungi
causing
vascular
wilt
disease
leads
sequentially
to
growth
reduction
and
death
of
trees
(Kozlowski
et
al.
1962).
Water
deficits
may
also
predispose
trees
to
onslaughts
of
fungus
pathogens
and
insects
(Vita
1961,
West
1979).
II.
Sources
of
Water
Although
trees
obtain
water
primarily
from
the
soil,
they
also
absorb
some
as
dew
or
fog.
Atmospheric
moisture
is
sometimes
ecologically
significant
as
in
parts
of
the
United
States,
Central
America,
Chile,
Mexico
and
Israel.
In
deserts
dew
often
forms
during
the
night
and
reduces
trans-
pirational
water
losses
and
plant
water
deficits
during
the
day
(Waisel
1958,
1960,
Duvdevani
1964,
Gindel
1973).
In
the
northwestern
United
States
accumulated
dew
amounted
to
15
to
20%
of
the
water
evaporated
from
a
large
Pseudotsuga
menziesii
tree
(Fritschen
&
Doraiswamy
1973).
In
the
fog
belt
of
northern
California
more
than
40
cm
of
fog
water
were
collected
under
Pseudo-
tsuga
crowns
(Azevedo
&
Morgan
1974).
Fog
interception
is
important
in
maintaining
dense,
luxuriant
cloud
forests
on
the
upper
windward
slopes
of
the
Sierra
Madre
Oriental
of
eastern
Mexico
(Vogelmann
1973).
In
the
Atacama
Desert
of
northern
Chile
where
precipitation
is
neglig-,
ible
growth
of
Prosopis
tamarugo
is
sustained
by
atmospheric
moisture
(Went
1975).
Water
supply
and
tree
growth.
Part
I
Water
deficits.
59
Water
needs
of
trees
are
at
least
partly
fulfilled
by
water
stored
in
leaves,
stems,
roots,
and
reproductive
structures.
Water
is
withdrawn
along
a
water
potential
(1,D)
gradient
from
tissues
nearest
the
site
of
evaporation.
Water
reserves
in
leaves
are
usually
small
but
may
be
important
in
control
of
stomatal
aperture
(Running
1980b).
In
seedlings
water
reserves
in
roots
are
im-
portant
in
preventing
severe
water
deficits
in
leaves.
In
large
forest
trees,
however,
the
sapwood
of
stems
is
a
more
important
water
reservoir.
The
total
water
storage
capacity
of
a
Pseudotsuga
menziesii
forest
was
estimated
at
270
m
3
/ha
(26.7
mm).
About
two-thirds
of
the
water
available
for
transpiration
was
in
the
stem
sapwood
and
less
than
5%
in
the
phloem,
cambium,
and
foliage
(Waring
et
al.
1979).
Withdrawal
of
water
from
the
sapwood
is
greatest
in
the
spring
and
early
summer.
Depletion
of
sapwood
water
starts
about
as
soon
as
transpiration
begins,
consistent
with
the
hypothesis
that
the
most
readily
available
sources
of
water
are
used
first.
The
availability
of
sapwood
water
reserves
is
shown
by
the
long
time
lag
in
propagation
of
tid
from
transpiring
leaves
to
the
roots.
Stems
of
severely
water-stressed
trees
shrank
less
per
bar
of
daily
IP
depression
of
shoots
than
did
stems
of
trees
with
moderate
water
deficits,
reflecting
differences
in
amounts
of
sapwood
water
(Pereira
&
Kozlowski
1978).
Fruits
often
function
as
water
reservoirs
(Kozlowski
1965)
as
shown
by
their
shrinkage
at
times
of
high
transpiration.
Young
fruits
often
shrink
less
than
old
fruits
either
because
the
former
store
little
water
or
because
of
a
low
transpiring
leaf
area
on
the
plant.
Natural
root
grafting
among
compatible
forest
trees
is
very
common
(Kozlowski
&
Cooley
1961)
and
provides
a
mechanism
for
transport
of
some
water
from
one
tree
to
another
(Bormann
1966).
For
example,
suppressed
Pseudotsuga
menziesii
trees
obtained
water
from
dominant
trees
to
which
they
were
connected
by
functional
root
grafts
(Eis
1972).
III.
Measurement
of
Water
Deficits
The
water
balance
of
trees
has
been
variously
characterized
by
wilting,
tissue
moisture
content,
relative
water
content
(RWC)
(also
called
relative
turgidity),
saturation
deficit
(SD),
and
water
potential
(P).
As
leaves
dehydrate
and
lose
turgor
they
eventually
wilt,
with
old
leaves
usually
wilting
first.
Wilting
varies
as
to
degree
and
may
be
incipient,
temporary,
or
permanent..
Incipient
wilting,
characterized
by
slight
decrease
of
turgor,
usually
does
not
cause
drooping
of
leaves.
Incipient
wilting
grades
into
temporary
wilting,
characterized
by
visible
drooping
of
dehydrated
leaves
during
the
day,
followed
by
rehydration
and
recovery
from
wilting
during
the
night
when
stomata
are
closed
and
evaporative
demand
declines.
During
sustained
periods
of
soil
drying,
temporary
wilting
grades
into
a
state
of
permanent
wilting
in
which
plants
do
not
recover
turgidity
at
night.
According
to
Slatyer
(1967),
permanent
wilting
occurs
when
the
tks
of
the
leaf,
root,
and
soil
around
the
roots
are
equal
and
turgor
pressure
is
zero.
This
is
because
water
movement
into
roots
occurs
only
when
of
roots
is
lower
than
IP
of
soil.
Permanently
wilted
plants
can
recover
turgidity
only
when
water
is
added
to
the
soil.
Prolonged
permanent
wilting
usually
kills
most
plants
(Kramer
1969).
Wilting
is
sometimes
used
to
determine
when
internal
water
deficits
develop
in
plants.
Species
like
Prunus
serotina,
Cornus
spp.,
and
many
delicate-leaved
shrubs
wilt
early
during
a
drought.
Unfortunately
visible
wilting
of
some
plants
reflects
a
very
advanced
and
severe
water
stress
con-
dition.
Leaves
of
some
species
do
not
wilt
even
when
severely
dehydrated
because
their
leaves
contain
large
amounts
of
rigid
sclerenchyma
tissue.
For
example,
leaves
of
Ilex
and
Pinus
are
permeated
with
abundant
lignified
tissue
and
do
not
wilt
even
after
their
parenchyma
cells
have
lost
turgor.
When
Elaeis
guineensis
trees
are
severely
dehydrated
they
do
not
wilt
visibly
because
of
their
fibrous
leaves,
thick
hypodermis,
and
well-developed
cuticle.
An
early
symptom
of
water
deficit
in
excised
Elaeis
leaves
is
rolling
of
leaflets
and
loss
of
sheen.
The
leaves
change
colour
to
brown
as
they
dehydrate,
but
they
do
not
wilt
readily
(Rees
1961).
For
these
reasons
observations
of
wilting
do
not
allow
for
quantitative
studies
of
water
deficits
in
trees.
Tissue
moisture
content
as
%
of
fresh
or
dry
weight
is
often
very
deceptive
and
unsatisfactory
as
an
index
of
water
deficit.
Fresh
weight
of
succulent
tissues
often
varies
widely
over
short
periods
and
large
changes
in
water
content
per
unit
of
tissue
result
in
only
small
changes
in
%
of
fresh
weight.
Water
content
as
%
of
dry
weight
is
also
unsatisfactory
because
it
does
not
necess-.
arily
reflect
changes
in
hydration
of
protoplasm
or
physiological
activity.
Changes
in
moisture
content
on
a
dry
weight
basis
may
reflect
changes
in
amount
of
water,
dry
weight
change,
or
both.
Large
changes
in
dry
weight
of
cert'in
tissues
may
result
from
photosynthesis,
respiration,
or
translocation.
Progressive
cell
wall
thickening
in
growing
leaves
also
accounts
for
dry
weight
increase.
Seasonal
decreases
in
percentage
moisture
of
angiosperm
leaves
were
traceable
largely
to
in-
crease
in
dry
weight
and
not
to
a
decrease
in
actual
water
present.
Early
in
the
season
moisture
content
(as
%
dry
weight)
decreased
rapidly
while
the
actual
amount
of
water
in
the
rapidly
expanding
leaves
decreased.
However,
in
most
trees
the
actual
amount
of
water
in
leaves
changed
little
from
mid-June
to
late
summer
but
moisture
content
(as
%
dry
weight)
decreased
progress-
ively.
In
current-year
needles
of
pines
the
moisture
content
(as%
dry
weight)
decreased
progressively
60
Forestry
Abstracts
1982
Vol.
43
No.
2
whereas
in
one-year-old
needles
it
increased
at
first
and
then
stabilized.
These
differences
were
caused
largely
by
dry
weight
changes
from
carbohydrate
translocation
from
older
needles
into
current-year
needles
(Kozlowski
&
Clausen
1965).
Percent
moisture
of
1-year-old
Picea
engelmannii
needles
was
highest
in
March
and
low
before
buds
opened
in
June.
The
decline
was
largely
the
result of
an
increase
in
dry
matter.
The
dry
weight
of
needles
then
decreased
while
%
moisture
increased
from
June
until
early
August.
Subsequent
fluctuations
in
dry
matter
and
water
com-
ponents
were
small
(Gary
1971).
Early
in
the
season
moisture
content
(as
%
dry
weight)
of
Picea
glauca
and
Larix
laricina
cones
increased
primarily
because
of
increased
water
uptake.
Later
in
the
season
the
moisture
content
of
cones
declined
when
their
dry
weight
increased
rapidly,
but
the
actual
amount
of
water
remained
relatively
stable
(Clausen
&
Kozlowski
1965).
In
fruits
of
Acer
rubrum,
Quercus
rubra,
Prunus
serotina,
and
P.
virginiana,
%
moisture
content
was
variously
influenced
by
water
uptake
or
loss,
change
in
dry
weight,
or
both
(Chaney
&
Kozlowski
1969b).
These
observations
emphasize
that
variations
in
%
moisture
content
must
be
interpreted
cautiously
and
in
relation
to
both
water
weight
and
dry
weight
changes
in
tissues.
Both
SD
and
RWC
compare
the
water
content
of
fresh
tissue
with
the
saturated
moisture
content
of
the
same
tissue,
usually
obtained
by
allowing
the
tissues
to
rehydrate
while
floating
on
water.
The
limitation
of
these
measures
is
that
leaves
of
one
species
may
be
fully
turgid
at
a
moisture
content
found
in
wilted
leaves
of
another
species,
or
in
leaves
of
different
ages
on
the
same
tree.
Most
modern
investigators
quantify
water
deficits
in
plants
by
water
potential
(>P),
the
differ-
ence
between
chemical
potential
of
water
in
a
system
and
that
of
pure
free
water.
Water
potential
quantifies
the
capacity
of
water
to
do
work
in
comparison
to
the
work
an
equal
mass
of
pure
free
water
would
do.
Total
4,
in
plant
tissue
or
soil
consists
of
several
components
including:
(1)
solute
potential
(1P
s
)
arising
from
dissolved
solutes
in
water,
(2)
pressure
potential
(14)
origin-
ating
from
turgor
in
plant
cells
or
tension
in
xylem
elements,
and
(3)
matric
potential
(tp
m
)
arising
from
capillary
or
colloidal
forces
by
soil
colloids,
cell
colloids,
or
cell
walls.
Both
4i
s
and
have
negative
values
but
is
positive
except
rarely
when
wall
pressure
is
negative.
In
xylem
elements
tPp
is
negative,
0,
or,
under
the
influence
of
root
pressure,
positive.
In
plant
cells
iii,
the
sum
of
hi
s
,
O
m
,
and
gyp,
is
negative
except
in
fully
turgid
cells
when
it
is
zero
(Kramer
&
Kozlowski
1979).
Water
potential
is
measured
in
energy
per
unit
volume
and
is
commonly
re-
ported
in
bars
or
megapascals
(MPa).
1
Knowledge
of
4/
in
any
two
plant
parts
or
in
the
soil
and
plant
enables
an
investigator
to
determine
which
way
water
will
move
(i.e.
from
a
region
of
high
to
one
of
low
4,),
or
if
it
will
move
at
all.
Water
potential
can
be
determined
by
vapour
equili-
bration,
thermocouple
psychrometry,
or
the
Scholander
pressure
chamber.
For
good
discussions
of
methods
and
sources
of
error
in
determining
water
deficits
in
plants
the
reader
is
referred
to
Barrs
(1968)
and
Slavik
(1974).
IV.
Development
of
Water
Deficits
Tree
crowns
expose
extensive
leaf
surfaces
which
lose
large
amounts
of
water
by
transpiration.
Salix
and
Populus
trees
along
stream
banks
may
lose
a
precipitation
equivalent
of
up
to
100
cm
of
water
in
a
single
growing
season
(Spurr
&
Barnes
1980).
Individual
trees
in
the
overstorey
of
a
rain
forest
may
transpire
up
to
1200
litres
of
water
each
day
(Jordan
&
Kline
1977).
A
single
Eucalyptus
tereticornis
tree
in
a
botanical
garden
in
Brazil
lost
37
500
litres
of
water
each
year
(Villaca
&
Ferri
1954).
Even
under
much
less
than
ideal
conditions
forests
in
the
southern
United
States
may
lose
more
than
12
000
litres
of
water
per
hectare
per
day
(Zahner 1968).
Whenever
conditions
are
conducive
to
high
transpiration
rates
the
absorption
of
water
cannot
keep
pace
with
transpirational
losses,
thus
creating
shoot
water
deficits,
even
in
trees
growing
in
well-watered
soils.
For
example,
as
stomata
open
in
the
morning
the
rate
of
transpiration
increases
rapidly
but
appreciable
absorption
of
water
does
not
occur
until
the
decreasing
leaf
11/
produces
enough
ten-
sion
in
the
xylem
sap
to
overcome
the
several
resistances
to
water
flow
through
the
xylem
and
from
the
soil
to
the
root
xylem
(Kramer
&
Kozlowski
1979).
Transpiration
rates,
hence
rate
of
development
of
water
deficits
in
trees,
vary
widely
among
and
within
species
(Kozlowski
et
al.
1974,
Pallardy
1981),
depending
largely
on
differences
in
leaf
area,
root-shoot
ratio
(Pereira
&
Kozlowski
1977b),
and
size,
frequency,
and
responsiveness
of
stomata
(Kramer
&
Kozlowski
1979).
The
soil-tree-atmosphere
system
can
be
considered
a
continuum
through
which
water
moves
along
a
path
of
decreasing
water
potential.
There
is
resistance
to
water
movement
from
the
soil
to
the
root
with
much
resistance
at
the
soil-root
interface
(Faiz
&
Weatherley
1977).
Considerable
resistance
to
flow
is
also
present
across
the
living
cells
of
the
roots
(Kramer
1969).
After
entering
stems
of
ring-porous
trees
water
moves
upward
in
a
thin
layer
of
the
outer
sapwood,
primarily
in
the
larger
vessels
of
the
outermost
annual
ring.
In
contrast,
in
diffuse-
porous
trees
many
vessels
of
several
sapwood
rings
conduct
water
(Kozlowski
1961,
Kozlowski
11
bar
=
1.0
x
10
6
dynes/cm
2
=
1.0
x
10
5
newtons/m
2
=
0.987
atm.
=
1017
cm
water
=
75.0
cm
Hg
=
14.50
lb./in.
2
=
10
5
pascals
=
0.1
megapascals
(MPa)
=
100
joules/kg.
Water
supply
and
tree
growth.
Part
I
Water
deficits.
61
&
Winget
1963,
Chaney
&
Kozlowski
1977).
In
gymnosperms
many
tracheids
of
several
xylem
rings
of
the
sapwood
conduct
water
(Kozlowski
et
al.
1966,
1967).
Some
resistance
to
water
movement
resides
in
stems
but
there
is
much
variation
among
species
in
amounts
of
stem
resist-
ance.
For
example,
resistance
to
water
flow
was
several
times
greater
in
Picea
sitchensis
stems
than
in
those
of
Pinus
sylvestris
(Jarvis
1976).
In
a
Picea
sitchensis
stand
11J
values
of
--1.5
MPa
or
lower
frequently
occurred
in
the
upper
parts
of
the
canopy
even
when
the
soil
was
wet.
Gradients
of
tP
of
up
to
0.2
MPa
occurred
in
stems
and
even
larger
gradients
in
branches.
On
overcast
days
canopy
values
were
higher
and
the
gradients
smaller.
Resistance
in
the
stem
appeared
to
be
largely
responsible
for
the
large
drop
in
IP
between
the
roots
and
leaves
(Hellkvist
et
al.
1974).
Resistance
to
water
transport
also
varies
within
an
annual
ring
as
shown
by
lack
of
water
movement
in
small
vessels
of
ring-
porous
trees
(Chaney
&
Kozlowski
1977)
and
latewood
tracheids
of
gymnosperms
(Kozlowski
et
al.
1966).
Both
root
and
stem
resistances
to
water
transport
are
affected
by
infections
and
injuries.
Vascular
wilt
diseases
increase
resistance
by
inducing
formation
of
gums
and
tyloses
that
block
the
water
conducting
vessels
(Talboys
1968).
Resistances
to
water
transport
are
also
in-
creased
by
injury
to
roots
(Havranek
&
Tranquillini
1972)
and
stems
(Rundel
1973).
As
water
leaves
the
stem
and
moves
through
leaves
and
into
the
atmosphere
it
encounters
several
additional
resistances
including
mesophyll-,
intercellular
space-,
stomatal
or
cuticular-,
and
air
boundary
layer
resistances
(Kramer
&
Kozlowski
1979).
The
resistance
of
wet
soil
to
water
movement
is
low
because
only
small
forces
are
necessary
to
move
water
through
water-filled
soil
pores.
As
the
soil
dries
the
resistance
to
water
transport
increases,
sometimes
leading
to
the
conclusion
that
the
soil
resistance
becomes
the
greatest
re-
sistance
in
the
soil-plant
system
(Cowan
1965,
Nnyamah
et
al.
1978).
The
resistance
of
the
plant
has
often
been
assumed
to
remain
constant
while
soil
dries
(Taylor
&
Klepper
1975).
Some
investi-
gators
have
estimated
that
root
resistance
may
account
for
half
to
two-thirds
of
the
total
resistance
to
water
movement
(Roberts
1977,
Running
1980a).
However,
other
studies
with
herbaceous
plants
indicate
that
the
plant
becomes
the
largest
resistance
to
water
flow
as
soil
dries
(Newman
1969a,
1969b,
Reicosky
&
Ritchie
1976).
Unfortunately
several
of
these
conflicting
studies
incorporated
only
approximations
of
one
or
more
resistances
in
the
soil-plant
system.
However,
Blizzard
&
Boyer
(1980)
determined
the
resistance
to
water
movement
in
both
the
soil
and
in
Glycine
max
plants
directly
and
simultaneously
from
measurements
of
the
soil,
root,
and
leaf
and
water
flux
through
the
soil
and
plant
to
evaporation
sites
in
the
leaf.
As
the
soil
dried
water
flow
through
the
soil
and
plant
decreased
greatly
and
both
the
soil
resistance
and
plant
resistance
increased.
Significantly,
the
plant
resistance
was
greater
than
the
soil
resistance
over
a
wide
range
of
soil
moisture
contents.
Some
of
the
increased
plant
resistance
to
water
transport
as
soil
water
is
depleted
results
from
increased
tension
on
water
in
xylem
vessels
and
appears
to
be
associated
with
cavitation
of
water
columns
(Milburn
&
Johnson
1966).
Changes
in
plant
resistance
may
also
occur
in
the
roots
as
the
soil
dries,
probably
from
root
suberization
and
losses
in
viability
(Kramer
1950).
Much
of
the
increased
total
soil
resistance
results
from
an
increase
at
the
soil-
root
interface
(Taylor
&
Klepper
1975)
associated
with
a
decrease
in
soil-root
contact
as
a
result
of
root
shrinkage
(Huck
et
al.
1970,
Weatherly
1976).
The
important
study
of
Blizzard
&
Boyer
(1980)
emphasizes
a
need
for
more
quantitative
research
on
the
resistance
of
water
flow
through
the
soil
and
in
ring-porous,
diffuse-porous,
and
non-porous
trees
as
droughts
intensify.
Seasonal
variations
in
tfi
The
degree
of
water
deficit
that
develops
in
trees
varies
greatly
with
site,
species
and
cultivars,
duration
and
severity
of
periods
of
low
soil
moisture
conditions,
and
evaporative
conditions
in
the
atmosphere.
Very
low
values
of
develop
in
desert
plants
and
in
mesic
plants
on
sites
with
pronounced
dry
periods,
and
near
the
alpine
timberline
(Table
1).
Leaf
%,1)
varied
seasonally
for
Ilex
opaca,
Cornus
florida,
Liriodendron
tulipifera,
and
Acer
rubrum.
Cornus
developed
the
lowest
t/./
and
Ilex
the
highest
during
a
severe
midsummer
drought
in
North
Carolina
(Roberts
et
al.
1979).
In
northern
Wisconsin
leaf
tP
of
Acer
saccharum
and
Betula
papvrifera
varied
seasonally
and
was
lowest
during
midsummer
droughts
(Pereira
&
Koz-
lowski
1978).
Quercus
agrifolia,
Juglans
californica,
and
Heteromeles
arbutifolia
plants
growing
at
the
top
of
a
slope
had
lower
4/
values
than
plants
at
the
bottom.
Water
potentials
declined
and
differences
became
greater
as
the
season
progressed
and
the
soil
dried
(Syvertsen
1974).
In
several
lowland
tropical
forest
trees
in
Panama
(Cordia
alliodora,
Paramea
occidentalis,
Heisteria
concinna,
Macquira
costanticana,
and
Trichilia
cipo)
predawn
and
daytime
values
of
IP
increased
greatly
after
the
beginning
of
the
rainy
season.
Diurnal
patterns
of
stomatal
aperture
were
diverse.
Stomata
of
Cordia
closed
during
the
day
whereas
those
of
Faramea,
Heisteria,
and
Macquira
did
not.
Responses
of
Trichilia
were
more
complex.
Before
the
rainy
season
began
predawn
and
midday
4/
values
were
low
(-2.6
and
—3.9
MPa,
respectively)
and
leaf
resistance
was
high.
After
the
first
rain
predawn
by
increas-
ed
to
—0.4
MPa.
At
that
time
both
tp
and
stomata
responded
to
changes
in
vapour
pressure
deficit
(VPD)
(Fletcher
1979).
In
Citrus
leaf
%t)
decreased
in
a
specific
relation
with
increasing
values
of
VPD/leaf
resistance'
,
provided
that
soil
water
was
adequate
and
soil
temperature
not
too
low,
vPD/leaf
resistance
is
an
estimate
of
transpiration
per
unit
leaf
area.
If
VPD
between
leaf
and
air
is
measured
as
a
water
vapour
concentrat-
io
n
difference
(in
1.1g
water/cm
3
)
and
this
VPD
is
divided
by
the
leaf
resistance
to
water
vapour
diffusion
(in
s/cm),
the
result
is
called
trans-
Pitational
flux
density
(TFD)
and
has
transpirational
units
of
/..tg
H
2
0.
cm-2.s'
Table
1.
Minimum
values
of
IP
reported
for
various
sites.
Species
(
Mpa)
Location
Source
Angiosperms
Acacia
harpophylla
6.0
Australia
van
den
Driessche
et
al.
(1971)
Acer
glabrum
2.0
Idaho,
USA.
Cline
&
Campbell
(1976)
Acer
rubrum
2.0
North
Carolina,
USA.
Roberts
et
al.
(1979)
Acer
saccharum
2.4
Wisconsin,
USA.
Pereira
&
Kozlowski
(1978)
Alnus
sinuata
1.6
Idaho,
USA.
Cline
&
Campbell
(1976)
Betula
papyri/era
1.6
Wisconsin,
USA.
Pereira
&
Kozlowski
(1978)
Cercidium
microphyllum
3.6
Sonora
desert,
USA.
Halvorson
&
Patten
(1974)
Cornus
Florida
2.2
North
Carolina,
USA.
Roberts
et
al.
(1979)
Ilex
opaca
1.5
North
Carolina,
USA.
Roberts
et
al.
(1979)
Larrea
divaricata
8.2
California,
USA.
Scholander
et
al.
(1965)
Malus
domestica
2.6
England
Goode
&
Higgs
(1973)
Physocarpus
malvaceus
3.0
Idaho,
USA.
Cline
&
Campbell
(1976)
Prosopis
glandulosa
6.0
Texas,
USA.
Haas
&
Dodd
(1972)
Prunus
serotina
2.0
West
Virginia,
USA.
Kochenderfer
&
Lee
(1973)
Pyrus
communis
2.1
Australia
Klepper
(1968)
Quercus
prinus
1.8
West
Virginia,
USA.
Kochenderfer
&
Lee
(1973)
Quercus
rubra
1.6
West
Virginia,
USA.
Kochenderfer
&
Lee
(1973)
Simmondsia
chinensis
6.2
Sonora
desert,
USA.
Halvorson
&
Patten
(1974)
Vitis
vinifera
1.9
Australia
Klepper
(1968)
Gymnosperms
Abies
balsamea
1.3
Wisconsin,
USA.
Pereira
&
Kozlowski
(1967a)
Picea
abies
2.2
Austria
Halbwachs
(1970)
Picea
engelmannii
(Krummholz
form)
9.0
Utah,
USA.
Hansen
&
Klikoff
(1972)
Picea
glauca
4.0
Alaska,
USA.
Cowling
&
Kedrowski
(1980)
Picea
sitchensis
1.8
Scotland
Hellkvist
et
al.
(1974)
Pinus
contorta
1.9
Australia
Klepper
(1968)
Pinus
monticola
2.5
Idaho,
USA.
Cline
&
Camp
bell
(1976)
Pinus
resinosa
1.8
Wisconsin,
USA.
Pereira
&
Kozlowski
(1976a)
Pinus
sylvestris
1.7
Sweden
Hellkvist
(1973)
Pseudotsuga
menziesii
2.0
Oregon,
USA
Waring
&
Cleary
(
I
967)
Tax
us
baccata
2.0
Austria
Richter
(1974)
Tsuga
canadensis
1.6
Connecticut,
USA.
Turner
&
de
Roo
(1974)
F
or
e
st
r
y
Ab
st
ra
ct
s
1
982
V
ol
.
4
3
N
o
.
2
Water
supply
and
tree
growth.
Part
I
Water
deficits.
63
regardless
of
season.
When
soil
LP
was
lower
than
—0.03
MPa
or
when
soil
temperature
was
lower
than
15
°
C
leaf
t//
was
lower
than
was
predicted
by
VPD/leaf
resistance.
These
differences
appeared
to
reflect
an
increase
in
flow
resistance
by
low
temperature
and
modification
of
both
resistances
and
LP
at
the
source
in
the
case
of
low
soil
moisture
(Elfving
et
al.
1972).
Near
Palm
Desert,
Cali-
fornia
the
seasonal
LP
pattern
in
Larrea
divaricata
closely
reflected
climatic
changes.
Values of
4/
were
lowest
during
the
hot
dry
season
and
highest
during
the
cool
and
moist
winter
months.
Dawn
varied
from
—5.2
MPa
in
September
to
—2.5
MPa
in
February
(Oechel
et
al.
1972).
Values
of
LD
of
the
krummholz
form
of
Picea
engelmannii
in
the
Wasatch
Mts.
of
Utah
were
—4.0
MPa
in
early
January,
declining
to
less
than
—9.0
MPa
in
March
and
April,
and
rising
to
—4.0
MPa
in
May.
These
low
values
appeared
to
reflect
a
condition
of
unavailable
water
in
frozen
soils
and
windy
desiccating
atmospheric
conditions
(Hansen
&
Klikoff
1972).
Larger
amplitudes
and
earlier
daytime
minima
in
twig
qi
of
4
Pinus
sylvestris
clones
were
found
during
the
summer
than
during
the
winter.
Plots
of
twig
tP
versus
irradiance
or
VPD
during
the
day
showed
hysteresis
because
of
simultaneous
influences
of
several
environmental
factors
on
tP.
The
relation
between
LP
and
potential
evaporation
rates
was
linear.
At
both
high
and
low
evaporation
rates
the
fastest
growing
clone
had
the
lowest
LP
and
the
slowest
growing
clone
had
the
highest
one
(Hellkvist
Parsby
1976).
Seasonal
changes
in
leaf
LP
of
8
Populus
clones
were
influenced
by
solar
radiation,
leaf
resistance,
evaporative
demand,
and
soil
moisture
content.
The
influence
of
soil
moisture
on
LP
was
greatly
modified
by
atmospheric
conditions
and
stomatal
resistance.
Rapid-growing
clones
exhibited
a
larger
initial
rate
of
decline
in
LP
with
TFD,
but
reduced
the
rate
of
decline
more
than
slow-growing
clones
as
TFD
increased
(Pallardy
&
Kozlowski
1981).
Variations
in
leaf
IP
of
Populus
clones
may
also
be
related
to
differences
in
stomatal
occlus-
ion
with
epicuticular
waxes
(Pallardy
&
Kozlowski
1980).
Diurnal
variations
in
IP
Typically,
leaf
LP
decreases
from
a
high
value
in
the
morning
to
a
midday
minimum
followed
by
an
increase
in
the
late
afternoon.
The
daily
maximum
LP,
which
depends
primarily
on
soil
water
availability,
occurs
at
or
near
dawn.
Such
patterns
have
been
shown
for
a
variety
of
angio-
sperms
(Klepper
1968,
Powell
1974,
Morrow
&
Mooney
1974,
Landsberg
et
al.
1975,
Hinckley
&
Bruckerhoff
1975,
Cline
&
Campbell
1976,
Connor
et
al.
1977,
Pereira
&
Kozlowski
1978,
Pallardy
&
Kozlowski
1
981,
Roberts
et
al.
1979,
Jones
&
Higgs
1979),
and
gymnosperms
(Waggoner
&
Turner
1971,
Hellkvist
et
al.
1974,
Cline
&
Campbell
1976,
Rook
et
al.
1976,
Running
1976,
Pereira
&
Kozlowski
1976a,
Tan
et
al.
1978,
Waring
et
al.
1979,
Landsberg
&
Jones
1981).
Over
a
fairly
wide
range
of
soil
moisture
contents
the
diurnal
changes
in
leaf
LP
are
associated
with
environmental
factors
influencing
transpiration
rates.
As
the
soil
dries
or
cools
excessively
the
relationships
weaken
(Hinckley
&
Bruckerhoff
1975).
Diurnal
changes
in
are
smaller
in
amplitude
at
the
base
of
the
stem
and
in
the
roots
and
occur
after
changes
in
leaf
IP
(Hellkvist
et
al.
1974).
Variations
in
LP
with
tree
height
are
discussed
by
Hinckley
et
al.
1978).
Shoot
LP
of
Acer
saccharum
and
Betula
papyrifera
trees
decreased
from
about
—0.3
MPa
in
the
morning
to
as
much
as
—2.0
MPa
for
Acer
and
—1.6
MPa
for
Betula
in
the
early
afternoon,
and
increased
in
the
late
afternoon
and
evening.
Shoot
1p
varied
with
species,
soil
moisture
availability,
VPD,
solar
radiation,
and
TFD
(Pereira
&
Kozlowski
1978).
In
Prosopis
glandulosa
diurnal
changes
in
IP,
which
often
exceeded
2
MPa,
were
associated
with
changes
in
evaporative
potential
of
the
atmosphere
(Haas
&
Dodd
1972).
Early
in
the
growing
season
early
morning
leaf
i
varied
little
between
young
Pinus
resinosa
and
P.
banksiana
plantation
trees
in
northern
Wisconsin.
However,
midday
LP
was
usually
higher
in
P.
banksiana.
As
the
season
progressed,
early
morning
ip
was
lower
and
decreased
more
during
the
day
in
P.
resinosa
than
in
P.
banksiana.
In
addition
stomata
of
P.
banksiana
usually
closed
early
in
the
day
(Pereira
&
Kozlowski
1977a).
Up
to
80%
of
the
variation
in
daily
c
of
Abies
concolor,
and
up
to
76%
of
that
in
Pinus
ponderosa,
was
accounted
for
by
using
irradiance,
VPD,
tree
height,
and
stomatal
aperture
as
independent
variables
in
regression
analysis
(Barker
1973).
After
dew
evaporated
from
Pinus
resinosa
trees
in
the
morning
the
lowest
leaf
LP
was
sometimes
reached
within
2
hours
(Sucoff
1972).
During
most
of
the
day
IP
was
below
—1.0
MPa.
Leaf
%,(i
of
different
whorls
and
needles
of
different
ages
varied
only
slightly.
However,
another
study
showed
that
the
diurnal
decrease
in
di
of
Abies
balsamea
and
Pinus
resinosa
was
gradual
and
the
lowest
values
were
reached
between
1
and
3
p.m.
(Pereira
&
Kozlowski
1976a).
Transplanting
Transplanted
trees
often
develop
very
severe
water
deficits
(Davies
et
al.
1973,
Kozlowski
1975).
Trees
that
are
moved
with
bare
roots
often
undergo
a
physiological
shock
following
desicc-
ation
of
roots
and
injury
to
absorbing
roots.
Roots
of
transplants
often
do
not
grow
fast
enough
to
absorb
sufficient
water
to
meet
transpirational
losses.
Exposure
of
bare-rooted
trees
to
drying
for
even
short
periods
may
rapidly
induce
water
stress,
resulting
in
reduced
growth
and
increased
mortality
(Kozlowski
&
Davies
1975a,
1975b).
Photosynthesis
and
transpiration
of
transplanted
5-year-old
Pinus
sylvestris
trees
were
reduced
by
as
much
as
half,
with
no
recovery
after
4
weeks
64
Forestry
Abstracts
1982
Vol.
43
No.
2
(Hallman
et
al.
1978),
further
emphasizing
transplanting
shock,
Gurth
(1970)
and
Lupke
(1976)
showed
that
a
15%
loss
in
weight
adversely
affected
growth
and
survival
of
transplanted
Picea
abies
trees.
Major
considerations
in
increasing
growth
and
survival
of
transplanted
trees
are
that
trans-
piration
should
be
reduced,
absorption
of
water
should
be
increased,
or
both
should
occur.
These
objectives
can
be
achieved
by
root
pruning,
site
preparation,
transplanting
broadleaved
trees
in
a
leafless
condition,
maintaining
favourable
root-shoot
ratios
by
root
pruning
periodically
in
the
nursery
to
stimulate
root
branching
(Rook
1973),
transplanting
under
favourable
environmental
conditions,
and
proper
handling
of
planting
stock
(Kozlowski
&
Davies
1975a).
Although
water
loss
of
trees
can
be
reduced
with
film-type
and
metabolic
antitranspirants
such
compounds
are
often
harmful
because
they
may
plug
stomatal
pores,
reduce
photosynthesis,
and
induce
injury.
The
effectiveness
of
antitranspirants
depends
on
the
species,
stage
of
development,
and
atmospheric
conditions.
Problems
in
use
of
metabolic
and
film-type
antitranspirants
are
discussed
by
Waisel
et
al.
(1969),
Kozlowski
&
Clausen
(1970),
Poljakoff-Mayber
&
Gale
(1972),
Lee
&
Kozlowski
(1974),
Chaney
&
Kozlowski
(1973,
1974),
Olofinboba
et
al.
(1974),
Davies
et
al.
(1974),
Davies
&
Kozlowski
(1974b,
1975a,
1975b),
Kozlowski
&
Davies
(1975a),
and
Kozlowski
(1976a).
V.
Responses
of
Trees
to
Water
Deficits
Water
deficits
influence
all
phases
of
tree
growth
and
are
probably
responsible
for
more
growth
loss
than
all
other
causes
combined
(Kramer
1980).
Tree
growth
is
reduced
both
directly
through
effects
on
cell
turgor
and
indirectly
through
the
intermediation
of
seed
germination
(Kaufmann
1969),
photosynthesis,
respiration,
mineral
nutrition
(Viets
1972),
enzymatic
activity
(Todd
1972,
Viera
da
Silva
1976),
hormone
relations
(Livne
&
Vaadia
1972,
Itai
&
Benzioni
1976),
and
nitrogen
metabolism
(Naylor
1972,
Sprent
1976,
Brandle
et
al.
1977).
It
is
often
difficult
to
determine
precisely
how
much
water
deficits
alone
influence
growth
because
of
the
confounding
effects
of
water
supply
with
those
of
other
environmental
factors.
Yield
is
a
measure
of
integrated
response
over
time
to
changes
in
rates
and
balance
of
physio-
logical
processes
that
are
regulated
by
several
fluctuating
and
interacting
environmental
factors.
The
effects
of
temperature
and
drought
on
tree
growth,
for
example,
are
closely
related.
High
temperatures
increase
transpiration,
thereby
inducing
shoot
water
deficits,
and
they
also
account
for
depletion
of
reserve
carbohydrates.
Low
temperature
induces
thermal
desiccation
injury
by
preventing
absorption
of
water
and
also
causes
chilling
and
freezing
injury
(Kramer
&
Kozlowski
1979).
This
selection
will
characterize
some
of
the
important
responses
of
trees
to
water deficits.
Seed
germination
Seed
germination
varies
widely
with
characteristics
of
seedbeds
that
influence
availability
of
water
to
seeds
(Winget
&
Kozlowski
1965).
Seeds
must
imbibe
water
amounting
to
2
to
3
times
their
dry
weight
in
order
to
initiate
the
metabolic
processes
leading
to
germination.
Although
most
seeds
with
permeable
seed
coats
can
germinate
in
soil
at
field
moisture
capacity,
germination
capacity
decreases
rapidly
as
soils
dry
(Kaufmann
&
Eckard
1977).
For
example,
for
each
reduction
of
0.1
MPa
in
soil
ti,
seed
germination
of
Chamaecyparis
obtusa
decreased
by
6%,of
Pinus
densi-
flora
by
2.5%,
and
of
Pinus
thunbergii
by
1.5%
(Satoo
1966).
Studies
with
polyethylene
glycol
(PEG)
solutions
showed
a
marked
reduction
in
germination
of
Pinus
ponderosa
seeds
at
a
t1.)
of
—0.4
MPa
and
Pinus
eldarica
at
—0.6
MPa.
Optimum
germination
occurred
at
—0.2
MPa
(Djavanshir
&
Reid
1975).
Reduction
of
of
solutions
to
—0.8
MPa
greatly
reduced
germination
of
Pinus
palustris
and
Pinus
elliotti
seeds.
No
seeds
germinated
at
stresses
of
—1.8
MPa
(Barnett
1969).
Shrinkage
of
plant
tissues
and
organs
Trees
increase
in
size
as
a
result
of
cell
division
and
expansion
and
they
also
shrink
and
swell
periodically
because
of
dehydration
and
rehydration.
The
reversible
changes
in
size
of
various
tissues
and
organs
are
sometimes
small
but
at
other
times
exceed
those
resulting
from
growth.
Both
seasonal
and
diurnal
shrinkage
and
swelling
of
leaves,
stems,
roots,
and
reproductive
structures
have
been
well
documented
(Kozlowski
1965,
1972a,
1972b,
Kramer
&
Kozlowski
1979).
During
progressive
drying
of
soil
the
turgor
of
plant
cells
decreases
daily
(although
there
is
some
rehydration
and
increase
in
turgor
of
plant
cells
during
the
night)
and
plant
tissues
shrink
progressively.
The
daily
rate
of
shrinkage
of
plant
tissues
during
a
drought
is
slowed
by
high
atmospheric
humidity
which
reduces
transpirational
water
loss.
As
a
drought
intensifies
plant
tissues
are
less
likely
to
regain
turgidity
and
to
expand
during
the
night.
Diurnal
shrinking
and
swelling
of
plants
occur
even
in
well
watered
plants
because
absorption
of
water
during
the
day
usually
does
not
keep
up
with
transpirational
losses.
At
night
the
rates
of
absorption
of
water
through
the
roots
and
transpiration
are
low
but
the
rate
of
absorption
is
somewhat
greater.
Hence,
there
is
a
tendency
for
nocturnal
rehydration
and
swelling
of
various
tissues
(Kramer
1969).
Leaves.
Leaf
thickness
of
four
Asian
tree
species
varied
up
to
13%
as
leaf
moisture
content
changed
(Meidner
1952).
In
Musa
acuminate
variations
in
leaf
thickness
were
correlated
with
changes
in
stomatal
opening
and
transpiration
rates
(Brun
1965).
Progressive
shrinkage
of
Prunus
Water
supply
and
tree
growth.
Part
I
Water
deficits.
65
cerasus
and
Citrus
mitis
leaves
occurred
over
a
rainless
period
of
several
days.
Leaves
shrank
appreciably
during
the
daytime
in
early
stages
of
a
drought
and
they
rehydrated
and
swelled
somewhat
during
the
night.
However,
the
rate
of
nightly
expansion
decreased
during
each
success-
ive
day
until
expansion
ceased
in
severely
dehydrated
leaves.
The
overall
trend
was
one
of
de-
creasing
leaf
thickness
as
the
soil
dried.
Irrigation
after
a
soil
drying
cycle
was
followed
by
rapid
resumption
of
leaf
expansion
at
night
and
shrinkage
during
the
day
(Chaney
&
Kozlowski
1969c,
1971).
As
Pinus
strobus
and
P.
nigra
var.
austriaca
leaves
were
dehydrated,
cells
of
the
chloren-
chyma,
endodermis,
and
transfusion
tissue
decreased
in
size
(Parker
1952).
Leaves
of
well-watered
Prunus
cerasus
plants
shrank
considerably
during
the
day
and
expanded
at
night.
Greater
shrinkage
was
associated
with
higher
VPD
of
the
air,
the
relation
holding
much
better
when
the
soil
was
wet
than
when
it
was
dry
(Chaney
&
Kozlowski
1969c).
In
Citrus
mitis
trees
that
were
watered
daily,
leaf
thickness
began
to
decrease
near
sunrise,
when
VPD
began
to
increase,
and
leaves
continued
to
shrink
until
late
afternoon
by
which
time
VPD
was
decreasing
Chaney
&
Kozlowski
1971).
Shrinkage
of
Citrus
leaves
during
the
day
became
noticeable
in
mid-
May,
increased
considerably
during
the
summer,
and
was
irregular
by
December.
Diurnal
shrinkage
of
leaves
was
very
responsive
to
relative
humidity
changes
below
75%
and
was
negligible
at
higher
relative
humidities
(Kadoya
et
al.
1975,
Kadoya
1977).
Diurnal
shrinkage
and
swelling
of
Citrus
leaves
were
also
influenced
by
soil
temperature
through
its
effects
on
absorption
of
water
(Kadoya
1978).
Stems.
Both
seasonal
and
diurnal
stem
shrinkage
occurs
in
forest
trees.
Progressive
stem
shrink-
age
for
weeks
to
months
during
rainless
periods
has
been
demonstrated
for
many
species
of
the
temperate
zone
and
tropics
(Kozlowski
1958,
1967a,
1968a,
1968b,
1968c,
1968d,
1970,
1972a,
1972b).
The
amount
of
stem
shrinkage
of
broadleaved
trees
during
a
severe
drought
in
August
in
New
Jersey
exceeded
total
growth
increment
up
to
that
time
(Buell
et
al.
1961).
Stems
of
mature
Pinus
strobus
trees
contracted
periodically
during
rainless
periods,
after
spring
supplies
of
soil
moisture
were
depleted
(Bormann
&
Kozlowski
1962).
Stem
shrinkage
during
a
soil
drying
cycle
was
much
more
gradual
in
Pinus
resinosa
seedlings
than
in
Fraxinus
americana
seedlings,
reflecting
much
higher
transpirational
water
loss
in
Fraxinus
(Ogigirigi
et
al.
1970).
In
southwestern
Colorado
late
April
and
May
rainfall
induced
swelling
of
Pseudotsuga
menziesii
and
Pinus
edulis
stems.
During
the
subsequent
long
dry
summer
the
stems
shrank
progressively,
except
for
slight
daily
rehydration
and
expansion
during
the
night.
Some
trees
continued
to
shrink
for
more
than
3
months.
After
a
fall
of
rain
in
September
the
trees
rapidly
rehydrated
and
their
stems
swelled
(Fritts
et
al.
1965).
In
the
state
of
Washington
Pseudotsuga
menziesii
stems
shrank
consistently
over
a
6-week
period
from
mid-July
to
the
end
of
August
(Dimock
1964).
In
Uganda
stems
of
the
rain
forest
species
Entandophragma
angolense
and
Lovoa
brownii
continued
to
shrink
for
2
to
3
months
(Dawkins
1956).
In
Costa
Rica
shrinkage
of
tree
stems
during
the
dry
season
varied
from
negligible
in
some
species
to
extensive
in
others.
The
amount
of
stem
shrinkage
during
the
dry
season
exceeded
net
annual
increment
in
Bursera
simaruba,
Calycophyllum
candidissimum,
Chomelia
spinosa,
Sapranthes
palanga,
and
Swietenia
macrophylla.
Complete
defoliation
did
not
prevent
dry-season
shrinkage
of
Bombacopsis
quinata,
Chomelia
spinosa,
Cochlospermum
vitifolium,
Guettarda
macrosperma,
Luehea
candida,
and
Tabebuia
neochrysantha
(Daubenmire
1972).
Swelling
of
tree
stems
following
a
period
of
drought
often
occurs
rapidly.
For
example,
the
circumference
of
Pinus
resinosa
stems
increased
rapidly
when
abundant
rain
fell
in
late
summer
after
a
drought
period.
Within
a
week,
however,
the
tree
stems
contracted
to
the
pre-rain
circum-
ference
(Kozlowski
&
Peterson
1962).
Droughted
Populus
tremuloides
trees
in
northern
Wisconsin
rehydrated
and
swelled
appreciably
within
30
minutes
after
a
cloudburst
(Kozlowski
&
Winget
1964).
When
Pinus
radiata
trees
were
unwatered
for
30
days
their
stems
shrank
very
gradually.
On
rewatering,
however,
stem
diameter
increased
rapidly
(Rook
et
al.
1976).
Tree
stems
usually
shrink
during
the
day
and
expand
at
night
as
a
result
of
changes
in
hydr-
ation.
The
amount
of
daily
stem
shrinkage
varies
with
species,
soil
moisture
supply,
leaf
area,
root-shoot
ratio
and
atmospheric
conditions
that
influence
transpiration
and
leaf
1p
(Kozlowski
&
Winget
1964,
Mitscherlich
1972,
Hinckley
et
al.
1974,
Braekke
&
Kozlowski
1975a,
1975b,
Braekke
et
al.
1978).
Diurnal
changes
in
stem
diameter
often
occur
rapidly
as
conditions
con-
trolling
transpiration
(e.g.
passage
of
dark
clouds,
sudden
rainstorms)
change
(Hinckley
et
al.
1974,
Pereira
&
Kozlowski
1976a).
The
amount
of
diurnal
stem
shrinkage
is
also
influenced
by
pathogenic
microorganisms
which
induce
formation
of
gums
and
tyloses
that
plug
the
water
conducting
vessels.
For
example,
inoculation
of
Quercus
ellipsoidalis
trees
with
a
spore
suspension
of
Ceratocystis
fagacearum,
the
incitant
of
oak
wilt
disease,
was
followed
by
a
reduction
in
the
amount
of
daily
stem
shrinking
and
swelling
(Kozlowski
et
al.
1962).
The
amount
of
stem
shrink-
age
often
varies
on
different
sides
of
the
same
tree
(Young
1952,
Pereira
&
Kozlowski
1976a)
and
is
usually
greater
in
the
upper
stem
than
in
the
lower
stem
(Braekke
&
Kozlowski
1975a,
1975b).
In
Pseudotsuga
menziesii
the
amplitude
of
diurnal
fluctuations
in
stem
diameter
was
greater
near
the
middle
of
the
crown
than
above
or
below
it
(Dobbs
&
Scott
1971).
Diurnal
shrinkage
of
tree
stems
lags
behind
the,
development
of
water
deficits
in
leaves.
For
example,
stem
shrinkage
at
breast
height
lagged
behind
reduction
in
leaf
tif
by
1
to
2
hours
in
66
Forestry
Abstracts
1982
Vol.
43
No.
2
Acer
saccharum
(Pereira
&
Kozlowski
1978),
2
hours
in
Pseudotsuga
menziesii
(Waggoner
&
Turner
1971,
Zaerr
1971),
and
3
to
4
hours
in
Pinus
resinosa
(Pereira
&
Kozlowski
1976a).
In
Wisconsin
the
amount
of
daily
stem
shrinkage
varied
in
the
following
declining
order:
Pinus
resinosa
;
Populus
tremuloides
;
Quercus
ellipsoidalis.
In
all
these
species
there
was
little.
diurnal
stem
shrinkage
early
in
the
growing
season
when
soil
moisture
supplies
were
high
and
the
leaves
were
small.
As
leaves
expanded
and
transpiration
increased,
the
amount
of
diurnal
stem
shrinkage
increased.
Late
in
the
summer
when
the
soil
was
dry
the
tree
stems
became
severely
dehydrated
and
the
daily
amplitude
of
stem
shrinkage
declined
greatly
(Kozlowski
&
Winget
1964).
In
Ohio
the
amount
of
diurnal
stem
shrinkage
of
Fagus
grandifolia
was
low
in
May
and
early
June,
but
increased
greatly
by
July
when
transpiration
rates
were
high
(Fritts
1958).
Pinus
resinosa
stems
shrank
much
more
each
day
than
those
of
Betula
papyrifera.
Daily
stem
shrinkage
and
swelling
continued
in
P.
resinosa
after
seasonal
cambial
growth
stopped
but
this
was
not
the
case
for
B.
papyrifera
(Braekke
&
Kozlowski
1975b).
Diurnal
stem
shrinkage
was
greater
in
Acer
saccharum
than
in
Betula
papyrifera
stems
(Pereira
&
Kozlowski
1978).
Roots.
Tree
roots,
like
stems,
shrink
during
the
day
and
expand
during
the
night.
In
California
daily
shrinkage
of
Pinus
radiata
roots
began
later
than
in
the
stem
during
November
to
March.
Later
in
the
growing
season
there
was
little
difference
in
time
of
shrinkage
of
roots
and
stems
(MacDougal
1936).
Extensive
daily
shrinkage
of
Gossypium
roots
was
shown
by
Huck
et
al.
(1970).
Fruits
and
cones.
Diurnal
shrinkage
and
expansion
have
been
recorded
in
a
variety
of
angio-
sperm
fruits
including
those
of
Acer
rubrum
(Chaney
&
Kozlowski
1969a),
Citrus
spp.
(Bartholo-
mew
1926,
Rokach
1953,
Kaufmann
1972,
Chaney
&
Kozlowski
1971),
Corylus
cornuta
var.
cornuta
(Chaney
&
Kozlowski
1969a),
Juglans
nigra
(MacDougal
1924),
Malus
spp.
(Harley
&
Masure
1938,
Tukey
1959,
1962),
Persea
americana
(Schroeder
&
Wieland
1956,
Klepper
1968),
Prunus
cerasus
(Kozlowski
1968c),
Prunus
pennsylvanica,
P.
serotina
(Chaney
&
Kozlowski
1969a),
Prunus
persica
(Magness
et
al.
1935),
Pyrus
communis
(Ackley
1954,
Klepper
1968),
and
P.
serotina
(Endo
1973a).
During
the
afternoon
of
a
hot
and
dry
day
fruits
of
Citrus
limon
decreased
in
diameter
but
recovered
turgidity
at
night.
The
fruits
lost
35%
more
water
than
detached
fruits,
emphasizing
that
fruits
attached
to
the
tree
were
a
water
reservoir
(Bartholomew
1926).
Maximum
size
of
Persea
americana
fruits
occurred
at
8
or
9
a.m.
As
transpiration
increased
and
water
was
with-
drawn
from
the
fruits
their
diameter
decreased
and
was
minimal
at
approximately
2
or
3
p.m.
(Schroeder
&
Wieland
1956).
Citrus
mitis
fruits
began
to
contract
daily
around
9
a.m.,
at
least
an
hour
after
leaf
shrinkage
started.
Transpiration
apparently
resulted
in
a
1p
gradient
from
the
fruit
to
the
leaves
and
water
was
withdrawn
from
the
fruit
along
a
11,
gradient
(Chaney
&
Kozlow-
ski
1971).
Such
a
free
energy
gradient
was
found
between
fruits
and
leaves
of
Pyrus
communis
(Klepper
1968).
In
Citrus
sinensis
fruits
diurnal
changes
in
exocarp
IP
were
closely
correlated
to
fluctuations
in
leaf
tP
(Kaufmann
1972).
Daily
shrinkage
of
Citrus
mitis
fruits
continued
until
near
sunset,
several
hours
after
leaf
contraction
stopped.
The
time
between
initiation
of
leaf
expansion
and
of
fruit
expansion
was
influenced
by
conditions
affecting
transpiration.
When
VPD
was
high
the
lag
between
initiation
of
fruit
expansion
over
leaf
expansion
was
greater
than
when
VPD
was
low
during
the
day
(Chaney
&
Kozlowski
1971).
Daily
shrinkage
and
swelling
of
Prunus
cerasus
fruits
depended
on
the
degree
of
fruit
water
deficit
and
atmospheric
conditions
con-
trolling
transpiration
(Kozlowski
1968c).
The
amount
of
diurnal
shrinkage
and
swelling
of
Japanese
pear
fruits
varied
with
fruit
devel-
opment
as
follows:
Stage
I
(early
growth)
growth
exceeded
shrinkage;
Stage
II
(retarded
growth),
growth
was
small,
shrinkage
marked;
Stage
III
(maximum
growth),
growth
rapid,
shrinkage
similar
to
that
in
Stage
II;
Stage
IV
(nearly
mature),
growth
continuous
during
the
day
and
night,
essenti-
ally
no
shrinkage
(Endo
1973a).
Diurnal
contraction
of
pear
fruits
did
not
occur
on
rainy
days
or
when
trees
were
sprinkled
with
water
(Endo
&
Ogasawara
1975).
Shading
reduced
diurnal
contraction
of
pears
(Endo
1975).
High
night
temperatures
accelerated
diurnal
shrinkage
and
swelling,
with
the
effects
greater
in
young
than
in
old
fruits
(Endo
1973b).
High
night
temperat-
ures
after
the
end
of
August
not
only
increased
shrinkage
of
pears
but
also
reduced
yield
(Endo
1974).
Young
fruits
often
shrink
less
than
old
fruits.
For
example,
Shamouti
orange
fruits
did
not
shrink
during
the
day
until
early
June,
by
which
time
the
fruits
were
not
as
large
as
ripe
olives.
Thereafter
fruits
contracted
during
the
day
and
expanded
during
the
night
(Rokach
1953).
The
failure
of
young
fruits
to
shrink
is
also
related
to
low
transpiration.
During
early
stages
of
devel-
opment
of
Prunus
cerasus
fruits,
leaves
were
not
fully
expanded,
plant
water
stresses
were
low,
and
fruits
did
not
show
daily
shrinkage.
Such
fruits
could
be
induced
to
shrink
in
the
middle
of
the
day,
however,
by
subjecting
the
plants
to
severe
droughts
(Kozlowski
1965).
Both
seasonal
and
diurnal
shrinkage
and
expansion
occur
at
certain
developmental
stages
of
gymnosperm
cones.
Early
in
their
development
cones
generally
show
progressive
increase
in
diameter
with
little
or
no
super-imposed
shrinkage
in
the
middle
of
the
day.
In
a
mid-stage
of
development
cones
shrink
appreciably
during
the
day
and
expand
at
night.
As
cones
approach
maturity
they
dehydrate
rapidly
and
show
predominantly
continuous
shrinkage
(Kozlowski
Water
supply
and
tree
growth.
Part
I
Water
deficits.
67
1972a).
Such
patterns
have
been
shown
for
cones
of
Pinus
resinosa
(Dickmann
&
Kozlowski
1969a,
1969b),
Picea
glauca,
and
Pinus
banksiana
(Chaney
&
Kozlowski
1969d).
Significance
of
shrinking
and
swelling
of
tissues
Periodic
changes
in
turgor
and
associated
shrinking
and
swelling
of
plant
tissues
have
several
important
biological
implications.
For
example,
loss
of
turgor
of
guard
cells
reduces
stomatal
aperture
which
in
turn
inhibits
photosynthesis
and
prevents
further
desiccation
of
leaves.
Shrink-
age
and
swelling
of
tissues
and
organs
also
affect
movement
of
gases
into
and
out
of
stems
(Hook
et
al.
1972);
measurement
of
growth;
dissemination
of
spores,
pollen,
and
seeds;
flow
of
latex
and
oleoresins;
absorption
of
water
and
ions;
and
extent
of
injury
(Kozlowski
1972a).
Because
shrinkage
and
swelling
of
stems
are
superimposed
on
irreversible
growth
changes,
estimates
of
xylem
increment
obtained
by
measuring
changes
in
stem
diameter
are
sometimes
erroneous.
Both
seasonal
and
diurnal
changes
in
stem
diameter
due
to
hydration
may
exceed
the
increment
resulting
from
cambial
growth.
For
example,
stems
of
Acer
negundo
seedlings
shrank
so
much
during
a
drought
that
stem
diameter
changes
due
to
cambial
activity
were
com-
pletely
masked
(Kozlowski
1967a,
1967b).
A
severe
drought
in
August
induced
so
much
stem
shrinkage
in
several
species
of
forest
trees
that
their
diameters
were
lower
than
they
were
before
the
growing
season
started
(Buell
et
al.
1961).
Although
this
is
an
extreme
example
of
seasonal
stem
shrinkage
it
emphasizes
the
need
for
caution
in
quantifying
cambial
growth
by
the
use
of
dendrometers.
It
is
particulary
difficult
to
determine
accurately
with
dendrometers
and
dendro-
graphs
when
seasonal
cambial
activity
begins
and
ends.
Because
stems
of
Pinus
radiata
trees
were
constantly
shrinking
and
swelling
Fielding
&
Millet
(1941)
concluded
that
they
could
not
deter-
mine
by
the
use
of
dendrometers
when
cambial
activity
began
or
ended.
Braekke
&
Kozlowski
(1975a)
showed
that
Pinus
resinosa
stems
continued
to
shrink
progressively
for
several
weeks
after
cambial
activity
ceased.
Dendrographs
(Fritts
&
Fritts
1955)
are
often
more
useful
than
dial
gauge
dendrometers
for
estimating
cambial
growth
over
short
periods
of
time.
This
is
because
dial
gauge
dendrometers
do
not
provide
a
continuous
record
of
shrinking
and
swelling.
In
contrast,
the
trend
of
cambial
growth
increment
can
be
estimated
over
several
days
by
connecting
daily
peaks
or
valleys
of
continuous
dendrograph
traces
of
stem
diameter
changes,
provided
the
amplitude
of
shrinkage
and
swelling
does
not
change
much
from
day
to
day
(Kozlowski
1972a).
Often
the
amount
of
daily
stem
shrinkage
greatly
exceeds
the
increment
traceable
to
cambial
activity.
For
example,
the
amount
of
shrinkage
of
Pinus
canariensis
stems
in
one
day
approximately
equalled
the
amount
of
cambial
growth
increment
in
5
days
(Holmes
&
Shim
1968).
Daily
shrinkage
of
Picea
spp.
and
Pinus
resinosa
stems
was
sometimes
equal
to
a
week's
cambial
growth
increment
(Kern
1961,
Braekke
&
Kozlowski
1975a,
1975b).
Various
imbibition,
cohesion,
and
turgor
mechanisms
associated
with
shrinkage
and
swelling
that
account
for
seed
dispersal
were
reviewed
by
Kozlowski
(1972a).
Progressive
dehydration
of
gymnosperm
cones
and
associated
cone
scale
movements
lead
to
seed
dispersal.
When
seeds
are
mature
the
cones
dehydrate
rapidly,
as
in
Picea
mariana,
P.
glauca,
Larix
laricina
(Clausen
&
Kozlowski
1965a),
Pinus
banksiana
(Beaufait
1960),
Pseudotsuga
menziesii
(Ching
&
Ching
1962),
Picea
glauca
(Cram
&
Worden
1957),
and
Abies
grandis
(Pfister
1957).
The
opening
of
cones
on
drying
is
related
to
the
greater
shrinkage
of
tissues
on
the
upper
cone
scale
surface
compared
with
tissues
on
the
lower
cone
scale
surface.
The
scales
open
when
stresses
in
the
scales
become
greater
than
the
cohesive
forces
between
them
(Harlow
et
al.
1964,
Allen
&
Wardrop
1964).
After
cones
have
opened
once
they
often
subsequently
close
and
reopen
with
changes
in
relative
humidity
(Fielding
1947).
Flow
of
latex
and
oleoresins.
Flow
of
latex
from
rubber
trees
is
regulated
by
internal
water
balance
and
therefore
varies
with
season,
site
conditions,
and
time
of
day
(Buttery
&
Boatman
1976).
Latex
flow
is
reduced
by
dry
weather.
In
the
Ivory
Coast
seasonal
latex
yield
reached
a
maximum
just
after
the
major
rainy
season
in
May
and
June.
Yield
was
also
higher
in
November
to
early
December
after
a
minor
rainy
season
(Ribaillier
1971).
Latex
flow
is
usually
greater
in
the
morning
than
in
the
afternoon.
Diurnal
changes
in
flow
of
latex
of
Hevea
brasiliensis
are
highly
correlated
with
changes
in
transpiration
rates.
When
Hevea
trees
were
tapped
at
intervals
throughout
a
24-hour
period,
latex
yields
were
similar
when
tapping
was
performed
between
2000
and
0700
hours.
For
trees
tapped
during
the
day
yield
declined
slowly
and
at
1300
hours
it
was
70%
of
the
night-tapping
yield.
Tapping
after
1300
hours
in-
creased
yield
again
(Pardekooper
&
Sookmark
1969).
Diurnal
changes
in
latex
yield
often
follow
changes
in
VPD.
Smaller
differences
between
night
yield
and
day
yield
occur
on
days
with
low
VPD
than
on
days
with
high
VPD.
The
low
turgor
in
the
middle
of
the
day
is
the
result
of
water
movement
from
the
phloem
tissue
to
the
xylem
when
transpiration
is
high.
This
requires
that
the
IP
in
the
xylem
decreases
during
transpiration,
water
can
move
freely
between
the
xylem
and
phloem,
and
the
latex
vessel
system
operates
as
an
osmometer
(Buttery
&
Boatman
1976).
Flow
of
oleoresins
from
tapped
pine
trees
depends
on
turgor
of
epithelial
cells
which
line
the
resin
ducts.
When
trees
dehydrate
as
a
result
of
high
transpiration,
turgor
of
epithelial
cells
de-
creases
and
the
rate
of
exudation
of
oleoresins
declines
(Kozlowski
1972a).
Oleoresin
exudation
pressure
(OEP)
varied
with
factors
controlling
transpiration,
particularly
temperature,
light
68
Forestry
Abstracts
1982
Vol.
43
No.
2
intensity,
and
humidity.
In
Pinus
ponderosa
OEP
was
highest
at
dawn
and
subsequently
decreased,
only
to
increase
in
the
late
afternoon
and
during
the
night.
During
a
progressive
drought
OEP
declined
as
soil
moisture
was
depleted
(Vite
1961).
Absorption
of
water
and
ions.
Models
of
transport
of
ions
and
water
through
the
soil-plant-air
continuum
usually
assume
that
roots
have
constant
diameters
and
that
contact
between
roots
and
soil
does
not
change
with
time
(Nye
1966,
Greenwood
1969,
Klute
&
Peters
1969).
However,
as
roots
shrink
the
contact
between
the
root
and
soil
decreases
so
that
transport
of
ions,
which
is
restricted
to
areas
of
contact,
is
also
reduced.
During
the
middle
of
the
day
water
movement
to
roots
decreases
as
more
of
it
has
to
take
place
across
vapour
gaps
(Bernstein
et
al.
1959).
Shrinkage
of
roots
also
influences
the
cross-sectional
area
available
for
upward
movement
(Huck
et
al.
1970).
Stomatal
aperture
Stomata
begin
to
close
when
the
turgor
of
guard
cells
decreases.
Stomata
usually
close
during
relatively
early
stages
of
leaf
water
deficit,
often
long
before
leaves
wilt
(Kozlowski
1976b).
Turgor
changes
that
control
stomatal
aperture
appear
to
be
caused
by
gain
or
loss
of
ions,
prim-
arily
potassium.
During
stomatal
opening
a
net
influx
of
potassium
ions
from
an
external
solution
of
adjacent
cells
has
been
shown
for
many
species
(Allaway
&
Milthorpe
1976).
Stomata
of
epider-
mal
strips
of
Vicia
faba
opened
only
if
they
were
floating
on
solutions
containing
potassium
(Fischer
&
Hsiao
1968).
The
stomata
of
Acer
saccharinum
seedlings
supplied
with
a
nutrient
solution
containing
potassium
were
wider
open
than
those
supplied
with
the
same
nutrient
but
without
potassium
(Noland
&
Kozlowski
1979).
Both
histochemical
and
microprobe
studies
show
a
linear
dependency
of
stomatal
opening
on
potassium
content
of
guard
cells
(Humble
&
Raschke
1971,
Raschke
1975).
Unfortunately,
the
mechanism
controlling
movement
of
potassium
ions
into
and
out
of
guard
cells
is
at
best
imperfectly
understood.
Current
evidence
indicates
that
potassium
uptake
by
guard
cells
may
involve
a
passive
or
active
mechanism,
or
both.
Production
of
organic
acids
in
cells
and
export
of
hydrogen
ions
could
increase
electrochemical
potential
in
cells
to
cause
passive
uptake
of
potassium
ions.
Alternatively
potassium
uptake
against
an
electro-
chemical
potential
gradient
might
be
mediated
by
an
active
potassium
pump
(Allaway
&
Mil-
thorpe
1976).
When
water
stressed
plants
are
irrigated
the
closed
stomata
sometimes
open
slowly
or
not
at
all
even
after
leaf
turgor
is
restored.
For
example,
when
stomata
of
Citrus
jambhiri
were
closed
by
drought,
several
days
elapsed
after
rewatering
before
stomatal
opening
occurred
(Kaufmann
&
Levy
1976).
The
rate
of
photosynthesis
of
Pseudotsuga
menziesii
trees
that
had
been
subject-
ed
to
drought
remained
low
for
several
days
after
the
trees
were
rewatered
(Zavitkovski
&
Ferrell
1970).
Following
irrigation
the
stomata
of
previously
stressed
Fraxinus
americana,
Acer
rubrum
and
Ulmus
americana
seedlings
opened
to
pre-stress
levels
but
stomata
of
Acer
saccharum
and
Cornus
amomum
did
not
(Davies
&
Kozlowski
1977).
Midday
closure
of
stomata.
Three
daily
patterns
of
stomatal
aperture
have
been
reported:
(1)
stomata
open
in
the
morning,
close
for
a
period
during
the
middle
of
the
day,
reopen
in
the
afternoon,
and
finally
close
late
in
the
day
as
the
light
intensity
decreases,
(2)
stomata
open
early
in
the
morning,
close
rather
early
in
the
day
and
remain
closed
until
the
following
morning,
or
(3)
on
days
of
low
VPD,
abundant
soil
moisture,
and
light
overcast
conditions
stomata
remain
open
until
evening
as
photosynthetically
active
radiation
declines.
Stomatal
closure
during
the
middle
of
the
day
has
been
reported
for
many
species
of
forest
and
orchard
trees
(Kramer
&
Kozlowski
1979).
Although
midday
stomatal
closure
has
been
attributed
to
several
causes
an
important
factor
is
the
lag
of
water
absorption
behind
transpir-
ation.
This
induces
leaf
dehydration
and
reduction
of
leaf
IP
to
a
critical
level
associated
with
stomatal
closure.
When
leaf
water
deficits
are
not
severe
midday
stomatal
closure
may
he
in-
duced
by
high
temperatures
causing
increase
of
CO
2
concentration
in
intercellular
spaces
of
the
leaves
(Kozlowski
1976b).
There
are
many
examples
of
midday
stomatal
closure
and
only
a
few
will
be
given.
In
Pinus
resinosa
daily
patterns
of
stomatal
opening
and
closing
varied
with
environmental
conditions.
In
the
absence
of
internal
water
deficits
stomatal
aperture
varied
only
slightly
during
the
day;
when
VPD
was
high
stomata
tended
to
close
in
the
early
afternoon
(Pereira
&
Kozlowski
1976a).
On
days
when
water
deficits
in
shoots
of
mature
Acer
saccharum
and
Betula
papyrifera
trees
were
not
severe
stomatal
aperture
was
rather
stable
throughout
the
day.
During
droughts,
however,
stomata
closed
early
in
the
afternoon
and
reopened
in
the
late
afternoon
(Pereira
&
Kozlowski
1978).
Midday
stomatal
closure
on
both
leaf
surfaces
occurred
in
Populus
clones
when
VPD
was
high
(Pallardy
&
Kozlowski
1981).
When
the
soil
was
charged
with
water,
and
air
humidity
was
high,
stomata
of
Malus
trees
remained
open
during
the
day.
When
soil
moisture
content
was
high,
and
humidity
was
low,
stomata
closed
before
noon.
With
a
reduced
soil
moisture
supply
the
daily
duration
of
stomatal
opening
was
reduced.
When
the
soil
was
at
or
near
wilting
percentage
the
stomata
usually
did
not
open
at
all
(Magness
et
al.
1935).
Water
supply
and
tree
growth.
Part
I
Water
deficits.
69
In
midsummer
Cornus
Florida
and
Liriodendron
tulipifera
developed
patterns
of
pronounced
midday
stomatal
closure
as
a
drought
period
progressed,
whereas
only
slight
midday
closure
occurred
in
Acer
rubrum
and
in
young
and
old
leaves
of
Ilex
opaca
(Roberts
et
al.
1979).
In
Coffea
arabica
the
stomata
were
partly
open
early
in
the
morning
and
closed
by
midday
(Bier-
huizen
et
al.
1969).
Midday
stomatal
closure
in
Elaeis
guineensis
trees
in
Nigeria
occurred
during
the
latter
half
of
the
dry
season
(from
November
to
April).
With
onset
of
the
rainy
season
diurnal
closure
of
stomata
disappeared.
During
the
dry
season
there
was
a
rapid
early-morning
opening
of
stomata
until
about
9
a.m.
Stomata
then
began
to
close
gradually
until
about
2
p.m.,
after
which
time
slight
opening
occurred
until
4
p.m.
Then
late-afternoon
stomatal
closing
began
and
was
com-
pleted
very
soon
after
sunset.
After
any
significant
rain,
daily
stomatal
closure
became
less
pro-
nounced
and
remained
slight
or
disappeared
until
the
effect
of
the
rain
on
leaf
turgor
disappeared
and
afternoon
closure
was
again
evident
(Rees
1961).
Additional
examples
of
midday
stomatal
closure
are
given
by
Hinckley
et
al.
(1978)
and
Kramer
&
Kozlowski
(1979).
Location
and
age
of
leaves.
The
time
of
stomatal
closure
during
increasing
leaf
water
deficits
may
vary
with
leaf
age,
as
well
as
stomatal
size
and
locatign.
Stomata
of
shade
leaves
are
more
sensitive
than
those
of
sun
leaves,
and
stomata
of
young
leaves
often
close
faster
than
those
of
old
leaves
in
response
to
water
stress.
For
example,
young
leaves
of
Eucalyptus
marginata
trans-
pired
about
20%
less
(per
unit
of
leaf
area)
than
mature
leaves
on
sunny
days
because
of
more
effective
stomata!
closure
in
the
young
leaves
(Doley
1967).
Similar
responses
were
found
in
young
and
old
leaves
of
Eucalyptus
stuartiana
(Henrici
1946).
In
Banksiana
menziesii
and
Stir-
lingia
latifolia
mature
leaves
transpired
more
per
unit
of
fresh
weight
than
young
leaves,
but
the
variations
decreased
as
the
leaves
developed.
The
differences
were
related
to
less
effective
stomatal
responses
of
old
leaves
subsequent
to
lignification
and
cutinization
(Grieve
1956).
Large
and
small
stomata
on
the
same
leaf
may
also
react
somewhat
differently
to
water
stress.
For
example,
responses
of
different-sized
stomata
of
Betula
papyrifera
varied
with
environmental
changes.
When
light
intensity
or
water
status
was
altered,
large
stomata
tended
to
open
first
and
close
last
(Waisel
et
al.
1969).
Burrows
&
Milthorpe
(1976)
discussed
this
subject
in
more
detail.
Sometimes
the
differences
in
sensitivity
of
young
and
old
stomata
to
stress
are
more
a
function
of
environmental
preconditioning
than
of
leaf
age.
For
example,
premature
leaf
senescence
and
decreased
sensitivity
of
stomata
of
Robinia
pseudoacacia
and
Quercus
rubra
to
environmental
changes
were
induced
by
drought
(Hinckley
1973,
Lassoie
&
Chambers
1976).
Stomatal
closure
and
leaf
41.
The
ti
of
leaves
can
vary
over
a
considerable
range
(above
a
critical
value)
without
a
marked
effect
on
stomatal
aperture
(Jarvis
1980).
For
example,
stomatal
aperture
of
Pinus
resinosa
remained
stable
during
the
day
when
leaf
varied
from
-0.5
to
-1.5
MPa
when
radiation
was
high
(Turner
&
Waggoner
1968).
When
a
critical
leaf
cli
is
reached
stomata
begin
to
close.
Differences
among
species
in
sensitivity
of
stomata
to
leaf
water
deficits
have
been
reported
for
both
angiosperms
and
gymnosperms.
In
angiosperms
rapid
stomatal
closure
occurred
at
the
following
values
of
leaf
IP:
Betula
papyrifera,
-1.5
MPa;
Populus
grandidentata,
-1.7
MPa;
Quercus
coccinea,
MPa;
Prunus
serotina,
-2.3
MPa
(Federer
1977);
Acer
circinatum,
-1.5
MPa;
Quercus
alba,
-2.5
MPa
(Lassoie
&
Scott
1977,
Hinckley
&
Bruckerhoff
1975);
Acer
saccharum,
-1.7
MPa;
Quercus
rubra,
-1.85
MPa;
Quercus
alba,
-2.3
MPa;
Quercus
velutina,
-2.4
MPa
(Phelps
et
al.
1976),
Malus
sylvestris,
-1.9
MPa
(West
&
Gaff
1976).
Stomata
of
mesic
species
generally
close
at
higher
values
of
leaf
tp
than
do
those
of
desert
species
(Kramer
&
Koz-
lowski
1979).
In
gymnosperms
stomata
closed
at
the
following
values
of
leaf
IP:
Pinus
contorta,
-1.46
MPa;
Pinus
ponderosa,
-1.65
MPa;
Picea
engelmannii,
-1.6
MPa;
Pseudotsuga
menziesii,
-1.9
MPa;
Abies
grandis,
-2.51
MPa
(Lopushinsky
1969);
Pinus
resinosa,
-1.8
MPa
(Pereira
&
Kozlowski
1976a);
Pseudotsuga
menziesii,
-2.0
MPa;
Pinus
ponderosa,
-1.8
MPa
(Running
1976);
Pinus
radiata,
-1.1
MPa
(Rook
et
al.
1977),
Picea
sitchensis
-2.0
MPa
(Beadle
et
al.
1978),
Pinus
sylvestris,
-0.8
MPa
(Jarvis
1980).
The
critical
leaf
li
for
stomatal
closure
reported
for
different
species
should
not
be
taken
too
seriously
because
the
value
varies
for
different
clones
and
cultivars
(Pallardy
&
Kozlowski
1979b)
and
because
the
response
of
stomata
to
leaf
water
deficit
is
modified
significantly
by
such
factors
as
light
intensity,
CO
2
content
of
intercellular
spaces,
air
humidity,
wind,
age
of
leaf,
osmotic
adjustment,
etc.
(Davies
et
al.
1974,
Davies
&
Kozlowski
1975c,
Kozlowski
&
Pallardy
1979).
Beadle
et
al.
(1978)
reported
that
stomata
of
Picea
sitchensis
closed
at
leaf
1p
values
of
-1.6
to
-2.7
MPa
at
different
times
of
the
year.
Nevertheless,
in
the
same
tree
there
is
a
general
relation
between
leaf
and
stomatal
aperture,
after
the
stomata
begin
to
close
(Hinckley
et
al.
1978).
In
both
herbaceous
and
woody
amphistomatous
plants
stomata
of
the
upper
(adaxial)
and
lower
(abaxial)
leaf
surfaces
often
close
at
different
critical
values
of
leaf
0.
For
example,
in
Gossypium
plants
undergoing
drought,
the
adaxial
stomata
closed
earlier
than
abaxial
stomata
(Sharpe
1973).
Abaxial
stomata
of
Eucalyptus
camaldulensis
seedlings
closed
gradually
at
tif
values
between
-0.8
and
-1.2
MPa.
Those
of
the
adaxial
surface
closed
rapidly
at
IP
values
near
-0.9
MPa
(Pereira
&
Kozlowski
1976b).
Adaxial
stomata
of
2
Populus
clones
were
more
sensitive
than
abaxial
stomata
to
changing
vapour
pressure
deficit
and
light
intensity
(Pallardy
&
Kozlow-
ski
1979a).
For
a
general
discussion
of
environmental
and
biological
control
of
stomatal
aperture
see
Pospisilova
&
Solarova
(1980).
Water
supply
and
tree
growth.
Part
I
Water
deficits.
7
1
relative
humidity,
photosynthesis
was
lower
than
when
soil-moisture
contents
were
high
(Tran-
quillini
1963).
Similarly,
low
soil
moisture
supplies
depressed
photosynthesis
of
Pinus
densiflora,
but
the
degree
of
inhibition
varied
with
air
humidity
(Negisi
&
Satoo
1954).
Relative
humidity
of
the
air
also
controlled
the
rate
of
photosynthesis
of
citrus
(Kriedemann
1968).
Increasing
the
relative
humidity
from
28
to
86%
almost
doubled
the
rate
of
photosynthesis
(Ono
et
al.
1978).
These
studies
emphasize
the
importance
of
relating
photosynthesis
to
plant
water
deficits
rather
than
to
soil
water
deficits.
Plant
and
photosynthesis.
In
many
mesic
species
the
rate
of
photosynthesis
begins
to
decline
when
leaf
ti
becomes
only
slightly
negative;
plants
adapted
to
grow
in
arid
regions
generally
maintain
photosynthesis
at
much
lower
ii
values
than
do
mesic
plants.
Photosynthesis
of
Alnus
oblongifolia
and
Fraxinus
pennsylvanica
decreased
when
leaf
i/i
dropped
to
—1.0
MPa;
of
Simm-
ondsia
chinensis,
Vaquelinia
californica,
and
Juniperus
deppeana
to
between
—1.0
and
—2.0
MPa;
of
Larrea
divaricata
to
—2.0
MPa;
and
Acacia
greggii
to
—2.3
MPa
(Chabot
&
Bunce
1979).
Photo-
synthesis
of
Sassafras
albidum
began
declining
when
leaf
>Ji
reached
—1.2
to
—1.4
MPa
and
was
negligible
at
—2.1
MPa
(Bazzaz
et
al.
1972).
Decrease
in
photosynthesis
of
Pseudotsuga
menziesii
began
when
shoot
IP
dropped
to
near
—1.0
MPa
and
at
—3.5
MPa
the
rate
was
negligible
(Brix
1972).
Simmondsia
chinensis
plants
in
the
desert
maintained
measurable
photosynthesis
at
—7.0
MPa
(Al-Ani
et
al.
1972).
In
Acer
negundo
from
a
streamside
habitat
photosynthesis
ceased
at
shoot
tp
values
1.0
to
1.5
MPa
higher
than
species
of
a
xeric
community
(Quercus
gambelli,
Art-
emisia
tridentata,
and
Purshia
tridentata)
(Dina
&
Klikoff
1973).
There
is
considerable
ecotypic
and
clonal
variation
in
photosynthetic
response
to
leaf
water
deficits.
For
example,
photosynthesis
of
Simmondsia
chinensis
plants
from
a
coastal
course
was
reduced
by
half
when
leaf
V/
decreased
to
—2.0
MPa.
In
plants'
from
desert
populations
a
com-
parable
reduction
in
photosynthesis
of
Simmondsia
did
not
occur until
decreased
to
—3.6
MPa
(Al-Ani
et
al.
1972).
Photosynthesis
of
plants
grown
in
moist
soil
and
humid
air
was
greatly
reduced
when
plant
IP
dropped
to
—2.0
MPa.
However,
photosynthesis
of
plants
growing
under
severe
drought
in
Death
Valley,
California
declined
very
little
at
a
11)
of
—2,9
MPa.
Photosynthetic
inhibition
of
the
well-watered
plants
was
caused
largely
by
non-stomatal
factors
but
this
was
not
the
case
for
plants
growing
under
severe
stress
(Mooney
et
al.
1977).
Because
stomata
of
various
Populus
clones
respond
differently
to
water
stress
(Ceulemans
et
al.
1978,
Pallardy
&
Kozlowski
1979a,
1979b,
Pallardy
1981)
photosynthesis
may
also
be
expected
to
vary.
Causes
of
reduction
in
photosynthesis
Some
of
the
reduction
in
photosynthesis
during
drought
results
from
increased
resistance
to
diffusion
of
CO
2
to
the
chloroplasts
and
some
is
caused
by
a
reduction
in
photosynthetic
cap-
acity.
As
a
drought
develops
photosynthesis
is
reduced
early
by
stomata!
closure
and,
as
leaves
become
more
severely
dehydrated,
the
photosynthetic
process
is
eventually
inhibited
(Luukkanen
1978).
In
the
longer
term
water
stress
also
reduces
total
photosynthesis
by
restricting
development
of
new
leaf
area
(Kramer
&
Kozlowski
1979).
Resistance
to
CO
2
diffusion.
The
availability
of
CO
2
to
chloroplasts
is
limited
by
various
resistances
in
its
inward
diffusion
path,
including
those
associated
with
the
boundary
layer,
cuticle,
stomata,
and
mesophyll.
Mesophyll
resistance
is
an
aggregation
of
several
different
resistance
components
including:
(1)
gaseous
diffusion
resistance
of
the
intercellular
spaces,
(2)
resistance
to
solution
of
CO
2
at
air-water
interfaces,
(3)
resistance
to
movement
of
CO
2
in
the
liquid
phase,
and
(4)
"carboxylation
resistance"
attributable
to
enzyme
activity
and
light
reactions.
Because
CO
2
enters
leaves
largely
through
stomatal
pores,
stomatal
aperture
may
be
expected
largely
to
regulate
photosynthesis.
In
addition
increased
mesophyll
resistance
has
been
proposed
as
a
possible
cause
of
reduced
photosynthesis
in
water-stressed
plants.
Stomatal
aperture
and
photosynthesis.
The
importance
of
stomatal
closure
in
regulating
photo-
synthesis
has
often
been
shown
by
parallel
reduction
of
photosynthesis
and
transpiration
as
a
drought
developed.
Temporary
midday
reductions
in
photosynthesis
commonly
occur
(Kozlowski
1957,
1962,
Mooney
et
al.
1966,
Kozlowski
&
Keller
1966,
Hodges
1967,
Hanson
&
Dye
1980,
Tenhunen
et
al.
1980)
and
are
often
correlated
with
stomatal
closure.
Stomata
usually
close
earlier
in
the
day
as
available
soil
moisture
decreases
and
this
reduces
photosynthesis
in
many
species.
Troughton
&
Slatyer
(1969)
showed
that
photosynthesis
of
Gossypium
was
inhibited
by
mild
water
stress
but
the
photosynthetic
mechanism
was
not
adver-
sely
affected
until
much
later
when
the
leaves
were
severely
desiccated.
Kriedemann
(1971)
recorded
close
correlation
between
stomatal
closure
and
decrease
in
photosynthesis
of
citrus
leaves
during
a
period
of
soil
drying.
As
stomata
close
in
the
middle
of
the
day
the
rate
of
photosynthesis
also
decreases,
often
by
a
corresponding
amount.
Rates
of
photosynthesis
and
transpiration
of
Pinus
taeda
were
highly
correlated
during
the
day,
indicating
that
stomatal
closure
at
low
leaf
111
reduced
photosynthesis.
Regehr
et
al.
(1975)
found
very
close
correlation
between
stomatal
aperture
and
photosynthesis
in
Populus
deltoides
and
obtained
good
estimates
of
photosynthesis
by
measuring
stomatal
aper-
ture
with
a
porometer.
Lakso
(1979)
found
a
linear
relationship
between
stomatal
resistance
and
photosynthesis
of
Malus
trees.
Forestry
Abstracts
1982
Vol.
43
No.
Photosynthesis
of
the
same
species
may
be
affected
differently
by
stomatal
closure
because
of
variations
in
responses
of
plants
grown
in
different
environments
(Van
Volkenburgh
&
Davies
1977).
Variations
in
stomatal
sensitivity
of
Gossypium,
and
probably
in
photosynthesis,
to
de-
creasing
leaf
Ii
declined
as
follows:
Growth
chamber
plants
;
greenhouse
grown
plants
;
field
grown
plants
(Davies
1977).
Mesophyll
resistance
and
photosynthesis.
During
a
developing
drought
photosynthesis
is
some-
times
reduced
more
than
can
be
attributed
to
observed
changes
in
stomatal
aperture
(Hari
et
al.
1975,
Beadle
&
Jarvis
1977,
Luukkanen
1978).
This
has
focused
attention
on
the
possible
import-
ance
of
effects
of
environmental
stresses
on
mesophyll
resistance
to
CO
2
diffusion.
For
example,
Wuenscher
&
Kozlowski
(1970)
showed
that
both
stomatal
and
mesophyll
resistances
impeded
CO
2
transfer,
and
variation
in
either
resistance
affected
net
photosynthesis.
At
high
light
intensity
variations
in
mesophyll
resistance
were
more
important
than
those
in
stomatal
resistance.
At
low
intensity
stomatal
resistance
limited
net
photosynthesis
of
the
shade
intolerant
species
Quercus
velutina.
Photosynthesis
of
desiccated
plants
may
be
reduced
by
a
decrease
in
the
carboxylation
process,
including
changes
in
both
light
and
dark
reactions
of
photosynthesis
or
by
an
increase
in
respiration
(Boyer
1976a,
1976b).
Water
stress
caused
an
increase
in
dark
respiration
of
Pinus
taeda
leaves
(Brix
1962)
and
Picea
abies
needles
(Luukkanen
1978).
Other
investigators
(Puritch
1973,
Boyer
1976a,
Kramer
&
Kozlowski
1979)
reported
two
other
patterns
of
respiratory
response
to
water
deficits:
(1)
decreases,
(2)
transient
increases
followed
by
decreases
as
water
deficit
increased.
A
change
in
carboxylation
or
respiration
is
associated
with
variation
in
mesophyll
resistance
and
is
largely
independent
of
change
in
the
rate
of
transpiration
(Gaastra
1959).
Much
evidence
shows
that
mesophyll
resistance
to
gas
exchange
of
forest
and
fruit
trees
is
increased
by
water
stress
(Hsiao
1973,
Hansen
1971,
Slavik
1975).
Redshaw
&
Meidner
(1972)
calculated
that
about
half
the
reduction
in
photosynthesis
of
Nicotiana
during
a
drought
was
associated
with
stomatal
closure
and
the
rest
with
increased
mesophyll resistance
or
CO
2
evolution.
Osonubi
&
Davies
(1980)
attributed
the
reduced
rate
of
photosynthesis
of
droughted
Betula
pendula
and
Gmelina
arborea
seedlings
largely
to
increased
mesophyll
resistance
especially
at
high
temperatures.
Luukkanen
(1978)
found
a
similar
relationship
in
Picea
abies.
Kriedemann
(1971)
found
high
correlation
between
stomatal
resistance
and
mesophyll
resistance
of
orange
leaves.
He
also
emphasized
the
importance
of
a
nonstomatal
effect on
CO
2
diffusion
because
water
stress
greatly
increased
the
CO
2
compensation
value
in
lemon
leaves.
Hoare
&
Barrs
(1974)
attributed
about
half
of
the
water-stress-induced
decrease
in
photosynthesis
of
Washington
navel
orange
to
increased
leaf
mesophyll
resistance.
When
Bunce
(1977)
compared
12
forest
species
from
a
wide
range
of
habitats,
mesophyll
resistance
increased
with
lower
leaf
;G.
The
changes
in
mesophyll
resistance
paralleled
decreases
in
photosynthesis
and
increases
in
stomatal
resistance.
Chloroplast
activity
and
photosynthesis.
Boyer
(1976a,
1976b)
cautioned
that
although
photo-
synthesis
and
transpiration
(or
stomatal
resistance)
are
often
highly
correlated
it
should
not
be
assumed
that
photosynthesis
of
plants
undergoing
water
stress
is
regulated
solely
by
restricted
CO
2
diffusion.
Water
deficits
may
reduce
photosynthesis
by
inhibiting
chlorophyll
formation.
Chlorophyll
synthesis
is
quite
sensitive
to
water
both
prior
to
leaf
senescence
and
also
after
desiccation
induces
leaf
senescence
(Bourque
&
Naylor
1971).
Additionally,
chlorophyll
content
of
leaves
and
photo-
synthesis
are
well
correlated
(Keller
&
Wehrmann
1963,
Keller
&
Koch
1962,
1964).
Other
factors,
e.g.
changes
in
hormonal
growth
regulators,
may
also
influence
photosynthesis
(Kramer
&
Koz-
lowski
1979).
Several
studies
show
that
chloroplast
activity
is
affected
by
leaf
desiccation.
Oxygen
evolution
of
chloroplasts
isolated
from
droughted
plants
varied
over
a
wide
range
of
tji
values.
Oxygen
evolution
decreased
as
soon
as
photosynthesis
in
leaves
was
affected.
Correlations
were
also
high
between
chloroplast
activity
and
photosynthesis
of
intact
leaves
over
a
wide
range
of
leaf
IP
values
(Boyer
&
Bowen
1970,
Keck
&
Boyer
1974).
Boyer
(1976a)
concluded
that
since
chloro-
plast changes
occurred
at
the
same
values
of
leaf
q
that
induced
stomatal
closure
photosynthesis
could
be
decreased
by
either
factor
and
the
inhibitory
importance
of
either
factor
might
vary
as
environmental
conditions
changed.
At
low
light
intensity,
for
example,
photosynthesis
is
decreased
by
reduced
chloroplast
activity;
at
higher
light
intensity
stomatal
control
becomes
more
important.
After-effects
of
water
deficits
on
photosynthesis.
When
droughted
plants
are
rewatered
the
rate
of
photosynthesis
may
recover
fully
or
it
may
be
reduced
for
a
long
time,
depending
on
the
severity
of
the
water
stress
and
plant
species
(Davies
&
Kozlowski
1977).
The
failure
of
photo-
synthesis
of
water
stressed
plants
to
recover
after
irrigation
may
reflect
injury
to
stomata,
chloro-
plasts,
and
roots,
or
it
may
be
associated
with
leaf
abscission
(Kramer
&
Kozlowski
1979).
Midday
depression
in
photosynthesis,
generally
associated
with
temporary
stomata!
closure,
often
lasts
for
only
a
few
hours.
Progressive
soil
drying
over
a
period
of
days
may
damage
the
stomata
so
they
will
not
open
fully
after
irrigation,
or
it
may
injury
the
photosynthetic
mechanism.
The
rate
of
photosynthesis
of
Pseudotsuga
menziesii
that
had
been
subjected
to
drought
remained
low
even
after
the
stressed
trees
had
been
watered
(Zavitkovski
&
Ferrell
1970).
After-effects
of
a
prolonged
drought
on
stomatal
aperture
and
photosynthesis
of
Camellia
sinensis
may
last
for
weeks
to
months
(Squire
1978).
When
water-stressed
Pinus
taeda
trees
were
irrigated,
the
rate
of
Water
supply
and
tree
growth.
Part
I
Water
deficits.
73
photosynthesis
was
reduced
for
a
long
time,
largely
because
the
root
tips
had
been
injured
by
drought
(Brix
1962).
Shoot
growth
Water
deficits
inhibit
both
leaf
growth
and
internode
expansion.
Water
deficits
reduce
leaf
area
by
inhibiting
initiation
of
leaves
as
well
as
their
subsequent
enlargement.
Mild
droughts
during
mid-
and
late-summer
when
Pinus
strobus
buds
were
forming
reduced
the
number
of
needle
primordia
by
40%
(Zahner
1968).
The
number
of
needle
fascicles
on
Pinus
resinosa
trees
was
proportional
to
the
frequency
of
irrigation
during
the
previous
growing
season
(Clements
1970),
further
emphasizing
the
importance
of
water
stress
in
leaf
formation.
Most
of
the
reduction
in
leaf
area
as
a
result
of
drought
appears
to
result
from
slowing
of
cell
expansion.
Because
cell
enlargement
depends
on
cell
turgor,
the
elongation
of
cells
is
very
sen-
sitive
to
desiccation.
Water
stresses
that
are
too
mild
to
close
stomata
and
inhibit
photosynthesis
will
reduce
cell
expansion.
Water
stress
directly
and
physically
reduces
growth
by
lowering
cell
turgor,
because
cell
enlargement
in
response
to
change
in
plant
water
balance
occurs
too
rapidly
to
he
mediated
by
metabolism.
Furthermore,
when
mildly
stressed
plants
are
rewatered
resumpt-
ion
of
leaf
elongation
may
be
evident
within
seconds
(Hsiao
1973).
As
water
stress
intensifies
the
rate
of
cell
enlargement
decreases
rapidly
at
first
and
then
more
gradually.
Very
low
turgor
pressures
result
in
reversible
cell
expansion.
As
desiccation
proceeds
the
irreversible
cell
enlargement
characteristic
of
growth
stops
when
turgor
pressure
is
still
positive.
In
many
species
leaf
enlargement
is
so
sensitive
to
water
deficit
that
it
may
occur
only
during
the
night
(Boyer
1968).
Whereas
normal
diurnal
changes
in
leaf
dehydration
do
not
greatly
affect
final
leaf
size,
de-
siccation
for
long
periods
results
in
small
leaves
(Boyer
1976a,
1976b).
For
example,
leaves
of
droughted
Golden
Delicious
and
Jonared
apple
trees
were
much
smaller
than
those
of
well-watered
trees
(Simons
1956).
Daily
watering
over
a
30-week
period
resulted
in
more
than
4
times
the
leaf
area
in
Granny
Smith
apple
trees
and
14
times
the
leaf
area
in
Delicious
trees
than
was
produced
by
irrigating
every
3
weeks.
Some
of
the
reduction
in
leaf
area
by
withholding
water
was
the
result
of
early
senescence
and
shedding
of
leaves
(Chapman
1973).
Water
stress
also
reduced
final
leaf
size
in
grapefruit
(Levy
et
al.
1978).
Soil
drying
reduced
needle
elongation
in
Pinus
sylvestris
(Sands
&
Rutter
1959)
and
Pinus
strobus
seedlings
(Lister
et
al.
1967).
Needles
of
irrigated
Pinus
resinosa
trees
expanded
for
a
longer
time,
grew
faster,
and
were
40%
longer
than
needles
of
unirrigated
trees
(Lotan
&
Zahner
1963).
Elongation
of
Pinus
taeda
needles
was
directly
proportional
to
leaf
Ii,
with
growth
at
a
of
—1.02
MPa
less
than
half
of
that
at
—0.36
MPa
(Miller
1965).
The
rate
of
needle
elongation
in
Pinus
radiate
was
closely
related
to
osmotic
potent-
ial
of
the
rooting
solution
(Sands
&
Correll
1976).
Water
deficits
also
reduce
leaf
area
by
accelerat-
ing
leaf
senescence
and
inducing
early
abscission.
In
citrus,
for
example,
leaf
abscission
is
one
of
the
most
obvious
responses
to
desiccation
(Marsh
1973,
Kaufmann
1977).
Height
growth
and
elongation
of
internodes
of
lateral
shoots
are
often
affected
differently
than
leaf
expansion
by
water
deficits
(Kozlowski
1978).
A
summer
drought
may
or
may
not
influence
current-year
height
growth
depending
on
when
the
drought
occurs
and
on
the
inherent
pattern
of
shoot
elongation
of
the
species
affected.
An
important
feature
of
shoot
elongation
of
temperate
zone
trees
is
that
it
begins
early
in
the
frost-free
season
and
is
completed
well
before
the
frost-free
season
ends.
There
is,
however,
much
variation
among
species
in
seasonal
duration
of
shoot
elongation.
Height
growth
of
some
temperate
zone
species
is
completed
in
less
than
a
third
of
the
frost-free
season;
in
others
height
growth
may
take
twice
as
long
(Kramer
1943).
Hence,
late
summer
droughts
do
not
influence
current-year
height
growth
(or
elongation
of
lateral
shoots)
of
species
that
completed
shoot
elongation
early
(e.g.
before
the
drought
occurred).
In
species
with
"fixed"
shoot
growth
(e.g.
northern
pines,
Fagus,
Picea)
the
winter
bud,
which
formed
in
year
n,
expands
very
rapidly
during
the
early
part
of
the
frost-free
season
in
year
n+1.
In
Pinus
resinosa
height
growth
(and
elongation
of
lateral
shoots)
is
completed
by
early
July,
although
the
needles
continue
to
elongate
beyond
that
date
(Kozlowski
1963,
Kozlowski
et
al.
1973).
Other
species
that
complete
height
growth
before
or
near
midsummer
include
Pinus
con-
torta
Horton
1958),
P.
strobus
(Husch
1959),
Picea
glauca,
P.
abies
(Kozlowski
&
Ward
1961),
and
Tsuga
heterophylla
(Owens
&
Molder
1973).
In
species
with
"free"
growth
(e.g.
Populus,
Betula,
Liriodendron)
shoot
growth
involves
expansion
of
those
shoot
parts
that
were
present
in
the
winter
bud
as
well
as
initiation
and
ex-
pansion
of
additional
leaves
while
the
shoot
is
elongating
(Kozlowski
&
Clausen
1966).
Shoot
(internode)
elongation
in
this
group
occurs
over
a
much
longer
part
of
the
summer
than
it
does
in
species
with
fixed
growth.
In
southern
pines
of
the
United
States
and
many
tropical
pines
and
broadleaved
trees
the
shoots
expand
in
recurrent
flushes.
This
involves
formation
and
expansion,
in
the
same
growing
season,
of
a
series
of
buds
at
the
tip
of
the
same
shoot.
In
recurrently
flushing
pines
most
indi-
vidual
shoots
flush
2
or
3
times
but
as
many
as
7
buds
have
been
reported
to
form
and
expand
on
a
shoot
in
the
same
growing
season.
An
important
feature
of
recurrently
flushing
species
is
that
seasonal
internode
elongation
occurs
for
a
longer
time
than
in
species
exhibiting
fixed
growth.
74
Forestry
Abstracts
1982
Vol.
43
No.
2
Shoot
elongation
of
some
tropical
pines
may
occur
more
or
less
continuously
throughout
the
year
(Kozlowski
&
Greathouse
1970).
Whereas
late-summer
droughts
often
have
little
effect
on
current
year
shoot
elongation
of
species
with
fixed
growth
(shoot
elongation
completed
by
midsummer)
they
generally
reduce
the
amount
of
shoot
elongation
of
species
exhibiting
free
or
recurrently
flushing
growth.
For
example,
height
growth
of
Pinus
resinosa
was
not
significantly
influenced
by
late
summer
drought
(Lotan
&
Zahner
1963).
In
Pinus
taeda,
however,
late
summer
drought
reduced
the
number
of
growth
flushes
and
greatly
inhibited
height
growth
(Zahner
1962).
Late
summer
droughts
generally
have
carry-over
effects
in
the
subsequent
year
on
height
growth
of
species
with
fixed
growth.
In
Pinus
resinosa
favourable
water
supplies
in
late
summer
resulted
in
development
of
large
buds
containing
many
needles.
In
the
following
year
the
large
buds
expanded
into
long
shoots.
Hence,
it
was
possible
to
predict
ultimate
shoot
length
from
the
size
of
the
winter
bud
(Kozlowski
et
al.
1973).
Other
investigators
emphasized
that
favour-
able
water
supplies
during
the
year
of
bud
formation
determine
ultimate
shoot
length
of
species
with
fixed
growth
to
a
greater
degree
than
do
conditions
during
the
year
of
bud
expansion
into
a
shoot
(Friesner
&
Jones
1952,
Clements
1970).
Wood
production
Cambial
tissues
are
under
considerable
water
stress
almost
daily
during
the
growing
season
because
of
great
tensile
forces
that
develop
in
the
adjacent
mature
xylem
(Zahner
1968,
Stewart
et
al.
1973).
Up
to
90%
of
the
annual
variation
in
xylem
increment
of
forest
trees
has
been
attributed
to
water
deficits
in
arid
regions
and
up
to
80%
in
humid
regions
(Zahner
1968).
Several
aspects
of
cambial
activity,
including
division
of
fusiform
cambial
cells
and
xylem
mother
cells
as
well
as
enlargement
and
differentiation
of
cambial
derivatives,
are
very
responsive
to
changes
in
water
balance.
Evidence
for
extreme
sensitivity
of
cambial
activity
to
water
deficits
comes
from
correl-
ations
of
xylem
increment
with
rainfall
or
available
soil
water
(Fritts
1976)
as
well
as
from
thin-
ning
and
irrigation
studies
(Kramer
&
Kozlowski
1979).
Water
deficits
in
trees
are
often
critical
when
light
and
temperature
are
optimal
for
growth
(Kozlowski
1955,
1957,
1958,
1968a,
1968b,
1968d,
1969,
1979,
Kozlowski
et
al.
1962).
Variations
in
width
of
annual
rings
with
climate,
and
particularly
rainfall,
are
the
basis
for
dendrochronology
(Fritts
1976,
Creber
1977).
The
number
of
xylem
cells
produced,
seasonal
duration
of
xylem
production,
time
of
initi-
ation
of
latewood,
and
duration
of
latewood
production
all
depend
on
the
water
balance
of
trees.
Both
direct
and
indirect
effects
of
water
deficits
appear
to
be
involved.
Cambial
activity
is
inhib-
ited
directly
by
water
deficits