■ - pi

- ; |

I .

*£'&. \ ' r

I

4^** !*

7*

.?■;

v.-

IL^SR CRGSsJ|[
ERSI%OF CQY

^k» if/ "^5-81

afeLy^

r^^5S

yL/'lS

TRANSPIRATION AND ITS CONTROL
BY STOMATA IN A PINE FOREST

PAUL E. WAGGONER and NEIL C. TURNER

BULLETIN OF THE CONNECTICUT AGRICULTURAL
EXPERIMENT STATION, NEW HAVEN • No. 726, MAY 1971

Digitized by the Internet Archive

in 2011 with funding from

LYRASIS members and Sloan Foundation

http://www.archive.org/details/transpirationitsOOwagg

4-3

CONTENTS

Summary 3

Chapter 1 Transpiration and Its Control— An Introduction 5

Chapter 2 A Description of Plantations, Weather,

and Treatments 7

Chapter 3 The Foliage, Its Distribution and Effect on Shade 12

Chapter 4 The Stomata and Their Control 20

Chapter 5 Water Potential in the Foliage 31

Chapter 6 Bole Contraction, A Reflection of Water Potential 38

Chapter 7 Growth of the Boles, Interaodes, and Needles 48

Chapter 8 The Soil Water 59

Chapter 9 Evaporation from the Forest 66

Chapter 10 Transpiration and Its Control— A Discussion 82

Acknowledgments 84

Literature Cited 85

Cover photo: A red pine canopy.

SUMMARY

The loss of water by a forest and the possible reduction of this loss
by partial closure of the microscopic pores, or stomata, in the foliage were
investigated for 5 years in two red pine plantations in eastern Connecticut.

The seasonal and diurnal variations in stomatal resistance were ob-
served and related to light and shade. The water potential in the foliage
changed in a daily cycle, and stomatal resistance was increased by de-
hydration. The boles or trunks of the trees shrank as the foliage dried
each day, but the shrinkage lagged a few hours behind the cycle of
radiation and foliage dehydration. Growth of the boles and elongation
of the internodes and needles followed a normal pattern. The disap-
pearance of soil water, mostly via evaporation, equaled insolation and
also evaporation from an open pan of water.

Spraying with phenylmercuric acetate doubled stomatal resistance in
all needles present during the springtime spray; later spraying scarcely
affected the stomata in new needles. Bole growth was unaffected in the
first year, but was decreased in subsequent years. After 3 years of spray-
ing, the foliage was substantially decreased. Before any defoliation,
however, the change in stomatal resistance promptly reduced the drying
of foliage and shrinkage of the bole. Further, it retarded the loss of
soil water significantly, and well within the quantity predicted from a
comprehensive model of energy and water exchange within the forest.
The removal of about a third of the foliage late in the experiment also
decreased evaporation.

Thus a partial closure of the microscopic pores without destruction
of vegetation altered the energy balance and conserved water in the soil.
Over winter this yielded about 25 mm extra water to the groundwater
that feeds the reservoirs.

TRANSPIRATION AND ITS CONTROL
BY STOMATA IN A PINE FOREST

PAUL E. WAGGONER and NEIL C. TURNER

Chapter 1
TRANSPIRATION AND ITS CONTROL-AN INTRODUCTION

Connecticut's rainfall, 1200 mm in an average year, is modest, but
adequate. Of this 1200 mm received by precipitation, only about half
enters the lakes and streams, while the remaining half returns to the
atmosphere as evaporation from the forests and farmland. A portion of
this evaporates directly from the wet foliage, but the largest portion
enters the soil and re-enters the atmosphere after passage through the
plant. Thus in Connecticut, which has a million hectares of forests, a
very large portion of the total rainfall in the state returns to the atmos-
phere from forests.

The droughts in the 1960's demonstrated the close balance between
the supply and demand for water in the northeastern U.S.A. With an
increase in both the population and the per capita use of water, the
demand for water continues to grow. Indeed the water requirements
are expected to be 25% above the present requirements within the next
decade (U.S. Bureau of the Census, 1970). In the light of this, an un-
derstanding of water loss by vegetation is valuable, and means of con-
trolling this loss need investigation.

This bulletin reports such a study. Since a substantial proportion of
the evaporation occurs from forested land, experimental sites were
established in two plantations of red pine in eastern Connecticut. We
did not simply study the evaporation from a forest and the environmental
and plant factors contributing to the loss, although this was indeed
accomplished, but we also endeavored to manipulate the vegetation to
reduce water loss, conserve groundwater, and ultimately increase the
streamflow to the reservoirs.

The surface of the vegetation is perforated by many small pores, the
stomata, through which the water extracted from the soil finally escapes
into the atmosphere. Since the stomata are strategically located between
the wet leaf interior and the dry atmosphere, they were a key part of
our study. We observed the diurnal and seasonal pattern of their be-
havior and their response to light and dehydration of the foliage. Further,
our attack on the control of water loss was at this junction between
leaf and atmosphere.

Independently Ventura (1954) and Zelitch (1961) discovered that
a number of biochemical compounds closed stomata without general

6 Connecticut Experiment Station Bulletin 726

toxicity to foliage. Subsequently one of these compounds, phcnylmcr-
curic acetate (PMA), was shown to decrease transpiration relatively
more than photosynthesis in leaves of tobacco (Zeliteh and Waggoner,
1962), maize (Shimshi, 1963a), and cotton (Slatycr and Bierhuizen,
1964), and to decrease soil water loss more than growth in sunflower
(Shimshi, 1963b) and a grass sward (Davenport, 1967). Moreover, pre-
liminary studies revealed that PMA was effective in reducing the tran-
spiration of isolated pine trees (Keller, 1966; Waggoner, 1967); there-
fore, we employed PMA in our study.

We must emphasize that our objective was not to demonstrate that
PMA was a suitable chemical for treatment of watersheds in order to
reduce water loss, but to demonstrate that closing stomata could indeed
reduce evapotranspiration from a forest. Even though PMA was known
to damage foliage at hot temperatures (Wadsworth, 1960), it was the
logical choice at that time because its closure of stomata had been
demonstrated. More recently a non-toxic plant hormone, abscisic acid,
has been shown to close stomata ( Mittelheuser and Van Steveninck,
1969; Jones and Mansfield, 1970) and, with more work, might today be
our logical choice since it would not introduce a heavy metal into the
environment.

Although manipulation of the stomata had been shown to reduce
transpiration effectively in individual plants in pots, the effectiveness
in a forest was very much in question. Thus using the data available
in 1956, Penman ( 1956 ) concluded that meteorological parameters were
of primary importance in determining the "potential" evaporation from
vegetation, while plant factors had little effect. We accepted the pre-
dominance of weather in changing evaporation but attempted to reveal
the smaller effects of vegetation upon actual evaporation by frequent
observations in replicate adjacent plots. Our primary goal was to de-
termine whether small increases in stomatal resistance would reduce
the evaporation and measurably increase the water in the soil. However,
as our treatments ultimately produced forest canopies differing by one
third in leaf area, we also used this opportunity to study, in the final
year, the effect of leaf area on evaporation.

Other attempts at increasing the water yield from a forested water-
shed have taken the drastic route of killing or removing the trees
(Hewlett and Hibbcrt, 1961). Conversion of the mature forest to low
vegetation increased the supply of water by 130 to 400 mm at Coweeta
and by 100 to 300 mm at a number of other sites in the Northeast (Lull
and Reinhart, 1967). Such drastic treatment, however, creates seasonal
surges in streamflow and increases the risk of erosion. Stomatal closure
does not suffer from these faults.

The following chapters describe in detail the observations and results
from the 5-year study at Voluntown in eastern Connecticut. Two brief
reports of some of our observations from the first 2 years of this study
have already been published (Waggoner and Bravdo, 1967; Turner and
Waggoner, 1968); we now report on the complete 5 years of observations.

Transpiration and Its Control by Stomata in a Pine Forest 7

Our primary concern was the conservation of soil water by reducing
evaporation from the foliage. But we also sought to understand the
mechanisms governing water loss by vegetation. Thus we observed the
amount and distribution of foliage since it first absorbs and then employs
the energy to evaporate water within the needles. We next examined
the numerous stomatal pores, the variation in conductance as they were
shaded, aged, or were sprayed with the PMA. Since the vapor loss
changes the hydration, or water potential, within the foliage, we ob-
served the diurnal change in the hydration within the canopy and the
changes created by treatment with the stomata-closing PMA. Bole con-
traction is another measure of plant dehydration (Worrall, 1966) that
we used for a more widespread measure of plant hydration. Finally, we
measured the normal pattern of growth of the trees and studied whether
treatment affected this pattern.

Soil water content was our most numerous observation. We employed
this to follow the pattern of water depletion and replenishment through-
out the season, and to determine the effect of shrunken stomata on these
changes. From a knowledge of the rainfall we were able to use the soil
water observations to calculate the evaporation from the forest and
compare the untreated and treated foliage with theoretical changes in
evaporation.

Chapter 2
A DESCRIPTION OF PLANTATIONS, WEATHER, AND TREATMENTS

The study of water loss and its control by phenylmercuric acetate
(PMA) was conducted in two red pine (Pinus resinosa Ait.) plantations
in the Pachaug State Forest at Voluntown, in eastern Connecticut. The
forest, which comprises a mixture of hardwood and pine stands, is lo-
cated on abandoned farmland. The first plantation, known to the for-
esters as the Daniels lot, was planted as seedlings at 2 m X 2 m intervals
in an abandoned cornfield in 1931. For a decade it has yielded needles
for the nearby Pachaug State Forest Nursery. The second plantation,
known as the Bitgood lot, was established in 1944 by transplanting seed-
lings at approximately 2 m X 2 m intervals. The Daniels lot is 25 km and
the Bitgood lot is 33 km north of Long Island Sound.

The soil at the first plantation is a Windsor loamy sand deposited in
gentle slopes along the bank of the Myron Kinney Brook. The bulk or
apparent density of the soil at 15 cm depth varied from 1.27 to 1.40 at
ten sites, while at 30 cm it varied from 1.30 to 1.58.

On one day in 1965, 540 liters of water were applied to a square meter
of the soil. Two days later, the soil beneath was examined to a depth
of 3 meters, Table 1. The number of roots was observed by counting the
number visible on 2700 cm2 of soil profile. We found that the loamy
fine sand of the top soil was underlain by a coarser stratum at about 1

8

Connecticut Experiment

Station

Bulletin 726

Table 1. The

roots and

soil in the first plantation, 12

August 1965.

Depth,

Roots,*

Water,

Bulk

Silt,

Clay,

cm

dm-2

%d

density

%e

%e

Texture'

AMHCS

15

15.0

20

1.32

16

2.6

Ifs

10

30

4.0

18

1.30

16

1.2

lfs

10

60

3.8

11

1.48

8

2.3

fs

8

90

1.0

11

1.52

4

1.5

fs

8

120

0.5

16

1.67

11

0.8

fs

9

150

0.8

28

1.50

57

3.1

sil

21

180

Manyb

32

1.55

68

2.7

sil

24

210

0.3

26

1.55

58

1.2

sil

22

240

c

11

1.42

18

1.5

lvfs

11

270

9

1.37

42

0.8

vfsl

18

300

18

1.39

41

0.8

vfsl

17

a Roots dm-2 is number of roots visible on a soil profile between 0 and 15 cm depth

or in 30 cm centered at 30, 60, . . ., 210 cm depth.
b Too numerous to count.
c No observations below 210 cm.
dv/v
e w/w

'U.S. Department of Agriculture (1951).
6 Available moisture-holding capacity, per cent v/v calculated from texture (Hill,

1959).

meter. (This gravelly layer was later evident throughout the plantation
as it plagued our driving of tubes for moisture measurement.) The
gravelly layer, of course, held less water and had less available for roots.
It had fewer roots than the soil above or below. Near 2 meters lay a
silty stratum that held much water, had more available water, and grew
an abundance of roots. A lens of silt laid down in a prehistoric ripple
and 150 cm underground in 1965 had 75% silt and an available moisture-
holding capacity of 25%; the entire zone, of course, had an average of
only 57% silt.

The soil beneath 215 cm was sandier and had only 10 to 20% available
moisture-holding capacity as calculated from its texture (Hill, 1959).
Clearly the water poured upon the surface two days before had not
reached this depth— and neither did the wide pit needed for counting
roots reach 3 meters.

The soil of the second plantation, only 10 km northeast of the first,
was also a Windsor loamy sand, but it was not examined in the same
detail because its soil water content was not observed.

In 1966 the trees at the first plantation were about 15 meters in height
and had a mean diameter of 18 cm at 140 cm from the ground (Fig 1A).
In 1968 the stems had a mean cross sectional area of 48 rrr per hectare
or 209 square feet per acre. The plantation, which extended at least 30
meters beyond the area we used, was surrounded by hardwood forests.

Within the first plantation, 16 rectangular plots were laid out. The
plots were 7 to 12 meters wide and 27 to 75 meters long. They encom-

Transpiration and Its Control by Stomata in a Pine Forest

A

Fig. 1. A, View of first plantation in 1966; B, Spraying tree crowns with PMA.

10 Connecticut Experiment Station Bulletin 726

passed about a thousand trees. The plots generally extended down a
gradual slope and then a short distance up a somewhat steeper slope.
No runoff was ever seen, however. In 1968 an infection by the brown
root and butt rot (Fomes annosus [Fr.] Cke. ) caused us to abandon
four plots, and in the final two years, only 12 plots were used.

The trees at the second plantation had grown to about 11 meters tall
by 1968. Well within the plantation, which was also surrounded by
hardwood forest, 114 trees were selected in pairs. They had a mean
diameter of 17 cm at 140 cm height.

The weather during the experiments is described by the monthly
precipitation and the monthly evaporation from a U.S. Weather Bureau
Class A pan. The precipitation (Table 2) was measured at the head-
quarters of the Pachaug State Forest which is near Voluntown, 5 km
north of the first and 7 km southwest of the second plantation. The
evaporation (Table 3) was measured at Coventry, Connecticut, and
Kingston, Rhode Island, which are about 40 km northwest and 30 km
southeast of the plantations. The observations are presented as averages
of the two locations. We are grateful to the loyal observers who faith-
fully provided these climatic data.

From Tables 2 and 3 we see that during the months of June to Sep-
tember, in which the majority of our observations were concentrated,
1966 and 1969 had near normal amounts of precipitation and evapora-
tion: high evaporation in July 1966 and low precipitation in June 1969
are two exceptions. The 1967 season was, however, wetter and duller
than normal, whereas 1968 was noticeably drier and had greater than
average evaporation during July and August.

The trees were sprayed once in 1966, thrice in 1967, and twice in 1968
with phenylmercuric acetate (PMA), which shrinks or narrows stomata.
The PMA was obtained as 'Nildew AC30', a 30% solution manufactured

Table 2. Precipitation (mm) at Pachaug State Forest Headquarters.

20-year

Month

1966

1967

1968

1969

Mean

1

71

33

86

36

92

2

86

64

28

56

89

3

61

157

168

86

108

4

30

91

46

114

102

5

119

150

91

91

87

6

86

66

109

28"

60

7

86

107

18

130

83

8

48

69

48

64

88

9

137

102

69

107

96

10

89

48

51

58

87

11

132

79

135

132

117

12

66

180

155

231

111

Observations after May 1969 taken at Pachaug State Forest Nursery.

Transpiration and Its Control by Stomata in a Pine Forest 11

Table 3. Mean evaporation (mm) from a pan at Coventry, Connecticut, and
Kingston, Rhode Island.

13-year

Month

1966

1967

1968

1969

Mean

5

125

146

125

153a

136

6

136

150a

125

147

141

7

190

114

161

125

146

8

138

105

141

138

126

9

93

109

104

84

94

10

68

69a

71a

59a

71

a Evaporation at Kingston only.

by Naftone Inc., 425 Park Avenue, New York, N.Y. 10022. A 300 ppm
or approximately thousandth molar solution of PMA was prepared; this
is approximately the concentration that shrank stomata of tobacco
(Zelitch and Waggoner, 1962) and other plants mentioned in Chapter
1. One of two wetting agents manufactured by Rohm and Haas,
Philadelphia, was also mixed in the solution. In 1966 and 1967, 1000 ppm
of Triton B-1956, a water insoluble nonionic modified phthalic glycerol
alkyd resin, was used. But in 1968, 500 ppm of Triton X-100, a water
soluble nonionic isooctyl phenoxy polyethoxy ethoxylate, was used in
place of the Triton B-1956. Since Triton X-100 is water soluble, it pro-
duced a solution of PMA rather than the dispersion produced by
Triton B-1956.

The solution was applied from the ground by a hydraulic sprayer jet
that could strike one crown at a time (Fig. IB). The wetting of the
uppermost foliage was verified by examining shoots pruned from the
crowns. In 1966 after one spray, analysis of foliage by L. G. Keirstead of
the Station's Analytical Chemistry Department showed 4 to 19 ppm of
mercury in the treated and 0.1 or less ppm in the adjacent untreated
foliage.

The first plantation was sprayed on 2 June 1966; 8 June and 19 July
1967; and 4 June and 15 July 1968. It was also sprayed on 1 June 1967,
but that Nildew was a year old, the behavior of the pine stomata and
bioassay with tobacco leaf discs (Zelitch, 1961) indicated that the
Nildew had deteriorated, and the pines may be considered as having
received only two effective sprays in 1967. The second plantation was
sprayed only on 4 June and 15 July 1968.

The plots of trees sprayed were selected as follows. The 16 plots in
the first plantation were divided into eight pairs or replicates. One of
each pair was selected for treatment by the flip of a coin. The area
sprayed was 0.4 hectare, it encompassed 570 trees, and at each spray
it received 3500 to 4000 liters of solution in 1966 and 1967 and 2800 to
3200 liters in 1968.

In the second plantation and on the first occasion in 1968, a randomly-
selected individual tree of each of 45 pairs was treated. On the second
occasion, 22 of those untreated and 23 of those treated on the first oc-

12 Connecticut Experiment Station Bulletin 726

casion were sprayed. In addition, one of each of ten new pairs was
sprayed. On the second occasion, four more trees were sprayed with
wetting agent alone; subsequently these behaved and appeared the
same as untreated trees and were considered to be untreated. Conse-
quently, the number of trees was:

Untreated — 37

First treatment — 23

Second treatment — 32

Both treatments - 22

Total - 114

Since damage of the foliage had been observed in greenhouse studies
(Waggoner, 1967), we carefully checked the foliage after each spraying.
It was not affected at either plantation except when sprayed on 4 June,
1968 with the PMA and Triton X-100. A few sprayed needles at both
plantations became chlorotic and fell.

We now go on in Chapter 3 to describe the trees further by showing
the leaf area, defoliation, and absorption of illumination in the first
plantation.

Chapter 3
THE FOLIAGE, ITS DISTRIBUTION AND EFFECT ON SHADE

First, foliage on a tree was weighed. A tree with a diameter similar
to the plantation mean was found, needles were stripped from the tree,
and the needles were weighed according to their year of development.
The total dry weight of foliage of 9 kg corresponds to a foliage weight
of 16,750 kg ha-1 at the observed density of 1880 trees per ha.

The sample tree had 17 live whorls. A few of the needles that had
emerged 5 years earlier were still present on the lower seven whorls,
but the bulk of the foliage was less than 5 years old in the lower canopy
and less than 4 years old in the fourth to tenth whorls. The sample tree
had 28,000 or 2.4 kg of 1-year-old fascicles, 45,000 or 3.4 kg of 2-year-
old, and 35,000 or 3.1 kg of older fascicles.

Next, a broader sample of foliage was obtained by trapping fallen
needles. This both verified the quantity of foliage on the untreated
forest and revealed whether our treatment with PMA altered the quan-
tity. In the first plantation we placed buckets or traps, 19.0 cm in
diameter, on the ground south and 60 cm from the base of six trees with-
in each of the eight treated and eight untreated plots. In the second
plantation traps were placed beneath 114 trees. In June 1968 the 96
trees at the first plantation were reduced to 72 as the four infected plots
were abandoned (Chapter 2). The traps were placed in the first plan-
tation on 15 September 1967 and in the second plantation on 4 June
1968. The number of needles caught in the traps was observed peri-

Transpiration and Its Control by Stomata in a Pine Forest 13

odically throughout the year. The traps were emptied after each ob-
servation. To analyze the foliage caught in the 19 cm diameter traps we
simply calculated the mean and standard error of the number of fascicles
(i.e. pairs of needles) collected between two observations. This simpli-
fication avoided complications caused by missing observations and
variable numbers of replicates, and the standard errors were still suf-
ficiently small to reveal differences between treatments.

The normal pattern of needle fall in red pine plantations is seen in
the first and third columns of Table 4 and by the accumulation of the
same numbers of fallen needles in Fig. 2. A few needles fall between
April and mid-September, but most fall between mid-September and
late November. A few fall during the winter.

The validity of the 19-cm-diameter traps as samplers can be tested
by comparing the amount that they caught with the annual growth of
needles measured on Connecticut red pine by G. R. Stephens (unpub-
lished). In the first plantation the catch was 137 fascicles in 1968-1969
and 127 in 1969-1970. Since these fascicles weighed 63 mg each, the an-
nual catches were 3000 and 2800 kg ha-1. In the second plantation, the
catches were 174 and 146 in the two years. Since these also weighed
about 63 mg each, the annual catches were 3800 and 3200 kg ha-1. All

Table 4. Number of fascicles caught by the 285 cm2 traps during periods ending on dates
tabulated. A single observation of dry weight in mg is shown in parentheses on 9 July 1968.
The traps were placed on 15 September 1967 at the first plantation and 4 June 1968 at
the second plantation, and emptied after each observation.

First Plantation

Second Plantation

Date

Treatment

Treatment

Untreated

Treated

None

First

Second

Both

1967

Oct. 25

66.0±2.0

112.0±4.0

Dec. 7

16.0±1.0

16.0+1.0

1968

May 7

16.0±1.0

16.0+1.0

June 3

3.3±0.3

2.7+0.3

July 9

4.5±0.5

9.4+0.8

5.8+0.6

14.0+1.0

5.4+0.6

15.0+1.0

(July 9

282±31

633+28

369+33

958+110

321 ±29

978+94)

Aug. 8

3.0±0.5

7.6+0.8

3.9±0.3

8.8+0.6

4.7+0.3

8.6±0.6

Sept. 9

4.6+0.8

12.0+0.7

5.2+0.5

16.0+1.0

6.3+0.6

19.0+1.0

Oct. 18

58.0+2.0

101.0+4.0

44.0±2.0

74.0+4.0

46.0+3.0

69.0+4.0

Nov. 25

44.0±2.0

45.0±2.0

65.0±3.0

80.0+4.0

71.0+4.0

78.0+4.0

1969

April 17

23.0±1.0

18.0+1.0

50.0+3.0

67.0±3.0

54.0+4.0

56.0+3.0

July 11

7.3±0.5

3.9±0.3

6.7+0.4

7.6+0.6

8.0+0.6

7.7±0.7

Sept. 24

15.0±1.0

10.0+1.0

13.0+1.0

13.0+1.0

12.0+1.0

12.0+1.0

Nov. 19

79.0±3.0

55.0±1.0

70.0+3.0

66.0+4.0

62.0+3.0

69.0+4.0

1970

April 28

26.0±1.0

19.0±1.0

56.0+3.0

54.0+3.0

61.0+4.0

50.0±3.0

14

Connecticut Experiment Station

Bulletin 726

400

300 400
DAYS

500

700

Fig. 2. Fascicles accumulated in traps beneath untreated trees ( ) in

the first ( ▲ ) and second ( % ) plantations, and beneath twice treated
trees ( ) in the first (A) and second (O) plantations. Days accumu-
lated from 4 June 1968. Shading identifies autumn and winter.

four catches were well within the range of 2000 to 6000 kg ha-1 annual
growth of needles that Stephens weighed on other Connecticut red pine.
Further, the greater amount of foliage in the second plantation, which
was evident to the eye, is confirmed by the catch of fallen needles.

The effect of the PMA treatment can now be examined. In both plan-
tations the spray approximately doubled the fall of needles between
spraying in June 1968 and mid-October 1968. The results in the second
plantation show the first spray caused the defoliation. On the other
hand, the fall between October 1968 and November 1968, i.e. with no
further spraying, was little if any greater beneath sprayed than un-
sprayed trees. Considering one whole year, which we take as 4 June
1968 to 11 July 1969, we see the normal fall of 140 to 180 fascicles is
increased to 200 to 270 by spraying. This increase is a third to a half of
the normal fall. Presumably if this increase were continued for three
years, as during 1966-68 in the first plantation, the standing crop of
foliage would be decreased a third to a half.

On six occasions during 1967-68 the average weight of the fascicles
caught in the traps under the PMA treated trees was 63.4 mg, while
that of the unsprayed fascicles was 63.2 mg. This equality of weight
suggests that the foliage that fell into the traps was from all whorls of
the treated trees and not simply from the lower or upper portions of
the crown.

To establish the source of the needles caught in the traps more direct-
ly, 10 trees from the first plantation, and 16 trees Irom the second plan-
tation were sampled on four occasions between June 1968 and November
1969. On each occasion a single branch from the fourth to sixth whorls
from the top of each tree was removed, and the fascicles removed and
counted by year of origin. On each date a branch from the same whorl
was sampled. On the first sampling date, the bulk of the foliage was

Transpiration and Its Control by Stomata in a Pine Forest 15

Table 5. Number of fascicles per cm of internode (± standard error) on branches
from the fourth to sixth whorls from treated and untreated trees. The trees were
sampled in the four months as noted at A, B, C and D, and fascicles were tallied
according to year of origin.

First plantation

Second

plantation

Year of

Treatment

Treatment

Origin

Untreated

Treated

None

First

Second

Both

A. June 1968

1967

10.0±0.6

10.6±0.2

10.0±0.4

o

0

a

1966

11.3±0.4

9.8±1.0

11.1±0.4

o

a

e

1965

7.6±0.6

1.9±1.0

5.8±1.0

»

a

a

1964

1.2±0.1

0

0.4±0.2

*

*

«

B.

September 1968

1968

12.2±0.3

11.4±0.7

9.5±0.6

9.6±0.4

8.5±0.1

8.9±0.8

1967

10.3±0.5

9.7±0.6

10.6±0.9

8.9±0.6

10.1±0.4

7.8±1.0

1966

10.5±0.5

8.3±1.0

9.9±1.0

5.0±2.0

9.3±1.0

4.4±1.0

1965

5.9±0.4

0.4±0.3

4.1±1.0

0.4±0.5

3.7±2.0

0.1±0.1

1964

0.7±0.5

0

0

0

0

0

C.

November 1968

1968

11.5±1.0

10.6±1.0

9.9±0.9

9.6±0.5

8.8±0.5

9.3±0.6

1967

10.0±0.5

7.8±0.8

9.7±0.6

6.0±4.0

8.8±0.8

8.1±0.9

1966

9.6±2.0

0.2±0.1

4.6±2.0

3.6±4.0

4.7±4.0

1.1±0.4

1965

2.6±1.0

0

0.3±0.4

0

0.2±0.3

0.2±0.2

1964

0

0

0

0

0

0

D.

November 1969

1969

10.4±0.5

11.2±1.0

8.8±0.5

9.3±2.0

9.9±1.0

7.3±2.0

1968

11.7±0.4

10.5±2.0

9.1±0.9

9.4±1.0

8.5±0.8

10.0±0.8

1967

6.1±1.0

1.8±0.7

5.2±2.0

5.0±2.0

5.6±1.0

6.1±3.0

1966

0.4±0.3

0

0.3±0.3

0

0.6±0.5

0

1965

0

0

0

0

0

0

* Spray treatment not yet applied.

3 years old or less, with a few 4-year-old needles still attached (first
and third columns, Table 5); this is consistent with our earlier findings
from the sampled tree. Table 5 shows that the 3-year-old fascicles were
fewer per cm than were the 1- and 2-year-old fascicles, and it was the
3- and 4-year-old fascicles that were shed during September and Octo-
ber. Surprisingly, the number of fascicles per cm of internode was similar
in both plantations even though as noted earlier, the foliage appeared
more luxurious at the second plantation. As we shall show later, this was
largely the result of longer internodes, not longer or denser fascicles.

The additional needles caught in the traps under the trees treated
with PMA were clearly the older needles (Table 5). Prior to treatment
in 1968, the treated trees at the first plantation had shed almost an
extra year's needles. Treatment in 1968 increased the loss so that by

16

Connecticut Experiment Station

Bulletin 726

Fig. 3. Four-year-old branches from the first plantation sampled on 7 May
1969. The left branch was from a treated tree and the right branch from
an untreated tree. The numbers indicate the age in years of the internodes.

November 1968, after the needle fall for that year, more than an extra
year's needles had been shed. This is clearly seen in the sampled branches
shown in Fig. 3. When the treatment with PMA was omitted for one
year, the needle fall was smaller in the treated plots (Table 4), but the
foliage did not completely recover to that in the untreated trees (Tabic
5, November 1969).

At the second plantation, the foliage sprayed on 4 June 1968 had
fewer older needles by September 9. After the needle fall in 1968 there
were slightly fewer needles on the trees treated on 4 June. By contrast,
the second spray had no effect, and by November 1969 there were no
differences in numbers of fascicles per cm of internode between treated
and untreated trees, whatever the date of treatment. Apparently the

Transpiration and Its Control by Stonxata in a Pine Forest 17

effect of treatment with PMA was progressively more damaging, but
the foliage quickly recovered in density once treatment was terminated.

Plant pathologists, who use phenyl mercurial compounds to eradicate
fungal infection in foliage, are familiar with its yellowing of foliage in
warm weather (Wadsworth, 1960). Because of the noticeable discolora-
tion of foliage by the first spray in 1968, the damage to the needles was
evaluated before and after the period of maximum needle fall in 1968
( Table 6 ) . The percentage of the needle length of 25 fascicles that was
not green was observed. The yellowing or browning of the needles in-
creased with age. Treatment with PMA increased yellowing of the
older needles by September; these were the needles subsequently shed,
so that by November, after the major needle fall, the yellowing of the
remaining needles was not statistically different from the controls.

Shade or illumination beneath the crown of the trees should provide
a further indication of how much foliage the trees bore and how much
was removed by the sprays. The difference between illumination in a
clearing and on the forest floor should roughly indicate the foliage area.
This indication is only approximate because the absorption coefficient
for light is not exactly known and because wood as well as needles
absorbs light. On the other hand, the absorption coefficients of wood
of treated and untreated trees should be essentially the same, and the
difference between illumination beneath the two sorts of trees should
more precisely indicate any defoliation. Hence in the first plantation
and beneath overcast skies at 1100, 1200 and 1300 hours E.S.T. on 9
September 1968, and about 1000 on 21 June and 12 July 1969, illumina-
tion was measured beneath 72 trees by a foot-candle meter (Weston
Instrument Co., 614 Frelinghuysen Avenue, Newark, New Jersey 07114).

Table 6. Percentage (± standard error) of needle length that was not grteen. Samples
of 25 fascicles were taken from fourth to sixth whorls and from four or five years
of origin.

First p]
Treal

antation Second plantation

Year of

ment Treatment

Origin

Untreated

Treated None First

Second

Both

A. September 1968

1968

0

0 0.8±0.6 17.6±20.0

1.6+2.0

0.1+ 0.1

1967

2.9±1.0

2.8± 1.0 3.4+1.0 4.7+ 4.0

3.4+4.0

15.4±13.0

1966

6.5+2.0

10.6+ 3.0 5.2+2.0 15.9+ 8.0

7.7+3.0

13.8+ 4.0

1965

7.1+2.0

40.4+15.0 10.8+7.0

5.4+2.0

63.2+ 6.0

1964

10.5±5.0

B. November 1968

1968

0.6+0.4

0.9+ 0.4 1.7+0.5 1.4+ 0.8

1.4+0.6

1.7+ 1.0

1967

2.5±0.5

2.6+ 0.9 4.0+0.9 6.1+ 2.0

4.6+4.0

10.2+ 7.0

1966

4.0+2.0

6.2+ 4.0 1.8+0.4 3.0+ 3.0

4.1+2.0

7.6± 6.0

1965

6.2+4.0

18 Connecticut Experiment Station Bulletin 726

These were the same trees whose needles were trapped. Treated and
untreated trees were observed alternately, occasional observations were
made in clearings, and an entire set of observations required 30 to 45
minutes.

On the first day, illumination beneath the untreated trees was 30 to
40% of that in a clearing. On the last day it was about the same. The
absorption of illumination in a canopy has often been represented by
Beer's Law:

-l0gc (Ib/Ia) = kFu (1)

where I is the illumination above (a) and below (b) untreated foliage
of projected area Fu with an extinction coefficient k. The observed 30 to
40% transmission corresponds to a logarithm of — 1, making the estimate
of Fu equal to 1/k. The range of k, albeit in flat foliage, is 0.3 to 1.0
(Saeki, 1963), making the estimate of projected leaf area (Fu) equal
to 1 to 3. This was indeed the range of areas observed by G. R. Stephens
(unpublished) in several Connecticut plantations of red pine.

This estimate of the projected area can be compared with the area
calculated from the weight of foliage on the tree. The annual catch of
needles, in 72 traps and two years, was 2900 kg ha-1 in untreated trees.
Table 5 shows that these trees held 3 to 4 years of needles. Therefore,
multiplying 2900 kg ha-1 by 3Ja, we estimated that the trees had 10,000
kg ha-1 of foliage. This is less than the 16,750 kg ha-1 of foliage upon
the sampled median-diameter tree, but since it is based on a much
greater number of observations, we preferred it in calculating the pro-
jected leaf area.

The projected foliage area was calculated from the 10,000 kg ha-1
weight of foliage by multiplying by the mean of the two fascicle widths
(0.12 and 0.15 cm), the length (12 cm) and dividing by the 63 mg per
fascicle. This gave an area of 2.6 ha ha-1; the range of areas established
by the absorption of illumination, therefore, compared favorably with
this estimate.

Since we shall later need an estimate of foliar surface area, it is now
calculated. Red pine needles arc produced in pairs from a bundle sheath;
a pair of needles constitutes a fascicle. Each needle has one flat surface
which faces the flat surface of the other needle in the fascicle. A cross
section through a needle is similar to a cross section through half an
ellipse. To calculate the surface area (A) of a red pine fascicle in cm2,
each fascicle can be treated as two halves of an elliptical cylinder:

A = L [ ( ^VD^W) + 2W (2)

whore L is the length of the needle, W is the lesser fascicle diameter
measured at J2 L, and D is the greater fascicle diameter when the two
needles arc placed together and measured at I2 L. Treating the fascicle
as a cylindrical ellipse, rather than a tapering cylindrical ellipse, over-

Transpiration and Its Control by Stoinata in a Pine Forest 19

estimates the true surface area by less than 6%. The average surface
area of 150 fascicles was 7.8 cm2. This corresponds to a foliar surface
of 12.4 cm2 of epidermis per cm2 of land, well within the range of 5 to
17 in other Pinus spp. (Tadaki, 1966).

Having established the projected and total surface area of the un-
treated forest, we return to our observations of illumination. The com-
parison of treated and untreated trees is summarized in Table 7. It is
clear from the mean values of illumination beneath treated and un-
treated trees presented in the upper two lines that more light was trans-
mitted by the treated trees.

If we write Beer's Law for It and Ft, the illumination beneath and
foliar area of treated canopies, and then subtract this second expression
from the one for untreated foliage, we obtain:

log, (It/I.) = k'(Fu - Ft) = k Fu (Fu - Ft) /Fu (3)

The mean differences in the logarithm of illumination between each
part of treated and untreated trees are given in the third line of Table 7.
These differences were much greater than their standard errors, estab-
lishing the statistical significance of the difference in defoliation and
transmission of illumination caused by the spray. The differences were
about 0.2 in logarithmic units, which indicates that about a fifth more
light was transmitted through sprayed than unsprayed foliage. This
observation must be reconciled with earlier data. Observations of the
foliage itself, which have already been presented, indicate that by late
1968 a half to one third of the foliage in the first plantation had been
removed by spraying, i.e.

0.3<(FU - Ft)/Fu<0.5

Furthermore, the shade beneath the untreated trees has already been
shown to correspond to:

kFu = 1.0

This kFu includes, of course, the shade of the branches and would have
to be changed to, say, 0.6 for foliage alone. The product of 0.3 to 0.5

Table 7. Mean illumination (I) in foot-candles beneath untreated (u) and treated
(t) trees near midday in cloudy weather, and the mean logarithm of their ratio.

1968

Date

1969

1200h

September 9
1300h

1400h

June 21
HOOh

July 12
lOOOh

Untreated
Treated

L0ge—-

lu

370

473

.24±.04

182

242

.28±.06

181

220

.19±.04

764

959

.22±.03

338

408

.19±.04

20 Connecticut Experiment Station Bulletin 726

defoliation and the 0.6 for kFu then makes loge (It/la) equal to 0.2,
which is consistent with the mean observed in Table 7.

Having established the weight and distribution of foliage on the trees
and estimated the foliage area by several means, in the next chapter we
turn to the porosity of the foliage, which controls how much water will
be evaporated per unit area of foliage.

Chapter 4
THE STOMATA AND THEIR CONTROL

Since the cuticle of foliage is only slightly permeable to water, tran-
spiration is controlled— or so it logically seems— by the frequency and
varying dimensions of the microscopic pores or stomata that connect
the moist interior of a needle or leaf to the arid atmosphere outside. The
stomatal conductance or its reciprocal, stomatal resistance, is therefore
a fundamentally important characteristic in the hydrology of the forest.
The resistance has a further importance in our investigations because
the treatment with PMA was designed to alter the evaporation or hy-
drology by altering stomatal resistance.

The discussion of stomata proceeds in the following steps. First,
the needles and the arrangement of pores are described. Next the
peculiar wax in them is measured. Our method of measuring stomatal
resistance is described. Then the normal diurnal, height, and seasonal
courses of stomatal resistance are shown. As these courses are examined,
the effects of light, dehydration, and PMA upon stomatal resistance are
revealed.

The stomata occur in rows along the needle on both the flat and
convex surfaces (Fig. 4A); they number approximately 5000 per cm2.
The two guard cells surrounding the stomatal pore are found at the
bottom of a pit approximately 20 microns deep. The pits are filled with
wax which is readily stained with Sudan IV ( Fig. 4B ) . The wax cannot
be observed in the youngest stomatal pits deep within the protective
bundle sheath at the base of the needle, but each pit is filled with wax
when it emerges from the sheath. It is, therefore, probably loosely packed
flakes of cuticular wax, which enter the stomatal cavity as the needle
emerges from the bundle sheath.

When needles were dipped in chloroform for 30 seconds, 0.01 g of
wax was removed per g (dry wt) of needle or 7 X 10-5 g of wax per
cm2 of needle; this includes cuticular wax and that within the stomatal
cavity. The quantity of wax recovered by this standard procedure did
not vary during the growing season or with the age of the needles.
Since we shall see that stomatal resistance varies from day to night, we
assume that the wax in the stomata docs not greatly alter stomatal
control of transpiration.

Transpiration and Its Control by Stomata in a Pine Forest 21

IL Jam f/A "" '7/,a<kS'///a>. *' //. 1 P

///AimM

Fig. 4. A, View of the surface of a red pine needle, showing the rows of
stomata; B, Cross section of a pine stomata stained with Sudan IV.

The presence of wax in the stomatal cavity prevents measurement of
stomatal aperture directly under the microscope. Furthermore, the pres-
ence of the wax and the depth of the cavity precludes the use of silicone
rubber impressions (Sampson, 1961; Zelitch, 1961). These limitations
have led to the development of infiltration methods and porometer
methods for measuring stomatal aperture in conifers. The degree of in-
filtration by a solvent or dye at atmospheric pressure can be used as a
measure of stomatal aperture (Oppenheimer and Engelberg, 1965). Fry
and Walker (1967) described a pressure infiltration method, which is
more sensitive. The development of the diffusion porometer for the
measurement of relative stomatal aperture in broad leaves (Wallihan,
1964) was followed by its modification for both broad leaves and pine
needles (Kramer, 1969; Turner et ah, 1969).

In 1966 we estimated stomatal apertures by Oppenheimer and Engel-
berg's infiltration technique. Immediately after cutting, sunlit needles
were dipped into a solution containing 8 g of crystal violet dissolved in
20 ml of ethanol to which 90 ml of ether and 90 ml of chloroform had

22

Connecticut Experiment Station

Bulletin 726

been added. After 2 to 5 minutes the needles were washed in secondary
butanol and the differences between trees estimated by ranking the
needles from the most to the least infiltrated.

In the years following 1966, we estimated the stomatal aperture with
a ventilated diffusion porometer (Turner et ah, 1969; Turner and Par-
lange, 1970). The porometer has two basic parts: an acrylic chamber
into which pine needles can be inserted (Fig. 5) and a meter to show
the changes in humidity within the chamber. With a bunch of five fas-
cicles inserted 4 cm into the chamber, the time for a fixed change in
humidity was observed. The observed times can be used as a measure
of relative stomatal aperture or converted into absolute values of the
epidermal resistance to diffusion of water vapor. Turner and Parlange
( 1970 ) have shown that the stomatal resistance to diffusion ( rs ) can be
accurately calculated from the observed times (t):

W

l0ge

( *-* oo C0 )

(Coo -Ct)

[Vs + Vc]|-rp (4)

where A is the surface area of the needles within the chamber of volume
Vc. V8 is the effective volume of the sensor, i.e. the capacity of the
sensor to absorb water; and C^, C0 and Ct are the relative humidities
within the chamber when saturated ( C ^ = 100 ) , at t = 0, and at time t,
respectively. C0 and Ct are given by the manufacturer ( Hygrodynamics,
Inc., 949 Selim Road, Silver Spring, Md., U.S.A. 20910) of the lithium
chloride humidity sensor for different temperatures and electrical re-
sistances of the sensor. The porometer resistance rp, which is largely
governed by the boundary layer resistance, was obtained from dead
needles dipped in water containing a small quantity of wetting agent.
Both V8 and rp are temperature dependent as given in Table 8. Vc was
32 cm3. A, calculated from Equation 2 and corrected for the 6% de-
crease due to the taper of the needles, was 12.4 ±0.2 cm2.

Fig. 5. Side view of ventilated diffusion porometer for use with pine
needles. A, humidity sensor and head thermistor; B, fan; C, plunger con-
taining desiccant; D, motor; E, rubber coupling; F, scaled healing; G,
closed-cell sponge ruhher; II, electrical connector.

Transpiration and Its Control by Stoniata in a Pine Forest 23

Table 8. Effect of temperature on the effective sensor capacity (Vg) and the poro-
meter resistance (r,,) for pine needles.

Temperature

15C

20C

25C

30C

35C

v8

137
1.72

110
1.66

83
1.62

56
1.57

29
1.52

Sampling for the determination of stomatal resistance occurred at
intervals throughout the season. In 1966 infiltration resistances were
observed on four occasions, 23 June, 7 July, 26 August and 9 September.
In 1967 stomatal resistances were measured with the porometer ap-
proximately every two weeks from 1 June until 25 October, whereas only
occasional observations were made in 1968 and 1969.

For an individual tree the calculated stomatal resistances are means
of at least duplicate samples on needles of the same year. Where needles
of more than one year were sampled from the same branch, samples
were taken within 5 minutes of removal of the branch from the tree; a
change in stomatal resistance was not observed within this period. To
establish the seasonal trends and treatment effects on stomatal resistance
an average of eight (minimum of three) trees in each treatment were sam-
pled on any one occasion. Only branches from the fourth to sixth whorls
from the tree-top were sampled. On a number of days in 1967 the diurnal
variation in resistance was observed in both upper (fourth to sixth) and
lower (below the tenth) whorls from a pair of trees, one sprayed with
PMA and one unsprayed.

Immediately prior to sampling, the radiation above the branch to be
sampled was measured with an omnidirectional meter constructed from

2 silicon photovoltaic cells placed back-to-back inside a table-tennis
ball that acted as a diffuser. This was placed near the cutter of the pole
pruner used to sample the foliage. The meter was calibrated with an
Eppley pyrheliometer. Radiation measurements were obtained only
in 1967.

The diurnal change in stomatal resistance from just before sunrise to

3 hours after sunset was observed during the three 24-hour periods be-
tween 28 June and 27 July 1967 (Fig. 6). The stomata opened quickly as
the sun rose in the morning, and the resistance declined to a minimum of
15 to 20 sec cm-1 in the upper branches at midday. Resistances in-
creased again after sunset when the stomata closed. In the lower
branches, i.e. those below the tenth whorl, the resistances were more
variable, were generally higher than in the better lit upper branches,
and reached a minimum resistance of approximately 24 to 27 sec cm-1
at midday (Fig. 6).

The new foliage which emerged in July gave us an opportunity to
compare the diurnal change in stomatal resistance of newly-emerged
needles and those one year older. Fig. 7 shows that between 0800 and

24

Connecticut Experiment Station

Bulletin 72G

75

50

<-> 125 -

75

*

en

UJ 50

25 -

<>♦ 28- 29 JUNE
o» 11-12 JULY
□■ 26 - 27 JULY

Sunrise

Sunset

12 16

TIME, E.S.T.

20

24

Fig. 6. Diurnal variation in stomatal resistance of foliage from the upper

six whorls (closed symbols), or whorls below the tenth (open symbols)

on three dates in 1967.

1800 hours E.S.T. the resistance in the newly-emerged needles was ap-
proximately half that of needles a year older; the newly-emerged needles
had resistances of 7 to 12 sec cm-1, whereas in the 1-year-old needles
the resistances were between IS and 24 cm sec-1.

The effects of PMA on the diurnal change in stomatal resistance arc
shown in Fig. 8. The year-old fascicles had received one spray of PMA
8 weeks before and a second spray 1 week before these observations of
stomatal resistance. The newly-emerged fascicles had not emerged at
the time of the first spray and therefore had received only the PMA
applied 1 week earlier. Both treated and untreated needles closed their
stomata at night; but during the day, the treated year-old needles had
resistances approximately twice as great as the untreated needles: be-
tween 0800 and 1800 hours E.S.T. the resistances ranged between 30
and 60 sec cm-1 in the treated needles, whereas the resistances were

Transpiration and Its Control by Stomata in a Pine Forest

I50r

25

125-

75>

50-

25-

o« I I- 12 JULY
am 26-27 JULY

Sunrise

Newly emerged Sunset

12 16 20

TIME, E.S.T.

24

Fig. 7. Diurnal variation in stomatal resistance of newly-emerged (open
symbols) and 1 -year-old (closed symbols) fascicles on two dates in 1967.

between 20 and 25 sec cm-1 in the untreated ones (Fig. 8A). On the
other hand, the newly-emerged needles were unaffected by the PMA
(Fig. 8B).

Light is a primary determinant of stomatal aperture ( Lof tf ield, 1921 ) ,
and its effect can be seen in the diurnal change. We can construct the
daily march of radiation from our observations of omnidirectional radi-
ation above the sampled foliage on 28 to 29 June, 11 to 12 July and 26
to 27 July (Fig. 9). In the upper six whorls the radiation increased
slowly at first, then more rapidly as the sun climbed and sunlight pene-
trated into the stand, and reached a maximum 2 hours before noon. The
reverse was true in the afternoon. The sampled branches lower than the
tenth whorl were poorly illuminated throughout the day except for a
brief period early in the afternoon.

We used the diurnal changes in radiation to establish the relation
between radiation and resistance. In Fig. 10A the resistance in upper
foliage of untreated trees is identified by solid symbols and establishes
a standard response to radiation, which is the lower, solid curve. The
resistance changed little over a wide range of radiation, increasing mark-
edly only when radiation fell below 0.2 cal cm-2 min-1.

26

Connecticut Experiment Station

Bulletin 726

c\i 2

«

— _ eg

<u ts

c

r--^

.-I K

rt

V

s^***

i*S

rt

C\J

C

o

o S

*S CI

CO

CD >»

c

N1

o

c SJ

CVJ

c
o

o c

to

„D

Q s

,.H3 DBS '3DNV1SIS3H

The behavior of stomata in the lower foliage is illustrated in Fig. 10B
and can be compared to the standard response which is reproduced
from Fig. 10A. The resistance is considerable in the lower foliage both
because radiation is usually dim and because resistance is greater at
the same radiation. Nevertheless, the stomata in lower foliage are func-
tional and do open somewhat. The greater resistance of stomata in

Transpiration and Its Control by Stomata in a Pine Forest 27

0*28-29 JUNE
o* 11-12 JULY
□■26-27 JULY

<

<_> 0.5

Sunrise

Sunset

12 16 20

TIME, E.S.T

24

Fig. 9. Diurnal variation of omnidirectional radiation of the upper (closed
symbols) and lower (open symbols) foliage in 1967.

lower foliage has been observed in other plants (Turner, 1969; Monteith
and Bull, 1970; Turner and Incoll, 1971).

PMA, like location in the canopy, also alters the response of stomata
to radiation. Thus, treated stomata open and close with changing light,
but they do not open as widely as do untreated ones (Fig. 10A).

Water is another determinant of stomatal resistance, and its effect
can also be seen in the diurnal course. Since the next chapter is devoted
to plant hydration, however, the effect of water upon stomata is ex-
amined there, and we now turn to the seasonal course of stomatal re-
sistance.

The seasonal variation in stomatal resistance can be seen from the
measurements of stomatal resistance in the untreated trees obtained on
several dates between early June and late October 1967. Prior to the
emergence of the new fascicles, the resistance of the year-old fascicles
was 11 to 15 sec cm-1, but after the emergence of the new fascicles the
resistance of the year-old fascicles increased to 19 to 22 sec cm-1 (Fig.
11). Upon emergence, the new fascicles had a resistance half that in
the year-old ones for two reasons: those newly-emerged shaded the
older ones, and the new fascicles had a lower resistance than the fully
exposed year-old ones.

The resistances of both the new and year-old fascicles remained rea-
sonably constant from July to early September, but then both increased
in late October. Since the observations were restricted to the hours when
radiation changed little (Fig. 9) and were generally restricted to days
with clear or hazy sunshine, this constancy was expected. By October
the air temperatures were 5C lower than those in September and 11C

28

Connecticut Experiment Station

Bulletin 726

Ul

>-

>

z

-1

_l

3

Z>

z>

~3

~3

~3

0>

CM

N

CM

—

CM

CO

_

tf>

C\J

—

(VJ

LU

>

Z

_l

Z>

3

-3

"3

(T>

(VJ

C\J

—

CO

_

CO

—

-O-^

ss£

oq

>.

' t

• •

o

_l

CD

£<

b

<

C5

o 3

C

*? 0

z~

'•"*

<«

o

C/3

O

S

m
CM

i-
<

r2

Q CO

P^ O t/3

O

Q

U

s -r "o

<

9

cr

*"*

—.a i,

*^

J2 w m

B

o .

O -Q |

O

o

^ a* o

« p o

ca

H3 0)

GO

OJJ

W3

Of) +rf -4-»

o

o c

IH

"o s s

^-" o

o

rt

i_o

o

3

o rt >

—

S

m

T

s

Z

o

3 -fl

m

2

_o

t^_, -3 OO

O

M

c3

ca 2

s

-3

to .« >

rt ot -*

c_>

rt

— U _

O
in

_i
<

O

per fo
Is) tre
1 from

o

u

CJ

z

u

*-§§

o

W

a 1

m

K

d

■n w

CM

<

iH

e c

.S «

o

Q

tb

a

fa < —

i.W3 D3S '30NVlSIS3y

lower than those in July; wc presume that the increase in resistance on
the final sampling date occurred because of the cool soil and air tempera-
tures prevailing by this date (Stalfelt, 1962).

We have already seen how PMA increased the stomatal resistance of
the 1966 needles during daylight; wc now follow its effects throughout
the season. As observed with other species ( Keller, 1966; Davenport,
1967), the effectiveness of PMA in closing stomata was temporary. On

Transpiration and Its Control by Stomata in a Pine Forest 29

40

20

JUNE JULY

AUG

SEPT OCT

Fig. 11. Seasonal trend of stomatal resistance in newly-emerged (O) an<^

1-year-old ( 0 ) untreated foliage. Arrow indicates the approximate date

of emergence of new needles.

23 June and 7 July 1966, infiltration of the organic dye was greater in
untreated than treated needles, but thereafter no evidence of stomatal
shrinkage could be found on either the newly-emerged or year-old
needles.

In 1967, regular observation with the porometer showed that the ef-
fectiveness of PMA was greatest immediately after spraying, but it
quickly diminished (Fig. 12) so that one month after the first spray
the difference in stomatal resistance between the treated and untreated
trees was no longer statistically significant. The second spray of 1967
again increased the resistance of the year-old needles. This second treat-
ment with PMA increased the resistance more than the first so that on
the day after spraying the treated needles had a resistance 2.5 times
that of the untreated ones. This renewed difference also declined with
time, but even at the end of the sampling period, some three months
after re-spraying, the resistance of the treated needles was still 25%
greater than the untreated, a difference which was statistically signifi-
cant (p^O.001). However, the PMA failed to close the stomata of the
newly-emerged needles (Fig. 12), as observed earlier (Fig. 8B).

Simple wetting tests suggested that our inability to change the sto-
matal resistance of the newly-emerged needles was caused by inade-
quate wetting of the foliage. Therefore, in 1968 the water insoluble
Triton B-1956 was replaced by the water soluble Triton X-100 wetting
agent. Although adding Triton X-100 to the PMA significantly (p^0.05)
increased the stomatal resistance of the newly-emerged needles from
8.0 to 9.5 sec cm-1 in trees at the first plantation and 12.7 to 15.9 sec
cm-1 at the second plantation, this was a small increase compared to
that observed in older needles. Furthermore, the effectiveness upon new
needles of PMA with Triton X-100 disappeared within three weeks (see
Table 9).

We noted earlier that the smaller resistance of the newly-emerged
fascicles, relative to the year-old fascicles, persisted throughout 1967
until the final sampling on 25 October. When first sampled on 4 June

30

Connecticut Experiment Station

Bulletin 726

2 50

200

150

100

Sproyed Sprayed

I- year-old

Newly-emerged
/ \

y

/

/

i

JUNE J U LY

AUG

SEPT OCT

Fig. 12. Seasonal trend of stomatal resistance of newly-emerged ( O ) ar>d
1 -year-old ( % ) needles treated with PMA, relative to the stomatal re-
sistance of the untreated trees.

1968, the 1966 needles still had a higher resistance of 12.0 sec cm-1
compared to the 9.6 of the 1967 needles, but the difference could not
be substantiated statistically. To understand the effect of age better,
the stomatal resistance of needles from different years was also studied
on 20 to 21 June 1968 (Table 10). Stomatal resistance of untreated needles
increased with age, with the greatest increase between the 0- and 1-year-
old foliage (Tables 9 and 10). These observations from the second site

Table 9. Stomatal resistance (sec cm-1), with standard error, of foliage with different
years of origin, one day and three weeks after the second spray. The second
plantation. Six replications.

Age of
Foliage

None

Treatment
First Second

Both

A. 16 July 1968

Newly-cmcrged
1 -year-old
2-year-old
3-year-old

13±1
34±1
37 ±4
38±2

30±1
40±6
50±9

B. 8 August 1968

ie± i

69±16

70±10

107±13

87 ±22
118±13
293±13

Newly-emerged
1 -year-old

12±1
26±2

43 ±3

12± 1
30± 4

40± 4

Transpiration and Its Control by Stomata in a Pine Forest 31

Table 10. Stomatal resistance (sec cm-1), with standard error, of untreated and
treated foliage of different ages on 20 and 21 June 1968. Nine replications.

Age

Treatment

of foliage

Untreated

Treated

Newly-emerged

11±1

1 -year-old

17±2

33± 3

2-year-old

22±2

40± 6

3-year-old

23±2

65±10

4-year-old*

28±3

No foliage

* Four replicates only

were substantiated by similar observations at the first plantation. The
differences cannot be entirely attributed to age since the older needles
were presumably more poorly illuminated than the younger ones; the
difficult task of measuring the radiation receipt of needles of different
ages was not undertaken.

Tables 9 and 10 also show the effect of age upon the effectiveness of
PMA: the older the needles, the more effective the spray in closing the
stomata. The observations in Table 10 were made in the second plan-
tation, and the foliage had received only a single application of PMA
18 days earlier. The second spraying of the foliage also had a greater
effect on the older needles (Table 9). Although the stomatal closure
caused by the first spray was no longer evident when the second spray
was applied, the first treatment evidently predisposed the stomata to
greater closure on re-spraying. Moreover, the effect of a double spray
persisted longer than a single spray as revealed on 8 August (Table 9);
by this date the early senescence of those needles sprayed on 4 June
was already evident as an increased stomatal resistance.

Having seen the effect of light or hour, season and PMA upon sto-
matal resistance, we next examine water potential in the foliage, which
both affects stomatal resistance and is affected by the resistance.

Chapter 5
WATER POTENTIAL IN THE FOLIAGE

If the demand for water for transpiration becomes excessive, or the
resistances to flow in the pathway of water movement within the plant
or soil become too great, the deficit of water in the foliage can reduce
leaf turgor to the point at which stomata close. Thus we studied the
development of water potential in the foliage, called foliage potential
for short, and studied its effect on stomatal aperture.

The development of the pressure chamber (Scholander et al., 1964,
1965) provided a simple method for measuring foliage potential in the
field. The technique employs a strong, stainless steel chamber that can

32

Connecticut Experiment Station

Bulletin 726

-so-

98

: ;;r , itjt

Fig. 13. Pressure chamber system.
Left. Schematic diagram of system. A, cylinder of nitrogen gas; B, re-
duction valve; C, pressure gauge (0-210 kg cm-"); D, pressure gauge
(0-70 kg cm-2); E, metering valve; F, G, shut-off valves; II, pressure
gauge (0-70 kg cm-"); I, pressure gauge (0-21 kg cm-2); J, pressure

chamber; K, exhaust valve.
Right. Detail of pressure chamber. Front view. A, cover; B, D, inter-
changeable sealing plates; C, compression gland; E, chamber top; F, cylin-
drical chamber with inlet (lower) and exhaust (upper) ports; G, chamber
base, which is welded to F. Dimensions are in centimeters.

withstand pressure greater than 70 kg cm-2, a cylinder of nitrogen gas
under pressure, pressure gauges and a reduction valve (Fig. 13); the
system has been fully described by Turner et al. (1971). The cut stem
of a needle-bearing twig is sealed into the top of the chamber. Pressure
from the cylinder of nitrogen is applied to the twig until the meniscus
of the xylem sap returns to the cut surface, and the balancing pressure
is noted (Fig. 14). Boyer (1967), Kaufmann (1968 a, b) and De Roo
(1969, 1970) have shown that the balancing pressure is nearly equivalent
to the foliage potential measured with a thermocouple psychrometer for
a range of species. II. C. Dc Roo of The Connecticut Agricultural Ex-
periment Station confirmed that this was also true for red pine.

The diurnal pattern of potential in the upper foliage during three 24-
hour periods of clear weather is presented in Fig. 15. The potentials
encountered on the three days were not identical, but they all show a de-

Transpiration and Its Control by Stomata in a Pine Forest 33

5^*^ '. »., ,-\

' " v-"1' * "*_ ' tO

Fig. 14. Ben-Ami Bravdo and Dwight B. Downs using the pressure chamber

at Veluntown.

creasing potential in the morning with a slow increase in water potential
after 1400 hr E.S.T. as observed by others in different species (e.g.
Waring and Cleary, 1967). At sunset the foliage potentials were still
as low as — 6 to — 8 bars, but the trees continued to recover overnight
and had reached potentials as high as — 2 bars by daybreak. The mini-
mum foliage potential observed during the three days was — 14.7 bars.

34

Connecticut Experiment Station

Bulletin 726

Or

-4

8 -

-12

TIME, E.S.T.
8 12 16 20

24

16 L Sunrise

Sunset

♦ 28-29 JUNE

• II - 12 JULY
■ 26-27 JULY

Fig. 15. Diurnal variation in the water potential of foliage from the upper
six whorls of untreated trees on three dates in 1967.

If we select one of the three periods, we see that the water potential
of the lower foliage was higher (less negative) than that in the upper
foliage at all times during the day (Fig. 16); the data from the other

TIME, E.S.T
8 12 16

20

24

Sunset

Fig. 16. Diurnal variation in the potential of foliage from the upper six
whorls (♦) and foliage from below the tenth whorl (O) on 28 to 29

June 1967.

Transpiration and Its Control by Stomata in a Pine Forest 35

days verified this. Since we have already seen that the radiation incident
upon the lower branches was dim (Fig. 9) and the stomata did not
open as widely in the lower as in the upper foliage (Fig. 6), it is not
surprising that foliage potentials were less negative in the lower than
in the upper foliage. This potential difference with height, of course,
drives the transpiration stream in red pines as it does in other trees
(Scholander et al, 1964).

Next we examine whether stomatal resistance was affected by a de-
crease in water potential. To do this stomatal resistance was related to
the potential of the upper foliage on those occasions when the radiation
was greater than 0.2 cal cm-2 min-1, i.e. when stomatal resistance was
not affected by dim light (Fig. 10A). The range of potentials encoun-
tered when radiation was above 0.2 cal cm-2 min-1 was — 4 to — 15
bars. Fig. 17 shows the newly-emerged needles were sensitive to water
stress; the stomatal resistance doubled from 7.5 to 15 sec cm-1 as the
foliage potential decreased from — 4 to — 15 bars (r2 = 0.52). However,
the year-old needles, as best we can tell from the variable data, were
unaffected. Thus, if PMA increases the foliage potential, as well we
might expect, this should cause the stomata of the newly-emerged foliage
to open and thus act against the stomatal closing properties of the
chemical itself. This might be an additional reason for the ineffective-
ness of the spray on the newly-emerged foliage (Figs. 8B, 12).

During the season the mean foliage potential of the unsprayed trees
varied between —7 and —14 bars (Fig. 18A). Any seasonal trend that
there might have been was obscured by the variability in evaporative
conditions encountered. Evaporation from a Class A pan 3 km away
at the Pachaug State Forest Nursery varied from 2.3 to 6.9 mm on the
days that foliage potentials were measured; the foliage potentials were
roughly correlated with the daily evaporation (Table 11).

We also tested whether the leaf area had an effect on the water po-
tential developed by the foliage. On 6 and 7 May 1969, when the dif-
ference in leaf area between the treated and untreated trees was greatest
but the leaf porosity was similar, the mean foliage potential in 8 un-

Table 11. Daily pan evaporation and mean foliage potential of unsprayed trees at
various sampling dates throughout the 1967 season.

Date

Foliage Potential,
bars

Pan Evaporation,
mm

June 28

-14.1

6.9

July 12
18
19
26

- 6.8

- 8.7

- 8.6
-13.1

2.3
3.6
1.8
4.8

August 11
24

- 8.8
-13.0

2.3
3.3

September 12

-13.4

4.3

36

Connecticut Experiment Station

-i i 1 r

A A

year-old

Newly-emerged

Bulletin 726

25

20

70

m

CO

5 U>

n

m

CO

m
n

o

POTENTIAL. BARS

Fig. 17. Effect of foliage potential on the stomatal resistance of the newly-
emerged ( A ) and 1-year-old ( A ) needles when the radiation exceeded
0.2 cal cm-2 min- \

treated trees was —13.4 bars whereas in the treated trees it was 2 bars
higher at — 11.4 bars. Clearly, the shedding of one third of the fascicles
(Table 5) reduced the stress on the foliage.

PMA increased stomatal resistance in year-old needles (Fig. 8A) and
this increased the foliage potential ( Fig. 19 ) . During daylight the in-
crease in foliage potential from treatment with PMA was greater in the
lower branches than the upper branches: in the upper branches it was
increased by a quarter, whereas in the lower branches it was increased
by half. This suggests that PMA, which was equally effective in closing
stomata of upper and lower foliage, slightly increased the needle tem-
perature of the upper foliage that received the direct insolation. Such

Transpiration and Its Control by Stomata in a Pine Forest

JUNE JULY AUG SEPT

37

o

-3

-6

-9 I-

-12

-15
120

<
i-

UJ

o 100

Q.

UJ
>

<

_l
UJ

ce

OCT

Sprayed

Sprayed

IB I

i I

Untreated

80

Sprayed

1

i i
Sprayed

1

1

B.

i '

/ l

s

,o-~~~

-

___0

\
\

0-.

--<r'

X

y

JUNE

JULY

AUG

SEPT

OCT

Fig. 18. Seasonal trend in foliage potential, A, in untreated trees (♦) and

trees sprayed with PMA ( O ) ; B, in treated trees relative to the unsprayed

controls. The bars denote the standard error of difference.

a rise in temperature would increase evaporation by increasing the
gradient of water vapor from within the leaf to the atmosphere, and thus
partly compensate for the decrease in stomatal width. That this compen-
sation was only partial is clear from the actual increases in foliage
potential.

We have now seen from the diurnal study that PMA increased the
foliage potential throughout the daylight hours. Also we noted that the
effect of PMA on stomatal closure was transient (Fig. 12). Therefore,
we expect that the effect of PMA on foliage potential will also be
transient. This was indeed found to be the case. Foliage potential was
first measured on 28 June 1967, 20 days after spraying with PMA. At
that time, the potential of the sprayed foliage was a significant ( p^0.05 )
1.5 bars higher than in the unsprayed trees (Fig. 18A). However, by
the time of the subsequent sampling the effect of PMA was no longer
evident. Spraying the trees again on 19 July 1967 increased the potential
of the affected foliage (Fig. 18B), but the effectiveness of the PMA

38

Connecticut Experiment Station

Bulletin 726

-4

-8

12 •

-16

TIME, EST.
4 8 12 16 20 24

*"*

N Treated
Untreated \ b

Sunrise

Sunset

Fig. 19. Diurnal variation in potential of untreated ( ■ ) and PMA-treated
(□) foliage on 26 to 27 July 1967.

declined with time until the 0.5 bar difference at the final sampling on
25 October was no longer statistically significant.

These variations in foliage potential are reflected in the diurnal shrink-
age of the boles or trunks of the trees, which is examined in the next
chapter.

Chapter 6
BOLE CONTRACTION, A REFLECTION OF WATER POTENTIAL

The water in the xylcm of the tree trunk or bole is under tension, and
this tension manifests itself in a contraction of the bole that can be
measured. We have shown in the preceding chapter that several bars
of pressure must be applied to the outside of an excised shoot to replace
the tension in the xylem that was previously attached to the shoot, and
one should not be surprised to find that the change from the small 3
bars of tension at dawn to the 15 bars of tension at midday cause nearly
100 microns or 0.1 mm of shrinkage of the sapwood and phloem of a
bole. Since this shrinkage can be easily measured on many trees, it is
our most frequent and widely replicated measurement of the water
status of the trees.

Dcndromcters for the measurement of radial growth and diurnal
shrinkage of boles have taken two forms, viz: (a) band dendrographs
in which changes in the length of a metal band around the bark of the
tree at a fixed height arc recorded, and ( b ) dial gauge dcndromcters in
which the distance between the outer surface of the bark and a fixed

Transpiration and Its Control by Stomata in a Pine Forest 39

screw mounted in the heartwood is measured with a sensitive depth
gauge. MacDougal (1921) described several of the early band dendro-
graphs and used a modified version himself for continuous recording
of tree perimeters. The introduction of the dial gauge dendrometer can
be traced back to Reineke (1932); Daubenmire (1945) improved the
earlier model by replacing the single hook screwed into the heartwood
with three screws. Both types of dial gauge dendrometer have also been
modified for continuous recording (Fritts and Fritts, 1955; Klemmer,
1969). Tree bands are sensitive to temperature fluctuations and insensi-
tive to shrinkage. On the other hand, growth and shrinkage do vary on
different sides of the tree, and thus bands give a better average of total
bole growth than does the dial gauge (Bormann and Kozlowski, 1962).
The dendrometer that we employed is illustrated in Fig. 20. First, the
loose bark is removed from a bole at a height of 140 cm. Then a brass
ring is cemented to the sound bark. Next, three nails are driven into
the heartwood of the tree on the perimeter of a circle around the ring
(Fig. 20A). The movement in and out of the ring relative to the heads
of the nails reflects the shrinkage of the sapwood and phloem and is
measured with the depth gauge or "dendrometer" shown in Fig. 20B.
This particular dendrometer was developed in earlier investigations by
B. Bravdo and R. M. Samish and introduced into the Voluntown ex-
periments by Bravdo.

Dendrometer stations, i.e. rings and nails, were established on six trees
within each of the 16 plots, 8 treated and 8 untreated, of the first plan-
tation and on each of the 114 trees of the second plantation. In 1968
four plots at the first plantation that were infected with butt rot were
discarded; this reduced the number of dendrometer stations from 96
to 72. Generally measurements were taken near 1300 hours E.S.T. on
the first day and at 0700 and 1300 hours on the second day. The shrink-
age on the first day was calculated as the difference between the first
two of the trio. About an hour was required to complete all observations
in a plantation, and treated and untreated trees were measured alter-
nately.

The reproducibility of the measurements, including both the errors
of measurement and the variability of shrinkage among trees can be
judged by the standard errors of the mean shrinkages observed in the
first plantation on 32 occasions in 1967. From time to time a ring would
become loose and a measurement be omitted during its replacement;
thus the number of treated or untreated trees measured fell as low as
38 in the first set of observations, but usually the number of observations
averaged was near the ideal of 48. The mean shrinkage on the 32 oc-
casions ranged from a low of 5 to a high of 112 microns. The standard
error of these means ranged from 1 to 4 microns. Thus differences as
small as 10 microns between occasions or treatments were clearly sig-
nificant.

The response of the shrinkage and the dendrometer indications to the

40

Connecticut Experiment Station

Bulletin 726

A.

Fig. 20. A, Dendrometer station; B, Dial gauge dendrometer, used to
measure radial growth and bole contraction.

weather can be seen in Fig. 21. Since the shrinkage is attributed to the
excess of evaporation from the foliage above the supply from the roots,
it should increase with evaporation. On 21 days when dendrometer
measurements were made in 1967, the evaporation from a Class A pan
was measured between mid-morning and the following morning at the
Pachaug State Forest Nursery, 3.5 km from the first plantation. The
shrinkage of the boles on these 21 days is plotted in Fig. 21 as a function
of the evaporation. Clearly the shrinkage increases about 1.2 microns
with each 1 mm increase in evaporation; the correlation coefficient is 0.70.
The shrinkage of the boles should also be related, perhaps more
closely, to the radiation. That is, the evaporation from the foliage is
affected by both the warming of the foliage and the opening of stomata
that is increased by sunlight, while evaporation from a pan may be
greatly affected by the nighttime humidity and ventilation that have

Transpiration and Its Control by Stomata in a Pine Forest 41

80

70

60

o

cc 50

o

40

30

20

20 30 40

EVAPORATION, MM

50

60

Fig. 21. Relation between shrinkage of boles in untreated trees at the first

plantation and evaporation on the following day from a pan at the Pachaug

State Forest Nursery, 3.5 km from the plantation.

little effect upon the foliage when stomata are sealed by darkness. Thus
it is not surprising to find in Fig. 22 that the shrinkage is correlated even
more closely (r is 0.75) with radiation than with evaporation. The radi-
ation on the 25 days is the insolation indicated by a bimetallic pyrhelio-
graph (Belfort Instrument Company, Baltimore, Md.) also at the Pachaug
State Forest Nursery.

The effect of the weather was also evident in the diurnal course of
shrinkage. On four days in 1967 the shrinkage of trees on half of the
plots in the first plantation was measured at intervals throughout the
day. Fig. 23 shows the course on cloudy 22-23 June and clear 26-27 July.
The maximum shrinkage on the clear day was just twice that on the
cloudy day. Maximum shrinkage occurred between 1400 and 1600 hours
E.S.T., and minimum shrinkage was observed between 0600 and 0700
hours E.S.T. as observed previously by Haasis (1934). Thus bole con-
traction 140 cm from the ground lagged 3 hours behind the radiation
incident upon the upper foliage ( Fig. 24 ) except that the bole was still
some 20 microns smaller than its maximum diameter 3 hours after sunset.

42

Connecticut Experiment Station

Bulletin 726

80

70

60

o

50

cr

o

5

.

40

UJ

<s>

<

i£

30

^

cc

I

20

100 200 300 400 500
INSOLATION, CAL CM"2 DAY-1

600

Fig. 22. Relation between bole shrinkage of untreated trees at the first

plantation and insolation measured with a bimetallic pyrheliograph at the

Pachaug State Forest Nursery, 3.5 km from the plantation.

The relation between bole contraction and incident radiation pre-
sumably arises from its effect on stomatal opening and heating of the
foliage that, in turn, affect the water potential of the foliage. Indeed,
Worral ( 1966 ) showed that a good correlation between diameter and
water status did exist under equilibrium conditions. However, our first
attempt to correlate bole shrinkage and foliage potential under the dy-
namic situation of the field was not entirely successful (Fig. 25A).
Clearly, bole contraction at 140 cm from the ground lagged behind the
potential of the upper foliage on the same trees; a 2 hour lag was evident
(Fig. 25B). The lag is, of course, an interesting function of the relative
water potentials of atmosphere and soil, relative conductivities of soil
and plant organs and water capacity of the bole. This could be expressed
more explicitly in a simulator or model. Unfortunately, however, the
size of the various parameters is so poorly known that building the
simulator seems premature. The important thing is that bole shrinkage
reflects foliage potential, and the lag in bole shrinkage demonstrates a

Transpiration and Its Control by Stomata in a Pine Forest 43

100

22-23 JUNE
26-27 JULY

Sunset

Cleor

Fig. 23. Diurnal variation in bole shrinkage in unsprayed trees on a clear
( ■ ) and a cloudy ( A ) day in 1967.

considerable reservoir of available water within the foliage and bole of
the tree (MacDougal, 1921).

We next examined whether the shrinkage was a function of the nature
of the trees. Did shrinkage vary with trunk diameter or the expansion
of bole? The 14 linear correlation coefficients between trunk diameter
and the shrinkage on seven occasions in 1966 in the treated and in the
untreated trees were calculated. Only one of the 14 coefficients attained
the 5% significance level, leaving us convinced that big and little trees
shrink about the same amount. Since roots, xylem and crown grow in
about the same proportions, it is not surprising that big stems with big
crowns shrink no more than little stems with little crowns.

Also the relation between the shrinkage and the growth of the trees
was sought since it seemed reasonable that a rapidly growing tree might
have more tissue that could be shrunken by the diurnal drying than
would a slowly growing tree. Thirty-seven untreated trees had sound
dendrometer measurements of both expansion from 10 May to 6 July
1967 and shrinkage on 6 July. The expansion varied from 170 to 1860
microns and the shrinkage from 30 to 100 microns. The correlation be-
tween shrinkage and expansion in both these trees and the corresponding
treated ones was, unfortunately, negligible. Thus the shrinkage of the
boles is a function of weather and foliage potential but not of the range
of the bole size and expansion encountered in the Voluntown plantations.

44

Connecticut Experiment Station

Bulletin 726

100

80

60

40

20

A22-23 JUNE
♦28-29 JUNE
■ 26-27 JULY

A. No lag

♦

100

80

60

40

20-

♦ ■

B. 3 Hour lag

■ ♦

OLl

0 0.25 0.50 0.75 1.00 0 0.25 0.50 0.75 1.00

RADIATION, CAL CM-2 MIN"1 RADIATION, CAL CM"2MIN_I

Fig. 24. Relation between bole shrinkage and radiation measured near the foli-
age for three dates in 1967. A, shrinkage related to radiation observed at the
same time; B, shrinkage related to the radiation observed 3 hours earlier.

■

UNTREATED

a

TREATED

100

100

A. N

o lag

B 2

Hour lag

■ ■

80

■

80

33

■

■

Z
7Z

■

■

60

■

60

>

m

■ D

D

D

40

-

■

40

2
o

□

D

■

DD

■

o

■

D

■

2

d*

-

20

■

D

20

(/)

O

D

0 '

[£■

D

n

-12

-8

-4

0

2

-8

-4

0

POTENTIAL,

BARS

POTENTIAL

BARS

Fig. 25. Relation between bole shrinkage and foliage potential in PMA-
treated (D) and untreated (■) trees. A, shrinkage related to potential
observed at the same time; B, shrinkage related to the foliage potential

2 hours earlier.

Transpiration and Its Control by Stomata in a Pine Forest 45

Finally, we examined the effect of PMA on shrinkage. Since the rela-
tion between bole contraction and foliage potential was the same in
PMA treated and untreated foliage (Fig. 25), we can confidently use
our observations of the effect of PMA on bole contraction to provide a
broad demonstration of the effect of stomatal control on the foliage
potential of the forest. The relation between shrinkage in the treated
and untreated trees was observed in three ways: first, for a few ex-
emplary occasions; second, the ratio between the two for entire seasons;
and finally, the change in the ratio during a season.

No observations were made before the first spray in 1966, but shortly
afterward the relation between the two sorts of trees was typified by
shrinkage of 95±4 microns in the untreated and 72±4 in the treated.
By late August, however, the difference had disappeared. Before the
first spray in 1967, shrinkages of 92 ±3 and 82 ±4 were observed in un-
treated and treated trees; but after two sprays, they became 75±3 and
57±3, for example. In 1968, a difference between the two was evident
both before and after the sprays. Finally, in 1969 when the trees were
untreated, the shrinkages were 75 ±4 and 66 ±4 on 11 July.

In the second plantation, three observations were made before the
first spray, confirming the equality of shrinkage in the two sets of trees.
After spraying, shrinkages of 94 ±5 and 84 ±4 represented the untreated
and treated. The second spray had no detectable effect.

The ratio for an entire season can be seen in Fig. 26. The season is
1967, when observations were most numerous. The average ratio, shown
by the line, is 0.75. The observation before spraying is the point to the
left of the curve and at the top, showing that the trees to be treated a
second year shrank fully 90% as much as the untreated. The point to the
right of the curve near the top is the first observation after spraying,
when the ratio fell to only 0.72.

The mean ratios for all occasions in the four years are presented in
Table 12. The results from the first plantation seem easily interpreted.
The relative shrinkage of the treated trees diminished over the first three
years as treatment was repeated, stomata shrank, and especially in 1968,
foliage was diminished. Then, with no spraying in 1969, the formerly
treated trees shrank more nearly like the untreated ones. In the second
plantation the first spray decreased shrinkage to 90% in the absence and
87% in the presence of a second spray. Surprisingly, it persisted into the
second year.

The final view of the diurnal shrinkages is of their annual courses,
which reveal the onset and disappearance of treatment effect. These
courses are plotted in Fig. 27. Occasions when the untreated trees shrank
less than 40 microns were discarded because small values can cause
erratic jumps in the ratios. Then the ratios of treated to untreated were
calculated and plotted in the Figure together with the best fitting, two-
term polynomial, relating ratio to date.

In 1966 the effect of treatment and its disappearance towards season's
end are clearly visible. In 1967 the single observation before treatment

46

Connecticut Experiment Station

Bulletin 726

lOOr

0 20 40 60 80 100

UNTREATED SHRINKAGE, MICRONS

Fig. 26. Relation between bole shrinkage in PMA-treated and untreated
trees at the first plantation in 1967.

can be seen as the high ratio at the left, but it scarcely influenced the
polynomial curve; the treated trees then shrank less than the others
throughout the season. In 1968, the three observations before spraying
influenced the polynomial curve, and the onset and decay of the effect
of treatment are clearer than in 1967. Finally, in 1969 when sprays were
omitted, the formerly sprayed trees continued to shrink less than those
that were never treated; but the diminution in shrinkage was less than
in the years of spraying. We have no reason not to ascribe the small
diminution of shrinkage in 1969 to the lesser amount of foliage following
five sprays.

Our observations of the daily shrinkage of the treated trees, whose
transpiring stomata had been somewhat shrunk by PMA, can be sum-
marized quickly. Treatment was promptly followed by less shrinkage,
and in successive years of treatment the decreased shrinkage persisted

Transpiration and Its Control by Stomata in a Pine Forest 47

APRIL MAY JUNE JULY AUG SEPT OCT

or

x

CO

00
60

1967

Sprayed Sprayed

•

•• •

••
•
•

_ •

llJ 100
>

i-
<

UJ

or

60

00

Sprc
1968

yed Spr<

Dyed

- -i

60

00

60

00

1969

• - •_

•

__._•

60

•

i i i i i i

00

60

APRIL MAY JUNE JULY AUG SEPT OCT

Fig. 27. Seasonal course of daily bole shrinkage in treated trees relative to
the shrinkage in untreated trees during the 4 years of observation.

48 Connecticut Experiment Station Bulletin 726

Table 12. The annual mean ratios of diurnal shrinkages in treated trees to shrinkage
in untreated trees.

First

Plantation

Second

Plantation"

Number

of

Number oi

E

Year

occasions

Ratio

occasions

Ratio

1966

8

0.82

1967

32

0.75

1968

14

0.71

12

0.90

1969

6

0.88

5

0.90

° Trees sprayed first time versus no spray.

further into the summer than did our evidence of stomatal closure. Fi-
nally, the decrease persisted slightly a full year after treatment, prob-
ably because of the loss of foliage removed by the repeated sprays.

In the next chapter we go on to show how the early morning obser-
vations with the dendrometer were used to measure the radial growth
of the tree bole. In addition we observed the annual radial growth by
means of increment cores and measured the extension of the internodes
and needles.

Chapter 7
GROWTH OF THE BOLES, INTERNODES, AND NEEDLES

Growth is intrinsically interesting: we want to know whether the red
pines in the two plantations behave in the normal pattern (Kienholz,
1934). Even more interest, however, attaches to learning whether treat-
ment with the stomata-closing chemical materially decreased growth.

Two mechanisms for decreased growth following treatment with PMA
can be postulated. First, and most subtle, stomatal shrinkage could ob-
struct the intake of carbon dioxide and hence photosynthesis and growth.
This effect upon photosynthesis is, however, logically, and in fact in
several plants, relatively less than the effect upon evaporation (Zelitch
and Waggoner, 1962). Fig. 8 shows that evaporation from a needle is de-
creased by half or less, and Chapter 9 will show that evaporation from
the plantation is decreased less than a tenth during an entire season.
Hence growth is likely retarded little by the increased resistance to car-
bon dioxide intake following stomatal shrinkage.

The second mechanism of growth retardation by PMA is less subtle
but undoubtedly more important: foliage became visibly chlorotic and
fell following PMA treatment in June 1968. Traps revealed some de-
foliation as early as 1967. This damage to the already modest supply of
foliage in the first plantation was bound to affect growth.

The growth of the trees was measured in two ways: the expansion of
the boles or trunks and the elongation of the branches and needles. The

Transpiration and Its Control by Stomata in a Pine Forest 49

< Id
UJ h-
(E <

/

O

1

•\

1

\°

-

CO
<£>
<7>

~^u~-_°

ori

\N\N 'HlMOidO

Fig. 28. Expansion of the boles of untreated ( % ) and PMA-treated ( O )

trees at the first plantation. A, 1966 and 1967; B, 1968; C, 1969. Arrows

denote time of spraying with PMA.

50

Connecticut Experiment Station

Bulletin 726

expansion of the boles was measured by the day-to-day changes in the
same dendrometer that was used to measure shrinkage from hour-to-hour,
and also by cutting cores from the boles of the same trees at the same
height ( 140 cm ) as the dendrometer stations and measuring the annual
growth of wood microscopically. We begin with the dendrometer ob-
servations, which have just been used in the preceding chapter, and
follow the annual course of expansion of the bole as well as the annual
total growth.

The courses for the first plantation for the four years of the experiment
are shown in Fig. 28, parts A, B and C. The standard errors of the
increase between two observations were generally 5 to 20 microns. The
salient features of the annual course can be seen in 1967: the boles
expand rapidly during May through August, and the wood beneath the
dendrometer expands 1 to 1/2 mm per year. Year 1966 provides the great-
est departure from this pattern, but the smaller expansion is largely
caused by the late beginning of observations in June rather than by the
drought of that year. The greater growth in the second, more vigorous
and younger plantation is illustrated in Fig. 29.

The increasing inhibition of growth by PMA is evident in Fig. 28.
In the first plantation, the expansion of the treated trees was scarcely
less than the untreated ones in the first year, 1966. After the second
spray on 9 June 1967, however, the treated trees grew markedly less,
ending the year with a third less expansion. The outcome of the third
year was much the same. In the fourth year, although the trees were
unsprayed, the formerly treated ones did have less foliage, and they
grew only about half as much as untreated ones.

2 5
2.0

o
010

0 5
0

V

/

A

MAY JUNE JULY AUG SEP OCT

Fig. 29. Expansion of the boles of untreated trees at the first (♦) and
second plantation ( O ) in 1968.

Transpiration and Its Control by Stornata in a Pine Forest 51

CJ

O

Q.

UJ

(/)

CD

3

<

>-

_l

-~-"-M

Z>
—>

o:

w s

Ul

T3 'J3

2
z>

CD

— D

p o

i CD

>-

<

Ȥ CO

2

Ph 1

H3 >h

_l

a is

cr

c3-<

Q.

""Z*ci

<
>

#CO

^~-S

2 ffl Ph

cS .„ _,

goo -5

■3 CO -r

C05 f

3 H

60

M-l ~ rt

o << .5

h-

,„ >»

U

lag

O

o o a

JSM

c3

Q.

UJ

+2 e3

CO

y-.'a

o

_. T3

o

o C

3

.2 o

<

C CD

as w

>-

w-5

_l

Z>

o"S

- )

CO

tf)

UJ

,S>$

2

fe £

3

~5

>

<

^

_l

ce

CL

<

IAIIAI 'HIMOdO

52

Connecticut Experiment Station

Bulletin 726

The outcome of the spraying of the second plantation revealed the
effect of PMA more quickly, Fig. 30. That is, the increment immediately
after the first spray was only 267±15 microns in the treated trees, while
the untreated ones expanded 433±18 microns in 17 days. The second
spray had no effect. At season's end the treated trees had expanded only
70% as much as the untreated ones. Next, the annual total, rather than
course, of growth is observed in the thickness of new wood at the same
height as the dendrometer stations.

From increment cores taken on 19 November 1969 the radial growth
of the xylem, hereafter simply referred to as xylem growth, was observed
for five years at the first plantation and three years at the second plan-
tation. The normal year-by-year variation in the xylem can be seen from
the untreated trees in Fig. 31; in all years except 1966, a year with lower
rainfall and higher evaporation than average (Chapter 2), the xylem
growth in the untreated trees was between 0.9 and 1.0 mm at the first
plantation. Spring growth, characterized by large tracheids with light
colored walls, and summer growth, characterized by small tracheids
with dark walls, each comprised half the annual growth. The increment
cores confirmed that the annual growth at the second plantation was
twice as great as the annual growth at the first site (Fig. 32). But growth,
primarily spring growth, declined progressively over the three years in
the untreated trees, presumably as the canopy closed.

The xylem growth in the trees at the first plantation sprayed with
PMA became progressively smaller in each succeeding year (Fig. 31),
confirming the observations of the dendrometer. Both spring and summer
growth were affected by spraying; spring growth was affected to a
lesser extent than summer growth. Examination of the increment cores

1.0

0.8

X* 0 6

2 04

02-

965 1966 1967 1968 1969

Fig. 31. Radial growth of the xylem in untreated (U) and PMA-treatod (T)

trees at the first plantation for the 5 years prior to sampling in November
1969. Hatching denotes spring growth and shading denotes summer growth.

Transpiration and Its Control by Stomata in a Pine Forest 53

also showed that the immediate reduction in bole expansion observed
with the dendrometer after the first spray at the second plantation oc-
curred in the xylem; spring growth was reduced considerably less than
summer growth ( Fig. 32 ) . The immediate effect of the spray on summer
growth presumably arose from the damage by this one spray (Chapter
3) and is not typical, Certainly the second spray at the same site re-
duced growth less (Fig. 32).

The xylem growth in the year before spraying, i.e. 1965 in plantation
one and 1967 in plantation two, was, by chance, not exactly the same in
the trees selected for treatment as those in the controls (Figs. 31 and
32). Therefore, growth in the various treatments was compared relative
to growth in the year before spraying (Table 13). This showed that
the increasing effect of the spray on growth at the first plantation was
clearly significant in all years after the first. The immediate reduction
in growth resulting from the first spray at the second plantation was
also statistically highly significant; even the lesser reduction in growth
produced by the second spray could be substantiated statistically.

The 1967 and 1968 xylem growth of trees at the second plantation
was measured in both October 1968 and November 1969; the same
trees were sampled at the same height, but at different aspects of the
bole, on each occasion. Neither the 1967 nor 1968 growth changed in
width between the two sampling dates (Table 14); this showed that
cambial growth was completed by October and that no compression of
the xylem occurred.

Phloem growth was not affected by PMA. The phloem and cortex,
produced in 1968 and observed on 17 October 1968 in trees at the
second plantation, measured 1.96 ±.06 mm in the untreated trees and

2.0

1.6
- 1.2

o 1.8

0 I 2

0.4

0

1967

Fig. 32. Radial growth of the xylem at the second plantation for the 3
years prior to sampling in November 1969. Untreated (0) trees and trees
treated with PMA on the first ( 1 ) , second ( 2 ) and both ( 3 ) occasions in
1968. Hatching denotes spring growth and shading denotes summer growth.

54 Connecticut Experiment Station Bulletin 726

Table 13. Xylem growth of the bole as a percentage of xylem growth the year
before treatment. Measured from increment cores.

First Plantation

Second

Plantation

Year of
Growth

Treatment
Untreated Treated

None

Treatment
First Second

Both

1966
1967
1968
1969

90± 5
110± 9
117±12
114±13

80±4 n.s.
63±4t
56±4i-
48±4t

96 ±3
81±4

66±4i

42±3t

S7±3°
67 ±4°

81±7°
60±9°

° Significantly different from the untreated trees at the 5% level.
t Significantly different from the untreated trees at the 0.1% level.
n.s. Not significantly different from the untreated trees.

1.82 ±.08 mm, 1.84 ±.08 mm and 2.02 ±.10 mm in the trees sprayed on
the first, second and both occasions, respectively. The measurement of
phloem and cortex growth from previous years was difficult because of
uneven compression of the dead tissue and secondary cambial activity.
The bole expansion measured by the dendrometer was consistently
greater than the xylem growth measured from the increment cores in both
sprayed and unsprayed trees. Avoiding 1966 since the dendrometer sta-
tions were not established until well into the growing season, we see
that the bole expansion observed by the dendrometer in the untreated
trees in the first plantation was 22% greater than the xylem growth in
1969 and 68% greater in 1968; the mean difference for both plantations
over 3 years was 47%. The dendrometer, which measures growth between
the heartwood (into which the nails were driven) and the brass ring
of the dendrometer station, includes the growth of the phloem, cortex
and bark. Since treatment with PMA did not restrict phloem growth,
the difference between the bole expansion measured by the dendrometer
and xylem growth measured from increment cores should be even great-
er in treated trees. This was indeed true: the bole expansion was 37%
to 130% (with a mean of 72%) greater than xylem growth. The difference

Table 14. Comparison of the xylem growth (mm), measured from increment cores
at the same height and from the same trees, in October 1968 and November 1969.
Plantation two.

Date
Sampled

None

Treatment
First Second

Both

Oct. 1968
Nov. 1969

Oct. 1968
Nov. 1969

1.93 ±.08
2.02 ±.09

1.78 ±.09
1.91±.09

A. 1967 Xylcm

2.06±.14
2.04±.14

B. 1968 Xylem

1.28±.09
1.29±.()9

1.95±.12

2.04 ±.09

1.71±.12
1.80±.ll

1.84±.15

1.68±.18

1.15±.13
1.21±.ll

Transpiration and Its Control by Stomata in a Pine Forest 55

between the measured bole expansion and xylem growth was, however,
smaller than the phloem growth measured from the increment cores in
October 1968. This presumably arose from compaction of older phloem
as new phloem was produced; as noted earlier, the xylem is not com-
pressed.

Radial or cambial growth, which is often studied because of its eco-
nomic importance, is only one measure of the total production of the
tree; we also studied the growth of the internodes and needles. Extension
of the internodes, which begins earlier than cambial growth, reaches a
maximum in early- to mid- June and has ceased by mid-August ( Kienholz,
1934), and it apparently uses stored carbohydrates rather than current
assimilates (Larson, 1964). Needle growth, which begins about 2 weeks
later, is more dependent on current photosynthesis (Larson, 1964).

During 1968 and 1969 we observed the internode and needle growth
at both plantations. Ten trees at the first plantation and sixteen trees
at the second plantation were sampled on three dates in 1968 and again
in November 1969. A single branch per tree from the same whorl, in the
fourth to sixth whorls from the top of the tree, was sampled on each
date. The internode length for all years subsequent to 1964 in the first
and subsequent to 1966 in the second plantation was measured on lat-
eral branches. Further, the needles were removed, counted and the
length and dry weight of 25 undamaged fascicles obtained.

The rapid growth of the leader is shown in Fig. 33. When the trees

Fig. 33. Growth in length of the 1968 leader in 1968 and 1969 at the first
(♦) and second (O) plantations. Bars denote twice the standard error

of the mean.

56 Connecticut Experiment Station Bulletin 726

were sampled on 4 June 1968, the 1968 internode was still elongating;
at that time its length was 10.1 cm (range 7.0 to 12.5 cm), but by the
second sampling on 9 September 1968, the extension of the leader was
complete and no further increase in length was observed in the subse-
quent year; this confirms previous observations in red pine in New
England (Kienholz, 1934). Fig. 34 shows that the internode length
tended to decrease in succeeding years, and extension growth of the
pines at the second plantation was about 50% greater than in the trees
at the first plantation.

A few measurements of internodes in 1967 revealed no effect of spray-
ing with PMA. Similarly, PMA did not affect the internode growth in
1968, but a significant reduction was observed in 1969 (Table 15). This
is some two years after we first observed a marked decline in bole ex-
pansion at the first plantation (Table 13). At the second plantation,
radial growth was more quickly reduced by PMA (Fig. 30 and Table
13), and similarly the internode lengths were more quickly affected
thereby PMA (Table 15).

Needle lengths in our experimental plots are summarized in Fig. 35.
The length of the fascicles increased in succeeding years, the converse
of the observed decrease in internode length, Fig. 34. When first meas-

Table 15. Internode lengths as a percentage of the length in the year before treat-
ments were applied. Standard errors are also given. The trees were sampled in the
four months noted and the lengths are tabulated according to the year of growth.

First Plantation

Second

Plantation

Year of

Treatment

Treatment

Growth

Untreated

Treated None

First

Second

Both

A. June 1968

1967

79± 4

77± 5

1966

95± 4

111± 5

B. September 1968

1968

59± 4

52± 5 85±7

86± 4

92±18

86± 1

1967

83± 8

75± 6

1966

94± 6

116±14

C. November 1968

1968

71± 6

76±28 87±3

89± 4

92± 6

83± 2

1967

90± 6

82±21

1966

92±11

99±17

D. November 1969

1969

68± 3

46± 5 94±5

71±11

63±17

73± 8

1968

60± 3

50± 4 83±2

85± 7

71± 8

79±13

1967

81± 4

78± 5

I960

99± 4

101±10

Transpiration and Its Control by Stomata in a Fine Forest 57

6961

9961

I

4961

8961

'

Z96r

8961

Z96I

00

| 8961

Z96I

1 1

1 I

i i i

> en
o co

uj
-J

> oo o_

°£ ^

CO

UJC0Q
2 CO

o

ro

O
CO

e;

696)

9961

> CD

■BPP

Z96I

o <jo

9961

2cn

TKfc^^*y«^^^.

9961

>00UJ

1

8961

r

Z96I

k-i

9961

202 ^

<

i

S96I

CO
UJ

Q_00<
UJ CO °

1

8961

2.961

r

9961

co en

1 : ■.,*

5961

1

8961

w 00

1

Z96I

2S

<

1

9961

=*S

L

9961

1

o

ro

o o
oo —

IAID 'H19N31

C

)

Fig. 34. Internode lengths of untreated branches on three dates in 1968

and November 1969. The year of origin of the internode is noted. Branches

from the same whorl were sampled on each occasion: A, first plantation;

B, second plantation.

58

Connecticut Experiment Station

Bulletin 726

20

0L

First Plantation

CO

to
9) 0>

I

Second Plantation

20

J0

NOV NOV NOV NOV

1968 1969 1968 1969

DATE SAMPLED

Fig. 35. Length of needles at the first and second plantations measured
in November 1968 and November 1969. The year of origin of the needles

is noted.

ured on 16 July, the 1968 needles were still growing. On that date the
needle lengths were 10.9 ±0.4 cm on unsprayed and 10.0 ±0.1 cm on
treated trees at the first plantation. Needle growth had ceased by Sep-
tember, and no difference in length was observed between September
and November in needles of any age. PMA had no effect on needle
length at any time.

In summary, the dendrometer revealed 1 to VA mm expansion of the
xylem, phloem and bark in the boles during May through August in
the first plantation and more in the second plantation. Cores taken from
the boles showed that about two-thirds of this increment was xylem.
Although this growth in the first plantation was scarcely affected by
the first treatment with PMA, by the fourth year it was decreased to
half. In the second plantation, where the first spray caused defoliation,
the growth was promptly decreased. Extension of new internodes, on
the other hand, was much less affected and extension of needles was
not affected at all by PMA, making its total effect upon growth much
less than its dramatic halving of xylem growth in the last year.

Thus, PMA has shrunken stomata and will, in the next chapter, allow
us to test whether stomatal management can alter evaporation from a
real forest. PMA, as we have seen, affects more than stomata; it reduces
the surface area of foliage on the trees, and it can cause chlorosis, de-
foliation and decreased growth as observed in 1968. We must be careful
not to confuse the effects of defoliation and damage on evaporation with
the effect of stomatal shrinkage. Since treatment with PMA was omitted
in 1969, a year with little foliage, we can separate this effect from that
of stomatal closing on evaporation. Moreover the damage and poor
growth encountered in the later years was not characteristic of the first
year, which therefore gives us a measure of the effect of stomatal closure

Transpiration and Its Control by Stomata in a Fine Forest 59

alone on evaporation. Because foliar injury is a peculiarity of PMA, we
can go on to examine the reduction of evaporation by stomatal closure,
and use this as a measure of the saving of water that an ideal stomatal
closer would achieve.

Chapter 8
THE SOIL WATER

The water contents of eleven 30-cm-deep strata of soil were meas-
ured, and they show the supply of water available to the roots, which
supplied the water to the tree trunks and the transpiring foliage. The
change in water content of each stratum showed which soil region was
the favored source of water for the roots. Later these same observations
of water content will be employed to estimate the evaporation from the
forest. In this chapter, however, we only explain the method of meas-
urement, show the water content, and explore the causes for extraction
of water from one stratum rather than another.

The volumetric percentage of water in the soil was observed by the
neutron scattering technique (see, for example, Van Bavel et at, 1961),
and the percentage was converted into the amount of water stored in
the soil by multiplying by the depth of soil in the stratum. The change
in storage was indicated by subtracting observations at, say, weekly
intervals.

In the neutron scattering technique a source of fast neutrons is low-
ered into the otherwise undisturbed soil through an access tube. The
neutrons are slowed by collisions with the hydrogen atoms in the soil,
which are mostly in water molecules. Then a detector attached to the
source reveals the number of slow neutrons produced by the collisions
and hence the volumetric concentration of hydrogen atoms and water
molecules. The instrument was a Model 105A neutron moisture meter
manufactured by Troxler Electronic Laboratories, Raleigh, North Caro-
lina. The source of neutrons was 50 or 100 mc of americium beryllium
and the number of radiation events was counted for 30 or 60 seconds.
Five to twenty thousand events were detected. The active center of the
device was determined, and observations were made at 30, 60, . . .,
330 cm.

The variation in the observations, i.e. the proportion of radiation events
counted or counts per minute (cpm), is caused by variation in the
emission of neutrons, in the duration of the counting, in the position of
the probe, and in the variation from place to place of water and other
hydrogenous molecules in the soil. The variation in the water content
from place to place is by far the greatest of these sources of variation
(Hewlett et ah, 1964), making estimation of water content less precise
than the estimation of the change in water content at a given place.

60

Connecticut Experiment Station

Bulletin 726

Thus, on 27 August 1969 the mean content in 36 sites at a depth of 90 cm
was 6470 cpm with a standard error of 1174, whereas the change in
content during the preceding 47 days was 1211 with a standard error of
only 81.

The calibration of the probe in terms of water content is somewhat
troublesome, mainly because the neutrons are scattered through a larger
heterogeneous volume than the usual volume of a gravimetric sample.
Hence, we employed the manufacturer's standard calibration of the
probe in cadmium chloride (Van Bavel et ah, 1961), checked the func-
tion of the instrument at least daily by an observation in a shield of
organic matter, and converted cpm to volumetric water content by a
standard divisor. Nevertheless, we tested the standard calibration by
comparing the cpm to the water content of a core removed as the access
tube was inserted in the plantation soil and found these observations
agreed with the standard calibration as well as might be expected in
the variable soil (Fig. 36). The observed calibration was somewhat
above the standard. The slope of the observed calibration, however,

i

1 1 1 1

16

V /

14

-

/ /

_

v •/

o
o

/ /

o

12

-

Soil Calibration / /

X

(r2--0.94) / /

(-

10

-

/ /
/ /

3

/ '

5

8

-

/ v

/ /

or

LU

/ /^ Standard Calibration

0_

6

/ /

CO

/

2

4

' •

V

o

•

u

/

2

n

1

i i i i

10 15 20

WATER CONTENT, %

25

30

Fig. 36. The relation between volumetric water content and counts per
minute by the neutron moisture meter in the plantation soil. The dashed
line is the standard calibration obtained by submerging the probe in cad-
mium chloride solutions.

Transpiration and Its Control by Stomata in a Pine Forest 61

was insignificantly different from the standard. Because of the uncer-
tainty of calibration with small samples of soil and because the slope
alone is used in the critical estimates of change in soil water, the stand-
ard calibration was used.

The arrangement of the rectangular plots of the first plantation was
described in Chapter 2. The 16 plots generally extended down a gradual
slope and then a short distance up a somewhat steeper slope. At the
highest location in each plot, a tube was inserted 360 cm into the soil,
and at the lowest, another was inserted to the same depth. Four other
tubes were inserted to a depth of 210 cm in each plot. Each tube was
located about a half meter from a sound tree surrounded by other
sound trees; the former were the trees on which the dendrometer sta-
tions were established and the needle fall trapped. Thus in the 16 plots,
96 observations of soil moisture were made at 30 cm, 96 at 60 cm, . . .,
96 at 180 cm, 32 at 210 cm, . . ., and 32 at 330 cm. As reported earlier,
in the final two years a fourth of the plots were abandoned. This com-
pletes the description of the mechanics of observation, and we turn now
to the observations of soil moisture beneath the untreated trees during
four seasons.

Observations, as in Table 16, were taken on the occasions identified
by the small arrows beneath the depictions of each year in Fig. 37. The
water content on intervening days was interpolated linearly. The spots
in Fig. 37 were darkened by multiple printing as the soil became wetter.
Thus in the fourth stratum, 120 cm deep, the water content on 1 June

1967 was 24% (Table 16) and is heavily overprinted, whereas by 25
October 1967 the water content was only 16% and is not as heavily
overprinted.

Several features of soil water use and distribution are apparent. The
drought of the summer of 1966 (Chapter 2) caused a deepening deple-
tion of water, which was then somewhat recovered in the upper layers by
September rains. The alternate wet and dry months of 1967 caused
wetting and drying of the upper strata, while the drying of lower strata
proceeded much as in 1966. The protracted dry spell beginning in July

1968 dried the soil downwards in a regular pattern. Finally the drought
of June 1969 caused the upper soil to dry, but since observations were
not taken soon after the July rains, an extended drought was interpolated
for the upper soil.

The usual pattern of first depletion of water from upper layers, despite
roots extending to fully 2 meters ( Table 1 ) , is clearly evident in Fig. 37.
The soil at 330 cm remained very moist. On the other hand, the sandy
strata just above the lowest strata dried during September and October,
while the finer-textured soil of the seventh stratum remained relatively
moist; this behavior is as likely caused by the exchange— or lack of
exchange— between strata of soil as by extraction by roots.

The great number of observations available provide an excellent op-
portunity for testing hypotheses about factors affecting loss of water

62

Connecticut Experiment Station

Bulletin 726

"T3

S3 £

(J 08

o .2

s

r

X

V

N

B

H

c

O w

OJOniflONTfCINHffl

i— ii-ir-(oa<Nc<i(Nc<i

.— li-Hi— I i-H i-H i— I (N W (M W IN

HHHHHH(M!M(MM(N

COOCDCOCOOlOTftrtCOOO
HHHHHHNN1MMIN

HHHHH(MO]IMMCqCi|

hhhhhNNIMNNN

(C®tOO)ffiOOOWW^O!
HHHrtHOKSNMNM

©J.-|i-Ii-I.-iC<1(N<N<NOJ<N

(Oinwooo'-'OicDtowo)

HHHHcqlN(SM«NN

OCOCDOc^WOOOt^CDO
F-tr-trHOJcaOJCOCNOJCNCO

nwHcq^w5HO)(C(oo

HHOlNolMtlMMMIN

©oooooooooo

nOffllNlfloOH^NOO

HHHciMMnn

^r

o

J

J

CI

-1

05

J

J

t- J

co O

^^

3 o

i- j

6

E

a

o I
•a o

Transpiration and Its Control by Stomata in a Pine Forest 63

MAY

JUNE JULY

AUG

SEPT

OCT

1966

4-
7-

10-

T...timini

TTTTM»lftflf

n.Titti riimtHilllM

limtlUHHHtHtHnittM
Htllill If 11 it I I It ! t tHHHt

1967

4-

7-
10-

!l..r.....,,HIH!IHI!HHHIIIimHtt.. ....1..HHHI

miu.i.ii.iiiHimitiHitiWii miium mm »»!♦♦♦«♦

HIimilHIItltlttllKtltlHtllltimtTMTiMiTm.mmmmiiTillHII
MIMUdlHIIIIIIillHIIHIIIIIHIIHIIIIHIIHHimilltllltHllltHIHI
HIIIIHIIIIIHItlllHIIMIHIHIIIIIIIIItHlimimillHIHtHmimtH
MIIIIHIHmilHIHII»HIHIHIIIIIIIIHIIIfllllHllim»milltlllltlt
IIINMIIINIMIIIHIIIINIIIIIIIIIIIIIIII9IIIHIIHIII1HIHIUIIIIIIIIII

1968

1 i i

1 i i i A i Mi

i i

lllHIIIHHIHUHIIIIIIHHIHt,,,

uiiitiihiiiiiiiiiiiiiiiiiniiim,,..,,,,,

miiH!l!HIIIIIHIIIIhlllHimiHH,,„„...,1,Trimrrm

IIIIIIIIIHIHIIIIIIIIIHIIIIHIIII|j||l||||inillTtt>.I.,r[Til.T>lTrT.ri

liifiiitftiiiiiiifffi«ffffffMfiifiiMfifff*iiiiitiffittiiifiT>ITTITTTT>»T>T>TT

iiiiiHiiiiiiiiiiiiiiiiiiiimiiiiiiiiiiiHiMHHiimMimiiutiimiHiii'i

IIIIIIHIIIIIIIINHIHIIIHHNHIHIHHIIHHtllllHI If f IMtlf f III iiiiii fft tt»**K

1969

i 1

-4
-7

-10

tttttttttittttttTrrrrr itifTnin

IHHIHHIIIItlttHmTm. itiiiiitiimintiniii

IIHIillllll ff II lilllttt t tii llmrmmir »i» mmin rrrtttil iiiiii »♦♦

MiffJfftflfffffitttMiiitnitttttitiitttttiitttttttttttttttttttttitttt
flffftfffftf MtttfMtlfttliiMfiif f fiiff if iff ffifffiiflfifflftfffttMtt

4
-7
-10

MAY

JUNE

JULY

AUG

SEPT

OCT

Fig. 37. The water in 11 soil strata measured by neutron probe at the times

indicated by the short arrows beneath the panels depicting the 4 years.

First plantation. The overprinting was increased with increasing moisture.

(Program by T. Siccama, Yale Forestry School.)

from different strata. We are not, however, examining a controlled ex-
periment with a few, great forces at work; therefore, correlation will be
our tool, and only the simplest and clearest hypotheses should be
examined.

The simple hypothesis examined is that the loss of water from the
volume of soil examined by the probe, a sphere about 30 cm across, is
a function of the loss P from the volume during the period just past,
the amount of water I in the volume at the beginning of the present
period, the amount of water B above the volume, and the depth Z of
the volume below the soil surface. The rationale for the effect of past
loss, P, is that a large past loss may indicate that many active roots

64 Connecticut Experiment Station Bulletin 726

Table 17. The correlation coefficients among four independent variables: I\ the
past loss; I, the amount of water in the 30 cm spherical volume; B, the amount of
water above the volume; and Z, the depth of the volume below the soil surface.
From 1152 observations of group LL in Table 16.

I B Z

P, past loss -0.07 -0.42 -0.48

I, water in volume 0.2S 0.25

B, water above 0.94

occupy the volume. Since the available moisture-holding capacity does
not vary greatly ( Table 1 ) , ample water I in the volume allows ample
loss. Since lower roots extract water more rapidly after upper ones have
depleted their supply (Hunter and Kelley, 1946), sparse water B above
the volume should increase the loss from the volume. Finally the ration-
ale for the depth Z is that more roots arc found near the surface, and
thus more water should be lost there. Z was taken in strata of 30 cm,
i.e. 1, 2, . . ., 6, and the lower depths were omitted because fewer ob-
servations were made there.

The 1967 observations provide the most numerous data. They were
divided into three groups, which arc identified at the bottom of Table
16. The group L occurred in the earlier part of the season, and they
are three periods of loss of soil water. Two out of three periods were
preceded by gains. The G group are the four periods of gain. Finally the
four periods LL are later periods of loss, all preceded by losses. In all
periods the loss in cpm, either positive when moisture decreased or
negative when it increased, was standardized to a loss per day. To
eliminate the effect of the weather, which is not under study in this
section, the mean loss for all depths for the period was subtracted; for
example, if the mean loss for a period was 100 cpm per day and the
water in a particular soil volume decreased 90 cpm per day, the de-
pendent variable or loss was — 10 cpm per day for that particular soil
volume.

Observations from 1968 were also examined. The periods of 20 June
to 9 July, 9 to 15 July, and 15 July to 8 August are called L68 because
they were times of loss preceded by losses. A fourth period, 8 August to
17 October, was examined separately.

The correlations among the four independent variables arc shown in
Table 17 for four periods of loss preceded by losses (LL, Table 16).
The correlation between B and Z is obvious: the deeper the volume, the
more the water above. The table also shows that the deeper the volume
—and the more the water above— the greater is the amount of water I in
the volume. Finally, the deeper the volume, the' smaller is the past loss P.

To segregate the independent effects of P, I, B and Z upon losses of
water from the soil, we employed a multiple regression analysis. Tin-
standard partial regression coefficients in Table 18 are designed to show
the relative potency of the four independent factors upon the dependent
variable, water loss. Being "partial" they show, for example, the el I eel

Transpiration and Its Control by Stomata in a Pine Forest 65

Table 18. Standard partial regression coefficients of the loss of water from a volume
of soil on four independent variables: P, I, B, Z.

Number of

Independent variables

Group

observations

P

I

B

Z

G67

1152

0.30

0.10

-0.43

1.13

L67

864

-0.46

0.03

-0.22

-0.06

LL67

1152

0.12

0.24

-0.14

-0.41

L68

648

-0.04

0.13

-0.08

-0.58

of P independent of I, B and Z. Being "standard," they permit the com-
parison of the effectiveness of, say, B with Z although B has a standard
deviation of over 10 thousand while Z varies from only 1 to 6. The co-
efficients of Table 18 are, of course, dimensionless.

The first line of the Table pertains to periods when moisture near the
soil surface was increased more by rain than it was depleted by evapora-
tion. The 0.30 coefficient indicates that volumes that had lost consider-
able moisture P in the past, indicating a zone of active roots, continued
to lose water at a fast rate. However, the water content I had little
effect under the conditions of a gain in water. As expected, the strata
of soil with little water B above lost more water than those with plenty
of water above them. Finally, the positive relation between depth Z
and loss indicates that shallow volumes gained water.

During periods of loss in spring and summer, L67 (L, Table 16), the
volumes of soil that had gained more water also lost more as shown by
the — 0.46 relating P to loss. Further the shallow depths did not gain
more, as might be expected under these conditions.

The third and fourth lines in Table 18 represent periods of decreasing
soil moisture within longer periods of drying. Clearly, past loss P and
the amount of water B above matter little at these times, while the loss
is greater at shallow depths. Finally, during late 1967 the soil moisture
I in a volume began to limit the loss, as shown by the coefficient 0.24.

The seasonal change in the influence of I caused us to examine the
change during 1968 of the simple correlations between loss and P, I and
B for single periods of observation throughout a season in which the
soil water progressively declined (Table 19). Because B and Z are
closely correlated, Z was not considered separately but unquestionably is
confounded with B.

During the early summer, water is lost more rapidly from those vol-
umes in which prior losses P were greater, but this correlation is lost
with declining soil moisture. On the other hand, the correlation between
loss and the water content I in a volume grows as the soil becomes drier
during the late summer and early autumn. Finally, the relation between
B (or Z) and loss changes direction; i.e. in the early season greater
losses occur from the shallower volumes with little water above them,
while in the later periods the greater losses occur from the deeper vol-
umes with more water above them.

Thus, the hypothetical relations between loss and the four independ-

66 Connecticut Experiment Station Bulletin 726

Table 19. The seasonal course of the correlations between 216 observations of
change of moisture in volumes of soil and three dependent variables — P, I, and B —
and the mean water content, 1968.

Period

Variable

20 Jun-

9 Jul

9 Jul-

15 Jul

15 Jul-
8 Aug

8 Aug-
17 Oct

P
I
B

0.50

0.05

-0.79

0.61
-0.24
-0.61

-0.31
0.46
0.24

0.19
0.66
0.60

Mean water con-
tent, % v/v

20

17

16

14

ent variables depended upon whether the soil was gaining or losing
water and whether it was wet or dry. Losses were greater where past
losses were greater if the soil was moist but not recently moistened.
Losses were greater where moisture was greater if the soil was dry. The
moisture above a volume had rather little effect, but depth had much,
with shallow volumes losing much water when the entire soil profile was
moist and deep volumes losing more when the profile was drier.

We shall see in the next chapter that PMA reduced the evaporation
from the forest. Logically, and in fact, this increased the amount of
water in the soil under the treated trees; the increases were small, how-
ever, compared with the total water in the soil. Since it is more mean-
ingful to discuss the effect of PMA on evaporation, rather than on soil
water content, we shall defer our discussion of PMA and soil water, and
first explain the use of our observations of soil water to calculate
evaporation.

Chapter 9
EVAPORATION FROM THE FOREST

Evaporation is, in a sense, the culminating subject that the preceding
chapters lead to. It will be estimated from the soil moisture observations
described in Chapter 8 and from rainfall observations. These estimates
will be presented after a theoretical framework is given for the effects
of weather and foliar area and stomata upon evaporation.

Weather is surely the predominant controller of the evaporation from
a heavy and moist canopy of leaves. Thus, Thornthwaite ( 1948 ) and
Penman (1956) were able to estimate "potential" evaporation from the
weather alone, neglecting the complications of differences among plants.
We shall examine the effect of weather upon evaporation from the
Voluntown pines, both theoretically and by observation.

Beneath the maximum set by potential evaporation, however, there
is scope for significant differences in evaporation caused by differences

Transpiration and Its Control by Stomata in a Pine Forest 67

among plants. Some of the differences were incorporated into the Pen-
man (1956) equation: it has a place for altered reflectivity and rough-
ness. Thus a dark, rough forest was expected to have a rapid potential
evaporation; in fact, Penman (1967) calculated that conifers on a lysi-
meter in Castricum, Holland evaporated fully an eighth more than poten-
tial transpiration from a very dark stand of plants. The Penman ( 1956 )
equation even included a term for an average stomatal resistance for
the entire canopy, but there was as yet no explicit rule for how the
stomatal resistances in the several levels of the canopy would be com-
bined with foliar area to produce the average resistance.

Experiments by foresters, who cut and pruned trees, showed that
severe changes in leaf area would increase the flow of water downstream
(e.g. Hewlett & Hibbert, 1961; Lull & Reinhart, 1967), but many ques-
tioned whether smaller differences— as between broad-leaved and conifer
trees— would affect evaporation and hence the yield of water to streams.

Chemicals that would shrink stomata without general toxicity to foli-
age were discovered by both Ventura (1954) and Zelitch (1961).
These proved powerful tools in revealing any role of stomata as well as
any effect of small differences between plants in the control of evapora-
tion. Thus, decreases in evaporation through partial stomatal closure were
soon produced in several cases of individual leaves or plants standing
alone, as reviewed in Chapter 1. Eventually, evaporation from a barley
crop was decreased by stomatal closure (Waggoner et al., 1964). When
an entire watershed covered by a hardwood forest was sprayed from
the air with a stomata-closing chemical, however, the stomata on the
undersides of the leaves were not reached by the spray and the experi-
ment failed (Waggoner and Hewlett, 1965).

Thus two tasks remained: to develop a model of energy exchange and
evaporation within a canopy that will logically relate foliar area and
stomatal resistance to evaporation, and to show experimentally whether
doubling the resistance of the microscopic stomata as in Fig. 8 will
significantly change evaporation and thus the hydrologic cycle. During
the years of the Voluntown experiment these tasks were accomplished
and have been partially reported; the model is reported briefly here as
a guide to the experimental results, and then the results are fully reported.

The essence of the model or simulator of energy exchange (Waggoner
and Reifsnyder, 1968; Waggoner et al., 1969b) is shown in Fig. 38. A
forest is divided into (n-1) strata of foliage and a lowest stratum of
stems and soil surface. In the Figure, n is 4. Each stratum absorbs a
net S, calories per second of radiation per cm2 of land. The subscript i
identifies the stratum and may be 1, 2, . . ., n. At steady state the foliage
attains a temperature of 6X that causes a loss of hi sensible heat and
Vi potential heat or evaporation that balances the gain Si. At the soil
surface, the current of energy G into the soil helps balance the budget.
The loss of sensible heat hi is determined by the 6„ the temperatures
Ta at the top of the canopy and Ts near the soil, the boundary layer

68

Connecticut Experiment Station

Bulletin 726

■ ■-'.•.' *<

is

*

G

e.^

Fig. 38. The conception of energy exchange in a forest (see text).

resistances r„ and the resistances Rj of the bulk air. The evaporation Vi
is determined by the B\ (under the assumption that the vapor pressure
is saturated beneath the stomata), the humidities VPD and cs at the
top of the canopy and near the soil, the Ri, and the qi. The q, is the sum
of X\ and stomatal resistance. Since all differences in temperature or
humidity are proportional to the products of resistances and either hi or
Vi, linear equations can be written. These equations plus the balance
of energy in each stratum comprise 3n equations that can be solved for
n evaporation currents as well as n currents of sensible heat, (n— 1) foli-
age temperatures, and G.

A modification of the algebra permits specifying G, v„ and h,„ the
flux of heat into the soil and the fluxes of evaporation and sensible heat
from the soil, instead of Ts and e*. Since conceiving the currents G, v„
and h„ for a hypothetical day is easier than conceiving Ts and es, this
modification is employed here.

Evaporation is now calculated for a forest with the following char-
acteristics. The temperature, humidity, net radiation, wind and dif-
fusivity at canopy top are tabulated in Table 20 for a hypothetical but
realistic clear day. The flux of heat into the litter-covered soil was taken
from Schubert's observations in a pine woodland ( Baumgartner, 1965).

Transpiration and Its Control by Stomata in a Pine Forest 69

Table 20. Weather at the top of the simulated pine forest and heating of the soil:
a hypothetical clear day.

Vapor

Diffu-

Temper-

pressure

Net

sivity

Heat into

ature,

deficit,

radiation,

Wind,

(X1000),

soil,

Hour

C

mm Hg

mly sec-1

cm sec-1

cm2 sec-1

mly sec-1

0

20

2

-1

100

2

0.4

2

20

0

-1

100

2

0.4

4

19

0

-1

100

2

0.3

6

19

4

-1

100

2

0

8

23

8

11

150

10

0.6

10

26

12

15

200

16

1.0

12

27

14

16

300

20

1.0

14

27

14

15

300

20

0.8

16

25

11

11

300

16

0.4

18

22

4

-1

200

4

0.3

20

20

2

-2

100

2

0.5

22

18

2

-2

100

2

0.5

Evaporation and condensation at the soil surface were assumed to be
negligible. A cloudy day was simulated by quartering radiation, flux of
heat into the soil and daytime warming of the air, and by halving the
vapor pressure deficits of Table 20.

In June 1969 the trees were 1540 cm tall, the live canopy 810 cm tall
and the needles 1.4 mm wide. The projected leaf area was 2.6 cm2 cm-2
and the leaf surface area was 12.4 cm2 cm-2 (Chapter 3). To simulate
defoliation by treatment in 1968, the outcome of a loss of a third of the
foliage was also calculated. The foliage was distributed vertically as the
Gaussian curve (Stephens, 1969).

The wind and diffusivity were extinguished exponentially through the
canopy as a function of relative height and an extinction coefficient
(Uchijima, 1962). In other canopies the coefficient has ranged from 2.5
in clover to 4.2 in 10-year-old pine (Brown and Covey, 1966). Because
the open space among the stems at Voluntown was great and foliage
area was small, a coefficient of 3, i.e. a decrease to 5 per cent through
the canopy, was assumed. Further, the diffusivity throughout the stem
stratum was set equal to that in the lowest canopy stratum, and a mini-
mum of 244 cm2 sec-1 was established.

The illumination was extinguished exponentially with the projected
foliage area (Saeki, 1963) and an extinction coefficient of 0.6, which
causes only slightly more absorption of illumination than our observation
(Chapter 3) of 70 per cent absorption of illumination in a leaf area of
2.6. A somewhat smaller coefficient, 0.5, was employed for the extinction
of net radiation.

Since stomatal resistance rs varies with illumination, it was calcu-
lated as

rB = rin / erf (I/C) (5)

70

Connecticut Experiment Station
I5|

Bulletin 726

5h
3

0 2 4 6 8 10

RELATIVE SCALE

Fig. 39. The midday distribution, in relative terms, of slomatal resistance
(rB), diffusivity (K) and foliage through the simulated pine.

where r,„ is a minimum stomatal resistance of 15 sec cm-1 (Figs. 6 and
10), I is insolation absorbed per projected foliage area, C is 0.005, and
erf is the error function. The insolation at canopy top was taken as 1.54
times net radiation. The C of 0.005 fits the observations of stomatal re-
sistance and insolation, Fig. 10A. The rm was taken as 30 in treated
foliage, Fig. 8A; an r,„ of 7.5 represented newly formed foliage, Fig. 7;
and an rm of zero represented wet foliage. Since the stomatal resistances
above pertain to each square centimeter of leaf surface (Equation 4),
the foliar surface area, and not projected area, per stratum was used
to obtain the ri of Fig. 38.

The distributions or profiles of the foliage, diffusivity and stomatal
resistance arc exemplified in Fig. 39. The profile of stomatal resistance
is similar to those observed by Turner (1969), and its unusual shape
arises from the non-linearity between resistance and light (Equation 5);
the resistance profile in Fig. 39 is for midday.

The outcome of different stomatal resistances at 1200 hr on a clear
day is shown by the partitioning of energy between evaporation and
sensible heal, Fig. 40, and by the accompanying temperatures ol loliage
and air, Fig. 41. In Fig. 40 seven pairs of horizontal bars show the
exchange of energy in each of the six canopy strata and at the soil. The
upper one of each pair shows the outcome when the minimum stomatal

Transpiration and Its Control by Stomata in a Pine Forest 71

15

CH

^

^m^m^

:: yzmZZa.

. v .T

(-" 9

x

UJ

x 7

E

15
30

mw/WW/ZTZTZl

'A I

V/////A

rzr

teg^^

0 12 3 4 5

MLY SEC"1

Fig. 40. The midday partition of the gain of radiation (S) into sensible

heat (h), evaporation (v) and the ground (G) when the minimum stomatal

resistance (rm) is 15 or 30 sec cm-1.

resistance is 15 sec cm-1, and the lower shows the outcome when it is
30 sec cm-1.

The net income or gain of radiation is indicated by the lengths of the
bars at each level or stratum. The sum of the six Si in the canopy is 11.5
mly sec-1, leaving 4.5 to be absorbed by the soil. The 4.5 is divided
between 1.0 into the ground and 3.5 sensible heat into the air. The
sensible heat h is indicated in the Figure by hatching and the ground
flux G by shading. This division at the soil is the same for both stomatal
conditions because ground flux was specified and S was calculated
from the constant foliar area.

Within the canopy, however, stomatal conditions matter. Essentially,
increasing stomatal resistance decreases evaporation v and increases the
loss of sensible heat h, and in Fig. 40 this is shown by an increase
in the hatched bars for an r,„ of 30 compared to 15 sec cm-1.

The greater loss of sensible heat when stomatal resistance increases,
of course, is accompanied by a warming of foliage relative to air and a
warming of the air within the canopy relative to the air above. These
changes are depicted for midday in Fig. 41. Relative to air above, the
air within the canopy is 0.26C warmer in the highest and 3.90C warmer
in the lowest stratum when r„, is 15 sec cm-1. Doubling rm warms the

72

Connecticut Experiment Station

Bulletin 726

x

CD

7

0 I 2

TEMPERATURE

3 4 5

DIFFERENCE, C

Fig. 41. Midday air temperature within canopies having an rm of 15 and
30 sec cm-1. The line at the left of each shaded portion is the air tempera-
ture, relative to the air temperature at the canopy top. The width of the
shaded areas is proportional (2x) to the warming of the foliage above
local air temperature and, at a given height, is proportional to the loss
of sensible heat (h).

air 0.09C in the highest and 0.79C in the lowest stratum. At the same
time, the difference between foliage and local air temperature increases
from a maximum of 0.13 to 0.17C in the second stratum, and this is de-
picted by a widening of the shaded areas in Fig. 41.

If stomatal resistance is halved, as in new foliage, the outcome is, of
course, reversed. Evaporation is increased by 30 to 100 per cent in the
several strata. In the bottom half of the canopy, this rapid evaporation
consumes more energy than radiation provides, the foliage becomes
cooler than the nearby air, and the needed energy is gained as sensible
heat. If the foliage were wet, as after a rain or dew, the foliage would
become still cooler, still more energy would be gained as sensible heat,
and evaporation from the whole canopy would exceed the receipt of
net radiation.

Clearly, midday, which was considered until now, is the time of greatest
effect of stomatal change. To show the outcome of changed weather
or foliar area or stomata upon soil moisture, however, we must add the
evaporation V| over height and over the day. The effect of halving or
doubling minimum stomatal resistance rm in the full canopy of 2.6 cm2
cm-2 is shown in the course of evaporation during a clear 24 hours,
Fig. 42. During daylight the differences among minimum stomatal re-
sistances affect evaporation. At night, however, real and simulated, treat-
ed and untreated, old and young stomata all close; then the differences
among minimum stomatal resistances do not affect evaporation.

At this point, we interject a comment about the outcome of zero
stomatal resistance because one wants to know how rapidly dew or in-

Transpiration and Its Control by Stomata in a Pine Forest 73

12 r

u

LxJ
If)

>■

<

or
O
a.
<

>

UJ

Fig. 42. The diurnal course of simulated evaporation from a pine planta-
tion with minimum stomatal resistances of 7.5, 15 and 30 sec cm.-1.

tercepted rain would evaporate. This evaporation of water on the foliage
ranges from 1 mm hr_1 (100 mly sec-1) at 0800 hours on a cloudy day
to 4.8 on a clear midday, providing an ample route for the prompt de-
parture of any dew or intercepted rainfall and explaining how the
evaporation of intercepted precipitation may be a third of the evapora-
tion from a forest ( Baumgartner, 1967).

It is now time to summarize the various evaporations by setting out
in Table 21 the total quantities lost in 24 hours. A canopy of young
foliage on a clear day would transpire 5.8 mm of water, consuming about
85 per cent of the net radiation. Any evaporation from the soil surface
would be added to this as would much of the evaporation of any inter-
cepted water. Cloudy weather, as defined here, would decrease the
daily evaporation to about a fourth, which is about the same as the
specified change in net radiation. Doubling stomatal resistance would
decrease evaporation about 45 per cent, whether one begins with an
rm of 7.5 or 15 sec cm-1 and whether the weather is clear or cloudy.

Table 21. Simulated daily evaporation (

mm).

Projected Leaf Area
Minimum stomatal re-
sistance, sec cm-1

7.5

2.6
15

30

7.5

1.7
15

30

Clear daya
Cloudy day"

5.8
1.2

3.3
0.7

1.8
0.4

4.6
1.0

2.6
0.5

1.4
0.3

a As specified in Table 20. Net radiation was equivalent to 7 mm evaporation.
b One quarter the net radiation and soil heat flux and one half the vapor pressure
deficit of the clear day. Otherwise the same as the clear day.

74 Connecticut Experiment Station Bulletin 726

The final calculation concerns the effect of the one-third decrease in
foliage that was apparent by late 1968 and caused by repeated spraying.
According to Table 21 this defoliation would decrease daily evaporation
about a fifth under both cloudy and clear conditions.

We now have a logical framework and predictions for the effects of
several factors upon evaporation and can analyze the changes in evapora-
tion that will be derived from soil and rain observations.

Evaporation from the forest was calculated as the sum of the loss in
water in the top 330 cm of soil plus the rainfall. These two parameters
are tabulated in Tables 22-24. The estimate of evaporation will be high
if considerable water leaches below 330 cm and will be affected by the
variability in soil water content and rainfall. The error due to leaching
is unknown. The variability caused by soil water content differed from
one period to another, but an example is provided by the change from
1 to 7 June 1967: there was a 44.1 mm change in the upper six strata
with a standard error of 1.3 mm, and 6.0±2.9 change in the lower five
strata. The variability due to rain was reduced but not eliminated by
using rainfall observations made at the Pachaug State Forest Nursery
3 km away rather than those made at the Pachaug State Forest Head-
quarters 5 km away (Table 2).

During the four years of observation, loss could be estimated from
28 pairs of observations representing 28 periods of 6 to 70 days, and
the mean rates for each period are shown in Fig. 43. They range from
7 mm day-1 during eight bright days in June to about 2.5 mm day-1
in the autumn or during eight cloudy days in August 1967. The smooth
curve in the Figure is the best-fitting, two-term polynomial. It is a rea-
sonable estimate of the climatic normal, since it was calculated from
four years of observations. The mean evaporation from the pan at Kings-
ton (Table 3) is also shown on the Figure, and it is essentially the same
as the estimated mean evaporation from the forest.

Next, the causes of the variations about the mean curve were explored.
The evaporation from the pan is an integration of factors that differ
between clear and cloudy days, and hence the evaporation from the
forest per day during a period was related to the daily loss from the
pan during the same period. The method of presenting the relation be-
tween loss of water and evaporation can greatly affect the appearance
of the correlation. Sometimes evaporation from a field is accumulated
over a season and compared in a graph to an estimate of the evaporation,
in this ease the accumulated evaporation from a pan. Since both values
inevitably increase through the season, one sees an illusory correlation
of observation and estimate. A more critical test is to compare the
evaporation per day from the plantation and pan. This method exposes
tin- difficulties in a glaring light, but it is nevertheless the test used in
Fig. 44.

The few observations ol 1966 and the more numerous ones of 1967
show the significant effect of the evaporation from the Kingston pan

Transpiration and Its Control by Stomata in a Pine Forest 75

MAY

JUNE JULY AUG SEPT OCT

Fig. 43. The course of water lost from untreated trees at the first plantation
during the four seasons ( % ) and the mean pan evaporation at Kingston ( □ ) .

upon the loss of water from the forest. One adjustment was necessary
before the correlation coefficients of 0.73 in 1966 and 0.55 in 1967 were
obtained: the 5 July 1967 observation was made during a 3-day rain,
and hence, the evaporation was better estimated for the 2-week period
including 5 July than for the week before and after 5 July. The corre-
lations show the sort of difference between relatively clear and cloudy
periods that we expect, but the correlation is not close.

A comparison of the insolation at the Pachaug State Forest Nursery,
3 km from the plantation, and pan evaporation at Kingston, 30 km from
the plantation, shows that occasionally cloudiness or persistent fog at
Pachaug reduced the daily insolation, while clear skies at Kingston
caused rapid evaporation, and vice-versa. Since evaporation is logically
related to net radiation (Penman, 1956), which varies with and is about
0.8 of insolation (Shaw, 1956), and since this is also true when other
weather factors vary accordingly (Table 21), we tested whether the
water loss from the plantation was more closely related to insolation at
the nearby Nursery than to the pan evaporation at further Kingston. The
relation between insolation, converted into equivalent evaporation by
dividing by 60 cal mm-1, is shown in Fig. 45. The closer correlation
(r = 0.73) between insolation upon the Nursery and water loss in 1967
than the correlation ( r = 0.55 ) between Kingston evaporation and water
loss ( Fig. 44B ) indicates that some deviation was caused by the distance

76

Connecticut Experiment Station

Bulletin 726

1

• \

i i

•

-

•
• •

•
•

-

~

•

•
•

"

•

Is-

CD

00

i

i

i

i

i

v -

1 1 \

Is-

<x» in ^ ro eo

i-AVQ IAIIAJ 'SS01 d31\7M

<

O

- ro .— '«

O

1 \

•

i

1

1

i 1

-

•

-

•

•

-

\. •

-

CD

- CD

CT)

<

i

i

<

a

cr

o

O

a

a_

u

o

CM

%

«

Ld

>

2*

3^

<
Q.

cd „

C CD

o
>

O CD

2 i— '

c „

>S<
Pi

to 0

•is £

m

J8-S

—

E *■

V

O w

<

* a

<t

u

— -3

<
or
O
a

>

— CL

I }

I I

to

Transpiration and Its Control by Stomata in a Pine Forest 77

2 3 4

INSOLATION.

MM DAY'

Fig. 45. 1967 loss of water from the first plantation related to insolation
upon the Pachaug State Forest Nursery.

between the plantation and observatory and that water loss is indeed
closely related to evaporative demand. Thus, if one accepts that the loss
of water from the forest should logically vary with either the evaporation
from an open pan or the radiant energy, then the correlations of the
three Figures are recommendations for our system of estimating the
evaporation from the forest.

The equality of evaporation from the forest and either pan evaporation
or insolation, however, is another matter because Table 21 shows that
theoretically evaporation would be less than net radiation (which was
7 mm day-1 in the example). Three causes for the rapid rates estimated
for the forest can be suggested. First, slow leaching beyond the deepest
soil observation and slow evaporation from the soil both undoubtedly
occur and would have to be added to the estimates of Table 21. Second,
the forest reflects little radiation and the net radiation receipt is likely
a larger fraction of the insolation than the ratio of 0.8 observed over
grass by Shaw (1956). Third, rain fell during many of the periods and
the intercepted rain plus any dew distilled from the soil depart rapidly
as we have calculated earlier and also observed (Waggoner et at, 1969a).

78 Connecticut Experiment Station Bulletin 726

Thus our estimates of evaporation arc rapid partly because of the
downward escape of water, but they must also be taken as confirmation
of the rapid evaporation from conifers already observed at Castricum
(Penman, 1967) and in England (Ruttcr and'Fourt, 1965).

The savings in evaporation caused by treatment, Tables 22-24, were
calculated by subtracting the change in soil water beneath treated trees
from the change beneath untreated trees. The greater changes, savings
and number of samples were in the upper six strata, and hence their
means and standard errors are tabulated separately from the lower five
strata. These relative estimates are more certain than the estimates of
absolute evaporation rates. That is, runoff, leaching of water below
330 cm and variation in rainfall between observatory and forest may
affect the estimate of evaporation at one site. But when we subtract
the change in soil water at one site from that at an adjacent one, the
average runoff, leaching and rainfall for the forest will have no effect.

Thus the index of water conservation through stomatal change, which
is the change in soil water beneath untreated trees versus that beneath
treated trees, can be depicted, as in Fig. 46, with confidence that we
are not being misled. The Figure shows a blank for instances when the
change in soil water at one of the eleven depths of observation was no
less beneath treated than untreated trees. Conversely, when treatment
decreased the loss, the appropriate space in the Figure is printed over
more as the saving is greater.

In the first panel of the Figure, which is devoted to 1966, one can
see the early savings in the two upper strata. As the upper soil dried
(Fig. 37) the early, upper saving was exhausted and a saving appeared
in the third, fourth and fifth strata. In 1967 and 1968 a second spray
was applied, and the saving in the upper five strata grew regularly or
was maintained. In 1969 with no spray but only two-thirds of the foliage
remaining, a lesser saving appeared gradually. Small savings also ap-
peared in lower strata in all years, but it is best to examine the variability
of the measurements before accepting Fig. 46 as convincing evidence
of a savings.

The standard error of the savings as well as the size of the savings
relative to evaporation are given in Tables 22, 23, and 24. Clear evidence
of a saving in evaporation and thus soil water promptly followed the
first shrinkage of stomata on 3 June 1966, Table 22. In the upper six
strata, where observations were numerous, the 10 mm saving was several
times as large as its standard error of 2 mm. The small saving in the
lower strata was insignificant. The sum of the savings, 11 mm, was 18
per cent of the total evaporation from the forest during the 2 weeks.
The saving in the upper strata during the next 3 weeks was again sig-
nificant. When added to an insignificant loss in the lower soil, however,
the total saving was only 6 per cent of evaporation. In 1966 a second
spray was not applied, new foliage overshadowed the old during July,
and no further significant savings were made, but neither were the
initial savings lost. Consequently, the total saving from 3 June to 4 Oc-

Transpiration and Its Control by Stomata in a Pine Forest 79

MAY

JUNE JULY

AUG

SEPT

OCT

1966

4-

7-

10-

1967

4-
7
10

1968

.illlMHlKIHOHHKHHKIHII

-7

-10

1969

iiiliiliilligiKtiiiiil

-4
-7

-10

MAY

JUNE

JULY

AUG

SEPT

OCT

Fig. 46. The saving of water in 11 soil strata following treatment of the

foliage to shrink stomata. First plantation. The arrows above the panels

depicting the 4 years denote the times of spraying.

tober was significant in the upper strata, and the saving in all strata
was 3 per cent of the total evaporation of the season. The failure to
save more water later in the summer corresponds to the time when dif-
ferences in dye infiltration through the stomata no longer differed be-
tween treated and untreated foliage (Chapter 4) and corresponds to
the return of the diurnal shrinkage of the boles of the treated trees to
the normal, greater shrinkage of the untreated ones (Fig. 27). It also
confirmed that the earlier saving was caused by a transitory cause, the
stomatal shrinkage, and not by a permanent change, such as discolora-
tion of needles or defoliation.

In 1967 two sprays were applied, and significant savings of 13 to 25
per cent in evaporation followed, Table 23. Consequently the saving

80

Connecticut Experiment Station

Bulletin 726

Table 22. The water budget and saving of water by stomatal shrinkage in 1966.
Standard error shown for saving.

Loss

Saving

fall,

soil,

Upper,"

Lower,1

Relative,"

Period

mm

mm

mm

mm

%

Jun

3-

- Jun

16

37

28

10±2

1±5

18

Tun

17-

-Jul

8

22

64

10±4

-5±4

6

Jul

9-

-Aug

11

64

138

-7±5

1±5

Aug

12-

- Sep

2

30

62

3±3

-3±3

Sep

3-

-Oct

4

152

-49

5±4

0±8

Sum

20±7

-6±12

3

Overwinter0

-24±8

0±15

a Upper soil is 0 to 195 cm and lower is 195 to 330 cm.

b Relative loss is the saving in upper and lower soil divided by the sum of rainfall

and loss from soil beneath untreated trees.
c 4 Oct 1966 to 1 Jun 1967.

Table 23. The water budget and saving of water by stomatal shrinkage in 1967.
Standard error shown for saving.

Period

Loss

Saving

fall,

soil,

Upper,"

Lower,"

Relative,1"

mm

mm

mm

mm

%

0

38

5±2

2±3

18

3

50

9±1

-2±3

13

60

-12

6±2

-1±4

10

47

-25

1±1

-2±1

6

32

0±2

3±2

8

23

2

-3±1

-3±2

9

17

5±1

1±2

25

39

2

2±1

1±1

8

34

3

1±1

-1±1

8

15

0±1

2±1

13

1

28

-1±1

-1±1

19

57

0±1

0±2

4

15

0±2

0-+-2

135

-30

8±3

-1±4

6

34±6

— 2±9

5

-26±5

10±10

Jun 1-
Jun 8-
Jun 15-
Jun 29-
Jul 6-
Jul 13 •
Jul 20 -
Jul 27-
Aug 3-
Aug 11-
Aug 17-
Aug 24-
Sep 7
Sep 13

Jun 7
Jun 14
Jun 28
Jul 5
Jul 12
Jul 19
Jul 26
Aug 2
Aug 10
Aug 16
Aug 23
Sep 6
Sep 12
Oct 25

Sum

Overwinter'

a Upper soil is 0 to 195 cm and lower is 195 to 330 cm.

b Relative loss is the saving in upper and lower soil divided by the sum of rainfall

and loss from soil beneath untreated trees.
' 26 Oct 1967 to 30 Apr 1968.

Transpiration and Its Control by Stomata in a Pine Forest 81

Table 24. The water budget and saving of water by stomatal shrinkage in 1968
and 1969. Standard error shown for saving.

Period

Rain-
fall,
mm

Loss
from
soil,
mm

Saving

Upper,
mm

Lower,
mm

Relative,"

1968

May 1 - Jun 4

Jun 5 - Jun 20

Jun 21 -Jul 9

Jul 10 - Jul 15

Jul 16 - Aug 8

Aug 9 -Oct 17

Sum
Overwinter0

1969

May 6 - Jun 4

Jun 5 -Jul 11

Jul 12 -Aug 26

Aug 27 -Sep 24

Sum

118

45

13±3

-16±9

32

12

14±2

6±3

46

47

41

4±2

1±2

6

0

32

2±1

0±1

6

46

64

7±4

-1±4

5

124

125

19±5

3±11

9

46±8

9±11

10

-55±8

4±11

89

34

3±3

-7±5

29

171

6±4

-1±5

2

187

-6

14±4

1±10

8

96

26

5±3

18±11

16

29 ±6

13+12

6

a Upper soil is 0 to 195 cm and lower is 195 to 330 cm.

b Relative loss is the saving in upper and lower soil divided by the sum of rainfall

and loss from soil beneath untreated trees.
c 18 Oct 1968 to 5 May 1969.

between 1 June and 25 October was fully 32 mm or 5 per cent of the
total evaporation, although no significant saving was made in the weeks
well past the spraying, showing that defoliation did not cause the earlier
saving.

Table 24 shows the outcome of the fewer observations of 1968 and
1969. The first spray of 1968 was followed by a significant 46 per cent
decrease in evaporation, the second by a smaller saving. By the end of
1968, however, defoliation was growing substantial, and it must be
considered.

An index of defoliation is provided by a summary of Table 5 and
Fig. 34. The number of fascicles per cm is multiplied by the length of
internode for each year's growth and added over the years. The ratios
of these indices for treated versus untreated trees and for the four times
of observation were: June 1968, 0.83; September 1968, 0.72; November
1968, 0.50; November 1969, 0.63. Thus, the saving in early 1968 is un-
doubtedly caused largely by stomatal shrinkage, but the continued saving
in late 1968, even in the autumn, is undoubtedly caused by defoliation.

Since no sprays were applied in 1969, it provides us with a clear test
of the effect of defoliation. The index of the preceding paragraph shows
that the treated trees began the season with about half the foliage of

82 Connecticut Experiment Station Bulletin 726

the untreated trees and ended the season with about two thirds. The
consequence was a 6 per cent decrease in evaporation for the season.
The saving in the upper soil was clearly significant. Since we calculated
that a defoliation of one third would theoretically save one fifth of the
evaporation, it is easy to ascribe the 6 per cent saving at the end of the
experiment to defoliation.

The seemingly countless observations of soil water saving can be
quickly summarized. Stomatal shrinkage, which the theoretical calcula-
tions predicted would decrease evaporation, in fact decreased evaporation
by 14 mm in 1966 and 32 mm in 1967. The decreases rose near the
theoretical savings rarely and only briefly after spraying, but they were
statistically significant and had not disappeared by the end of the sum-
mer. Therefore, less rain was needed overwinter to return the soil beneath
treated trees to field capacity, allowing 24 mm in 1966-67 and 16 mm
in 1967-68 to percolate into streams and reservoirs.

The dramatic 55 mm saving of 1968 was caused by defoliation as well
as stomatal shrinkage, but even then shrinkage played a role because
the saving was only 42 mm in 1969 when defoliation was greater. The
saving by defoliation was never observed to be more than 6 per cent.

The most important conclusion remains: by shrinking the microscopic
stomata only enough to double stomatal resistance and this for only part
of the season, evaporation is logically and actually substantially de-
creased. Thus, subtle changes in vegetation, as well as weather, can
clearly have profound effects upon the hydrologic cycle.

Chapter 10
TRANSPIRATION AND ITS CONTROL-A DISCUSSION

The interconnections among many of the characters that we have
observed can be depicted as in Fig. 47. Simplifying a diagram by H.
Harssema and F. W. T. Penning de Vries and employing the symbols
of Forrester ( 1969 ) , we indicate a rate by a valve, a level by a rectangle,
a controlled flow by an arrow, and information by a dotted line. Rain
replenishes the level of soil water. This supplies the root and stem,
which in turn hydrates the foliage. Finally, foliage loses water to the
atmosphere at a rate influenced by stomata. The dashed lines for "infor-
mation" merely indicate the effect, e.g. of light upon stomata, as wc
shall next explain in detail.

Our data from the forests demonstrate several of these connections.
Let us begin at the top of the Figure. The atmosphere, either through
dryness or through leaf-warming radiation, increases evaporation ( Figs.
44 and 45). Stomatal closure, however, decreases evaporation (Fig. 46).
Stomatal closure can be caused by PMA (Fig. 8A) or shade (Fig. 6).
Closure can also be caused by a decreased level of water in the foliage

Transpiration and Its Control by Stomata in a Pine Forest 83

Fig. 47. A simplified scheme for the hydraulic system of a forest.

(Fig. 17), and that decreased level can follow the diurnal opening of
stomata and aridity of the day (Fig. 15). On the other hand, stomatal
closure by PMA allows the foliar water level to increase (Fig. 19).

By midday the loss from foliage has exceeded the supply to the roots,
and the level of water in the stem is decreased as shown by contraction
of its diameter (Fig. 23). This contraction, of course, lags behind the
dehydration of the foliage (Fig. 25). A moister, duller day, of course,
causes a smaller shrinkage of the stem than does a drier, brighter day
(Figs. 21, 22 and 23). Similarly, stomata-closing PMA decreases the
diurnal shrinkage of the stem (Fig. 26).

Finally, we reach the soil. The level of water in the soil is decreased
at a rate depending upon the atmosphere (Figs. 44 and 45). Significantly,
the loss is slower when the stomata have been closed by PMA, allowing
an overwinter loss below that is added to groundwater and eventually
a stream (Tables 23, 24 and 25).

Thus the small changes in stomatal resistance do, indeed, have a
discernible effect upon the hydration and water movement throughout
the entire plant and soil. The measured reduction in evaporation of 14
to 32 mm, although clearly discernible, is not large compared with the
total annual evaporation of 600 mm. However, it did cause a similar
increase in the water yield to groundwater, and thus, ultimately to the
reservoirs. Moreover, in Connecticut, the amount of water lost to the
atmosphere by evaporation is about equal to the amount of water that

84 Connecticut Experiment Station Bulletin 726

replenishes the streams and reservoirs. In more arid regions the propor-
tion of water lost by evaporation is greater than that which goes to
groundwater, and an equivalent saving by stomatal closure could cause
a more significant yield to the streams.

The 16 to 24 mm increased yield of water observed in this study is
less than the 100 to 300 mm obtained in the eastern United States by
the more drastic method of removing the forest (Lull and Reinhart,
1967). However, without later cutting and thinning, the increased yield
of water soon disappears (Hewlett and Hibbert, 1961). Thus the rela-
tively small savings by stomatal closure, with its advantage of leaving the
forest intact, may in time prove attractive.

Before stomatal closure can be considered practical, another com-
pound must be found, however. We emphasized at the beginning that
PMA was used to establish the principle that stomatal closure would
change the hydrologic cycle, even in a forest, and that PMA itself was
an impractical compound. The principle has been established, and PMA
remains impractical. Its impracticality derives from two characteristics.
First, mercury is not suitable for widespread use. Second, in hot weather
or after repeated application PMA will damage foliage and reduce
growth. This damage, of course, prevents the advantageous increase in
water use efficiency, which stomatal closure should cause theoretically
and has caused in single leaves (e.g. Zelitch and Waggoner, 1962).

Other stomata-closing compounds have, of course, been discovered
that might avoid these difficulties. Decenylsuccinic acid (Zelitch, 1964)
is an example, but its closure is not consistent or persistent. In Chapter 1
we mentioned that the plant hormone, abscisic acid, has closed stomata.
In an unpublished test, however, we found that pine stomata were closed
3 hours but not 4 days after spraying with abscisic acid.

Thus the search for a non-toxic compound must go on. In the mean-
time, however, the Voluntown experiments have shown that stomatal
closure in a real stand of vegetation can decrease evaporation, prom-
ising benefits of water conservation in the soil and improved hydration
in the tissue if a non-toxic stomata closer is found.

ACKNOWLEDGMENTS

The staff of the Connecticut State Park and Forest Commission at
Pachaug welcomed us into their forests and buildings and helped us
gladly throughout 5 years of the experiments. In 1966 Ben-Ami Bravdo,
now at the Hebrew University, Rehovot, helped mightily to start the
experiments. In 1967 R. C. Ellis made the extensive observations of soil
water and the dendrometcrs. J. E. Begg, C. R. Carlson, D. B. Downs,
D. E. Hill, Mrs. B. Kennedy, and P. F. Tomlinson also made valuable
contributions. This study was supported, in part, by the Melntire-Stennis
Fund.

Transpiration and Its Control by Stomata in a Pine Forest 85

LITERATURE CITED

Baumgartner, A. 1965. The heat, water and carbon dioxide budget of plant cover:
methods and measurements, p. 495-512. In Methodology of Plant Eco-Physi-
ology. Proc. of the Montpelier Symposium. UNESCO, Paris.

Baumgartner, A. 1967. Energetic bases for differential vaporization from forest and
agricultural lands, p. 381-389. In W. E. Sopper & H. W. Lull (eds.), Forest
Hydrology. Pergamon Press, Oxford.

Bormann, F. H. and T. T. Kozlowski. 1962. Measurements of tree ring growth with
dial-gage dendrometers and vernier tree ring bands. Ecology 43:289-294.

Boyer, J. S. 1967. Leaf water potentials measured with a pressure chamber. Plant
Physiol. 42:133-137.

Brown, K. W. and W. Covey. 1966. The energy-budget evaluation of the micro-
meteorological transfer processes within a cornfield. Agr. Meteorol. 3:73-96.

Daubenmire, R. F. 1945. An improved precision dendrometer. Ecology 26:97-98.

Davenport, D. C. 1967. Effects of chemical antitranspirants on transpiration and
growth of grass. J. Exp. Botany 18:332-347.

De Roo, H. C. 1969. Leaf water potentials of sorghum and corn, estimated with
the pressure bomb. Agron. J. 61:969-970.

De Roo, H. C. 1970. Leaf water potentials of tobacco, estimated with the pressure
bomb. Tob. Sci. 14:105-106.

Forrester, J. W. 1969. Urban Dynamics. M.I.T. Press, Cambridge, Mass. p. 13.

Fritts, H. C. and E. C. Fritts. 1955. A new dendrograph for recording radial changes
of a tree. Forest Sci. 1:271-276.

Fry, K. E. and R. B. Walker. 1967. A pressure-infiltration method for estimating
stomatal opening in conifers. Ecology 48:155-157.

Haasis, F. W. 1934. Diametral changes in tree trunks. Carnegie Inst. Wash. Pub.
450. 103 p.

Hewlett, J. D. and A. R. Hibbert. 1961. Increases in water yield after several types
of forest cutting. International Ass. Sci. Hydrol. 6:5-17.

Hewlett, J. D., J. E. Douglass and J. L. Clutter. 1964. Instrumental and soil mois-
ture variance using neutron scattering method. Soil Sci. 97:19-24.

Hill, D. E. 1959. The storage of moisture in Connecticut soils. Conn. Agr. Exp.
Station. Bull. 627. 30 p.

Hunter, A. S. and O. J. Kelley. 1946. A new technique for studying the absorption
of moisture and nutrients from soil by plant roots. Soil Sci. 62:441-450.

Jones, R. J. and T. A. Mansfield. 1970. Suppression of stomatal opening in leaves
treated with abscisic acid. J. Exp. Botany 21:714-719.

Kaufmann, M. R. 1968a. Evaluation of the pressure chamber technique for esti-
mating plant water potential of forest tree species. Forest Sci. 14:369-374.

Kaufmann, M. R. 1968b. Evaluation of the pressure chamber method for measure-
ment of water stress in citrus. Proc. Amer. Soc. Hort. Sci. 93:186-190.

Keller, T. 1966. t)ber den Einfluss von Transpirationshemmenden Chemikalien
( Antitranspirantien ) auf Transpiration, COa— Aufnahme und Wurzelwachstum
von Jungfichten. Forstwiss. Centr. 85:65-79.

Kienholz, R. 1934. Leader, needle, cambial and root growth in certain conifers and
their interrelations. Bot. Gazette 96:73-92.

Klemmer, L. 1969. Die Periodik des Radialzuwachses in einem Fichtenwald und
deren meteorologische Steuerung. Universitat Miinchen, Meteorologisches Insti-
tut, Wissenschaft iche Mitteilungen Nr 17. 85 p.

Kramer, P. J. 1969. Plant and Soil Water Relationships: A Modern Synthesis. Mc-
Graw-Hill, New York. p. 320-322.

Larson, P. R. 1964. Contribution of different-aged needles to growth and wood
formation of young red pines. Forest Sci. 10:224-238.

Loftiield, J. V. G. 1921. The Behaviour of Stomata. Carnegie Inst. Wash. 104 p.

Lull, H. W. and K. G. Reinhart. 1967. Increasing water yield in the Northeast by
management of forested watersheds. U.S. Forest Service Research Paper. NE-66.
45 p.

86 Connecticut Experiment Station Bulletin 726

MacDougal, D. T. 1921. Growth in trees. Carnegie Inst. Wash. Pub. 307. 41 p.
Mittelheuser, C. J. and R. F. M. Van Steveninck. 1969. Stomatal closure and inhibi-
tion of transpiration by RS-abscisic acid. Nature 221:281-282.
Monteith, J. L. and T. A. Bull. 1970. A diffusive resistance porometer for field use.

II. Theory, calibration and performance. J. Appl. Ecol. 7:623-638.
Oppenheimer, H. R. and N. Engelberg. 1965. Mesure du degre d'ouverture des

stomates de coniferes methodes anciennes et modernes, p. 317-323. In Meth-
odology of Plant Eco-Physiology. Proc. of the Montpelier Symposium. UNESCO,

Paris.
Penman, H. L. 1956. Evaporation— an introductory survey. Neth. J. Agr. Sci. 4:9-29.
Penman, H. L. 1967. Evaporation from forests: a comparison of theory and obser-
vation, p. 373-380. In W. E. Sopper and H. W. Lull (eds.), Forest Hydrology.

Pergamon Press, Oxford.
Reineke, L. M. 1932. A precision dendrometer. J. Forestry 30:692-699.
Rutter, A. J. and D. F. Fourt. 1965. Studies in the water relations of Pinus sylvestris

in plantation conditions. III. A comparison of soil water changes and estimates

of total evaporation on four afforested sites and one grass-covered site. J. Appl.

Ecol. 2:197-209.
Saeki, T. 196^. Light relations in plant communities, p. 79-94. In L. T. Evans (ed.),

Environmental control of plant growth. Academic Press, N.Y.
Sampson, J. 1961. A method of replicating dry or moist surfaces for examination by

light microscopy. Nature 191:932-933.
Scholander, P. F., H. T. Hammel, E. D. Bradstreet and E. A. Hemmingsen. 1965.

Sap pressure in vascular plants. Science 148:339-346.
Scholander, P. F., H. T. Hammel, E. A. Hemmingsen and E. D. Bradstreet. 1964.

Hydrostatic pressure and osmotic potential in leaves of mangroves and some

other plants. Proc. Nat. Acad. Sci. U.S. 52:119-125.
Shaw, R. H. 1956. A comparison of solar radiation and net radiation. Bull. Amer.

Meteorol. Soc. 37:205-206.
Shimshi, D. 1963a. Effect of chemical closure of stomata on transpiration in varied

soil and atmospheric environments. Plant Physiol. 38:709-712.
Shimshi, D. 1963b. Effect of soil moisture and phenylmercuric acetate upon sto-
matal aperture, transpiration and photosynthesis. Plant Physiol. 38:713-721.
Slatyer, R. O. and J. F. Bierhuizen. 1964. The effect of several foliar sprays on

transpiration and water use efficiency of cotton plants. Agr. Meteorol. 1:42-53.
Stalfelt, M. G. 1962. The effect of temperature on opening of the stomatal cells.

Physiol. Plant. 15:772-779.
Stephens, G. 1969. Productivity of red pine. 1. Foliage distribution in tree crown

and stand canopy. Agr. Meteorol. 6:275-282.
Tadaki, Y. 1966. Some discussions on the leaf biomass of forest stands and trees.

Government For. Exp. Sta., Tokyo. Bull. 184, p. 135-161.
Thornthwaite, C. W. 1948. An approach toward a rational classification of climate.

Geogr. Rev. 38:55-94.
Turner, N. C. 1969. Stomatal response to illumination in three contrasting canopies.

Crop Sci. 9:303-307.
Turner, N. C. and L. D. Incoll. 1971. The vertical distribution of photosynthesis in

crops of tobacco and sorghum. J. Appl. Ecol. 8: (In press).
Turner, N. C. and J.-Y. Parlange. 1970. Analysis of operation and calibration of a

ventilated diffusion porometer. Plant Physiol. 46:175-177.
Turner, N. C. and P. E. Waggoner. 1968. Effects of changing stomatal width in a

red pine forest on soil water content, leaf water potential, bole diameter, and

growth. Plant Physiol. 43:973-978.
Turner, N. C, H. C. De Roo and W. H. Wright. 1971. A pressure chamber for the

measurement of plant water potential. Conn. Agr. Exp. Sta. Special Soils Bull.

33, 9 p.
Turner, N. C, F. C. C. Pedersen and W. II. Wright. 1969. An aspirated diffusion

porometer for field use. Conn. Agr. Exp. Sta. Special Soils Bull. 29, 12 p.

Transpiration and Its Control by Stomata in a Pine Forest 87

Uchijima, Z. 1962. Studies on the microclimate within plant communities. I. On

the turbulent transfer coefficient within plant layers. J. Agr. Meteorol., Tokyo,

18:1-19.
U.S. Bureau of the Census. 1970. Statistical Abstract of the United States. 91st

edition. U.S. Government Printing Office. Washington, D.C. p. 171.
U.S. Department of Agriculture. 1951. Soil Survey Manual. U.S. Dept. Agr. Hand-
book 18. U.S. Government Printing Office. Washington, D.C. 503 p.
Van Bavel, C. H. M., D. R. Nielson and J. M. Davidson. 1961. Calibration and

characteristics of two neutron moisture probes. Soil Sci. Soc. Amer. Proc,

25:329-334.
Ventura, M. M. 1954. Acao de inibidones enzimaticos sobre a transpiracao e o

comportamento dos estomatos. II. Acao do Arsenito, 2,4-dinitrofenol e verde de

Janus sobre folhas isolades de Stizolobium aterrimum, Pip. o Trac. Rev. Brasil

Biol. 14:153-161.
Wadsworth, D. F. 1960. Chemical control of disease affecting turf on golf greens.

Golf Course Reporter 28(3):26-30.
Waggoner, P. E. 1967. Transpiration of trees and chemicals that close stomata,

p. 483-489. In W. E. Sopper and H. W. Lull (eds.), Forest Hydrology. Per-

gamon Press, Oxford.
Waggoner, P. E. and B. Bravdo. 1967. Stomata and the hydrologic cycle. Proc.

Nat. Acad. Sci. U.S. 57:1096-1102.
Waggoner, P. E. and J. D. Hewlett. 1965. Test of a transpiration inhibitor on a

forested watershed. Water Resources Res. 1:391-396.
Waggoner, P. E. and W. E. Reifsnyder. 1968. Simulation of the temperature, hu-
midity and evaporation profiles in a leaf canopy. J. Appl. Meteorol. 7:400-409.
Waggoner, P. E., J. E. Begg and N. C. Turner. 1969a. Evaporation of dew. Agr.

Meteorol. 6:227-230.
Waggoner, P. E., G. E. Furnival and W. E. Reifsnyder. 1969b. Simulation of the

microclimate in a forest. Forest Sci. 15:37-45.
Waggoner, P. E., J. L. Monteith and G. Szeicz. 1964. Decreasing transpiration of

field plants by chemical closure of stomata. Nature 201:97-98.
Wallihan, E. F. 1964. Modification and use of an electric hygrometer for estimating

relative stomatal aperture. Plant Physiol. 39:86-90.
Waring, R. H. and B. D. Cleary. 1967. Plant moisture stress: evaluation by pressure

bomb. Science 155:1248-1254.
Worrall, J. 1966. A method of correcting dendrometer measures of tree diameter

for variation induced by moisture stress changes. Forest Sci. 12:427-429.
Zelitch, I. 1961. Biochemical control of stomatal opening in leaves. Proc. Nat. Acad.

Sci. U.S. 47:1423-1433.
Zelitch, I. 1964. Reduction of transpiration of leaves through stomatal closure in-
duced by alkenylsuccinic acids. Science 143:692-693.
Zelitch, I. and P. E. Waggoner. 1962. Effect of chemical control of stomata on

transpiration and photosynthesis. Proc. Nat. Acad. Sci. U.S. 48:1101-1108 and

1297-1299.

THE CONNECTICUT

AGRICULTURAL EXPERIMENT STATION

NEW HAVEN, CONNECTICUT 0650*

U~H A. J^J^-^

POSTAGE PAID
United Statts Deportment of Agriculture

V^-1

32

Wsl

w\

ft WM

% 4 9

eft *

1 ^r r.

':■: *v \

WM

■. • ^

University of
Connecticut

Libraries

39153029115146