Planta (1984)162:276-282

P l a n t a 9 Springer-Verlag 1984

Equilibrium freezing of leaf water and extracellular ice formation in Afroalpine ' giant rosette' plants* Erwin Beck 1, Ernst-Detlef Schulze 2, Margot Senser 3 and Renate Scheibe 1 1 Lehrstuhl Pflanzenphysiologie and 2 Lehrstuhl Pflanzen6kologie der Universitiit, Universit/itsstrasse 30, D-8580 Bayreuth, and 3 Botanisches Institut der Universitfit, Menzinger Strasse 67, D-8000 Miinchen 19, Federal Republic of Germany

The water potentials of frozen leaves of Afroalpine plants were measured psychrometrically in the field. Comparison of these potentials with the osmotic potentials of an expressed cellular sap and the water potentials of ice indicated almost ideal freezing behaviour and suggested equilibrium freezing. On the basis of the osmotic potentials of expressed cellular sap, the fractions of frozen cellular water which correspond to the measured water potentials of the frozen leaves could be determined (e.g. 74% at - 3 . 0 ~ C). The freezing points of leaves were found to be in the range between 0 ~ C and - 0 . 5 ~ C, rendering evidence for freezing of almost pure water and thus confirming the conclusions drawn from the water-potential measurements. The leaves proved to be frost resistant down to temperatures between - 5 ~ and - 1 5 ~ C, as depending on the species. They tolerated short supercooling periods which were necessary in order to start ice nucleation. Extracellular ice caps and ice crystals in the intercellular space were observed when cross sections of frozen leaves were investigated microscopically at subfreezing temperatures. Abstract.

Key words: Equilibrium freezing - Freezing tolerance - Leaf (water potential) - Rosette plant (Afroalpine) - Water potential.

Introduction

Overwintering plants of temperate climates seasonally increase and decrease in frost hardiness : shortening of daylength and lowering of the ambient temperatures induce frost resistance whereas dehardening is usually observed when temperature * Dedicated to Professor Dr. Hubert Ziegler on the occasion of his 60th birthday Symbols: T = temperature; ~ = water potential

and photoperiod increase. In the alpine regions of tropical high mountains the seasons are much less pronounced than in the temperate climate and the amplitude of the diurnal temperature oscillation by far exceeds that of the annual changes. As a consequence of the relatively short period of nocturnal cooling, the air temperatures, at least in the plant-covered alpine zone, rarely drop below - 12 ~ C (Beck et al. 1982) but above 4000 m altitude frost may occur almost every night. It is obvious that tropical alpine plants must have developed mechanisms of permanent frost hardiness allowing for reasonable photosynthesis immediately after thawing of the leaves (Schulze et al. 1984). They contain high concentrations of colligatively efficient solutes, such as sucrose (Beck et al. 1982) which are considered to be beneficial with respect to freezing of the cells. Previously Beck et al. (1982) described mechanisms of freezing avoidance of Afroalpine 'giant rosette' plants (belonging to the genera Dendrosenecio and Lobelia) protecting the meristematic leaf bud in the center of the rosette via the formation of a so-called 'night bud' from the outer rosette leaves. However, these adult outer rosette leaves are usually stiffly frozen for a few hours per night. In the present work an attempt was made to obtain insight into the freezing process of these leaves in order to advance understanding of their freezing tolerance. For this purpose the water status of frozen and thawed leaves has been measured in situ with a leaf psychrometer and has been related to their freezing behaviour. To the authors' knowledge this is the first report of water-potential measurements of frozen leaves. Materials

and methods

Measurements of water potentials. The measurements were performed with adult rosette leaves of the Giant Groundsel Dendrosenecio keniodendron (R.E. & Th. C.E. Fries) B. Nord. and

E. Beck et al. : Equilibrium freezing in Afroalpine plants of Lobelia keniensis R.E. & Th. C.E. Fries. The data reported in this communication were collected during an ecological research project performed at Teleki valley (Mt. Kenya, Kenya) during February and March 1983 at an altitude of 4200 m above sea level. Additionally, plants grown in a greenhouse (17~ C day, 10-]4 ~ C night temperature) were used for collecting cellular sap. For measurements of leaf water potentials a Wescor leaf psychrometer (CR 51 Wescor, Inc., Logan, Utah, USA) was used in the psychrometric mode. Routinely a cooling period of 5 s was employed. The sensors were mounted on the upper side of the leaves, where they remained for at least 24 h. In this paper only those measurements will be reported, in which the sensors were shaded from direct irradiation by sunlight and where the offset control was routinely performed (Shackel 1984). The utilization of psychrometers at subzero temperatures has been shown by Van Haveren (1972). The output of the sensor was calibrated with solutions of known osmolality ( - 1 . 0 and -3.0 MPa) at different temperatures between + 20 ~ C and - 8 ~ C in a controlled growth cabinet in the laboratory. In this case, and for determination of the osmotic potential of expressed cellular sap, a droplet of the solution was put onto a disc of filter paper in the psychrometer chamber. For measurements of osmotic potentials greater than - 3 . 0 MPa, the cooling period was increased up to 25 s in order to obtain the typical psychrometer response curves. Simultaneously with the leaf water potentials the leaf temperatures were measured with Siemens (Berlin, West Germany) heat conductors as described elsewhere (Beck et al. 1982).

Preparation of expressed leaf cellular sap. Cellular sap was produced in the laboratory from 7.5 g of leaves ofD. keniodendron which were freshly harvested from greenhouse-grown plants. After removal of the large midribs the leaf material was ground in liquid nitrogen. The frozen powder was transferred into 4-ml tubes, thawed and centrifuged at 150000 g (Beckman ultracentrifuge; Mfinchen, FRG) for I h. Part of the supernatant cellular sap ~ was directly used for determination of the osmotic potential and freezing point while another part was concentrated by lyophilization in order to establish a calibration curve of the dependence of the osmotic potential on the concentration of the expressed cellular sap.

Determination of freezing-point depression and frost hardiness of leaves. A Siemens heat conductor was mounted on the surface of 0.8-cm 2 pieces of freshly harvested leaves which were then embedded in styrofoam. The device was sealed in a plastic bag and placed in ice-salt mixtures of various concentrations. The leaf temperature was monitored every ] 0 s. Frost hardiness of the leaves was determined with excised leaf pieces. The leaf pieces were exposed at the final temperature for 4 h and after slow rewarming (in ice) were kept in moist Petri dishes and inspected during another 4 d for necroses. A limit temperature of - 5 ~ C in this test indicates that less than 50% of the samples had developed necroses, whereas at the next lower temperature ( - 8 ~ C) more than 50% of the samples showed frost damage.

Conductivity measurements. Conductivity of expressed cellular sap was measured with a device consisting of a 9-V battery, a milliamperemeter and two platinum electrodes which, apart from their lower ends, were sealed in a glass rod of l cm diameter. A calibration curve was established with NaC1 solutions of various concentrations. 1 Since the cell-wall volumina accounted for not more than 6.5% of a turgescent leaf cell and because the major leaf vein had been removed, the expressed sap should approximately represent the cellular sap

277

Results and discussion 1. The water status of the frozen leaves Theory. It is generally a c c e p t e d t h a t i n t r a c e l l u l a r f o r m a t i o n o f ice crystals is i n s t a n t l y fatal (Levitt 1978). T h u s freezing o f w a t e r in a f r o s t - t o l e r a n t leaf is a s s u m e d to o c c u r o u t s i d e the p r o t o p l a s t s . At subfreezing temperatures a physico-chemical e q u i l i b r i u m b e t w e e n extracellular ice a n d the cell as a n o s m o t i c s y s t e m is s u g g e s t e d ( e q u i l i b r i u m freezing; Olien 1978) w h i c h is g o v e r n e d b y the w a t e r p o t e n t i a l o f ice. T h e latter decreases (i.e. bec o m e s m o r e negative) w i t h d e c r e a s i n g t e m p e r a tures. T h e r e f o r e , u n d e r c o n d i t i o n s o f e q u i l i b r i u m freezing, a d e c r e a s e o f the s u b z e r o t e m p e r a t u r e results in a d e c r e a s e o f the w a t e r p o t e n t i a l o f the cell, i.e. in a d e h y d r a t i o n o f the p r o t o p l a s t a n d a c o n c e n t r a t i o n o f the cellular solutions. A l t h o u g h d e h y d r a t i o n o f leaf cells w i t h dec r e a s i n g s u b z e r o t e m p e r a t u r e s has b e e n r e c o r d e d f o r leaves o f v a r i o u s species (see e.g. R a j a s h e k a r a n d B u r k e 1982), e q u i l i b r i u m freezing has n o t une q u i v o c a l l y b e e n d e m o n s t r a t e d b e c a u s e the w a t e r p o t e n t i a l o r the c o r r e s p o n d i n g w a t e r - v a p o r pressure o f f r o z e n leaves c o u l d n o t be m e a s u r e d . T h e p r e s e n t p a p e r r e p o r t s o n leaf w a t e r p o t e n tials o f f r o z e n leaves. P r o v i d e d t h a t e q u i l i b r i u m freezing h a d o c c u r r e d in these p l a n t s , it is e x p e c t e d t h a t the w a t e r p o t e n t i a l (}[]leaf (T)) at a given t e m p e r a t u r e (7) equals the w a t e r p o t e n t i a l o f ice (~ice(r)) a n d , b e c a u s e o f the loss o f t u r g o r d u e to d e h y d r a t i o n o f the cell ( ~ p - - , 0 ) , s h o u l d also be similar to the o s m o t i c p o t e n t i a l o f the cells ( ~ ( r ~ ) - T h u s , the w a t e r r e l a t i o n s at e q u i l i b r i u m freezing m a y be described by Eqn. 1 : ~LI'tice(T)= ~leaf(T) ~' ~ ( r ) "

(gqn. 1)

T h e v a l i d i t y o f E q n . ] c a n be e x a m i n e d b y c o m p a r i s o n o f the w a t e r p o t e n t i a l s ( o s m o t i c p o t e n t i a l s ) o f a n expressed cellular sap with t h o s e o f the f r o z e n leaves a n d o f ice, respectively, at different s u b z e r o t e m p e r a t u r e s . I f E q n . 1 is valid a n d if the d e p e n d e n c e o f the o s m o t i c p o t e n t i a l o n the c o n c e n t r a t i o n o f the e x p r e s s e d cellular sap is k n o w n , the f r a c t i o n o f f r o z e n l e a f w a t e r c a n be estimated. F o r this p u r p o s e the osmolalities o f the cellular solutions, c o r r e s p o n d i n g to the w a t e r p o t e n t i a l s , h a v e to be calculated. A c c o r d i n g to R a j a s h e k a r a n d B u r k e (1982) a h y p e r b o l i c c o r r e l a t i o n b e t w e e n the a m o u n t o f f i o z e n (or liquid) leaf w a t e r a n d t e m p e r a t u r e is c h a r a c t e r i s t i c o f e q u i l i b r i u m freezing. S u c h a f u n c t i o n b e c o m e s linear w h e n f r o z e n (or liquid) leaf w a t e r is p l o t t e d versus I / T i n s t e a d o f versus T.

278

E. Beck et al. : Equilibrium freezing in Afroalpine plants i

i

~

i

J

-10-

-8-

i

i

Table 1. D e h y d r a t i o n o f leaf cells caused by freezing o f cellular water in the outer free space as calculated from the osmotic potentials o f various concentrations o f expressed cellular sap and o f the measured water potentials o f frozen and thawed leaves in situ

i

A

3

B

Species

Ternperature T [~

8-/,-

Water potential (MPa) o f frozen leaf as normalized to + 20 ~ C

Water potential (MPa) o f leaf subsequent to thawing or o f expressed cellular sap as normalized to + 20 ~ C

Percent o f leaf water frozen at T

-- 1.94 --

---0.98

50 -

-

-0.95

-

-- 3.75 --

---0.97

74 -

-

-0.95

-

- 5.61 --

- 1.00

83 -

-2-

0 {~ Fig. 1, W a t e r potentials o f frozen leaves [o] o f D. keniodendron (i, 2) and Lobelia keniensis (3) a n d o f frozen expressed cellular sap from D. keniodendron leaves [o] as c o m p a r e d with that o f ice (A, B) at subfreezing temperatures. Line A was calculated according to Rajashekar and Burke (1982) by Eqn. 2,

~ioe [MPa] = 1.22. T [~

(Eqn. 2)

line B was obtained using Eqn. 3 (Critical tables; W a s h b u r n and West 1928) ~ioe [MPa]=(1.16_+0.04)-T[~

(Eqn. 3)

Results. During field measurements at Mt. Kenya (Kenya), water potentials of Den&osenecio kenioden&on and Lobelia keniensis leaves which were stiffly frozen at dawn could be measured. The water potentials were found to depend on the temperature and to range from - 1 . 8 ( - 1 . 5 ~ C) up to - 5 . 2 ( - 4 . 4 ~ C) MPa (Fig. 1). The water potentials in the frozen state were considerably more negative than those measured with the same leaves during the day which never exceeded - 2 . 0 MPa (Schulze et al. 1984). In order to investigate whether equilibrium freezing can be demonstrated in leaves of nocturnally frozen Afroalpine plants, the water potentials were compared with those of ice. The latter were calculated according to the two equations presented by Rajashekar and Burke (1982). Figure 1 shows good coincidence of the water potentials of the frozen leaves and of ice, thus verifying the first term of Eqn. 1 and indicating equilibrium freezing. To examine the second term of Eqn. 1, the osmotic potentials (as measured psychrometrically) of frozen expressed cellular sap of D. keniodendron leaves were compared with the water potentials of the frozen leaves. Figure 1 demonstrates good agreement and thus indicates that the water poten-

Dendrosenecio keniodendron Leaf Leaf Expressed cellular sap

-- 1.5 +4.3 +20.0

Dendrosenecio keniodendron Leaf Leaf Expressed cellular sap

-- 3.0 +9.5 +20.0

Lobelia kenienst;s Leaf Leaf

- 4.4 + 3.0

tials of the frozen leaves can be interpreted as their osmotic potentials. Since this comparison encompasses data from field-grown plants on the one hand and from a greenhouse plant on the other, the validity of the results has to be demonstrated: it is provided by the finding that the water potential of leaves as measured just after thawing (when Up should still be negligible) was identical with the osmotic potential of unfrozen expressed cellular sap at the same temperature (Table 1). The coincidence of the water potentials of frozen leaves and of the osmotic potentials of frozen expressed cellular sap with the water potentials of ice indicates almost ideal freezing characteristics of the cell solution and of the leaves as well. Ideal freezing behaviour of the expressed cellular sap is also indicated from the agreement between the measured osmotic potential (0.95 MPa) and that calculated from the freezing-point depression (A T , , = - 0 . 7 3 ~ C) by Eqn. 4 which resulted in 0.96 MPa (T= + 20 ~ C).

E. Beck et al. : Equilibrium freezing in Afroalpine plants

A Tm[~ T[~ (Eqn. 4) 223.7 From the freezing-point depression, an average concentration of the cellular sap of 0.39 osmol was calculated. According to conductivity measurements, one third of the solutes must be of ionic nature (K § being the predominant cation) and two thirds uncharged substances, predominantly sucrose (Beck et al. 1982) which is a well-known membrane cryoprotectant (Steponkus et al. 1977). With Lobelia, the water potential of the justthawed leaf was found to be similar to that of Dendrosenecio (Table 1) but the osmotic potential of the expressed cellular sap could not be measured. However, similar relations between charged and uncharged solutes were found in a leafhomogenate as in the cellular sap of Dendrosenecio. Therefore, the same correlation of cell-sap concentration versus osmotic potential was used for both plant species for the determination of the fraction of frozen cellular water. Equilibrium freezing predicts an increase of the portion of frozen leaf water with decreasing subfreezing temperatures which results in an increase of the solute concentrations within the cell. Under this condition, the water potential of the tissue is predictable from that of ice from which in turn the osmolality of the cell solution can be calculated. However, in order to allow for eventual deviations from ideal solutions, the relation between osmotic potential and solute concentration of expressed cellular sap (D. keniodendron) was determined experimentally. The resulting hyperbolic relation, which upon transformation showed good linearity (Fig. 2), again indicated the presence of an almost ideal solution even in the concentrated sap. From this relation, the fractions of frozen and unfrozen water could be determined for each temperature as shown in Table 1. When the portion of unfrozen water (Lr) was plotted versus 1/T, a straight line resulted with an intercept K on the LT-axis of 0.08 g water per g dry weight (Fig. 3). This fraction is equivalent to 1.2% of the total leaf water and according to Eqn. 5 represents that portion of water which cannot be frozen at 0 ~ (Rajashekar and Burke :1982).

279 i

% [MPa] =

L r = (L o - K)" A T m+ K.

r[~

(Eqn. 5)

In Eqn. 3, L o is the amount of liquid water pet g fresh weight at 0 ~ C and T m represents the melting point (freezing-point depression, o C) of the cellular sap. A T~, as calculated from the regression line of Fig. 3 ( - 0 . 7 9 ~ was found to be only

i

i

i

.E 0.831.0-

t

~ #e

\

455

X

o

083 1.o

/

/

S

5

; 157

~ 1.25-

/

/

~- 1.67"5

l

2.5-

O

.

O

-

60

t9

5.010,00 0

1'.0

210

3'0

7

0 i i r 80J 100 i 120 , O0 20 40

410 -q/'~

5'.0

s'0

1

7'0

810

(MPa)

Fig. 2. Correlation between the concentration of an expressed cellular sap of D. keniodendron leaves and its osmotic potential (gJ~) at + 20 ~ C. Various concentrations were prepared by lyophilization and appropriate dilution of the expressed cellular sap

,

o:~

~

i

i

i

i

i

i

K=OZ)SgH20(gDWFI

I-

/ 00

-0h 42 -0'.3 -d4 -05 -o'6 -d7 {oc1-1

Fig. 3. Freezing curve for Afroalpine 'giant rosette' leaves of D. keniodendron (1, 2) and L. keniensis (3). Because of the few data, L r (liquid water content) was plotted vs 1/T. A straight line (regression line with r 2= 0.999) is indicative of a hyperbola in the L T vs T plot. K represents the amount of water which cannot be frozen as crystalline ice

negligibly higher than that measured with the expressed cellular sap ( - 0.73 ~ C). Freezing curves similar to that shown in Fig. 3 indicate almost ideal freezing behaviour and have been found with cereal crowns (Marcellos and Burke 1979; Gusta et al. 1975; Rajashekar and Burke 1982), Solanurn species (Chen et al. 1976), red-osier dogwood stems (Harrison et al. 1978) and onion epidermal tissue (Palta et al. 1977). They are typical for plants which do not contain a deeply supercooled aqueous compartment as may occur in woody or sclero-

280

E. Beck et al. : Equilibrium freezing in Afroalpine plants

Table 2. Freezing-point depressions, supercooling and frost tolerance of four species of Afroalpine 'giant rosette' plants. The water potentials corresponding to the various limit temperatures of frost tolerance were calculated from the osmotic potential of ice at the respective temperatures assuming equilibrium freezing, n.d. = Osmotic potential of expressed cellular sap was not determined Species

Date

Actual freezing-point depression of leaf (~

Supercooling necessary for leaf water to freeze (~

Limit temperature (T) of frost tolerance (~

Water potential gt{r) corresponding to T (MPa)

Percentage of total water frozen at T

(n=m) Dendroseneeio keniodendron

1983 Febr. 24

-0.174-0.1 (n=5)

--7.0

-5

--6.0

84

Dendr osene cio brassiea

1983 March 4

--0.37_+0.3 (n=3)

--5.5

-5

-6.0

n.d.

Lobelia keniensis

1983 March 5

-0.234-0.1 (n=3)

-5.0

-10

-11.0

n.d,

Lobelia telekii

1983 Febr. 22

-0.17•

-3.7

-15

-17.4

n.d,

(n=9)

phyllous species (Anderson et al. 1983; Gusta et al. 1983). It is noteworthy that the cellular solute concentrations, as calculated from the water potentials of frozen leaves of the Afroalpine species studied, cause freezing-point depressions of the cellular solutions which were similar to the measured leaf temperatures. Since freezing of a solution requires at least moderate supercooling, equilibrium freezing prevents freezing of the protoplasts.

2. Freezing point depressions and frost hardiness of leaves The idea of equilibrium freezing of leaves of Afroalpine 'giant rosette' plants could be confirmed by studies of their freezing behaviour. When determined with intact leaves or excised leaf pieces, average freezing-point depressions of four Afroalpine 'giant rosette' plants (Dendrosenecio keniodendron, D. brassica, Lobelia keniensis, L, telekii) have been reported between - 1 . 0 ~ and - 4 . 2 ~ C (Beck et al. 1982). In these experiments, freezing points close to 0 ~ had also sometimes occurred. In spring 1983, with leaves of L. keniensis, L. telekii and D. brassica freezing points between 0 and - 0 . 5 ~ were consistently found whereas those of D. keniodendron leaves ranged from - 0 . 1 ~ C to - 9 ~ C. In the light of the results of the water-potential measurements, the freezing points close to 0 ~ C are regarded to be the correct ones, indicating freezing of pure water or of very diluted solutions, which may result from ions in the extracellular free space (Cosgrove and Cleland

Dendrosenecio keniodendron 25-

February 18/19,1983

2015-

oJ 10-

0

P

B, /

.... -,

18

20

2'2

2'&

i

&

,

g

g

1()

12

1l.

lg

18

Time (h)

Fig. 4. Oendrosenecio keniodendron, Diurnal course of the leaf temperature (solid line) as compared with the air temperature (dashed line). Two freezing exotherms were observed: the first (A) resulted from adhering water droplets whereas the second (B) indicated freezing of the leaf water

1983). Freezing points lower than about - 0 . 5 ~ C must then be interpreted as apparent freezing points resulting from a freezing exotherm which was too small to heat the whole leaf tissue to the actual freezing point of water (Larcher 1984). With excised leaf pieces, freezing exotherms close to 0 ~ C were found after moderate supercooling down to - 5 or - 7 ~ C (Table 2). With intact leaves attached to the plant at the natural stand, similar observations were made (Fig. 4). This level of supercooling did not injure the leaves. However, with those freezing exotherms of D. keniodendron leaf pieces, resulting in large freezing-point depressions, supercooling down to - 1 4 ~ C was observed

E. Becket al. : Equilibrium freezingin Afroalpineplants

Fig. 5A, B. Photomicrographsof cross sections of a frozen leaf of D. keniodendron. A Ice caps (all arrows) on the surface of the cells and in intercellular spaces (0- B Cellular structure of the same area. Bar = 25 gm which was harmful to the leaves. This is further evidence that in those determinations the ' t r u e ' freezing points were not measured. Frost resistance of excised leaf pieces was determined in the field for the four Afroalpine 'giant rosette' species listed above at temperature intervals of --3 ~ C. Both Lobelia species were considerably more frost hardy than the groundsels (Table 2). During extracellular freezing of water, the temperature of initial frost damage ( - 5 ~ C) of D. keniodendron leaves would be equivalent to an osmotic potential of - 6.0 MPa, as calculated from the water-vapor pressure of ice at - 5 ~ C normalized to +20 ~ C. Such a potential is equivalent to an expressed cellular sap containing only 16% of the original water. For comparison, in frost-hardy cereals the fraction of frozen leaf water at temperatures at which frost injury starts, amounted to a similar proportion, namely 23-21% of the total water content (Gusta et al. 1975). For the other three species listed in Table 3,

281

Fig. 6A, B. Photomicrographs of cross-sections in the palisade layer of a frozen leafof D. keniodendron. A Ice crystals(arrows) coating the intercellular spaces (0- B Cellular structure of the same area. Bar=25 ~m such calculations cannot be made since expressed cellular sap was not collected. The moderate extent of supercooling which was necessary to start freezing of the leaves (Table 2, Fig. 4) together with the relatively small extent of frost resistance and the finding that the freezing points of leaves were close to that of pure water confirms the conclusion regarding near-ideal freezing behaviour drawn from the investigations of the water potentials. 3. Anatomical studies

From frozen leaves of D. keniodendron at predawn and subfreezing air temperatures ( - 3 ~ C), crosssections were prepared and microscopically inspected in the field. Intact cells could only be observed in rather thick sections; therefore, photomicrographs had to be taken of the same area at several focussing levels (Figs. 5, 6). Considerable amounts of ice were detected in the intercellular spaces or as icecaps on the surface of the cells, providing additional evidence for extracellular freezing of water. The crystals disappeared when the slides were heated above 0 ~ C, thus indicating

282

their nature of ice. Ice crystals could not unequivocally be identified within the space between the cell wall and the more or less plasmolysed protoplast. Thus freezing of excreted cell water appears to occur predominantly in the intercellular space. From the dimensions and arrangements of the cells in turgescent spongy and palisade parenchyma, it was calculated that the transport of 85% of the cellular water into the adjacent intercellular spaces would still not fill them completely. The spongy structure of the mesophyll tissue exhibiting abnormally large intercellular spaces even in the palisade layers may be a prerequisite for these leaves to cope with the almost regular nocturnal freezing. It should be noted that the leaves of the Afroalpine 'giant rosette' plants, by nyctinastic movements, form a nocturnal 'night bud' which insulates the central leaf bud from freezing (Beck et al. 1982). These nyctinastic movements may, at least partially, be caused by the decrease of turgor during freezing of cellular water in the intercellular spaces. The authors wish to thank Professor John O. Kokwaro, Botany Department of the University of Nairobi for his generous and tireless help which rendered their field work possible. In addition they acknowledge the assistance of the Kenyan Ministry of Wildlife and Tourism and in particular of the warden of Mr. Kenya National Park, Mr. Mungai. The assistance of W. Keil, R. Zimmerman and Dr. P. Ziegler with the field experiments is gratefully acknowledged. For fruitful discussions, the authors are very indebted to Professor W. Larcher, Botanisches Institut der Universit/it Innsbruck, Austria and to Professor J. Boyer, Department of Botany, University of Illinois, Urbana, USA.

References Anderson, J.A., Gusta, L.V., Buchanan, D.W., Burke, M.J. (1983) Freezing of water in Citrus leaves. J. Am. Soc. Hortic. Sci. 108, 397-400 Beck, E., Senser, W., Scheibe, R., Steiger, H.M., Pongratz, P. (1982) Frost avoidance and freezing tolerance in Afroalpine 'giant rosette' plants. Plant Cell Environ. 5, 215-222 Burke, M.J., Rajashekar, C., George, M.F. (1983) Freezing of plant tissues and evidence for large negative pressure potentials. Plant Physiol. 72 Suppl., 44 Chen, P.M., Burke, M.J., Li, P.H. (1976) The fi'ost hardiness of several Solanum species in relation to the freezing of

E. Beck et al. : Equilibrium freezing in Afroalpine plants water, melting point depression, and tissue water content. Bot. Gaz. (Chicago) 137, 313-317 Cosgrove, D.J., Cleland, R.E. (1983) Solutes in the free space of growing stem tissues. Plant Physiol. 72, 326-331 Gusta, L.V., Burke, M.J., Kapoor, A.C. (1975) Determination of unfrozen water in winter cereals at sub-freezing temperatures. Plant Physiol. 56, 707-709 Gusta, L.V., Tyler, N.J., Chen, T.H.-H. (1983) Deep undercooling in woody taxa growing north of the - 40 ~ C isotherm. Plant Physiol. 72, Suppl., 44 Harrison, L.C., Weiser, C.J., Burke, M.J. (1978) Freezing of water in red-osier dogwood stems in relation to cold hardiness. Plant Physiol. 62, 89%901 Larcher, W. (1984) K/iRe und Frost. In: Handbuch der Pflanzenkrankheiten, Bd. 1, Sorauer, P., ed. Parey, Berlin (in press) Levitt, J. (1978) An overview of freezing injury and survival, and its interrelationships to other stresses. In: Plant cold hardiness and freezing stress, pp. 3-15, Li, P.H., Sakai, A., eds. Academic Press, New York San Francisco London Marcellos, H., Burke, M.J. (1979) Frost injury in wheat: ice formation and injury in leaves. Aust. J. Plant Physiol. 6, 513-521 Olien, C.R. (1978) Analyses of freezing stresses and plant response. In: Plant cold hardiness and freezing stress, vol. 1, pp. 37-48, Li, P.H., Sakai, A., eds. Academic Press, New York San Francisco London Palta, J.P., Levitt, J., Stadelmann, E.J., Burke, M.J. (1977) Dehydration of onion cells: a comparison of freezing vs. desiccation and living vs. dead cells. Physiol. Plant. 41,273-279 Rajashekar, C , Burke, M.J. (1982) Liquid water during slow freezing based on cell water relation and limited experimental testing. In: Plant cold hardiness and freezing stress, vol. 2, pp. 211-220, Li, P.H., Sakai, A., eds. Academic Press, New York London Schulze, E.-D., Beck, E., Scheibe, R., Ziegler, P. (1984) Carbon dioxide assimilation and stomatal responses of Afroalpine 'giant rosette' plants. Oecologia (Berlin) (in press) Shackel, L.A. (1984) Theoretical and experimental errors for in-situ measurements of plant water potential. Plant Physiol. 75 (in press) Steponkus, P.L., Garber, M.P., Myers, S.P., Lineberger, R.D. (1977) Effects of cold acclimation and freezing on structure and function of chloroplast thylakoids. Cryobiology 14, 303-321 Van Haveren, B.P. (1972) Measurements of relative vapor pressure in snow with thermocouple psychrometers. In: Psychrometry in water relation research, pp. 178-184, Brown, R.W., Van Haveren, B.P., eds. Utah Agricultural Experiment. Station, Utah State University Washburn, E.W., West, C.J., eds. (1928) International critical tables of numerical data, physics, chemistry and technology, vol. 3. McGraw-Hill, New York Received 5 May; accepted 15 June 1984