leaf water relations characteristics of lupinus angustifolius and l. cosentinii

Oecologia (1990) 82:114-121 Oecologia �9 Springer-Verlag 1990

Leaf water relations characteristics of Lupinus angustifolius and L. cosentinii C.R. Jensen* and I.E. Henson** CSIRO Dryland Crops and Soils Research Unit, Laboratory for Rural Research, Private Bag, P.O. Wembley, W.A. 6014, Australia

Summary. Lupins (Lupinus angustifolius and L. cosentinii) growing in 32 1 containers in a glasshouse were exposed to drought by withholding water. Leaf water potential (~ul), and leaf osmotic potential (~/s) were measured daily as soil water became depleted. Leaf water relations were further assessed by a pressure-volume technique and by measuring q/~ and relative water content of leaves after rehydration. Analysis by pressure-volume or cryoscopic techniques showed that leaf osmotic potential at saturation (~'~oo) decreased from - 0 . 6 MPa in well watered to - 0 . 9 MPa in severely droughted leaves, and leaf water potential at zero turgor (~'zt) decreased from about - 0 . 7 to -1 .1 MPa in well watered and droughted plants, respectively. Relative water content at zero turgor (RWCzt) was high (88%) and tended to be decreased by drought. The ratio of turgid leaf weight to dry weight was not influenced by drought and was high at about 8.0. The bulk elastic modulus (e) was approximately halved by drought when related to leaf turgor potential (g/p) and probably mediated turgor maintenance during drought. The latter was found to be negatively influenced by rate of drought. Supplying the plants with high levels of K salts did not promote adjustment or turgor maintenance.

Key words: Lupinus - Pressure volume curves - Turgor maintenance - Osmotic adjustment - Tissue elasticity

The broad leafed sand plain lupin (Lupinus cosentinii) oc- curs naturally in coastal areas of Morocco with limited occurrences in southern Spain and Portugal, Sicily, Corsica and Tunisia. It is also naturalised on the coastal sandplain soils of Western Australia where it is grazed (Gladstones 1974). The narrow leafed lupin (L. angustifolius) is more widespread around the coastal areas of the Mediterranean Sea, extending inland to about 1500 m in North Africa and Iberia (Gladstones 1974). It is grown widely as a crop, par- ticularly in Western Australia.

Lupin plants have a deeply penetrating tap root system enabling the plants to take up water at considerable depth

* Permanent address and address for offprint requests: Department of Soil and Water and Plant Nutrition. The Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK-1871 Frederiks- berg C, Copenhagen, Denmark ** Present address: 4 Lorong Serai Dua, Taman Cheras, 56/00 Kuala Lumpur, Malaysia

at maximum of about 2.2 m on the coarse textured soils (Hamblin and Hamblin 1985). Earlier investigations (Hen- son et al. 1989b) showed that stomatal conductance and net photosynthesis of lupins were significantly reduced even if only a mild plant water deficit was imposed without mea- surable changes in either leaf water potential (~ut), leaf turgor potential (gtp) or leaf relative water content (RWC).

In the present study we evaluate the leaf water relations characteristics of both L. cosentinii and L. angustifolius and the extent of osmotic adjustment and turgor maintenance when the plants are exposed to soil water stress. Also we examine some of the factors which may influence these processes, specifically the rate of water deficit imposition and potassium supply. The degree of osmotic adjustment has been found to be affected by rate of stress development, with high rates causing less adjustment than low rates (Turner and Jones 1980; Morgan 1984). Furthermore, increased potassium supply has significantly increased solute concentrations in leaves of cereals (Jensen 1982, ! 985), and beans (Jensen 1981 ; Mengel and Arnecke 1982). The impli- cations of the findings of the present and earlier investigations (Henson et al. 1989b) on drought tolerance of lupins are discussed.

Materials and methods

Plants of the narrow-leafed lupin Lupinus angustifolius L. and the broad-leafed sandplain lupin (Lupinus cosentinii Guss.) were grown as described previously (Henson et al. 1989a) in a naturally-lit, temperature-regulated glasshouse. All experiments were done in Perth, Western Australia. Growing conditions in the glasshouse were: photoperiod 11 h, night/day temperature 15~176 relative humidity 55-65%, maximum irradiance 1600 gmol quanta m -2 s-1 of photosynthetically active radiation.

Plants were grown in 32 1 plastic containers holding 49 kg oven dry weight of a coarse, sandy soil. The soil had a bulk density of 1.58 x 106 g m 3 and a water content at field capacity of about 0.16m 3 m -3 which decreased to about 0.03 m 3 m 3 when the plants wilted permanently. Fertilizers were mixed into the soil prior to use to provide major nutrients and trace amounts of copper, zinc, boron, molybdenum and cobalt as specified for lupins by Farr- ington and Pate (1981). The seeds were inoculated with Bradyrhizobium. In two experiments with L. cosentinii, plants were supplied with two contrasting levels of potassium (K) in the soil; " l ow" K, 27 gg g-1 naturally present

in the soil or "h igh" K, 240 ~tg g- 1 following the addition of potassium sulphate. In all other experiments plants received 107 ~tg K g-1 as a supplement to the background level.

Four plants were grown in each pot. In each experiment plants were either watered daily to restore soil water to about 0.16 m 3 m - 3 or drought was imposed by withholding water. This was done at 27 to 65 days after sowing. For each treatment 4-6 replicates were used. In all cases the soil surface was covered with aluminium foil to prevent water loss by evaporation. No toxic effects were observed from the aluminium when compared with non-covered pots.

L e a f water relations measurements by three approaches (i-iii)

Different approaches were used to characterize leaf water relations and assess the degree of osmotic adjustment.

(i). Leaf water relations and osmotic adjustment were assessed from pressure-volume curves obtained using a pressure chamber (Turner 1988). Leaves were at dawn cut under distilled water and rehydrated in a dark humid box for 2 to 4 h. The droughted leaves were always rehydrated for 4 h. Drying of the leaves between ~uz measurements took place on the bench and leaf water changes were determined by weighings preceding the ~a determination.

Values for leaf turgid weight (TW) was calculated following the procedure of Ladiges (1975). Fresh weight (FW) and turgid weight (TW) were used to determine the turgid weight (TW)/dry weight ratio (TW/DW) and the relative water content:

RWC =- (FW-- DW)/(TW-- DW). (1)

The bulk elastic volumetric modulus (e) can be defined by the equation (e.g. Hellkvist et al. 1974):

= d~up/dV x V (2)

where dl//p is the change in turgor pressure for an infinitesi- mal change in symplastic cell volume (V) or tissue water content when the apoplastic water fraction is low. Both the findings of a high TW/DW ratio, Table 1, (Campbell et al. 1979), and the absence of negative turgor values when severely droughting the plants (Fig. 5) indicated a low fraction of apoplastic water (Wenkert 1980). Hence V and d V may be replaced by RWC and d(RWC), respectively, i.e. :

e = d ~ v / d ( R W C ) x RWC (3)

By plotting ~u~ versus RWC-~ (type I transformation: c.f. Richter 1978), curves were obtained by regression relat- ing ~u~ to RWC -1 with an initial non-linear portion followed by a linear section, where the beginning of the linear portion indicates the leaf water potential at the turgot loss point (~uzt) or the RWC at zero turgor (RWCzt), and the extrapolated linear part the osmotic potential of the turgid state (~uslOO) of the leaves or with opposite sign ~u v when ~ul=0 and R W C = I (Fig. 1). The turgor potential (q/v above the turgot loss point) was calculated as the difference between ~ul and leaf osmotic potential (~us) :

~up = q& - ~s (4)

Assuming an exponential function exists between ~up and RWC-~ (Fig. 1), then at R W C = 1 the maximum turgot (q/p(ma~)) is reached. In the pressure range of positive turgot, the turgor can be described as (Stadelmann 1984):

115

In q/p--ln I//p(max):fl (RWC - 1 - 1 ) (5)

where t , the sensitivity factor of elasticity, relates exponential changes in turgor to changes in RWC and was determined as the regression coefficient for individual determina- tions of RWC-1 versus In ~Up according to Stadelmann (1984; equation 2). The bulk elastic modulus (e) at a given turgor pressure was calculated as (Stadelmann 1984):

e = - q~p-fl 1/RWC (6)

(ii). Leaves were excised at dawn and rehydrated as for the pressure-volume technique. The leaves were then sampled for osmotic potential (~'s) and relative water content (RWC) as described below. The osmotic potential obtained was corrected to that at 100 per cent RWC (~U~loo) where ~Usloo = q/s x RWC. This approach neglects any dilution of cell water with extracellular water and equates RWC with relative symplastic water content. At values of RWC close to 1, however, the errors introduced by these approxima- tions are minimal when the apoplastic water fraction is small.

RWC (equation 1) was determined by floating leaflets on distilled water for 4 h at room temperature (about 22 ~ C) under dim light. The leaflets of L. cosentinii were partly slit along the mid-rib to aid water uptake. The turgid weight was then measured after blotting and the samples were dried for 24 h at 80 ~ C and the dry weight determined.

(iii). Leaves were sampled daily at midday (1100 to 1300h Australian Western Standard Time) or daily throughout the photoperiod from watered and droughted plants at three levels of stress during the development of a drying cycle. The uppermost fully expanded leaves were selected as previously described (Henson etal. 1989b). Leaves were enclosed in a polyethylene bag and then rapidly detached and transferred to a pressure chamber (Turner 1988) for measurement of q&. Other leaves were rapidly frozen in liquid nitrogen for subsequent measurement of Iffs.

~u~ was measured using either a single, middle leaflet or an entire leaf. The latter included a short length of petiole. In initial experiments, ~ul of droughted plants measured on single leaflets was consistently higher than that found using whole leaves; the difference increasing as ~ua declined. As the use of ~u~ measured on single leaflets gave rise to anomalously high values of leaf turgor potential of droughted leaves (derived using the relation: qJl = q&+ ~Up), values of ~u~ obtained with whole leaves were used. In experiments where q/1 was measured only on leaflets, these values were converted to values of whole leaf ~u~ by use of an empirically derived regression equation.

For measurement of ~us, leaf material was rapidly frozen in liquid nitrogen, later thawed for about 15 minutes and the sap expressed with a plastic syringe. Varying the thaw- ing time between 1 and 60 minutes had no effect on the resultant value of ~us- q& was measured either by thermocou- ple hygrometry using a Wescor C52 chamber connected to a Wescor HR33T microvoltmeter (Wescor Inc., Logan, Utah, USA) or by measurement of freezing point depres- sion with a Roebling micro-osmometer (H. Roebling, Ber- lin, West Germany).

Osmotic potential o J apoplastic water

Apoplastic water was obtained from leaves by applying a small (0.2 MPa) overpressure to leaves inserted into the

116

pressure chamber (Ackerson 1982). The exuding sap was retained by a plastic tube fitted over the pro t ruding petiole and sap was successively recovered with a syringe as 25 gl aliquots. Tota l solute concentra t ion of the sap was measured using the micro-osmometer . As the first a l iquot of sap collected may have been contamina ted by cut cells (al- though the petiole s tump was first r insed with deionised water before collection), results presented are for the second al iquot only.

Results

Measurements o f "long t erm" leaf water relations by the three approaches (i-iii)

During the drying cycle ~u~ decreased by 0.08 to 0.1 MPa d a y - ~ unless othervise indicated. The plants were consid- ered severely stressed in that they were wilted, and the lower leaves were yellowing or had abscised. Soil water content had declined from 0.16 to 0.03 m 3 m -3. Midday ~ux was about 1.0 M P a below values for the watered plants for both species (Table 1).

i. The pressure-volume (PV) technique was used to assess differences in leaf water relat ions between watered and droughted plants at the end o f a drought ing period. In Fig. I pooled PV measurements for several leaves are given as leaf water potent ia l (g~) versus R W C - 1 in watered and droughted leaves of L. cosentinii and L. angustifolius. The results of the analysis are summarised in Table 1. The turgid weight/dry weight rat io was high and similar in both species and unaffected by drought . R W C at zero turgor (RWCz0 was about 87-90% for watered plants of both species and about 83-86% for droughted plants. The estimates of ~u~ at full hydra t ion (~U~zoo) were between - 0 . 5 2 to - 0 . 5 9 M P a in watered plants and about - 0 . 9 0 to - 0 . 9 2 M P a in droughted plants. The difference in ~%~oo between watered and droughted plants, which estimates the level of osmotic adjus tment induced by drought , was about 0.3 in L. cosentinii and 0.4 M P a in L. angustifolius. An adjustment of similar magni tude was found when est imating the difference in ~Us at zero turgor ({Uzt) between watered and droughted plants.

F o r the droughted plants the osmotic relations followed Boyle van ' t Hoff ' s law for p lant cells. Thus for L. angustifolius Nzt x RWCzt = -- 0.92 which equals the est imate o f ~gsloo by the regression curve (PV, {U~aoo; Table 1). F o r watered plants, q4aoo est imated by the regression curves; i.e. - 0 . 5 2 and - 0 . 5 9 M P a for L. angustifolius and L. cosentinii, respectively (Table 1); were about 0.1 M P a higher than calculated by the van HofFs law. Similar discrepancies have been noted in other p lant mater ia l (Kim and Lee-Stadelmann 1984).

The l imitat ions of the type I t ransformat ion (q/t versus R W C - ~ ) opposed to the more common 1/{u~ versus R W C plot (type II t ransformat ion) , namely as poin ted out by Tyree and Richter (1982) that with a large fraction of apoplastic water the former t ransformat ion becomes non-l inear (causing error in est imating {u~ and q4aoo by linear extrapolation), was investigated by doing both t ransformations. Similar values of ~U~oo and ~uz~ was obta ined by both trans- format ions indicat ing a low apoplast ic water fraction. Therefore the use of the type I t ransformat ion is not expected to cause considerable errors for this material .

0i ,x A

-0.4 .

8~176 o 1 "

-0.8 ~ t ~ ~ ~-:2.... o �9 o o

[3- ~q.._| o

-6 k,%

r - i i i i

"E 1.0 1.1 1.2 1.3 1.4 1.5

o_ O- Ooo

~, _o B

~ - 0 . 4 -

-o.s [_ " ~ o

-1.2 I "*'~"'~--" "

/ I I I I I

1.0 1.1 1.2 1.3 1.4 1.5

Re la t i ve w a t e r con ten t q Fig. 1 A, B, Pressure-volume curves for fully expanded leaves of L. cosentinii (A) and L. angustifolius (B). (o) indicates fully watered plants and (o) plants which have been slowly droughted to a pre- dawn leaf water potential of - t . 6 _+ 0. l MPa. Arrows indicate turgor loss points. Curves were fitted by a linear function below, and by a exponential function (equation 5) above, the turgor loss point (indicated by a arrow), respectively, on results of four to six leaves. For L. cosentinii, watered, ~'1 = 1.15 - 1.70 x RWC 1, P<0.0001, r=0.92; fl= --52.9+4.1 (s.e. estimate), P<0.0001, r = 0.92. For L. cosentinii, droughted, ~ul = 0.033 - 0.94 x RWC- 1 p < 0.0009, r=0.67; fl= -22.6-+1.34, P<0.0001, r=0.95. For L. angustifolius, watered, qJl =0.86-1.38 x RWC -1, P<0.001, r=0.81 ;

= - 35.5 + 2.43, P < 0.0001, r = 0.91. For L. angustifolius, droughted, gq = 0.035 - 0.95 x RWC- 1, p < 0.0009, r = 0.65 ; fl = - 17.0-+0.72, P<0.0001, r=0.96

The exponential factor (fl), which relates the sensitivity of elasticity changes to changes in RWC, was significantly changed by drought (Table 1). A t full turgor e was at maximum and tended to be decreased by drought (Table i). A t R W C close to one e was at the highest level in watered plants of both species (Fig. 2 B). e increased p ropor t iona l with increase in {up (Fig. 2 C) and the increase was steepest in watered plants. The curves of Fig. 2 C indicate an approximate halving o f e in droughted plants for the ~up range from 0 to about 0.5 MPa. RWCzt and e tended to be at the highest and fl at the lowest level in L. cosentinii in both watered and droughted plants (Table 1, Fig. 1, and Fig. 2 B and C).

Merging the results of several PV curves when calculating leaf water characterist ics by PV analysis may not be reliable enough when calculating elastic behavior (Kikuta and Richter 1986). Therefore the logari thmic t ransforma- tion model (equation 5) was also applied to PV curves of individual leaves. Similar results in respect to fitting da ta by the model were obtained as when used on merged da ta (Fig. 1), as i l lustrated in Fig. 3 where calculated and " o b - served" ~Up values are shown in a typical watered and droughted leaf of L. angustifolius. At high R W C (>0 .98) the observed {up values are lower than est imated by the model.

Table 1. Water relations characteristics of leaves as determined by pressure-volume technique (PV) or expressed sap (cryoscopic) technique (ES). ~t0o, osmotic potential at full hydration; ~ , osmotic potential at zero turgot; RWC~t, RWC at zero turgor; fl, sensitivity factor of elasticity; e . . . . maximum bulk modulus of elasticity; TW/DW, turgid weight/dry weight ratio. Values are based on data of four to six leaves

117

L. angustifolius L. cosentinii

Watered Droughted Difference Watered Droughted Difference

Midday ~ul (MPa) -0 .64 -1 .59 ***x 0.95 -0 .68 -1 .69"** 1.01 ES,~s~.oo (MPa) --0.69 --0.91 *** 0.22 --0.57 --0.94** 0.37 PV,~too (MPa) -0 .52 -0 .92** 0.40 --0.59 --0.90** 0.31 PV,~zt (MPa) --0.72 --1.11"** 0.39 -0 .75 -1 .06"** 0.31 PV,RWC,t (%) 86.9 82.6 4.3 90.0 85.6 4.4 PV,fl --35.5 --17.0"** -18 .5 --52.9 -22.6*** -30 .3 PV,~m,,~ (MPa) 18.4 15.7 2.7 29.5 20.4 9.1 PV,TW/DW 7.7 7.7 0 8.3 8.3 0

1, , , **, *** Significantly different at the 0.05, 0.01 and 0.001 probability levels, respectively. The standard errors of the mean (n=4-6) or of the estimate from the pooled regression curves (Fig. 1) were used in a t test or for confidence limits to determine the level of significance between means or estimates, respectively (Snedecor and Cochran 1967)

0.8 0_

~' o.4 3 b--

24 --~ CL ~E

~-~ 16

1.1J 8

0 0,8

A

L. cosentinii /// /

B

L. cosentinii

/ /

- / ~ ,

0.9 1.0

L /

L. angustifolius / /

L. angustifolius

./ / /

/ /

0.9 1.0 Relat ive wa te r c o n t e n t

>,

bo O

LLI

cosentinii /

/

16 / / /

/ /

/ 8 /

/

i

0 0.L

L. angustifolius

/ /

i i r i i

0.8 0 0.4 Turgot po tent ia l (MPo)

/ /

/ /

i

0.8

Fig. 2A-C. Estimates of turgor potential (~,p) and bulk elastic mo- dolus (e) as a function of RWC or qJp in L. cosentii and L. angustifolius. The fully drawn curves indicate watered plants and the broken curves droughted plants, respectively. The estimates were derived from the regression curves of Fig. 1 as outlined by Stadelmann (1984)

/ 0.8 I

"S I - [3._ / .

0.6 / . - / .

c f /

-3o_ 0.4 . / ~ /L_

/ /o o~ / o/o

0.2 " S o / / ~ , / o /

va "~ o /

0 - - ~ ~ ~,~" ,.~ .. o . , ,

0.85 0.90 0.95 1.00 Rela t ive w a t e r con ten t

Fig. 3. Turgor potential (~p) as a function of RWC for a typical single leaf of a watered (o) and a droughted (e) plant of L. angustifolius. The curves were fitted by the exponential function (equation 5) to the calculated observations. Maximum ~,p was derived from extrapolation of the q/1 on I /RWC regression line to RWC = 1 from results below the turgor loss point

ii. Resul ts ob ta ined af ter measu r ing ~'s on rehydra ted samples and cor rec t ing to 100 per cent R W C to ob ta in ~stoo are also shown in Tab le 1 (ES method) . The values o f q/sioo were s imilar to those ob ta ined with the PV m e t h o d in L. cosentinii. In watered plants o f L. angustifolius, ~q~oo de te rmined by the ES m e t h o d was a b o u t 0.17 M P a lower than by the PV m e t h o d ; however , in the d r o u g h t e d plants s imilar values were found.

iii. The re la t ionship be tween g/s and qJl measured "in s i tu" at m i d d a y dur ing d rough t showed two pat terns . W h e n d rough t stress deve loped slowly (by a fall in ~1 o f < 0 . 1 0 M P a d a y - l ) , ~s decl ined in paral le l wi th ~1, as ~1 fell f rom - 0.6 to - 1.1 M P a , to ma in t a in a tu rgor potent ia l at least equ iva len t to tha t in watered plants (Fig. 4A) . In these cases Np was ma in t a ined posi t ive to a value o f ~,1 as low as abou t - 1 . 3 M P a . W h e n stress d e v e l o p m e n t was more rap id ( > 0.14 M P a d a y - t ) , there was no such adjust- men t and [/]p became close to zero when ~ t was a b o u t - 1 . 1 M P a (Fig. 4B). qJl o f wa te red plants was at a slightly

118

-1.8 Water potential (MPa)

-1.4 -1.0

q-

-0.6

/ r

B

-0./,

-0.8

-1.2 ~:

-6

-1.6 o

._o -0.6 -~

E

-1.0

-1./,

-1.8 Fig. 4A, B. Relationship between osmotic potential (~u,) and water potential (~ua) of leaves of L. cosentinii during (A) "slow" development of water stress (<0.10 MPa day t, A~tt/Atime ) and (B) "fast" development of water stress (>0.14MPa day - t , A~uz/ Atime). In (A) data are pooled from three, and in (B), from two experiments. The curves show main trends for watered (o) and droughted (o) plants, respectively. Data points are means of four to six leaves; bars indicate 2 x pooled s.e. mean. The 1:1 relationship between q/1 and gt~ (~,p =0) is indicated by the diagonal lines

lower level at the rapid rate of droughting. The rapid rate was obtained by performing the experiments during a period with comparat ively high evaporat ive demands inducing a greater leaf water deficit even in fully watered plants.

~ at midday during a drought ing treatment, consistently gave lower values for L. angustifolius than for L. cosentinii. Thus, over most of the range of water potent ia l at which positive turgor was maintained, ~,p was higher in L. angustifolius than in L. cosentinii (Fig. 5).

Fur the rmore ~p declined approximate ly linearly with ~,~ during the first hal f of the day as leaves par t ia l ly dehy- drated. Ext rapola t ion o f this relat ionship to zero qJp pro- vided estimates of ~ at zero tu rgo t of - 0 . 8 7 to - 0 . 9 4 MPa for watered plants and - 0 . 9 2 to - 1 . 5 3 M P a for droughted plants. The difference in ~ at zero ~p between watered and droughted plants increased from zero M P a (Fig. 6A) to 0.58 MPa (Fig. 6C) as drought stress intensi- fied. Hysteresis in the relat ionship between qJp and ~ was evident for watered plants and for droughted plants under mild stress, but leaves of the most severely stressed plants (Fig. 6C) did not regain turgor during the day, and hence hysteresis could not be ascertained in this case. The hysteresis indicates accumulat ion of solutes in the leaves during the pho toper iod (Morgan 1984).

F r o m the diurnal changes in ~p versus ~ , the max imum ~p (~p at ~ = 0 or full turgor) and hence ~ o o could also be ascertained assuming the relat ionship to be l inear up

0.5

0./, O_ ~E

0.3 �9 �9 r

o_ 0.2

~- 0.1 '~ ~ !

0 -1.2 -1.0 -0.8 -0.6 -0./,

Water potential (MPa) Fig. 5. The relationship between leaf turgor potential (~p) and water potential (N1) determined by sampling daily at midday during the development of water stress for L. angustifolius (A) and L. cosentinii (A). Lines are fitted linear regressions of ~p on ~1; for L. angustifolius r=0.9t , Np=0.625 +0.56 x gtl, and for L. cosentinii r = 0.90, Np = 0.362 + 0.37 x gq ; both significant at P < 0.001. Data points are means of four leaves

0.8

0.6

0.4 "-5 o_

o.2

g 0.8 Q.

,- 0.6 0 o')

0.4

/Z+ -0.8 -0.4

-5

B// 0 -0.8 - O . L ~ O .

/ 02 S

'/-7, ..": 0 . . . . t. , , ,

-1.6 -1.2 -0.8 -0.4 0 Water potential (MPa)

Fig. 6A-C. The relationship between leaf turgor potential (grp) and water potential (~t) of L. cosentinii at different times of day for watered (o) and droughted (o) plants at three stages (A-C) of increasing water stress. Data points are means of six leaves; bars indicate _+2xpooled s.e. mean. Diagonal arrows indicate order of sampling and vertical arrows indicate the point of zero turgot derived by extrapolation of the linear relationship between ~,p and ~1 for points on the descending part of each plot

to full turgor. F o r control leaves, using the lower (descending) por t ion of each curve, ~Ustoo ranged from - 0 . 7 3 to - 0 . 7 6 MPa. These values were lower than those obtained when leaves were detached and rehydrated (Table 1) and may reflect loss of solutes in the leaves during the 2-4 h rehydra t ion period.

Effect o f potassium supply on leaf water relations o f lupin

Varying soil K supply to plants of L. cosentinii about 9-fold at two growth stages resulted in a 2.4-fold mean difference

Table 2. Effect of potassium (K) supply on the contribution of the ion to total osmotic potential of leaves (~u~) of watered and droughted plants of L. consentinii at two growth stages

119

Treatment K Leaf water Leaf osmotic ~sk ~'sk/~s supply potential potential (MPa) total

(MPa) (MPa)

a) Vegetative plants

Watered

Droughted

b) Flowering plants

Watered

Droughted

low -- 0.97 • 0.04 -- 1.06 • 0.03 -- 0. l0 + 0.007 0.094 high -- t.00 • 0.02 - 1.08 • 0.02 -- 0.19 • 0.009 0.176

low -- 1.80 -t- 0.08 -- 1.93 • 0.06 - 0.17 • 0.011 0.088 high - 1.46 • 0.07 - 1.67 • 0.04 - 0.28 • 0.019 0.168

low -- 0.97___ 0.02 -- 1.15 • --0.07 • 0.002 0.061 high - 1.01 • 0.02 - 1.12 • 0.02 - 0.25 • 0.001 0.223

low - 1.61 • - 1.66+_0.05 - 0.10 • 0.008 0.060 high -- 1.51 • 0.05 -- 1.61 • 0.04 -- 0.33 • 0.013 0.205

Data are means_+s.e.m, for the last two days of a drought treatment; n=6; gtsk= ~ due to K, assuming 40 mmol K=0.1 MPa

Table 3. Water potential and solute concentrations and osmotic potentials of apoplastic water collected from leaves of watered and droughted plants of L. cosentinii

Treatment Leaf water Solute Osmotic potential concentration potential (MPa) (mosmoles) (MPa)

Watered - 0.70 • 0.03 9.7 __ 1.4 - 0.024 Droughted - 0.95 • 0.02 9.3 -I- 0.9 - 0.023

Data are means of 12 leaves• The solute potential in MPa was calculated assuming 40 mosmoles = 0.1 MPa

in K concentra t ion in the leaves. Generally, this failed to affect significantly leaf osmotic potentials (Table 2). How- ever, as a result of high K supply, ~ul of droughted plants tended to be higher near the end o f the drought t reatment (Table 2). This was p robab ly a consequence of the slower rate of water use by the high K plants (results not presented). There were, however, only negligible effects of K on ~Up. The contr ibut ion o f K to the total osmotica was greater with high K, but contr ibut ions of other solutes changed to compensate for the al terat ions in K supply.

Solute potential of apoplastic water

Apoplas t ic water expressed from detached lupin leaves using the approach of Ackerson (1982), had an osmotic potential higher than - 0 . 0 3 MPa and an osmolal i ty less than 2.7 per cent of that of bulk leaf sap (Table 3), assuming a mean value for the lat ter for watered plants of 363 mosmoles ( ~ - 0 . 9 MPa). There was no effect of drought on solute concentrat ions in the apoplas t ic water.

Discussion

Linear, exponential and power functions, have been used to describe the relat ionship of ~up versus R W C and ~up versus ( R W C - I - 1 ) , e.g. Kiku ta and Richter (1986). In the present study an exponential function was used to observations of ~,1 versus R W C - ~ of several leaves in the same

plot, or ~,p versus R W C for a single leaf. Deviat ion from the calculated curvilinear form near water sa tura t ion was evident in both plots. We suggest that the deviat ion may be ascribed inaccuracies, i.e. small overest imations of chamber balance pressure near saturat ion, when using the pressure chamber technique for q/1 determination. How- ever, the approach is justified by the fact that ~up(~x) (determined by ext rapola t ion of the regression line for ~u~ versus R W C - a beyond the tu rgot relaxat ion point to the ~up value for R W C = 1) seemingly fits well into the rest o f the curvilin- early calculated regression curves. Similar deviat ions possi- bly caused by an over-est imation of chamber balance pressure near sa tura t ion were seemingly present in leaves of Phaseolus vulgaris for which the same curve fitting technique (equation 5) was used (Kim and Lee-Stadelmann 1984; Fig. 3).

Midday and diurnal measurements of ~ul and ~us at three levels o f stress, and PV analysis, all indicated that lupin exhibited turgor maintenance at slow drought ing rates and could mainta in positive turgor down to gq values of - 1.1 to - 1 . 5 MPa. Turgor maintenance was induced par t ly by solute accumulat ion (Table 1), a confi rmat ion o f earlier findings by Turner et al. 1987, and par t ly by increase of elasticity. Our results indicated that e was a linear function of ~,p and at similar ~Up values e was about halved in droughted plants o f both species. The doubling o f leaf elasticity of droughted plants resulted in positive turgors being mainta ined at lower R W C values (Fig. 2A) and at lower qJl values (Fig. 1) in droughted plants. Fur thermore , osmotic adjus tment and decrease of e together caused turgot maintenance to improve in droughted lupin leaves when R W C decreased (Fig. 2A). Normal ly one would expect e to increase with drought as thicker and more rigid cell walls are expected under such growth conditions. However, decreases in e due to drought have been repor ted in field- grown beans (Elston et al. 1976) and in pot -grown wheat plants (Kikuta and Richter 1986) exposed to a similar dura- t ion of water stress as the lupin plants repor ted here. The question therefore arises: what tr iggered the decrease of

during drought? In the lupins, similar T W / D W ratios were found in watered and droughted plants indicat ing that cell size had not changed. Also, lupins have a sensitive sto-

120

matal response to drought preventing assimilates contribut- ing very much to osmoregulation. Transport of solutes from the cell walls could have contributed to osmoregulation while decreasing the rigidity of the cell walls. While this point is rather speculative it emphasizes the need for further investigations on cell structure in relation to stress acclima- tion, as also mentioned by Kikuta and Richter (1986).

The high e values in leaves of L. cosentinii indicating more rigid cells in that species, suggest that differences in cell wall structure may exist among species of Lupinus. Simi- lar observations were made by Turner et al. (1987).

The absence of negative turgors even in severely stressed plants (Fig. 4) suggested that dilution errors in estimating ~,~ were generally negligible, though whether this resulted from a low apoplastic water fraction, insufficient mixing of apoplastic and symplastic water, or other reasons (Wen- kert 1980) is not clear. The estimates of qJ~loo by PV and sap extraction methods were, moreover, in good general agreement (Table 1). However, ~Usloo and ~uzt estimates obtained from diurnal measurements (Fig. 6), where gt~ and ~u~ were evaluated "in situ", were lower than when these were measured by techniques involving rehydration of leaves. This also applied to the zero turgor points obtained by plotting mid-day values of ~u~ against ~uz (e.g. Fig. 4A) and may, as mentioned above, indicate solute changes during rehydration of the leaves when the ES- and PV-method were used. On the whole, however, the differences between methods were not large and all techniques pointed to the occurrence of only limited adjustment in leaves of lupin depending on rate and degree of stress.

Estimation of cell turgot, when calculated as the difference between bulk leaf ~ul and ~'s, could also be incorrect if the apoplastic water in the leaf contained an appreciable concentration of solutes relative to that within the cell. This would cause water withdrawal from the cell and hence a lowering of ~,p (Greenway and Munns 1980; Oertli 1984). Such an increase in apoplastic solute concentration during drought could account for the closure of lupin stomates in the absence of measm'able changes in ~uI (Henson et al. 1989b). However, measurements of solute potential of expressed apoplastic water failed to indicate any marked change following drought and hence it seems unlikely that solute accumulation in the apoplast is involved in responses of stomata of lupin to water deficits.

Watered plants of L. angustifolius had a lower V~ than L. cosentinii (by about 0.1 MPa) and consequently had higher turgors (Figs. 2 and 5) as previously reported by Turner et al. (t987). Typical values of gtp for lupins at midday seldom exceeded 0.3 MPa and leaves of L. cosentinii generally "functioned" with midday values of qJp of about 0.2 MPa (Fig. 6). Despite such low turgors, stomatal conductance and rates of CO2 assimilation were high in lupins (Henson et al. 1989a; Turner and Henson 1989). Because of the low ~,p of the watered plant, it is likely that physiological processes in lupin such as growth, stomatal opening and CO2 assimilation would be especially sensitive to decreases in ~1, and this seems to be the case (Henson et al. 1989b).

The ability of lupins to adjust osmotically in response to water deficits was limited; AqJslOO being at most about 0.3 to 0.4 MPa between watered and droughted plants (Ta- ble 1). Also, a capacity to adjust appeared to be lacking or small in experiments with cowpea (Shacket and Hall 1983; Muchow 1985), Phaseolus species (Markhart 1985)

and soybean (Turner etal. 1978; Muchow 1985), and poorly developed in other grain legumes (Muchow 1985). Adjustment in th pasture legume Siratro was ascribed to changes in osmotic volume rather than to solute accumulation per se (Wilson et al. 1980). However, in some legume species osmotic adjustment may be quite appreciable (Lud- low 1980; Ludlow et al. 1983; Flower and Ludlow 1986). Hence, it may be better not to generalize until a wider range of species have been examined.

There appears to be variation between lupin species in ability to osmotically adjust (Turner et al. 1987). In the present study there was evidence for the ~'s of leaves of watered plants being lower in L. angustifolius than in L. cosentinii, also found by Turner et al. (1987), though this led to only a small difference in the g'l at zero ~,p (Figs. 1 and 5). Turner et al. (1987) found that L. atlanticus and L. pilosus were capable of greater adjustment than several other lupin species. Forseth and Ehleringer (1982) found no significant adjustment by the desert lupin, L. arizonicus.

Besides the limited level of turgor and osmotic adjustment, other leaf characteristics also indicated a limited ability of lupin leaves to withstand severe drought. Positive turgors of leaves were only maintained at a slow droughting rate to qJl of about -1 .1 to - 1 . 5 MPa and at ~1 values of - 1 . 5 to - 2 . 0 MPa the leaves turned yellow and died. In a comparison of lupins with wheat, we found (Henson et al. 1989b) that in wheat positive turgors could be maintained down to ~,z values of - 2 . 2 MPa and tropical grasses have been found able to survive extremely low qJl values of -10 .0 to -12 .0 MPa (Wilson et al. 1980).

Similarly the high RWCzt and TW/DW values are characteristics of a low drought resistance. Drought resistant species might be expected to be adapted to large losses of water without loss of turgot (Jane and Green 1983) and the leaf cells would be small and thick walled causing low TW/DW ratios (Cutler et al. 1977).

The high TW/DW values of lupin leaves (7.7 8.3), as compared to other species (Campbell et al. 1979), were not influenced by stress. This contrasts with findings of Turner et al. (1987), who found that osmotic adjustment in roots and leaves of lupins were correlated with decreases in the TW/DW ratio. This may probably be due to the longer period of time during which stress was applied in their experiments. Hence, the leaves may have developed under stress conditions causing small cells with thicker cell walls, so decreasing the TW/DW ratio of the leaves.

The PV technique did not allow us to estimate accu- rately the amount of apoplastic water present. However, the apoplastic water values in herbaceous species range from about 10-30 volume %, and are associated with the amount of cell wall material (Boyer 1967). Due to the high TW/DW ratio of lupin we assume the apoplastic water fraction to be low in lupin. This assumption is supported by findings of Campbell et al. (1979) for potato, which has a TW/DW ratio of 6.5 and a value of apoplastic water fraction of 0.05.

The seemingly low drought tolerance of lupins may be related to their drought "avoidance" strategy whereby water loss is restricted due to a very sensitive stomatal response to drought (Forseth and Ehleringer 1982; Henson et al. 1989b; Turner and Henson 1989). This early closure of stomata would restrict COz assimilation which might limit the solutes available to promote adjustment.

Supplying higher levels of K did not promote osmotic

12t

ad jus tment , which agrees wi th some studies (Wilson and L u d l o w 1983), bu t confl icts wi th o thers (Jensen 1981, 1982 and 1985; Menge l and A r n e k e 1982). Po t a s s ium did n o t appea r to be a n i m p o r t a n t solute with respect to osmot ic ad jus tment in leaves o f a range o f t ropica l l egumes ( F o r d 1984).

Acknowledgements. This work was carried out while the authors were Visiting Scientist and Research Scientist, respectively, at CSIRO Dryland Crops and Soils Research Unit, Floreat Park, Perth, Western Australia. We thank CSIRO for the use of facilities for conducting these experiments. We also thank Ms J.M. Burke for excellent practical help with the experiments and Dr. N.C. Turner for helpful comments on the manuscript. CRJ is grateful to the Danish Veterinary and Agricultural Research Council (Pro- ject No. 13-3479) for granting study leave during which this work was carried out.

References

Ackerson RC (1982) Synthesis and movement of abscisic acid in water-stressed cotton leaves. Plant Physiol 69:609-613

Boyer JS (1967) Matric potentials of leaves. Plant Physiol 42:213-217

Campbell GS, Papendick RI, Rabie E, Shayo-Ngowi AJ (1979) A comparison of osmotic potential, elastic modulus, and apoplastic water in leaves of dryland winter wheat. Agron J 71:31-36

Cutler JM, Rains DW, Loomis RS (1977) The importance of cell size in the water relations of plants. Physiol Plant 40:255-260

Elston J, Karamanos AJ, Kassam AH, Wardsworth RM (1976) The water relations of the field bean crop. Phil Trans R Soc Lond B 273 : 581-591

Farrington P, Pate JS (1981) Fruit set in Lupinus angustifolius cv. Unicrop. I. Phenology and growth during flowering and early fruiting. Aust J Plant Physiol 8 : 293-305

Flower DJ, Ludlow MM (1986) Contribution of osmotic adjustment to the dehydration tolerance of water-stressed pigeonpea (Cajanus cajan (L.)millsp.) leaves. Plant Cell Env 9:33-40

Ford CW (1984) Accumulation of low molecular weight solutes in water-stressed tropical legumes. Phytochemistry 23:1007-1015

Forseth IN, Ehleringer JR (1982) Ecophysiology of two solar tracking desert winter annuals. II. Leaf movements, water relations and microclimate. Oecologia (Berlin) 54:41-49

Gladstones JS (1974) Lupins of the Mediterranean region and Africa. Western Australia Department of Agriculture Bull 26 : 1-48

Greenway H, Munns R (1980) Mechanisms of salt tolerance in nonhalophytes. Ann Rev Plant Physiol 31:149-190

Hamblin AP, Hamblin J (1985) Root characteristics of some tem- perate legume species and varieties on deep, free-draining Ent- isols. Aust J Agric Res 36:63-72

Hellkvist R, Richards GP, Jarvis PG (1974) Vertical gradients of water potential and tissue water relations in Sitka spruce trees measm'ed with the pressure chamber. J Appl Ecol 11 : 63%667

Henson IE, Jensen CR, Turner NC (1989a) Influence of leaf age and light environment on the leaf gas exchange of lupins and wheat. Physiol Plant (submitted)

Henson IE, Jensen CR, Turner NC (1989b) Leaf gas exchange and water relations of lupins and wheat. I. Shoot responses to soil water deficits. Aust J Plant Physiol 16:401-413

Jane GT, Green TGA (1983) Utilisation of pressure volume techniques and non-linear least squares analysis to investigate site induced stress in evergreen trees. Oecologia (Berlin) 57 : 380-390

Jensen CR (1981) Influence of water and salt stress on water rela- tionships and carbon dioxide exchange of top and roots in beans. New Phytol 87:285-295

Jensen CR (1982) Effect of soil water osmotic potential on growth and water relations in barley during soil water depletion. Irrig Sci 3:111-121

Jensen CR (1985) Potassium induced improvement of yield response in barley exposed to soil water stress. Irrig Sci 6 : 117-129

Kim JH, Lee-Stadelmann OY (1984) Water relations and cell wall elasticity quantities in Phaseolus vulgaris leaves. J Exp Bot 35:841-858

Kikuta SB, Richter H (1986) Graphical evaluation and partitioning of turgot responses to drought in leaves of durum wheat. Planta 168:36-42

Ladiges PV (1975) Some aspects of tissue water relations in three populations of Eucalyptus viminalis Labill. New Phytol 75:53 62

Ludlow MM (1980) Stress physiology of tropical pasture plants. Trop Grasst 14:13(~145

Ludlow MM, Chu ACP, Clements RJ, Kerslake RG (1983) Adap- tation of species of Centrosema to stress. Aust J Plant Physiol 10:119-130

Markhart AH (1985) Comparative water relations of Phaseolus vulgaris L. and Phaseolus aeutifolius Gray. Plant Physiol 77:113-117

Mengel K, Arneke W-W (1982) Effect of potassium on the water potential, the pressure potential, the osmotic potential and cell elongation in leaves of Phaseolus vulgaris. Physiol Plant 54: 402-408

Morgan JM (1984) Osmoregulation and water stress in higher plants. Ann Rev Plant Physiol 35:299-319

Muchow RC (1985) Stomatal behaviour in grain legumes grown under different soil water regimes in a semi-arid tropical environment. Field Crops Res 11:291-307

Oertli JJ (1984) Water relations in cell walls and cells in the intact plant. Z Pflanzenernaehr Bodenk 147:187-197

Richter H (1978) A diagram for the description of water relations in plant cells and organs. J Exp Bot 29:1197 1203

Shackel KA, Hall AE (1983) Comparison of water relations and osmotic adjustment in sorghum and cowpea under field conditions. Aust J Plant Physiol 10:423-435

Snedecor GW, Cochran WG (1967) Statistical Methods. Sixth Ed. The Iowa State Univ Press, Ames, USA

Stadelmann EJ (1984) The derivation of the cell wall elasticity function from the cell turgor potential. J Exp Bot 35:859-868

Turner NC (1988) Measurement of plant water status by the pressure chamber technique. Irrig Sci 9:289-308

Turner NC, Jones MM (1980) Tnrgor maintenance by osmotic adjustment: a review and evaluation. In: Turner NC, Kramer PJ (eds) Adaptation of Plants to Water and High Temperature Stress. John Wiley, New York, pp 8%103

Turner NC, Henson IE (1989) Comparative water relations and gas exchange of wheat and lupins in the field. In : Kreeb KH, Richter H, Hinckley TM (eds) Structural and Functional Re- sponses to Environmental Stresses: Water Shortage. SPB Aca- demic Publishing: The Hague, pp 293 304

Turner NC, Begg JE, Rawson HM, English SD, Hearn AB (1978) Agronomic and physiological responses of soybean and sorghum crops to water deficits. III. Components of leaf water potential, leaf conductance, 14CO2 photosynthesis, and adaptation to water deficits. Aust J Plant Physiol 5:179-194

Turner NC, Stern WR, Evans P (1987) Water relations and osmotic adjustment of leaves and roots of lupins in response to water deficits. Crop Sci 27:977-983

Tyree MT, Richter H (1982) Alternate methods of analysing water potential isotherms. Some cautions and clarifications. II. Curvi- linearity in water potential isotherms. Can J Bot 60:911-916

Wenkert W (1980) Measurement of tissue osmotic pressure. Plant Physiol 65 : 614~617

Wilson JR, Ludlow MM, Fisher M J, Schulze E-D (1980) Adapta- tion to water stress of the leaf water relations of four tropical forage species. Aust J Plant Physiol 7 : 207-220

Wilson JR, Ludlow MM (1983) Time trends of solute accumulation and the influence of potassium fertilizer on osmotic adjustment of water-stressed leaves of three tropical grasses. Aust J Plant Physiol 10: 523-537

Received March 15, 1989 / Accepted August 30, 1989

leaf water relations characteristics of lupinus angustifolius and l. cosentinii

Documents