Directory UMM :Data Elmu:jurnal:P:PlantScience:PlantScience_Elsevier:Vol159.Issue2.2000:

In vitro selection and characterisation of a drought tolerant clone

of

Tagetes minuta

M.A.-H. Mohamed

a,b

_{, P.J.C. Harris}

a

_{, J. Henderson}

a,

_*

a_{School of Natural and En6ironmental Sciences,Co6entry Uni6ersity,Priory Street,Co6entry,CV1 5FB,UK} b_{Department of Horticulture,Faculty of Agriculture,Mania Uni6ersity,El-Mania61111,Egypt}

Received 5 July 1999; received in revised form 22 May 2000; accepted 7 July 2000

Abstract

AromaticTagetesplants produce secondary products which have a biological activity against a wide range of micro-organisms, insects and nematodes.Tagetesoils are also used as pharmaceuticals and as flavour components in the food industry. This study aimed to use somaclonal variation to select drought tolerant plants ofTagetes. Cotyledons cultured on MS medium containing 3 mg l−1_{IAA and 10 mg l}−1_{BA (callus growth medium; CGM) with 60 mM mannitol died. Shoot clumps developed on CGM}

for 6 months and then subcultured onto CGM containing 80 mM mannitol also died. Four shoots were regenerated from 72 shoot clumps on1

2MS medium containing 0.5 mg l−

1_{IAA (shoot growth medium; SGM) after culturing on CGM without mannitol}

for 6 months and then on CGM with 60 mM mannitol for 3 months. Twelve shoots developed from 72 shoot clumps on SGM after culture for 9 months on CGM. Significant variations were observed in biomass amongst regenerated clones when cultured on medium containing mannitol. After growth in greenhouse conditions for 2 months, one clone developed from shoot clumps selected on medium with mannitol exhibited a significant tolerance in vitro in medium containing 90 mM mannitol; this medium completely inhibited growth of control plants. This clone had significantly higher proline content and soluble sugars than the non-stress-selected clone when cultured on medium containing 0 or 30 mM mannitol. When tested for drought tolerance (growth at 40% soil field capacity) in the greenhouse for 2 months, this clone showed a significant tolerance compared with other regenerated and control plants and revealed lower water potential, greater accumulated biomass and a higher relative growth rate. © 2000 Elsevier Science Ireland Ltd. All rights reserved.

Keywords:Marigold; Water stress; Somaclonal variation; Regeneration

www.elsevier.com/locate/plantsci

1. Introduction

The importance of Tagetes minuta L. (Mexican marigold) (Asteracea) as a source of volatile oils and thiophenes as well as its practical uses have been reviewed [1 – 3]. Volatile oils of T. minuta

have a biological activity against a wide range of microorganisms and insects. They are also used in perfumes and as flavour components in many food products. The plants have a suppressive effect on

free-living nematodes and have been used as an intercrop or in rotation to protect crops [4].

Water deficit is a major component of environ-mental stresses such as drought, salinity and low temperature, and 40 – 60% of the agricultural land around the world suffers from drought [5,6]. Breeding for water stress tolerance by traditional methods is a time consuming and inefficient proce-dure [7]. In vitro culture may be used to obtain drought-tolerant plants assuming that there is a correlation between cellular and in vivo plant re-sponses [8]. This method is based on the induction of genetic variation among cells, tissues and/or organs in cultured and regenerated plants. Al-though there are genetic, biochemical and physio-logical constraints in obtaining stress-tolerant

Abbre6iations:BA, 6-benzyladenine; CGM, callus growth medium;

FC, field capacity; IAA, indole-3-acetic acid; RGR, relative growth rate; SGM, shoot growth medium.

* Corresponding author. Tel.: +44-24-76888632; fax: + 44-24-76888702.

E-mail address:janey.h@coventry.ac.uk (J. Henderson).

0168-9452/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 8 - 9 4 5 2 ( 0 0 ) 0 0 3 3 9 - 3

plants through in vitro culture, Nobars [6] pointed out that this technique has been successfully used to produce stress-tolerant plants from several species.

The accumulation of proline and soluble sugars as an osmotic tolerance mechanism has been widely observed in many species. Proline may provide osmoregulation and stabilisation of proteins and membrane during stress [9]. Soluble sugars, which accumulate in the vacuole, are a major organic solute, involved in osmotic adjust-ment when plants are exposed to drought [9].

Sucrose, mannitol or sorbitol were studied by Sabili et al. [10] as osmotic stress agents on in vitro-grown Chrysanthemum morifolium. Sucrose failed to elicit consistent osmotic stress symptoms and enhanced both shoots and root growth. The osmotic potential of the tissue paralleled the in-crease of mannitol or sorbitol concentrations on culture medium. The harmful effect of the osmotic stress agent occurred in both shoot proliferation and rooting stages. Sorghum bicolor was tested in vitro for drought stress by Dunca et al. [11]. Results showed that osmotic stress applied to in vitro cultures reduced regeneration ability. How-ever, by screening the regenerated plants under field conditions they obtained stress-tolerant re-generates with higher yield under stress conditions than that of their parents.

Regeneration of plants displaying an increased tolerance to environmental stress is an important goal for the biotechnological improvement of many plant species [6]. Therefore, the aim of this investigation was to induce somaclonal variation in regenerated plants in order to select drought-tolerant clones of T. minuta, which grows as a cash crop in some developing countries [2,3], using a rapid in vitro regeneration protocol [12].

2. Materials and methods

2.1. General materials and methods

Sterile seedlings and cotyledon-derived shoot clumps were obtained and maintained as reported previously [12]. Unless stated, 42 explants from each treatment were cultured in seven Phytatrays II (Sigma) with 100 ml medium. Shoot and root fresh and dry weights for plants developed in vitro were measured for 10 randomly selected plantlets

after gently removing the medium from the roots. The fresh weights for individual shoots and roots were assessed immediately after blotting the plantlets on tissue, and the dry weights determined after drying to a constant weight in an oven at 80°C and cooling to room temperature in a desic-cator. Plants were adapted to greenhouse condi-tions of 25 – 31°C and 16-h photoperiod, as reported previously [12]. The significance of differ-ences in shoot, leaf, stem, plant fresh and dry weights, proline and soluble sugars content and water potential were tested by Analysis of Vari-ance using Minitab for Windows Version 10.5 computer programme.

2.2. Optimising mannitol concentration for osmotic stress

This preliminary experiment was carried out to optimise mannitol concentrations which could then be used as a selective agent for osmotic stress. Shoot tips of 2-week-old sterile seedlings were cultured on 1

2-strength MS medium [13] containing

2% sucrose, 0.5 mg l−1 _{IAA (shoot growth}

medium; SGM) with 0, 20, 40, 60, 80 or 100 mM mannitol. After 4 weeks, the number of segments which developed plantlets and the fresh and dry weight of both shoots and roots were measured. In order to study the ability of cotyledon ex-plants to regenerate shoots on medium containing mannitol, 40 cotyledons from 1-week-old sterile seedlings were cultured onto eight Petri dishes each with 25 ml MS medium containing 3% su-crose, 3 mg l−1_{IAA, 10 mg l}−1_{BA (callus growth}

medium; CGM) with 0 or 60 mM mannitol. After 1 month, the developing shoot clumps were trans-ferred onto SGM to enhance adventitious shoot growth.

2.3. Shoot regeneration from cotyledon explants on medium containing mannitol

Cotyledons were cultured on CGM as stated previously. The developing shoot clumps were subcultured onto CGM six times at monthly inter-vals (Fig. 1, step 1). Following that, ca. 2-g pieces of shoot clumps were subcultured onto fresh medium containing 0, 60 or 80 mM mannitol for 3 months at monthly intervals (Fig. 1, step 2).

Pieces of shoot clumps ca. 2 g, previously cul-tured on CGM and CGM with 60 mM mannitol

were subcultured onto SGM containing 0 or 60 mM mannitol, respectively. In this trial, 72 shoot clumps were cultured on 12 Phytatrays II for each treatment. Shoot clumps were subcultured twice onto a fresh medium at monthly intervals (Fig. 1, step 3). Developing shoots were micropropagated as separate clones by single node culture on SGM for 2 months (Fig. 1, step 4). Plants were hardened off under greenhouse conditions (Fig. 1, step 5). For control plants, seeds were sown in the greenhouse. After 2 months, shoot tips from each of 30 plants was excised. Segments were surface-sterilised with 10% sodium hypochlorite solution (BDH) with a few drops of Tween 20 for 30 min, followed by several washes with sterile distilled water. Nodes were cultured in SGM for 1 month before transferring to the greenhouse (Fig. 1, step 4).

2.4. In 6itro screening of clones for drought stress

After growing the clones for 2 months in the greenhouse, segments with the uppermost two identifiable nodes and the shoot tips were excised and surface-sterilized. Single-node segments were

cultured on SGM with 0 or 60 mM mannitol and 500 mg l−1 _{carbenicillin, to eliminate bacterial}

contamination (Fig. 1, step 6). After 1 month the percentage of nodes that developed into plantlets was assessed and shoot and root fresh and dry weights were measured. The remaining plantlets that developed on medium without mannitol were returned to the greenhouse (Fig. 1, step 7).

Clones were grown in the greenhouse for 2 months before retesting for osmotic stress in vitro. Two intermediate clones developed in non-stress conditions and two clones including a drought-tol-erant clone and non-toldrought-tol-erant clone which devel-oped from callus cultured on mannitol-containing medium were tested for osmotic stress in vitro. Control plants (Fig. 1, step 4) were tested in the same way as the clones. First and second upper-most nodes were surface-sterilised and cultured on SGM containing 500 mg l−1 _{carbenicillin and 0,}

30, 60 or 90 mM mannitol (Fig. 1, step 8). After 1 month, shoot and root fresh and dry weights were assessed. The remaining plants grown on manni-tol-free medium were micropropagated by shoot tips for 1 month and then transferred to the greenhouse.

2.5. Proline and soluble sugars determination

Plants were grown in the greenhouse for 2 months then, first and second uppermost nodes of stress-selected clone, PM3, and non-stress-selected clone, P4, were cultured as in Section 2.4 on SGM containing 0 or 30 mM mannitol. After 1 month the new developing shoots were used to measure the proline content and soluble sugars. Samples of 1 g from each of 15 individual plantlets were used to measure proline content as described by Bates et al. [14]. For soluble sugars, 15 individual plantlets from of the each clones were dried at 80°C, and 0.1 g from each individual plantlet was used to measure soluble sugars following the method reported by Jermyn [15].

2.6. In 6i6o screening of clones for drought stress

After 2 months of growth in the greenhouse, shoot tip segments of two clones, P4 selected on mannitol-free medium and PM3 selected on man-nitol-containing medium, and control plants were cultured in vitro for a further month before being returned to the greenhouse. Two weeks later

Fig. 1. Protocol designed to select drought-tolerant clones of

T. minuta. CGM, callus growth medium=MS+3% su-crose+10 mg l−1_{BA+3 mg l}−1_{IAA; SGM, shoot growth}

medium=1

2MS+2% sucrose+0.5 mg l

Fig. 2. Effect of mannitol concentrations on shoot fresh (A) and dry (B) and root fresh (C) and dry (D) weights on plants developed from shoot tip segments ofT.minutaafter 1 month of culture. The bars show LSD for all pair comparisons at

P=0.05.

Drought stressed plants were watered to 40% FC. After 2 months, plant heights were measured and then plants were cut at the soil surface. WP was measured for individual plants using 10-cm long shoot tips immediately after cutting. Individual leaf and stem fresh weights were assessed immedi-ately and the dry weights were assessed after air drying for 2 weeks. The relative growth rates (RGR) for plant fresh and dry weights were calcu-lated by the formula:

RGR (g g−1 _day−1₎

=logeW 2−logeW1/T2−T1

whereWrepresent weight andTindicates harvest-ing time [18].

3. Results

3.1. Optimising mannitol concentration for osmotic stress

All segments cultured on medium with mannitol remained green, even those that did not show any further growth. Fig. 2 shows that mannitol had an affect on plantlets grown from cultured segments. On medium with 80 and 100 mM mannitol some segments developed only shoots that were too small to measure. Only 68, 45, and 33% of shoot tips cultured on medium containing 40, 60 and 80 mM mannitol, respectively, developed into plantlets. Plantlets developing on medium contain-ing 60 mM mannitol had shoot fresh and dry weights of 53 and 52%, and root fresh and dry weights of 71 and 59%, respectively, of those of plantlets developed on mannitol-free medium. The equivalent values for plantlets grown on medium with 80 mM mannitol were 30, 38, 55 and 42%. All cotyledons cultured in CGM containing 60 mM mannitol became swollen and necrotic and did not survive. However, cotyledons cultured in CGM without mannitol increased in size after 1 week in culture and callus was initiated from the basal end before developing adventitious shoots. When 1-month-old shoot clumps were transferred from CGM to SGM for a month, an average of 20 shoots explant−1 _{was regenerated. Media with 60}

and 80 mM mannitol were therefore used to apply osmotic stress in vitro and to assess the drought stress of micropropagated plants in vitro.

plants were transplanted into 12 cm diameter plas-tic pots filled with 650 g air-dried clay soil. One week later, 10 plants were randomly selected from each clone and control plants. Individual plant heights were assessed and water potential (WP) measured using the pressure-chamber technique [16]. Individual shoot fresh weights were assessed immediately. Shoot dry weights were measured after air drying for 2 weeks.

Plants were arranged in a randomised block design with three blocks, each block with 18 plants from each clone and the control plants. Within each block, half of the plants of each type were well-watered to 100% field capacity (FC) and half exposed to drought conditions at 40% FC. Plants were watered every 3 days. Field capacity of the soil in pots was assessed before the treatment as reported by Tuomela [17]. Plants grown in soil with 40% FC were irrigated for the first time 6 days after starting water stress treatment. Before watering, three pots from each treatment in each block were randomly selected and weighed. The amount of water required to restore the soil to 100% FC was added to the non-stressed group.

3.2. Shoot regeneration from cotyledon explants on medium containing mannitol

After culture for 6 months on CGM without mannitol to encourage somaclonal variation, shoot clumps were transferred to CGM containing 0, 60 or 80 mM mannitol (Fig. 1, step 2). Two weeks after transfer of shoot clumps onto CGM containing 60 mM mannitol, vitrified swollen shoots developed but died at this stage although the differentiated clumps remained green. No shoots elongated from shoot clumps cultured on the CGM. After a month, a little decrease in growth was observed but all shoot clumps cultured in medium containing 60 or 80 mM mannitol survived with some necrotic tissue forming espe-cially on medium containing 80 mM mannitol. After subculturing onto fresh CGM for another month, shoot clumps cultured in medium contain-ing 60 mM mannitol were healthy but 80% of shoot clumps cultured on medium containing 80 mM mannitol became necrotic and died. In the second subculture, all shoot clumps cultured onto medium containing 80 mM mannitol died. After 3 months, 36 and 12 shoot clumps out of 72 clumps cultured on CGM containing 0 and 60 mM man-nitol, respectively, survived. These surviving shoot clumps were transferred onto SGM with 0 or 60 mM mannitol for shoot clumps previously cul-tured on CGM with 0 or 60 mM mannitol, respec-tively, to stimulate shoot growth (Fig. 1, step 3). Some of the shoot clumps cultured onto SGM became necrotic and stopped growth regardless of the medium used to maintain the shoot clumps during the previous 9 months. Many of the shoots developed from shoot clumps cultured on medium containing mannitol died at an early stage after reaching approximately 0.5 cm in length. The number of shoots developed from clumps at this stage was 12 from five shoot clumps in mannitol-free medium and four shoots from two shoot clumps in medium containing mannitol. Shoots developed on mannitol-free medium were named as P1 – P12, and those developed in stressed medium named as PM1 – PM4. Two shoots from those developed from non-stressed shoot clumps were discarded because of contamination.

Developing shoots were excised and microprop-agated as clones on SGM (Fig. 1, step 4). Shoot tip segments continued growth as a main stem but the two axillary buds on nodal segments developed

into shoots and all axillary buds on developing shoots sprouted. After 2 months of micropropaga-tion all shoots became weak, developing small narrow leaves and long nodes. To achieve normal growth, plants were transferred in vivo to a green-house (Fig. 1, step 5). All transferred plantlets survived and developed normal shoots within a month.

3.3. In 6itro screening of clones for drought tolerance

Nodes from greenhouse-grown plants were ex-cised from each clone and cultured on SGM con-taining 0 or 60 mM mannitol (Fig. 1, step 6). The percentage of surviving explants which developed new shoots varied from 75 to 100% in mannitol-free medium and from 30 to 100% in medium containing mannitol. However, although the per-centage survival of segments on medium contain-ing mannitol for clones P2, P6, P8, PM1 and PM2 was 100%, the developing shoots and roots were too small to obtain fresh/dry weight measurements (Fig. 3).

There were significant differences among clones in both shoot fresh and dry weights and in the response of these parameters to mannitol (Figs. 4 and 5). There were also significant differences among clones in root fresh and dry weights (data not presented). Plantlets from clone PM3 cultured on medium containing mannitol gave shoot fresh and dry weights of 298 and 27 mg−1_{, respectively.}

These weights did not differ from those for plantlets grown on mannitol-free medium.

Plantlets from four of the clones developed on medium without mannitol were transferred to the greenhouse for 2 months (Fig. 1, step 7). These clones and control plants were tested for drought tolerance by culturing nodal explants on SGM containing 0, 30, 60 or 90 mM mannitol (Fig. 1, step 8). Percentage survival was not a good indica-tor of osmotic stress as most cultured nodes, even in medium containing the highest mannitol con-centration, remained green while showing no fur-ther growth. Fig. 6 shows that shoot fresh and dry weights were significantly decreased by all manni-tol concentrations for all plants except for PM3 clone, which previously showed a high tolerance to drought. There was no difference in biomass when the drought-tolerant clone PM3 was grown on medium containing 0 or 30 mM mannitol. Plants

of two clones, one from the non-stress selected shoot clumps and one from stress-selected shoot clumps were very sensitive to osmotic stress based on shoot fresh and dry weight measurements. A drought-sensitive regenerated clone P4, from non-stress-selected shoot clumps, the drought tolerant clone PM3 which developed from stress-selected shoot clumps, and the control plants grown on SGM without mannitol, were all cloned from shoot tips and returned to the greenhouse to test their response to drought stress in vivo.

3.4. Proline and soluble sugars content

Neither the proline content nor soluble sugars were significantly increased in either clones when 30 mM mannitol was added to the growth medium (P=0.11). Overall plantlets of clone PM3 exhibited a significantly higher proline content (PB0.001) than plantlets of P4, but a significant

difference between clones could not be shown for the 0 and 30 mM mannitol individually. Overall PM3 plantlets grown in medium containing 0 or

Fig. 3.T.minutanodal segments of (from left to right) PM3 on SGM, PM3 on SGM containing 60 mM mannitol and P10, P7, P3 and PM1 on SGM containing 60 mM mannitol, after 1 month in culture. The scale is in centimeters.

Fig. 4. Shoot fresh weight for T. minutaclones after 1 month of growth on medium containing mannitol. Clones have been arranged in descending order by weight of plants grown on medium with mannitol as a percentage of control plant weight. These values are shown for each clone. The bar shows the LSD for all pair comparisons atP=0.05. Clones P1 – P10 were developed on mannitol-free CGM and clones PM1 – PM4 were developed on CGM with mannitol.

Fig. 5. Shoot dry weights for T. minuta clones after 1 month of growth on medium containing mannitol. Clones have been arranged in descending order by weight of plants grown on medium with mannitol as a percentage of control plant weight. These values are shown for each clone. The bar shows the LSD for all pair comparisons atP=0.05. Clones P1 – P10 were developed on mannitol-free CGM and clones PM1 – PM4 were developed on CGM with mannitol.

30 mM mannitol had significantly higher soluble sugars content than P4 plantlets (Fig. 7).

3.5. In 6i6o testing of clones for drought tolerance

When plants were exposed to the drought stress in the greenhouse the bottom leaves of the stressed plant, especially the control and the non-stress-se-lected clone P4, started to dry out after a month of drought stress. Before starting drought stress, con-trol, non-stress-selected clone P4 and the stress-se-lected clone PM3 had a WP of −0.121 to −0.125 MPa, with no significant difference among them. Two months after drought stress, plants of the stress-selected clone PM3 had a WP of −0.178 MPa, significantly lower than the WP of plants of the non-stress-selected clone P4 or the control plants grown in soil at 40% FC. In soil at 100% FC there was no significant difference between control, non-stress selected P4 and stress-selected PM3 plants (Table 1).

Plant height and plant fresh and dry weights did not differ significantly between the clones or be-tween the clones and the control before water stress was applied or after 2 months growth in soil at 100% FC. Plant height and fresh and dry weights were significantly reduced by drought (PB0.001 for all

of these parameters). Plants of the stress-selected clone PM3 grown in soil with 40% FC had a plant height and plant fresh and dry weight significantly greater than that of control or non-stress selected P4 plants (Table 1). Fresh and dry weights of stress-selected plants PM3 grown in soil at 40% FC

were 38 and 41%, respectively of the unstressed control plants, whereas the equivalent values for control plants were 19 and 22%, and for the non-stress-selected clone P4 were 22 and 19%.

Fig. 6. Shoot fresh (A) and dry (B) weights for T. minuta

clones after 1 month of growth on medium containing manni-tol. The bar shows the LSD for all pair comparisons at

P=0.05. Missing bars represent zero value. Clones P1 and P4 were developed on mannitol-free CGM and clones PM2 and PM4 were developed on CGM containing mannitol.

Fig. 7. Proline (A) and soluble sugars content (B) ofT.minuta

clones after 1 month of growth on medium containing 0 or 30 mM mannitol. The bar shows the LSD for all pair compari-sons atP=0.05.

stress-selected clone PM3 had RGR (fresh wt.) which was significantly higher than for the non-stress-selected P4 clone (152%) and control plants (142%). Also, stress-selected plants had RGR (dry wt.) which was significantly higher than for P4 (129%) and control plants (143%) (Table 1).

4. Discussion

The data show an inhibitory effect of mannitol on shoot fresh and dry weights. Also, cotyledons cultured on medium containing 60 mM mannitol showed no callus or shoot development. Mannitol has an inhibitory effect on plant growth by lower-ing the water potential of the medium, so cultured explants are unable to take up water and nutrients from the medium. Lipavska and Vreugdenhil [19] reported that the decrease in osmotic potential of the culture medium by mannitol caused a decrease in the dry matter accumulation of wheat embryos, rape seedlings and potato stem segments grown in vitro.

Shoot clumps cultured on medium containing mannitol became necrotic and died after 3 months of culture on medium containing 80 mM manni-tol. A reduction in callus growth due to mannitol has been reported previously in Brassica culture [20]. After 1 month of culture, 20 adventitious shoots developed from one shoot clump whereas after 9 months of culture on mannitol-free medium, only 12 shoots developed from a total of 72 shoot clumps. Thus, shoot clumps cultured on Relative growth rates (RGR) of 0.056 to 0.057 g

g−1 _day−1 _{(fresh wt.) and 0.06 to 0.07 g g}−1

day−1 _{(dry wt.) were calculated for the plants in}

soil with 100% FC, with no significant difference between clones and control plants. Significant dif-ferences between the different clones and between clones and control plants grown in soil with 40% FC were recorded (PB0.001). Plants from the

Table 1

Effect of drought stress on water potential and growth characteristics of regenerated clones and control plants ofT.minutaafter 2 months of drought stressa

RGR–DW Plant DWe

Plant FCb_{(%) WP}c _{(MPa) Plant height Plant FW}d _RGRf_–FW

(g g−1_day−1₎

(g g−1_day−1₎

(g plant−1₎

(cm) (g plant−1₎

0.042 0.028

Control 40 −0.142 27 6.8 1.6

7.3 0.056

100 −0.114 61 36.2 0.068

0.041

1.5 0.029

P4 40 −0.141 27 8.2

37.8 7.7 0.057

Clone 100 −0.112 65 0.068

0.053 0.040

3.0

PM3 40 −0.178 32 14.5

0.067 0.056

Clone 100 −0.121 61 37.8 7.2

0.002 0.003

LSD 0.05 0.013 4.5 3.8 0.74

a_{Each value is the mean of 24 measurements.} b_{Field capacity.}

c_{Water potential.} d_{Fresh weight.} e_{Dry weight.}

medium without osmotic stress for 9 months lost some of their regeneration ability. Regeneration ability is commonly decreased by repeated subcul-ture of tissue in many plants. The reasons for loss of regenerative ability are not clearly understood and may be due to somaclonal variations, loss of a substance promoting the regeneration of fresh tis-sue, or epigenetic changes [21].

After growing clones without stress for 3 months, a significant change in the effect of mannitol on plant biomass was recorded when they were recul-tured in vitro. Some clones continued to grow in the presence of mannitol but others lost their potential to grow. This suggests that some clones had only a physiological adaptation to osmotic stress and that the plants lost their tolerance when the stress factor was removed from the medium. Similar physiological adaptations to stress conditions had been reported by Nabors [6].

Clones developed from non-stressed callus showed significant variations in growth when ex-posed to osmotic stress. Somaclonal variation among in vitro regenerated plants has been reported by Larkin and Scowcroft [22]. From fourT.minuta

clones regenerated under water stress, one clone exhibited greater shoot and root fresh and dry weights on medium containing mannitol than on mannitol-free medium.

Both the drought-tolerant-selected clone PM3 and the non-stress-selected clone P4 grown on medium with 30 mM mannitol had higher proline content and soluble sugars, but not significantly so, than the same plants grown on mannitol-free medium. The accumulation of proline and soluble sugars in osmotically stressed plants of neither clone appears insufficient to relieve water stress. Hanson and Hitz [23] reported that, in many plants during moderate water stress, there was a 10- to 100-fold increase of free proline in leaf tissue. Alian et al. [24] showed that there was no correlation between the accumulation of proline and drought stress toler-ance in tomato. Therefore, the drought-tolerant-se-lected clone of T. minuta in this study may have another strategy to tolerate water stress.

Results suggested that the drought-tolerant-se-lected clone had a higher capacity to maintain membrane stability and/or low water potential and greater growth when grown on medium with low water potential. When this clone was tested in the greenhouse for drought, it yielded a higher biomass than other clones. This indicates that applying

water stress during callus growth was efficient in selecting a drought tolerant plant. Hasissou and Bouharmont [25] reported similar findings for the culture of Triticum durum where, out of 30 plants surviving water stress, 13 exhibited tolerance to drought compared with unselected controls.

In conclusion, one clone of T. minuta selected under drought appeared to show drought tolerance and had higher proline content and soluble sugars than the non-stress-selected clone even after growth in non-stress conditions for 6 months. Also, in vivo results showed that it had a higher yield and RGR under water stress than other regenerated clones or plants grown from seeds. Under non-stress condi-tions, the stress-selected clone exhibited similar growth characteristics to other tested plants. Fur-ther studies are required to investigate the mecha-nism in drought tolerance of the regenerated clones and the inheritance of this trait through sexual reproduction.

Acknowledgements

This investigation was financially supported by The Ministry of Higher Education, Egypt.

References

[1] R.T. Neher, The ethnobotany ofTagetes, Econ. Bot. 22 (1968) 317 – 325.

[2] J.A. Soule,Tagetes minuta: A potential new herb from South America, in: New Crops Proceeding of the Second National Symposium New Crop Exploration. Research & Commercial Utilisation, Indiana 1991, Wiley, New York, 1993, pp. 649 – 654.

[3] P. Vasudevan, K. Suman, S. Satyawati, Tagetes: A multipurpose plant, Bioresour. Technol. 7 (1997) 29 – 35. [4] J. Kimpinski, W.J. Arsenault, Nematodes in annual ryegrass, marigold, mustard, red clover and soybean, Forage Notes 37 (1994) 52 – 53.

[5] E.A. Bray, Plant response to water deficit, Trends Plant Sci. 2 (1997) 48 – 54.

[6] M.W. Nabors, Environmental stress resistance proce-dure and applications, in: J.D. Philip (Ed.), Plant Cell Line Selection, VCH, Weinheim, 1990, pp. 167 – 185. [7] K. Dorffling, H. Dorffling, G. Lesselich, In6itro

selec-tion and regeneraselec-tion of hydroxyproline-resistant lines of winter wheat with increased proline content and increased frost tolerance, Plant Physiol. 142 (1993) 222 – 225.

[8] R.H. Smith, S. Bhaskaran, F.R. Miller, Screening for drought tolerance in sorghum using cell culture, In Vitro Cell Dev. Biol. 21 (1985) 541 – 545.

[9] H. Thomas, Drought resistance in plants, In: S.D. Amarjit, K.B. Ranjit (Eds.), Mechanisms of Environ-mental Stress Resistance in Plants, Harwood, 1997, pp. 1 – 42.

[10] R.A. Shibli, M.A.L. Smith, L.A. Spomer, Osmotic ad-justment and growth response of three Chrysanthemum morifoliumRamat. cultivars to osmotic stress inducedin 6itro, Plant Nutr. 15 (1992) 1373 – 1381.

[11] R.R. Duncan, R.M. Waskom, M.W. Nabors, In 6itro

screening and field evaluation of tissue-culture-regener-ated sorghum (Sorghum bicolor L. Moench) for soil stress tolerance, Euphytica 85 (1995) 373 – 380.

[12] M.A.-H. Mohamed, P.J.C. Harris, J. Henderson, An efficient in 6itro regeneration protocol for Tagetes minuta, Plant Cell Tissue Organ Cult. 55 (1999) 211 – 215.

[13] T. Murashige, F. Skoog, A revised medium for rapid growth and bioassays with tobacco tissue cultures, Phys-iol. Plant. 15 (1962) 473 – 497.

[14] L.S. Bates, R.P. Waldern, I.D. Teare, Rapid determina-tion of free proline from water stress studies, Plant Soil 39 (1973) 205 – 207.

[15] M.A. Jermyn, Increasing the sensitivity of the anthron method for carbohydrate, Anal. Biochem. 68 (1975) 332 – 335.

[16] P.F. Scholander, H.T. Hammel, E.D. Bradstreet, E.A. Hemmingsen, Sap pressure in vascular plants: Negative hydrostatic pressure can be measured in plants, Science 148 (1965) 339 – 3416.

[17] K. Tuomela, Leaf water relations in six provenances of

Eucalyptus microtheca: a greenhouse experiment, Forest Ecol. Manage. 92 (1997) 1 – 10.

[18] R. Hunt, Basic Growth Analysis: Plant Growth Analysis for the Beginners, Unwin Hyman, London, 1990. [19] H. Lipavska, D. Vreugdenhil, Uptake of mannitol from

the medium by in 6itrogrown plants, Plant Cell Tissue

Organ Cult. 45 (1996) 103 – 107.

[20] G. Gangopadhyay, S. Basu, S. Gupta, In6itroselection

and physiological characterization of NaCl- and manni-tol-adapted callus lines in Brassica juncea, Plant Cell Tissue Organ Cult. 50 (1997) 161 – 169.

[21] E.F. George, Plant Propagation by Tissue Culture. Part 1: The Technology, 2nd edn, Exegetic, Somerset, UK, 1993.

[22] P.J. Larkin, W.R. Scowcroft, Somaclonal variation — a novel source of variability from cell cultures for plant improvement, Theor. Appl. Genet. 60 (1981) 197 – 214. [23] D. Hasissou, J. Bouharmont, In 6itro selection and

characterization of drought-tolerant plants of durum-wheat (Triticum durum desf), Agronomy 14 (1994) 65 – 70.

[24] A. Alian, A. Altman, B. Heuer, Genotypic difference in salinity and water stress tolerance of fresh market tomato cultivars, Plant Sci. 152 (2000) 59 – 65.

[25] A.D. Hanson, W.D. Hitz, Metabolic responses of meso-phytes to plant water deficits, Annu. Rev. Plant Physiol. 3 (1982) 163 – 203.

3.2. Shoot regeneration from cotyledon explants on medium containing mannitol

After culture for 6 months on CGM without mannitol to encourage somaclonal variation, shoot clumps were transferred to CGM containing 0, 60 or 80 mM mannitol (Fig. 1, step 2). Two weeks after transfer of shoot clumps onto CGM containing 60 mM mannitol, vitrified swollen shoots developed but died at this stage although the differentiated clumps remained green. No shoots elongated from shoot clumps cultured on the CGM. After a month, a little decrease in growth was observed but all shoot clumps cultured in medium containing 60 or 80 mM mannitol survived with some necrotic tissue forming espe-cially on medium containing 80 mM mannitol. After subculturing onto fresh CGM for another month, shoot clumps cultured in medium contain-ing 60 mM mannitol were healthy but 80% of shoot clumps cultured on medium containing 80 mM mannitol became necrotic and died. In the second subculture, all shoot clumps cultured onto medium containing 80 mM mannitol died. After 3 months, 36 and 12 shoot clumps out of 72 clumps cultured on CGM containing 0 and 60 mM man-nitol, respectively, survived. These surviving shoot clumps were transferred onto SGM with 0 or 60 mM mannitol for shoot clumps previously cul-tured on CGM with 0 or 60 mM mannitol, respec-tively, to stimulate shoot growth (Fig. 1, step 3). Some of the shoot clumps cultured onto SGM became necrotic and stopped growth regardless of the medium used to maintain the shoot clumps during the previous 9 months. Many of the shoots developed from shoot clumps cultured on medium containing mannitol died at an early stage after reaching approximately 0.5 cm in length. The number of shoots developed from clumps at this stage was 12 from five shoot clumps in mannitol-free medium and four shoots from two shoot clumps in medium containing mannitol. Shoots developed on mannitol-free medium were named as P1 – P12, and those developed in stressed medium named as PM1 – PM4. Two shoots from those developed from non-stressed shoot clumps were discarded because of contamination.

Developing shoots were excised and microprop-agated as clones on SGM (Fig. 1, step 4). Shoot tip segments continued growth as a main stem but the two axillary buds on nodal segments developed

into shoots and all axillary buds on developing shoots sprouted. After 2 months of micropropaga-tion all shoots became weak, developing small narrow leaves and long nodes. To achieve normal growth, plants were transferred in vivo to a green-house (Fig. 1, step 5). All transferred plantlets survived and developed normal shoots within a month.

3.3. In 6itro screening of clones for drought tolerance

Nodes from greenhouse-grown plants were ex-cised from each clone and cultured on SGM con-taining 0 or 60 mM mannitol (Fig. 1, step 6). The percentage of surviving explants which developed new shoots varied from 75 to 100% in mannitol-free medium and from 30 to 100% in medium containing mannitol. However, although the per-centage survival of segments on medium contain-ing mannitol for clones P2, P6, P8, PM1 and PM2 was 100%, the developing shoots and roots were too small to obtain fresh/dry weight measurements (Fig. 3).

There were significant differences among clones in both shoot fresh and dry weights and in the response of these parameters to mannitol (Figs. 4 and 5). There were also significant differences among clones in root fresh and dry weights (data not presented). Plantlets from clone PM3 cultured on medium containing mannitol gave shoot fresh and dry weights of 298 and 27 mg−1_{, respectively.} These weights did not differ from those for plantlets grown on mannitol-free medium.

Plantlets from four of the clones developed on medium without mannitol were transferred to the greenhouse for 2 months (Fig. 1, step 7). These clones and control plants were tested for drought tolerance by culturing nodal explants on SGM containing 0, 30, 60 or 90 mM mannitol (Fig. 1, step 8). Percentage survival was not a good indica-tor of osmotic stress as most cultured nodes, even in medium containing the highest mannitol con-centration, remained green while showing no fur-ther growth. Fig. 6 shows that shoot fresh and dry weights were significantly decreased by all manni-tol concentrations for all plants except for PM3 clone, which previously showed a high tolerance to drought. There was no difference in biomass when the drought-tolerant clone PM3 was grown on medium containing 0 or 30 mM mannitol. Plants

of two clones, one from the non-stress selected shoot clumps and one from stress-selected shoot clumps were very sensitive to osmotic stress based on shoot fresh and dry weight measurements. A drought-sensitive regenerated clone P4, from non-stress-selected shoot clumps, the drought tolerant clone PM3 which developed from stress-selected shoot clumps, and the control plants grown on SGM without mannitol, were all cloned from shoot tips and returned to the greenhouse to test their response to drought stress in vivo.

3.4. Proline and soluble sugars content

Neither the proline content nor soluble sugars were significantly increased in either clones when 30 mM mannitol was added to the growth medium (P=0.11). Overall plantlets of clone PM3 exhibited a significantly higher proline content (PB0.001) than plantlets of P4, but a significant difference between clones could not be shown for the 0 and 30 mM mannitol individually. Overall PM3 plantlets grown in medium containing 0 or

Fig. 3.T.minutanodal segments of (from left to right) PM3 on SGM, PM3 on SGM containing 60 mM mannitol and P10, P7,

P3 and PM1 on SGM containing 60 mM mannitol, after 1 month in culture. The scale is in centimeters.

Fig. 4. Shoot fresh weight for T. minutaclones after 1 month of growth on medium containing mannitol. Clones have been

arranged in descending order by weight of plants grown on medium with mannitol as a percentage of control plant weight. These values are shown for each clone. The bar shows the LSD for all pair comparisons atP=0.05. Clones P1 – P10 were developed on mannitol-free CGM and clones PM1 – PM4 were developed on CGM with mannitol.

Fig. 5. Shoot dry weights for T. minuta clones after 1 month of growth on medium containing mannitol. Clones have been arranged in descending order by weight of plants grown on medium with mannitol as a percentage of control plant weight. These values are shown for each clone. The bar shows the LSD for all pair comparisons atP=0.05. Clones P1 – P10 were developed on mannitol-free CGM and clones PM1 – PM4 were developed on CGM with mannitol.

30 mM mannitol had significantly higher soluble sugars content than P4 plantlets (Fig. 7).

3.5. In 6i6o testing of clones for drought tolerance When plants were exposed to the drought stress in the greenhouse the bottom leaves of the stressed plant, especially the control and the non-stress-se-lected clone P4, started to dry out after a month of drought stress. Before starting drought stress, con-trol, non-stress-selected clone P4 and the stress-se-lected clone PM3 had a WP of −0.121 to −0.125 MPa, with no significant difference among them. Two months after drought stress, plants of the stress-selected clone PM3 had a WP of −0.178 MPa, significantly lower than the WP of plants of the non-stress-selected clone P4 or the control plants grown in soil at 40% FC. In soil at 100% FC there was no significant difference between control, non-stress selected P4 and stress-selected PM3 plants (Table 1).

Plant height and plant fresh and dry weights did not differ significantly between the clones or be-tween the clones and the control before water stress was applied or after 2 months growth in soil at 100% FC. Plant height and fresh and dry weights were significantly reduced by drought (PB0.001 for all of these parameters). Plants of the stress-selected clone PM3 grown in soil with 40% FC had a plant height and plant fresh and dry weight significantly greater than that of control or non-stress selected P4 plants (Table 1). Fresh and dry weights of stress-selected plants PM3 grown in soil at 40% FC

were 38 and 41%, respectively of the unstressed control plants, whereas the equivalent values for control plants were 19 and 22%, and for the non-stress-selected clone P4 were 22 and 19%.

Fig. 6. Shoot fresh (A) and dry (B) weights for T. minuta

clones after 1 month of growth on medium containing manni-tol. The bar shows the LSD for all pair comparisons at

P=0.05. Missing bars represent zero value. Clones P1 and P4 were developed on mannitol-free CGM and clones PM2 and PM4 were developed on CGM containing mannitol.

Fig. 7. Proline (A) and soluble sugars content (B) ofT.minuta

clones after 1 month of growth on medium containing 0 or 30 mM mannitol. The bar shows the LSD for all pair compari-sons atP=0.05.

stress-selected clone PM3 had RGR (fresh wt.) which was significantly higher than for the non-stress-selected P4 clone (152%) and control plants (142%). Also, stress-selected plants had RGR (dry wt.) which was significantly higher than for P4 (129%) and control plants (143%) (Table 1).

4. Discussion

The data show an inhibitory effect of mannitol on shoot fresh and dry weights. Also, cotyledons cultured on medium containing 60 mM mannitol showed no callus or shoot development. Mannitol has an inhibitory effect on plant growth by lower-ing the water potential of the medium, so cultured explants are unable to take up water and nutrients from the medium. Lipavska and Vreugdenhil [19] reported that the decrease in osmotic potential of the culture medium by mannitol caused a decrease in the dry matter accumulation of wheat embryos, rape seedlings and potato stem segments grown in vitro.

Shoot clumps cultured on medium containing mannitol became necrotic and died after 3 months of culture on medium containing 80 mM manni-tol. A reduction in callus growth due to mannitol has been reported previously in Brassica culture [20]. After 1 month of culture, 20 adventitious shoots developed from one shoot clump whereas after 9 months of culture on mannitol-free medium, only 12 shoots developed from a total of 72 shoot clumps. Thus, shoot clumps cultured on Relative growth rates (RGR) of 0.056 to 0.057 g

g−1 _day−1 _{(fresh wt.) and 0.06 to 0.07 g g}−1 day−1 _{(dry wt.) were calculated for the plants in} soil with 100% FC, with no significant difference between clones and control plants. Significant dif-ferences between the different clones and between clones and control plants grown in soil with 40% FC were recorded (PB0.001). Plants from the Table 1

Effect of drought stress on water potential and growth characteristics of regenerated clones and control plants ofT.minutaafter 2 months of drought stressa

RGR–DW

Plant DWe

Plant FCb_{(%) WP}c _{(MPa) Plant height Plant FW}d _RGRf_–FW

(g g−1_day−1₎

(g g−1_day−1₎

(g plant−1₎

(cm) (g plant−1₎

0.042 0.028

Control 40 −0.142 27 6.8 1.6

7.3 0.056

100 −0.114 61 36.2 0.068

0.041

1.5 0.029

P4 40 −0.141 27 8.2

37.8 7.7 0.057

Clone 100 −0.112 65 0.068

0.053 0.040

3.0

PM3 40 −0.178 32 14.5

0.067 0.056

Clone 100 −0.121 61 37.8 7.2

0.002 0.003

LSD 0.05 0.013 4.5 3.8 0.74

a_{Each value is the mean of 24 measurements.}

b_{Field capacity.} c_{Water potential.} d_{Fresh weight.} e_{Dry weight.}

medium without osmotic stress for 9 months lost some of their regeneration ability. Regeneration ability is commonly decreased by repeated subcul-ture of tissue in many plants. The reasons for loss of regenerative ability are not clearly understood and may be due to somaclonal variations, loss of a substance promoting the regeneration of fresh tis-sue, or epigenetic changes [21].

After growing clones without stress for 3 months, a significant change in the effect of mannitol on plant biomass was recorded when they were recul-tured in vitro. Some clones continued to grow in the presence of mannitol but others lost their potential to grow. This suggests that some clones had only a physiological adaptation to osmotic stress and that the plants lost their tolerance when the stress factor was removed from the medium. Similar physiological adaptations to stress conditions had been reported by Nabors [6].

Clones developed from non-stressed callus showed significant variations in growth when ex-posed to osmotic stress. Somaclonal variation among in vitro regenerated plants has been reported by Larkin and Scowcroft [22]. From fourT.minuta clones regenerated under water stress, one clone exhibited greater shoot and root fresh and dry weights on medium containing mannitol than on mannitol-free medium.

Both the drought-tolerant-selected clone PM3 and the non-stress-selected clone P4 grown on medium with 30 mM mannitol had higher proline content and soluble sugars, but not significantly so, than the same plants grown on mannitol-free medium. The accumulation of proline and soluble sugars in osmotically stressed plants of neither clone appears insufficient to relieve water stress. Hanson and Hitz [23] reported that, in many plants during moderate water stress, there was a 10- to 100-fold increase of free proline in leaf tissue. Alian et al. [24] showed that there was no correlation between the accumulation of proline and drought stress toler-ance in tomato. Therefore, the drought-tolerant-se-lected clone of T. minuta in this study may have another strategy to tolerate water stress.

Results suggested that the drought-tolerant-se-lected clone had a higher capacity to maintain membrane stability and/or low water potential and greater growth when grown on medium with low water potential. When this clone was tested in the greenhouse for drought, it yielded a higher biomass than other clones. This indicates that applying

water stress during callus growth was efficient in selecting a drought tolerant plant. Hasissou and Bouharmont [25] reported similar findings for the culture of Triticum durum where, out of 30 plants surviving water stress, 13 exhibited tolerance to drought compared with unselected controls.

In conclusion, one clone of T. minuta selected under drought appeared to show drought tolerance and had higher proline content and soluble sugars than the non-stress-selected clone even after growth in non-stress conditions for 6 months. Also, in vivo results showed that it had a higher yield and RGR under water stress than other regenerated clones or plants grown from seeds. Under non-stress condi-tions, the stress-selected clone exhibited similar growth characteristics to other tested plants. Fur-ther studies are required to investigate the mecha-nism in drought tolerance of the regenerated clones and the inheritance of this trait through sexual reproduction.

Acknowledgements

This investigation was financially supported by The Ministry of Higher Education, Egypt.

References

[1] R.T. Neher, The ethnobotany ofTagetes, Econ. Bot. 22 (1968) 317 – 325.

[2] J.A. Soule,Tagetes minuta: A potential new herb from South America, in: New Crops Proceeding of the Second National Symposium New Crop Exploration. Research & Commercial Utilisation, Indiana 1991, Wiley, New York, 1993, pp. 649 – 654.

[3] P. Vasudevan, K. Suman, S. Satyawati, Tagetes: A

multipurpose plant, Bioresour. Technol. 7 (1997) 29 – 35. [4] J. Kimpinski, W.J. Arsenault, Nematodes in annual ryegrass, marigold, mustard, red clover and soybean, Forage Notes 37 (1994) 52 – 53.

[5] E.A. Bray, Plant response to water deficit, Trends Plant Sci. 2 (1997) 48 – 54.

[6] M.W. Nabors, Environmental stress resistance proce-dure and applications, in: J.D. Philip (Ed.), Plant Cell Line Selection, VCH, Weinheim, 1990, pp. 167 – 185. [7] K. Dorffling, H. Dorffling, G. Lesselich, In6itro

selec-tion and regeneraselec-tion of hydroxyproline-resistant lines of winter wheat with increased proline content and increased frost tolerance, Plant Physiol. 142 (1993) 222 – 225.

[8] R.H. Smith, S. Bhaskaran, F.R. Miller, Screening for drought tolerance in sorghum using cell culture, In Vitro Cell Dev. Biol. 21 (1985) 541 – 545.

[9] H. Thomas, Drought resistance in plants, In: S.D. Amarjit, K.B. Ranjit (Eds.), Mechanisms of Environ-mental Stress Resistance in Plants, Harwood, 1997, pp. 1 – 42.

[10] R.A. Shibli, M.A.L. Smith, L.A. Spomer, Osmotic

ad-justment and growth response of three Chrysanthemum

morifoliumRamat. cultivars to osmotic stress inducedin 6itro, Plant Nutr. 15 (1992) 1373 – 1381.

[11] R.R. Duncan, R.M. Waskom, M.W. Nabors, In 6itro

screening and field evaluation of

tissue-culture-regener-ated sorghum (Sorghum bicolor L. Moench) for soil

stress tolerance, Euphytica 85 (1995) 373 – 380.

[12] M.A.-H. Mohamed, P.J.C. Harris, J. Henderson, An

efficient in 6itro regeneration protocol for Tagetes

minuta, Plant Cell Tissue Organ Cult. 55 (1999) 211 – 215.

[13] T. Murashige, F. Skoog, A revised medium for rapid growth and bioassays with tobacco tissue cultures, Phys-iol. Plant. 15 (1962) 473 – 497.

[14] L.S. Bates, R.P. Waldern, I.D. Teare, Rapid determina-tion of free proline from water stress studies, Plant Soil 39 (1973) 205 – 207.

[15] M.A. Jermyn, Increasing the sensitivity of the anthron method for carbohydrate, Anal. Biochem. 68 (1975) 332 – 335.

[16] P.F. Scholander, H.T. Hammel, E.D. Bradstreet, E.A. Hemmingsen, Sap pressure in vascular plants: Negative hydrostatic pressure can be measured in plants, Science 148 (1965) 339 – 3416.

[17] K. Tuomela, Leaf water relations in six provenances of

Eucalyptus microtheca: a greenhouse experiment, Forest Ecol. Manage. 92 (1997) 1 – 10.

[18] R. Hunt, Basic Growth Analysis: Plant Growth Analysis for the Beginners, Unwin Hyman, London, 1990. [19] H. Lipavska, D. Vreugdenhil, Uptake of mannitol from

the medium by in 6itrogrown plants, Plant Cell Tissue

Organ Cult. 45 (1996) 103 – 107.

[20] G. Gangopadhyay, S. Basu, S. Gupta, In6itroselection

and physiological characterization of NaCl- and manni-tol-adapted callus lines in Brassica juncea, Plant Cell Tissue Organ Cult. 50 (1997) 161 – 169.

[21] E.F. George, Plant Propagation by Tissue Culture. Part 1: The Technology, 2nd edn, Exegetic, Somerset, UK, 1993.

[22] P.J. Larkin, W.R. Scowcroft, Somaclonal variation — a novel source of variability from cell cultures for plant improvement, Theor. Appl. Genet. 60 (1981) 197 – 214.

[23] D. Hasissou, J. Bouharmont, In 6itro selection and

characterization of drought-tolerant plants of

durum-wheat (Triticum durum desf), Agronomy 14 (1994) 65 –

70.

[24] A. Alian, A. Altman, B. Heuer, Genotypic difference in salinity and water stress tolerance of fresh market tomato cultivars, Plant Sci. 152 (2000) 59 – 65.

[25] A.D. Hanson, W.D. Hitz, Metabolic responses of meso-phytes to plant water deficits, Annu. Rev. Plant Physiol. 3 (1982) 163 – 203.