Effects of soil resistance to root penetration on leaf expansion ...

Journal of Experimental Botany, Vol. 47, No. 304, pp. 1651-1662, November 1996Journal ofExperimentalBotany

Effects of soil resistance to root penetration on leafexpansion in wheat (Triticum aestivum L.): composition,number and size of epidermal cells in mature blades

Gerrit T.S. Beemster1 and Josette Masle123

1 Environmental Biology Group, Research School of Biological Sciences, Institute of Advanced Studies,Australian National University, GPO Box 475, 2601 ACT, Canberra, Australia

^^Cooperative Research Centre for Plant Sciences, Australian National University, GPO Box 475, 2601 ACT,Canberra, Australia

Received 12 December 1995; Accepted 22 July 1996

Abstract

Wheat seedlings (Triticum aestivum L.) were grown onsoils with contrasted resistances to root penetration(measured as penetrometer resistance, flj. High fl.reduced the rates of leaf appearance and expansion.Although the duration of expansion was increased,mature leaves were smaller. Underlying changes in leafanatomy were investigated on cleared mature leaves,focusing on the epidermes. Three leaves were ana-lysed: leaves 1 and 3 which started their developmentin the embryo, and leaf 5 which was initiated on theseedling, after imposition of contrasted soil conditions.In all leaves, high R, caused a reduction in mature cellsizes, lengths and widths, and a shift in the relativeproportions of functionally different cell types, with adecrease in the relative proportions of stomata andassociated cell types (interstomatal and sister cells)and an increase in the proportions of unspecializedelongated epidermal cells and of trichomes. In leaves3 and 5 the number of cellular files across the bladewas also reduced, while in leaf 1 it was similar at thetwo /?.. These differences between leaves are attrib-uted to differences in their developmental stage whenroot stress was first perceived. Remarkably, Rm had noeffect (leaf 1) or relatively small effects (leaves 3 and5) on the total number of cells per file, suggesting thatthis parameter is either largely insensitive to variationin root environment, or is programmed at the outsetbefore stress was perceived at the apex.

Key words: Wheat, anatomy, mature epidermis, rootimpedance.

Introduction

Soils with high resistance to root penetration (R,) reduceplant growth. Many studies have concentrated on rootresponses (Barley, 1962; Taylor and Ratlin", 1969; Goss,1977). However, leaf growth is also severely affected(Masle and Passioura, 1987; Masle, 1990). In fact, in thelong term (scale of a few weeks) carbon accumulation inwheat seedlings with impeded roots may be more reducedin leaves than in roots (Masle et al., 1990; Masle, 1992).Leaf emergence is delayed; blades expand more slowlyover a longer period of time and mature blades are bothnarrower and shorter (Masle and Passioura, 1987; Masle,1990, 1992).

The aim was to examine the cellular bases of thesewhole leaf growth responses. The present paper reportsobservations of the anatomy of mature blades from wheatplants {Triticum aestivum L.) grown on soils with con-trasted R,. Effects of R, on epidermal cell dimensions andnumbers, and on the organization of these cells into fileswere investigated. Companion papers address effects onthe kinetics of cell partitioning and cell expansion in theepidermis of elongating blades and on apical developmentaround the time of leaf initiation. Reasons for focusingon epidermal cells were two-fold: firstly leaf expansion isconstrained by the extensibility of the epidermis(Kutschera et al., 1987; Kutschera, 1989). Secondly, inwheat, mesophyll cells are lobed in an irregular fashionso that tracing the contour of neighbouring cells is difficultand not always reliable.

Grass leaves are convenient experimental systems foranatomical studies. Most of the leaf length is generated

'To wtiom correspondence should be addressed. Fax: +61 6 249 4919. E-mail: masle©rsbs-central.ami.edu.au

O Oxford University Press 1996

Dow

nloaded from https://academ

ic.oup.com/jxb/article/47/11/1651/600336 by guest on 20 Septem

ber 2022

1652 Beemster and Masle

in a basal growth zone, physically well-defined and organ-ized into cellular files parallel to the leaf axis. New cellsare formed by transverse divisions at the base of the leafand displaced to more distal positions by the formationand expansion of cells at more basal locations. A fewmillimetres from the base of the leaf, cells lose their abilityto divide and only expand, reaching their mature sizewhile still enclosed in the sheath of the preceding leaf(Boffey et ai, 1980; Volenec and Nelson, 1981). Celldifferentiation follows a basipetal gradient and continueswell after the cell has reached its final dimensions (Boffeyet ai, 1980).

Materials and methods

Growth conditions

Wheat plants (Triticum aestivum L. cv. Egret) were grown incylindrical PVC pots (200 mm high; 87 mm in diameter). Fourseeds (24-28 mg) were sown in each pot at constant depth(15 mm). The soil used for this experiment was a silty loanmade of 5% coarse sand, 40% fine sand, 31% silt, 19% clay,and 4.5% organic matter and was part of the batch of soil usedby Masle and Passioura (1987) in earlier experiments on growthresponses to R,. This soil was fertilized before use by additionof 2.2 g finely ground super phosphate kg"1 dry soil and 123 mgnitrogen provided as a mixture of KNO3 and NH4NO3, or ofCa(NO3)2 and NH4NO3 or as NH4NO3 alone. After thoroughmixing, soil water content was brought to 0.24 or 0.22 g H3Og"1 dry soil and homogeneously packed to a bulk density of1.12 and 1.42 gem"3, respectively, resulting in 0.9 and 6.6 MPasoil penetrometer resistances, respectively, in a first run(experiment A, see below), and 0.9 and 7.7 MPa, in a secondrun (experiment B) and soil matric potentials ranging from 0.01to 0.1 MPa.

The pots were placed in a growth chamber providing aphoto- and thermoperiod of 11 h with day/night temperaturesof 18/15°C, 600 fimol quanta m~2 s"1 irradiance, and a relativehumidity of 80%. After germination the soil was covered witha layer of white plastic beads in order to limit soil waterevaporation and thereby variations in R,. Each pot was weigheddaily, and lost water was added one to three times a day inorder to maintain soil water content within 1% of its originalvalue. Ten days after sowing, seedlings were thinned to twoper pot.

All three nutrient regimes ensured plentiful nutrient supply(data not shown) and gave similar plant growth and morpho-logy. Results below are, therefore, from bulked data. Incontrast, increasing R, had dramatic effects on plant size anddevelopmental stage, throughout the experiment with, forexample, a 35% increase in the phyllochron (interval betweenthe emergence of succesive leaves) from 4.1 to 5.6 d, a 50%drop in leaf elongation rate, and a 20-30% decrease in finalleaf length. These differences were similar to those reported byMasle and Passioura (1987, Fig. 4; Table 3) for similar combina-tions of soil densities and water content. These authorsconcluded that, in their experiment, the effects of soil resistancecould not be attributed to a restricted oxygen or water supplyto the roots. Water deficit is unlikely to have been a confoundingfactor in the present experiment either, despite the slightlyhigher air temperature and daily irradiance applied. The lowestleaf water potentials measured at the end of the experimentmid-afternoon before watering, were indeed similar at the two

R, (-0.83±0.71 and -0.83±0.67 MPa at low and high R,,respectively). Furthermore, in concurrent experiments run underthe same environmental conditions as here, growth responsesto R, were insensitive to 0.4 MPa variation in leaf waterpotential (from 0.6 to 1 MPa) whether induced by decreasingair humidity or by pressurizing the roots (Masle, unpublishedresults).

Leaf preparation

In a first run, fully expanded blades of leaves 1 and 3 wereharvested on five plants at each R,. As their analysis revealedunexpected leaf position effects, a second batch of plants wasgrown under the same conditions (run 2) from which a thirdleaf, leaf 5, was harvested. Leaf 1 and leaf 3 start theirdevelopment in the wheat embryo (Hayward, 1938) while leaf5 is initiated after germination, i.e. in this experiment, severaldays after the roots had experienced low or high soil resistance.Upon harvest, blade length (Z,) and area (a) were measured.The whole blade was then cleared for light microscopy, using amodification of the technique described by Clarke (1959).Blades were immersed in boiling methanol for about 30 minuntil all chlorophyll was removed, and then in lactic acid untilanalysis. This gentle clearing procedure partly destroys thecytoplasm and removes chlorophyll without damaging cellwalls. The cleared leaves were mounted on a light microscope(Zeiss axioscope) fitted with a Panasonic video camera (modelWV-CL 702E). Maximum leaf width (W) was measured mid-way along the cleared blade and was similar to that measuredbefore clearing. Morphometric analysis of the epidermis wasperformed on video images (total magnification 190x) usingthe morphometric program MTV (Garr UpdegraftTDatacrunch,1991) on a PC equipped with a video-card.

Morphometric analysis of blade epidermis

The epidermis of a wheat leaf contains a variety of cells. Filesof different cell types are arranged in a regular pattern (Percival,1921; Stebbins and Shah, 1960; Esau, 1977; Silvy, 1982) whichis closely associated with the location of veins in the mesophyll(Fig. 1). Between two adjacent veins, there were usually twostomatal rows, comprising stomatal complexes (guard cells andassociated subsidiary cells, labelled st in Fig. 1) separated byinter-stomatal cells (i). Adjacent to these files were files ofwider 'sister cells' (s) so-called because they derive from thesame 'mother cells' as the subsidiary cells (Tomlinson, 1974,1994). The two inner files of sister cells, located between twoadjacent stomatal rows were separated by several files ofunspecialized, long and narrow 'elongated' cells (e) on the bladeabaxial side, enlarged and shorter 'bulliform' cells (b) on theadaxial side. One or two outer files above larger veins weremade of sclerenchymatous cells (scl). A number of shorterelongated cells and also of sister cells were associated withtrichomes, especially at high R, (cells labelled e, and s,,respectively, in Fig. 1).

The effects of R, or leaf position on blade anatomy were,therefore, analysed in terms of total number of files across theblade, number of files of the various types described above, celldensity (total number of cells per unit area, and number ofcells of a given type), and cell dimensions. To allow for possibletemporal or ontogenetic variations in the effects of R, on leafhistogenesis the above parameters were determined at severallocations along and across the blade on three replicated leaves.For these leaves, blade width, number of veins, number andwidth of cellular files were determined at 10 locations evenlydistributed along the blade. At one of these locations (distance0.4Z. from the ligule) cell dimensions were measured in three

Dow



ber 2022

Fig. 1. Typical anatomy of the abaxial (a) and adaxial (b) epidemicsof a wheat leaf. Symbols denote the various epidermal cell types whichwere distinguished in the present study: b, Bulliform cell; e, elongated,non-specialized cell; st, stomata; i, interstomatal cell; s, sister cells; t,trichomes; scl, schJerenchyma cell. Subscript t for elongated and sistercells denotes association with a trichome.

different interveinal areas selected on one side of the mid-vein(second area from the blade mid-rib and margin, and areabetween the central veins). At each of three other locationsalong the blade (0.2, 0.6 and 0.8L) measurements were restrictedto the second interveinal area from the mid-rib. The length andgreatest and smallest widths of 10 cells of each type weremeasured. For the two other replicates, cell dimensions weremeasured at one location only (second interveinal area from

Leaf anatomy and root stress 1653

mid-rib and distance 0.4L from ligule). Cell densities were alsodetermined at that same location for all five replicates. A seriesof overlapping images of a 3-mm-long segment were printedwith a video-graphic printer (Sony, model UP-811). Individualcells were included in the counts when their basal transversewall was within the segment. Sclerenchyma cells located abovethe veins were omitted because their contours were often hardto follow. Cells of this type contributed little to the totalepidermal area and their proportions relative to files of othertypes were little affected by R,. The width of the observedsegment was estimated by averaging the widths measured atfive equidistant points along its length, and used to calculatethe total area to which cell counts had to be referred. The totalnumber of cells per file was also counted along one file ofbulliform cells, sister cells and stomatal/interstomatal cells infive replicated leaves per treatment. The three files were selectedclose to the mid-rib, among files which could be unambiguouslytraced from the ligule to the tip of the blade. AH parameterswere determined on the two epidermes. When the effects of soilconditions or leaf position were similar for each epidermis, dataare presented for the abaxial epidermis only.

Statistical ana/vs/s

The effect of R, on the relative proportions of files of varioustypes and on cell densities was analysed by log-linear modellingof contingency tables, using the statistical package Genstat 5(version 2.1; Lawes Agricultural Trust (RothamstedExperimental Station)). All other data were analysed withANOVA analysis of variance using the same statistical software.

Results

Whole leaf dimensions

High Rt significantly reduced the area of all mature blades(P< 0.001; Fig. 2a) by c. 35% in leaves 1 and 3, and 47%in leaf 5. This reduction in area reflected a decrease inboth leaf length (21, 16 and 27% for leaves 1, 3, and 5,respectively (Fig. 2b)) and leaf width (Fig. 3). Maximumblade width measured at a distance of 0.4L from theligule was decreased by 30, 20 and 38% in leaves 1, 3,and 5, respectively (Fig. 3). This reduction was represent-ative of that measured at other locations within the basaltwo-thirds of the blade, but was greater than reductionat more distal locations (Fig. 3). Soil resistance modifiedleaf shape with blades being almost rectangular at highR, (e.g. leaf 1, Fig. 3), while having a marked triangularshape at low R,. The greater response of leaf 5 comparedto the two other leaves may only be apparent as the highR, for leaf 5 (experiment B) was 7.7 MPa penetrometerresistance compared with 6.6 for leaves 1 and 3 (experi-ment A, see Materials and methods). The elongation ratesof leaves 1 and 3 on the plants from which leaf 5 wassampled also responded more to R, than those of leavessampled in the previous run and shown in Figs 2 and 3(Table 1).

Number of cell files: cell width

However, there were striking true leaf position effects onthe cellular bases of the overall changes in blade dimen-

Dow



ber 2022


a.

15-

10-

25-

2 0 -

cCD

• 1 5 -

10-

b.

Leaf position

Fig. 2. Variations in the size of mature blades with soil resistance andleaf position: (a) blade area; (b) blade length. Leaves grown at low andhigh R, are described by grey and black bars, respectively. Error barsdenote the standard error of the mean (/i = 5).

sions with R, described above. In leaf 1, blades fromplants grown at high Rt were narrower only because cellsof all types were much narrower (Fig. 4). The number ofcellular files constituting the blade was indeed similar atthe two R, (Fig. 5). In leaves 3 and 5, in contrast, bladesgrown at high R, were made of fewer cell files (Fig. 5).Cells within these files were also narrower, but somewhatless than in leaf 1; the reduction in cell width was spatiallyuneven across and along the blades and not alwayssignificant (Fig. 4).

1.0-1

0.8-

S 0.6 H

* 0.4-cd0

0.2-

0.0-0.0 0.2 0.4 0.6

Distance from ligule

0.8 1.0

Fig. 3. Variations in the width of mature blades with soil resistance(open and solid symbols for low and high R,, respectively), leaf position(circles, squares and triangles for leaves 1, 3 and 5, respectively), anddistance from the ligule expressed as a proportion of total blade length.

Table 1. Comparison of leaf elongation rates (mm d'1) betweenplants grown in two successive runs for analysis of leaves 1 and3 (run 1) and leaf 5 (run 2)

Numbers are thetime data, during

R, Leaf 1

Low R,

Run 1 24.1Run 2 24.2

slope of thethe linear t

High* .

14.212.9

: regression line fitted to leaf length versus)hase of leaf elongation.

Leaf 3

Low R,

31.531.9

High/?,

23.721.7

Leaf 5

Low R, High R,

25.8 15 4

Figure 4 shows that there was significant spatial vari-ation in cell width within a blade, especially at low R,,with cells becoming narrower towards the tip of the bladeor its margins (circles in Fig. 4) compared to more centrallocations. There were also variations associated with celltype and formation of trichomes. For example, elongatedcells tended to be wider than sister cells. More noticeably,elongated cells which, within a file, were preceded andfollowed by a trichome were significantly narrower thanthose preceded and followed by another elongated cell(comparison of diamonds and squares in Fig. 4).However, when expressed in relative terms, the effects ofRt on cell width were of similar magnitude at differentlocations on the blade and for different cell types.

There was a high correlation between number of cellfiles and number of veins across the blade, i.e. the number

Fig. 4. Effects of soil resistance (open and closed symbols for low and high R,, respectively) on epidermal cell widths in leaves 1, 3 and 5. Data areshown for various cell types for the abaxial epidermis, except for bulliform cells (present on the adaxial epidermis only). Widths of elongated cells,sister cells and interstomatal cells on the adaxial side of the blade showed a similar pattern of variation. Data points (D , • ) correspond to meanvalues (±s.e., n = 3 or 5 leaves) at four equidistant locations along the blade for leaves 1 and 3, one location only for leaf 5, all within the secondinterveinal area from the mid-rib. At position 0.4/. widths are also shown for cells located in the second interveinal area from the blade margins( O , • ) . For elongated cells, data refer to cells which were not associated with trichomes except at one location where cells associated withtrichomes are also shown ((O, • ) to be compared with ( • , • ) ) .

Dow



ber 2022

Leaf 1Elongated cells

35 A

30

25 A

20 A

15-1

35 n Sister cells

25 A

15H

10 J

5I 30

O25-I

20 H

15H

10J

40

35

X A

25

20

Interstomatal cells

15-L-r

Bulliform cells

0.4 0£

35 A

30

25 A

2OA

15

25 H

15H

10-1

35

20 H

10-1

35 A

30 H

25 H

20-^

15-OS


Leaf 3 Leaf 5

35

204

15-1

30H

10-1

30

20

35

02 tt4 06 OB


150.4

Dow



ber 2022


<D 3 0 0T3O5

200-

8 ioo-

J3

Z 0

I

0.0 0.2 0.4 0.6

Distance from the llgule

0.8 1.0

Fig. 5. Variations in the number of cellular files constituting the abaxialepidermis of mature blades as a function of soil resistance, leaf positionand location along the blade. Symbols are as in Fig. 3.

of cell files between two veins varied little with either R,or leaf position (overall mean number = 11.7). However,variation in R, caused qualitative differences in the relativeproportions of different types of files: the number of'ground tissue' files (Esau, 1977), made of elongated,bulliform and sclerophyllous cells, was increased by highR, while the number of stomatal rows and associated filesof sister cells remained unaffected. This shift was especi-ally marked for the adaxial epidermis (Fig. 6) where filesof stomata and sister cells represented 46% of all cell filescompared with 50% at low R, (P <0.001). Furthermore,high /?, caused a significant increase in the number offiles with trichomes which represented 27% of all fileswith elongated cells and 56% of all files with sister cells,as opposed to 12% and 37%, respectively, at low Rt.

The number of cell files and veins across the bladeincreased from leaves 1 to 3 to 5 (Fig. 5). These increasesbore no relation to variations in cell width, for whichthere was no consistent ranking between leaves (Fig. 4).They were large enough to cause a significant increase inblade width of successive leaves (Fig. 3), even when cellwidths were decreased (as from leaf 3 to 5, at low R,(Fig. 4)).

Cell length: number of cells along a file

Mature cell lengths are described in Fig. 7. They werespatially even more variable than cell widths, in anapparently more or less random fashion rather thanaccording to a positional gradient along or across theblade. Elongated cells were 2 to 4 times longer than othercells except when associated with a trichome (Fig. 7) andcells of all types were longer on the abaxial than adaxialepidermis (data not shown). In leaf 1 high R, caused amarked reduction in cell length which was of similarmagnitude for all cell types (20-25%) apart from bulli-form cells (0-10% depending on location along the blade).

scIDH

10-1

8-

6-

4-

1 1

fc

•S 10-,

Ez

8-

6-

4 -

4

Abaxial

10

4 6 8

Number of files between veins at low R

10

Fig. 6. Effects of soil resistance on the average number of cellular filesbetween adjacent veins in the adaxial and abaxial epidermes. Datapoints correspond to individual blades of leaf I (D, • ) , leaf 3 (O, • )and leaf 5 (A, A). Open symbols represent the number of stomatalfiles and related files of sister cells, solid symbols represent the numberof all other files.

At the one location of leaf 5 where cell dimensions weremeasured, cell lengths were also reduced (Fig. 7). Incontrast, in leaf 3, only elongated cells were on averageshorter at high than low Rt.

The total cell numbers per file in the mature blade areshown in Table 2. Consistent with the differences in lengthbetween cell types shown in Fig. 7, the number of cellscounted along adjacent files of different type varied.Remarkably, however, in all three types of files shown inTable 2, cell number was unaffected (in leaves 1 and 5),or only decreased by about 10% (7 to 17% in leaf 3) byhightf,.

Cell densities: cell indices

The overall effect of variations in cell length and widthand in the relative proportions of files of various typesdescribed above was reflected in variations in cell densities(number of cells per unit area). Soil resistance was themain source of variation of cell densities. Figure 8 shows

Dow



ber 2022

c

75O

Leaf 11000

800-

600-

400-

200-

Bonoated cells

150-

100-

50J

400-,

350

300-

250-

200-

150-

100-1—r

BullHorm cells

1000-

800-

600

400-

200-

350-

300-

200-

150-

Leaf 3


Leaf 51000

803-

600

400

200

CM

30CH

250-

200

150-

100-1

250 -,

200-

150-

100-

50J

350-

300-

200-

150-

—,—10a08

5

0402 0.4 OB OiJ 02 O4 06


Fig. 7. Effect of soil resistance on epidermal cell lengths in leaves 1, 3 and 5 for the same cell types and same leaves as in Fig. 4. Symbols are as in Fig. 4.

Dow



ber 2022


Table 2. Total number of cells counted along single files from ligule to blade tip on the abaxial and adaxial epidermes

Numbers in brackets denote standard errors of the means (n = 5 leaves) Asterixes denote statistically significant differences between R, (P<\%(***); 1%<P<5%(*»); 5%</><10%(*)).

Leaf

1

3

5

Epidermis

AbaxialAdaxialAbaxialAdaxialAbaxialAdaxial

Interstomatal

Low R,

370(13)528 (19)705 (6)898 (20)

cells

Highfl,

353(14)519 (14)583(34)784 (42)

P

• • *

• »*

Sister cells

Low R,

338 (13)475 (15)557 (10)760 (11)968(13)

High*.

386(12)456(10)520(35)667 (25)982 (13)

P

*

**•

Bulliform cells

Low R,

354(9)

559(16)

974(13)

HighK,

314(7)

503 (36)

826(16)

P

• *

• • •

* • •

that total densities were significantly greater at high thanlow /?,. This effect was especially marked in leaf 1 (68%increase on average as opposed to 54% and 38% in leaves3 and 5, respectively) and, in that leaf, was greatertowards the base of the blade and in central files comparedto more distal or marginal locations. In leaf 3, celldensities were spatially more uniform. Densities of trich-omes and of elongated cells were the most increased (100to 200%, Fig. 9) while the densities of stomata andinterstomatal cells were less affected (25% at the most).The increase in the density of sister cells was intermediate(Fig. 9).

Changes in the relative proportions of various cell typeswith R, can be quantified by comparing cell indices inwhich the number of cells of a given type is weighed bythe abundance of these cells relative to the entire popula-tion (Fig. 10). By definition, these indices take valuesranging between 0 and 1 and the total of the indicescalculated for a given sample equals 1. Figure 10a showsthat indices for trichomes and elongated cells were signi-ficantly increased at high R, while indices for stomataand derived cells were reduced. These shifts were notsolely caused by the large increase in trichome number.When trichomes were excluded from the total number ofcells for the calculation of cell indices (Fig. 10b), therewas still a clear trend to an increase in elongated cellindices (10 to 100%) and a decrease in stomatal indices(10 to 50%). Within a leaf, sister cells were the mostabundant (c. 30% of all cells). As expected from thegenesis of the stomatal complex (Tomlinson, 1974), thedensities of sister cells, stomata, and interstomatal cellswere in a stable ratio close to a 2:1:1 ratio. Elongatedcells were in similar or slightly lower number than stomataand interstomatal cells while, on the adaxial side of theblade, bulliform cells were significantly fewer.

Discussion

Whole leaf growth responses to variations in soil mechan-ical resistance have been well documented (Masle andPassioura, 1987; Masle, 1992). The present study confirmsthat high R, restricts leaf expansion and mature leaf sizein wheat seedlings and provides new information on someof the anatomical modifications underlying theseresponses or associated with them. Changes in the size,number, type and arrangement of epidermal cells wereanalysed in mature blades of the first, third and fifthleaves of the main stem. This showed that high R, has anumber of effects on leaf histogenesis.

High R, caused a reduction in the sizes of mature cells,both in length and width, a decrease in the number ofcellular files constituting the blade, and a shift in therelative proportions of functionally different cell types.Surprisingly, the total number of cells in a file was notor comparatively little affected.

The occurrence and magnitude of the above effectsdepended on leaf position and also showed significantvariation with location on the blade and cell type. Forleaf 1, reduction in blade dimensions at high R, wasmostly due to mature cells being smaller (Figs 4, 7), whileboth the number of cell files and number of cells along afile were similar at the two R, (Fig. 6; Table 2). In leaf 3,however, the reduction in cell dimensions was less markedor even often not significant, especially lengthwise, whilethe number of cell files was significantly reduced. In leaf5 both cell sizes and number of cell files were decreased.As mentioned earlier, the data for that leaf are not ascomplete as for leaves 1 and 3 and may, in part, beconfounded by a run effect (see Materials and methods).However, later experiments in which all three leaves weresampled from the same plants confirmed the above results,

Fig. 8. Total epidermal cell densities for leaves 1, 3 and 5 at low and high R, (grey and black bars, respectively); (a) longitudinal variation withinthe second interveinal area from the mid-rib; (b) lateral variation at a distance 0.4i. from the ligule. Labels on the x-axis refer to the location ofcounts, described by the distance from the ligule in panel (a) (0 = ligule; 1 = tip of the leaf) and by the distance from the mid-rib in panel (b) (0 =mid-rib; 1=0.5 WO.

Dow



ber 2022

a. Longitudinal variation


b. Lateral variation

350

350-I

300-

250-

200-

150-

100-

50-

0--

350 -i

0.2 0.4 0.6 0.80.5 0.8

Distance from midrib

Leaf 5

0 2 0.4 0.6 0 8


Dow



ber 2022


100n 1-4/ 1:2.

EEtn

Itc"CO

.C

"ffl

slty

cCD•a

ell

O

80-

60-

40 -

20 -

1

/

/

' D •

= - ! a

•' » . /J y7 v

• AO V O/

20 40 60 80 100

Cell density at low Rt (colls mm'2)

Fig. 9. Effects of R, on the densities of various epidermal cell types inthe adaxial and abaxial epidemics (open and solid symbols, respectively);( l l ) trichomes ; ( • ) elongated cells; (O) interstomatal cells, (V)sister cells; (A) stomata; (O) bulliform cells. Data points representaverage densities (n = 5) for each leaf position (I, 3 and 5) at location0 4L along the blade and in the second interveinal area from the mid-rib.

0.4-

0.3-

0.2-

0 .1-

0.0-

• /

/

/D4O

:

11 /

K 0.0 0.1 0.2 0.3 0.4

0.4- k

0.3-

0.2-

0 1 -

0.00.0 0.1 0.2 0.3

Cell index at low Ft.

0.4

Fig. 10. Cell indices at low and high R, calculated from the cell densitiesshown in Fig. 9, i.e. number of cells of a given type as a proportion ofthe total number of cells (a) including trichomes (b) excludingtrichomes. Symbols are the same as in Fig. 9.

including the smaller sensitivity of cell sizes to R, in leaf3 than in leaves 1 and 5. In wheat, the three first leavesare initiated in the embryo on the mother plant (Hayward,1938). In the present experiment, therefore, leaf 1 andleaf 3 initially developed in similar conditions for allplants, from initiation to the time of radicle emergencewhen differences in R, were first perceived, i.e. for at theleast 2.5 and 0.5 plastochrons, respectively. In contrast,leaf 5 was initiated on the very young seedlings afterroots had already experienced contrasted soil conditionsfor some time. Three to four days after germination, i.e.about 1 d after seedling emergence, the primordium ofleaf 5 was indeed only 25-50 ^m long. It is thereforeconcluded that the above differences in the response ofsuccessive leaves to Rs partly reflect differences in theirstage of development when variation in R^ was perceived.Consequently, the data suggest that at that stage the finalnumber of formative divisions for the epidermis wasalready determined in leaf 1, but not in leaves 3 and 5.In addition to these interactions with leaf ontogeny,variation in the response of successive leaves may also becaused by time-dependent changes in the metabolic effectsof high R^. Earlier studies with the same experimentalsystem have shown that in an initial period, reduced leafexpansion at high R^ could not be explained by reducedphoto-assimilates, nutrients or water supply, and thatsignal (s) elicited in the roots were acting as the primarytrigger (Masle and Passioura, 1987; Masle, 1990; Masleet al., 1990). The nature of these signals and the cascadeof metabolic events that they generate in leaf meristemsare still uncertain. Abscisic acid and related compoundsor ethylene appear as likely candidates (Apelbaum andYang, 1981; Parker and Ford, 1982; Zhang and Davies,1990; Tardieu et al., 1992; Spollen et al., 1993). However,Masle et al. (1990) also provided evidence that in asecond period, leaf growth becomes carbon limited, butwith a lower responsiveness to sugar concentrations thanat low /?,. From their study, it can be inferred that, inthe present experiment, such a limitation would havedeveloped only after leaf 1 was fully expanded, but whileleaves 3 and 5 were expanding.

The size of bulliform cells was the least reduced byhigh R, while that of elongated cells was the most (Figs3, 7). For this latter type this size reduction was to alarge extent related to the increased frequency of trichomeformation (Fig. 9). Trichomes are small, hair bearingcells produced by a final asymmetrical division of somemeristematic cells (Stebbins and Shah, 1960). Such asym-metrical divisions were seen at the distal end of the leafmeristem in files of both sister cells and elongated cells(Beemster et al., 1996). Trichomes were derived from thesmallest daughter cell while the biggest daughter celldifferentiated into a non-specialized elongated or sistercell which was much shorter and narrower than thoseproduced by symmetrical division (Figs 4, 7). Such pairs

Dow



ber 2022

of cells were scattered along whole files, with no apparentpattern, and no consistency between files. Adjacent cellsin neighbouring files were not necessarily smaller thancells preceding or following them in the file. Togetherthese observations indicate that asymmetrical partitioningwas associated with a shorter cell cycle than symmetricalpartitioning or with locally prolonged meristematic activ-ity. Both interpretations would lead to smaller cell sizesat maturity and it is known that division rates in leafmeristems vary both spatially and temporally (Poethigand Szymkowiak, 1995; Beemster et al., 1996).

As a result of reduced cell sizes, blades grown at highK, were characterized by increased cell densities.Following from the above, this was mostly due to anincreased density of trichomes and of elongated cells;however, the densities of sister cells and bulliform cellswere also increased. Stomatal densities were usually theleast increased (8-60%), or even not significantly differentfrom those at low R, (abaxial epidermis, leaves 1 and 3).Earlier studies with the same genotype had shown that,on a leaf area basis, stomatal conductance was decreasedby high R, (Masle and Passioura, 1987). The presentdata show that this effect was the direct result of reducedstomatal aperture. Significant decreases in stomatal indexindicate that this involved R, effects on stomatal develop-ment. Analysis of herbarium specimens showed thatstomatal densities declined by about 40% on average overthe last two hundred years concurrently with an increasein atmospheric CO2 concentration from about 280-340/xmol mol"1 (Woodward, 1987). Our data, together withother reports of variation in stomatal densities withdrought or growth irradiance shows that caution is neededbefore interpreting this observation in terms of directCO2 effects on stomata and cell expansion.

As a whole the effects of high i?, on the anatomy ofmature leaves are typical of the syndrome associated withwater shortage. Reduced cell sizes leading to increasedcell densities, increased frequency of trichome formation,reduced stomatal index have also been reported in anumber of experiments in response to drought (Zalenski,1904; Yapp, 1912; McCree and Davis, 1974; Quarrie andJones, 1977; Zagdanska and Kozdoj, 1994) or to treat-ments that influence the water status of developing leaves(Jones, 1985; Ristic and Cass, 1991). These similaritiesreinforce the proposition formulated by Masle andFarquhar (1988) that plants may perceive high R, as awarning of a shortage of water, and that this induces anumber of morphological and physiological adaptationsenabling conservation of water. Masle and Passioura(1987) suggested that plants growing at high .ft, arereminiscent of bonsai plants. The present study does notgive support to that analogy; in their comparative studyof a number of bonsai plants Korner et al. (1989) indeedfound that mature leaves were made of fewer and oftenlarger cells. In the wheat plants used here, high Rt had


totally (leaf 1) or partially (leaves 3 and 5) oppositeeffects.

What do these changes in the anatomy of mature leavestell us of the effects of R, on leaf development? Thenumber of cells in the blade and the distribution of theirsizes are the product of the relationships between therates of two elemental processes that took place in themeristem and growth zone of the expanding leaves,namely cell wall expansion, and cell partitioning (Green,1976). There is no doubt that high R, imposed limitationson cell surface expansion. This was clearly the case inleaf 1, where blade elongation was slower (Table 1) andproceeded for longer at high R, despite the total numberof cells which contributed to that elongation being similar(Table 2). It is also clear that the number of formativedivisions, which gives rise to the number of parallel cellfiles constituting the blade was very sensitive to rootimpedance. Interestingly, the number of proliferative divi-sions along these files, which are realized later in leafdevelopment, was comparatively insensitive to it. Thisresult suggests that this is a genetically programmedcharacter whose expression is largely independent ofenvironmental conditions. Gunning et al. (1978) showedin Azolla that the entire development of the root,expressed in terms of the number of divisions the apicalcell will undergo, and of spatial and temporal patterns ofdivision, is rigidly programmed from the outset.Alternatively, the number of proliferative divisions insuccessive leaves may be set very early in the life of theseedling in relation with some early events at the apicalmeristem which, in the present experiment, occurredbefore the root stress was perceived. In any case, thesedata indicate that number of formative divisions andnumber of proliferative divisions are not directly related,and suggest that these two kinds of divisions may corre-spond to two fundamentally different kinds of cell cycles.Later experiments (Beemster et al., 1996) showed that inleaves 3 and 5, high R, had significant effects on thenumber of cells per file in the leaf meristem and on thecell cycle duration. This makes the insensitivity to R, ofthe final number of cells per file in the mature bladeseven more remarkable. As a whole these results raise thefollowing questions: when and how are the number offormative and proliferative divisions determined? To whatextent may the constraint that they impose on final plantsize be amenable to manipulation?

Acknowledgements

We are grateful to Mr RC Cunningham for statistical adviceand for performing the log-linear modelling of contingencytables. We also thank Dr Sally Stowe and Mr David Vowlesfrom the EMU at RSBS for their help with the microscopy andmorphometric equipment.

Dow



ber 2022


References

Apelbaum A, Yang SF. 1981. Biosynthesis of stress ethyleneinduced by water deficit. Plant Physiology 68, 594-6.

Barley KP. 1962. The effects of mechanical stress on the growthof roots. Journal of Experimental Botany 13, 95-110.

Beemster GTS, Masle J, Williamson RE, Farquhar GD. 1996.Effects of soil resistance to root penetration on leaf expansionin wheat {Triticum aestivum L.): kinematic analysis of leafelongation. Journal of Experimental Botany 47, 1663-78.

Boffey SA, Sellden G, Leech RM. 1980. Influence of cell age onchlorophyll formation in light-grown and etiolated wheatseedlings. Plant Physiology 65, 680-4.

Clarke J. 1959. Preparation of leaf epidermis for topographicstudy. Stain Technology 35, 35-9.

Esau K (ed). 1977. Anatomy of seed plants, 2nd edn. New York:John Wiley & Sons.

Goss MJ. 1977. Effects of mechanical impedance on rootgrowth in barley (Hordeum vulgare L.). I. Effects on theelongation and branching of seminal root axes. Journal ofExperimental Botany 28, 9 6 - 111.

Green PB. 1976. Growth and cell pattern formation on an axis:critique of concepts, terminology and modes of study.Botanical Gazette 137, 187-202.

Gunning BES, Hughes JE, Hardham AR. 1978. Formative andproliferative cell divisions, cell differentiation and develop-mental changes in the meristem of Azolla roots. Planta143, 121-^4.

Hayward HE (ed). 1938. The structure of economic plants. NewYork: Macmillan Publishers.

Jones HG. 1985. Adaptive significance of leaf development andstructural responses to environment. In: Baker DA, DaviesWJ, Ong CJC, eds. Control of leaf growth. Society forExperimental Biology, Seminar series 27. CambridgeUniversity Press, 155-73.

K6rner C, Menendez-Riedl SP, John PCL. 1989. Why arebonsai plants small? A consideration of cell size. AustralianJournal of Plant Physiology 16, 443-8.

Kutschera U, Bergfeld R, Schopfer P. 1987. Co-operation ofepidermis and inner tissues in auxin-mediated growth ofmaize coleoptiles. Planta 170, 168-80.

Kutschera U. 1989. Tissue stresses in growing plant organs.Physiology Plantarum 77, 157-63.

Masle J. 1990. Growth and stomatal behaviour: response tosoil resistance to root penetration. In: Davies WJ, Jeffcoat B,eds. Importance of root to shoot communication in the responsesto environmental stress. Monograph 21, British Society forPlant Growth Regulation, 95-113.

Masle J. 1992. Genetic variation in the effects of root impedanceon growth and transpiration rates of wheat and barley.Australian Journal of Plant Physiology 19, 109-25.

Masle J, Farquhar GD. 1988. Effects of soil strength on therelation of water-use efficiency and growth to carbon isotopediscrimination in wheat seedlings. Plant Physiology 86, 32-8.

Masle J, Farquhar GD, Gilford RM. 1990. Growth and carboneconomy of wheat seedlings as affected by soil resistance topenetration and ambient partial pressure of CO2. AustralianJournal of Plant Physiology 17, 465-87.

Masle J, Passioura JB. 1987. The effect of soil strength on the

growth of young wheat plants. Australian Journal of PlantPhysiology 14, 643-56.

McCree KJ, Davis SD. 1974. Effect of water stress andtemperature on leaf size and on size and number of epidermalcells in grain sorghum. Crop Science 14, 751-5.

Parker ML, Ford MA. 1982. The structure of the mesophyll offlag leaves in three Triticum species. Annals of Botany49, 165-76.

Percival, J. 1921. The wheat plant. A monograph. London:Duckworth & Co Publishers.

Poethig RS, Szymkowiak EJ. 1995. Clonal analysis of leafdevelopment in maize. Maydica 40, 67-76.

Quarrie SA, Jones HG. 1977. Effects of abscisic acid and waterstress on development and morphology of wheat. Journal ofExperimental Botany 28, 192-203.

Ristic Z, Cass DD. 1991. Leaf anatomy of Zea mays L. inresponse to water shortage and high temperature: a compar-ison of drought-resistant and drought-sensitive lines. BotanicalGazette 152, 173-85.

Silvy A. 1982. La premiere feuille des plantules d'orge. I.Morphogenese les epidermes. Canadian Journal of Botany60, 2864-76.

Spol'en WG, Sharp RE, Saab IN, Wu Y. 1993. Regulation ofcell expansion in roots and shoots at low water potentials.In: Smith JAC, Griffiths H, eds. Water deficits. Plantresponses from cell to community, Oxford: BIOS ScientificPublishers Limited, 37-52.

Stebbins GL, Shah SS. 1960. Developmental studies of celldifferentiation in the epidermis of monocotyledons. II.Cytological features of stomatal development in theGramineae. Developmental Biology 2, 477-500.

Tardieu F, Zhang J, Davies WJ. 1992. What information isconveyed by an ABA signal from maize roots in drying fieldsoil? Plant, Cell and Environment 15, 185-91.

Taylor HM, Ratliff LF. 1969. Root elongation rates of cottonand peanuts as a function of soil strength and soil watercontent. Soil Science 108, 113-19.

Toralinson PB. 1974. Development of the stomatal complex asa taxonomic character in the monocotyledons. Taxonomy23, 109-28.

Tomlinson PB. 1994. Monocotyledons—towards an understand-ing of their morphology and anatomy. Advances in BotanicalResearch 3, 208-92.

Volenec JJ, Nelson CJ. 1981. Cell dynamics in leaf meristemsof contrasting tall fescue genotypes. Crop Science 21, 381-5.

Woodward FI. 1987. Stomatal numbers are sensitive to increasesin CO2 from pre-industrial levels. Nature 327, 617-18.

Yapp RH. 1912. Spiraea ulmaria L., and its bearing on theproblem of xeromorphy in marsh plants. Annals of Botanv26, 518-70.

Zagdanska B, Kozdoj J. 1994. Water stress-induced changes inmorphology and anatomy of flag leaf of spring wheat. AdaSocietal is Botanicorum Poloniae 63, 61—6.

Zalenski V. 1904. Materials for the study of the quantitativeanatomy of different leaves of the same plant. Mim PolytechInst Kiev 4, 1-203.

Zhang J, Davies WJ. 1990. Changes in the concentration ofABA in xylem sap as a function of changing soil water statuscan account for changes in leaf conductance and growth.Plant, Cell and Environment 13, 277-85.

Dow



ber 2022

Effects of soil resistance to root penetration on leaf expansion ...

Documents

Transcript of Effects of soil resistance to root penetration on leaf expansion ...