Relative tolerances of wild and cultivated barley to infection by Blumeria graminis f.sp. hordei (Syn. Erysiphe graminis f.sp. hordei). II—the effects of infection on photosynthesis

Relative tolerances of wild and cultivated barleys to infection

by Blumeria graminis f.sp. hordei (Syn. Erysiphe graminis f.sp. hordei).

I. The effects of infection on growth and development

A. Akhkhaa,*, D.D. Clarkeb

aBotany Research Laborarory, Division of Biochemistry and Molecular Biology, Institute of Biomedical and Life Sciences,

University of Glasgow, Glasgow G12 8QQ, Scotland, UKbDivision of Environmental and Evolutionary Biology, Institute of Biomedical and Life Sciences, The Graham Kerr Building,

University of Glasgow, Glasgow G12 8QQ, Scotland, UK

Accepted 20 May 2003

Abstract

The two lines of wild barley, B19909 and I-17-40 and the cultivated barley, cv. Prisma used in this investigation were found to be the most

susceptible to infection of 25 wild and four cultivated barley lines when exposed to the local population of Blumeria graminis f.sp. hordei. All

three lines were susceptible during the early stages of growth but expressed some level of adult plant resistance although this level of

resistance was significantly higher in line I-17-40 than in either of the other two.

The relative tolerances of the lines to the mildew were determined by comparing the effects of infection on their growth and development

in growth cabinet experiments. Mildew developed more slowly on line I-17-40 than on the other two lines and by the third week after

inoculation, when mildew cover on B19909 and cv. Prisma had reached about 27%, only about 15% of the green leaf area of line I-17-40 was

covered. Mildew continued to increase on line B19909 and cv. Prisma so that 6 weeks after inoculation it covered 40% of their leaf-blades.

On line I-17-40 30% of the green leaf area was colonised by 4 weeks after inoculation but because of adult plant resistance coupled with the

loss of the earlier infected leaves through senescence mildew cover then reduced falling to 15% by 6 weeks. Although total mildew biomass,

measured as conidial production was higher on line B19909 than on cv. Prisma all its growth parameters were reduced less indicating that it

was the more tolerant line. Conidial production on the lower susceptible leaves of line I-17-40 was slightly lower than on cv. Prisma yet the

reaction of these leaves to infection was the same on both lines indicating that tissues of I-17-40 were slightly less tolerant than those of the

cultivated barley. However, during the later stages of growth when its upper leaves expressed high levels of ‘adult plant resistance’ dry matter

production in this line increased to levels higher even than in the controls. This capacity for compensatory photosynthesis ensured that by the

end of growth few differences in any of the measured growth parameters between infected and uninfected plants of line I-17-40 were

significant.

The greater tolerance of line B19909 over the other two lines and of cv. Prisma over line I-17-40 during the early stages of growth appears

to be due to a lower sensitivity to infection of those processes which regulate dry matter accumulation and its distribution around the plant.

q 2003 Elsevier Ltd. All rights reserved.

Keywords: Powdery mildew; Erysiphe; Blumeria; Wild barley; Cultivated barley; Hordeum vulgare; Hordeum spontaneum; Tolerance

1. Introduction

Even low levels of microbial infection can cause

significant reductions in the growth and yield or reproduc-

tive output of crop plants. In contrast, wild plants often

support relatively high levels of infection without their

growth and reproductive output appearing to be affected to

an equivalent extent [1,2,9,13]. Thus wild plants appear to

possess a greater ability to endure or tolerate microbial

attack than crop plants. Tolerance is here defined as the

ability of a plant to endure levels of infection that cause

greater impairment of growth and yield or reproductive

output in other plants of the same or similar species [2].

That tolerance could be an important component of the

survival strategy of wild plants to microbial attack and even

0885-5765/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.

doi:10.1016/S0885-5765(03)00072-9

Physiological and Molecular Plant Pathology 62 (2003) 237–250

www.elsevier.com/locate/pmpp

* Corresponding author. Tel.: þ44-141-330-6171; fax: þ44-141-330-

4620.

E-mail address: [email protected] (A. Akhkha).

to consumer attack in general is borne out by both empirical

[1,5,9] and theoretical studies [8,12]. One clear advantage of

tolerance over resistance in a host’s survival strategy is that,

unlike resistance, it does not impose a selection pressure on

the parasite for increased virulence or aggressiveness [8].

Certainly the boom and bust cycles so evident in some

major agricultural crops resulting from the deployment of

major resistance genes does not appear to be a feature of

wild plant populations.

Despite the recent interest in tolerance little is known in

detail of its physiological basis. However, the fact that

cultivars of crops often suffer more severely from consumer

attack than native plants indicates that some understanding

may be obtained by comparing the reactions of crop

cultivars with lines of their wild relatives to infection by a

common parasite. To this end Sabri and Clarke [9]

compared the reactions of a line of wild oat (Avena fatua)

with those of two cultivars of the cultivated oat (A. sativa) to

infection by the powdery mildew fungus Blumeria graminis

f.sp. avenae. This study showed that although the wild oat

supported the development of a higher level of mildew

biomass, measured as conidial production, than either

cultivated oat, its growth and development was reduced

the least. In fact, one of the cultivars supported little more

than half the amount of mildew growth as the wild oat yet its

growth and reproductive output were reduced to a

significantly greater extent. The reductions in growth were

attributable to both reductions in photosynthesis and to

changes in the pattern of distribution of photosynthates to

different parts of the plant but the least changes in both

occurred in the more tolerant wild oat. Clearly, the two

physiologic systems of the wild oat were less sensitive to the

activities of the mildew than those of the two cultivated oats.

However, although A. fatua is closely related to A. sativa

it is not its direct ancestor and so lines of the two species are

likely to differ genetically from each other in a range of

systems other than those specifically determining tolerance.

These differences could also affect host reactions to the

mildew fungus and thus make the interpretation of studies of

the effects of the fungus on host physiological systems

somewhat problematic. Thus, a comparison of the reactions

of cultivars of a crop with those of a direct ancestor should

be easier to interpret because of their greater genetic

similarity. For this reason wild (H. spontaneum) and

cultivated barley (H. vulgare) could provide suitable

material. Although wild and cultivated barley have been

given different specific epithets they are both self-pollinat-

ing diploids ð2n ¼ 14Þ and where they occur naturally

together they freely hybridize to produce hybrids which are

fully fertile [16], their chromosomes pairing well with

normal segregation [4]. Thus they clearly belong to the same

biological species with H. spontaneum being the direct

ancestor of the cultivated forms [4]. In this paper we report a

comparative study of the reactions of susceptible lines of

both species to infection by B. graminis f. sp. hordei (syn.

Erysiphe graminis f. sp. hordei).

2. Materials and methods

2.1. Plant material

The cultivated barley (H. vulgare), cv. Prisma, was

obtained from Twyfords Seeds, Banbury, Oxfordshire, UK

while the wild barley (H. spontaneum), lines, B19909 and

I-17-40 were kindly provided by the John Innes Centre,

Norwich Research Park and the Scottish Crop Research

Institute, Invergowrie, Dundee respectively.

These three lines were used in the experiments because

they proved to be among the most susceptible of 25 wild

barley lines and 4 cultivated barleys to the local population

of powdery mildew in greenhouse tests. In addition, none of

the lines expressed any necrotic reaction to infection that

could be indicative of major gene resistance to one or more

of the local mildew races. However, all lines, particularly

line I-17-40 expressed some reduction in the level of

infection during the latter stages of growth which is

indicative of adult plant resistance.

2.2. Conditions for growth

Although H. spontaneum is a naturally inbreeding

species, both wild lines were inbred for two generations,

starting from single grains, to ensure that the populations

used in experiments were uniform for morphological

characteristics.

Grains of the cultivated barley (cv. Prisma) were

germinated on moist filter paper in trays. Because grains

of wild barley possess a high level of dormancy their grains

were first chilled in a refrigerator at 4 8C for 7 days to break

this dormancy before placing on moist filter paper in trays to

germinate in the same way as the cultivated barley. About 1

week after germination, seedlings of a uniform size were

transplanted into 12.7 cm plastic pots containing a mixture

of equal parts of horticultural grade sand (Silvaperl sand)

and perlite (Celite), both from William Sinclair Horticul-

tural Ltd, Firth Rd, Lincoln, LN6 7AH UK). This medium

has good water and nutrient retention and supports good

growth while allowing the root systems to be cleanly

harvested with little loss. The growth medium in each pot

was covered with a circular sheet of black plastic to limit

evaporation and to cut down algal growth.

After potting, the seedlings were placed in a growth

cabinet which was illuminated by Kolorarc high-pressure

mercury vapour lamps giving 130 mmol quanta m22 s21 at

plant level during a 16 h photoperiod. The temperature

within the cabinet was maintained at 20 ^ 2 8C with a

relative humidity ranging between 65 and 80%. A nutrient

solution [15] was applied to each pot 3 times a week for the

first 8 weeks of growth and then once a week until the end of

the experiment. In between applications water was provided

as required.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250238

2.3. Inoculation

The plants remained free from infection in the growth

cabinets until mildew inoculum was introduced. When the

third leaves of the seedlings were fully expanded, about 2

weeks after transplanting, they were inoculated by shaking

heavily infected barley plants, over them at daily intervals

over a period of 7 days. The inoculum plants were produced

and maintained in the greenhouse. The control plants were

maintained free from mildew in the growth cabinets by

applying a 0·05% aqueous solution of Benomyl as a soil

drench weekly. Both the inoculated and the control plants

were placed in a random arrangement in the growth cabinet

and their positions changed weekly to ensure even growth.

2.4. Mildew assessment

Two methods were used to estimate the amount of

mildew growth on infected leaves. Firstly, the percentage of

the leaf blade colonised by the mildew was estimated using

a set of standard diagrams [6] Secondly, the number of

conidia present on each leaf blade was counted using the

procedure described by Sabri and Clarke [9].

2.5. Growth analysis

Plants were harvested at weekly intervals up until the

time the flag leaves had fully expanded. At this stage the

remaining plants were left for the grain to ripen in order to

obtain yield data. At each harvest, after measuring primary

tiller height, all tillers were excised from the plant at ground

level. The leaf blades were excised from their leaf sheath

and the area (green leaf blade area; GLA) of each leaf blade

on the main tiller and of all leaf blades together on each

subsidiary tiller was measured using a Delta-T flat bed

scanner (Delta-T Devices Ltd. 128 Low Road, Burwell,

Cambridge CB5 0EJ, England). The root systems were

recovered from the sand and perlite mixture using a Delta-T

root washer. The root systems were separated by hand into

the seminal and nodal root system. The total lengths of each

root system, the number of laterals on the roots, and the

mean diameters of the roots were determined using the

software provided with the Delta-T scanner.

3. Results

3.1. Mildew development and its effect on plant growth

3.1.1. Mildew development

Small isolated mildew colonies were first visible on the

inoculated plants from about 5 days after inoculation. From

then on the colonies progressed steadily on cv. Prisma and

line B19909 so that by 6 weeks after inoculation, when the

experiment was terminated, about 45% of their GLA

had become colonised (Fig. 1A). In contrast, mildew

development on line, I-17-40, progressed more slowly so

that only about 30% of its GLA were colonised by the

fourth week after inoculation. From then on the colonised

GLA declined as the mildew failed to develop to any

significant extent on the upper leaves and the lower leaves

were lost through senescence. By 6 weeks after inoculation

only about 15% of the upper leaves of line I-17-40 were

colonised. Higher numbers of conidia were produced on the

lower leaves than on the upper leaves of all lines with the

highest production on each leaf occurring on line B19909

(results not shown). The lowest levels of production overall

both on the lower leaves but particularly on the upper

leaves, occurred on line I-17-40.

The total number of conidia which had been produced on

all tillers of each line by each harvest are given in Fig. 1B.

The total number produced on line B19909 was significantly

higher than that on cv. Prisma. Almost all conidial

production on line I-17-40 occurred within 4 weeks of

inoculation. After this stage, because of the loss of the lower

leaves through senescence, mildew growth was restricted to

the upper resistant leaves where very little additional

production occurred. A higher level of conidial production

occurred during the experimental period on line B19909

(5·5 £ 108) than on cv. Prisma (4 £ 108) and the least

production occurred on line I-17-40 (1·5 £ 108).

3.2. Effects of infection on the growth and morphology

of the plant

3.2.1. Tillers

Mildew infection significantly reduced primary tiller

extension on but not on line, I-17-40 (Fig. 2A). The number

of tillers produced by cv. Prisma and line B19909 was also

significantly reduced (Fig. 2B).

Infection did not reduce the number of leaves produced

on the primary tillers of any line (11-12 on all lines) but it

did reduce the total GLA present at each stage of growth

(Fig. 2C). The reductions in GLA were the result of a

combination of processes. Firstly, in cv. Prisma and line

B19909, although not in line I-17-40, infection delayed the

emergence of each leaf. Secondly, and in all three lines it

reduced the expansion of the blades of those leaves that

were not fully developed at the time the infection had

become established. Finally, it increased the rate of

senescence of the leaves, particularly of the lower more

heavily infected leaves. However, infected leaves of line

B19909 senesced more slowly than those of the other two

lines did. Uninfected plants of all three lines also lost their

lower leaves through senescence but each leaf was lost

about 2 weeks later than the equivalent leaf on the infected

plants. The reduction in the GLA of the subsidiary tillers on

each line followed the same pattern as on the primary tiller

except that, because most tillers developed after inoculation,

the reductions were greater (Fig. 2D).

The reduced development of the tiller systems on the

infected plants of cv. Prisma and line B19909 was reflected

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250 239

in reduced dry weights (Fig. 3A). Dry matter accumulation

in the tillers of both lines, relative to that in the uninfected

plants, began to decline from soon after inoculation with the

greatest reduction in cv. Prisma. Dry matter accumulation

also declined in line I-17-40, beginning soon after

inoculation. However, with the expansion of the upper,

resistant, leaves the rate of accumulation began to increase

again to reach higher levels even than in the uninfected

controls so that by 6 weeks after inoculation the difference

in dry matter content between infected and control plants

was no longer significant.

Infection had marked effects on unit leaf rate (ULR),

which is a measure of the photosynthetic efficiency of the

GLA (Fig. 3B). The ULR followed similar ontogenetic

progressions in the uninfected plants of all three lines

reducing from high levels in the seedling stage to a

minimum when the plants were between 7 and 8 weeks of

age. The ULR also declined in the infected plants of all lines

but over a different time scale reaching a minimum about

a week earlier than in the uninfected plants. This minimum

level was then followed, in all three lines, by a dramatic rise

to a maximum well above that in the uninfected plants. The

maximum level reached in the B19909 and cv. Prisma was

then followed by a further fall and by the last harvest their

ULR were well below the levels in the uninfected plants. In

contrast, the maximum level attained in line I-17-40 was

followed by a slow decline and by the last harvest its ULR

was still much higher than in the uninfected plants.

Infection induced changes in the ontogenetic pro-

gressions of leaf-blade weight ratio (LWR), leaf-blade

area ratio (LAR) and specific leaf-blade area (SLA) in some

of the lines (Fig. 4A–C). Although infection had little effect

on the LWR of line I-17-40 it increased it slightly in cv.

Prisma and significantly increased it in line B19909. The

increased LWR in the infected plants indicates an increase

in the dry matter content per unit leaf area, probably due to

the mildew biomass developing over the leaf blade,

particularly in line B19909, as well as the retention of

Fig. 1. Development of B. graminis f.sp. hordei on wild line B19909 (W), cv. Prisma (V) and wild line I-17-40 (S). (A), percentage leaf area colonised; (B),

cumulative number of conidia produced per plant by each harvest. Each value is the mean ^ SE of determinations on three plants.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250240

more photosynthates in the leaves. The reductions in LAR in

cv. Prisma and line I-17-40 in response to infection indicate

that their GLA was smaller relative to the weight of the

whole plant than that of the uninfected controls, i.e.

the leafiness of the plants was reduced. On the other hand

the LAR of the most tolerant line B19909 was not affected

by infection. Reductions in SLA were mainly due to the

more rapid loss of the lower leaves by senescence although

the accumulation of more dry matter in the relatively

smaller upper leaves could also contribute to this reduction.

Fig. 2. Effects of infection by B. graminis f.sp. hordei on the growth of wild line B19909, cv. Prisma and wild line I-17-40. (A), primary tiller height; (B)

number of tillers per plant; (C), total green leaf blade area on the primary tiller; (D), total green leaf blade area on all tillers. (V), infected plants; (S), uninfected

plants. Each value is the mean ^ SE of determinations on three plants.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250 241

3.2.2. Root systems

Infection had no effect on the total number of seminal

roots produced per plant because this character was already

determined in the embryo, but it did reduce the number of

lateral roots produced by the seminal roots and the total

length of the whole system (Fig. 5A and B). Although

the mean diameters of the seminal root systems were not

reduced, because of the reductions in length the total surface

areas were reduced with the greatest reduction in cv. Prisma

(Fig. 5C).

Infection significantly affected the development of the

nodal rood system in all lines but again to different extents

Fig. 2 (continued )

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250242

in each. In B19909 the number of nodal roots 6 weeks

after inoculation was reduced by about 50% (from 40 to 27)

while in cv. Prisma it was reduced by about 60% (from 101

to 26). The number was also reduced by nearly 50% in line

I-17-40 (from 68 to 36) 6 weeks after inoculation. However,

the reduction in the number of nodal roots per plant was

clearly related to the reduction in tillering since the mean

number of nodal roots per tiller was not changed by

infection in either of the wild lines and was only slightly

decreased in cv. Prisma. Infection caused a large reduction

Fig. 3. Effects of infection by B. graminis f.sp. hordei on dry matter production by wild line B19909, cv. Prisma and wild line I-17-40. (A), shoot dry weight;

(B), unit leaf rate. (V), infected plants; (S), uninfected plants. Each value is the mean ^ SE of determinations on three plants.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250 243

Fig. 4. Effects of infection by B. graminis f.sp. hordei on the development of leaf tissue by wild line B19909, cv. Prisma and wild line I-17-40. (A), leaf weight ratio; (B), leaf area ratio; (C) specific leaf area. (V),

infected plants; (S), uninfected plants. Each value is the mean ^ SE of determinations on three plants.

A.

Akh

kha

,D

.D.

Cla

rke/

Ph

ysiolo

gica

la

nd

Mo

lecula

rP

lan

tP

ath

olo

gy

62

(20

03

)2

37

–2

50

24

4

in the number of laterals produced on the nodal roots in cv.

Prisma (Fig. 6A) but, apart from slightly delaying develop-

ment in the early stages of growth on line I-17-40, little

effect was evident on either of the two wild lines. The total

length of the nodal root system (Fig. 6B) was reduced in all,

with the greatest reduction in cv. Prisma and the least in line

B19909. The mean diameter of the nodal roots was also

reduced (results not shown) and because of this and the

reductions in total root length, the total surface area of

the nodal root systems of all lines was reduced (Fig. 6C) but

the least reduction occurred in line B19909.

The reduced development of the root systems of the

infected plants was reflected in their reduced dry matter

content (Fig. 7A). Dry matter accumulation in the root

systems of infected plants of cv. Prisma had more or less

ceased by 2 weeks after inoculation but it was still

continuing, up to the last harvest in the two wild lines,

although at a reducing rate particularly in line B19909.

The changes in the root:shoot ratios (Fig. 7B) also show

that infection reduced dry matter accumulation in the

roots more than in the shoots of both cv. Prisma and line

I-17-40 although these reductions became significant

much later in line I-17-40 than in cv. Prisma. The effect

of infection on the root:shoot ratio of line B19909 was

rather erratic between harvests and the differences were

generally not significant.

Fig. 5. Effects of infection by B. graminis f.sp. hordei on the development of the seminal root system by wild line B19909, cv. Prisma and wild line I-17-40.

(A), Total number of lateral roots; (B), Total root length; (C), Total surface area. (V), infected plants; (S), uninfected plants. Each value is the mean ^ SE of

determinations on three plants.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250 245

3.3. The effects of infection on the development

of yield structures

There was so much variation in the number of fertile

tillers produced per plant in each line that none of the

differences between infected and uninfected plants were

significant. Despite the significant reductions in total plant

biomass in all lines the proportion of that biomass converted

to grain, i.e. the harvest index, was significantly reduced

only in cv. Prisma and in this line it was reduced by 43%.

Infection had no significant effect on the total number of

grain produced on each tiller in either of the two wild lines

although it reduced it on all tillers, even the main tiller, of

cv. Prisma. However, it had no effect on the size of

the individual grains since thousand-grain weight was not

affected by infection in any line.

4. Discussion

In this study we have compared the reactions of three

susceptible lines of barley, two wild lines and one cultivated

line to infection by B. graminis in order to determine the

relative tolerances of the three lines to infection and to

determine something of the underlying basis of tolerance.

Comparing wild and cultivated lines of a crop can be

Fig. 6. Effects of infection by B. graminis f.sp. hordei on the development of the nodal root system by wild line B19909, cv. Prisma and wild line I-17-40. (A),

total number of lateral roots; (B), total root length; (C), total surface area. (V), infected plants; (S), uninfected plants. Each value is the mean ^ SE of

determinations on three plants.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250246

difficult because although they may belong to the same

species through selection for agronomic performance during

domestication, the cultivated forms will have accumulated a

number of characteristics which are not features of the wild

forms. This is a common feature of all cereals including

wild and cultivated barley [3]. Thus cv. Prisma, differed

morphologically from the two wild barley lines in several

respects. Firstly it produced fewer tillers with leaves that

Fig. 7. Effects of infection by B. graminis f.sp. hordei on (A), root dry weight and (B), root: shoot ratio of wild line B19909, cv. Prisma and wild line I-17-40.

(V), infected plants; (S), uninfected plants. Each value is the mean ^ SE of determinations on three plants.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250 247

were much broader with a larger total GLA than either of the

two wild lines. Secondly it developed a higher number of

nodal roots per tiller than either of the wild lines. Further

more, in contrast to the wild forms it possessed a non-shatter

rachis. However, despite these differences, the ontogenetic

progressions of the rates of dry matter accumulation,

evidenced in the ULR of lines, B19909 and cv. Prisma,

were almost identical to each other indicating a remarkable

similarity in their general developmental and physiological

systems. Surprisingly, the ontogenetic progression of dry

matter accumulation in the second wild line, I-17-40, was

rather different from that of the other two lines, a difference

which may be related to its relatively high level of adult

plant resistance.

Mildew development progressed in a similar manner on

each tiller of B19909 and cv. Prisma although line B19909

supported the higher level of development both in terms of

leaf area colonised and in the number of conidia produced.

However, despite the greater susceptibility of line B19909

all its measured growth parameters were reduced less than

in cv. Prisma and so clearly it possesses the greater level of

tolerance of the infection. This greater tolerance appears to

be due to the fact that its leaf tissue remained green and

turgid much longer after infection than that of cv. Prisma

thus allowing photosynthesis and in consequence mildew

growth and conidial production to continue over a longer

period of time. The previous studies of the relative

tolerances of wild and cultivated oats to mildew infection

also showed that infected leaf tissues of the wild oat staying

green and turgid longer than those of the cultivated oats

despite supporting the higher level of mildew development

[10]. Clearly, the leaf tissues of the wild oat were also less

sensitive than that of the cultivated oats to the mildew

infection.

The early stages of growth of the adult-plant-resistant

line I-17-40 were also reduced by infection. In fact,

although its lower leaves supported little more than half

the level of mildew development as leaves in similar

positions on the tillers of cv. Prisma, they senesced just as

rapidly. Thus the tissues of these leaves, appear to possess a

lower level of tolerance of the mildew than those of cv.

Prisma. However, mildew development on the upper,

highly-resistant, leaves was severely restricted and during

this latter stage of infection growth increased again to such

an extent that by the end of the experiment any differences

in most growth parameters between infected and uninfected

plants were no longer significant.

In contrast to its effects on the vegetative growth of the

B19909 and cv. Prisma infection had little or no effect on

most of the parameters of reproductive growth, particularly

the size and weight of individual grains. This was also found

in the earlier study of the effects of mildew infection on wild

and cultivated oats [10]. However, many workers have

shown that the reproductive output of plants, particularly the

weight or size of grains or seed are very much less sensitive

to disruption by stress factors than are vegetative growth

parameters [11].

The reduced growth, particularly vegetative growth, of

infected plants is clearly largely due to reductions in the

amounts of photosynthates produced in their tissues. Net

photosynthesis, of which ULR is a measure, followed

similar ontogenetic progressions in the uninfected plants of

each line but these progressions were drastically affected by

infection. The initial decline was probably due to the

decreasing proportion of photosynthetic to non-photosyn-

thetic tissue as more structural tissue was formed in the

developing plant. However, after reaching a minimum level

the ULR began to increase again, probably due to the

establishment of new sinks for photosynthate in the

developing reproductive structures. In the uninfected plants

the ULR of all three lines declined from high levels in the

seedling stage as the ratio of non-photosynthetic to

photosynthetic tissue increased but this decline occurred

more rapidly in the infected plants. This more rapid decline

in the infected plants must have been due to the disruptive

effects of the developing mildew fungus although the rate of

decline was not related to the level of mildew development.

Thus, the rate of decline was similar in both line B19909

and cv. Prisma even though B19909 supported the higher

level of mildew development. Surprisingly, the most rapid

decline occurred in line I-17-40 even though its leaves were

supporting the lowest level of mildew development. This

again indicates that the photosynthetic apparatus in the

lower leaves of this line is more sensitive to mildew

development even than it is in the lower leaves of cv.

Prisma.

After falling to a minimum level in both infected and

uninfected plants the ULR increased again but surprisingly

it increased to higher levels in the infected than in the

uninfected plants particularly in line I-17-40. This increase

in each of the three lines probably reflects the development

of compensatory photosynthesis in the newly emergent

upper leaves to supply the photosynthate sinks in the

developing yield structures. In line B19909 and cv. Prisma

these newly emergent leaves were initially lightly infected

but as the mildew developed on them the ULR began to fall

again to levels well below those of the earlier minimum. The

sharpest and greatest fall occurred in cv. Prisma despite the

fact that its upper leaves supported lower levels of infection

than those of line B19909. In contrast to the other two lines

little mildew developed on the upper leaves of tillers of line

I-17-40 and its ULR continued at a higher level even than in

the uninfected plants of this line for the remainder of the

experiment. The continued high ULR in the upper leaves of

this line clearly explains how infected plants were able to

generate sufficient photosynthate during the later stages of

growth to ensure that overall, their final growth and yield

was not significantly less than that of the uninfected

controls.

Similar changes in ULR or net photoynthesis in response

to powdery mildew infections have been observed in

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250248

groundsel [1] and in wild and cultivated oats [10]. In these

species the ULR of infected plants fell to relatively low

levels beginning soon after inoculation but increased again

later, presumably to meet the requirements of the develop-

ing yield structures. In the wild and cultivated oats, as in the

barleys in this study, the increase was lower in the least

tolerant cultivated oat, cv. Peniarth, than in the more

tolerant cultivated oat, cv. Lustre, or in the wild oat. These

increases in ULR in infected plants, particularly of the wild

plants, to higher levels than in the uninfected plants, indicate

a capacity for compensatory photosynthesis in the upper

leaves. This capacity may be a significant factor in tolerance

since it would enable the infected plant to make up for

reduced production occurring as the result of infections

during earlier stages of growth.

In addition to reducing dry matter production, infection

also changed the proportions distributed to the different

parts of the plant. This is evident in the changes in the

various growth ratios found, particularly in the root:shoot

ratio but also in the various leaf ratios, LWR, LAR and

SLA. However, just as with dry matter production the

magnitude of the changes in these ratios was not related to

the level of mildew infection. In general, line B19909

showed smaller changes in the proportions of dry matter

allocated to its different organs than cv. Prisma and in

some cases smaller changes than in line I-17-40. Changes

have been noted in growth ratios in several other plant

species in response to infection. For example, Sabri and

Clarke [9] noted changes in various leaf ratios in wild and

cultivated oats in response to infection by B. graminis

f.sp. avenae but, as in this study, these changes were

much less marked in the wild line than in the cultivated

lines. Changes in several leaf ratios were also noted by

Ben-Kalio and Clarke, [1] in the native weed, groundsel

(Senecio vulgaris), in response to infection by Erysiphe

fischeri. However, these changes were also much smaller

than have been observed in this study to occur in the

cultivated barley or in the earlier study on cultivated oats

[9]. Changes in the root:shoot ratio have been reported to

be a common response of many plant species to infection

by a wide range of pathogens [see Ref. [14] for a review]

and it has been suggested that such changes are an

inevitable consequence of infection [7]. However, almost

all studies reporting changes to root:shoot ratios have

involved crop plants. Most studies which have examined

the responses of wild plants to infection [1,9] have shown

either little change or a much smaller change than in crop

plants. This indicates that the large changes often reported

for crop plants are not an inevitable consequence of

infection but probably reflect a level of intolerance of

infection. Clearly, an ability to maintain normal patterns

of translocation of photosynthates around the plant is

likely to be a significant factor in tolerance.

In conclusion, it is clear that wild line B19909 was

more tolerant of the mildew infection than cv. Prisma.

This greater tolerance of the former line appears to result

from the relatively lower sensitivity of its metabolic

processes to disruption by infection since photosynthesis

and the patterns of translocation of photosynthates around

the plant were disrupted less than in the cultivated barley.

Additionally, both wild lines, but particularly line I-17-40,

appear to have a greater capacity for compensatory

photosynthesis than the cultivated barley. In line B19909

this could result from the general lower sensitivity of the

photosynthetic apparatus to infection since its leaves

remained green and turgid much longer after infection

than those of the cultivated barley. However, this is

unlikely to be the case for line I-17-40 since its lower

leaves appeared to be even more sensitive to the mildew

than those of the cultivated barley. In this line

compensatory photosynthesis appears to occur largely in

the upper highly resistant and lightly infected leaves. Line

I-17-40 is an interesting line since the ‘adult-plant

resistance’ in its upper leaves permitted high levels of

compensatory photosynthesis and this appeared more than

adequate to make up for the reduced production during

the earlier susceptible stages of growth. Tolerance of the

infection that occurs during the early vegetative phases of

growth is not a feature of this line. Thus tolerance of

infection perhaps does not provide a selective advantage

to the early stages of growth of lines if any reductions in

production caused by infection are made up during the

later adult-plant resistance stages. This would certainly be

the case if tolerance has a metabolic cost.

References

[1] Ben-Kalio VD, Clarke DD. Studies on tolerance in wild plants: effects

of Erysiphe fischeri on the growth and development of Senecio

vulgaris. Physiological Plant Pathology 1979;14:203–11.

[2] Clarke DD. Tolerance of parasites and disease in plants and its

significance in host–parasite interactions. In: Ingram DS, Williams

PH, editors. Advances in Plant Pathology, vol. 5. London: Academic

Press; 1986. p. 161–98.

[3] Evans LT. Crop Evolution Adaptation and Yield. Cambridge:

Cambridge University Press; 1993.

[4] Harlan JR. Barley. In: Simmonds NW, editor. Evolution of Crop

Plants, Cambridge University Press; 1976. p. 93–8.

[5] Harry IB, Clarke DD. The effects of powdery mildew (Erysiphe

fischeri) infection on the development and function of leaf tissue

by Senecio vulgaris. Physiological Plant Pathology 1992;40:

211–24.

[6] Large EC, Doling DA. The measurement of cereal mildew and its

effect on yield. Plant Pathology 1962;80:47–57.

[7] Parbery DG. A consideration of tolerance to parasites in plants.

Annals of Applied Biology 1978;89:373–7.

[8] Roy BA, Kirchner JW. Evolutionary dynamics of pathogen resistance

and tolerance. Evolution 2001;54:51–63.

[9] Sabri N, Clarke DD. The relative tolerances of wild and cultivated

oats to infection by Erysiphe graminis f.sp. avenae 1. The effects of

infection on vegetative growth and yield. Physiological and Molecular

Plant Pathology 1995;49:405–21.

[10] Sabri N, Dominy PJ, Clarke DD. The relative tolerances of wild and

cultivated oats to infection by Erysiphe graminis f.sp. avenae 11.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250 249

The effects of infection on photosynthesis and respiration. Physio-

logical and Molecular Plant Pathology 1996;50:321–35.

[11] Silverside JW. Introduction to Plant Population Ecology. London:

Longman Group Ltd; 1982.

[12] Stowe KA, Marquis RJ, Hochwender CG, Simms EL. The

evolutionary ecology of tolerance to consumer damage. Annual

Review of Ecology and Systematics 2000;31:565–95.

[13] Tarr SAJ. Principles of Plant Pathology. London: Macmillan Press;

1972.

[14] Walters DR. Shoot: root interrelationships: the effects of

obligately biotrophic fungal pathogens. Biological Reviews

1985;60:47–79.

[15] Walters DR, Ayres PG. Effects of powdery mildew disease on uptake

and metabolism of nitrogen by roots of infected barley. Physiological

Plant Pathology 1980;17:369–79.

[16] Zohary D, Hopf M. Domestication of Plants in the Old World, 2nd ed.

Oxford: Oxford University Press; 1994.

A. Akhkha, D.D. Clarke / Physiological and Molecular Plant Pathology 62 (2003) 237–250250